Interspecific competition prevents the proliferation of social cheaters in an unstructured environment

Abstract

Bacterial communities are intricate ecosystems in which various members interact, compete for resources, and influence each other's growth. Antibiotics intensify this complexity, posing challenges in maintaining biodiversity. In this study, we delved into the behavior of kin bacterial communities when subjected to antibiotic perturbations, with a particular focus on how interspecific interactions shape these responses. We hypothesized that social cheating-where resistant strains shield both themselves and neighboring cheaters-obstructed coexistence, especially when kin bacteria exhibited varied growth rates and antibiotic sensitivities. To explore potential pathways to coexistence, we incorporated a third bacterial member, anticipating a shift in the dynamics of community coexistence. Simulations and experimental bacterial communities confirmed our predictions, emphasizing the pivotal role of interspecific competition in promoting coexistence under antibiotic interference. These insights are crucial for understanding bacterial ecosystem stability, interpreting drug-microbiome interactions, and predicting bacterial community adaptations to environmental changes.

Keywords: antibiotic impact; community coexistence; interspecific dynamics; kin bacteria; social cheating.

Figure 1

Growth rate determined intraspecific interaction outcomes in environments free from antibiotic interference; (A) C. testosteroni CNB-2/∆LuxR exhibited a higher growth rate than KF-1 across inoculum densities of 10⁵ to 10⁹ CFU/ml; effective growth rates of monocultures inoculated at different initial cell densities (mean ± SD, n = 6) were shown; the per capita growth rate referred to the population growth rate normalized by the initial population size; (B) co-culture of CNB-2/∆LuxR and KF-1 was diluted by a factor of 10 each day (1/10 of previous day’s culture transferred to fresh medium, with constant volume); (C) fast-growing CNB-2/∆LuxR dominated and drove slow-growing KF-1 to extinction (n = 9).

Figure 2

Sulfamethoxazole (SMX) exposure triggered intraspecific social cheating; (A) under 200 μg/l SMX exposure, KF-1 showed a higher growth rate than CNB-2/∆LuxR across inoculum densities of 10⁵–10⁹ CFU/ml; effective growth rates of monocultures at different initial cell densities (mean ± SD, n = 6) were shown; the per capita growth rate referred to the population growth rate normalized by the initial population size; (B) co-culture of CNB-2/∆LuxR and KF-1 was performed using serial dilution, revealing two stable states; (C) within these states, cooperator KF-1 dominated the community at low initial cell density, while cheater CNB-2/∆LuxR dominated at higher initial cell density; pink highlights 24-hour cycle under SMX exposure (n = 9).

Figure 3

Understanding community dynamics using a modified Lotka–Volterra (LV) model; (A) we applied a refined LV competition model, accounting for the per capita growth rate, self-inhibition, initial cell density, intraspecific interaction, and antibiotic-induced death rate; (B) the phase diagram depicted the model’s predictions on how antibiotic exposure influences community dynamics; the horizontal axis represented the growth rate of the Cheater B, while the vertical axis indicated the initial cell density; representative time series further illustrated these dynamics over time; as the initial population increased, community shifts were evident: moving from a dominance of Cooperator A to a dominance of Cheater B.

Figure 4

Regulator species enhanced species coexistence and biodiversity; phase diagram demonstrated profound influence of interaction strength (A–C) and growth rate (D) on community coexistence and dynamics; the simulation encompassed three categories of bidirectional interactions: competition (where both α_AC and α_BC were positive), mutualism (where both α_AC and α_BC were negative), and exploitation (where either α_AC or α_BC was positive).

Figure 5

Coexistence dynamics of three members in a well-mixed environment during antibiotic perturbation; (A) three-way interactions among KF-1, CNB-2/∆LuxR, and Pseudomonas aeruginosa; growth rates of three sulfamethoxazole (SMX) response phenotypes of P. aeruginosa (PAO1, PAO1-LuxR, and PAO1-Sul1) were measured under two conditions: (B) absence and (C) presence of 200 μg/l of SMX; the per capita growth rate referred to the population growth rate normalized by the initial population size. Equation depicted the relationship, while R² indicated the linear correlation between growth rate and initial cell density. In all regression analyses, the t-test yielded P-values <.05; (D) coexistence performance of a three-species community was evaluated using a serial dilution protocol; (E) and (F) showed the comparison of mono- and pairwise co-cultures; population size was quantified in CFUs/ml over time for monocultures and pairwise co-cultures; significance levels (^*P < .05) were assessed using the Wilcox test each cycle (Table S6 and S7); (G) our hypotheses proposed that increased interspecific interactions play a crucial role in promoting species coexistence in unstructured, well-mixed habitats when bacterial species encountered antibiotic perturbations.