A gel-forming antioxidant eye drop for photoreceptor protection in retinitis pigmentosa

Abstract

Retinitis pigmentosa (RP) is a chronic genetic condition that leads to progressive loss of photoreceptor cells and vision. While gene therapy is available for a small subset of patients with specific mutations, developing a therapeutic that broadly targets the cellular stresses that lead to cell death could address a major unmet need. One such option would be utilizing antioxidant therapies to neutralize reaction oxygen species (ROS) in the retina that underlie RP progression. Here, we describe an approach for delivering the antioxidants N-acetyl cysteine (NAC) or N-acetyl cysteine ethyl ester (NACET) with a gel-forming eye drop previously demonstrated to provide therapeutic drug delivery in the posterior segment in animals. We demonstrated therapeutic protection of photoreceptor structure and function in a chemically-induced rat model of RP (48% increase in photopic b-wave amplitude), as well as some limited protection in an aggressive genetic mouse model (rd10) of retinal degeneration (~ 31% increase in photopic b-wave amplitude) with once daily application. However, antioxidants have inherent stability issues when stored in solution, so we investigated the use of additional excipients to improve stability and retain potency. While a promising approach, future work to address product stability and efficacy in larger eyes is needed for further development.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at 10.1007/s13346-026-02070-4.

Introduction

Retinitis pigmentosa (RP) is a genetic disorder affecting approximately 1 in 4,000 individuals that leads to loss of photoreceptors and decrease in visual acuity [1, 2]. RP is one of the most common causes of low-light blindness in individuals aged 20 to 60 [1]. Despite the prevalence of RP, there are few effective treatments, each with limitations. The only FDA approved therapy for RP is Luxturna® (Spark Therapeutics), a subretinal gene therapy to correct RPE65-induced retinal degeneration. While an important step forward, only a small fraction of RP patients in the United States and Europe have RPE65 mutations [3]. Unfortunately, with about 3,100 known causative mutations in over 50 different genes, each gene target would require a unique treatment [4]. Visual prosthetics may provide ambulatory vision for patients, but there has been limited advancement in these technologies following the discontinuation of the Argus II retinal prosthesis system [5]. A therapy that improves photoreceptor survival and function independent of genetic mutation could be a significant advancement in the clinical management of RP.

RP is characterized by the loss of rod and cone photoreceptors in the retina due to many compounding factors, including oxidative damage. Rod photoreceptors are highly specialized components of the nervous system responsible for processing light stimulus and are essential for scotopic (dim light) vision. Rods are also highly metabolically active, essential in maintaining tissue oxygen levels in the retina, and greatly outnumber cones ~ 95:5 [6]. When genetic mutations in RP lead to loss of rods, oxygen consumption in the retina decreases, leading to excess unconsumed oxygen in the retina, resulting in hyperoxia [7]. This leads to increase of superoxide radicals and consequently oxidative damage on cones [8]. It has also been shown that lowering hyperoxia in the retina after rod death, by placing the RP mice in a hypoxic chamber result in decreased oxidative damage on cones, and better cone vision [9]. An alternative approach could be to lower the oxidative damage by lowering the free radical produced as the rods die. Glutathione (GSH), a glutamine-cysteine-glycine tripeptide, is the body’s most abundant free radical acceptor. GSH can reduce quantities of ROS to modulate tissue homeostasis and balance cell death vs. cell proliferation pathways [10]. However, when this pathway is overwhelmed in the presence of excess retinal oxygen concentrations, GSH decreases, glutathione disulfide (GSSG) levels rise, and oxidative damage occurs [11].

To restore the redox balance between GSH/GSSG, oral supplementation of the cysteine and GSH precursor, N-acetyl cysteine (NAC), has shown efficacy in protecting photoreceptor survival and retinal function in both animal models and Phase 1 clinical trials when administered orally [12, 13]. The benefits of NAC supplementation for RP are two-fold, as NAC not only feeds into GSH biosynthesis pathways following de-acetylation to cysteine but is also able to directly neutralize ROS through its free thiol. However, since NAC suffers from poor oral bioavailability, large doses are required to provide a therapeutic benefit [13]. A multisite, randomized controlled Phase 3 trial is underway to confirm the therapeutic benefits of 1,800 mg of NAC taken twice daily [14]. While oral NAC would be the first mutation-independent therapy available for RP patients, additional options with less frequent and more targeted dosing would be advantageous to patients managing a life-long chronic disease.

