Proofreading mechanism of Class 2 CRISPR-Cas systems

Aset Khakimzhan, David Garenne, Benjamin I. Tickman, Jason Fontana, James Carothers, Vincent Noireaux

View at source ↗PDF ↗Preprint1 citations
Preprint: not peer reviewed. Results may change in later versions.

Abstract

CRISPR systems experience off-target effects that interfere with the ability to accurately perform genetic edits. While empirical models predict off-target effects in specific platforms, there is a gap for a wide-ranging mechanistic model of CRISPR systems based on evidence from the essential processes of target recognition. In this work, we present a first principles model supported by many experiments performed in vivo and in vitro. First, we observe and model the conformational changes and discrete stepped DNA unwinding events of SpCas9-CRISPR target recognition process in a cell-free transcription-translation system using truncated guide RNAs, confirming structural and FRET microscopy experiments. Second, we implement an energy landscape to describe single mismatch effects observed in vivo for spCas9 and Cas12a CRISPR systems. From the mismatch analysis results we uncover that CRISPR class 2 systems maintain their kinetic proofreading mechanism by utilizing intermittent energetic barriers and we show how to leverage the landscape to improve target specificity.

Abstract: arXiv descriptive metadata (CC0). Thank you to arXiv for use of its open access interoperability. This service was not reviewed or approved by, nor does it necessarily express or reflect the policies or opinions of, arXiv.

Record

Archive ID
arxiv:2004.09237
Published
Apr 20, 2020
DOI
10.1088/1478-3975/ac091e
arXiv
2004.09237
Categories
q-bio.BM, physics.bio-ph
Provenance
arxiv

Topics