Intrastromal Gene Therapy Prevents and Reverses Advanced Corneal Clouding in a Canine Model of Mucopolysaccharidosis I

Abstract

Mucopolysaccharidosis type I (MPS I) is an autosomal recessive lysosomal storage disease characterized by severe phenotypes, including corneal clouding. MPS I is caused by mutations in alpha-l-iduronidase (IDUA), a ubiquitous enzyme that catalyzes the hydrolysis of glycosaminoglycans. Currently, no treatment exists to address MPS I corneal clouding other than corneal transplantation, which is complicated by a high risk for rejection. Investigation of an adeno-associated virus (AAV) IDUA gene addition strategy targeting the corneal stroma addresses this deficiency. In MPS I canines with early or advanced corneal disease, a single intrastromal AAV8G9-IDUA injection was well tolerated at all administered doses. The eyes with advanced disease demonstrated resolution of corneal clouding as early as 1 week post-injection, followed by sustained corneal transparency until the experimental endpoint of 25 weeks. AAV8G9-IDUA injection in the MPS I canine eye with early corneal disease prevented the development of advanced corneal changes while restoring clarity. Biodistribution studies demonstrated vector genomes in ocular compartments other than the cornea and in some systemic organs; however, a capsid antibody response was detected in only the highest dosed subject. Collectively, the results suggest that intrastromal AAV8G9-IDUA therapy prevents and reverses visual impairment associated with MPS I corneal clouding.

Graphical Abstract

A new gene therapy approach wherein AAV vectors are deposited into the cornea prevented and reversed corneal clouding in MPS I canines. The results support a promising treatment for restoring clear vision in MPS I patients and a strategy to treat other lysosomal storage disorders with corneal clouding resulting in compromised vision.

Introduction

Mucopolysaccharidosis type I (MPS I) is a rare autosomal recessive metabolic storage disease caused by a deficiency of alpha-l-iduronidase (IDUA), a ubiquitous enzyme that breaks down intra-cellular and extra-cellular glycosaminoglycans (GAGs). In the absence of IDUA, GAG accumulation results in enlarged cells and tissues manifesting as a progressive multisystem disease with severe mental retardation, ischemic cardiac disease, and usually death in childhood. In addition, MPS I ocular abnormalities include corneal clouding, retinopathy, optic nerve damage, and glaucoma. Of these, corneal clouding is responsible for the highest incidence of impaired vision associated with MPS I,¹ which is evident during the first year of life. For instance, 79% of patients with Hurler’s syndrome, which is the most severe form of MPS I, have impaired vision. Further, all patients studied with ages ranging from 5 months to 23 years demonstrated varying degrees of corneal clouding with a direct correlation between age and severity.²

MPS I corneal disease has been characterized as a disease of the corneal stroma.^3,4 Lacking the ability to break down heparin and dermatan sulfates, the MPS I corneal stroma accumulates GAGs within the cell and in the extracellular matrix, thereby disrupting the precise collagen fibril size and alignment. The GAG-induced fibril disorganization, in addition to increased corneal fibrosis and myofibroblast proliferation (as indicated by increased α-smooth muscle actin staining), results in corneal clouding and reduced vision.3, 4, 5 MPS I retinopathy and optic nerve edema can be improved by bone marrow or hematopoietic stem cell transplantation (HSCT).⁶ However, there is no effective treatment for MPS I corneal clouding because the high incidence of cornea transplant rejection in MPS I patients discourages this practice as a standard of care.⁷

