Methods

Zebrafish husbandry

All vertebrate animal work was performed at the facilities of the University of Utah, CBRZ. This study was approved by the Office of Institutional Animal Care & Use Committee (IACUC # 18-2008) of the University of Utah’s animal care and use program.

Cloning and transcription of CRISPR-Cas systems and anti-CRISPR

CRISPR-Cas and anti-CRISPR plasmids were ordered from Addgene, thanks to gifts from many investigators (Hou et al. 2013; Gagnon et al. 2014; Ran et al. 2015; Kleinstiver et al. 2015; Pawluk et al. 2016; Moreno-Mateos et al. 2017; Rauch et al. 2017). Open reading frames were amplified using PCR with primers that add overlapping ends corresponding to the pCS2 vector, and subcloned into the pCS2 vector using NEBuilder HiFi DNA Assembly (NEB). All oligo sequences are available in Table S2. Plasmids were miniprepped (Zymo) and confirmed via Sanger sequencing. Each vector was linearized using NotI restriction digest (NEB). Capped mRNA was synthesized using the HiScribe SP6 kit (NEB) and purified using the RNA Clean & Concentrator kit (Zymo). All constructs generated in this study are available at https://www.addgene.org/James_Gagnon/.

Generation of sgRNAs and crRNAs

SpyCas9 tyr and tbxta sgRNAs were synthesized using EnGen sgRNA Synthesis Kit (NEB). SauCas9, Sth1Cas9, and Nme1Cas9 sgRNAs were synthesized as previously described, with modifications (Gagnon et al. 2014). Briefly, gene-specific and constant oligos were designed for overlap extension for template synthesis. A reaction containing 2 ul constant oligo (5 uM), 2 ul gene-specific oligo (5 uM), 12.5 ul 2X Hotstart Taq mix, and 8.5 ul water was cycled on a thermocycler using this protocol - 95°C for 3 mins, then 30 cycles of (95°C 30 seconds, 45°C 30 seconds, 68°C 20 seconds), followed by 68°C for 5 minutes. Templates were run on a 1% TAE agarose gel to confirm correct band size, and purified using the DNA Clean and Concentrator kit (Zymo). sgRNAs or crRNAs were transcribed using the HiScribe T7 High Yield RNA Synthesis kit (NEB) or the MEGAscript T7 Transcription kit (Thermo Fisher), and purified using the RNA Clean & Concentrator kit (Zymo), or by using phenol chloroform RNA extraction. AspCas12a and LbaCas12a tyr crRNAs were transcribed following a previously-described protocol (P. Liu et al. 2019), or chemically synthesized (Synthego). Nme2Cas9 sgRNAs were generated using a previously published protocol (Amrani et al. 2018; Edraki et al. 2019). The sgRNAs or crRNAs were then pooled into a single mix at equal molarities (∼600 ng/ul for sgRNAs, ∼1800 ng/ul for crRNAs). For SpyCas9 we pooled 4 or 5 sgRNAs, SauCas9 we pooled 5 sgRNAs, Nme1Cas9 we pooled 10 sgRNAs, Nme2Cas9 we pooled 10 sgRNAs, Sth1Cas9 we pooled 3 sgRNAs, and LbaCas12a we pooled 3 crRNAs into the final pools we used to make injection mixes. All oligo sequences are in Table S2.

Cas9 CRISPR-Cas injection mixes

We assembled microinjection mixes in the following order in 1.5 ml tubes: 1 ul of 1 M KCl (Sigma-Aldrich), 0.5 ul phenol red (Sigma-Aldrich), 1 ul of a mix of sgRNAs, generated as described above. This pre-mix was briefly vortexed and centrifuged to bring the solution to the bottom of the tube. Then 1 ul Cas mRNA (∼300 ng/ul), 1 ul of 20 μM Cas protein (SpyCas9 and SauCas9 from NEB; Nme2Cas9 was purified as previously described (Edraki et al. 2019)), and/or 1 ul of anti-CRISPR mRNA (∼500 ng/ul) was added to the tube, and the mix vortexed and centrifuged again. Nme2Cas9 injection mixes were incubated at 37C for 5 minutes, then kept on ice until ready. 1-2 nl was injected into the cell of a zebrafish zygote. Nme2Cas9 and Sth1Cas9 were injected into tyr +/- embryos for more sensitive mutation detection, and wildtype embryos for T7E1 assay, as shown in Figure S3.