We hypothesize that a local drug delivery strategy could provide more effective and targeted therapeutic effect with less frequent dosing. However, achieving therapeutic accumulation in the posterior segment following conventional eye drop has proven difficult and is often considered the “holy grail” of ocular drug delivery. Various anatomical barriers, including the cornea, tear film, and blinking initially limit intraocular accumulation in the anterior segment [15, 16]. For drug that is absorbed into the conjunctiva following topical application, only approximately 0.1% of the dose may reach the retina through transscleral diffusion after much of the drug is lost via systemic absorption through the conjunctival blood vessels [17–19]. To increase accumulation in the posterior segment after topical administration, we previously developed a novel hypotonic in situ gel-forming eye drop formulation [20]. The approach combines hypotonicity and thermosensitive gel-forming polymers (e.g., Pluronic F127) formulated below their critical gel concentration (CGC), allowing for rapid and uniform water absorption-driven spreading over the ocular surface. Equilibration of osmotic pressure across the ocular surface leads to the formation of a thin, uniform, drug-laden gel layer, providing sustained ocular drug absorption. Importantly, we have confirmed therapeutically relevant drug accumulation in the posterior segment in both rabbits and pigs with as few as five days of once daily dosing [20, 21]. Prior study has also identified N-acetyl cysteine ethyl ester (NACET), an esterified and lipophilic NAC prodrug, as a promising alternative for reducing intracellular ROS [22, 23]. Here, we describe studies aimed at combining the antioxidants NAC and NACET with our gel-forming eye drop as a potential treatment for RP. An eye drop formulation that provides sustained intraocular antioxidant delivery has the potential to address a critical unmet need in preventing the progression of RP, though future work is required to ensure product stability and confirm therapeutic drug delivery in larger species.

Materials and methods

Materials

ARPE-19 cells (Lot # 70,049,881) were purchased from ATCC (Manassas, VA). T75 flasks were purchased from Sarstedt (Nümbrecht, Germany). DMEM:F12 (1:1) ((+) pyruvic acid, (+) HEPES, (+) phenol red) was purchased from Thermo Fisher (Waltham, MA). Phosphate buffered solution (PBS) and trypsin with 2.21 mM EDTA were purchased from Corning (Manassas, VA). Penicillin–Streptomycin (P/S) was purchased from Quality Biological (Gaithersburg, MD). Fetal Bovine Serum (FBS) was purchased from Fisher Scientific (Hampton, NH). 10 × Dilution Buffer and DCFDA/H₂DCFDA Cellular ROS Assay Kit were purchased from Abcam (Cambridge, UK). Tert-Butyl hydroperoxide (TBHP), 2′−7′-Dichlorodihydrofluorescein diacetate (DCFDA), dimethyl sulfoxide (DMSO) for molecular biology, N-acetyl-L-cysteine (NAC), Pluronic F127, paraquat dichloride dihydrate, tannic acid, ethylenediaminetetraacetic acid (EDTA), ascorbic acid/sodium-L-ascorbate, sodium hydroxide (NaOH), hydrochloric acid (HCl), Mayer’s hematoxylin, and Eosin Y alcoholic solution were purchased from Sigma Aldrich (St. Louis, MO). N-acetyl-L-cysteine ethyl ester (NACET) was purchased from Cayman Chemicals (Ann Arbor, MI). N,N’-Diacetyl-L-cystine (diNAC) was purchased from Targetmol (Wellesley Hill, MA). 96 well black with clear flat bottom tissue culture treated plates were purchased from Costar (Kennebunk, ME). Proparacaine hydrochloride 0.5% was purchased from Akorn, Inc (Lake Forest, IL). Tropicamide 0.5% and phenylephrine 10% ophthalmic solutions were obtained from the Johns Hopkins Pharmacy. Titan 3 0.45 μm PTFE hydrophilic syringe filters were purchased from Thermo Fisher (Rockwood, TN). Steri-dropper 10 mL sterile dropper bottles were purchased from The Medi-dose Group (Ivyland, PA). 20 mL vials with red pressure relief caps were purchased from Chemglass Life Sciences (Vineland, NJ).

Cell culture for ARPE-19 cells

ARPE-19 cells were cultured in a T-75 cell culture flask in DMEM:F12 media supplemented with 10% Fetal Bovine Serum and 1% Penicillin/Streptomycin according to ATCC guidelines and as previously described [24]. Cells were cultured in a 37 °C cell culture incubator with 5% CO₂ atmosphere until full confluency. The culture media was removed, trypsin was added, and the cells were placed in the incubator until detached. Trypsin was neutralized following addition of 4X volume of fresh culture media and centrifugation at 300 xg for 3 min to pellet cells. The pellet was resuspended in 3 mL of fresh culture media, and 1 mL was subsequently added to a fresh T-75 flask for further propagation. ARPE-19 cells were used for experimentation from passage 3 through 25.

