Number of target sites calculation for different Cas nucleases in zebrafish and C. elegans genomes
Genome sequences from C. elegans (v. WBcel235) and D. rerio (v. GRCz11) were downloaded from Ensembl database 43.
PAM sites (Supplementary Table 1) were searched by an in-house script written in PERL language. It is based on the fuzznuc function of EMBOSS package to show the coordinates of the sites in both strands 44. Genomic positions were converted into GTF format for later comparison to the different types of regions of the genome using bedtools 45. The count of sites by genomic position was depicted by the ggplot2 3.3.0 library from R programming language.
Target and gRNA design in zebrafish
Target (protospacers) and gRNAs for SpG and SpRY were designed using an updated version of the algorithm CRISPRscan (www.crisprscan.org) tool 13. Regular expressions for SpG (NGN) and SpRY (NAGN) were added for on-target and off-target searches. On-target scores were evaluated using the same CRISPRscan scoring algorithm used for SpCas9. Off-targets with mismatches restricted, or not, to the target “seed” are reported similarly, to SpCas9 46,47. The CFD score 22 used to characterize potential off-targeting was kept identical to score protospacers but adapted to score PAMs of potential off-targets. SpG and SpRY matching PAMs were set a 1.0 score instead of using the original (lower than 1.0) Doench et al. score characterized for SpCas9. Protospacers in the different loci were selected without predicted off-targets 46, and within functional domains of the protein and in exons in the first half of the ORF with the exception of gRNA albino 7 which approximately maps in the last third part of the ORF. To evaluate activity of the Cas9 variants in the NGG PAMs, we used two gRNAs (albino a and albino b) as previously described 13. All the information about the targets is detailed in Supplementary Table 1.
In vitro transcription and gRNA generation
gRNAs were generated as previously described 48. gRNA DNA templates were amplified by fill-in PCR. Briefly, a 52-nt oligo (sgRNA primer), containing the T7 promoter, the 20 nt of the specific gRNA DNA binding sequence (spacer) starting with two Gs and a constant 15-nt tail for annealing, was used in combination with an 80-nt reverse oligo to add the gRNA invariable 3′ end (universal primer). Here, all spacers used in zebrafish experiments started by 5’ GG (Supplementary Fig. 1a) allowing 100% match between target and in vitro transcribed gRNAs. A 117 bp PCR product was generated following these parameters: 30 seconds at 98°C, 30 cycles of 10 seconds at 98°C, 30 seconds at 51°C, and 30 seconds at 72°C, and a final step at 72°C for one minute. PCR products were purified using FavorPrep™ GEL/PCR Purification kit (Favorgen) columns and approximately 120–150 ng of DNA was used as template for a T7 in vitro transcription (IVT) reaction (AmpliScribe-T7-Flash transcription kit from Epicentre). In vitro transcribed gRNAs were DNAse-treated using TURBO-DNAse for 20 min at 37 °C and precipitated with sodium acetate/ethanol and resuspended in RNAse and DNAse free water. gRNAs were visualized in 2% agarose stained with ethidium bromide to check for RNA integrity, quantified using the Qubit RNA BR Assay Kit (ThermoFisher, Q10210), and stored in aliquots at −80°C.
Target and gRNA design in C. elegans
For experiments in the dpy-10 locus, the dpy-10 gRNA with NGG PAM used in co-CRISPR was used as the reference sequence 17. Mismatches were then introduced into the protospacer at one (+1) or five (+5) nucleotides upstream of the PAM (Fig. 1f). Meanwhile, a protospacer with an NGH PAM was chosen based on proximity to the reference NGG sequence to maintain the cut site within the RXXR domain of dpy-10 which is responsible for the production of the dominant dumpy and roller phenotypes, such as that of the cn64 allele 49. Five targets with distinct PAM requirements were selected for wrmScarlet: one NGG, one NGH, and three NAN (Fig. 3a). For HDR experiments in swsn-4, usp-48, trx-1, W05H9.1, and cep-1, gRNAs were selected based on the proximity of the DSB from the desired edit site (Fig. 4d). crRNAs and ssODNs were purchased in tubes from IDT as 2 nmol ALT-R crRNAs and 4 nmol ultramers, respectively, and resuspended in 20 μl or 40 μl of nuclease-free duplex buffer (30 mM HEPES, pH 7.5; 100 mM potassium acetate. IDT, Cat. No. 11-01-03-01), respectively, to yield a stock concentration of 100 μM. A list of all crRNAs used can be found in Supplementary Table 1.
