Simple and efficient CRISPR/Cas9-mediated targeted mutagenesis in Xenopus tropicalis

Abstract

We have assessed the efficacy of the recently developed CRISPR/Cas (clustered regularly interspaced short palindromic repeats/CRISPR-associated) system for genome modification in the amphibian Xenopus tropicalis. As a model experiment, targeted mutations of the tyrosinase gene were verified, showing the expected albinism phenotype in injected embryos. We further tested this technology by interrupting the six3 gene, which is required for proper eye and brain formation. Expected eye and brain phenotypes were observed when inducing mutations in the six3 coding regions, as well as when deleting the gene promoter by dual targeting. We describe here a standardized protocol for genome editing using this system. This simple and fast method to edit the genome provides a powerful new reverse genetics tool for Xenopus researchers.

FIG. 1

Strategy for CRISPR/Cas-mediated genome modification. (a) Schematic representation of experimental procedure. We found that conventional plasmid subcloning methods for making sgRNA templates was time-consuming and inconvenient; instead all sgRNAs used in this study are transcribed in vitro from double-stranded DNA templates that were made by PCR (except the sgRNA for tyr target site 1, which used a cloned template, pDR274-Xt-tyr). This PCR strategy uses a 5’ oligonucleotide (primer) that begins with the T7 promoter (shown as an orange line in this schematic; alternatively, T3 or SP6 promoters could be used) and contains the genomic target sequence (shown as a blue line in this schematic; note the genomic target sequence must begin with a G for proper transcriptional initiation using the T7 promoter) and a 3’ oligonucleotide (primer) that partly overlaps the 5’ primer and contains the sgRNA backbone sequence required for proper folding of sgRNAs. This rapid, easy way to make sgRNA templates by PCR takes less than two days to produce sgRNAs to inject once oligonucleotides are received. During the preparation of this manuscript, we have noticed that similar strategies have also been reported [e.g., (Bassett et al. 2013)] and thus this method has general versatility. The resulting sgRNAs were co-injected with Cas9 mRNA into one cell-stage Xenopus embryos. The bottom part of the scheme shows how sgRNA and Cas9 work to cleave the target site in the genome. Briefly, the sgRNA forms a complex with the Cas9 protein and identifies the target site via complementary basepairing. The Cas9 nuclease subsequently cleaves the genomic DNA at the target site, just upstream of the PAM sequence. (b) Targeting strategy for the two genes described in this study. Both tyr and six3 genes were targeted in exon 1 to cause frame-shifts after the translation initiation codon (ATG). Two independent sgRNAs were tested for both genes. The six3 gene was furthermore targeted in the proximal promoter region with two sgRNAs simultaneously to cause deletion of the promoter region. Bent arrows (blue for tyr, orange for six3 exons and brown for the six3 promoter) show sgRNA targets. Arrows (black for tyr, green and grey for six3) indicate genomic PCR primers for mutation analyses. Note that the 3’ primer for the six3 gene is common for both sets of PCR reactions but colored differently for clarity. Drawings are not to scale.

FIG. 2

Successful targeting of the tyrosinase gene caused albinism in Xenopus embryos. (a) Different dose combinations of Cas9 mRNA and sgRNA were tested. The severity of the phenotype was directly dependent on the amounts of RNAs injected. bacCas9, the original bacterial-codon Cas9 mRNA (which shows examples with a weak phenotype; a patchy loss of pigmentation in the RPE is indicated by white arrowheads). humCas9, the modified Cas9 using humanized-codons. sgRNA, targeting tyrosinase gene (first target, see Fig. 3b for location relative to ATG). The toxicity of sgRNA seems to vary depending on its sequence and also on the batch of embryos. For example, in one specific experiment, tyrosinase sgRNA (target1) was overall relatively non-toxic, with more than 75% of injected embryos developing normally one day after injection at all tested doses shown here (76.9-97% survivors [n=19-33] as opposed to 84.8% of uninjected embryos [n=46]), whereas in other experiments, survival was between 50-60% (see Supporting Information Table 1). (b) Representative results of conventional sequencing assays (top three panels). Each single embryo injected with indicated RNAs was lysed and the targeted genomic region was PCR-amplified; amplicons were then directly sequenced (DSP assay, see the text). Perturbation of peaks on the 3’ side of the PAM region (red arrows) suggests in-del events happen in-between the target sequence (shaded in purple) and the PAM region (shaded in red). PCR amplicons were re-cloned and sequenced to show the profile of individual mutations found in mosaic individuals (bottom sequence alignments). Dashes (-) indicate gaps. Lowercase green characters indicate insertions or substitutions. The numbers in parentheses indicate the frequency of each mutation pattern seen in total numbers of sequenced clones.