In vitro reactive oxygen species (ROS) assay

ARPE-19 cells were cultured as above then seeded at a concentration of 2.5 × 10⁴ cells/well (200 µL) in fresh culture media in 96-well black walled, clear bottom tissue culture treated plates for fluorescence detection. Cells were allowed to adhere for 24 h before experimentation. The ROS assay was performed according to DCFDA/H₂DCFDA Cellular ROS Assay Kit instructions. Briefly, ARPE-19 cells were incubated with 10 µM DCFDA in 1X Supplemental Buffer (1X Buffer + 10% Fetal Bovine Serum) for 45 min. Cells either received 200 µM TBHP alone or co-incubation with NAC, NACET, or diNAC at a range of concentrations, where 200 µM had been previously demonstrated to be the maximum concentration to not significantly affect cell viability in ARPE-19 cells [25]. ROS quantification was performed on a Spark® Multimode Microplate Reader (Tecan Life Sciences, Mannedorf, Switzerland) at an Ex/Em of 485/535 hourly for 8 h. Between measurements, the assay plate was kept in a 37 °C cell culture incubator.

Preparation of NAC and NACET gelling eye drop formulations

The gel-forming vehicle contained 12% (w/w) Pluronic F127 as previously described [20, 21]. Briefly, Pluronic F127 was prepared at 20% (w/v) in distilled water and storage at 4 °C until fully dissolved. 12% F127 was prepared after 6:4 (w/w) dilution in Milli-Q water. To prepare equimolar formulations of NAC and NACET for use in in vivo studies, solubility screening was first performed on the more lipophilic NACET. The maximum solubility of NACET in 12% F127 was determined to be 21.5 mg/mL (112 mM), which corresponds to a molar equivalent of 18.35 mg/mL NAC. Thus, eye drops were prepared by adding 21.5 mg/mL of NACET and 18.35 mg/mL of NAC to 12% (w/w) F127 and allowed to dissolve overnight. Eye drop formulations were stored at 4 °C and used for up to two weeks.

Animal ethics statement

All procedures, protocols, and medications were approved by the Johns Hopkins University Animal Care and Use Committee and were performed in accordance with the Association for Research in Vision and Ophthalmology statement for use of animals in ophthalmic and vision research, and the National Institutes of Health guide for the care and use of laboratory animals. Animals were housed in a pathogen-free facility in micro-isolator cages. Veterinarians who are members of the Animal Services Department oversaw the care and handling of animals and performed regular inspections. Sentinel animals were used to monitor for pathogens and all positive screening tests led to immediate quarantine. Quarantines were not lifted until remedial action was taken and 2 subsequent screening tests were negative. Veterinarians were also available for consultation and care of animals that were sick. Attempts to minimize pain and distress in animals were taken. For injections and ERG, animals were anesthetized with ketamine/xylazine. For eye drop administration, animals were gently restrained without the need for anesthesia. Prior to injections, the ocular surface was sterilized with povidone-iodine, and 0.5% proparacaine hydrochloride anesthetic eye drops were applied. Bacitracin zinc ointment was applied to the injection site for infection control. Animals were monitored over the course of experimentation and confirmed that animals were awake and alert following anesthesia. All euthanasia methods were approved by the Panel of Euthanasia of the American Veterinary Medical Association.

Recording of electroretinograms (ERGs) in mice and rats

Scotopic and photopic ERGs were recorded using an Espion ERG machine (Diagnosys LLC, Littleton, MA) as previously described [9]. Prior to scotopic ERG, mice and rats were dark adapted overnight. Mice and rats were anesthetized with subcutaneous ketamine/xylazine, the corneal surface was numbed with 0.5% proparacaine drops, and pupils were dilated with sequential drops of 0.5% tropicamide and 2.5% phenylephrine HCl. Mice and rats were placed on a 37 °C heating pad for the duration of measurement, platinum electrodes were placed on the corneal surface, and a reference electrode was placed subcutaneously on the scalp between the eyes and a ground electrode was placed into the tail of mice and rats. The head of the rodents was centered under a ganzfeld bowl illuminator and right and left eye measurements were measured concurrently. Scotopic ERGs of rats were recorded at 7 light intensities ranging from −4.00 to −0.60 log cd-s/m². Following scotopic ERGs, rods were photobleached with a background light of 30 cd/m² for 10  min before recording photopic ERGs. Photopic ERGs of rats and mice were recorded at 3 light intensities ranging from 0.60 to 1.39 log cd-s/m². For each flash intensity, 5 readings were taken and averaged.

Gel-forming antioxidant eye drop efficacy in an intravitreal paraquat (PQ) model of RP in rats

Retinal degeneration was induced in 4-week-old Brown Norway rats following intravitreal injection (3 µL) of 2 mM PQ. Injections were performed using a Hamilton syringe fitted with a 30-gauge needle following pre-puncture of the sclera with a 30-gauge needle superiorly. Pressure equilibration was allowed for 30 s following injection to avoid backflow. Efficacy of the eye drop formulations was evaluated via the following dosing scheme: 3 once-daily loading doses of either NAC, NACET or vehicle only (5 µL) were administered bilaterally prior to intravitreal PQ injection. Brown Norway rats that did not undergo any treatment (PQ or eye drops) were also included as a healthy untreated control group. Daily eye drops were administered for one week following PQ injection, and scotopic and photopic ERGs were performed as above.