SpG and SpRY zebrafish codon optimized constructs and mRNA generation
pT3TS_zCas9 (Addgene, 46757) 50 was modified to generate two plasmids encoding the SpG and SpRY zebrafish codon-optimized variants: pT3TS_zSpG and pT3TS_zSpRY include D1135L/S1136W/G1218K/E1219Q/ R1335Q/T1337R and A61R/L1111R/D1135L/S1136W/G1218K/E1219Q/ N1317R/A1322R/R1333P/R1335Q/T1337R modifications, respectively. All mutations were generated using site-directed mutagenesis (QuikChange Multi Site-Directed Mutagenesis Kit, Agilent Technologies) and primers used are detailed in Supplementary Table 1. The final pT3TS_zSpG and pT3TS_zSpRY constructs were confirmed by sequencing.
SpG and SpRY mRNA were in vitro transcribed from DNA linearized by XbaI (1 μg) using the mMESSAGE mMACHINE™ T3 kit (Invitrogen, Thermo Fisher). In vitro transcribed mRNAs were DNAse treated using 1 μl TURBO-DNAse for 20 minutes at 37 °C and purified using RNeasy Mini Kit (Qiagen). mRNA product was quantified using NanoDrop (Thermo Fisher) and stored in aliquots at −80°C.
SpG and SpRY purification
The two Cas9 variants: SpG (D1135L/S1136W/G1218K/E1219Q/R1335Q/T1337R) and SpRY (A61R/L1111R/D1135L/S1136W/G1218K/E1219Q/N1317R/A1322R/R1333P/R1335Q/T1337R) were cloned into the pET-28b-Cas9-His, with DNA gBlocks (IDT) encoding for the mutated regions.
The different Cas9 proteins were expressed in E. coli (DE3) using the auto-induction method, by growing for 4h at 37 °C, followed by 20h expression at 25 °C. Cells were harvested by centrifugation at 4000 × g for 10 min. Cell pellets were suspended in equilibration buffer (20 mM Tris pH 8.0, 500 mM NaCl) plus protease inhibitors (Roche) and lysed using a high pressure Emulsiflex. The cell debris were removed by centrifugation at 30000 × g for 30 min. The supernatant was purified using a HisTrap column (Cytiva) pre-equilibrated with buffer A (20 mM Tris pH 8.0, 500 mM NaCl). After sample loading, the columns were washed with buffer A plus 50mM imidazole and eluted with buffer A plus 500 mM imidazole. The eluted protein was concentrated with Amicon Ultra 50 Kda filters (Millipore) and loaded into a Superdex 200 10/300 size-exclusion column (Cytiva) equilibrated with SEC buffer, 20 mM HEPES pH 7.4, 500 mM KCl, and 1 mM DTT. The eluted sample was dialyzed against storage buffer 20 mM Tris pH 7.4, 200 mM KCl, 10mM MgCl2, and 10% glycerol, concentrated to 2 mg/mL, snap frozen in liquid nitrogen, and stored at −80°C.
RNP in vitro test
The gRNA was prepared by pre-annealing 3.2 μL of 32 μM ALT-R tracrRNA (IDT, Cat. No. 1072532) and 1 μL of 100 μM crRNA with 5.8 μL of nuclease-free duplex buffer (IDT) at 95°C for 5 minutes. Dilutions of the components, namely gRNA, nuclease, and PCR product were prepared at 300 nM, 900 nM, and 90 nM, respectively. Then, the RNP complex was assembled by incubating 9 μL of guide RNA with 3 μL of nuclease in 12 μL of nuclease-free H2O with 3 μL of 10x Cas9 reaction buffer (New England Biolabs, #B0386) at 37 °C for 15 minutes. Alt-R® S.p. Cas9 Nuclease V3 (IDT, Cat. No. 1081058) was used as commercial Cas9. 3 μL of the DNA substrate (PCR product) containing the target site was then added to achieve a final molar ratio of nuclease, guide RNA, and target site of 10:10:1 (90 nM:90 nM:9 nM). The 30-μl reactions were incubated at different temperatures (15, 25, 37, and 50 °C) for 60 minutes. To release the DNA substrate from the RNP complex, 1 μl Proteinase K (20 mg/mL) was added to the reaction and incubated at 56°C for 10 minutes. The cleaved products were analyzed through agarose gel electrophoresis using a 2% gel stained with SYBR® safe DNA gel stain (ThermoFisher Scientific, Cat. No. S33102).
Wild-type zebrafish embryos were obtained through natural mating of AB/Tübingen AB/Tu zebrafish of mixed ages (5–18 months). Selection of mating pairs was random from a pool of 20 males and 20 females. All experiments involving zebrafish conform to national and European Community standards for the use of animals in experimentation and were approved by the ethical committees from the University Pablo de Olavide, CSIC, and the Andalusian Government. Zebrafish wild-type strains AB/Tübingen (AB/Tu) were maintained and bred under standard conditions 51. All experiments were carried out at 28 °C, a temperature allowing optimal zebrafish development.
