|Ariel Weinberger|| at 11:00
Affiliation: University of California at Berkeley
Modeling Predicts Ancestral Viral Blooms Drive Old End Uniformity In CRISPR Loci
A major barrier to demonstrating virus-host coevolution in microbial systems is that most microbes are uncultivable. Metagenomic snapshots of Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR), a newly discovered microbial adaptive immune system against viruses and plasmids, offer cultivation-independent serial recordings of immunogenic, virus-derived ‘spacers.’ We present a population-scale mathematical model reconstructing the dynamics of CRISPR-driven virus-host coevolution from metagenomic snapshots of CRISPR loci across natural and laboratory populations. Our computationally evolved CRISPR patterns predict experimentally-measured loci, showing how virus-host coevolution drives the strikingly consistent pattern of old-end clonality and new-end diversity measured across microbial populations. Simulations predict that CRISPR old ends are conserved against metagenomically-recorded blooms of ancestral viral elements. CRISPR may be the first immune system tuned against persistent, temperate viruses, suggesting temperate phage therapy to select for CRISPR-laden microbes that block plasmid-borne pathogenicity and antibiotic resistance.