To perform outer nuclear layer (ONL) thickness measurements, following euthanasia, a mark was placed on the superior position with a cauterizing pen on both the right and left eye to ensure proper orientation throughout processing. Eyes were exposed to 15% followed by 30% sucrose gradient and embedded in OCT media and frozen. 12 µm thickness sections were collected through the optic nerve to ensure consistent depth and location across all groups. The sections were stained with hematoxylin and eosin and images were taken with a Nikon ECLIPSE Ni Upright Microscope (Nikon, Melville, NY) and ONL thickness at six locations relative to the optic nerve were measured: 25% (S3), 50% (S2), and 75% (S1) of the distance between the superior pole and the optic nerve and 25% (I3), 50% (I2), and 75% (I1) of the distance between the inferior pole and the optic nerve.

Gel-forming antioxidant eye drop efficacy in rd10 mice

Rod degeneration in rd10 mice is rapid after mice reach 2 weeks of age (Postnatal day 14, P14), around when pups open their eyes and topical eye drop treatment can be initiated. There is a significant loss of rods by 5 weeks of age (P35), around the time when cone degeneration begins, making evaluation of cone function by photopic b-wave ERG a desirable output for protecting photoreceptor survival with benchmarks at P35, P50 and P65 [9]. Pups from rd10 litters were randomly assigned to respective treatment groups. Mice were administered a single topical drop (2.5 µL) of NAC (18.35 mg/mL) or NACET (21.5 mg/mL) gel-forming drops, or gel-forming vehicle alone bilaterally daily starting from when their eyes opened (P13-15) through P65. Scotopic and photopic ERG measurements were conducted at the same light intensities as above with 5 measurements averaged per flash intensity at P35, P50 and P65.

Chemical stability studies of NAC in vitro

NAC samples were prepared at 10 mg/mL (61.28 mM) in water and 12% F127 to evaluate oxidation/dimerization rate with various additives (tannic acid (TA), ascorbic acid (AA), ethylenediaminetetraacetic acid (EDTA)) to reduce oxidation with prolonged storage. Samples were prepared with TA (30, 150, and 300 µg/mL) [26], EDTA (100, 500, and 1000 µg/mL) [27], and AA (10, 50, and 100 µg/mL) [28] due to their applications in reducing oxidation and scavenging free radicals in both ophthalmic and non-ophthalmic settings. Tannic acid and ascorbic acid were soluble in water, however EDTA was solubilized under basic conditions following minimal addition of 10 M NaOH. NAC samples were prepared without pH correction (~ pH 2.5) and at physiologic pH (pH 7) with 1 M NaOH and 1 M HCl using a Mettler Toledo FE20 Benchtop pH Meter (Columbia, MD). A subset of samples was also prepared in degassed water under oxygen-free conditions in an anaerobic glovebox (Coylab) or under aerobic conditions following bubbling under N₂ gas for 1 h with continuous stirring in a round bottom flask. Anaerobic samples were stored in glass screw top vials and N₂-purged samples were transferred to pressure-relief scintillation vials kept under N₂ atmosphere via 27-gauge insulin syringe. Samples were prepared in triplicate and were stored in the dark at 42 °C with 25% relative humidity (RH) in an accelerated degradation chamber, as 2 weeks storage under these conditions is equivalent to approximately 6 weeks storage under ambient conditions [29, 30]. The potency of samples following storage was then evaluated in the in vitro ROS assay in ARPE-19 cells as described above to evaluate impact of storage duration and additive on NAC oxidation. Further, NAC and diNAC content was evaluated via HPLC to determine the extent of dimerization. The HPLC mobile phase was constituted by 95:5 (v/v) water:acetonitrile containing 0.1% trifluoroacetic acid with a flow rate of 1 mL/min with a run time of 8 min at a λ_max = 200 nm. Chromatographic separation was achieved using a Kinetex reversed phase C18 column (150 × 4.6 mm, 5 µm particle size). NAC concentration was determined following interpolation to a previously constructed standard curve.

Statistical analysis

Statistical analysis was performed in GraphPad Prism. Statistical analysis for ERG readings were done by comparing scotopic or photopic amplitudes between NAC, NACET and Vehicle, at each light intensity using Mann–Whitney non-parametric analysis. P values ≤ 0.05 were considered statistically significant. Analyses for ROS assays were performed by one way or two-way ANOVA followed by Tukey’s multiple comparison test with differences considered significant at p < 0.05.

