Thread by @JoshCofsky

New pre-print! While #CRISPR-Cas9 searches through the genome, how does it actually "read" DNA sequence information, which is normally hidden within the double helix? We determined structures of Cas9 in the act of DNA interrogation to find out! Thread below 👇 (1/8)

A key step in RNA-guided target search is breaking open DNA base pairs next to the PAM. It'd be great to see what that process looks like, but it's difficult to observe because it occurs so quickly (estimated lifetime <30 ms). (2/8)

We used an old chemical trick from the Verdine Lab to staple Cas9 to a piece of DNA that has a PAM but is not complementary to the guide RNA, holding the interrogation complex together long enough for us to do experiments on it. (3/8)

Using cryo-EM, we caught the complex in two different states—one in which the DNA is linear (a putative DNA scanning state), and one in which the DNA is bent/twisted in a way that exposes nucleobases to be checked for complementarity to the guide RNA. (4/8)

Structural modeling suggests that Cas9 may engage a twisting motion between its two lobes to manipulate the DNA in this way. (5/8)

My favorite part: we also revived some old-school solution biochemistry experiments (cyclization efficiency and permanganate reactivity measurements) to make sure the EM data weren't fooling us. (6/8)

The speed with which Cas9 performs these bending actions may determine how fast it can find its target. Also, Cas9 may interact differently with pre-bent DNA sequences (such as those wrapped around histones). Both important considerations for genome editing efficiency! (7/8)

Thanks to @berkeleyMCB and all my wonderful co-authors (Kasia Soczek, @gavinjknott, @NogalesLab, @doudna_lab) for making these discoveries possible and teaching me a thing or two about structural biology 🧬 (8/8)