Evaluation of Pregabalin Bioadhesive Multilayered Microemulsion IOP-Lowering Eye Drops

Abstract

In spite of available treatment options, glaucoma continues to be a leading cause of irreversible blindness in the world. Current glaucoma medications have multiple limitations including: lack of sustained action; requirement for multiple dosing per day, ocular irritation and limited options for drugs with different mechanisms of action. Previously, we demonstrated that pregabalin, a drug with high affinity and selectivity for CACNA2D1, lowered IOP in a dose-dependent manner. The current study was designed to evaluate pregabalin microemulsion eye drops and to estimate its efficacy in humans using in silico methods. Molecular docking studies of pregabalin against CACNA2D1 of mouse, rabbit, and human were performed. Pregabalin microemulsion eye drops were characterized using multiple in vivo studies and its stability was evaluated over one year at different storage conditions. Molecular docking analyses and QSPR of pregabalin confirmed its suitability as a new IOP-lowering medication that functions using a new mechanism of action by binding to CACNA2D1 in all species evaluated. Because of its prolonged corneal residence time and corneal penetration enhancement, a single topical application of pregabalin ME can provide an extended IOP reduction of more than day in different animal models. Repeated daily dosing for 2 months confirms the lack of any tachyphylactic effect, which is a common drawback among marketed IOP-lowering medications. In addition, pregabalin microemulsion demonstrated good physical stability for one year, and chemical stability for 3–6 months if stored below 25°C. Collectively, these outcomes greatly support the use of pregabalin eye drops as once daily IOP-lowering therapy for glaucoma management.

Graphical Abstract

1. Introduction

In spite of available treatment options, glaucoma continues to be a leading cause of irreversible blindness in the world. Trends predict that by 2040, as many as 111.8 million people worldwide will have glaucoma[1] and many of those will be legally blind due to optic nerve (ON) damage[2]. Several risk factors are known for this disease[3–5], with elevated intraocular pressure (IOP)[2, 3, 6–8] and fluctuations in IOP[9, 10] being the only modifiable risk factors linked to the development and progression of glaucoma[2, 3, 6–8]. As such, the first line standard of care for all forms of adult-onset glaucoma is treatment with IOP-lowering medications delivered topically as eye drops. Current glaucoma medications have multiple issues, the most significant of which are: 1) lack of sustained action; 2) requirement for multiple dosing per day; 3) direct ocular irritation; 4) possible systemic and ocular side effects; and 5) limited options for drugs with different mechanisms of action. Previously, we discovered that pregabalin (PRG), an antagonist with high specificity and affinity for CACNA2D1 could be used as a new IOP-lowering drug for the treatment of glaucoma. This was discovered based on the identification of the calcium voltage-gated Cacna2d1 gene as a modulator of IOP. In our previous studies, PRG lowered IOP in a dose dependent manner when delivered in an aqueous solution in mice[11]; however, its efficacy was low and would require multiple dosings per day to maintain a reduced IOP for a full 24 h. To eliminate potentially harmful IOP fluctuations that occur with a short acting drug and to offer a fully effective IOP-lowering with once daily dosing, we engineered a transparent novel water-in-oil-in-water (W/O/W) microemulsion (ME) formulation with extended release and bioadhesive properties. Our previously published data demonstrate that a single drop of PRG ME induced a markedly reduced IOP that returned to baseline at ~33 h after a single application[12]. Thus, it is plausible that PRG ME eye drops could serve as a promising once daily new glaucoma therapeutic and drug delivery system.

