Leaders in Pharmaceutical Business Intelligence (LPBI) Group

Evolutionary Change

Larry H. Bernstein, MD, FCAP, Curator

LPBI

Change accelerates stalled evolution

biomolbioandco    https://biomolbioandco.wordpress.com/2015/11/17/change-accelerates-stalled-evolution/

Evolution can get stuck. When no mutations are available that can improve the fitness of an organism, evolution cannot proceed. However, using environmental fluctuations that are ubiquitous in nature, evolution can proceed. In the preprint that we posted earlier on the bioRxiv and which has recently been published in PNAS we describe how a transcription factor and its binding site, constraint in the different constant environments, can evolve in fluctuating environments.

Stalled evolution

The accessibility of phenotypes in different environments depends on the wiring of the genotype-phenotype map and the translation of this map into fitness. When sub-optimal genotypes are surrounded by valleys in the adaptive landscape, neighboring genotypes are inaccessible and evolution is unable to proceed by single mutational step-by-step positive Darwinian evolution (Figure 1).

To investigate the connectedness of the fitness increasing genotypes in sequence space we used the lac regulatory system in Escherichia coli. Decades of work have elucidated the physiology of the lac regulatory system. And it is known that a few base pairs in the operator DNA, and a few amino acid residues in the transcription factor are responsible for specificity of binding of these operator-repressor pairs. We here experimentally constructed mutants with mutations in these specific residues (6 residues, yielding 2⁶=64 mutants in total) that together constitute 6!=720 direct trajectories.

Assessing the ability to repress the lac operon in one environment (without a lactose-sugar-derivative), and the ability to express the lac operon in the other environment (with a lactose-sugar-derivative), we find that none of the mutational trajectories allows continuous improvements along the mutational trajectories in one of the constant environments. Interestingly, we find that alternating between these environments doesallow for constant improvements along the mutational trajectories. We find that the shortest route towards the final genotype is of Hamming distance 6, there is thus a direct mutational trajectory from the starting sequence to the final combination of transcription-factor and binding site. With a computational method that describes the mutational and environmental transitions as a Markov process, we can further calculate the crossing rates from the initial to the final genotype for all trajectories in the landscape, allowing mutational detours. We find that the crossing rate is maximal when the environmental switches are on the order of the mutation rate (or the rate at which a sweep can be completed).

Tradeoffs are crucial for crossing the adaptive landscape in fluctuating environments

Cross-environmental tradeoffs are responsible for these continuous improvements. In the most extreme case, suboptimal peaks are translated into valleys, by which constraints are resolved. In the more subtle case, descending slopes are turned into ascending slopes upon environmental change, allowing adaptive trajectories to surf over the slopes with positive selective coefficients. Evolutionary constraints can thus be overcome by the environment-dependent ‘ratcheting’ that allows the crossing of otherwise inaccessible regions in sequence space (Figure 2).

We think that this research not only aids the fundamental understanding of ecological and evolutionary transitions in fluctuating environments, but that it can also help rethinking the evolutionary optimization of certain biotechnological processes (for instance the production of antibodies). And in addition, it cautions against the use of cyclic multi-drug protocols in clinical treatments, as these might potentially increase the speed of adaptation to the drugs, instead of halting it.

Breaking evolutionary constraint with a tradeoff ratchet
M.G.J. de Vos, A. Dawid, V. Sunderlikova, S.J. Tans
PNAS, (2015) | http://dx.doi.org/10.1073/pnas.1510282112