Self-organized division of labor and selection for long-term antiviral defense in CRISPR arrays.

CRISPR-Cas is an adaptive immune system that safeguards prokaryotes from viruses and other genetic parasites. The system uses short DNA sequences, called spacers, which are acquired during encounters with parasites and stored in the CRISPR array. The mechanisms behind spacer acquisition are well understood, but it is unclear whether CRISPR-Cas functions as a short-term or long-term defense mechanism.

The limited size of CRISPR arrays creates a trade-off between acquiring new spacers to address recent infections and retaining old spacers for long-term defense. Researchers developed a stochastic model to explore how CRISPR-Cas systems address this trade-off. The simulations show three evolutionary regimes, leading to short-term, long-term, and dual memory arrays.

The regime is determined by the ratio of two measurable parameters: the length of the CRISPR array and the time needed for transient epidemics to drop below basal endemic viral abundance. A new phenomenon, spontaneous division of labor, emerges in the dual memory regime. Spacers near the leader and distal ends of the CRISPR array specialize in short-term and long-term memory, respectively.

This specialization results in a U-shaped profile of the probability of encountering CRISPR targets within the local virome. This observation, which has been empirically observed in CRISPR arrays from the human gut microbiome, could guide future research on the environmental persistence of poorly understood viromes.