To corroborate the efficacy studies, we employed advanced molecular modeling techniques to investigate the interaction between PRG and mouse, rabbit, and human CACNA2D1. The 3D structure of CACNA2D1 for each species was subjected to molecular docking with PRG against their canonical binding site via Glide in Maestro (Schrödinger, Inc., USA). The resultant docked complexes were analyzed using PyMOL molecular visualization software to elucidate the binding mode of PRG with respect to CACNA2D1 in each species. This meticulous molecular modeling approach provided insights into the intricate molecular interactions underlying PRG - CACNA2D1 binding across diverse species, augmenting our understanding of their interaction dynamics. Likewise, in drug discovery and lead optimization, computational assays have emerged as powerful tools [13–19]. In addition to molecular docking, molecular dynamics simulations, founded on Newtonian principles, emerge as a powerful computational tool that considers protein structure flexibility and facilitate the study of protein-ligand interactions collected during simulations [20–23]. Molecular mechanics and the generalized Born surface area (MM/GBSA) method can then be used to predict the binding free energy of compounds with macromolecules[24–26]. In instances where the protein structure is unknown, as is the case in the current investigation, accurate prediction of target structure becomes imperative. Homology modeling emerges as a highly reliable computational method in this context [27]. The process entails the identification of a template for the target protein, typically a peptide sequence from another protein exhibiting significant sequence identity. Studies of this nature are also helpful in predicting the translatability of efficacy findings in preclinical models to humans.

In addition to the analysis of PRG binding to CACNA2D1, it is very important to determine the physical and chemical stability of PRG ME formulation as a new drug product. Subsequently, the recommended transport, shipping and storage conditions should be established before marketing a new pharmaceutical eye drops. Hence, the drug product should be exposed to a variety of temperatures for a specific time. To evaluate the stability of a new drug product, the International Conference on Harmonisation (ICH) of technical requirements for registration of pharmaceuticals for human use produced a specific guidance that defines the stability data package required for a registration application in the United States, the European Union and Japan.

The main goal in this investigation was to evaluate PRG ME formulation as a once daily glaucoma therapy. In the context of this study, we aimed to first create homology models of CACNA2D1 in diverse species and conduct molecular docking of the S-enantiomer of PRG obtained from PUBCHEM database using Glide in Maestro (Schrödinger, Inc., USA). We also determined the minimum concentration of PRG that produce the maximum reduction in IOP (the optimal PRG dose) and tested its IOP-lowering effect in Dutch belted rabbits and three mouse strains [C57BL/6J (B6), BXD14 and BXD44]. Furthermore, the corneal residence time of PRG ME formulation was evaluated by using a fluorescent version of the ME. Safety and efficacy profiles of PRG ME upon prolonged use was also evaluated. Long term, intermediate and accelerated stability of the ME formulation were investigated according to the ICH guidance storage conditions. By optimizing the in vivo bioadhesion characteristics of PRG ME with an excellent safety profile, we have demonstrated its efficacy as a promising once daily eye drops for glaucoma therapy without any decrease in the drug response upon prolonged use.

2. Materials and Methods

2.1. Materials

Pregabalin (PRG), was purchased from Nantong Chanyoo Pharmatech Co., Ltd (Nantong, China). Formic Acid (99+%, Optima^™ LC/MS Grade), methaqualone solution (1.0 mg/ml in methanol), propylene glycol, sodium chloride, phosphoric acid, methanol and acetonitrile (HPLC grade) were purchased from Fisher Scientific (Fair Lawn, NJ). Gift samples of Labrafac Lipophile WL1349, Capryol 90 and Labrasol were obtained from Gattefossé Corporation (Paramus, NJ). Soybean L-α-Lecithin was purchased from Calbiochem (Billerica, MA). Ethyl alcohol was purchased from Decon Labs, Inc. (King of Prussia, PA). Carbopol 981 was obtained as a gift sample from Lubrizol advanced materials, Inc. (Cleveland, OH). Ketamine hydrochloride (100 mg/ml) and Xylazine hydrochloride (100 mg/ml) were obtained from Covetrus North America (Dublin, OH). Acridine orange hydrochloride, Cremophore^® EL, sodium phosphate dibasic, potassium chloride, potassium dihydrogen phosphate and dipotassium hydrogen orthophosphate were purchased from Sigma-Aldrich (St. Louis, MO).