In a previous work, a human codon-optimized IDUA expression cassette was described as capable of restoring IDUA, both normal and supraphysiological levels, to primary MPS I patient fibroblasts.⁸ Furthermore, a rationally designed adeno-associated virus (AAV) capsid, AAV8G9, was deemed most efficient for human corneal gene delivery, compared with the natural AAV serotypes, following corneal intrastromal injection.⁸ Corneal intrastromal administration of AAV8G9 vectors in wild-type (WT) rabbits culminated in expansive transduction with transgene expression found in all major corneal layers spanning the epithelium, stroma, and endothelium.⁹ Although the human ex vivo and WT rabbit experiments optimized corneal gene delivery via AAV8G9-cytomegalovirus (CMV)-human IDUA (AAV-IDUA) and demonstrated no toxicity in multiple contexts,⁸ its ability to prevent or reverse MPS I corneal disease remained unknown. Therefore, the purpose of the study herein was to determine the efficacy of AAV-IDUA gene addition in the most relevant naturally occurring large-animal model, the MPS I canine.^2,10, 11, 12, 13, 14 The data obtained from the canine MPS I model demonstrate that: (1) intrastromal AAV vector injections in MPS I corneas were well tolerated, (2) intrastromal AAV-IDUA reversed MPS I corneal clouding in advanced disease, and (3) intrastromal AAV-IDUA prevented progression of early MPS I corneal disease while reversing the early changes. Remarkably, at all tested doses, phenotypic recovery was apparent within weeks following a single vector injection and was largely sustained to the study endpoint, 25 weeks post-injection. Other than a delayed-onset corneal edema, which was inconsistent and transient, no complications were noted, and biodistribution studies demonstrated that although vector genomes were detected in extraocular tissues, only the highest administered dose resulted in a capsid neutralizing antibody response. The collective efficacy and safety data of intrastromal AAV-IDUA in MPS I canine corneas warrant the continued examination of this technology as a potential single-dose cure to restore vision in MPS I patients affected with corneal clouding.

Results

To investigate the therapeutic efficacy and tolerability of AAV-IDUA gene therapy for MPS I corneal disease, we employed a naturally occurring canine model (IDUA^−/−).10, 11, 12, 13 Four MPS I-affected animals were included in the study; one younger animal (dog 1, age 9 months) had early disease, whereas the older three littermates (dogs 2–4, age 11 months) had advanced corneal disease. In the early disease, changes were limited to peripheral corneal neovascularization and diffuse stromal granularity consistent with GAG accumulation (Figure S1). The advanced corneal disease was characterized by neovascularization and diffuse thickening (Figure S1), where the former is a unique feature observed in advanced MPS I canine corneas, unlike the avascular MPS I human cornea.¹ Corneal clouding in all cases obscured the details of the anterior segment and reduced the fundus light reflexes. Ophthalmic examination prior to corneal injection determined intact corneal epithelium, normal tear production, and normal intraocular pressure in all dogs.

Single-strand AAV-IDUA vectors, containing a previously reported codon-optimized human IDUA cDNA,⁸ were injected into the canine corneal stroma with increasing volume (50–80 μL) in different animals, resulting in escalating viral dose ranging from 5e10 to 8e10 viral genomes (vg) per cornea (Table 1). In these experiments, the contralateral eye injected with AAV vectors encoding the green fluorescent protein (GFP) served as a control (Table 1).

Table 1.

Summary of the MPS I Animals, AAV Vector Injections, and Clinical Outcome

	Animal ID
	2		3		4		1
Animal research ID	I-709		I-710		I-712		I-744
Sex	F		M		M		F
Eye	R	L	R	L	R	L	R	L
Injection
Vector	AAV-GFP	AAV-IDUA	AAV-IDUA	AAV-GFP	AAV-GFP	AAV-IDUA	AAV-IDUA	AAV-GFP
Volume injected intrastromally (μL)	80	80	50	50	65	65	65	0
Viral dose (vg/cornea)	8 × 1010	8 × 1010	5 × 1010	5 × 1010	6.5 × 1010	6.5 × 1010	6.5 × 1010	5 × 1010
Intrastromal injection procedure	uneventful	uneventful	uneventful	uneventful	uneventful	additional intrastromal needle insertion	AC needle entry; additional intrastromal needle insertion	uneventful
Intracorneal hemorrhagea	+	+++	++	−	+	+++	−	−
Pre-injection
Age (months)	13		13		13		9
Corneal disease stage	advanced		advanced		advanced		early
Clinical scoreb	4	4	4	4	4	4	1	1
CCT (μm)	782	828	725	720	790	795	479	522
Endpoint (25 weeks post-injection)
Age (months)	19		19		19		15
Clinical scoreb	2c	1	0	4	4	0	1	3
CCT (μm)	570	629	493	867	802	543	465	583
Capsid antibody titer	1:64		−		−		−
Corneal Edema
Onset (weeks post-injection)	10	6	−	19	−	−	17	−
Duration (weeks)	3	3	−	3	−	−	2	−
Steroid treatment	topical	topical + systemic	−	topical	−	−	topical	−

