Is it me you like, or my bacteria?

When two populations of a species evolve in different directions — perhaps because they live on separate islands, with different food sources or different dangers — at some point individuals from the two populations become unwilling to mate with each other.  This can increase the rate at which the two populations diverge, and thus the chance that they will actually become separate species.   A recent paper (Sharon et al. 2010 Commensal bacteria play a role in mating preference of Drosophila melanogaster, PNAS doi: 10.1073/pnas.1009906107) suggests that developing a mating preference may be much easier than we thought, and can be due not to genetic change but to changes in the bacteria carried by the population.

Sharon et al. were building on classic experiments from 1989 in which Diane Dodd raised flies on different food sources (one population was fed starch, the other maltose) and after 25 generations found a significant mating preference.  (Food is not the only way to do this: Kilias and Alahiotis had already shown that maintaining populations at different temperatures could also lead to reproductive isolation.)  The authors set out to reproduce the Dodd experiment, but instead of feeding one population on maltose they kept their control flies on normal fly food (cornmeal/molasses/yeast, or CMY) while the other was fed on starch.  After 11 generations, they put all flies back on CMY for a generation, to make sure that the food itself was not a factor, and tested for mating preference.  They found one, and it was strong: of 38 matings, 29 were between flies fed the same food and only 9 were between flies fed different foods.

The big surprise came, though, when they did the experiment again and looked at earlier time points.  The mating preference showed up even after only 2 generations on different foods.  By now it was clear that the preference was not due to genetic changes in the flies themselves — a gene that caused the change in mating preference wouldn’t have had anywhere near enough time to spread through the population — and Sharon et al. began looking for alternative explanations.  They repeated their experiment again, this time supplementing both foods (starch and CMY) with broad-spectrum antibiotics.  This time, they got no significant mating preference, suggesting that the flies’ gut bacteria were responsible for the preference for mating with flies that had been fed the same food.  And indeed, if you take bacteria from discarded media (which is covered in fly excretions) and use them to infect antibiotic-treated flies, the reinfected flies now show the appropriate mating preferences.  The biggest difference between the microbiota of the CMY flies and those of the starch flies is that the starch flies have a much higher percentage of Lactobacillus plantarum — 26% vs 3%.  In fact, infection with a pure culture of L. plantarum, obtained from starch-fed flies, is enough to make antibiotic-treated flies act like starch-fed flies.

So the story so far is that when flies are fed on starch, for some reason L. plantarum is able to expand, becoming a much larger component of the fly microbiota.  In turn, this leads to a change in mating preference, making starch-fed flies more attractive to each other and less attractive to CMY-fed flies.  How?

It’s possible that this effect is mediated through pheromones.  Drosophila’s sex pheromones are complicated: most of the known ones are part of a class of compounds called cuticular hydrocarbons, which appear to act over short distances or when flies touch each other. These have recently become possible to quantitate by GC-MS.  There are volatile pheromones as well, which are even less well understood but probably equally important.  Sharon et al. used GC-MS to look at the patterns of cuticular hydrocarbons in starch-fed, CMY-fed and antibiotic-treated flies. Some of the putative sex hormones show clear differences between starch-fed and CMY-fed flies, and some of these differences disappear on antibiotic treatment.  While I wouldn’t say that there’s clear proof that cuticular hydrocarbons are responsible for the changes in mating preference, at least it’s a plausible mechanism for how gut microbiota could affect sexual behavior.  In flies, anyway.

The authors argue that these results support the “hologenome theory of evolution”, a term they coined to explain the role of bacteria in the evolution of corals.  In essence, the theory is that natural selection acts on the combination of the host plus its associated microorganisms; because your microorganisms can affect your phenotype (by, for example, allowing you to digest something you otherwise couldn’t), changes in the bacteria you carry can be an advantage that can be selected for.  The concept seems uncontroversial to me, but the question is whether such selection can be important over evolutionary timescales.  How different do the food sources have to be? How long do the differences have to be maintained before a mating preference starts to make a real contribution to genetic isolation?  In the wild, could this effect be overcome by the equivalent of a breath mint and a quick application of deodorant, or are the pheromone changes substantial enough to make the unfortunate starch-fed fly unconquerably unappealing?  In any case, this is clearly another reason to nurture your gut bacteria; if you maltreat them, not only can they make you fat (perhaps), they can also make you smelly.  I guess we knew that.

Sharon G, Segal D, Ringo JM, Hefetz A, Zilber-Rosenberg I, & Rosenberg E (2010). From the Cover: Commensal bacteria play a role in mating preference of Drosophila melanogaster. Proceedings of the National Academy of Sciences of the United States of America, 107 (46), 20051-6 PMID: 21041648