2.2. Animals

Dutch belted rabbits (equal numbers of males and females) were procured from Covance Inc., (NJ, USA) weighing ~1.5–2.2 Kg and were aged 6–12 months. Different mice strains used in the current study— (C57BL/6J) B6, BXD14 and BXD44—were obtained from Jackson Laboratory (Bar Harbor, ME). Six mice of each strain were used, aged 3–5 months and weighed 20–25gm. All procedures including rabbits and mice were approved by the Animal Care and Use Review Board of the University of Tennessee Health Science Center (UTHSC) and complied with the Association of Research in Vision and Ophthalmology (ARVO) Statement for the Use of Animals in Ophthalmic and Vision Research in addition to the guidelines for laboratory animal experiments (Institute of Laboratory Animal Resources, Public Health Service Policy on Humane Care and Use of Laboratory Animals).

2.3. Pregabalin Assay

A previously published reversed phase HPLC-UV method was used for quantification of PRG throughout the stability study [28]. An Agilent 1100 series HPLC system (Germany) was attached to Supelco kromasil C18 column (5 μm, 100°A, 4.0mm × 300mm). The mobile phase consisted of a mixture of methanol: acetonitrile: 0.02M dipotassium hydrogen orthophosphate (3:1:16, v/v/v) at a flow rate of 1ml/min. The effluent was monitored by a photodiode array detector and PRG was detected after a retention time of 5.4 min at a detection wavelength of 210 nm.

The ultra-high-performance liquid chromatography-tandem mass spectrometry (UPLC– MS/MS) method developed and validated by Pauly et al. (2013) [29] was verified and used for the quantification of PRG in all biological samples.

2.4. Molecular modeling of PRG interaction with mouse, rabbit, and human CACNA2D1

The 3D structures of rabbit CACNA2D1 (PDB: 7JPV) and human CACNA2D1 (PDB: 7VFS) were obtained from the protein data bank, whose electron microscopy resolutions were 3.40 Å and 2.80 Å, respectively [30, 31]. Because the 3D structure of mouse CACNA2D1 is not available in the protein data bank, its homology model was constructed as previously described [18]. Specifically, the voltage-gated calcium channel Ca(v)1.1 (PDB: 7MIX) shared ~ 94% sequence identity with the peptide sequence of mouse CACNA2D1, which was subjected to homology modelling using Prime in Maestro software (Schrödinger, Inc., USA). Notably, when the protein sequence shares greater than 30% identity with the peptide sequence of your target structure, predictions often approach the precision of low-resolution X-ray structures [31]. Fine-tuning of the model through post-modifications is frequently necessary to ensure heightened reliability of the predictions. For instance, the non-template loops with less than 7 residues were then refined with OPLS3 force field and VSGB solvation model. The model was further refined by using the protein preparation wizard tool in Maestro, which generated the ionization state at pH 7.0 ± 2.0 using the Epik module. The peptide sequence of CACNA2D1 was subjected to pairwise-sequence alignment using EMBOSS Needle tool to compare the sequence identities across all three species [32]. Next, in preparation for molecular docking, the 3D structure of PRG was obtained from PubChem, refined using LigPrep in Maestro with the OPLS3e force field, and was subsequently docked against the canonical binding site of CACNA2D1 for each species via Glide in Maestro [33]. The docked complexes were then transferred to PyMOL molecular visualization software (Schrödinger, Inc., USA) to analyze the binding mode of PRG with respect to CACNA2D1 in each species.

2.5. Quantitative structure property relationships (QSPR) analysis of PRG

The 3D structure of PRG was subjected to QSPR analysis using physics-based membrane permeability tool in Maestro [34, 35]. The physiochemical properties derived mathematically from QSPR were then used to predict the corneal permeability of PRG [14]. To calculate changes in free energy of partitioning (ΔG_o/w), we first obtained the distribution coefficient (cLogD at pH 7.4) and partition coefficient values (cLogP). The free energy of distribution value was then calculated based on theoretical equations adapted from the literature [14, 34, 36].