AAV-GFP, AAV8G9-CMV-GFP; AAV-IDUA, AAV8G9-CMV-IDUA; AC, anterior chamber.

^aIntracorneal hemorrhage score: −, no hemorrhage; +, a small single focus of hemorrhage; ++, small multifocal hemorrhages; +++, large area of coalesced multifocal hemorrhages.

^bCornea clinical score: 0 = clear, storage (−); 1 = storage (+), minimal vascularization; 2 = storage (+), midrange vascularization; 3 = storage (+), diffuse vascularization; 4 = storage (+), diffuse vascularization, crystalline deposits, fibrosis.

^cReduced perfusion of diffuse vasculature.

To determine whether AAV-IDUA can prevent the progression into advanced corneal disease, the younger MPS I canine with early corneal disease received a single intrastromal injection in each cornea (5e10 vg; Table 1). The deposition of the vector solution resulted in immediate opacity and increased thickness of the cornea consistent with focal edema (Figure S2), which is common to all clinical corneal intrastromal injections. Clarity and central corneal thickness (CCT) returned to pre-injection values within 24 h as observed clinically and by ultrasound biomicroscopy (Figure S2). Within 1–2 weeks of the injection, the AAV-IDUA-treated cornea regained clarity with loss of vascular perfusion that remained clear until the end of the study at 25 weeks (Figures 1A and 1C, animal 1; Table 1, animal 1).

Figure 1.

Restoration of Corneal Clarity following Intrastromal AAV-IDUA Injection in the MPS I Canine Cornea

(A) Corneal images prior to AAV injection and at the indicated experimental conclusion for all MPS I dogs with early (dog 1) and advanced (dogs 2–4) corneal diseases. (B) Long-term restoration of near-normal central corneal thickness (CCT) in the AAV-IDUA-injected cornea in dog 4. Data are presented as mean ± SD. (C) Clinical corneal vascularization score at pre-injection at the indicated time points post-injection. There is reduction in vascularization in all three corneas injected with AAV-IDUA at advanced disease (dogs 2–4), sustained throughout the follow-up period. Dog 1 that was treated at early disease developed minimal vascularization in the AAV-IDUA-injected cornea, whereas the AAV-GFP-injected cornea vascularized over time. (D) Retroillumination of the tapetal fundus in dog 3 at 7 weeks post-injection. Upon shining light on the animal’s face, there is a normal bright reflection from the tapetum lucidum of the right AAV-IDUA-injected eye, whereas the left AAV-GFP-injected eye has a dull and dim reflection. Indirect ophthalmoscopy reveals a clear view of the ocular fundus with crisp details of the optic disc and retinal vasculature in the right fundus. The left fundus appears hazy, consistent with the presence of MPS I corneal disease.

