Gene Therapy for Glaucoma by Ciliary Body Aquaporin 1 Disruption Using CRISPR-Cas9

Abstract

Glaucoma is a common cause of blindness, yet current therapeutic options are imperfect. Clinical trials have invariably shown that reduction in intraocular pressure (IOP) regardless of disease subtype prevents visual loss. Reducing ciliary body aqueous humor production can lower IOP, and the adeno-associated virus ShH10 serotype was identified as able to transduce mouse ciliary body epithelium following intravitreal injection. Using ShH10 to deliver a single vector CRISPR-Cas9 system disrupting Aquaporin 1 resulted in reduced IOP in treated eyes (10.4 ± 2.4 mmHg) compared with control (13.2 ± 2.0 mmHg) or non-injected eyes (13.1 ± 2.8 mmHg; p < 0.001; n = 12). Editing in the aquaporin 1 gene could be detected in ciliary body, and no off-target increases in corneal or retinal thickness were identified. In experimental mouse models of corticosteroid and microbead-induced ocular hypertension, IOP could be reduced to prevent ganglion cell loss (32 ± 4 /mm²) compared with untreated eyes (25 ± 5/mm²; p < 0.01). ShH10 could transduce human ciliary body from post-mortem donor eyes in ex vivo culture with indel formation detectable in the Aquaporin 1 locus. Clinical translation of this approach to patients with glaucoma may permit long-term reduction of IOP following a single injection.

Keywords: AAV; Aquaporin; CRISPR-Cas9; Glaucoma; ciliary body; gene editing; intraocular pressure.

Graphical abstract

Figure 1

Adeno-Associated Virus ShH10 Serotype Efficiently Transduces Ciliary Body Epithelium following Intravitreal Injection A total of 2 × 10¹⁰ genome copies of different AAV serotypes encoding GFP driven by the CMV promoter were injected into the vitreous cavity of one eye of each mouse. (A and B) Four weeks later, (A) ciliary body expression was compared using immunofluorescent sections, and (B) retinal transduction was examined by in vivo fundal fluorescence imaging. Only the ShH10 serotype demonstrated clear ciliary body GFP expression. Representative images from n = 8, two independent experiments. Scale bars, 25 μm.

Figure 2

Aquaporin 1 Is Expressed in the Mouse Ciliary Body and Can Be Targeted Using CRISPR-SaCas9 The mouse eye expresses AQP1 predominantly in the ciliary body, cornea, and RPE. (A–C) Representative western blot and pooled data from dissected tissues (A) for protein (B) and RNA (C) by quantitative PCR; n = 4–6 eyes. (D) Genomic map of exon 1 of mouse Aqp1 displaying sequence and binding location of tested SaCas9-compatible short guide RNAs (sgRNA) tested. (E) T7 endonuclease 1 assay for different plasmid transfected sgRNAs and the indel creation efficiency of each. sgRNA-labeled B and E were packaged into ShH10 serotype AAV vectors and co-cultured with mouse B6-RPE cells individually and in combination. (F and G) T7 endonuclease 1 assay (F) and quantitative PCR (G) for mAqp1 were performed. Aqp1 expression was significantly reduced using sgRNA E alone and in combination with B (MIX) compared with uninfected (UN) controls. Kruskal-Wallis test with Dunn’s multiple comparisons. Mean ± SD is shown. ***p = 0.001, ****p < 0.001. n = 10–12. Lamin B1 was used as loading control.

