How type II CRISPR-Cas establish immunity through Cas1-Cas2-mediated spacer integration.

Abstract

CRISPR (Clustered Regularly Interspaced Short Palindromic Repeats) and the nearby cas (CRISPR-associated) operon establish an RNA-based adaptive immunity system in prokaryotes1–5. Molecular memory is created when a short foreign DNA-derived prespacer is integrated into the CRISPR array as a new spacer6–9. Whereas the RNA-guided CRISPR interference mechanism varies widely among CRISPR-Cas systems, the spacer integration mechanism is essentially identical7–9. The conserved Cas1 and Cas2 proteins form an integrase complex consisting two distal Cas1 dimers bridged by a Cas2 dimer in the middle6,10. The prespacer is bound by Cas1-Cas2 as a dual forked DNA, and the terminal 3′-OH of each 3′-overhang serves as an attacking nucleophile during integration11–14. Importantly, the prespacer is preferentially integrated into the leader-proximal region of the CRISPR array1,7,10,15, guided by the leader sequence and a pair of inverted repeats (IRs) inside the CRISPR repeat7,15–20. Spacer integration in the most well-studied Escherichia coli Type I-E CRISPR system further relies on the bacterial Integration Host Factor (IHF)21,22. In Type II-A CRISPR, however, Cas1-Cas2 alone integrates spacer efficiently in vitro18; other Cas proteins (Cas9 and Csn2) play accessory roles in prespacer biogenesis17,23. Focusing on the Enterococcus faecalis Type II-A system24, here we report four structure snapshots of Cas1-Cas2 during spacer integration. EfaCas1-Cas2 selectively binds to a splayed 30-bp prespacer bearing 4-nt 3′-overhangs. Three molecular events take place upon encountering a target: Cas1-Cas2/prespacer first searches for half-sites stochastically, then preferentially interacts with the leader-side CRISPR repeat and catalyzes a nucleophilic attack that connects one strand of the leader-proximal repeat to the prespacer 3′-overhang. Recognition of the spacer half-site requires DNA bending and leads to full integration. We derive a mechanistic framework explaining the stepwise spacer integration process and the leader-proximal preference.