Together with Daan de Groot and Erik van Nimwegen, both in the Basel Biozentrum, and Age Tjalma, currently at AMOLF (Amsterdam), Frank was involved in a study on microbial phenotype switching which has recently been published in PNAS. Daan and Age previously worked in our lab.
A fundamental limitation causes microbes to continuously struggle for their survival and evolutionary persistence. They have finite biosynthetic resources for protein synthesis, including energy and intracellular space, which bounds their protein concentrations. Cellular tasks such as adaptation to new conditions and biosynthetic rates generally improve at higher concentrations of the catalyzing proteins. Thus, they cannot excel at all fitness-contributing tasks at the same time, as they trade off. This is illustrated by E. coli, which adapts poorly to new conditions when it grows fast at nutrient excess, as it expresses then mostly growth-supporting proteins instead of preparatory proteins for future conditions. It does not do so at nutrient limiting conditions. It also does not have sensors for all environmental conditions.
So, how can it then be adaptive and evolutionarily successful? They bet!
They bet in a very special way. An isogenic population of microbes, all having the same DNA, consists of individuals with slightly different behaviours (phenotypes). This is inevitable, due to the underlying stochasticity of biochemical processes — the required protein burden to overcome this does not pay off. They control this phenotypic diversity to profit from it, by controlling the phenotype switching rates, as these alternative phenotypes are each adapted to slightly different futures. Since these deviating phenotypes are not all perfectly adapted to the current environmental state, having too many maladapted ones is a fitness cost. How can a microbial genotype (a species) cope with this?
Well, the fast-growing phenotypes resemble the best bets and should not switch to another phenotype. Thus, in a population of isogenic microbes the fast growing ones should have a lower switching rate than those that perform less well.
Microbes should, after all, not change their bet when they are winning! Like we do!
In our recent PNAS paper, we performed a general theoretical analysis of the fitness benefit of this growth-rate dependent strategy of phenotype switching. This had not been done before.
We found that reduced phenotype switching rates as function of growth rate has an enormous fitness advantage. We expect therefore that this has evolved often, across microbial species. And since this does not involve active sensing of the environment, it is likely a potent adaptation strategy across many different environmental conditions. Indirect physiological evidence of E. coli also points into the direction that it exploits this strategy.