Gain-of-function experiments with bacteriophage lambda uncover residues under diversifying selection in nature

Maddamsetti, Rohan; Johnson, Daniel T.; Spielman, Stephanie J.; Petrie, Katherine L.; Marks, Debora S.; Meyer, Justin R.

The congruence we find between laboratory and natural evolution in λ contrasts with work showing that adaptation in Lenski's long-term experiment anti-correlates with natural protein variation in E. coli (Maddamsetti et al. 2017) . Presumably, the simplicity of Lenski's experiment leads the bacteria down evolutionary paths not taken in nature. By contrast, the dominant selection pressure in evolution experiments with bacteriophageattachment to bacterial host cells-is probably also a dominant selection pressure on wild phage. Notably, another study on bacteriophage has observed residues under positive selection in both the lab and nature (Wichman et al. 2000) , as have studies on canine parvovirus (Allison et al. 2014 ) and poliovirus (Stern et al. 2017) . Apparently, the reverse strategy-identifying candidate residues for changes in phenotypes in natural populations and then validation in the laboratory-does not work as well, at least in one notable case (Liu et al. 2017) . One reason put forth in Liu et al. (2017) for the failed nature-lab connection are overlooked effects of higher order epistasis (Weinreich et al. 2013) . Indeed, our methodology revealed that a four-way interaction among the J mutations was required to endow new function on OmpF, and so we would not have been able to uncover them by typical methods that only measure the effects of single or pairs of mutations.