Study design and validation of sgRNA-Cas9 ribonucleoprotein complex efficiency. (A) Graphic representation of the target region in ush2a exon 13. The sgRNA (targeting the complementary strand) and protospacer-adjacent motif (PAM) are indicated and allow for Cas9 cleavage in close proximity of position c.2312. The asymmetric antisense template for homology directed repair contains the c.2312G>T variant and a silent PAM-site change. (B) Graphic representation of the target region in ripor2 exon 14. The sgRNA is designed to direct Cas9 to cleave within the nucleotide sequence that we aimed to delete. The asymmetric antisense template for homology directed repair, containing the 12-nucleotide deletion and a silent PAM-site change, is shown in sense orientation. (C) High resolution melting (HRM) analysis profiles of individual embryos injected with the ush2a sgRNA-Cas9 ribonucleoprotein (RNP) complex. A representative example of the Sanger sequencing traces from a single embryo is shown on the right. The typical overlapping peaks occur around the indicated Cas9 cleavage site. (D) HRM analysis profiles of individual ripor2 sgRNA-Cas9 ribonucleoprotein (RNP) complex injected embryos. The typical overlapping peaks in the sanger sequencing traces, shown on the right, are indicative of sgRNA-Cas9 RNP activity.

Next-generation sequencing analysis of knock-in efficiency after sgRNA-Cas9 RNP and HDR template delivery. (A) Percentage of mutant reads in the genomic DNA of pooled embryos after injection of sgRNA, Cas9, and asymmetric antisense oligonucleotide templates. Mutant reads are quantified as the percentage of all non-wildtype reads. Data is expressed as mean (bar) and values of individual replicates (scatterplot). (B) Knock-in efficiency after injection of sgRNA, Cas9 and asymmetric antisense oligonucleotide template. Knock-in (KI) reads are quantified as percentage of total reads with correct knock-in with and without the silent variant in the protospacer adjacent motif (PAM). The percentage of correct knock-in reads with additional variants is provided as KI + indels. Examples of correct knock-in reads with and without PAM variants and indels are provided for (C) ush2a C771F RNP injected embryos, and (D) ripor2 del12 RNP injected embryos. The correctly-introduced nucleotide substitutions are indicated in red. Unintended nucleotide substitutions are indicated with a red font on orange background. Deletions are indicated in dark blue. The PAM sites are indicated on the wildtype sequences by a green box.

Knock-in efficiency after co-delivery of the xrcc6-targeting PMO. (A) Addition of the xrcc6-targeting PMO to the RNP injection mixtures results in alterative xrcc6 splicing. Both intron retention and exon 2-skipping are observed upon RT-PCR analysis of the xrcc6 transcript in PMO-injected embryos, but not control injected embryos. Replicate injections of C711F RNPs provide insight in variability of PMO-induced alterative xrcc6 splicing. (B) Quantification of knock-in efficiency after sgRNA-Cas9 RNP and template injection with and without co-injection of the xrcc6-targeting PMO. For ripor2, possibly tp53, but not ush2a, an increase in knock-in events is observed after co-injection of sgRNA-Cas9 RNPs, template oligonucleotide and the xrcc6-targeting PMO. Results are expressed as mean (bars) and individual replicates (scatterplot). Only reads that include the PAM variant are included in the quantification to ensure that all quantified correct knock-in reads are the project of homology directed repair. (C) Comparison of changes in knock-in efficiency between embryos of the same clutch of eggs. For ush2a C771F RNPs, addition of the PMO to the injection mixture leads to variable outcomes in knock-in efficiency between the replicate injections. Three out of the four replicate injections of ripor2 del12 RNPs with xrcc6-tarageting PMO display a higher knock-in efficiency, while for tp53 the addition of the PMO leads to an increase in knock-in efficiency in one of the two replicate injections.

Proposed strategy for the efficient generation of knock-in zebrafish. Step 1 entails the verification of sgRNA-Cas9 efficiency at creating genomic lesions using high resolution melting (HRM) analysis. Step 2 concerns the injection of sgRNA-Cas9 ribonucleoprotein complexes and the asymmetric antisense oligonucleotide template with the variant of interest. Next generation sequencing (NGS) is recommended to determine knock-in efficiency. In the situation of low knock-in efficiency, knockdown of non-homologous end-joining protein Ku70 (by co-delivery of an morpholino antisense oligonucleotide (PMO), targeting xrcc6) may offer a means to improve knock-in efficiency. Finally, injected embryos with sufficient knock-in reads can be raised to adults and screened for founders with the variant of interest in the germline (step 3). The white boxes in step 1 and step 2 described the key components injected into the zebrafish zygote. A detailed protocol can be found in Supplemental Document S1.