Increased privatization of a public resource leads to spread of cooperation in a microbial population

Abstract

The phenomenon of cooperation is prevalent at all levels of life. In one such manifestation of cooperation in microbial communities, some cells produce costly extracellular resources that are freely available to others. These resources are referred to as public goods. Saccharomyces cerevisiae secretes invertase (public good) in the periplasm to hydrolyze sucrose into glucose and fructose, which are then imported by the cells. After hydrolysis of sucrose, a cooperator retains only 1% of the monosaccharides, while 99% of the monosaccharides diffuse into the environment and can be utilized by any cell. The non-producers of invertase (cheaters) exploit the invertase-producing cells (cooperators) by utilizing the monosaccharides and not paying the metabolic cost of producing the invertase. In this work, we investigate the evolutionary dynamics of this cheater-cooperator system. In a co-culture, if cheaters are selected for their higher fitness, the population will collapse. On the other hand, for cooperators to survive in the population, a strategy to increase fitness would likely be required. To understand the adaptation of cooperators in sucrose, we performed a coevolution experiment in sucrose. Our results show that cooperators increase in fitness as the experiment progresses. This phenomenon was not observed in environments which involved a non-public good system. Genome sequencing reveals duplication of several HXT transporters in the evolved cooperators. Based on these results, we hypothesize that increased privatization of the monosaccharides is the most likely explanation of spread of cooperators in the population.IMPORTANCEHow is cooperation, as a trait, maintained in a population? In order to answer this question, we perform a coevolution experiment between two strains of yeast-one which produces a public good to release glucose and fructose in the media, thus generating a public resource, and the other which does not produce public resource and merely benefits from the presence of the cooperator strain. What is the outcome of this coevolution experiment? We demonstrate that after ~200 generations of coevolution, cooperators increase in frequency in the co-culture. Remarkably, in all parallel lines of our experiment, this is obtained via duplication of regions which likely allow greater privatization of glucose and fructose. Thus, increased privatization, which is intuitively thought to be a strategy against cooperation, enables spread of cooperation.

Keywords: CNV; SNV/indel; cheating; cooperation; microbe; public good system; sucrose; yeast.

Fig 1

(A) Sucrose utilization in S. cerevisiae is facilitated by the invertase Suc2p, which is secreted in the periplasm, where it hydrolyzes sucrose into glucose and fructose. Roughly 99% of the hydrolyzed monosaccharides are released into the media and are available to all members of the population. The remaining amount is retained by the cell that produces Suc2p. (B) Growth kinetics of BY4741 (wild type/cooperator) and suc2Δ (cheater) in (i) 0.2% sucrose, (ii) 0.2% glucose, (iii) 0.2% fructose, and (iv) 0.1% glucose and 0.1% fructose. The growth rates of the two strains are significantly different in sucrose (P-value <0.001) and statistically similar in glucose, fructose, and glucose-fructose (P-value >0.4), unpaired t-test. (C) After 1 day of growth in 0.2% sucrose, the ratio of frequency of cooperator BY4741 to the frequency of cheaters suc2Δ was approximately 3:7 in all six experimental lines. Cooperator frequency is in black; cheater frequency is in gray. The average and standard deviation of the six lines are shown in the figure. (D) The BY4741 and suc2Δ were grown in 0.2% glucose (Glu), 0.2% fructose (Fru), and 0.1% glucose + 0.1% fructose (Glu + Fru). After a day of growth, the ratio of frequency of cooperators to the frequency of cheaters was almost 1:1 across all media lines. Cooperator frequency is in black; cheater frequency is in gray. All experiments were performed in triplicate. The average and standard deviation are reported.

Fig 2

(A) The coevolution of cooperator BY4741 and cheater suc2Δ in 0.2% sucrose was started with 1:1 ratio of both genotypes. The frequency of cooperators in the co-culture significantly increased in the 200 generations (P-value < 0⁻¹⁰, unpaired t-test). Dotted line represents individual lines. The solid line represents the average of the six lines. (B) Growth kinetics of the ancestor cooperator (solid) and the coevolved cooperators (200 generations) of all six lines (dashed) in 0.2% sucrose. The growth rate of cooperators of lines are not statistically different from that of ancestor (P-value >0.5, unpaired t-test).

Fig 3

(A) In a competition assay conducted between coevolved cooperators (CCoop) (of all six lines) and ancestral cheaters (ACh), ~50% were cooperators. These results are significantly different than that from the competition between ancestor cooperator and ancestor cheater (P-value <10⁻⁶, unpaired t-test). (B) Competition assay between coevolved cheaters (CCh) (of all six lines) and ancestor cooperator (ACoop). The fraction of cheaters was ~70%. These results are not significantly different than the competition between ancestor cooperator and ancestor cheater (P-value ~0.35, unpaired t-test).

Fig 4

The frequency of the cooperator during coevolution of BY4741 and suc2Δ in different sugars: (A) 0.2% glucose: increase in the frequency of cooperators in the two lines (P-value <0.05, unpaired t-test); (B) 0.1% glucose + 0.1% fructose: there is no significant increase in the frequency of cooperators in any line (P-value >0.8, unpaired t-test); and (C) 0.2% fructose: there is no significant increase in the frequency of cooperators in any line (P-value >0.9, unpaired t-test).

Fig 5

The cooperator BY4741 was grown in 0.2% sucrose. Every 50 generations, competition assay was conducted with evolving cooperator (ECoop) and ancestor cheater (Ach) to quantify their frequency values. At the 200th generation, in the presence of ancestor cheater, the evolved cooperator of all three lines was performing better than the ancestor cooperator (P-value <0.0006 paired t-test). The ratio of frequency of cooperator to cheater does not vary significantly from the results observed in the competition between a coevolved cooperator (200th generation) and ancestor cheater (P-value ~0.16, unpaired t-test).

Fig 6

An increase in α leads to a non-linear increase in the fraction of cooperators in the population (x) which coexist with cheaters. The amount of sugar produced by each cell (r) has little impact on the proportion of the cooperators (x) when the sugar intake percentage (α) is constant.