To Cleave or not to Cleave: Predicting the Target Specificity of CRISPR-Cas Systems through Theoretical Modeling

Abstract

Many prokaryotes employ the CRISPR-Cas (Clustered Regularly Interspaced Short Palindromic Repeat – CRISPR associated) system to fend off attacks by hostile genetic elements. This adaptive immune system recognizes invaders based on their level of sequence complementarity with RNA transcribed from a library of past invasions stored at the CRISPR locus. To avoid infection, the CRISPR interference complex must be able to single out a short target sequence (20 – 40 nt) among a total of 105-106 nt in the cell, while at the same time recognizing targets that have evolved away from the record stored at the CRISPR locus. Despite the tremendous interest the CRISPR-Cas9 system has gained as a novel genome editing tool, the sequence preference of the interference complex remains poorly understood. Experiments have shown that it is not just the amount of mutations, but also their placement along the guide/target, that determine the level of interference.Through kinetic modeling we provide simple rules to assess sequence specificity based on mismatch patterns, and quantitatively explain the experimentally observed seed region over which no mismatches are permitted. Moving beyond sequence complementarity, we also quantify how changes in the conformation of the interference complex can serve to improve specificity during target recognition. Finally, by fitting our model to published experimental data, we give estimates for the base-pairing energies within the interference complex for various CRISPR systems.