Results

In vitro ROS assay

The antioxidant activity of NAC and NACET was assessed in an in vitro ROS assay in ARPE-19 cells using TBHP (200 µM) as the oxidative stress-generating reagent. A concentration-dependent increase in ROS level was observed with increasing TBHP from 50 µM to 200 µM (Supplementary Fig. 1). Further, when THBP was co-incubated with either NAC, NACET, or diNAC (two NAC molecules bridged via disulfide bond, masking the active thiol) for 9 h, a concentration-dependent reduction in ROS generation was observed (Fig. 1). ROS generation was reduced by 55.6% and 53.1% of the maximum (TBHP only) following co-incubation with 25 µM NAC and NACET, respectively. After increasing the concentrations of NAC and NACET to 50 µM, ROS generation was further reduced to 69.3% and 73.2% of the maximum, respectively. diNAC showed greatly reduced antioxidant activity, as treatment with 25 and 50 µM diNAC reduced ROS generation by only 20.8% and 12.7% relative to the maximum (Fig. 1).

Fig. 1.

Percent of maximum reactive oxygen species (ROS) generation as determined by fluorescent detection of ROS-sensitive dye 2',7'-dichlorodihydrofluorescein diacetate (DCFDA) following co-incubation of 200 µM tert-butyl hydroperoxide (TBHP) with NAC, NACET or diNAC with ARPE-19 cells. Data is represented as mean ± SEM, n = 6 for each group other than the TBHP group with n = 12. ***p < 0.001, ****p < 0.0001 by 2-way ANOVA

Gel-forming antioxidant eye drop efficacy in rats

To determine whether NAC or NACET in the gel-forming eye drop vehicle would provide protection of photoreceptor structure and function in vivo, we performed once daily dosing for a total of 3 doses prior to intravitreal PQ injection, then continued once daily administration for 7 doses after induction of retinal degeneration in Brown Norway rats. Intravitreal injection of PQ leads to oxidative stress-induced death of both rod and cone photoreceptors [31, 32]. At the conclusion of the dosing regimen, we performed ERGs and measured the scotopic b-wave (Fig. 2A) and photopic b-wave (Fig. 2B) amplitudes after treatment with NAC, NACET, or vehicle only. Rats treated with either NAC and NACET (112 mM) showed functional protection to photoreceptors in PQ-induced retinal degeneration. NAC was protective at the two highest scotopic flash intensities, with ~ 46% and 61% increase compared to vehicle, respectively (Fig. 2A), whereas NACET was protective at the two highest flash intensities with 43% and 48% increase as compared to vehicle in photopic ERGs. (Fig. 2B). We then evaluated the impact of PQ treatment on outer nuclear layer (ONL) thickness to examine the extent of structural protection of retinal photoreceptors. Following enucleation, eyes were fixed, frozen, sectioned and H&E stained to visualize the ONL. Rats receiving either NAC or NACET eye drop treatment had greater ONL thickness compared to vehicle treated rats, and resembled healthy, age-matched controls along both superior and inferior poles (Fig. 2C, D). However, there were no significant differences observed between NAC and NACET treated animals (Fig. 2C). No gross adverse ocular surface effects were observed with daily dosing of NAC or NACET.

Fig. 2.

Topical delivery of gel-forming drops containing NAC and NACET provide protection of photoreceptor function and structure in rats after intravitreal injection of the ROS generator paraquat (PQ). (A) Scotopic and (B) photopic b-wave amplitudes by ERG following once-daily topical NAC, NACET, or vehicle (12% F127 only) treatment for 3 days prior and 7 days after PQ injection (total 10 doses). Treatment with NAC showed significant protection of scotopic b-wave amplitudes at the two highest stimulus intensities compared to vehicle and treatment with NACET showed significant protection of photopic b-wave amplitudes at the two highest stimulus intensities compared to vehicle. (C) Measurement of the outer nuclear layer (ONL) thickness confirmed protection of both NAC and NACET on photoreceptor structure. Sections were collected and ONL thickness was measured at fixed distances from the optic nerve (ON) locations 25%, 50% and 75% of the distance along the superior (S1,2,3) and inferior (I1,2,3) poles. (D) Table of group comparisons and their statistical significance at different locations are shown in C. Each replicate was the average of measurements taken from 3 sequential sections of the same eye. Data represented as mean ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001 by Mann–Whitney test

Gel-forming antioxidant eye drop efficacy in rd10 mice

After confirming that topical delivery of NAC and NACET in gel-forming eye drops could provide functional and structural protection of photoreceptors in a chemically-induced model of ROS-driven retinal degeneration, we then sought to test efficacy in a genetically-induced model of RP. Rd10 mice possess a PDE6β mutation that initially leads to rod degeneration, followed by progressive cone loss [33]. Daily equimolar dosing of gel-forming drops (freshly prepared every 14 days) containing the polymer only (vehicle) or with 112 mM NAC or 112 mM NACET (21.5 mg/mL) was initiated when the eyes opened at P14. Both NAC and NACET eye drops provided no significant protection in photopic ERGs at P35, when cone degeneration begins compared to vehicle (Fig. 3A). At P50, a  mid stage of cone degeneration, daily NACET administration led to significant protection of cone function (~ 31%) compared to vehicle at the highest flash intensity in photopic b-wave amplitude, as compared to the vehicle in rd10 mice model of RP (Fig. 3B). By P65, a late stage of cone degeneration, cone function in mice receiving either NAC or NACET gel-forming eye drops resembled vehicle only controls (Fig. 3C). It is possible that the decline in therapeutic effect could have been due to the aggressiveness of the cone degeneration in the rd10 mice [34], or potential issues with storing the drug solution during experimentation.