2.6. Preparation of PRG multilayered ME eye drops

Blank ME was prepared using our previously published method with some modifications[12]. Briefly, the blank water in oil in water (W/O/W) ME was prepared in several steps, the first of which was the preparation of the primary W/O ME. A pseudo-ternary phase diagram was constructed using the water titration method to determine the ratios of the primary W/O ME components. In this method, labrafac lipophile WL1349 (oil phase) was mixed with surfactants mixture (capryol 90 and soybean lecithin, 1:1) in several ratios. The prepared mixtures were titrated with Milli-Q water until the appearance of the first turbidity that indicated the end of the ME zone. From this ME zone, one point was selected which provided the ratios of the primary W/O ME components. The second step was the preparation of the bioadhesive external aqueous phase by incorporating Carbopol 981 in a mixture of Milli-Q water with surfactants and cosurfactant (Labrasol, Cremophor EL and propylene glycol). The last step was the simple mixing of the primary W/O ME with the external aqueous phase to produce the final W/O/W ME. The ratio of the primary W/O ME that incorporated in the final multiple ME was determined by the titration method. PRG ME was prepared using the same method as the blank ME while incorporating 30% of the PRG in the internal aqueous phase and 70% in the external aqueous phase.

2.7. Pregabalin ME in vivo Evaluations

2.7.1. Dose finding and EC₅₀ determination in Dutch belted rabbits

PRG ME eye drops were prepared according to the previously mentioned method using seven different PRG concentrations (0.1, 0.2, 0.3, 0.4, 0.5, 0.6 and 0.7% w/w). Each formulation was evaluated using a single dose-response design. Briefly, 100 μl of each PRG ME formulation was applied into the inferior conjunctival sac of the right eye of Dutch belted rabbits (n = 5) while the left eye received the blank ME eye drops and served as a control using the previously published protocol [11, 37]. IOP was measured using a Tono-Pen AVIA Vet, (Reichert Technologies, Depew, NY) prior to treatment (baseline) and at hourly time intervals after application until the IOP returned to its baseline value. Area under the % IOP reduction versus time curves were calculated for each formulation and plotted against the corresponding PRG concentration. The EC₅₀ was calculated from the area under the curve versus PRG concentration graph as the PRG concentration corresponding to 50% of the highest area under the curve. For the purpose of comparison between different formulations of PRG-loaded ME, different pharmacodynamics (PD) parameters were calculated using GraphPad Prism 10 software (GraphPad Software Inc., San Diego, CA). PD data were statistically analyzed using one-way analysis of variance (ANOVA) test followed by Tukey’s multiple comparisons test. The calculated PD parameters include: maximum reduction in IOP (% IOP reduction); time required to reach maximum decrease in IOP (T_max); time required for IOP to return to baseline value (i.e., end of drug effect; T_end); and total area under the %IOP-vs-time curve (AUC). Results were reported as mean ± SEM.

2.7.2. Efficacy study on different mice strains

To test the efficacy of PRG as an IOP-lowering medication in rodents, three different mouse strains—C57BL/6J (B6), BXD14 and BXD44—were used in a single dose-response design study. These strains were selected because mice carrying the B parental allele of Cacna2d1—B6, BXD14 and BXD44—are more responsive to PRG than mice carrying the D parental allele [11]. Ten microliters of PRG ME (the selected concentration from the dose finding study) were instilled into the inferior conjunctival sac of the right eye in two separate applications of 5μl, each applied within 1min, while the left eye received the blank ME eye drops and served as a control (n = 6). IOP was measured using a Tonolab tonometer (Colonial Medical Supply, Franconia, NH) prior to treatment (baseline) and at hourly time intervals after application until the IOP returned to baseline value [38]. For comparison of PRG efficacy between different mice strains, different PD parameters were calculated, and statistically assayed using GraphPad Prism 10 software as mentioned above, under the dose finding study. The results were presented as mean ± SEM.

