COS 69-4 - The robustness of coevolutionary dynamics in CRISPR-mediated microbe-lytic virus systems

Background/Question/Methods

Microbial hosts are known to evade viral infection through a number of resistance mechanisms, including adaptive immunity obtained via the CRISPR system and its CRISPR-associated (Cas) proteins. Previous computational models have shown that lytic viruses and microbes with CRISPR-Cas immunity can exhibit characteristic coevolutionary dynamics exhibiting alternation between: i) viral escape and co-diversification of both partners, and ii) host control over viral proliferation, leading to viral extinction. The alternating dynamics of this inverse matching-allele model differ from the two most recognized temporal behaviors of antagonistic coevolution, namely fluctuating selection and arms-race dynamics, as observed in matching-allele and gene-for-gene models. To better incorporate extinction and invasion of new strains, we develop a fully stochastic model of CRISPR-mediated microbe-lytic virus coevolution. We computationally implement this model to examine the robustness of the alternating dynamics to parameter variation, and to identify the processes that generate the characteristic transitions. Patterns of clade and strain-level phenotypic variation that explain and anticipate dynamical transitions are also investigated.

Results/Conclusions

Resulting phase diagrams show that the alternating coevolutionary mode is robust within a broad range of ecological and evolutionary dimensionless parameter values. The diagrams exhibit an increasing trend in the probability of occurrence and the number of alternations with the probability of mutant virion production. For low spacer acquisition probabilities, the system exhibits successive strain replacement similar to that observed in phylodynamic pathogen models. Resource competition between microbes is a key process behind the alternating dynamics, that induces cascades in the dominance of different clades and anticipates viral escape. In turn, the characteristic non-stationarity of both microbial and viral strain diversity, and of the number of matches conferring protection, explain eventual viral extinction. Our results establish the generality of the alternating coevolutionary mode in microbe-lytic virus dynamics, and therefore, the feasibility of observing it in nature or laboratory settings. They further provide an expectation for phylogenetic patterns of CRISPR-mediated microbial and lytic-viral interactions, which can aid empirical inference of selective processes at play in complex microbe-virus coevolutionary systems.