Evolutionary and spatial controls on bacterial enzyme production

Background/Question/Methods

The breakdown of complex organic matter depends on the action of extracellular enzymes produced mainly by bacteria and fungi. These enzymes benefit the producer by enabling access to chemical resources that cannot pass through the cell membrane. However, enzyme producers are also vulnerable to cheating because enzymes act outside the cell and the reaction products may diffuse to competitors. Previous theoretical work predicted that cheating may drive enzyme producers to extinction under well-mixed conditions where all microbes have equal access to reaction products. These models also predict that limited diffusion promotes coexistence by allowing enzyme producers to recover a larger fraction of reaction products. We aimed to test these predictions in a laboratory system with Pseudomonas fluorescens bacteria under well-mixed and diffusion-limited conditions. Strains with or without the capacity to produce protease enzyme were grown either alone or in competition under different conditions. We used either complex or hydrolyzed protein as a growth substrate, and grew strains on liquid medium in shaken flasks (well-mixed) or on agar plates (diffusion-limited). The cultures were transferred to new medium every few days for up to 10 weeks, and we measured producer and cheater frequencies using plate counts.

Results/Conclusions

As predicted, we found that cheaters out-competed producer strains growing on complex protein under well-mixed conditions within 20 days. Producer frequency declined to zero, no protease enzyme was detectable in the medium, and culture growth declined >75%. We also observed this pattern in cultures where producer strains were grown alone on complex protein, suggesting that cheater strains evolved de novo in our experiment. In one experimental trial with mixed flasks, we observed a recovery in protease production and growth rates after 3 weeks, suggesting a genetic sweep that reestablished enzyme function. Results from diffusion-limited conditions were also broadly consistent with theoretical predictions. Cheater and producer strains coexisted for >10 weeks and there was no community-wide loss of enzyme function. Novel strains with intermediate levels of enzyme production began to emerge, suggesting that enzyme expression levels were evolving in a heterogeneous resource environment. Overall, our results imply that both evolution and spatial structure may increase diversity and maintain enzymatic function in bacterial systems.