Next, investigations were designed to determine whether AAV-IDUA can reverse diffuse pre-existing corneal clouding. Again, the AAV vector injections were well tolerated in the corneas of all three older subjects (subjects 2–4), and focal corneal edemas associated with the injection resolved to pre-injection values over a 22- to 48-h period (Figure S2). The only complication was mild bilateral blepharospasm upon recovery from general anesthesia with resolution within 24 h in the dog that received the highest injection volume per dose (animal 2; Table 1). Focal to multifocal intracorneal hemorrhages, indicative of rupture of pre-existing vasculature (Figure 1A, animal 2), developed in all but one cornea with advanced disease with subsequent resorption within ∼2 weeks. In all AAV-IDUA-treated corneas with advanced disease, resolution of MPS I-associated corneal clouding, as well as reduced vascular perfusion, became evident as early as day 7 post-injection (Figures 1 and S3). There was progressive reversal of corneal storage disease with maximum treatment effect achieved by 6 weeks post-injection unless transiently interrupted by the confounding events described below (Figures 1 and S3; Table 1). Clinical examination scores were improved in all AAV-IDUA-treated eyes (Table 1; Figure 1), with normalized CCT significantly reduced from pre-injection values (p < 0.0001) and from the contralateral control eye (p < 0.0001; Figure 1B; Table 1). At the experimental endpoint, CCT in the AAV-IDUA-treated eyes was not statistically different from the CCT of WT research dogs (560 ± 36 μM) of the same colony (Table 1). Clinical vascularization analysis demonstrated significant reduction and near absence of vascular perfusion in all AAV-IDUA-injected corneas (Figure 1C) with improved visualization of the intraocular structures (Figure 1D). Striking phenotypic reversal was observed in all AAV-IDUA-treated corneas for the entire study period of 25 weeks, until the humane endpoint (Figures 1 and S3; Table 1).

Between 6 and 19 weeks post-injection, corneal edema acutely developed in two AAV-IDUA- and two AAV-GFP-treated eyes of three dogs (Table 1; Figure S4). The edema typically started in the axial to paraxial region, often progressing to encompass the entire cornea (Figure S4). The clinical signs were transient and resolved over a 2- to 4-week period with use of topical corticosteroid, and in one dog with concurrent use of systemic corticosteroids (Table 1). Notably, the two eyes that developed the earliest incidences of corneal edema (6 and 10 weeks) had been injected with the highest volume (80 μL) and viral dose (8e10 vg) of AAV-IDUA or AAV-GFP (Table 1). Once the delayed-onset corneal edema was resolved, of the four corneas that were affected, the AAV-IDUA-injected corneas (N = 2) returned to the reversed phenotype with restored clarity and reduced vascular perfusion (Figures 1, S3, and S4).

Histological analysis relied on a previously described scoring scheme^9,15 that grades the degree of cellular infiltration, extent of fibrosis, and degree of neovascularization. All AAV-IDUA-treated corneas had reduced corneal thickness, neovascularization, and fibrosis (Figure 2A). Alcian blue staining to detect acidic polysaccharides as a functional indication of IDUA therapy was substantially reduced in all AAV-IDUA-treated corneas to levels similar to the asymptomatic heterozygous IDUA^+/− MPS I canine (Figure 2B).

Figure 2.

Postmortem Histological Analysis

Hematoxylin and eosin histology representative images of advanced disease MPS I corneas from a single subject treated with the indicated vector (A). The average histopathology scores of all patient corneas treated with indicated AAV vector were assessed by a blinded ocular toxicologist as described in the text (N = 4). Data are presented as mean ± SD. (B) Alcian blue staining of representative corneas treated in the indicated manner and blue pixel quantitation by ImageJ (N = 4). Data are presented as mean ± SD. p = 0.0215.

Transgene product staining demonstrated IDUA or GFP in strong abundance in all layers of corneas injected with AAV-IDUA or AAV-GFP, respectively (Figure 3). Compared with the non-pathological IDUA^+/− cornea, AAV-IDUA injected to MPS I (IDUA^−/−) corneas elicited supraphysiological IDUA levels at all administered doses (Figure 3A). Immunostaining of α-smooth muscle actin (αSMA), a common marker of myofibroblasts and hence fibrosis, revealed strong signals in all of the control (AAV-GFP) corneas, whereas there was marked reduction in αSMA⁺ cells in all AAV-IDUA-treated corneas (Figure 3B).

Figure 3.

Transgene Product and α-Smooth Muscle Actin Staining following AAV Vector Corneal Intrastromal Injection

Vector expression and distribution in the cornea were verified by immunofluorescence for IDUA or GFP protein (A). α-Smooth muscle actin staining was performed in the indicated canine backgrounds receiving the labeled AAV vector (B). The numbers represent each patient. Endo, endothelium.

