The evolution of evolvability

I will discuss the intriguing, albeit controversial, hypothesis that life has evolved to evolve. Controversy arises because such capacity would appear to require foresight on the part of natural selection. Additional discomfort stems from the fact that evolvability is a property of lineages — not of individuals — and thus requires selection to operate at levels above that of single entities. Motivated by desire for mechanistic insight we turned to experimental populations of bacteria and performed a three-year-long, real-time evolution experiment, in which lineages were required to compete for growth under alternating conditions, with success depending upon capacity to mutate between two phenotypic states. Lineages that failed to evolve target phenotypes within a set time went extinct and were replaced by successful lineages. During the course of the experiment a lineage emerged seemingly able to anticipate the future via localised hyper-mutation. I will describe the moment-by-moment birth of this “contingency” behaviour and document experiments that reveal an innate capacity for further evolution. I will conclude by emphasising the critical importance of selection working at two levels: while individual-level selection repeatedly drove cell populations between the same two phenotypic states, the genetic underpinnings of these phenotypes were free to diverge, fuelling an exploration of evolutionary potential, the consequences of which emerged on the timescale of lineages. Ultimately, this exploration generated the variation necessary for construction and cumulative refinement of a lineage-level adaptive trait. More generally, the study clarifies conditions by which evolvability can itself evolve adaptively.