CoChat
Sign Up
Cited by 454Evolution and Genetic DynamicsBacterial Genetics and BiotechnologyMicrobial Community Ecology and Physiology

The DOI matches its Crossref record; 76 of the 78 references it cites were found in a registry or an index, and 2 are not checked yet.

Quantifying the local adaptive landscape of a nascent bacterial community

Open access
Ascensao, Joao A., Wetmore, Kelly M., Good, Benjamin H., Arkin, Adam P., Hallatschek, Oskar·2023·Nature Communications
DOI 10.1038/s41467-022-35677-5PMC PMC9842643

Published version of a preprint

Open this paper in CoChat — free accountOpen PDF

Abstract

The fitness effects of all possible mutations available to an organism largely shape the dynamics of evolutionary adaptation. Yet, whether and how this adaptive landscape changes over evolutionary times, especially upon ecological diversification and changes in community composition, remains poorly understood. We sought to fill this gap by analyzing a stable community of two closely related ecotypes (“L” and “S”) shortly after they emerged within the E. coli Long-Term Evolution Experiment (LTEE). We engineered genome-wide barcoded transposon libraries to measure the invasion fitness effects of all possible gene knockouts in the coexisting strains as well as their ancestor, for many different, ecologically relevant conditions. We find consistent statistical patterns of fitness effect variation across both genetic background and community composition, despite the idiosyncratic behavior of individual knockouts. Additionally, fitness effects are correlated with evolutionary outcomes for a number of conditions, possibly revealing shifting patterns of adaptation. Together, our results reveal how ecological and epistatic effects combine to shape the adaptive landscape in a nascent ecological community.