Biodistribution outside the cornea detected that the highest amount of vector genomes was in the neighboring conjunctiva, with lower levels in the iris and retina (Table S1). Other tissues, including the liver, heart, and kidney, also tested positive for recombinant AAV genomes (Table S1). Despite this, only the highest dosed animal developed capsid neutralizing antibodies (1:64) at 25 weeks post-injection (Table 1).

Discussion

Corneal clouding resulting in impaired vision is commonly observed in lysosomal storage diseases, including MPS I, MPS IV, MPS VI, MPS VII, and cystinosis.^1,16 High-risk corneal transplantation is a therapeutic option for the impaired vision in these MPS patients; however, the high degree of transplant rejection discourages this procedure as a standard of care.¹⁷ In a previous work, a chimeric AAV capsid (AAV8G9) was validated for enhanced transduction in human corneas ex vivo.⁸ Following AAV-IDUA intrastromal injections, supraphysiological levels of α-l-iduronidase activity were observed with no associated toxicity.⁸ The study herein extends these data by confirming the hypothesis that AAV-IDUA injected into the MPS I canine cornea can prevent progression of corneal clouding and, beyond our expectations, completely reverse existing disease (Figure 1A, animal 2; Table 1). Strikingly, evidence of storage disease clearance was observed as early as 7 days post-injection and was largely sustained to 25 weeks in all the MPS I corneas injected with AAV-IDUA (Figure 1; Table 1; Figures S3 and S4A).

Metrics of ocular disease, including clinical exam scores, CCT, and corneal neovascularization (an indirect measure of canine MPS I pathology), were remarkably reduced and sustained over the course of the study, with transient interruption by the edematous incidence in a subset of corneas (discussed below; Table 1; Figures S4 and S5). Histological examinations support the clinical observations with markedly reduced cellular infiltrates, fibrosis, neovascularization, and Alcian blue staining (Figure 2) in all eyes administered AAV-IDUA therapy. Importantly, the therapeutic effect was correlated to the presence of IDUA in the corneas administered AAV-IDUA, which was observed at supraphysiological levels compared with the clinically normal IDUA^+/− control canine cornea (Figures 1 and 3). Unexpectedly, both GFP and IDUA were expressed persistently in the corneal epithelium, a dividing cell population predicted to lose episomal vector genomes during cell division. It is possible that extracellular particles in the cornea continually transduce the epithelia over time, and/or AAV vector genomes integrated in limbal stem cells, which are thought to produce/maintain the central basal epithelia; however, this was not observed in human cornea experiments following intrastromal injections ex vivo.^18,19 Our histological findings also revealed a direct correlation between reduced αSMA staining and IDUA therapy, consistent with a report suggesting that fibrosis, and not solely GAG accumulation, contributes to MPS I corneal clouding⁵ (Figure 2B).

All observed incidences of the delayed-onset corneal edema characterized by increased CCT occurred 6 weeks post-injection or later (Figures S4 and S5; Table 1). The rapid phenotypic reversal of storage clearance as early as 7 days post-injection indicates α-l-iduronidase production and function during the first week following AAV-IDUA injection. Early disease correction, along with the known kinetics of AAV transgene product-specific T cell responses in other tissues,²⁰ supports the reasoning that either GFP and/or IDUA is causative for the delayed-onset corneal edema. Also, the immunogen(s) leading to corneal edema could be a combination of both transgene-derived proteins and the AAV capsid. However, the high abundance of GFP or IDUA in the post-edematous eyes at experimental completion suggests the corneal edema was either not a cytotoxic T lymphocyte response to the foreign proteins or, alternatively, resulted in an incomplete clearance of the transduced cells. It remains a formal possibility that the delayed-onset edema could be resultant of the loss of ocular immune privilege prior to vector administration because MPS I canine cornea displayed vasculature that was ruptured during the injection (hence the transient intrastromal hemorrhage; Figures 1A, animal 2, and S3). Therefore, the development of transient corneal edema observed herein may not be relevant to human MPS I, where corneas are avascular and, additionally, the HSCT-treated patients are tolerized to the vector-encoded transgene product, IDUA.

