Engineering the cis and trans DNA cleavage of CRISPR-Cas12a orthologues

Abstract

CRISPR-Cas (Clusters of Regularly Interspaced Short Palindromic Repeats - CRISPR associated) systems are host defences of prokaryotes that have been harnessed for biotechnology applications, due to their programmable nuclease activities. CRISPR-Cas12a is the signature effector protein of type V-A CRISPR-Cas systems and is widely used for both gene editing and molecular detection. Cas12a orthologues have high structural similarity and low sequence similarity, and a conserved mechanism of CRISPR-RNA (crRNA)-guided double-stranded DNA cleavage with a single RuvC nuclease domain. Cas12a has a bi-lobed structure, where recognition (REC) and nuclease (NUC) lobes respectively recognise and cleave target DNA. Cas12a has highly specific target DNA cleavage (in cis), which is harnessed for genome engineering. After target DNA cleavage, Cas12a unleashes non-specific degradation of single stranded nucleic acids (in trans), which has been repurposed for molecular detection technologies. Three natural Cas12a orthologues (As - Acidaminococcus sp. BV3L6, Lb - Lachnospiraceae bacterium ND2006, Fn - Francisella tularensis subsp. novicida U112) have been well characterised, and widely used for genome editing and molecular detection. AsCas12a and LbCas12a are efficient genome editors of mammalian cells, LbCas12a has the most robust trans cleavage activity for molecular detection, while FnCas12a has both low trans cleavage and editing efficiency. This raises the question: what protein structure and sequence features drive these differences? Stable RNA and DNA interactions are crucial to the crRNA-programmable DNA cleavage activity of Cas12a, and unknown interactions drive non-specific trans cleavage. In this thesis, I used structural alignments of three Cas12a orthologues (AsCas12a, LbCas12a, FnCas12a) to inform mutagenesis of divergent structural motifs and key residues, to determine what drives their different cis and trans cleavage activities. Thus, I characterised in vitro and in vivo the role of crRNA binding moieties, of REC lobe-crRNA:DNA heteroduplex stacking interactions, and of protein-DNA electrostatic interactions near the RuvC nuclease domain. Although crRNA binding is required for gene editing activity, engineering stronger crRNA binding was not sufficient to rescue the low editing activity of FnCas12a. Furthermore, although relaxation of REC-heteroduplex stacking interactions quickened the rate-limiting step of target DNA cleavage, this caused weaker crRNA binding and decreased activity in vivo for the three Cas12a orthologues. A hitherto uncharacterised 'NUC loop' was identified as a divergent structural element between Cas12a orthologues. This anti-parallel beta sheet structure extends from the NUC lobe towards the REC, potentially interacting with the crRNA:DNA target heteroduplex. Disruption of this loop was deleterious to the function of FnCas12a and LbCas12a, while modestly enhancing the gene editing of AsCas12a. Finally, key residues in the NUC were found drive trans cleavage through electrostatic interactions, which were neutralised by increasing ionic strength. I exploited this observation to rationally design Cas12a orthologues and optimal buffer conditions with enhanced trans cleavage activity, decreasing DNA detection times by 2 to 3-fold. In addition, these engineered Cas12a orthologues had enhanced editing efficiencies in human cell lines. Overall, this inter-orthologue survey highlights the structural and sequence features that drive the different properties of Cas12a orthologues. This work details how DNA cleavage mechanisms subtly differ between orthologues and provides a blueprint for rational engineering of cis and trans cleavage.