A rapid and effective method for screening, sequencing and reporter verification of engineered frameshift mutations in zebrafish

Abstract

Clustered regularly interspaced palindromic repeats (CRISPR)/Cas-based adaptive immunity against pathogens in bacteria has been adapted for genome editing and applied in zebrafish (Danio rerio) to generate frameshift mutations in protein-coding genes. Although there are methods to detect, quantify and sequence CRISPR/Cas9-induced mutations, identifying mutations in F1 heterozygous fish remains challenging. Additionally, sequencing a mutation and assuming that it causes a frameshift does not prove causality because of possible alternative translation start sites and potential effects of mutations on splicing. This problem is compounded by the relatively few antibodies available for zebrafish proteins, limiting validation at the protein level. To address these issues, we developed a detailed protocol to screen F1 mutation carriers, and clone and sequence identified mutations. In order to verify that mutations actually cause frameshifts, we created a fluorescent reporter system that can detect frameshift efficiency based on the cloning of wild-type and mutant cDNA fragments and their expression levels. As proof of principle, we applied this strategy to three CRISPR/Cas9-induced mutations in pycr1a, chd7 and hace1 genes. An insertion of seven nucleotides in pycr1a resulted in the first reported observation of exon skipping by CRISPR/Cas9-induced mutations in zebrafish. However, of these three mutant genes, the fluorescent reporter revealed effective frameshifting exclusively in the case of a two-nucleotide deletion in chd7, suggesting activity of alternative translation sites in the other two mutants even though pycr1a exon-skipping deletion is likely to be deleterious. This article provides a protocol for characterizing frameshift mutations in zebrafish, and highlights the importance of checking mutations at the mRNA level and verifying their effects on translation by fluorescent reporters when antibody detection of protein loss is not possible.

Keywords: CRISPR; Cas9; Mutation; Reporter; SgRNA; Zebrafish.

Fig. 1.

Strategy for F1 mutation carrier screening and identification of their mutations. (A) Screening of F1 mutation carriers. Genotyping, mutation screening and sequencing in this strategy are illustrated by data from screening of pycr1a mutation carriers. F1 mutation carriers are screened by fin clipping, preparing DNA extracts and running PCRs followed by HMA and the sample results of screening six F1 fish are shown (2 and 6 are positive and marked with ‘+’). (B) Cloning restriction site-tagged PCR products from multiple mutation carriers. Positive mutation carriers are split into groups of four, separated in individual tanks and assigned identifiers (e.g. 1A) and the corresponding forward PCR primers with either AgeI, ClaI or SacII. PCRs with the assigned forward primers and common reverse primer were run on DNA extracts from F1 fish as per assignment, pooled and cloned using a TOPO-cloning procedure. (C) HMA analysis of pycr1a bacterial clones. HMA on colony PCR products from bacterial clones mixed with wild-type PCR products is performed and positive clones are identified. (D) Restriction analysis of M13 PCR products. PCR products from positive bacterial clones amplified with M13 primers were digested with enzymes, for which sites were inserted into forward PCR primers, and clones digestible with each of the enzymes are identified. (E) Sequencing analysis of selected pycr1a clones. The identified pycr1a plasmid clones corresponding to single F1 zebrafish were sequenced and analyzed both at the restriction site position and the mutation site showing complete agreement with previous assays.

Fig. 2.

Cloning strategy for producing mutation reporter constructs using pCS2+MCS-P2A-sfGFP vector. (A) Mutation reporter vector structure and cloning strategy. PCR products are amplified from cDNA with the primers designed to introduce the T3 polymerase promoter and PacI and AscI sites for cloning and to amplify the 5′ UTR and 200 codons of coding sequence for insertion into pCS2+MCS-P2A-sfGFP vector. (B) The multiple cloning site sequence (MCS) of pCS2+MCS-P2A-sfGFP contains BstBI, EcoRI, PacI, SphI and AscI restriction sites followed by the P2A sequence and sfGFP coding sequence. (C) Gene mutations in cloned cDNA fragments. Mutations identified at the level of cDNA in chd7 (deletion of 2 nt), hace1 (insertion of 8 nt) and pycr1a (deletion of 71 nt) are shown using alignment of mutant sequences to wild-type ones.

Fig. 3.

Mutation in pycr1a exon 3 disrupts predicted exonic splicing enhancers. (A) An insertion of CTGATGG (ins7) was introduced by CRISPR/Cas9 targeting at the sgRNA site 2 bp upstream (indicated by an arrow) from the PAM sequence. sgRNA target site, PAM, intron and exon sequences are indicated. (B) RT-PCR detection of deletion in pycr1a cDNA with the skipped exon 3. Amplification of pycr1a cDNA fragments for insertion into mutation reporter constructs shows evidence of exon skipping (cDNA fragment with deletion). (C) Exonic splicing enhancers disrupted by ins7 mutation. Exonic splicing enhancer predictions were performed using ESE Finder 3.0 and RESCUE-ESE. For the ESE Finder enhancers, the protein matrix used is indicated and for the RESCUE-ESE, the hexamer overlapping the insertion site is shown.

Fig. 4.

Expression of mutation reporters for chd7, hace1 and pycr1a mutations quantifies their frameshift efficiency. (A) Representative images of 16-18 hpf (∼16-somite stage) zebrafish embryos injected with the indicated mutation reporter mRNA from constructs made as described in Fig. 3 or uninjected. Fluorescence images of superfolder GFP (sfGFP) and TagRFP and merged images are shown. (B) Areas in the somite regions of these embryos were quantified by measuring fluorescence intensities of both sfGFP and TagRFP, adjusted for background in the corresponding channel, and the ratio of these intensities was calculated to account for injection variability and used for plotting the data and statistical analysis. For chd7 mutation reporters, the difference between wild-type (n=15) and 2-bp deletion mutant (n=12) mutation reporters was large and very significant (P=2.034e−15). Embryos injected with hace1 wild-type (n=10) and 8-bp insertion mutant (n=12) cDNA fragments did not show any significant difference. The pycr1a wild-type (n=41) and 71-bp deletion mutant (n=28) reporters had a small but significant difference (P=7.776e−13). ***P<0.001. The injections were performed twice on different days with very similar results and embryos from one of the injections were subjected to imaging and statistical analysis of fluorescent protein intensities ratios. Student's two-tailed t-test was performed for each experiment.

Fig. 5.

Phenotypes of chd7 and pycr1a mutants. The phenotype of chd7 del2 homozygous mutants was most clearly observed after 72 hpf and included smaller eyes, enlarged heart with edema and failure to inflate the swimbladder; images show fish at the 78 hpf stage as an example. For pycr1a del71 mutants, the observed abnormality was a pronounced developmental delay at 24 hpf, as shown in the image, that persists up to about 48 hpf and disappears afterwards.