Figure 3

CRISPR-Cas9-Mediated Disruption of Ciliary Body Aquaporin 1 Lowers Intraocular Pressure in the Mouse Intravitreal injection of 2 × 10¹⁰ genome copies of the ShH10 virus encoding an equal proportion of mAqp1 B and E sgRNA (MIX) was performed into one eye of each wild-type C57BL/6J mouse. After 3 weeks, (A) a representative T7 Endonuclease 1 assay demonstrates genomic DNA mutation in ciliary body dissected from treated eyes, but not in uninjected eyes (UN). (B) SaCas9 DNA is detectable by PCR only in ciliary body tissue from MIX eyes. (C) IOP is reduced by mAqp1 disruption by a mean of 2.9 mmHg; paired t test, n = 18 pairs. (D) IOP is not altered by control ShH10 CMV-GFP virus injection, one-way ANOVA with Holm-Sidak’s multiple comparison, three independent experiments, n = 50 eyes. (E and F) Representative western blot (E) and densitometry (F) showing reduced AQP1 protein in isolated ciliary body tissue; paired t test, n = 7 pairs. (G) Representative H&E-stained paraffin sections of ciliary body show no clear structural abnormalities; n = 6, representative images shown with ×2 original magnification inset. (H) Representative optical coherence tomography (OCT) scans of treated and control eyes. (I and J) No significant “off-target” increase in thickness is seen for either (I) cornea or (J) retina, paired t test, n = 9 pairs. All scale bars: 50 μm. Mean ± SD is shown. **p = 0.01, ****p < 0.001. ns, not significant.

Figure 4

Ciliary Body Aquaporin 1 Disruption Lowers Intraocular Pressure in Two Experimental Glaucoma Models and Prevents Ganglion Cell Loss Using a corticosteroid-induced ocular hypertension model, paired eyes were subsequently injected with ShH10-CMV-SaCas9-sgRNA B and E (MIX) or untreated (UN). (A) Intraocular pressure (IOP) 3 weeks later is reduced in treated eyes by a mean of 2.88 mmHg; paired t test, n = 11 from two independent experiments. Dotted line is 11.3 mmHg mean IOP before model induction. (B and C) Representative ex vivo ciliary body western blot (B) and pooled data (C) demonstrating reduced mAQP1 protein levels; paired t test, n = 11. The more acute microbead ocular hypertension model was employed with data shown pooled from three independent experiments of three to five mice per run. (D) Eyes were treated 1 week after microbead injection, which attenuated the increase in IOP. Two-way ANOVA, p = 0.0003, n = 12. (E) After 3 weeks post-virus injection, mean IOP reduction was 3.9 mmHg; paired t test, n = 12. Dotted line represents 12.7 mmHg mean baseline IOP. (F) Matched ex vivo ciliary body mAQP1 protein was reduced in treated eyes; paired t test, p = 0.0008, n = 12. Extending to 7 weeks, confirmed reduced ganglion cell loss in the treated group. (G and H) Representative examples of retinal flatmount staining for Brn3a (G), with ganglion cell quantification (H), a reference mean Brn3α⁺ ganglion cell count in wild-type (WT) untreated retina by flatmount, also provided; mean of six fields per eye quantified as mean ± SD per mm², two independent experiments; paired t test, n = 9 pairs. Mean ± SD is shown. Scale bars, 50 μm. **p < 0.01, ***p < 0.001, ****p < 0.0001.

Figure 5

Human Ciliary Body Expresses Aquaporin 1 and Can Be Targeted by ShH10 Vector to Permit CRISPR-Cas9-Mediated Gene Disruption Human ex vivo ciliary body from post-mortem donors was obtained and could be maintained in culture up to 7 days. (A) Representative western blot for aquaporin 1 (hAQP1) protein from an eye undergoing immediate dissection. (B) Pooled qPCR expression data of available tissue from several donors, n = 6 (only one globe contained cornea). AQP1 was enriched in ciliary body and corneal endothelium. Human ciliary body was placed into immediate culture with ShH10 virus expressing GFP under the control of the ubiquitous CMV promoter. (C) By 72 h, GFP expression was detected in the ciliary body epithelium above autofluorescence using live fluorescence microscopy. Identical exposure times were used. (D) At day seven of culture, GFP can be seen in the outer non-pigmented ciliary epithelium using confocal microscopy histological sections. Representative example from four independent cultures. Three human sgRNAs were generated targeting exon 1 of hAQP1 and (E) tested in 293T cells by plasmid transfection and T7 endonuclease 1 assay. (F) sgRNA K was selected and packaged into an ShH10 vector. Addition to 293T cells for 72 h produced detectable indel formation by T7 endonuclease 1 assay. (G) Co-culture of the same vector with ex vivo human ciliary body led to low but detectable indel formation. Mean ± SD is shown. Scale bars, 400 μm.