Steering the mutation spectrum to probe the limits of evolutionary forecasting

Forecasts in wrinkly-spreader (WS) systems anticipated pathway usage but failed at some molecular details, with Pentz et al. (2024) observing no wspF mutations and unexpected adhesive compensation. This raises whether deviations arise from selection (fitness effects) or mutational biases. The project proposes “mutation-spectrum steering” by installing low-level, locus-specific mutational pressure using base editors or retron-driven SNV donors at candidate targets (wspF, wspA/E, alternative diguanylate cyclase promoters) while evolving WS phenotypes in multiple Pseudomonas species. Editors are calibrated to slightly elevate mutation rates at target loci without overwhelming natural variation. Outcomes are tracked via deep targeted sequencing to quantify how enhancing access to a locus changes evolutionary trajectories and fitness distributions. Contrasts are made across species with different c-di-GMP circuits and in environments with or without competitor species. This approach is novel because it actively perturbs mutation spectra to separate supply-limited from selection-limited outcomes, directly testing a core assumption behind forecasting failures. It complements pathway-level predictions with mechanism-disentangling experiments. The impact includes establishing when and how mutation spectra can steer or rescue forecast accuracy, informing strategies to direct evolution (e.g., to hinder biofilm formation or resistance) and refining general models of predictability across related species.

References:

1. Forecasting of phenotypic and genetic outcomes of experimental evolution in Pseudomonas syringae and Pseudomonas savastanoi. Jennifer T. Pentz, Aparna Biswas, Bassel Alsaed, Peter A. Lind (2024). bioRxiv.