2.7.3. Efficacy study in comparison to a market leader commercial glaucoma medication

To compare the IOP-lowering efficacy of PRG ME eye drops to latanoprost eye drops as a commercially available market leader treatment for glaucoma[39], ten Dutch belted rabbits were divided into two groups (5 each). Each rabbit from the first group received 30 μl of PRG ME (the selected concentration from the dose finding study) into the inferior conjunctival sac of the right eye while the left eye received the blank ME eye drops and served as a control[11, 37]. Each rabbit from the second group received 30 μl of latanoprost eye drops (0.005% w/v) into the inferior conjunctival sac of the right eye while the left eye received latanoprost vehicle eye drops and served as a control. In both groups, the IOP was measured using a Tono-Pen AVIA Vet, (Reichert Technologies, Depew, NY) prior to treatment (baseline) and at hourly time intervals after application until the IOP returned to baseline value. For comparison of PRG efficacy with latanoprost, different pharmacodynamic parameters were calculated, and statistically assayed using GraphPad Prism 10 software as mentioned above, under the dose finding study. The results were presented as mean ± SEM.

2.7.4. Bioadhesion and corneal residence time assessment

The ability of the ME to adhere to the eye surface after topical application was evaluated using a fluorescent version of the ME so that it could be easily visualized. To prepare the fluorescent ME, PRG was replaced by acridine orange hydrochloride (AO) fluorescent dye of roughly equivalent physiochemical characteristics to PRG as a surrogate API in both internal and external aqueous phases of the ME. Dutch belted rabbits (n = 5) were dosed with 100μl of the fluorescent ME in one eye, while the fellow eye received 100μl of AO in PBS. Tears were collected and eyes were imaged immediately before the application (baseline) and at predetermined time points—5 minutes, 1, 6, 12 and 24 h—after application of the formulations. Eye photos were taken using ClearView-2 camera (Optibrand, Fort Collins, CO) under blue light source. The fluorescence intensities of images were quantified using ImageJ software. To increase the volume of the collected tears, 50μl of PBS was applied to the outer corner of each eye and collected from its other corner. Tears samples were assayed for AO contents by a validated spectrophotofluorometric method using microplate spectrophotofluorometer (μ-Quant Bio-Tek Instruments, Inc. Winooski, VT) at excitation/emission wavelengths of 490/520nm.

To evaluate the bioadhesion and corneal residence time of PRG ME, the same steps were repeated (with the exception of imaging) using PRG ME instead of AO ME. Briefly, Dutch belted rabbits (n = 5) dosed with 100μl PRG ME formulation in one eye while the other eye received 100μl PRG aqueous solution. At predetermined time points (baseline (0), 5 minutes, 1, 6, 12 and 24 h after application), tears samples were collected from both eyes. All the collected tear samples were kept at −80°C until further analysis. PRG contents of the tears were quantified using LC-MS/MS method with a limit of quantification (LOQ) of 2.5pg/mg[29].

