Bacterial rock-paper-scissors

In evolutionary terms, “non-transitive” competitive effects—those without strict competitive hierarchies in which the biggest, best, or meanest win—have been shown to underlie processes such as the diversity of mating tactics observed in side-blotched lizards. A paper published in the March 25 issue of Nature suggests a similar dynamic exists between competing strains of Escherichia coli in the intestines of their host mice, but others remain to be convinced.Benjamin Kirkup and Margaret A. Riley at the Department of Ecology and Evolutionary Biology, Yale University, built on in vitro and in silico studies of antibiotic-mediated antagonism between strains of E. coli. Some strains (C) produce colicins, narrow spectrum antibiotics that kill other, colicin-sensitive (S) strains. In turn, S strains outcompete colicin-resistant (R) strains, which close the circle by outcompeting C. This is analogous to the game of rock–paper–scissors, no one strategy can prevail over the other two: scissors cut paper, paper covers stone, and stone blunts scissors.The work is “exciting” said Brendan Bohannan, who led the Stanford University team that in 2002 demonstrated this rock–paper–scissors dynamic in vitro, because “it's making that link from laboratory studies to the field—if you think of the mouse as the field for E. coli.”Kirkup and Riley fed mice with one of three E. coli strains, C, S, or R. These mice were then housed in groups of three, each individual mouse harboring a different strain, and their feces examined over the ensuing 12 weeks to see which strains dominated as bacteria were transmitted between mice.The authors observed that each mouse almost always harbored only a single strain, but the identity of the dominant strain changed over time, with cycles of replacement tending to follow the sequence predicted by the rock–paper–scissors model. Strains disappeared from a cage and then reappeared when their turn came. These re-invasions may be the result of mutations or of reestablishment from tiny, undetectable refuge populations lurking somewhere in the gut or cage, suggested Kirkup and Riley.However, while the in silico and in vitro studies suggested that the diversity of strains is maintained over time, colicin-producing strains tended to be lost eventually from the in vivo system. Nevertheless, Kirkup points out that strain diversity was maintained for longer within cages than within single mouse hosts, demonstrating the importance of discrete habitat patches in maintaining microbial diversity.Bohannan points to the small number of habitat patches, or mice, in Kirkup and Riley's experiments. “On a plate, we saw much longer maintenance of diversity, probably because we had larger population sizes and more spatial constraints,” he told The Scientist. “So, rather than having dozens or hundreds of individual populations that are spatially constrained, they had just three.”Kirkup cites logistical and animal welfare considerations as the reasons for the small population sizes in his experiments. “Perhaps with a larger population of mice, you'd have even more cycles of extinction and re-invasion,” said Bohannan.The findings could have “enormous” implications for population genetics, said Kirkup. “We have that little factor w that appears in all our beautiful equations, which is the fitness of a strain in the environment… All your fitness values in population genetics are based on transitivity,” he said. “Every type of bacteria we've looked at has some sort of bacteriocin… so it's quite reasonable to predict that we can find this nontransitivity in every taxonomic group of bacteria.” In which case, population genetics needs a major overhaul so it can deal with nontransitive systems, Kirkup said.Not everyone agrees that the textbooks need rewriting just yet. “This probably would be described as a special kind of frequency-dependent selection,” said Rolf Hoekstra of Wageningen University in the Netherlands, who worked on the in silico simulations. “It fits quite well into a tradition of population genetic thinking.”Nick Colegrave, an evolutionary ecologist at the University of Edinburgh, also has reservations. He told The Scientist that it remains to be seen how stable such a system would be were evolution permitted to operate over a longer time-scale. “We see it there; we can make it happen, and it works. But if the system is kept running for a year or so, maybe… a strain [would appear] that can beat everyone,” he said.

Bacterial rock-paper-scissors

bacteriocins may promote, rather than eliminate, microbial diversity

Interested in reading more?

Become a Member of