Biodistribution experiments herein demonstrate that AAV vector genomes were found in other ocular tissues, as well as in liver, kidney, and heart in all dogs treated (Table S1). Because all dogs presented corneal neovascularization at the time of injection (Figures 1, S1, and S3), it is likely that AAV vectors, in part, disseminated via hematogenous. This would not be expected to occur in the avascular MPS I patient corneas following precise intrastromal vector deposition. In further support of this hematogenous hypothesis are unpublished data from rigorous toxicity experiments performed in WT rabbits. In those studies, AAV-IDUA vector genomes administered by an intrastromal injection to the avascular rabbit cornea failed to detect vector genomes outside of the ocular compartment, even at a higher dose than used herein (data not shown). Only dog 2, which received the highest dose and volume, developed neutralizing antibodies against the AAV8G9 capsid, consistent with the hypothesis that a threshold amount of AAV vectors in systemic circulation must be achieved to elicit a capsid antibody response (Table 1). Based on our observations, the lowest volume tested herein restored excess IDUA compared with the clinically normal MPS I canine carrier (IDUA^+/−). It is likely that lower vector doses would prove similarly effective while minimizing concerns of systemic vector distribution, the AAV capsid antibody response, and the delayed-onset corneal edema observed in a subset of canine corneas treated herein.

Collectively, the data demonstrate the potential benefit of ocular anterior segment gene therapy for addressing the vision deficits associated with MPS I and other lysosomal storage diseases.

Materials and Methods

Animals

The MPS I canine research colony was maintained under NIH and US Department of Agriculture (USDA) guidelines for the care and use of animals in research, and all of the study protocols were approved by the University of Pennsylvania Institutional Animal Care and Use Committee (IACUC protocol #805221). Four MPS I dogs were included in the study. The IDUA genotype was confirmed by detection of homozygosity for the known mutation¹³ by PCR. Peripheral blood was collected prior to AAV injection and 6 and 25 weeks post-injection. Tear and cerebral spinal fluid (CSF) (1–2 mL) samples were collected 1 week before the endpoint of the study. At 25 weeks post-AAV injection, animals were humanely euthanized by administration of sodium pentobarbital (80 mg/kg intravenously [i.v.]). Tissues were collected and stored at −80°C until assessment.

Intrastromal Vector Injection

These experiments relied on previously described codon-optimized human IDUA cDNA⁸ (GenBank: NM_000203, 1,962 nucleotides) that generates a protein with 79% identity to canine alpha-L-iduronidase (IDUA). Single-strand AAV-IDUA and self-complementary AAV8G9-CMV-GFP (AAV-GFP; control) viral constructs were prepared by the University of North Carolina (UNC) Viral Vector Core as described previously⁸ and titered by qPCR and silver staining. The AAV-IDUA vector was injected into the corneal stroma of one eye, and AAV-GFP was injected in the contralateral eye. The identity of the two viral constructs was masked to the ophthalmic surgeon (K.M.). For injections, the dogs were placed under general anesthesia and positioned in dorsal recumbency, and the ocular surface was prepared with 1:50 iodine-povidone solution and sterile eye irrigation solution. A lid speculum and two stay sutures were placed to stabilize the globe.

The vector solution was loaded in an insulin syringe attached with a 28½G needle. Under an operating microscope, the tip of the microneedle was inserted into the cornea obliquely entering at one paraxial location and directed toward the axial cornea. The needle was inserted bevel up until the entire bevel reached the mid-stromal level. Following injection, the needle tip was held in place for approximately 1 min to minimize back-leakage from the injection tract. As the needle was removed, a cellulose sponge spear was used to apply slight pressure to the cornea overlying the injection site to further minimize leakage. In one cornea (animal 3, left eye), as the fluid was injected the needle tip was expelled from the cornea by counterpressure. In another cornea (animal 4, right eye), the needle tip inadvertently penetrated the Descemet’s and the endothelium, reaching the anterior chamber. In both incidences, a second entry was made into the cornea to inject any remaining vector solution.