Fig. 3.

Topical antioxidants provided partial protection of function of cone photoreceptors in rd10 mice as measured by the ERG photopic b-wave amplitude in mice at (A) P35, (B) P50 and (C) P65. Mice received once-daily treatment starting after their eyes opened (P14), and treatment was continued for the duration of the study. Treatment with NACET exhibited protection of photoreceptor function at P50 at the highest stimulus intensity compared to vehicle. Data represented as mean ± SEM. *p < 0.05 by Mann-Whitney test

Chemical stability studies of NAC in vitro

Long-term stability is essential in the development and manufacturing of aqueous eye-drop formulations. Due to the propensity of free thiols to form disulfide bonds in solution, we sought to evaluate the effect of preparation and storage, given the reduced antioxidant activity observed following NAC conversion to diNAC. We have also evaluated the antioxidant effect of NAC under different storage conditions in vitro. If NAC conversion were to occur during storage in solution, there may have been a negative impact on efficacy in rd10 mouse studies. Thus, stability studies under various storage conditions were conducted to quantify and mitigate NAC oxidation. Storage under accelerated degradation conditions (42 °C, 25% RH) was used to screen various antioxidants. After storage of NAC solution (10 mg/ml) for 2 weeks at 42 °C, there was approximately 40% NAC remaining (Fig. 4A), corresponding to only 15% reduction in ROS compared to maximum, whereas freshly prepared NAC provided a 61% reduction in ROS compared to maximum in ARPE-19 cells (Fig. 4B). Conversely, when EDTA was added at either 100, 500 or 1000 µg/mL, there was 75.2 to 83% NAC content with only approximately 1% diNAC in solution (Fig. 4A), leading to similar antioxidant potency in cells, resembling freshly prepared NAC stocks (Fig. 4B). Because EDTA stocks were prepared via basification and thiol reactivity is dependent on protonation state, NAC stability was also evaluated following the addition of equimolar NaOH, reducing diNAC formation by approximately 35% compared to unadjusted samples (Fig. 4A). While the addition of high concentration tannic acid (300 µg/mL) limited diNAC formation (10.3%) (Supplementary Fig. 2A), the corresponding reduction of ROS generation was 47% (Supplementary Fig. 2B), like that achieved with 1000 µg/mL EDTA supplementation (Fig. 4B). Interestingly, ascorbic acid reduced diNAC formation, but did not preserve the antioxidant effect of NAC. Ascorbic acid at 10 µg/mL reduced diNAC formation to 42.3%, and 100 µg/mL reduced diNAC formation to 17.1%. However, there was no evidence of further reduction of ROS generation with the addition of ascorbic acid (Supplementary Fig. 2C) when compared to freshly prepared NAC solution (Supplementary Fig. 2D). Although 300 µg/mL TA and EDTA had a similar impact on preserving NAC antioxidant activity when added to stored solutions, TA did not prevent NAC dimerization to the same extent. Thus, 1000 µg/mL EDTA was used in further stability testing.

Fig. 4.

Concentration and antioxidant potency of NAC following storage for 2 weeks under accelerated degradation conditions in aerobic media with and without EDTA (100, 500, or 1000 µg/mL). (A) Percent NAC and diNAC content in solution at day 7 and 14 compared to day 0 measurements. Data represented as the average of n = 3 replicates. (B) Percent of maximal reactive oxygen species (ROS) generation as determined by fluorescent detection of ROS-sensitive dye 2',7'-dichlorodihydrofluorescein diacetate (DCFDA) following co-incubation of 200 µM tert-butyl hydroperoxide (TBHP) with NAC with and without EDTA in ARPE-19 cells. Data represented as mean ± SEM, n = 6 for each group. ****p < 0.0001 by 2-way ANOVA