FIG. 3

The mutation of a second tyr gene target site also caused albinism. (a) Phenotype of injected embryos (bottom five embryos; top five embryos are uninjected sibling controls) with Cas9 mRNA (2.2 ng) and sgRNA (200 pg) for the second target site. (b) Representative mutation profiles from two embryos sequenced after cloning. The second target sequence (shaded in purple) and the PAM sequence (in red) are shown together with first target and PAM sequences (in gray; see Fig. 2) relative to the ATG (Met) site. The alignment shown is labeled as described in Fig. 2b.

FIG. 4

Summary of phenotype caused by mutations of six3 gene. All targeted mutations showed a similar phenotype. (a-e) Phenotype variations seen in st.40 embryos targeted with the sgRNA for coding region site 1. By this stage, the obvious phenotype is significantly reduced eye size, but a brain defect (see below) is not yet obvious. The phenotype is six3-specific because it was partially rescued by co-injection of six3 mRNA (b, c). Severity of phenotype was scored as +++ (most severe, no eye or tiny piece of eye), ++ (severe, small and malformed eye), + (modest, small relatively normal looking eye), and - (no or little phenotype) (a-c). As observed with the tyr target injections, toxicity of injected RNAs varied greatly depending on batch of embryos. For example, in different experiments, the percentage of normal embryos recovered after the same RNA doses were injected (six3 target 1) varied from 43% to 87% (see Supporting Information Table 1). (d) Chromatograms of the DSP assay from an uninjected embryo (WT) and one of rescued embryos shown in (b). Perturbation of peaks is seen around the PAM region (two-headed red arrow), suggesting the presence of mutated sequences. At later stages, the brain phenotype becomes obvious (e-g). (e) Examples of the phenotype of embryos (emb1 to emb3) targeted to delete the proximal promoter region at st.42. Using the primer as schematically shown in Fig. 1b (green arrows), genomic PCR of sibling uninjected wild-type embryo showed an expected ~ 1.2 kbp band, whereas embryos injected with sgRNAs had an extra ~ 0.6 kbp band (emb1-3, white arrows). One band (marked by *) was recloned and sequenced; the results are shown in Supporting Information Fig. S1. The promoter deletion injections were on average more toxic, perhaps due to higher total amounts of RNAs being injected, but survival rates as high as 68% were observed in individual experiments. (f) Phenotype examples of st. 42 embryos injected with sgRNA targeting coding region site 1 (left panel) and site 2 (right panel), lateral views. Compared to wild-type embryos (bottom, WT, round shaped head), mutant embryo heads had a more angular and flattened shape. The dorsal view (second panel) and frontal sections (third panel) of the same embryos (emb1-1, 1-2, 1-3) in first left panel clearly show narrow heads (compare two-headed white arrows between WT and mutants) and smaller brains (black arrowheads in frontal sections), especially the forebrain (red arrowheads in second panel, see also Fig. 4g) in the mutant embryos. The sequencing summary of mutated loci of emb1-1 and 1-3 in the top left panel and emb2-1, 2-2 in the top right panel are shown in Supporting Information Fig. S1. (g) A mild phenotype seen in embryos (target site 2) at st. 46. Top panel shows a dorsal view of embryos. Second panel shows horizontal sections of equivalent embryos shown in the top panel. At early stages, these embryos could be scored as wild type. Later observation of mutant embryos reveals that the nasal pits (red arrows) are fused (shown here) or shifted medially (not shown), and the telencephalon (red arrowheads) is smaller and fused. The last panel shows chromatograms of the DSP assay from one of the sectioned embryos shown above to confirm genomic mutations as evidenced by perturbation of peaks around PAM region (two-headed red arrow) in the mutant embryo.