Motivated by a desire to understand the mechanistic bases of evolutionary change much of our work involves the genetic analysis of mutant types arising from various selection experiments. But sometimes we do genetics because it’s simply fun. And once one has the bug then it’s hard to resist the temptation that comes from solving “what caused it” puzzles in much the way that a detective solves “who done it” crimes.
Genetics of niche specialistion
Our first forays into evolutionary genetics were motivated by the niche specialist mutants arising during the Pseudomonas adaptive radiation. The work was begun by Sophie Kahn, one of Rainey’s earliest D. Phil. students at Oxford (see Spiers et al 2002). She did a wonderful — pioneering — job, back in the days before we had a genome sequence and where identifying the genomic location of transposons required Southern blotting, genomic libraries and cloning pieces of impossibly large DNA using restriction enzymes whose cleavage positions relative to one another were unknown. Things have progressed: we now have the genome of the ancestor sequenced (Silby et al 2009) and armed with this reference it becomes possible to take advantage genome re-sequencing technologies, RNA-seq and Tn-seq to wonderful effect.
Over the years, while continuing to study the genetics of niche specialist mutants (e.g., Spiers et al 2002, Knight et al 2006, Bantinaki et al 2007, Ferguson et al 2009, McDonald et al 2009, Farr et al 2017), with a lot of important work on cellulose by Andrew Spiers, we have also delved into the mechanistic bases of stochastic switching genotypes that arose both through the reverse evolution experiment of Bertus Beaumont and through work on the evolution of multicellularity by Katrin Hammerschmidt and Caroline Rose — even the genetic switch of phage lambda with Dominik Refardt.
Genotype-to-phenotype map / rules of adaptive evolution
More recently, and arising from detailed study of the genetics of wrinkly spreader mutants, has come appreciation of the significance of the genotype-to-phenotype map and the role it plays in affecting the likelihood that mutant phenotypes are realised. Mike McDonald‘s work (McDonald et al 2009) showed that the function of regulatory components and their inter-connectivities is a contributor to parallel molecular evolution. Given this, it seemed reasonable to consider the possibility that better understanding of the genotype-to-phenotype map might deliver insight into rules by which new adaptive phenotypes arise (Lind et al 2015). This proved to be true as elegantly demonstrated by Peter Lind and Eric Libby who achieved this via a combination of experiment and mathematical modelling (Lind et al 2019).
At the present time Dave Rogers, Michael Barnett and Joanna Summers continue the core of our evolutionary genetics with focus on the phenotype-to-genotype map and the biases it brings to evolutionary change, and on the genetics of developmental regulation arising during the course of the life cycle experiment looking at the origins of multicellularity. On top of this Dave Rogers has spearheaded re-construction of tools built last century so that they now function far more effectively. He’s also developed a host of new tools for analytical genetics of Pseudomonas.
Returning to past work — much of which remains unpublished — that has proved important in many ways is the reverse evolution experiment mentioned above. The experiment involved repeated propagation of 12 replicate lines in two contrasting environments (static and shaken microcosms). As soon as a new adaptive mutant arose in for example the static (unshaken) microcosm it was isolated and re-introduced into a shaken microcosm where it encountered an environment to which it was maladapted. A further compensatory mutation was therefore required, which once obtained, allowed the mutant to be isolated and returned to a fresh unshaken microcosm. The schema is shown below.
Remarkably most lines continued through eight full reversals and many were continued through further rounds by Andy Farr. The fact that lineages could continually reverse their phenotype (by spontaneous mutation), without a reduction in fitness, gave substance to ideas about two-phase life cycles that proved central to our work on nascent developmental programmes and the of evolution of multicellularity (Hammerschmidt et al 2014).
It is not completely true that all work from this experiment remains unpublished (and Rainey vows to have the paper submitted by the end of 2019): we have dissected two lines that gave rise to stochastic switching (bet-hedging) types. The first was part of Bertus’s primary publication (Beaumont et al 2009) with molecular details that took years to understand and threw us (and Jenna Gallie in particular) into the depths of despair and central metabolism (Gallie et al 2015, Gallie et al 2019). Not put off by either despair or central metabolism, Philippe Remigi delved further and deeper, and discovered unexpected connections between switching and ribosomes (Remigi et al 2019). A notable feature of Philippe’s work was connection of the switching behaviour that arose during course of the reverse evolution experiment to its ecological significance in the ancestral type (in the wild), where, we think, provisioning of ribosomes occurs during stationary phase in order to ensure rapid exit from slow growing (or no-growth) states.