Harnessing accurate non-homologous end joining for efficient precise deletion in CRISPR/Cas9-mediated genome editing

Abstract

Background: Many applications of CRISPR/Cas9-mediated genome editing require Cas9-induced non-homologous end joining (NHEJ), which was thought to be error prone. However, with directly ligatable ends, Cas9-induced DNA double strand breaks may be repaired preferentially by accurate NHEJ.

Results: In the repair of two adjacent double strand breaks induced by paired Cas9-gRNAs at 71 genome sites, accurate NHEJ accounts for about 50% of NHEJ events. This paired Cas9-gRNA approach underestimates the level of accurate NHEJ due to frequent + 1 templated insertions, which can be avoided by the predefined Watson/Crick orientation of protospacer adjacent motifs (PAMs). The paired Cas9-gRNA strategy also provides a flexible, reporter-less approach for analyzing both accurate and mutagenic NHEJ in cells and in vivo, and it has been validated in cells deficient for XRCC4 and in mouse liver. Due to high frequencies of precise deletions of defined "3n"-, "3n + 1"-, or "3n + 2"-bp length, accurate NHEJ is used to improve the efficiency and homogeneity of gene knockouts and targeted in-frame deletions. Compared to "3n + 1"-bp, "3n + 2"-bp can overcome + 1 templated insertions to increase the frequency of out-of-frame mutations. By applying paired Cas9-gRNAs to edit MDC1 and key 53BP1 domains, we are able to generate predicted, precise deletions for functional analysis. Lastly, a Plk3 inhibitor promotes NHEJ with bias towards accurate NHEJ, providing a chemical approach to improve genome editing requiring precise deletions.

Conclusions: NHEJ is inherently accurate in repair of Cas9-induced DNA double strand breaks and can be harnessed to improve CRISPR/Cas9 genome editing requiring precise deletion of a defined length.

Keywords: Accurate NHEJ; CRISPR/Cas9; Genome editing; Knockout; Paired gRNAs; Precise deletion; Targeted in-frame deletion; Templated insertions.

Fig. 1

Repair of Cas9-induced DSBs by NHEJ is inherently accurate. a The sGEJ reporter. NHEJ repair of paired DSBs induced by I-SceI or Cas9-gRNAs are indicated. Repair products are divided into four groups based on DSBs induced at either or both target sites by I-SceI or Cas9-gRNA in the reporter. Groups I, II, III, and IV represent, respectively, NHEJ for DNA cleavage simultaneously at both target sites, only at the first target site, only at the second target site, and individually at two target sites as indicated. b–d Analysis of I-SceI- or paired gRNA-guided Cas9-induced NHEJ in the sGEJ reporter. The normalized editing efficiency (b) was calculated as ratios of edited events to total reads and normalized by transfection efficiency. The frequency of group I, group II, group III, and group IV (c) was calculated as ratios of reads from each group to total edited reads. The frequency of accurate NHEJ, deletion, insertion, and InDel (d) was calculated as ratios of reads from each category in group I to total group I reads. Bars represent the mean ± standard deviation (SD) of three independent experiments. For normalized editing efficiency (b), Student’s paired t-test between I-SceI and gsGEJ2 + 2a *P = 0.025. e Analysis of NHEJ induced by paired Cas9-gRNAs at 70 endogenous genome sites from mouse and human cells. The editing efficiency without transfection efficiency, the frequency of group I in edited events, and the frequency of accurate NHEJ in group I events from each site were analyzed. Bars represent the mean ± SD. f–h Analysis of NHEJ induced by paired Cas9-gRNA at mPIF1, mLDHA, and two hp53 sites. The normalized editing efficiency (f), the frequency of each group in edited events (g), and the frequency of each category in group I events (h) were calculated. Bars represent the mean ± SD of three independent experiments

