AAV Gene Therapy for MPS1-associated Corneal Blindness

Abstract

Although cord blood transplantation has significantly extended the lifespan of mucopolysaccharidosis type 1 (MPS1) patients, over 95% manifest cornea clouding with about 50% progressing to blindness. As corneal transplants are met with high rejection rates in MPS1 children, there remains no treatment to prevent blindness or restore vision in MPS1 children. Since MPS1 is caused by mutations in idua, which encodes alpha-L-iduronidase, a gene addition strategy to prevent, and potentially reverse, MPS1-associated corneal blindness was investigated. Initially, a codon optimized idua cDNA expression cassette (opt-IDUA) was validated for IDUA production and function following adeno-associated virus (AAV) vector transduction of MPS1 patient fibroblasts. Then, an AAV serotype evaluation in human cornea explants identified an AAV8 and 9 chimeric capsid (8G9) as most efficient for transduction. AAV8G9-opt-IDUA administered to human corneas via intrastromal injection demonstrated widespread transduction, which included cells that naturally produce IDUA, and resulted in a >10-fold supraphysiological increase in IDUA activity. No significant apoptosis related to AAV vectors or IDUA was observed under any conditions in both human corneas and MPS1 patient fibroblasts. The collective preclinical data demonstrate safe and efficient IDUA delivery to human corneas, which may prevent and potentially reverse MPS1-associated cornea blindness.

Figure 1. Restoration of IDUA activity in MPS1 patient fibroblasts by AAV gene therapy.

(A) Schematic diagram of the AAV-IDUA optimized vector construct. (B) Cell lysates (Left panel) and supernatants (Right panel) from AAV2 infected and not-infected fibroblasts, normal human fibroblasts (NHF) or MPS1 fibroblasts, were analyzed for IDUA protein expression. Detection of β-actin was performed as a loading control of total protein in cell lysates. All the experiments were performed in triplicate. (C) Functional activity of IDUA protein was obtained for cell lysates and supernatants from AAV2 infected and not-infected fibroblasts, NHF or MPS1 fibroblasts. For cell lysates, the nmoles of 4-MU were normalized to one hour reaction and mg total protein. For cell supernatants, the nmoles of 5-MU were normalized to 10 μl of the sample. All the experiments were performed in triplicate. 4-MU, 4-methylumbelliferone; CMV, cytomegalovirus promoter; ITR, inverted terminal repeats; GFP, Green fluorescence protein.

Figure 2. AAV capsid serotype evaluation in human cornea.

(A) Human corneas were injected with a self-complementary CMV-GFP cassette encapsidated in AAV serotypes 8, 9, or the 8/9 chimeric 8G9. Seven days later Western blot was used to detect GFP. PBS corresponds to a human cornea injected with only PBS (vehicle control). Detection of β-actin was performed as a loading control. (B) Human cornea pieces were incubated with 1 × 10¹⁰ viral genomes of each serotype: 8, 9, or the chimeric 8G9, but as single stranded viruses. Seven days following viral infection, the tissues were collected and the protein lysates were obtained for quantification of GFP. Human corneas incubated with the medium but replacing the viral addition for PBS were used as negative controls (PBS). Absorbance measurements at 450nm were normalized to total protein concentration. (*corresponds to p ≤ 0.05 when performing T-tests comparing normalized absorbance 450nm values for AAV8G9 with AAV8 or AAV9; N = 5). (C) Single-strand AAV8G9-CMV-GFP was injected into the stroma of human corneas and harvested for histology 7 days later. Left- Immunofluorescence image showing the distribution of recombinant AAV8G9-GFP transduction across a human cornea section. Human corneas injected with PBS served as the negative control. Images were obtained with 10× objective and assembled by stitch processing. The scale bar is equal to 2000 μm. Right- Different areas of the same stained human cornea taken with 20× objective. The scale bar is equal to 100 μm. Hoechst was used for nuclei counterstain. (D) A representative section of the corneas in (C) stained with GFP (green) and the indicated cell marker (red). The scale bar is equal to 10 μm. Hoechst was used for nuclei counterstain.

Figure 3. Restoration of IDUA activity in human corneas by AAV8G9-opt-IDUA.

(A) A representative section of a normal non-injected cornea stained with IDUA antibody (red) and a cell marker (green). Hoechst was used for nuclei counterstain. The scale bar is equal to 10 μm. (B) Western blot detecting IDUA amounts produced after injection of AAV8G9-opt-IDUA. Administration of AAV8G9-GFP served as a vector infection control. B-actin serves as a loading control. (C) Functional activity of IDUA protein obtained from human corneas 7 days post-injection of AAV8G9-opt-IDUA. AAV8G9-GFP served as the negative control. The nmoles of 4-MU were normalized to one hour reaction and mg total protein. PBS corresponds to a human cornea injected with only the vehicle control. The experiment was performed in triplicate.

Figure 4. Distribution of IDUA protein following AAV-8G9-opt-IDUA transduction.

(A) Immunofluorescence image showing the distribution of IDUA protein across a human cornea section seven days post-injection of AAV8G9-opt-IDUA or PBS (vehicle). The images were obtained with 10× objective and assembled by stitch processing. The scale bar is equal to 2000 μm. Hoechst was used for nuclei counterstain. (B) Cross section of the same stained human corneas but taken with a 20× objective. The top figure corresponds to the negative control of staining, which was treated in the same manner as the other samples but without the primary antibody. Hoechst was used for nuclei counterstain. The scale bar is equal to 100 μm. The experiment was performed in duplicate.

Figure 5. No apoptosis following AAV8G9-opt-IDUA injection in human corneas.

(A) Human corneas injected with PBS, AAV8G9-GFP, or AAV8G9-opt-IDUA were processed 7 days later for Tunel staining. Top, images taken at 10×. Bottom, images taken at 20×. As a positive control, human corneas injected with PBS where nuclease treated. The scale bar is equal to 100 μm. (B) Quantitation of pixel area for total staining performed in (A). A minimum of 7 areas per cornea for each treatment were counted using SigmaScan Pro software. No statistical difference between cytotoxicity for corneas injected with AAV8G9-GFP or AAV8G9-opt-IDUA was observed (p > 0.05).