Correction of the auditory phenotype in C57BL/6N mice via CRISPR/Cas9-mediated homology directed repair

Abstract

Background: Nuclease-based technologies have been developed that enable targeting of specific DNA sequences directly in the zygote. These approaches provide an opportunity to modify the genomes of inbred mice, and allow the removal of strain-specific mutations that confound phenotypic assessment. One such mutation is the Cdh23 (ahl) allele, present in several commonly used inbred mouse strains, which predisposes to age-related progressive hearing loss.

Results: We have used targeted CRISPR/Cas9-mediated homology directed repair (HDR) to correct the Cdh23 (ahl) allele directly in C57BL/6NTac zygotes. Employing offset-nicking Cas9 (D10A) nickase with paired RNA guides and a single-stranded oligonucleotide donor template we show that allele repair was successfully achieved. To investigate potential Cas9-mediated 'off-target' mutations in our corrected mouse, we undertook whole-genome sequencing and assessed the 'off-target' sites predicted for the guide RNAs (≤4 nucleotide mis-matches). No induced sequence changes were identified at any of these sites. Correction of the progressive hearing loss phenotype was demonstrated using auditory-evoked brainstem response testing of mice at 24 and 36 weeks of age, and rescue of the progressive loss of sensory hair cell stereocilia bundles was confirmed using scanning electron microscopy of dissected cochleae from 36-week-old mice.

Conclusions: CRISPR/Cas9-mediated HDR has been successfully utilised to efficiently correct the Cdh23 (ahl) allele in C57BL/6NTac mice, and rescue the associated auditory phenotype. The corrected mice described in this report will allow age-related auditory phenotyping studies to be undertaken using C57BL/6NTac-derived models, such as those generated by the International Mouse Phenotyping Consortium (IMPC) programme.

Fig. 1

Design of two targeting strategies to recover normal splicing/function of the C57BL/6NTac Cdh23 gene. a Design 1 utilises a 121 bp single-stranded oligonucleotide donor (ssODN_U1) in combination with two single guide RNAs (sgRNA_U1 and sgRNA_D1), which flank the Cdh23 ^ahl allele. The full ssODN_U1 donor sequence is shown below the C57BL/6NTac Cdh23 ^ahl sequence (uppercase denotes exonic sequence and lowercase denotes intronic sequence), with homology indicated by dashed lines, the Cdh23 ^ahl allele with an arrow, and two base changes shown in red text. The two synonymous base substitutions are designed to prevent further modification by the CRISPR/Cas9 following repair. The corrected Cdh23 ^753A>G allele is the allele found in inbred mouse strains that do not demonstrate age-related hearing loss (ARHL). The final corrected Cdh23 ^753A>G(U1) gene sequence closely matches that found in these non-ARHL inbred strains at the nucleotide level, except for the two synonymous base substitutions (c.724A > T and c.725G > C; red text). The Cdh23 ^753A>G(U1) product is predicted to be identical to the Cdh23 protein found in non-ARHL mouse strains. b Design 2 also utilises a 121 bp ssODN (ssODN_U2) in combination with two single guide RNAs, sgRNA_D1 also used in design 1 and sgRNA_U2, which lies across the Cdh23 ^ahl locus. The full ssODN_U2 donor sequence is shown below the C57BL/6NTac Cdh23 ^ahl sequence, with homology indicated by dashed lines, the Cdh23 ^ahl allele with an arrow, and one intronic base change shown in red text. The intronic base substitution is designed to prevent further modification by the CRISPR/Cas9 following repair. The final corrected Cdh23 ^753A>G(U2) gene sequence is identical to that found in non-ARHL inbred strains at the nucleotide level, with only an intronic base substitution (c.753 + 9c > t; red text). The Cdh23 ^753A>G(U2) product is predicted to be identical to the Cdh23 protein found in non-ARHL mouse strains