Fig. 2

Accurate NHEJ is influenced by frequent + 1 templated insertions (TI), not directly by the orientations of paired Cas9 or the distance between target sites of paired Cas9. a Correlation between the frequency of accurate NHEJ and the frequency of + 1 TI (blue dot) or + 2 TI (red rhombus). TI template insertions with no additional mutations at repair junctions. b Comparisons between accurate NHEJ and accurate NHEJ combined with + 1 and + 2 template insertions (i.e., Accurate+TI). The difference indicates the underestimation of accurate NHEJ due to + 1 and + 2 TI. Student’s paired t-test P < 0.0001. c Correlation between the frequency of group I and the frequency of accurate NHEJ (blue dot) or the frequency of accurate NHEJ+TI (red rhombus). d Four different Cas9-gRNA orientations guided by paired PAMs. W/W, W/C, C/C, and C/W orientations were defined by the position of paired PAMs on either the Watson strand (W) or the Crick strand (C). e The frequency of Accurate or Accurate+TI from four orientations of 71 endogenous genome sites were calculated and summarized. Bars represent the mean ± standard deviation (SD). In the Accurate group, one-way ANOVA P < 0.0001; post LSD pairwise comparisons: **P < 0.01 for W/W vs W/C; ***P < 0.0005 for W/C vs C/W; *P < 0.05 for W/C vs C/C; not significant (NS) for the others. In the Accurate+TI group, one-way ANOVA NS. f The frequency of + 1 TI from four orientations of 71 endogenous genome sites were calculated and summarized. Bars represent the mean ± SD. One-way ANOVA P < 0.0001; post LSD pairwise comparison test ***P < 0.0005 for W/W vs W/C, ***P < 0.0005 for W/C vs C/W, ***P < 0.0005 for W/C vs C/C; not significant (NS) for the others. g, h Correlation between the frequency of accurate NHEJ (blue circle) or the frequency of Accurate+TI (red rhombus) and the distance between two cleavage sites at 71 genome sites (g) or 29 genome sites with the same W/W orientation (h)

Fig. 3

Validation of a reporter-less NHEJ assay in cells deficient for XRCC4. a Deletion of the intervening sequence between two target sites by paired Cas9-gRNA. Cells were transfected with expression plasmids for single or paired sgRNAs and Cas9 and genomic DNA was purified 72 h post-transfection and amplified by primers flanking the cutting sites. The PCR amplicons were subjected to agarose gel electrophoresis. The distance between two cleavage sites was 57 bp at the LDHA site and 33 bp at the ROSA26 site as indicated and the deletion of the intervening sequence indicates simultaneous Cas9 cleavage. b–d Analysis of NHEJ induced by paired Cas9-gRNA at the LDHA and ROSA26 sites in isogenic XRCC4^+/+ and XRCC4^−/− mouse ES cells. The normalized editing efficiency (b), the frequency of each group in edited events (c), and the frequency of each category in group I events (d) were calculated. The normalized editing efficiency represents the efficiency of overall NHEJ, including accurate and mutagenic NHEJ. Bars represent the mean ± standard deviation (SD) of three independent experiments. For normalized editing efficiency (b), Student’s paired t-test between XRCC4^+/+ and XRCC4^−/− P = 0.0028 at the LDHA site. e Deletion length distributions of Group I ‘Del’ events in isogenic XRCC4^+/+ and XRCC4^−/− mouse ES cells. The reads were combined by three independent experiments. At the LDHA site and the ROSA26 site, each dot represents 100 reads and 20 reads, respectively. The median deletion length is indicated, and deletion distributions demonstrate a shift towards longer deletions in cells lacking XRCC4 (Mann–Whitney test, ****P < 0.0001). f Frequency of accurate NHEJ among group I (left) and frequency of deletions with different deletion length in “Del” events of group I NHEJ (right) in either isogenic XRCC4^+/+ or XRCC4^−/− mouse ES cells. Del NHEJ events were grouped into 58–63 bp and > 63 bp at the LDHA site and 34–39 bp and > 39 bp at the ROSA26 site. The respective reads and frequencies were summarized in the inset and compared by a χ² test with P values indicated. g Frequency of microhomology usage at different lengths in deletion-only group I events of either isogenic XRCC4^+/+ or XRCC4^−/− mouse ES cells