We then evaluated the effect of storage medium and minimization of O₂ content in combination with EDTA on NAC stability after 2 weeks storage at 42 °C. To reduce oxygen content, we prepared solutions by eliminating headspace in the storage vessel, attempting to purge with N₂, or degassing and preparing samples under an inert atmosphere. Removal of headspace alone reduced diNAC formation from 76.5% to 34.9% in aerobic stocks after 2 weeks (Fig. 5A). Although bubbling under N₂ gas has been reported to reduce soluble O₂ [35], NAC and diNAC content remained comparable to samples stored without N₂ purging (Supplementary Fig. 3). We also confirmed that polymer incorporation in storage media did not affect NAC stability, as preparation in the gel-forming vehicle (12% F127) did not greatly affect NAC and diNAC content when compared with preparation in water with and without EDTA addition (Supplementary Fig. 3). However, complete removal of O₂ through preparation in anaerobic media minimized oxidation of NAC to diNAC with and without EDTA addition, with 1.5% and 5.3% diNAC content, respectively (Fig. 5A). Lastly, the NAC activity was preserved following 2-week storage of anaerobically prepared NAC solutions at 42 °C, reflected by comparable potency for reducing ROS in ARPE-19 cells (Fig. 5B). Across all storage conditions, the addition of 1000 µg/mL EDTA most significantly reduced the formation of diNAC in solution under the storage conditions tested.

Fig. 5.

Concentration and antioxidant potency of NAC following storage for 2 weeks under accelerated degradation conditions in aerobic and anaerobic media with and without 1000 µg/mL EDTA and headspace minimization. (A) Percent NAC and diNAC content in solution at day 7 and 14 compared to day 0 measurements. Data represented as the average of n = 3 replicates. (B) Percent of maximal reactive oxygen species (ROS) generation as determined by fluorescent detection of ROS-sensitive dye 2',7'-dichlorodihydrofluorescein diacetate (DCFDA) following co-incubation of 200 µM tert-butyl hydroperoxide (TBHP) with NAC with and without EDTA stored under anaerobic conditions. Data represented as mean ± SEM, n = 6 for each group. ****p < 0.0001 by 2-way ANOVA

Discussion

Treatments for RP are desperately needed. Although Luxturna® is approved for use in correcting RPE65-induced retinal degeneration, only 1–2% of RP patients and 6–16% of Leber congenital amaurosis (LCA) patients have mutations that are treatable using this approach [36, 37]. The heterogeneity of mutations that lead to RP has led to a variety of approaches that attempt to address larger portions of the patient population. There have been various strategies reported preclinically, including the intravitreal injection of stanniocalcin-1, an antiangiogenic peptide that protected cone function and reduced oxidative stress in P23H pigs [38] and subretinal injection of AAV-based RPGR gene therapies in dogs [39]. However, the development of therapies for globally preventing photoreceptor death regardless of the mutation would be a great advancement in the management of RP. Oral administration of NAC was previously reported to improve photoreceptor function in rd10 mice [12] and in Phase 1 clinical trial [13], outlining a new promising minimally invasive route for RP treatment. However, NAC requires multiple doses per day to overcome low oral bioavailability [40]. Because RP is a chronic condition, the development of a minimally invasive therapeutic that reduces frequency of application with local administration is appealing to reduce risk of non-adherence.

Although eye drops are widely used for the treatment of disorders affecting the anterior segment, there is yet to be successful clinical development of a topical eye drop that achieves therapeutically effective delivery to the retina. Clinical trials have evaluated once per day formulations [41], as well as formulations that were applied 3 times per day [42] and up to 6 times per day [43] without evidence of therapeutic effect. The development of new delivery strategies that promote therapeutic accumulation in the posterior segment with efficacy comparable to injectable therapies but without onerous dosing regimens is necessary. We have previously reported the development of a hypotonic, gel-forming eye drop vehicle that provided therapeutic drug delivery to the posterior segment in rabbits and pigs [20]. By formulating Pluronic F127 below its critical gelling concentration (CGC) in a hypotonic vehicle, application to the ocular surface led to the formation of a thin, uniform, drug-laden gel layer [20]. This translated to therapeutic accumulation of sunitinib malate in both rabbits and pigs with once daily administration, and thus provided a greater reduction in choroidal neovascularization (CNV) spot size after laser-induced rupture of Bruch’s membrane compared to once daily sunitinib malate in a conventional liquid vehicle [20]. We hypothesized that loading water soluble antioxidants into the hypotonic, gel-forming vehicle could provide therapeutic scavenging of ROS in the posterior segment, a potential therapeutic opportunity for treating RP. Topical administration of NAC and NACET lessened the extent of photoreceptor function and structure degeneration in the paraquat-induced rat and rd10 mouse models of RP with once daily application. Partial preservation of cone function at P50 in rd10 mice was especially encouraging with a topical eye drop, as most rods are lost by ~ P30. The oxidative stress caused by dying rods then causes degeneration of the cones, which at only 2% of the photoreceptors in the mouse retina, become overwhelmed by the oxidative damage [44, 45]. However, achieving therapeutic accumulation in the retina of a larger eye (e.g. rabbit) is more challenging than in mice or rats [46]. Before investing in future work testing the efficacy of topical antioxidant eye drops to treat RP, improvements in product stability and the target product profile must first be made.