Ophthalmic Examination and Imaging

Before and after injections, ophthalmic examinations were performed by a board-certified veterinary ophthalmologist (K.M.) while the identity of the injected vectors remained masked. Examinations included slit-lamp biomicroscopy (SL-17; Kowa, Tokyo, Japan), tonometry (TonoVet, Icare, Finland), and when possible, indirect ophthalmoscopy. Following vector injections, examination was performed at <1, 24, and 48 h, and at 1 and 2 weeks post-injection, then every 1–4 weeks until the endpoint of the study at 25 weeks post-injection. Ultrasound biomicroscopy (48 Hz, UBM Plus; Accutome, Malvern PA, USA) was used for in vivo transverse imaging of the cornea, as well as measurement of the CCT, and was performed pre-injection, at <1, 24, and 48 h post-injection, and periodically thereafter. A hand-held fundus camera (Genesis-D; Kowa, Tokyo, Japan) was used to photograph the en face corneal, facial, and ocular fundus images. Clinical vascularization score relied on the following scoring system: 0, no vascular perfusion; 1, minimal vascular perfusion, limited peripheral neovascularization; 2, reduced vascular perfusion; 3, diffuse neovascularization; and 4, diffuse neovascularization with increased perfusion.

Vector Biodistribution

At necropsy, ocular and systemic tissues were promptly dissected and frozen on dry ice for AAV vector biodistribution analysis. Small samples of brain tissues were dissected from the frontal, temporal, and occipital cortices. Samples were stored at −80°C until the time of analysis.

Vector biodistribution was examined as described previously.^19,21 In short, DNA was isolated from tissues using the QIAmp DNA Mini Kit (QIAGEN). Using the Roche Universal Probe Finder web-based software v.2.53, a unique probe site is the common CMV promoter 5′-ctgctcct-3′ for Probe #68, using an amplicon sequence defined as 5′-ggtggtggtgcaaatcaaagaactgctcctcagtggatgttgcctttacttctaggcctg-3′ (forward primer 5′-ggtggtggtgcaaatcaaa-3′ and reverse primer 5′-caacggaaatgaagatccggac-3′). The assay was carried out following kit instructions, and the crossing point signal was quantified using a plasmid as a standard curve. Results were normalized to total DNA detected in samples carried out by the NanoDrop 2000 DNA concentration function using the mean of three replicate measurements.

Histology and Immunohistochemistry

Cornea tissues were isolated post-mortem and preserved for 18–24 h with neutral-buffered formalin (NBF). After fixation, tissues were transferred to 70% ethanol and submitted to the Center for Gastrointestinal Biology and Disease (CGIBD) histology core facility at UNC for routine paraffin embedding, sectioning, hematoxylin and eosin (H&E) staining, and Alcian blue staining. For immunofluorescence, 5-μm sections of embedded cornea tissues underwent deparaffinization by way of two incubations in xylene, 10 min each, and rehydration by sequentially incubating tissue sections in ethanol solutions that followed a standard gradient of decreasing ethanol concentrations. Tissue sections then underwent a heat-induced epitope retrieval (HIER) procedure at 98°C for 15–20 min to ensure epitope unmasking of all proteins cross-linked during the paraffinization procedure. Specifically, sections stained for IDUA (orb157615, 1:50; Biorbyt), GFP (gfp-1020, dilution 1:500; Aves), and α-smooth muscle actin (NBP1-30894, 1:100; Novus) were immersed into pre-heated citrate-based antigen retrieval solution (pH 6.0; H-3300; Vector Laboratories). Slides were allowed to cool at room temperature for 20 min before being washed with Tris-buffered saline (TBS) and gently permeabilized with TBS plus 0.02% Tween 20 with gentle agitation. Non-specific binding sites within the tissues were blocked by incubation with 10% normal goat serum plus 1% BSA in TBS for 30 min at room temperature (RT). After blocking solution was removed, the tissues were then incubated overnight at 4°C with primary antibody diluted appropriately in TBS with 1% BSA. Following incubation, sections were washed two times, 5 min each, with TBS plus 0.02% Tween 20 with gentle agitation to remove all non-specifically bound primary antibody. A suitable fluorophore-conjugated secondary antibody was diluted according to the manufacturer’s suggestions in TBS plus 1% BSA, added to each tissue section, and incubated for 2 h at room temperature with gentle agitation. Finally, after slides were washed three times with TBS, 10 min each, coverslips were mounted, and the nuclei within the tissues were stained with ProLong Diamond Antifade Mountant with DAPI (P36971; Molecular Probes). All histological images were taken using an Olympus IX83 research inverted microscope with a 40× objective (Olympus, Tokyo, Japan) and processed using Olympus Life Science cellSans imaging software.

