Microbes in space: The importance of spatial organization in evolution

Abstract

My research uses a multi-prong approach to understand the ecological and evolutionary processes that drive and maintain diversity. Specifically, I examine how competition and spatial structure drive evolutionary change. Part 1. A population under selection to improve one trait may evolve a sub-optimal state for a second trait due to tradeoffs and other evolutionary constraints. How the duration of evolution under conditions favoring the first trait affects the capacity of a population to adapt when conditions change to favor the second trait is an open question. We investigated this question using isolates from a lineage spanning 60,000 generations of the Long-Term Evolution Experiment (LTEE) with Escherichia coli. In the LTEE, cells have access to a shared pool of resources, and the bacteria have evolved increased competitive ability and a concomitant reduction in numerical yield. Here, we shifted the focus of selection to numerical yield by propagating LTEE-derived populations in media-in-oil emulsions, in which cells grew in isolated patches with private resources. We found that the length of time evolving under shared resources (i.e., duration in the LTEE) did not affect the ability to re-evolve toward higher numerical yield. The evolution of numerical yield in emulsified populations commonly occurred through mutations in the phosphoenolpyruvate phosphotransferase system. These mutants exhibit slower uptake of glucose, which makes them poorer competitors for public resources. However, these mutants also produce smaller cells that release less carbon as overflow metabolites, which makes them more numerically productive when resources are private. Our results demonstrate that mutations that were not part of adaptation under one selective regime may enable access to ancestral phenotypes when selection changes to favor evolutionary reversion. Part 2. Chemical warfare in the microbial world is ubiquitous. One kind of chemical weapon microbes employ is a proteinaceous toxin called a bacteriocin. The best-studied bacteriocins are the colicins, produced by and active against Escherichia coli. Many colicin systems encode suicidal lysis genes, such that the producing cell releases the toxin through cell lysis. Released colicin kills sensitive competing cells allowing immune clones of the producer to capitalize on the liberated resources. However, a major group of colicin systems lack this lysis gene, and it has been unclear how such toxins were released from the producing cell. I explore a newly discovered union between such a colicin system and a prophage (bacterial virus that has been incorporated into the bacterium’s genome), where release of the colicin occurs via phage-encoded cell lysis. This union should be detrimental to susceptible cells, which would be hit twice: once by the colicin, once by the phage. Mathematical modeling shows that the success of this dual killing system depends on the prophage’s ability to produce infectious virions in a well-mixed environment. If the prophage is cryptic (encoding lysis, but not producing infectious phage), then this dual system is susceptible to invasion by “cheaters” that only possess the colicin system (and its immunity) but do not lyse. I use agent-based simulations to model this scenario in a spatially structured environment. Under spatial conditions, the social dilemma and is resolved when the population exhibits fluctuating rock-paper-scissors dynamics (where rock chases scissor, scissor chases paper, and paper chases rock), providing insights into what ecological conditions were necessary in order to maintain this union.