Overview

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1038/s41467-022-35677-5

Cites

76 of 78 references verified0 citation errors · 0 not checkable

Statistics

Figures

Fig. 1 Measuring mutational fitness effects. A Timeline of evolution in the ara-2 LTEE population, showing mutation accumulation and diversification into S and L around 6k generations, then clone sampling at 6.5k generations; data from47, jitter added to mutation fixation time for easier visualization. Hypermutator phenotype appeared around 2.5k generations59. B Schematic of transposon mutagenesis process to generate barcoded libraries of REL606, 6.5k S and L, as well as experimental procedure to observe barcode dynamics. The erlenmeyer flask clipart (flask-2 icon by DBCLS) shown here is used and modified under a CC-BY- 4.0 creative commons license. C Barcoded knockout mutant frequency trajectories in the evolutionary condition for each genetic background, colored by estimated fitness. All barcodes within a gene were summed together; shown are the trajectories from replicate 1 in the evolutionary condition for each genetic background (monoculture for REL606, together at ecological equilibrium frequency for S and L; representative of both replicates). D Overall distributions of fitness effects in the evolutionary condition for each genetic background. The majority of knockouts were neutral, so only genes that were called as significantly non-neutral were included (see methods section “Probabilistic model of read count trajectories and fitness inference”). E Replicate–replicate correlation of estimated fitness effects. F Comparison of knockout fitness effects across genetic backgrounds, which are generally uncorrelated. Points with a blue interior correspond to genes that were mutated (excluding S-SNPs) in 6.5k S/L relative to REL606 (sequencing data from ref. 46). Points with red outlines correspond to genes that were mutated in parallel in nonmutator LTEE populations (data from ref. 47). The correlation coefficients decrease slightly if we recompute them, excluding likely neutral genes (ρ = 0.14, 0.03, 0.03; top to bottom). In panels E, F, knockouts with high measurement noise (σs > 0.3%) were excluded (except for labeled genes), ρ is the weighted pearson correlation coefficient, and error bars represent standard errors, as calculated in methods section “Probabilistic model of read count trajectories and fitness inference”. Also in panels E, F, the “cloud” of points around 0 mostly represents likely neutral knockouts.
Fig. 1 Measuring mutational fitness effects. A Timeline of evolution in the ara-2 LTEE population, showing mutation accumulation and diversification into S and L around 6k generations, then clone sampling at 6.5k generations; data from47, jitter added to mutation fixation time for easier visualization. Hypermutator phenotype appeared around 2.5k generations59. B Schematic of transposon mutagenesis process to generate barcoded libraries of REL606, 6.5k S and L, as well as experimental procedure to observe barcode dynamics. The erlenmeyer flask clipart (flask-2 icon by DBCLS) shown here is used and modified under a CC-BY- 4.0 creative commons license. C Barcoded knockout mutant frequency trajectories in the evolutionary condition for each genetic background, colored by estimated fitness. All barcodes within a gene were summed together; shown are the trajectories from replicate 1 in the evolutionary condition for each genetic background (monoculture for REL606, together at ecological equilibrium frequency for S and L; representative of both replicates). D Overall distributions of fitness effects in the evolutionary condition for each genetic background. The majority of knockouts were neutral, so only genes that were called as significantly non-neutral were included (see methods section “Probabilistic model of read count trajectories and fitness inference”). E Replicate–replicate correlation of estimated fitness effects. F Comparison of knockout fitness effects across genetic backgrounds, which are generally uncorrelated. Points with a blue interior correspond to genes that were mutated (excluding S-SNPs) in 6.5k S/L relative to REL606 (sequencing data from ref. 46). Points with red outlines correspond to genes that were mutated in parallel in nonmutator LTEE populations (data from ref. 47). The correlation coefficients decrease slightly if we recompute them, excluding likely neutral genes (ρ = 0.14, 0.03, 0.03; top to bottom). In panels E, F, knockouts with high measurement noise (σs > 0.3%) were excluded (except for labeled genes), ρ is the weighted pearson correlation coefficient, and error bars represent standard errors, as calculated in methods section “Probabilistic model of read count trajectories and fitness inference”. Also in panels E, F, the “cloud” of points around 0 mostly represents likely neutral knockouts.CC BY · PubMed Central
Fig. 2 Statistical properties of DFEs as well as effects of individual mutations sensitively depend on environment. A Knockout fitness effects tend to have a larger magnitude when S and L are at ecological equilibrium versus when they are in monoculture. Line shown is a rolling average of fitness effects ± standard error. Error bars on points represent standard errors, as calculated in methods section “Probabilistic model of read count trajectories and fitness inference”. B Distribution of fitness effects across environments, where we only included knockouts that were called as significantly non-neutral. Please note that the DFEs of REL606 are on a different scale than S and L. C Illustration of how the sign of fitness effects changes across environments. Few mutations are unconditionally beneficial or deleterious, many are non-neutral only in one or few environments, and sign-flipping of fitness effects across environments is pervasive. REL606 only has 4 unique environments, compared to 6 for S and L. Size of circle is proportional to number of genes that fall into each class.
Fig. 2 Statistical properties of DFEs as well as effects of individual mutations sensitively depend on environment. A Knockout fitness effects tend to have a larger magnitude when S and L are at ecological equilibrium versus when they are in monoculture. Line shown is a rolling average of fitness effects ± standard error. Error bars on points represent standard errors, as calculated in methods section “Probabilistic model of read count trajectories and fitness inference”. B Distribution of fitness effects across environments, where we only included knockouts that were called as significantly non-neutral. Please note that the DFEs of REL606 are on a different scale than S and L. C Illustration of how the sign of fitness effects changes across environments. Few mutations are unconditionally beneficial or deleterious, many are non-neutral only in one or few environments, and sign-flipping of fitness effects across environments is pervasive. REL606 only has 4 unique environments, compared to 6 for S and L. Size of circle is proportional to number of genes that fall into each class.CC BY · PubMed Central
Fig. 3 Similarity of fitness effects between environments. A Clustering environments by using fitness effect correlation as a measure of similarity reveals which environments are the most functionally alike. For example, environments related to the putative ecotype niches--exponential acetate growth and glucose growth, for S and L, respectively—cluster with conditions where the ecotype is in the minority. The red and yellow dots indicate that the branch has ≥90% or ≥70% support respectively, computed via bootstrapping. B Principal components analysis (PCA) of our data, using (normalized) fitness effects as features (% variance). We see that L experiments cluster separately from the S and REL606 experiments, with the exception of the acetate exponential phase condition.