Cas12a CRISPR-Cas injection mixes

We assembled microinjection mixes in the following order in 1.5 ml tubes: 1 ul of 1 M KCl (Sigma-Aldrich), 0.5 ul phenol red (Sigma-Aldrich), 1.5 ul (LbaCas12a) or 2.5 ul (AspCas12a) of a mix of crRNAs, generated as described above. This pre-mix was briefly vortexed and centrifuged to bring the solution to the bottom of the tube. Then 1.5 ul of 50 μM LbaCas12a protein (NEB) or 1 ul of 63 μM AsCas12a Ultra protein (IDT) was added to the tube, and the mix vortexed and centrifuged again. Injection mixes were incubated at 37C for 5 minutes, then kept on ice until ready. 1-2 nl was injected into the cell of a zebrafish zygote. Injection mixes for multiplexed mutagenesis were generated by doubling the pre-mix and Cas protein volumes.

CRISPR mutagenesis assays

At 1 dpf, we screened to remove unfertilized or dead embryos. In most conditions, <10% of embryos exhibited toxicity, which we attribute mostly to injection artifact. At 2-3 dpf, embryos were scored for pigmentation loss into one of four categories: fully pigmented (100% pigmentation), mostly pigmented (51-99% pigmentation), mostly not pigmented (6-50% pigmentation), and not pigmented (0-5% pigmentation) (Figure 2B). T7 endonuclease 1 assay was performed following manufacturer’s protocol (NEB).

Imaging

Larvae were imaged at 3 dpf, with the exception of the embryos in Figure 3, which were imaged at 1 or 2 dpf. In cases where the larvae had not hatched from the chorion, they were manually dechorionated using tweezers. From a stock solution of 4 mg/ml of Tricaine (Sigma-Aldrich), we create a diluted solution of 0.0064 mg/ml in E3 buffer. The larvae were anesthetized in diluted Tricaine solution for 2 minutes. Once the larvae were immobile, they were moved onto a thin layer of 3% methylcellulose (Sigma-Aldrich) and oriented for a lateral or top view. Images of larvae were taken using a Leica M205FCA microscope with a Leica DFC7000T digital camera.

Barcode design

Well-edited CRISPR targets that were not present in the zebrafish genome were identified for SauCas9 and LbaCas12a using a combination of computational design, literature review, and experimental validation. SpyCas9 targets were used from previous barcode designs (McKenna et al. 2016). Five sites for each of the three systems were concatenated in an array, with a three nucleotide spacer between each target site, and cloned into a vector containing a myl7:GFP marker and Tol2 recognition sites for transgenesis. This transgenesis vector was named pTol2-HybridBarcode, and is available at https://www.addgene.org/James_Gagnon/.

Generation of transgenic zebrafish with a single-copy CRISPR barcode

To generate founder fish, 1-cell embryos were injected with Tol2 mRNA and pTol2-HybridBarcode vector DNA. Potential founder fish were screened for GFP expression in the heart at 30 hpf and grown to adulthood. Founder transgenic fish were identified by outcrossing to wild type and screening clutches of embryos for heart GFP expression at 30 hpf. A single-copy Tol2 transgenic line was identified from a single founder using copy number qPCR as previously described (McKenna et al. 2016). This barcode line was given the ZFIN line designation zj1Tg, expanded by out-crossing, and used for all experiments in this manuscript.

CRISPR barcoding

The barcode linezj1Tg was crossed to wild type or Tg(hsp70l:zCas9-2A-EGFP,5x(U6:sgRNA))a168Tg males (Raj et al. 2018) to generate double transgenic embryos. SauCas9 and/or LbaCas12a RNPs containing the appropriate guide RNAs were injected at the 1-cell stage with 2 nl of the following injection mix: 0.75 ul 1M KCl, 1.75 ul sgRNA/crRNA mix at molar ratios, 1.25 ul SauCas9 protein, 0.75 ul LbaCas12a protein, 0.5 ul phenol red. These embryos were screened for green hearts to identify the barcode transgene, and then heat shocked to induce SpyCas9 expression, as previously described (Raj et al. 2018), and grown to 2 days of age. Genomic DNA was extracted from individual embryos following the HotSHOT method.

Amplicon sequencing of barcodes

Amplicon sequencing libraries were prepped using two rounds of PCR, which completed the Illumina adapters and added dual 8bp indices that were unique to each sample, following previously published protocols (Gagnon et al. 2014; Komor et al. 2016). Libraries were pooled at roughly equimolar ratios, and sequenced on an Illumina MiSeq using 600-cycle v3 kits.

Computational analysis

Sequencing data was processed using the previously published GESTALT pipeline, with modifications to permit the use of Singularity as a container environment on our computing cluster. All code for post-processing and analyzing barcodes will be available at https://github.com/Gagnon-lab/.

Statistical analysis

All statistical tests performed were unpaired, two-sample, one-tailed Student t-tests. The Welch t-test provided by the “t.test” function in R was used to calculate p-values in Figure 4C,E.