To assess the ability of the ME to maintain the inherent high permeability of AO (BCS class I drug) and its ability to deliver it into deeper eye tissues, rabbits were euthanized 24 h after dosing with either AO in PBS or AO ME. Eyes were enucleated and fixed in 4% paraformaldehyde in PBS, pH 7.4 for at least 24 h. Fixed eyes were processed for paraffin embedding by dehydrating with a graded ethanol series consisting of 50%, 70%, 85%, 95% ethanol followed by three 100% ethanol incubations for 1 h each. The eyes were then transitioned to xylenes using 2:1, 1:1, 1:2 ratios of ethanol: xylenes, then incubated twice with 100% xylenes for 1 h each. Eyes were transitioned to paraffin using 2:1, 1:1, 1:2 ratios of xylenes: paraffin for 1 h each, followed by two 100% paraffin incubations for 4 h each. Eyes were then embedded in paraffin and sectioned on a rotary microtome to produce 8μm thick sections that were transferred to glass slides for fluorescent immunohistochemistry (FIHC). Eye sections were deparaffinized using 100% xylenes repeated twice, followed by 1:1 ethanol: xylenes, 100% ethanol repeated twice, then 95%, 85%, 70%, 50% ethanol, and ultrapure water repeated twice for at least 2 min per incubation. FIHC was performed on deparaffinized tissue using antigen retrieval for 90 min at 70°C in citrate buffer, pH 6.0 consisting of 10mM sodium citrate and 0.05% Tween-20. Slides were then washed twice in PBS, pH 7.4, and an aqueous barrier was drawn around the sections using a hydrophobic pen (Gnome Pens, Invignome). Slides were blocked for 30 min with goat blocking buffer consisting of 10% goat serum, 5% BSA and 0.5% TritonX-100 in PBS, pH 7.4. After blocking, primary antibodies were added in a 1:1 mix of PBS: goat blocking buffer and incubated overnight at 4°C in a humidified slide tray. The primary antibodies used were anti-CACNA2D1 (mouse IgG1,Santa Cruz Biotechnology, sc-271697) at a dilution factor of 1:100 and anti-beta actin (mouse IgG2b,ProteinTech, 66009–1-Ig) at a dilution factor of 1:200. After overnight incubation, primaries were removed from slides with three PBS washes for at least 10 min each. Goat anti-mouse IgG1 conjugated to HRP (ThermoFisher, A10551) and goat anti-mouse IgG2b conjugated to AlexaFlour 647 (ThermoFisher, A21242) were diluted 1:200 and 1:400, respectively, in a 1:1 solution of PBS: goat blocking buffer. Slides were incubated in secondary antibody for 1 h at room temperature before being removed with four PBS washes. CACNA2D1 labeling was enhanced using tyramide signal amplification (TSA) using the 568 tyramide 568 reagent (Click Chemistry Tools, 1541–2). A tyramide working solution was made by combining 10 mL tyramide 568 reagent and 10 mL 0.143% aqueous hydrogen peroxide per 1 mL reaction buffer (50mM Tris-HCl, pH 7.4). Slides were incubated for 2 min in the tyramide working solution before being washed 3 times with PBS. Slides were stained with a 1:10,000 dilution of 14.3 mM DAPI (ThermoFisher, D21490) for 1 h at room temperature, followed by 2 PBS washes. Slides were then mounted with Prolong Diamond Antifade Mountant (ThermoFisher, P36970). Z-stacks were taken using confocal microscope (Zeiss 980 LSM, Germany) with either 10X objective with 0.45 NA (cornea) or 20X objective with 0.8 NA (ciliary body and trabecular meshwork).

2.7.5. Efficacy, safety and drug-tissue biodistribution studies after 60 days of daily application

The efficacy and safety of PRG ME upon prolonged use were tested on Dutch belted rabbits. Five rabbits were dosed once daily by 100 μl PRG ME (the selected concentration from the dose finding study) in one eye, while the fellow eye received blank ME for 60 consecutive days. The IOP of both eyes was measured twice daily, immediately before the formulation application and at the time of maximum IOP reduction (T_max). Slit-lamp biomicroscopic and fundoscopic examinations were performed before and after study at days 0 and 61 to detect any ocular side effects that may happen due to repeated use of PRG ME.

Thereafter, the rabbits were euthanized on day 61. Both eyes, brain, in addition to select peripheral organs (lungs, heart, spleen, kidneys, liver) were collected. Eyes were dissected while fresh into different tissues (cornea, aqueous humor, iris, ciliary body, trabecular meshwork, lens, vitreous humor, retina, eye cup (sclera/choroid/retinal pigment epithelium) and optic nerve using our published method[40]. Blood samples were collected by cardiac puncture then separated into plasma while fresh. All tissues and fluids were kept at −80°C until further analysis[41]. Each cornea was divided into two halves, one of which was used to estimate its drug content, while the other was subjected to histological examination to detect any abnormality in the corneal tissues as a result of the daily ME application. The drug contents of the samples were evaluated using LC−MS/MS method with LOQ of 2.5pg/mg. Briefly, tissues were homogenized in a frozen methanol using tissue tearor (Biospec. Inc., Bartlesville, OK). Tissue homogenates were kept overnight at −20 °C to dissolve PRG and precipitate protein. The methanolic extract was separated by centrifugation at 14000 rpm for 15 min at −5°C and then evaporated until dryness. The residue was dissolved in methanol before being injected into the LC−MS/MS system (SCIEX-5500, Framingham, MA)[29].