We hypothesized that storing the gel-forming NAC eye drops under refrigeration while performing daily dosing for several weeks in rd10 mice may have led to decreased potency due to NAC dimerization in solution. We observed decreased ROS-scavenging capacity with increased dimerization of the thiol groups in solution (formation of diNAC) in TBHP-induced ARPE-19 cells, providing insight on the balance between GSH/GSSG and the capacity for reduction of oxidative stress. Additionally, the presence of higher diNAC concentrations may further reduce intracellular GSH stores, resulting in increased ROS levels compared to TBHP-alone that were noted. By preventing diNAC formation through addition of 1000 µg/mL EDTA, we were able to retain NAC potency through 2-week storage of NAC solutions at 42 °C. EDTA is also widely reported to serve as a corneal penetration enhancer resulting from improved transcellular permeability, potentially through chelating Ca²⁺ from epithelial and endothelial tight junctions [47, 48]. However, the effect of EDTA on ocular bioavailability of NAC requires further assessment in vivo. Nevertheless, observing greater than 10% reduction in NAC concentration under accelerated degradation conditions may still prove insufficient to move forward with product development. Further investigation of stabilization of the thiol group could be possible for further development as a solution, though oxygenation may be an issue with products used continuously over a month. Single use ampules may be more feasible.

Here, we demonstrate the potential of topical delivery of the antioxidants NAC and NACET for the treatment of RP, regardless of causative mutation. While the therapeutic efficacy of the topical NAC and NACET dosing was more evident in the rat model of paraquat-induced photoreceptor death, this model involves non-specific photoreceptor degradation due to the chemical insult. In contrast, the rd10 mouse models genetically-induced rod death that progressively leads to cone death, as in RP, though the model is highly aggressive and the photoreceptor degeneration is very rapid and difficult to treat. Whether it was the aggressiveness of the rd10 model, or the loss of potency of the stored solutions, or a combination that led to only partial protection, future work in animal models is warranted with more translationally viable compositions. Further, while we observed that the addition of EDTA improved NAC stability and potency from ~ 33 to ~ 85% under accelerated conditions, ideally less than 10% loss with careful evaluation of degradation products would be necessary to further development. Further, the inherent limitations of studying topical dosing in murine models necessitates further studies of pharmacokinetics and efficacy in larger species, as well as ocular safety and tolerability studies. Nevertheless, this work validates the potential of topical antioxidant delivery as a therapeutic strategy for RP. With further optimization to the formulation and dosing of antioxidant gel-forming eye drops, there is potential for a topical therapy to protect photoreceptor function and survival for the treatment of RP.

Conclusion

We report the development of antioxidant-loaded, gel-forming eye drops for the treatment of RP with once-daily application. Through formulation of water-soluble antioxidants NAC and NACET in a hypotonic, gel-forming vehicle, we demonstrated the potential for protection of retina structure and function in rodents. A formulation optimized for therapeutic antioxidant delivery to the retina has the potential to provide all RP patients a minimally invasive treatment option for slowing progressive photoreceptor death and vision loss. However, further work in maintaining antioxidant stability and formulation for daily use, as well as studies in larger species, is needed.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Conceptualization: M.B.A., Y.K., P.A.C., L.M.E.; Investigation: M.B.A., Y.K., K.M, T.A., M.K.; Methodology: M.B.A., Y.K., K.M.; Writing – original draft: M.B.A, Y.K., L.M.E. Writing – review and editing: M.B.A., Y.K, K.M, T.A., M.K., P.A.C., L.M.E.

Funding

This work was supported by the National Institutes of Health (NIH) (grant no. R01EY031041 and P30EY001765), The Robert H. Smith Family Foundation, an unrestricted departmental grant from Research to Prevent Blindness, and Marcella E. Woll and the Maryland E-Novation Initiative Fund to establish the Marcella E. Woll Professorship in Ophthalmology. M.B.A. was supported in part by a National Eye Institute Training Grant (grant no. T32EY007143).

Data availability

Data will be made available upon request.

Declarations

Ethics approval

All procedures, protocols, and medications were approved by the Johns Hopkins University Animal Care and Use Committee and were performed in accordance with the Association for Research in Vision and Ophthalmology statement for use of animals in ophthalmic and vision research, and the National Institutes of Health guide for the care and use of laboratory animals.

Competing interests

L.M.E. is a co-inventor on patents related to topical eye drops that are licensed to Novus Vision LLC. L.M.E. is a co-founder and on the Board of Directors for Novus Vision LLC. The terms of this arrangement are being managed by the Johns Hopkins University (JHU) in accordance with its conflict-of-interest policies. P.A.C. has no conflicts directly related to manuscript, but unrelated to the manuscript serves as consultant for Cove, ExgenesisBio, Exonate, Ltd, Genentech, Neuvasq, Merck & Co Inc, and Perfuse. He has equity in Cove and JHU receives research support from Cove, Genentech, Ocular Therapeutics, Oxford Biomedica, and RegenxBio. These interests are managed in accordance with JHU policies. The other authors declare no competing interests.