Fig. 4

Validation of a reporterless NHEJ assay in mouse liver. a Delivery of paired Cas9-gRNAs into mouse liver by hydrodynamic tail vein injection. b–d Analysis of NHEJ induced by paired Cas9-gRNA at the LDHA sites in mouse liver. The editing efficiency (b), the frequency of each group in edited events (c), and the frequency of each category in group I events (d) were calculated (Mann–Whitney test, not significant (NS)). e Deletion length distributions of group I “Del” events in mouse livers. The median deletion length is indicated, and deletion length distributions demonstrate no shift towards longer deletions between these liver specimens (Kruskal–Wallis test, NS, P = 0.1922). f Frequency of accurate NHEJ among group I (left) and frequency of deletions with different deletion lengths in Del events of group I NHEJ (right) in mouse livers. Del NHEJ events were grouped into 58–63 bp and > 63 bp at the LDHA site. The respective reads and frequencies are summarized in the inset. g Frequency of the microhomology usage at different lengths in deletion-only group I events of mouse liver

Fig. 5

Precise out-of-frame deletion mediated by accurate NHEJ promotes gene knockout. a Classification of the Common, Ideal, and Paired approaches. With paired Cas9-gRNAs, Common, Ideal, and Paired represent genome editing that depends, respectively, upon NHEJ without group I events, NHEJ for ideally all simultaneous Cas9 cleavage, and NHEJ including group I events. b Correlation between accurate NHEJ and out-of-frame mutations derived from the Common approach (black squares), the Ideal approach (blue squares), or the Paired approach (red circle). c The frequency of out-of-frame mutations between different groups as indicated. Bars represent the mean ± standard deviation (SD) of the out-of-frame frequencies. Data were analyzed by one-way ANNOVA followed by post-hoc LSD pairwise comparisons (**P < 0.005, ***P ≤ 0.0005, NS not significant). d Testing of paired Cas9-gRNA at exon 2 of mMDC1. Mouse ES cells were transfected with expression plasmids for single or paired gRNAs and Cas9 and genomic DNA were purified 72 h post-transfection and amplified by primers flanking the cutting sites. The PCR amplicons were subjected to agarose gel electrophoresis. The distance between paired Cas9-gRNA target sites is 52 bp as indicated. e The frequency of out-of-frame (i.e., knockouts) and in-frame MDC1 editing events induced by single or paired Cas9-gRNA. For knockout efficiency, Student’s paired t-test between g2 and g2 + 3 *P = 0.029; between g3 and g2 + 3 *P = 0.013. f The frequency of group I, group II, group III, and group IV events among those edited by Cas9 guided by paired gRNAs as indicated. g The frequency of accurate, deletion, insertion, and indel events among group I events induced by paired Cas9-gRNAs. h The HR reporter. Repair of an I-SceI- or Cas9-induced DSB by HR between sister chromatids generates wild-type GFP. i, j Percentage of I-SceI- (i) or Cas9- (j) induced GFP⁺ cells from HR reporter mouse ES cells transiently transfected with expression plasmids for single or paired gRNA-guided Cas9. Values are the mean ± SD of three independent experiments, each in triplicate. In I-SceI-induced HR (i), Student’s paired t-test between g2 and g2 + 3 **P = 0.0048; between g3 and g2 + 3 *P = 0.0154. In Cas9-induced HR assays (j), Student’s paired t-test between g2 and g2 + 3 ***P = 0.0004; between g3 and g2 + 3 *P = 0.0449. k Generation of MDC1 mutant clones by paired Cas9-gRNAs. Clones were picked, grown, and identified by Sanger sequencing. The frequencies of specific MDC1 mutations are indicated in parentheses. l MDC1 knockout was confirmed by western blotting using β-actin as a loading control. m Percentage of I-SceI-induced GFP⁺ cells from one MDC1^+/+ and three isogenic MDC1^−/− HR reporter clones (clones #4, #26, and #32). Bars represent the mean ± SD of three independent experiments, each in triplicate. Student’s unpaired t-test: ***P < 0.0001 between MDC1^+/+ and MDC1^−/−. WT wild type

