Genome editing in animals with minimal PAM CRISPR-Cas9 enzymes

Abstract

The requirement for Cas nucleases to recognize a specific PAM is a major restriction for genome editing. SpCas9 variants SpG and SpRY, recognizing NGN and NRN PAMs, respectively, have contributed to increase the number of editable genomic sites in cell cultures and plants. However, their use has not been demonstrated in animals. Here we study the nuclease activity of SpG and SpRY by targeting 40 sites in zebrafish and C. elegans. Delivered as mRNA-gRNA or ribonucleoprotein (RNP) complexes, SpG and SpRY were able to induce mutations in vivo, albeit at a lower rate than SpCas9 in equivalent formulations. This lower activity was overcome by optimizing mRNA-gRNA or RNP concentration, leading to mutagenesis at regions inaccessible to SpCas9. We also found that the CRISPRscan algorithm could help to predict SpG and SpRY targets with high activity in vivo. Finally, we applied SpG and SpRY to generate knock-ins by homology-directed repair. Altogether, our results expand the CRISPR-Cas targeting genomic landscape in animals.

Fig. 1. SpG and SpRY are active nucleases in zebrafish and C. elegans.

a Two gRNAs (a, b) targeting slc45a2 exon 1 in zebrafish (top). Experimental setup to analyze CRISPR-Cas9, SpG, and SpRY-mediated mutations in zebrafish by injecting one-cell-stage embryos (bottom). b Phenotypes obtained after the injection of the mRNA–gRNA duplex targeting slc45a2 (albino) showing different levels of mosaicism (albino-like (alb-like), severe, mild) compared to the WT. Lateral views (scale bar, 1 mm) and insets of the eyes (scale bar, 0.2 mm) of 48 h post-fertilization (hpf) embryos are shown. c Percentage of albino- phenotypes (panel b) in embryos 48 hpf. (n) total number of injected embryos. The results were obtained from at least two independent experiments. d Phenotypic evaluation at different concentrations of gRNAs and mRNAs. Stacked bar plots show the percentage of the phenotypes described in panel b. The results were obtained from at least two independent experiments. e gRNAs were complexed with purified proteins to form RNPs for in vitro and in vivo testing in C. elegans by microinjection. f Sequences of gRNAs targeting dpy-10 with distinct complementarity. RNP combinations comprised of each gRNA and WT SpCas9, SpG, or SpRY were tested in vitro. Top bands show uncleaved PCR product. Lower bands show cleaved products. g The dpy-10 matched and +5 gRNAs were tested for in vivo activity in C. elegans by injecting a single gonad arm. Each dot represents the editing efficiency in each P₀ that produced at least 100 F₁s. The results were obtained from two independent experiments, with both conditions carried out in parallel injections (Student’s t test p value). h Schematic representation of in vivo experiments in C. elegans using a gtbp-1::wrmScarlet reporter by screening for loss of fluorescence. dpy-10 gRNA was used for co-CRISPR. i An anti-wrmScarlet gRNA with NGG PAM was complexed with 1.3 µM of SpCas9, SpG, or SpRY to compare their in vivo efficiencies. In a separate experiment, the SpG-anti-wrmScarlet (NGG) RNP was injected at 8.0 µM. Each dot represents the editing efficiency in each P₀ that produced at least five Dpy or Rol F₁s. (One-way ANOVA followed by Tukey’s test for multiple comparisons p values).

Fig. 2. SpG and SpRY are active nucleases at minimal PAM targets in zebrafish.

a Diagram illustrating 15 gRNAs (blue) with NGH PAM in three zebrafish genes. Five gRNAs targeting exon 1, exons 3, 4 and 5 and exons 2 and 3 of slc45a2, slc24a5 and tbxta, respectively. Different NGH PAMs for each target are detailed in Supplementary Data 1. b mRNA SpG shows high activity in NGH sites. Individual gRNAs described in panel a were injected with the SpG mRNA (240 pg gRNA and 300 pg mRNA per embryo). Stacked barplots show the percentage of albino-like/golden-like (gol-like)/class III, severe/class II, mild/class I, and phenotypically wild-type (WT) embryos 48 hpf after injection. (n) total number of injected embryos. The results were obtained from at least two independent experiments. c Phenotypes obtained after the injection of the mRNA–gRNA duplex targeting slc24a5 showing different levels of mosaicism (golden-like, severe, mild) compared to the WT. Lateral views (scale bar, 1 mm) and insets of the eyes (scale bar, 0.2 mm) of 48 hpf embryos are shown. d Phenotypes obtained after injection of the mRNA–gRNA duplex targeting tbxta in zebrafish embryos (Lateral views). Levels of mosaicism compared to wild type (WT) were evaluated at 28 hpf. Class I: Short tail (least extreme). Class II: Absence of notochord and short tail (medium level). Class III: Absence of notochord and extremely short tail (most extreme). Scale bar, 0.5 mm. e Diagram illustrating 8 gRNAs (green) with NAN PAM in three zebrafish genes. Three gRNAs targeting slc45a2 exons 1 and 5, two gRNAs targeting slc24a5 exons 1 and 3 and three gRNAs targeting tbxta exons 1 and 2. Different NAN PAMs for each target are detailed in Supplementary Data 1. f mRNA SpRY shows high activity in NAN sites. Individual gRNAs described in panel e were injected with the SpRY mRNA (240 pg gRNA and 300 pg mRNA per embryo). Stacked barplots show the percentage of albino-like/golden-like/class III, severe/class II, mild/class I, and phenotypically wild-type (WT) embryos 48 hpf after injection. (n) total number of injected embryos. The results were obtained from at least two independent experiments.

