Abstract. Eigen and Schuster's concept of the error catastrophe due to high mutation rates led Loeb to propose that pathogens might be eradicated by increasing their mutation rates to achieve lethal mutagenesis. Around the same time, theory on the evolution of mutational robustness was developed at SFI and elsewhere, which led people to wonder whether pathogens could evade lethal mutagenesis by evolving mutational robustness. Models were tried with large neutral networks to see if mutational robustness could forestall lethal mutagenesis, but they could not. But the question remained open whether pathogens could discover some sort of fitness landscape where robustness would protect against lethal mutagenesis. Here this question is definitively answered: for all possible fitness landscapes with at least two fitnesses, population mean fitness decreases with per-base mutation rate to the rate where there is no inheritance at all, and there will always be an extinction threshold if random sequences are inviable. However, "error catastrophes" have nothing to do with this threshold --- in fact, the mean fitness can never drop precipitously, but decreases at ever more gradual rates with increasing mutation. An illustration of the disconnect between "error thresholds" and fitness decrease is provided with a "quasispecies yo-yo", where increasing mutation rates cause the composition of a genome to switch back and forth multiple times between two bases, all the while through these dramatic genotype changes the mean fitness shows a smooth decline. Spectral landscape theory, also developed at SFI, is newly applied to evolutionary dynamics, and provides bounds on the mean fitness that show a deep connection between mutational robustness and mutational relaxation times.
Collins Conference Room
US Mountain Time
Lee Altenberg, University of Hawaii at Manoa
Our campus is closed to the public for this event.