Treatment of monogenic and digenic dominant genetic hearing loss by CRISPR-Cas9 ribonucleoprotein delivery in vivo

Abstract

Mutations in Atp2b2, an outer hair cell gene, cause dominant hearing loss in humans. Using a mouse model Atp2b2^Obl/+, with a dominant hearing loss mutation (Oblivion), we show that liposome-mediated in vivo delivery of CRISPR-Cas9 ribonucleoprotein complexes leads to specific editing of the Obl allele. Large deletions encompassing the Obl locus and indels were identified as the result of editing. In vivo genome editing promotes outer hair cell survival and restores their function, leading to hearing recovery. We further show that in a double-dominant mutant mouse model, in which the Tmc1 Beethoven mutation and the Atp2b2 Oblivion mutation cause digenic genetic hearing loss, Cas9/sgRNA delivery targeting both mutations leads to partial hearing recovery. These findings suggest that liposome-RNP delivery can be used as a strategy to recover hearing with dominant mutations in OHC genes and with digenic mutations in the auditory hair cells, potentially expanding therapeutics of gene editing to treat hearing loss.

Fig. 1. Design of a genome-editing strategy to disrupt the Obl mutant allele.

a SpCas9 sgRNAs were designed to target the mutant Atp2b2 Obl allele, in which C2765 is changed to T (red). The protospacer (green arrows) of each Obl-targeting sgRNA contains a complementary A or T (red) that pairs with the mutation in the Obl allele, but that forms a mismatch with wild-type Atp2b2 allele. b In vitro Cas9:sgRNA-mediated Atp2b2 DNA cleavage. 50 nM of a 956-bp DNA fragments of Atp2b2 or the Atp2b2^Obl/+ was incubated with 300 nM of each of the three Cas9:sgRNA for 15 min at 37 °C. Expected cleavage products of 553-bp and 403-bp were detected in all samples. M = 100 bp ladder. This was repeated independently for 3 times with similar results. c Quantification of DNA cleavage in (b) by densitometry using imageJ. d Editing shown by the indel percentages in three unsorted adult mouse primary fibroblast cells (WT, Atp2b2^Obl/+ and Atp2b2^Obl/Obl) after nucleofection of the RNP complex of Cas9:Atp2b2-mut1 gRNA. Indels were only detected in the cells with the Obl allele. n = 3 biologically independent experiments. Values and error bars are mean ± SD. Source data are provided as a Source data file. e NGS reads of indel from (d). The Obl mutation was highlighted in yellow. The red arrow indicates the cutting site. The list of indels is not comprehensive, only the top representative indels are shown.

Fig. 2. In vivo gene editing by RNP injection in Atp2b2 Obl mutant mice.

a Schematic overview of in vivo indel analysis experimental design. b, c Indel frequencies on the Obl allele from injected Atp2b2^obl/+ or Atp2b2^+/+ organ of Corti samples 4 days after injection of Cas9-GFP:Atp2b2-mut1:Lipo2000 or from uninjected Atp2b2^obl/+organ of Corti. The most abundant editing events by small deletions at the Atp2b2 sequences, grouped by similarity, from the organ of Corti samples are shown on the right. The PAM sites were marked red. The Obl mutation is highlighted in yellow. The red arrow indicates the cutting site. Values and error bars are mean ± SD. d A schematic representation of nested PCR to detect large deletions resulting from injection of Cas9-GFP:Atp2b2-mut1:Lipo2000. Itr: intron; ex: exon. e Gel image of the nested PCR products from injected Atp2b2^obl/+, WT, and uninjected Atp2b2^obl/+ samples. Red asterisks indicate small fragments of varying sizes used for NGS, and the upper bands are the ~2 kb expected fragments. From injected WT and uninjected Atp2b2^obl/+ samples, the nested PCR did not produce any smaller fragments. This was repeated independently for 3 times with similar results. f Representative reads from multiple NGS analyses of smaller nested PCR products showed deletions that spanned the entire exon 24 with the Obl mutation. g Comparison of WT and Obl allele frequency from isolated Atp2b2^ob/+ hair cells showed a shift that increased the WT allele frequency and decreased the Obl allele frequency after in vivo injection of the Lipo:Cas9:Atp2b2-mut1 RNP complex compared to uninjected Atp2b2^obl/+ hair cells. The dots represented independently purified hair cell groups from five animals. Values and error bars are mean ± SEM. h Indel profiles and read abundance at the mRNA level after injection of Lipo:Cas9:Atp2b2-mut1 RNP complex in Atp2b2^ob/+ and WT mice. Indels were only detected in the mRNA of injected Atp2b2^ob/+ but not in the mRNA of injected WT ears. i The ratio of un-modified Obl mRNA relative to un-modified WT mRNA in uninjected and injected Atp2b2^ob/+ animals based on the NGS reads. Values and error bars are mean ± SD. j The normalized ratio of un-modified Obl and WT mRNA from (i). Values and error bars are mean ± SD, n = 3. Source data are provided as a Source data file.