Fig. 3 Similarity of fitness effects between environments. A Clustering environments by using fitness effect correlation as a measure of similarity reveals which environments are the most functionally alike. For example, environments related to the putative ecotype niches--exponential acetate growth and glucose growth, for S and L, respectively—cluster with conditions where the ecotype is in the minority. The red and yellow dots indicate that the branch has ≥90% or ≥70% support respectively, computed via bootstrapping. B Principal components analysis (PCA) of our data, using (normalized) fitness effects as features (% variance). We see that L experiments cluster separately from the S and REL606 experiments, with the exception of the acetate exponential phase condition.CC BY · PubMed Central
Fig. 4 Correlations between genes across environments. We observed that many pairs of genes have correlated fitness effects across environments, for example, A most genes of the fecABCDE operon. However, fecA is correlated with the other genes to varying degrees, depending on the genetic background. Error bars on points represent standard errors, as calculated in methods section “Probabilistic model of read count trajectories and fitness inference”. B We computed the pairwise correlation of fitness effects (cofitness) for all pairs of genes, and then clustered genes with a community detection algorithm55. We then rearranged the cofitness matrices by reordering genes based on “optimal” clustering of other genetic backgrounds. For each column, we ordered the genes based on the clustering of a given genetic background. For each row, we used the cofitness matrix for a given background. It is apparent that replotting the cofitness matrix using another strain’s clustering does not produce noticeable structure. C, D Cluster reassortment is not entirely random—pairs of genes C in the same operon and D that strongly interact with each other (high EcoliNet score), tend to stay in the same clusters across genetic backgrounds. In contrast, the cofitness of pairs of genes that are not in those categories appear to change in a way that is indistinguishable from random reassortment. In panels C and D, the abbreviations R, S, and L refer to REL606, and 6.5k S/L, respectively. Error bars represent standard errors.
Fig. 4 Correlations between genes across environments. We observed that many pairs of genes have correlated fitness effects across environments, for example, A most genes of the fecABCDE operon. However, fecA is correlated with the other genes to varying degrees, depending on the genetic background. Error bars on points represent standard errors, as calculated in methods section “Probabilistic model of read count trajectories and fitness inference”. B We computed the pairwise correlation of fitness effects (cofitness) for all pairs of genes, and then clustered genes with a community detection algorithm55. We then rearranged the cofitness matrices by reordering genes based on “optimal” clustering of other genetic backgrounds. For each column, we ordered the genes based on the clustering of a given genetic background. For each row, we used the cofitness matrix for a given background. It is apparent that replotting the cofitness matrix using another strain’s clustering does not produce noticeable structure. C, D Cluster reassortment is not entirely random—pairs of genes C in the same operon and D that strongly interact with each other (high EcoliNet score), tend to stay in the same clusters across genetic backgrounds. In contrast, the cofitness of pairs of genes that are not in those categories appear to change in a way that is indistinguishable from random reassortment. In panels C and D, the abbreviations R, S, and L refer to REL606, and 6.5k S/L, respectively. Error bars represent standard errors.CC BY · PubMed Central
Fig. 5 Fitness effects of beneficial genes are correlated with evolutionary outcomes. We explored if genes with beneficial knockout fitness effects are correlated with A establishment of a mutation in a gene, and B changes in gene expression over evolutionary time, relative to neutral knockouts. A Slopes from logistic models, with presence of a mutation in a gene as the response variable. The fitness effects were normalized by the median beneficial fitness effect, so that coefficients can be interpreted as the average difference in log-odds establishment between neutral knockouts and the ‘typical’ beneficial knockout. REL606 beneficial knockout fitness is positively correlated with gene establishment probability for most environments, but in different time intervals, potentially pointing to shifting targets of selection. Error bars represent standard errors, which were calculated as detailed in methods section “Genome evolution”. FDR-corrected p-values were obtained from the logistic regression model. B We compared the distributions of log-fold change in expression between genes with neutral knockout fitness effects, less beneficial effects (lower 50%), and more beneficial effects (upper 50%). We used the change in expression from 0k gens (REL606) to 40k gens (L), from 6.5k gens (S) to 40k gens (S), and from 6.5k gens (L) to 40k gens (L) for the REL606, 6.5k S, and 6.5k L panels, respectively. The expression change between ancestor and 40k L (left) is nearly identical to the expression change between ancestor and 40k S as well as other timepoints (Fig. S25). Beneficial knockout fitness in REL606 is generally positively correlated with increasing gene expression over time. In S and L, fitness in several environments--including the ecological equilibrium and acetate and glucose growth--is correlated with decreasing gene expression. The center line of the boxplot is the median; the limits of the box represents the first and third quartiles; the whiskers extend to one times the inter-quartile range; outliers are not shown. FDR-corrected Mann–Whitney U tests were used to compare conditions. Asterisks denote coefficients/comparisons that are significantly different from 0 (FDR correction; *p < 0.05, **p < 0.01, ***p < 0.001).
Fig. 5 Fitness effects of beneficial genes are correlated with evolutionary outcomes. We explored if genes with beneficial knockout fitness effects are correlated with A establishment of a mutation in a gene, and B changes in gene expression over evolutionary time, relative to neutral knockouts. A Slopes from logistic models, with presence of a mutation in a gene as the response variable. The fitness effects were normalized by the median beneficial fitness effect, so that coefficients can be interpreted as the average difference in log-odds establishment between neutral knockouts and the ‘typical’ beneficial knockout. REL606 beneficial knockout fitness is positively correlated with gene establishment probability for most environments, but in different time intervals, potentially pointing to shifting targets of selection. Error bars represent standard errors, which were calculated as detailed in methods section “Genome evolution”. FDR-corrected p-values were obtained from the logistic regression model. B We compared the distributions of log-fold change in expression between genes with neutral knockout fitness effects, less beneficial effects (lower 50%), and more beneficial effects (upper 50%). We used the change in expression from 0k gens (REL606) to 40k gens (L), from 6.5k gens (S) to 40k gens (S), and from 6.5k gens (L) to 40k gens (L) for the REL606, 6.5k S, and 6.5k L panels, respectively. The expression change between ancestor and 40k L (left) is nearly identical to the expression change between ancestor and 40k S as well as other timepoints (Fig. S25). Beneficial knockout fitness in REL606 is generally positively correlated with increasing gene expression over time. In S and L, fitness in several environments--including the ecological equilibrium and acetate and glucose growth--is correlated with decreasing gene expression. The center line of the boxplot is the median; the limits of the box represents the first and third quartiles; the whiskers extend to one times the inter-quartile range; outliers are not shown. FDR-corrected Mann–Whitney U tests were used to compare conditions. Asterisks denote coefficients/comparisons that are significantly different from 0 (FDR correction; *p < 0.05, **p < 0.01, ***p < 0.001).CC BY · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Research works printed but not linked: 1