C. elegans maintenance
We used the Bristol N2 strain as the wild-type background while the strain CER541 gtbp-1(cer149gtbp-1::wrmScarlet) IV was used for wrmScarlet knockout experiments. Worms were maintained at 15 °C, 20 °C or 25 °C on Nematode Growth Medium (NGM) plates seeded with Escherichia coli OP50 bacteria 52. All strains generated in this study are listed in Supplementary Table 1.
mRNA and RNP injections and image acquisition in zebrafish
mRNA injection mixes were prepared at two different concentrations by combining the mRNA of the variants, mSpG and mSpRY, and the gRNAs. One nL containing 150 pg of SpG or SpRY mRNA and 20 pg of gRNA were injected into one-cell stage embryos, similar to what has been previously described for SpCas9 13. To optimize SpG and SpRY activity, either 0.5 nL or 1 nL of a mixture containing 300 pg of mRNA and 240 pg of gRNA was injected. RNP injection mixes were prepared in 20 mM HEPES pH 7.4, 250 mM KCl, and 1 mM DTT by mixing the protein and gRNA at a ratio of 1:1.3. RNPs (6 μM) were incubated at 37 °C for 10 min and then kept on ice before use. One nL (6 fmol), or 0.5 nL (3 fmol) from the 6 μM solution was injected in one-cell stage embryos. In both cases, the mixtures were kept on ice, and any excess stored at −80°C for up to three freeze-thaw cycles maintained similar efficiency. Zebrafish embryo phenotypes were analyzed at 24 hours or at 48 hours, depending on the target gene, using an Olympus SZX16 stereoscope and photographed with a Nikon DS-F13 digital camera and further edited in Adobe Photoshop.
Microinjection in C. elegans
Injection mixes were prepared by combining Cas9 nuclease, tracrRNA, and crRNA, and incubated at 37 °C for 15 minutes. When necessary, ssODN repair templates were added after incubation and the mixture centrifuged at 13,200 rpm for 2 minutes to settle particulate matter. The injection mixes were kept on ice prior to loading of the needles and any excess stored at −20°C afterwards. Eppendorf Femtotips® capillary tips (Eppendorf, Cat. No. 930000035) for microinjection were loaded with 2 μl of the injection mix and fixed onto the XenoWorks Microinjection System (Sutter Instrument) coupled to a Nikon Eclipse Ti-S inverted microscope with Nomarski optics. Approximately 15–20 young adult hermaphrodites were injected for each experimental condition. The worms were fixed on 2% agarose pads with halocarbon oil in groups of five and were injected in one or both gonad arms. Injected worms were recovered in M9 buffer and were individually separated onto nematode growth medium (NGM) agar plates. The plates were incubated at 25 °C for three days.
RNP testing in C. elegans
Bristol N2 worms were injected with dpy-10 RNPs and screened for the presence of dumpy and/or roller phenotypes. The editing efficiency from each injected P0 was calculated by counting the proportion of dumpy or roller F1 progeny over the total number of F1 progeny laid by each P0 worm. Occasionally, injections targeting the dpy-10 locus included pCFJ90 (myo-2p:: mCherry) and pCFJ104 (myo-3p::mCherry) as co-markers to facilitate the screening of recessive Dpy phenotypes in the F2. On the other hand, CER541 worms harboring a homozygous gtbp-1::wrmScarlet fluorescent reporter were injected with anti-wrmScarlet RNPs combined with dpy-10 RNP as co-CRISPR marker. From each injected P0, between five to ten dumpy or roller F1s were separated and allowed to lay F2 progeny. The F2 progeny were then screened for wrmScarlet knockouts using a Nikon SMZ800 stereomicroscope linked to a Nikon Intensilight C-HGFI epi-fluorescence illuminator with an mCherry filter. The editing efficiency from each injected P0 was calculated by counting the proportion of separated F1 progeny that gave rise to non-fluorescent F2 worms. Non-fluorescent worms were indicative of indels arising from error-prone repair of DSBs.