Fig. 3. SpG and SpRY are active nucleases at minimal PAM targets C. elegans.

a Diagram illustrating one crRNA with NGH PAM (blue) and three crRNAs with NAN PAM (green) targeting wrmScarlet. The sequences of the crRNAs and PAMs are shown below. b Three distinct concentrations of SpG RNP, with the anti-wrmScarlet NGH gRNA 1, were tested in a strain expressing gtbp-1::wrmScarlet. Editing efficiency is defined as the number of F₁ worms exhibiting loss of fluorescence in the F₂ divided by the total number of separated dpy-10 co-edited F₁s. Each dot represents the editing efficiency in each individual P₀ that produced at least ten Dpy or Rol F₁s. All three conditions were carried out in parallel injections (One-way ANOVA followed by Tukey’s test for multiple comparisons p values). c Comparison of editing efficiencies with anti-wrmScarlet NGH gRNA 1 between SpCas9 and SpG, and SpG and SpRY in independent experiments at an RNP concentration of 8.0 µM. Editing efficiency, dots, and numbers are defined as in panel b. Conditions belonging to parallel injections are separated by a dashed line (Student’s t test p value). d In vitro analysis of three anti-wrmScarlet NAN gRNAs. Different RNP combinations were tested in vitro at 37 °C by incubating the RNPs with wrmScarlet PCR product. Top row of bands shows uncleaved PCR products and the specific cleavage products for each gRNA are specified in the figure. The gRNA appears as a faint band at approximately 100 bp. This experiment was performed once. e In vivo analysis of the three anti-wrmScarlet NAN gRNAs. A titration of three distinct SpRY RNP concentrations was performed for gRNAs 1 and 3, while gRNA 2 was tested at 8.0 µM only. The editing efficiency of SpCas9 in NAN sites was evaluated using gRNAs 2 and 3. Editing efficiency, dots, and numbers are defined in panel b. A dashed line separates conditions belonging to parallel injections (One-way ANOVA followed by Tukey’s test for multiple comparisons [NAN 1 and 3], Student’s t test [NAN 2] p values). The percentage of alleles edited in panels b, c, and e was calculated as previously described in Fig. 1i.

Fig. 4. Further SpG and SpRY optimization in vivo.

a Prediction of gRNA activity in vivo using CRISPRscan. Stacked barplots show the percentage of highly efficient gRNAs (blue) and not highly efficient gRNAs (orange) in two groups separated based on CRISPRscan scores (>66 and ≤66). Highly efficient gRNAs generate more than 50% of embryos with albino-like/golden-like/class III or severe/class II phenotypes and less than 10% phenotypically wild-type. Fisher test p value. b RNP can enhance SpG and SpRY activity in zebrafish. Stacked barplots show the percentage of albino-like/golden-like/class III, severe/class II, mild/class I, and phenotypically wild-type (WT) embryos 48 h post-fertilization (hpf) after injection. (n) total number of injected embryos. The results were obtained from at least two independent experiments. mSpG and mSpRY injections data from Fig. 2b, f, and RNP injections data from Supplementary Fig. 10. The χ2-test p value per comparison is shown. c Utility of SpG and SpRY for the insertion of DNA sequences via HDR. We used ssDNA (ssODN) as a donor template to produce a missense mutation in swsn-4, to introduce a gene fragment for nested CRISPR at usp-48, trx-1, W05H9.1, and to add a degron tag to cki-1. Information about targeted sequences, PAMs, DSB to edit distance, insert lengths, nuclease concentration used, and correct vs incorrect insertions obtained, is showed in Supplementary Data 2. d EG9615 and CER660 worms, which express SpCas9 and SpG endogenously in the germline (SpCas9e and SpGe), respectively, were injected with tracrRNA and a crRNA targeting dpy-10 with an NGH PAM. The fluorescent markers myo-2p::mCherry and myo-3p::mCherry were used as co-injection markers. F₁ worms expressing mCherry in the pharynx or body wall muscle were singled out and the appearance of Dpy progeny was screened in the F₂. Editing efficiency is defined as the number of F₁ worms that segregate Dpy progeny in the F₂ divided by the total number of separated F₁s. Each dot represents the editing efficiency in each individual P₀ that produced at least ten mCherry-expressing F₁s. All conditions were carried out in parallel injections (Student’s t test p value).