Fig. 3. In vivo lipid-mediated RNP delivery of Cas9:sgRNA complexes improves hair cell survival.

a Representative confocal microscopy whole mount images from uninjected (left) and Cas9:Atp2b2-mut1 sgRNA:Lipo2000 injected (right) Atp2b2^Obl/+ cochleae at P1 with the cochleae harvested at 2 months of age. PMCA2 labeled OHC (PMCA2/PVALB) was detected in the injected Atp2b2^Obl/+ cochlea and was missing in the uninjected contralateral control cochlea in the mid-base turn. In the mid-apical turn, OHCs were seen in injected and uninjected Atp2b2^Obl/+ cochleae. b, c Quantification of OHC (b) and IHC (c) survival in Atp2b2^Obl/+ mice 8 weeks after Cas9:Atp2b2-mut1 sgRNA:Lipo2000 injection (blue) compared to uninjected (red) contralateral ears across frequency regions. Values and error bars reflect the mean ± SEM of 4 mice (n = 4). Source data are provided as a Source data file. Paired t test was used for the analysis: *p < 0.05 and ****p < 0.0001.

Fig. 4. In vivo lipid-mediated delivery of Cas9:sgRNA complexes restores hair bundle morphology.

a Experimental workflow. Samples were harvested, processed and imaged at 4 weeks after injection. b Images of scanning electron microscopy (SEM) of uninjected Atp2b2^Obl/+ (left) and Cas9:Atp2b2-mut1 sgRNA: Lipo2000 injected Atp2b2^Obl/+ (right) outer hair cell bundles. This was repeated independently for 3 times with similar results. Scale bar, 10 µm. c High-magnification images of outer hair cell bundles from the insets of (b) of uninjected Atp2b2^Obl/+ (left) and Cas9:Atp2b2-mut1 sgRNA: Lipo2000 injected Atp2b2^Obl/+ cochlea (right). The arrow points to shorter stereocilia in an outer hair cell from an injected ear that are missing from an outer hair cell from an uninjected ear. This was repeated independently for 3 times with similar results. Scale bar, 2 µm. d Images of scanning electron microscopy (SEM) uninjected Atp2b2^Obl/+ (left) and Cas9:Atp2b2-mut1 sgRNA: Lipo2000 injected Atp2b2^Obl/+ (right) inner hair cell bundles. This was repeated independently for 3 times with similar results. Scale bar, 10 µm. e High-magnification images of inner hair cell bundles from the insets of (d) of uninjected Atp2b2^Obl/+ (left) and Cas9:Atp2b2-mut1 sgRNA: Lipo2000 injected Atp2b2^Obl/+ cochlea (right). This was repeated independently for 3 times with similar results. Scale bar, 2 µm.

Fig. 5. In vivo lipid-mediated delivery of Cas9:sgRNA complexes improves OHC function, hearing and auditory behavioral function.

a DPOAE thresholds in Atp2b2^Obl/+ ears injected with Cas9:Atp2b2-mut1 sgRNA:Lipo2000 (blue), uninjected Atp2b2^Obl/+ ears (red), and wild-type C3H ears (green) after 4 weeks of age. b ABR thresholds in Atp2b2^Obl/+ ears injected with Cas9:Atp2b2-mut1 sgRNA:Lipo2000 (blue), uninjected Atp2b2^Obl/+ ears (red) (the thresholds for each individual mouse could be found in Source data file), and wild-type C3H ears (green) after 4 weeks of age. c Representative ABR waveforms illustrating a threshold of 40 dB (red trace, left) at 16 kHz in a Cas9:Atp2b2-mut1 gRNA injected Atp2b2^Obl/+ inner ear compared to the threshold of 85 dB (red trace, right) of the uninjected contralateral ear of the same mouse 4 weeks after injection. d Amplitudes of ABR Wave 1 at 16 kHz in Cas9:Atp2b2-mut1 sgRNA:Lipo2000-injected ears (blue), uninjected ears (red), and wild-type C3H ears (green) after 4 weeks of age. e Startle responses in Cas9:Atp2b2-mut1 sgRNA:Lipo2000 injected mice (blue), uninjected mice (red) and wild-type C3H ears (green) at 8 weeks post injection. Statistical tests were two-way ANOVA with Bonferroni correction for multiple comparisons: **p < 0.01, ***p < 0.001, and ****p < 0.0001. Values and error bars are mean ± SEM.

Fig. 6. In vivo gene editing partially recovers hearing of digenic mutation origin.

a A schematic diagram of the experimental design. b ABR and c DPOAE thresholds in Atp2b2^Obl/+: Tmc1^Bth/+ ears injected with Cas9:Atp2b2-mut1:Tmc1-mut3:Lipo2000 (blue) and uninjected Atp2b2^Obl/+: Tmc1^Bth/+ ears (red) at 4 weeks post injection. d ABR and e DPOAE thresholds in Atp2b2^Obl/+: Tmc1^Bth/+ ears injected with Cas9:Atp2b2-mut1 sgRNA:Lipo2000 (blue) and uninjected Atp2b2^Obl/+:Tmc1^Bth/+ ears (red) at 4 weeks of age. f ABR and g DPOAE thresholds in Atp2b2^Obl/+:Tmc1^Bth/+ ears injected with Cas9:Tmc1-mut3:Lipo2000 (blue) and uninjected Atp2b2^Obl/+: Tmc1^Bth/+ ears (red) at 4 weeks post injection. **p < 0.01, ***p < 0.001, and ****p < 0.0001. Values and error bars reflect mean ± SEM.