  1. NCBI GEO. Evolution of gene expression during long term coexistence in a bacterial evolution experiment. https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE30639 (2012).

Cited by

Sorted by citations

See all 454

Related

See all

Metadata

Authors
Ascensao, J. A., Wetmore, K. M., Good, B. H., Arkin, A. P., & Hallatschek, O.
Affiliations
  • University of California, Berkeley (US)
  • Lawrence Berkeley National Laboratory (US)
  • Stanford University (US)
  • Leipzig University (DE)
Corresponding author
Hallatschek, Oskar
ORCID
Venue
Nature Communications
Publisher
Springer Science and Business Media LLC
Published
2023-01-16
Volume
14
Issue
1
Pages
248-248
ISSN
2041-1723 (ISSN-L)
DOI
10.1038/s41467-022-35677-5
PMC
PMC9842643
OpenAlex
W4316511358
Cited by
447
References
76
Citations by year
2026: 74 · 2025: 172 · 2024: 145 · 2023: 53
FWCI
73.88
Type
journal article
Language
en
Open access
Licence
Locations
  • Nature Communications (publishedVersion, cc-by)
  • bioRxiv (Cold Spring Harbor Laboratory) (acceptedVersion, cc-by-nc)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • eScholarship (California Digital Library) (submittedVersion, cc-by)
  • RePEc: Research Papers in Economics (submittedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
  • OSTI OAI (U.S. Department of Energy Office of Scientific and Technical Information) (submittedVersion)
Funders
  • U.S. Department of Energy (DE-AC02-05CH11231)
  • National Science Foundation (1555330)
Topics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Genetics › Evolution and Genetic Dynamics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Genetics › Bacterial Genetics and Biotechnology
  • Physical Sciences › Environmental Science › Ecology › Microbial Community Ecology and Physiology
Keywords
fitness landscape, Escherichia coli, long-term evolution experiment, gene knockout, transposon mutagenesis, invasion fitness, epistasis, ecological diversification, community composition, fitness effects, adaptive evolution
MeSH terms
Adaptation, Physiological, Escherichia coli, Mutation, Genetic Fitness, Ecotype

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review