2.7.6. Pilot pharmacokinetics study after ocular application

An exploratory pharmacokinetics (PK) study was conducted using a single dose-response design to determine the affinity of PRG to different eye tissues. Twenty-two Dutch belted rabbits, balanced between males and females were used in the study. All rabbits were six months old and were weighing 1.5–1.8Kg. Two naïve rabbits were used for construction of standard curves. The remaining 20 rabbits were randomly divided into two groups (10 each). Each rabbit of the first group received 30μl of 0.6% PRG ME in both eyes in the lower conjunctival sac while each rabbit of the second group received 30μl of 0.6% PRG aqueous solution in both eyes. At predetermined time points (1, 2, 3, 4 and 6 h) after application, two rabbits from each group were euthanized. Both eyes (n = 4), frontal & occipital lobes, optic chiasm and peripheral organs (n = 2) (heart, lung, liver, spleen and kidney) were collected. Eyes were dissected while fresh into different tissues—cornea, bulbar conjunctiva, palpebral conjunctiva, aqueous humor, iris, ciliary body, trabecular meshwork, lens, vitreous humor, retina, eye cup and optic nerve. Due to the bioadhesive nature of the ME, the whole eyes were rinsed with PBS before dissection to remove any attached formulation. Also, blood was collected by cardiac puncture then separated into plasma. All the collected tissues and fluids were kept at −80°C until assayed for their drug contents. All eye tissues and fluids were weighed before processing. Due to the large weight of the peripheral organs and plasma, only 200 mg were used to assay their drug contents. All tissues were homogenized in 2 ml cold methanol in ice bath using tissue tearor (Biospec. Inc., Bartlesville, OK). Tissue homogenates were kept overnight at −20°C to dissolve PRG and precipitate protein then centrifuged for 20 min at 14000 rpm at −4°C (Eppendorf AG, 5804-R, Hamburg, Germany). The methanolic extract was separated and evaporated until dryness. The residue was dissolved in 0.5 ml methanol and assayed for its PRG content using the previously described LC-MS/MS method (SCIEX-5500, Framingham, MA) with a LOQ of 2.5pg/mg[29].

2.8. Stability studies of PRG ME

To ensure the stability of the formulation during shipping, transport and storage, a full stability study was conducted according to ICH guidance[42]. For stability during shipping and transport, freeze-thaw and centrifugation tests were conducted. While, for stability during storage, different storage conditions were selected to cover all expected storage conditions. Because the formulation is new and its best storage conditions have not yet been determined, we selected a wide storage temperature range that could cover the long-term, intermediate and accelerated storage conditions for formulations intended to be stored either in ambient temperature or in refrigerator[42]. If the ambient temperature is the intended storage condition, 25°C for 12 months, 30°C for 6 months and 40°C for 6months could serve as long-term, intermediate and accelerated storage conditions, respectively. While, if refrigeration is the intended storage condition, 5°C for 12 months, 25°C for 6 months, could serve as long-term and accelerated storage conditions, respectively.

During storage, degradation of active pharmaceutical ingredients (APIs) in formulations commonly occurs through several pathways including acidic and/or basic degradation, oxidative degradation, aqueous hydrolysis, photolysis and thermal degradation. PRG is highly stable against aqueous hydrolysis, photolysis and thermal degradation, and relatively stable against acidic hydrolysis[43]. In contrast, it is unstable under basic and oxidative conditions[43]. To evaluate the physical and chemical stability of PRG ME eye drops, three batches of the formulation were aseptically prepared and packaged in sealed white Nalgene opaque LDPE dropper bottles (Fisher Scientific, Fair Lawn, NJ). The filled bottles were stored in thermostatically controlled chambers at four different temperatures (5 ± 3°C, 25 ± 2°C, 30 ± 2°C and 40 ± 2°C) [44]. Per ICH guidance [G1A (R2)], sealed containers do not require to be stored at controlled relative humidity during stability study. Thus, humidity was not controlled because the containers were sealed. In addition, to easily monitor any change in the physical appearance of the ME, the formulation was also packaged in an additional transparent glass bottle at each evaluated temperature. The three batches were evaluated initially and at specified time intervals—1, 2, 3 and 6 months for formulations stored at 30°C and 40°C and 1, 2, 3, 6, 9 and 12 months for formulations stored at 5°C and 25°C—regarding their physical appearance, pH, droplet size, PDI, zeta potential, drug content and in vitro drug release.