Fig. 6

Targeted in-frame deletions mediated by accurate NHEJ assists functional domain analysis in situ. a Correlation between accurate NHEJ and in-frame deletions generated by the Common (black triangles), Ideal (blue squares), and Paired approaches (red circles). b Domain structure of mouse 53BP1 protein. Oligomerization domain (OD), Tudor domain, nuclear localization sequence (NLS), and tandem BRCT domain are shown. Two sets of paired gRNAs respectively targeting the Tudor domain (exon 21) and OD domain (exon 18) are also indicated. Precise in-frame deletions were confirmed by Sanger sequencing of the PCR product for repair junction. c Detection of 53BP1 wild-type and in-frame mutants by western blotting using β-actin as loading control. WT wild type. d IR-induced foci formation of γH2AX (red) and 53BP1 (green) in 53BP1 WT clone and in-frame mutant clones. e–g Analysis of NHEJ induced by paired Cas9-gRNAs at the LDHA site in 53BP1 WT clone and in-frame mutant clones. The normalized editing efficiency (e), the frequency of each group in edited events (f), and the frequency of each category in group I events (g) were calculated. Bars represent the mean ± standard deviation (SD) of three independent experiments. h Deletion length distributions of group I “Del” events at the LDHA site in 53BP1 WT mouse ES clone and in-frame mutant clones. The reads were combined from three independent experiments. Each dot represents 20 reads. The median deletion length is indicated, and deletion length distributions demonstrate a shift towards longer deletions in 53BP1 in-frame mutant clones (Mann–Whitney test between WT and ΔTudor or ΔOD ****P < 0.0001). i Frequency of accurate NHEJ among group I (left) and frequency of deletions with different deletion length in Del events of group I NHEJ (right) in 53BP1 WT mouse ES clone and in-frame mutant clones. Del NHEJ events were grouped into 58–63 bp and > 63 bp. The respective reads and frequencies are summarized in the inset and compared by a χ² test with P values as indicated

Fig. 7

The Plk3 inhibitor GW843682X enhances accurate NHEJ in the repair of Cas9-induced DSBs. a, b Percentage of I-SceI- (a) or paired-Cas9- (b) induced GFP⁺ cells from sGEJ reporter cells transfected with siRNA against CtIP or Luciferase (Luc). Bars represent the mean ± standard deviation (SD) of three independent experiments, each performed in triplicate. In I-SceI-induced HR assays (a), Student’s paired t-test between siLuc and siCtIP **P = 0.0034; In Cas9-induced HR assays (b), Student’s paired t-test between siLuc and siCtIP **P = 0.0006. Depletion of CtIP was determined by western blotting using β-actin as a loading control. c Percentage of paired Cas9-induced GFP⁺ cells from sGEJ reporter cells treated with the Plk3 inhibitor GW843682X at concentrations of 1 μM, 3 μM, and 5 μM. Bars represent the mean ± SD of three independent experiments, each performed in triplicate. Student’s paired t-test between DMSO and 1 μM *P = 0.016; between DMSO and 3 μM **P = 0.003; between DMSO and 5 μM *P = 0.005. d–f Analysis of NHEJ induced by paired Cas9-gRNAs at the LDHA site in mouse ES cells treated with the Plk3 inhibitor at different concentrations. The normalized editing efficiency (e), the frequency of each group in edited events (f), and the frequency of each category in group I events (g) were calculated. Bars represent the mean ± SD of three independent experiments. For the normalized editing efficiency (d), Student’s paired t-test between DMSO and 1 μM **P = 0.006; between DMSO and 5 μM *P = 0.04. g Deletion length distributions of group I Del events at the LDHA site in mouse ES cells treated with the Plk3 inhibitor. The reads were combined from three independent experiments. Each dot represents 200 reads. The median deletion length is indicated, and deletion length distributions demonstrate a shift towards longer deletions in cells treated with the Plk3 inhibitor (Mann–Whitney test between DMSO and Plk3i ***P < 0.0001). h Frequency of accurate NHEJ among group I (left) and frequency of deletions with different deletion length in Del events of group I NHEJ (right) in mouse ES cells treated with Plk3i at different concentrations as indicated. Del NHEJ events were grouped into 58–63 bp and > 63 bp. The respective reads and frequencies are summarized in the inset and compared by a χ² test with P values as indicated