A qualitative histological corneal score was assigned to each slide with the examiner blinded to the treatment groups. The corneal histological scoring scheme was used as previously described^9,15 and graded the following areas: degree of cellular infiltrate, extent of fibrosis, and extent of neovascularization. Scores of 0–4 were assigned based on the following rubric: 0, no significant lesion; 1, low numbers of inflammatory cells, mild fibrosis, minimal neovascularization; 2, moderate, diffuse stromal inflammatory cell infiltrates, fibrosis, or neovascularization; 3, moderate to marked diffuse stromal inflammatory cell infiltrate, fibrosis, or neovascularization; and 4, marked, diffuse, cellular infiltrate, fibrosis, vascularization. Final histological score was the summation of infiltrate, fibrosis, and neovascularization scores.

Neutralizing Antibody Assay

Whole blood serum used for this assay was collected from dogs prior to AAV injection, 6 months after treatment, and stored at −80°C. HEK293 cells were plated at 25,000 cells/well in a 48-well plate with per well volume of 300 μL DMEM supplemented with 10% FBS and 1% penicillin-streptomycin (Pen Strep). Cells were cultured in a 5% CO₂ incubator set at 37°C 24 h prior to assay. Serum was then thawed, with 13 μL serum serially diluted with 13 μL PBS to obtain 1:1, 1:2, 1:4, 1:8, 1:16, 1:32, 1:64, and 1:128 dilutions in duplicate. 13 μL of scAAV8G9-Luciferase with a total 2.2e10 vg titer was added,⁸ mixed, and incubated with the serum dilutions for 2 h at 4°C, and each dilution replicate was added to a separate well. Total protein lysate was collected 48 h after treatment with Promega Lysis Buffer. Upon addition of luciferin, light emission was detected using the Perkins Mark II plate reader.

This experiment was repeated on three separate occasions.

Author Contributions

M.L.H., R.J.S., J.K., and K.M. conceived and were responsible for the experimental design. K.M., M.L.H., T.A.L., K.C., P.O., J.B., B.G., L.C., L.S., R.J.S., and J.K. contributed to experimental execution and/or data analysis. K.M., M.L.H., L.S., B.G., and T.A.L. contributed to manuscript writing/editing. M.L.H. and R.J.S. provided funding for the experiments.

Conflicts of Interest

R.J.S. and M.L.H. are co-inventors of a US patent regarding optIDUA as a therapy for MPS I ocular diseases. M.L.H. has also received royalties from Asklepios BioPharmaceutical related to patent #9447433. M.L.H. and B.G. are co-founders and shareholders of Bedrock Therapeutics, which holds an unrelated license from UNC. R.J.S. is the founder and a shareholder at Asklepios BioPharmaceutical and Bamboo Therapeutics, Inc. R.J.S. holds patents that have been licensed from UNC to Asklepios Biopharmaceutical, for which he receives royalties. R.J.S. has consulted for Baxter Healthcare and has received payment for speaking.