Engineering of a high-fidelity Cas12a nuclease variant capable of allele-specific editing

Abstract

CRISPR-Cas12a, often regarded as a precise genome editor, still requires improvements in specificity. In this study, we used a GFP-activation assay to screen 14 new Cas12a nucleases for mammalian genome editing, successfully identifying 9 active ones. Notably, these Cas12a nucleases prefer pyrimidine-rich PAMs. Among these nucleases, we extensively characterized Mb4Cas12a obtained from Moraxella bovis CCUG 2133, which recognizes a YYN PAM (Y = C or T). Our biochemical analysis demonstrates that Mb4Cas12a can cleave double-strand DNA across a wide temperature range. To improve specificity, we constructed a SWISS-MODEL of Mb4Cas12a based on the FnCas12a crystal structure and identified 8 amino acids potentially forming hydrogen bonds at the target DNA-crRNA interface. By replacing these amino acids with alanine to disrupt the hydrogen bond, we tested the influence of each mutation on Mb4Cas12a specificity. Interestingly, the F370A mutation improved specificity with minimal influence on activity. Further study showed that Mb4Cas12a-F370A is capable of discriminating single-nucleotide polymorphisms. These new Cas12a orthologs and high-fidelity variants hold substantial promise for therapeutic applications.

Fig 1. Analysis of Cas12a ortholog activities.

(A) Schematic of the GFP-activation assay. A 5-bp random sequence followed by a 24-bp protospacer (target sequence) is inserted between ATG and the GFP-coding sequence. The library is carried by a lentivirus vector. (B) crRNA/Cas12a-expressing plasmids are transfected into reporter cells. The GFP-positive cells are sorted out, and target sequences are PCR-amplified for deep sequencing. (C) Cas12a orthologs induce GFP expression. The percentage of GFP-positive cells was shown. The cells without transfection of Cas12a were used as a negative control. BF, bright field; GFP, green fluorescent protein.

Fig 2. Analysis of Cas12a PAMs.

(A) Deep sequencing reveals that Mb4Cas12a induced indels on the target site. PAM is shown in red; the GFP sequence is shown in green. (B) PAM WebLogos of Cas12a is generated based on deep sequencing data. (C) PAM wheels of Cas12a are generated based on deep sequencing data. PAM positions are shown on the wheel pictures.

Fig 3. Characterization of Mb4Cas12a with the in vitro cleavage assay.

(A) DNA cleavage of Mb4Cas12a at 37°C for 8 hours. The PAMs are shown on the top; the cleaved bands are indicated by green asterisks; the cleavage efficiency is shown below. (B) DNA cleavage of Mb4Cas12a at different temperatures for 1 hour. Temperatures are shown on the top; the cleavage efficiency is shown below. Ctr, control, the DNA fragment without incubation with Mb4Cas12a and crRNA. (C) Sanger sequencing traces from Mb4Cas12a-cleaved target show staggered 5-nt 5’ overhang. NTS, nontarget site; TS, target site. The cleaved target sites are indicated by red triangles.

Fig 4. Genome editing with Mb4Cas12a at endogenous sites.

(A) Schematic of the Cas12a and crRNA expression constructs. (B) The protein expression level of Cas12a was measured by western blot. Cells without Cas12a transfection were used as a negative control. (C) Comparison of Mb4Cas12a and LbCas12a activity at 13 endogenous loci (n = 3). The data represent the mean ± SD. Student t test, ns, means no significance, * p < 0.05. (D) Quantification of editing efficiency for Mb4Cas12a and LbCas12a. The data represent the mean ± SD; n = 3. Student t test, * p < 0.05. The numerical values underlying this figure can be found in S5 Table.

Fig 5. Test of Mb4Cas12a specificity by the GFP activation assay.

(A) The GFP-activation assay shows that Mb4Cas12a and LbCas12a substantially tolerate single-nucleotide mismatches. A schematic of the GFP-activation reporter is shown on the top; crRNAs are shown on the left; single-nucleotide mismatches are indicated in red. On, on-target crRNA. (B) Quantification of off-target efficiency. The off-target efficiency is normalized by on-target efficiency. (C, D) Test of 4 Mb4Cas12a variant activity at 2 endogenous loci EMX1 (C) and VEGFA (D). The data represent the mean ± SD; n = 3. Student t test, * p < 0.05. ** p < 0.01. *** p < 0.001. The numerical values underlying this figure can be found in S5 Table.

Fig 6. Test of Cas12a specificity with GUIDE-seq.

(A, B) The genome-wide off-target effects of LbCas12a, Mb4Cas12a, and F370A are analyzed by GUIDE-seq. On-target and off-target sequences are shown on the left. Read numbers for on-target and off-target sites are shown on the right. The reads number for off-target sites less than 50 generated by LbCas12a is not displayed. Mismatches compared to the on-target site are shown and highlighted in colour.

Fig 7. Mb4Cas12a-F370A allows allele-specific genome editing.

(A, B) Mb4Cas12a-F370A allele specifically disrupts the single-nucleotide polymorphism (SNP) at locus rs28385702 (A), rs4927193 (B), rs747000387 (C), rs45545732 (D). PAMs are shown in light blue; target SNPs are shown in red; crRNAs are shown in purple; indel efficiencies are shown on the right. Data represent mean ± SD. n = 3. (E) Summary of pathogenic sites or SNPs that are potentially disrupted by Mb4Cas12a-F370A. The data represent the mean ± SD; n = 3. Student t test, **** p <0.0001. The numerical values underlying this figure can be found in S5 Table.