Endogenous reporters for the usp-48, trx-1, and W05H9.1 loci were generated to test the efficiency of HDR by using SpG with NGN PAMs, and in the cep-1 locus by using SpRY with an NAN PAM. Using the Nested CRISPR approach 18, a C-terminal wrmScarlet fusion was made in usp-48, trx-1, and cep-1, while a transcriptional reporter was made for W05H9.1 by removing the entire coding sequence and replacing it with the GFP::H2B sequence. For each locus, an injection mix containing the SpG nuclease, target gene crRNA(s), dpy-10 crRNA, tracrRNA, and an ssODN repair template consisting of a partial wrmScarlet or GFP::H2B fragment flanked by two 35-bp homology arms were assembled and injected into Bristol N2 worms. dpy-10 co-edited F1 progeny were then separated individually or in pools onto NGM agar plates, allowed to lay F2 progeny, and genotyped via PCR. The overall editing efficiency was calculated by counting the proportion of F1 worms harboring the insertion of the correct size over the total number of genotyped F1 animals. Two independent homozygous lines for the step 1 insertion are then verified via Sanger sequencing and the complete wrmScarlet or GFP::H2B fragment is inserted via the nested CRISPR step 2 protocol using SpCas9 18. Finally, the R350C substitution in the swsn-4/SMARCA4 gene was accomplished via the introduction of a missense mutation using an ssODN repair template. A list of primers used for genotyping and sequencing can be found in Supplementary Table 1.
Generation of endogenous germline SpG -expressing C. elegans strains
EG9615 (oxSi1091mex-5p::Cas9(smu-2 introns) unc-119(+) II; unc-119(ed3) III), a strain carrying a transgene that expresses SpCas9 in the germline was a gift from Dr. Matthew Schwartz and Dr. Erik Jorgensen (unpublished). EG9615 hermaphrodites were injected with three crRNAs and three ssODN repair templates to introduce the six amino acid substitutions to convert SpCas9 to SpG. Each crRNA and ssODN repair template introduced the D1135L and S1136W, G1218K and E1219Q, and R1335Q and T1337R substitutions by pairs (Supplementary Table 1). The first round of injections contained all three crRNAs at a final concentration of 1 μM each, tracrRNA at 3.2 μM, the three ssODN repair templates at 2.2 μM each, pCFJ90 at 2.5 ng/μL, and pCFJ104 at 5.0 ng/μL. F1 progeny with visible mCherry expression in the pharynx or body wall were singled out and were genotyped via single worm lysis and PCR after laying F2 progeny. From the first set of injections, a strain with the D1135L and S1136W substitutions was successfully isolated. Then, a second round of injections was made over this strain to introduce the remaining substitutions. However, the remaining two crRNAs were combined with tracrRNA and IDT Cas9 v3 nuclease at a final concentration of 2.1 μM to form RNPs in an attempt to increase the editing efficiency. Worms were injected with this injection mixture and genotyped as previously described. The four remaining substitutions were successfully isolated and three independent lines were kept and frozen (CER658, CER659, and CER660).
Indel mutation sequencing
In zebrafish, five embryos per injection were collected at 24 hpf and genomic DNA was extracted following a protocol adapted from Meeker et al., 2007. Using this genomic DNA as template, a ~100 bp PCR product was obtained using the following parameters: 30 seconds at 98°C, 35 cycles of 10 seconds at 98°C, 30 seconds at 60°C, and 30 seconds at 72°C, and a final step at 72°C for two minutes. In C. elegans, individual worms were collected (F2 dumpies or F2 wrmScarlet knockouts from F1 heterozygotes) and genomic DNA was extracted via single worm lysis. Using this genomic DNA as template, a PCR product was amplified using a touchdown PCR program: 2 minutes at 98°C, 11 cycles of 15 seconds at 98°C, 15 seconds at 64°C (decrease by 0.5°C per cycle), and 30 seconds at 72°C, 24 cycles of 15 seconds at 98°C, 15 seconds at 59°C, and 30 seconds at 72°C, and a final step at 72°C for ten minutes.
PCR products were visualized on agarose gel, purified (QIAquick PCR purification, Qiagen) and sequenced. After sanger sequencing (Stabvida), indel mutations were identified and analyzed using the ICE tool (Synthego Co). Target sequences were aligned by mafft v7.271 using default options 54.
No statistical methods were used to predetermine sample size. The experiments were not randomized and investigators were not blinded to allocation during experiments and outcome assessment. Bar graphs show the average and are represented with S.E.M bars and violin plots are represented with individual data points and the median. Zebrafish phenotype or viability data in different injections conditions come from, at least, two independent experiments per figure panel. C. elegans data come from one or two independent experiments per figure panel with all data derived from parallel injections unless otherwise specified in the figure legend. Exact Fisher test, Student’s t-test or one-way ANOVA (all values shown are two-sided) with Tukey’s test for multiple comparisons were performed using SSPS (IBM, Armonk, NY, USA), Prism (GraphPad Software v9, La Jolla, CA, USA) or R programming language.
All relevant data are available from corresponding authors upon reasonable request.
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 PAM, 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.
We have characterized and optimized the activity of SpG and SpRY 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 efficient mutagenesis at regions inaccessible to SpCas9. We also found that the CRISPRscan algorithm can predict SpG and SpRY 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.