Fig. 2

Targeting events in F₀ mice and transmission to F₁ offspring. a Targeting events in an F₀ using design 1. The genotype of the F₀ was directly assessed by ear clip DNA analysis. This revealed the presence of two alleles: allele 1 corresponding to the legitimate HDR event and allele 2 corresponding to the WT allele. The genotype of the F₀ animal was confirmed through the transmission of both alleles to the F₁ population, several of which were identified as heterozygous for the legitimate repair (see trace). b Targeting events in an F₀ using design 2. The genotype of the F₀ was directly assessed by ear clip DNA analysis. This revealed the presence of three alleles: allele 1 corresponding to an NHEJ event consisting of a 24 nucleotide deletion; allele 2 corresponding to a combination of the intended repair and additional sequence changes (illegitimate repair) comprising modification of the two targeted nucleotides and a 3′ deletion of 24 nucleotides; and allele 3 corresponding to a legitimate repair. The genotype of the F₀ animal was confirmed through transmission of these three alleles to the F₁ population (F₁ allele 1 and F₁ allele 2). However, one additional allele was also identified in the F₁ population: ‘F₁ allele 4’ comprising WT sequence

Fig. 3

CRISPR-Cas9 mediated repair of the Cdh23 ^ahl allele in C57BL/6NTac mice preserves age-related high-frequency hearing and sensory hair cell stereocilia bundles. a ABR measurements from 24-week-old C57BL/6NTac mice (Cdh23 ^ahl/ahl) and their heterozygous ‘repaired’ C57BL/6NTac littermates (Cdh23 ^ahl/753A>G). As previously reported, by 24 weeks of age the WT Cdh23 ^ahl/ahl (n = 17) mice show elevated hearing thresholds (>60 dB SPL) for the 32 kHz stimulus, the highest frequency tested. However, at 24 weeks of age the Cdh23 ^ahl/753A>G (n = 13) littermates do not have elevated thresholds at 32 kHz, but instead display thresholds similar to those measured for the 8 and 16 kHz stimuli (~20–35 dB SPL). b By 36 weeks of age the Cdh23 ^ahl/ahlmice (n = 12) have very elevated hearing thresholds (≥80 dB SPL) for the 32 kHz stimulus, showing progression of the high-frequency hearing impairment. However, at 36 weeks of age the Cdh23 ^ahl/753A>G littermate mice (n = 9) still exhibit hearing thresholds within the normal range (~20–35 dB SPL) for all frequencies tested (8, 16 and 32 kHz). These results indicate that the CRISPR/Cas9 repaired Cdh23 allele is sufficient to preserve high-frequency hearing in C57BL/6NTac mice. ABR data analysed using an unpaired t test with Welch’s correction, and shown as mean ± standard deviation. c Scanning electron micrographs of the sensory epithelia in the apex, mid and base regions of the cochlea in WT C57BL/6NTac (Cdh23 ^ahl/ahl) and heterozygous repaired C57BL/6NTac (Cdh23 ^ahl/753A>G) mice at 36 weeks of age. Loss of outer hair cell (OHC) bundles is evident at the cochlear base of Cdh23 ^ahl/ahl mice. No loss of OHC bundles is evident in the age-matched Cdh23 ^ahl/753A>G littermate mice. d, e Cochleograms showing the number of inner hair cell (IHC) and OHC bundles present in the apex, mid and base regions of the cochlea in WT C57BL/6NTac (Cdh23 ^ahl/ahl) (n = 4) and heterozygous repaired C57BL/6NTac (Cdh23 ^ahl/753A>G) (n = 4) mice at 36 weeks of age. By 36 weeks of age, no significant loss of IHC bundles in any cochlear region of Cdh23 ^ahl/ahl or Cdh23 ^ahl/753A>G mice is observed. Significant loss of OHC bundles is found at the cochlear base of Cdh23 ^ahl/ahl mice, whereas no loss is found in the cochleae of Cdh23 ^ahl/753A>G mice. Hair cell count data analysed using an unpaired t test with Welch’s correction, and shown as mean ± standard error of the mean. *p < 0.05, **p < 0.01, ****p < 0.0001, ns not significant