Mutualism breakdown underpins evolutionary rescue in an obligate cross-feeding bacterial consortium

Abstract

Populations facing lethal environmental change can escape extinction through rapid genetic adaptation, a process known as evolutionary rescue. Despite extensive study, evolutionary rescue is largely unexplored in mutualistic communities, where it is likely constrained by the less adaptable partner. Here, we explored empirically the likelihood, population dynamics, and genetic mechanisms underpinning evolutionary rescue in an obligate mutualism involving cross-feeding of amino acids between auxotrophic Escherichia coli strains. We found that over 80% of populations overcame a severe decline when exposed to two distinct types of abrupt, lethal stress. Of note, in all cases only one of the strains survived by metabolically bypassing the auxotrophy. Crucially, the mutualistic consortium exhibited greater sensitivity to both stressors than a prototrophic control strain, such that reversion to autonomy was sufficient to alleviate stress below lethal levels. This sensitivity was common across other stresses, suggesting it may be a general feature of amino acid-dependent obligate mutualisms. Our results reveal that evolutionary rescue may depend critically on the specific genetic and physiological details of the interacting partners, adding rich layers of complexity to the endeavor of predicting the fate of microbial communities facing intense environmental deterioration.

Fig. 1. Testing how readily evolutionary rescue can occur in populations engaged in obligate mutualism.

a We utilized a pair of E. coli strains engaged in obligate mutualism based on amino acid exchange and exposed them to three treatments (No stress, salinity, and PNP) to observe if the consortium could avoid extinction through evolutionary rescue. b Abrupt environmental stress is expected to cause a decline in population density towards extinction, followed by three possible scenarios: extinction of both partners, adaptation of one partner that recovers while the other goes extinct, or adaptation and recovery of both partners.

Fig. 2. Evolutionary rescue prevents the extinction of bacterial consortium engaged in an obligate mutualism based on metabolic exchange.

a Population dynamics of the prototrophic strain (gray) and the obligate mutualism (green) in three different stress treatments (no stress, salinity, and PNP). The red and purple background indicate exposure to salinity or PNP, correspondingly. The experiment consisted of 48 independent populations for each treatment. b Box plots representing the transfer when consortia/populations begin to recover after stress exposure (OD₆₀₀ shifts from negative to positive trend), with n = 48 for salinity and PNP. P-value (***p < 0.01) was determined by a two-sided Mann–Whitney U-test (p-value ~ 10^-12). c Maximal rate of recovery computed as maximum change in population size per transfer for each experimental culture after stress exposure (OD₆₀₀/transfer), with n = 37 for salinity and n = 46 PNP. P-value (***p < 0.01) was determined by a two-sided Mann–Whitney U-test (p-value ~ 10⁻¹⁰). d Final population size of mutualistic communities (n = 48), calculated as the median of the last three transfers normalized relative to the median of the prototrophic strain at the final transfer. P-values (b, c) ***<0.001 (two sided Mann–Whitney U-test). All box plots display the interquartile range (IQR) of the data, with the horizontal line inside the box indicating the median. The whiskers extend to 1.5 times the IQR, showing the range of the data distribution. Source data are provided as a Source Data file.

Fig. 3. Only a single strain survives the stress and reverts to metabolic autonomy.

a Strains present at the end of the experiment. Colors indicate the presence of each strain in the consortium as detected by PCR. b Growth curves of recovered populations and of the auxotrophic ancestor of the ΔI strain in M9 without isoleucine addition. Different colors indicate the treatment from which the populations were isolated (purple from PNP, red from salinity and blue is the ancestor ΔI strain). Dots and error bars indicate the mean ± SD of measurements from three technical replicates of each of three evolutionary replicates (i.e. three different populations that were evolved in parallel during the evolutionary rescue experiment). Growth curves of individual replicates are included in Supplementary Fig. 3. Source data are provided as a Source Data file.

Fig. 4. The obligate mutualism is more susceptible to environmental stress than the prototroph.

Each panel shows the growth rates of prototrophic (gray) and mutualistic consortium (green) under different stressors: salinity (%), p-nitrophenol (PNP, mM), hydrogen peroxide (H₂O₂, µM), and spectinomycin (µM). Each box plot displays the interquartile range (IQR) of the data, with the horizontal line inside the box indicating the median. The whiskers extend to 1.5 times the IQR, showing the range of the data distribution (n = 4 for all boxplots except PNP (0) and H₂O₂ (0), where n = 6, and spectinomycin (0), where n = 3). The lines connecting the boxplots in each panel indicate the median growth rates for the prototrophic and mutualistic groups across different conditions. P-values (*p < 0.05, **p < 0.01, ns not significant) were determined by a two-sided Mann–Whitney U-test. Source data are provided as a Source Data file.

Fig. 5. Metabolic pathway mutations drive evolutionary rescue and mutualism breakdown.

a Growth rates and (b) yields of the mutualism, the prototrophic ancestor, and strains evolved in salinity or PNP. For this experiment, we included one population of the prototrophic strain (gray), six evolutionary replicates derived from salinity stress and five from PNP stress, and the ancestral M and I strains that constituted the mutualism. The data are presented as the mean ± SEM (n = 4). c, d Genes with mutations present in at least two replicate populations or occurring at a frequency > 0.3, represented by different colors (red for salinity and purple for PNP). Color shades indicate the frequency of each mutation, with darker shades indicating higher mutation frequencies. Specific mutations are detailed in Supplementary Tables 2 and 3. Source data are provided as a Source Data file.