Pervasive Selection for Cooperative Cross-Feeding in Bacterial Communities - PubMed (original) (raw)
Pervasive Selection for Cooperative Cross-Feeding in Bacterial Communities
Sebastian Germerodt et al. PLoS Comput Biol. 2016.
Abstract
Bacterial communities are taxonomically highly diverse, yet the mechanisms that maintain this diversity remain poorly understood. We hypothesized that an obligate and mutual exchange of metabolites, as is very common among bacterial cells, could stabilize different genotypes within microbial communities. To test this, we developed a cellular automaton to model interactions among six empirically characterized genotypes that differ in their ability and propensity to produce amino acids. By systematically varying intrinsic (i.e. benefit-to-cost ratio) and extrinsic parameters (i.e. metabolite diffusion level, environmental amino acid availability), we show that obligate cross-feeding of essential metabolites is selected for under a broad range of conditions. In spatially structured environments, positive assortment among cross-feeders resulted in the formation of cooperative clusters, which limited exploitation by non-producing auxotrophs, yet allowed them to persist at the clusters' periphery. Strikingly, cross-feeding helped to maintain genotypic diversity within populations, while amino acid supplementation to the environment decoupled obligate interactions and favored auxotrophic cells that saved amino acid production costs over metabolically autonomous prototrophs. Together, our results suggest that spatially structured environments and limited nutrient availabilities should facilitate the evolution of metabolic interactions, which can help to maintain genotypic diversity within natural microbial populations.
Conflict of interest statement
The authors have declared that no competing interests exist.
Figures
Fig 1. Schematic overview over the focal genotypes and their corresponding growth performance.
(A) Genes deleted from the genome of the prototrophic wild type (WT) strain of Escherichia coli to yield mutants that are auxotrophic for the amino acids arginine or lysine (AUX1, AUX2), a mutant that overproduces a mixture of both amino acids (OP), and cross-feeding genotypes that are auxotrophic for one, yet produce increased amounts of the respective other amino acid (CF1, CF2). (B,C) Experimentally determined growth performance of all focal genotypes in response to different concentrations of the amino acids (B) arginine and (C) leucine. Amino acids were applied in a mixture of amino acids mimicking the blend of amino acids produced by the overproducer. Growth rate (μ) per hour is the experimentally determined Malthusian parameter during 24h of growth. Lines represent fitted Monod kinetics for auxotrophic and cross-feeding genotypes and the calculated mean for the prototrophic (WT) strain (red) as well as the genotype overproducing amino acids (blue). See S1 Text and S3 Fig for further information.
Fig 2. Auxotrophic and cross-feeding genotypes are selectively favored under a broad range of conditions.
Shown are representative simulation results after 100 simulation steps. Parameters analyzed include the benefit-to-cost ratio where the experimentally determined values were computationally in- or decreased (x-axes), the degree of metabolite diffusion in the environment (y-axes) ranging from low (structured environment) to high (unstructured environment), and the environmental availability of amino acids (A,B) including (A) no amino acids are available in the environment, and (B) substantial additional availability of amino acids in the environment. Color-code of genotypes: red = wild type, blue = amino acid overproducer, yellow = amino acid auxotrophs (2 types), green = cross-feeding genotypes (2 types).
Fig 3. Prevalence of unilateral and bilateral cross-feeding.
Sizes of circles represent the mean proportion of (A, C) unilateral and (B, D) bilateral cross-feeding that included all interactions depicted in the schematic above. Colors of circles represent the mean connectivity between cell types that measures the co-occurrence of genotypes and thus reflects the probability of genotypes to display one or the other type of cross-feeding. 100 simulation runs were analyzed per parameter combination. Simulation results with (A, B) no or (B, D) a substantial addition of amino acids to the environment are displayed. Parameter combinations that were analyzed in each panel include the benefit-to-cost ratio (x-axes) and the degree of metabolite diffusion in the environment (y-axes) ranging from low (structured environment) to high (unstructured environment).
Fig 4. Metabolic cross-feeding increases genotypic diversity within bacterial populations.
Mean Shannon-Weaver diversity indices (H ± standard deviation) of simulated populations with varying benefit-to-cost ratios (BCR) are shown. Simulations were performed in the presence (blue line) or absence (red line) of environmentally supplemented amino acids in (A) spatially structured (i.e. low diffusion) and (B) unstructured environments (i.e. high diffusion). The dashed line indicates the maximally achievable diversity index for six genotypes. Asterisks indicate significant differences between the amino acid supplemented- and unsupplemented environment for a given BCR (FDR-corrected two-sample t-test: *** P < 0.001, ns: _P_ > 0.05, n = 50).
Fig 5. Population dynamics in environments without amino acid supplementation.
Repeated simulations (n = 100) are plotted for varying benefit-to-cost ratios (BCR) and degrees of amino acid diffusion (bold line: mean, shaded ribbon: standard deviation). All simulations start with a random distribution of all genotypes and undergo a specific dynamic alternation of genotype frequencies. Depending on the genotype’s strategy, it can repress, facilitate, or even outcompete others (see text for more details). Legend: red = wild type, blue = overproducing genotype, yellow = auxotroph (2 types), green = cross-feeder (2 types).
References
- Conrad R. Soil microorganisms as controllers of atmospheric trace gases (H2, CO, CH4, OCS, N2O, and NO). Microbiological Reviews. 1996;60(4):609–640. Available from: http://mmbr.asm.org/content/60/4/609.short. - PMC - PubMed
- Falkowski PG, Fenchel T, Delong EF. The microbial engines that drive Earth’s biogeochemical cycles. Science (New York, NY). 2008. May;320(5879): 1034–1039. Available from: http://www.ncbi.nlm.nih.gov/pubmed/18497287. 10.1126/science.1153213 - DOI - PubMed
- Torsvik V, Sørheim R, Goksøyr J. Total bacterial diversity in soil and sediment communities—a review. Journal of Industrial Microbiology. 1996;17:170–178. Available from: http://link.springer.com/article/10.1007/BF01574690. 10.1007/BF01574690 - DOI - DOI
- Parfrey LW, Knight R. Spatial and temporal variability of the human microbiota. Clinical Microbiology and Infection. 2012;18(SUPPL. 4):8–11. - PubMed
Publication types
MeSH terms
Substances
Grants and funding
Funding by the International Max Planck Research School (IMPRS) to KB, the Volkswagen Foundation to CKo and the Jena School of Microbial Communication (JSMC) to CKo, SS and SP is gratefully acknowledged. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
LinkOut - more resources
Full Text Sources
Other Literature Sources