Horizontal Gene Transfer Promotes Altruism

The origin of altruism is one of the enduring mysteries of evolutionary biology. While social insects like ants and bees are the mainstay of research into the evolution of altruism, cooperation between social bacteria is also a fruitful avenue of research. Many bacteria produce so-called "public goods" molecules, secreted chemicals which enhance the growth not only of the bacterium secreting them but also of its neighbours. These molecules are costly to produce, so it's surprising that bacteria produce them. Why take on the cost of making something that benefits your neighbours? What's to stop cheaters from taking advantage of the public goods molecules without producing any of their own? In a paper appearing in PNAS, a team of researchers based in France and the UK have offered an intriguing answer, suggesting that horizontal gene transfer between bacteria may solve both problems.

In addition to passing on their genes via reproduction, bacteria can also transfer a copy of them directly to other individuals; these 'horizontal' transfers spread the transfered gene across the community and transform its evolutionary tree into a tangled bush. Antibiotic resistance genes spread quite quickly by horizontal transfer, even hopping between species. It may seem odd for a bacterium to share its ability to resist antibiotics with its neighbours, since that reduces its competitive advantage, but from the perspective of the resistance gene, horizontal transfer is just another way to make more copies of itself -- that is, to increase its fitness. The tension between the different levels at which selection happens -- gene, individual bacterium, or population of bacteria cells -- is at the heart of the question of altruism. Selfish genes happily spread through horizontal transfer, and that transfer can proceed essentially unchecked so long as they don't harm the individuals that acquire them. But what happens when genes for cooperative behaviour get transfered? What's the fate of a selfish gene which is costly to the individual but beneficial to the group?

The team addressed this question by using a combination of computer models and genetically engineered bacteria to investigate how horizontal gene transfer affects the fate of a 'producer' gene -- that is, one which produces public good molecules. They found that a producer gene's fate depends on how the bacterial population is structured. In a single, well-mixed population -- one with no internal structure or barriers to horizontal transfer -- producer genes are out-competed by non-producers, even when both can be transferred horizontally. In other words, the cheaters win. The producer gene can move through the population, making bacteria secrete public goods molecules to everyone's benefit...but so can the non-producer gene, and individuals which have that gene will benefit from the public goods without paying any cost.

Things change if the population has some internal structure, with subpopulations that have little or no transfer between them. There are still cheaters and producers within each subpopulation, and the cheaters still do better within the subpopulation. But the subpopulations are also all competing with each other, and those with a higher proportion of producers grow faster than those dominated by cheaters. The population structure creates a tension between competition between individuals (i.e., within each subpopulation) and between subpopulations, and that tension between different levels of selection creates an opportunity for the producer gene to spread. As long as the conditions are right for competition between subpopulations -- that is, as long as there's some variation in the initial cheater/producer ratio between subpopulations and there isn't too much migration -- then the group-level selection will favour an increase in the producer allele even though it's at a disadvantage at the individual level. Horizontal transfer even offers some protection against migration, since the gene can make its way into immigrant cheaters and transform them into producers, though higher levels of migration will still overwhelm the group.

Group selection has something of a checkered history in evolutionary biology. The idea that a particular trait evolved "for the benefit of the group (or the species)" is a common mistake fueled by a naïve misunderstanding of evolutionary dynamics. Selection at the level of the individual will favour those who enjoy the benefits of being in a group without giving anything back -- cheaters. Eventually, they'll dominate the group, even if it would have been better for everyone to cooperate. Even though groups of cooperators will out-compete those made up of cheaters, the cooperative groups are vulnerable to being taken over by cheaters. Evolutionary biologists have borrowed a term from economics to describe this impasse, the 'tragedy of the commons'. Unless groups have a strongly cohesive identity, they're always vulnerable to cheating, which is favoured by individual-level selection.

In many social systems (including ants and bees), kinship provides the cohesion required to overcome the tragedy of the commons. As long as altruism is directed towards kin, its benefits stay within the group, and group-level selection prevails over individual-level selection. It turns out that horizontal gene transfer can play a similar role in bacteria. Horizontal transfer ensures that others in the group also have the producer gene, strengthening the role of group selection and enabling cooperative groups to out-compete those riddled with cheaters. The evolution of altruism depends on balancing the tension between selection at the levels of groups and individuals. The combination of horizontal gene transfer and population structure is powerful enough to pull that off, joining kinship as a mechanism for creating viable group selection dynamics. Given that, I'd like to leave you with an intriguing question to ponder: what about the combination of population structure and horizontal meme transfer?


Ref
Dimitriu, T et al. Genetic information transfer promotes cooperation in bacteria. PNAS; published ahead of print. (2014) 10.1073/pnas.1406840111