Each experiment was performed in triplicate, and the results were calculated as mean ± SEM. Data were statistically analyzed using one-way analysis of variance (ANOVA) test followed by Tukey’s multiple comparisons test compared to the initially measured parameters of fresh samples. Statistical calculations were carried out using GraphPad Prism 10 software (GraphPad Software Inc., San Diego, CA).

2.8.1. Physical Stability Tests

2.8.1.1. Physical stress tests

2.8.1.1.1. Freeze-thaw test

Three batches of the prepared 0.6% PRG ME eye drops were subjected to 3 freeze-thaw stress cycles, freezing at −20°C for 48h then thawing at 25°C for 48h then assessed for their physical stability by evaluating their physical appearance[45, 46].

2.8.1.1.2. Centrifugation test

Three batches of 0.6% PRG-ME were subjected to ultracentrifugation at different speeds (5000, 10000, 20000, 40000, 50000 and 60000 rpm) for 30min/each using ultracentrifuge (SORVALL, WX Ultra Series Centrifuge, USA) and observed for phase separation, precipitation, creaming or cracking[46, 47].

2.8.1.2. pH evaluation

The pH values of each ME formulation were measured using a pH meter (Corning Inc., Corning, New York) at 25°C. One gram of each formulation was dispersed in 20mL of Milli-Q water, and then the pH was measured. The experiment was repeated three times, and the results were presented as mean ± SEM.

2.8.1.3. Average droplet size, polydispersity index (PDI) and zeta potential

The droplet size, PDI, and zeta potential of PRG ME were evaluated using Zetasizer (Nanoseries, nano-ZS, Malvern Instruments Limited, UK) after suitable dilution according to our previously published protocol[12]. All measurements were performed at 25°C. The results of three independent test runs were presented as mean ± SEM.

2.8.2. Chemical stability

2.8.2.1. Drug content of PRG ME formulation

To determine the drug content of PRG ME formulation, 100 mg of the ME was accurately weighed in a stoppered 10ml volumetric flask and diluted with a combination of Milli-Q water and absolute ethyl alcohol (3:7). The flasks were shaken for 10min, and the solution was filtered using 0.22μm membrane filters (Millipore, Billerica MA) then assayed for the total PRG content by UV-HPLC method described before[28]. The experiment was repeated three times, and the results were presented as mean ± SEM.

2.8.2.2. Degradation kinetics

To determine the order of PRG degradation reaction during storage, the drug content data was plotted according to the equation for zero-order, first-order, or second-order reactions. The reaction order that best fit the data was selected, the half-lives and the shelf-lives were calculated from the corresponding equation[48].

2.8.2.3. In vitro drug release

To detect any change in the release behavior due to storage at different temperatures, drug release was studied according to our previously published protocol[12]. Briefly, fast micro-equilibrium dialyzers (1500μl capacity) were used to which semipermeable regenerated cellulose membranes with molecular weight cutoff 5,000 Da were attached (Harvard Apparatus Co., Holliston, MA). One-hundred microliters of PRG ME were placed in the donor chamber. PBS (1.5 ml at 35° ± 0.5°C) was used as a release medium in the receptor chamber. The dialyzer was kept in thermostatically controlled shaker (35° ± 0.5°C and 50 rpm). At predetermined time intervals ranging from 0.25 to 24h, the entire medium in the receptor chamber was withdrawn and replaced by 1.5mL of fresh warm PBS at 35°C. The concentration of released drug was then determined by UV-HPLC. The experiment was repeated three times, and the results were presented as mean ± SEM. The release kinetics was determined by plotting cumulative amount drug released against time (zero order, [A] = [A]₀ − K₀ t), square root of time (Higuchi, [A] = [A]₀ − K_H t_1/2)), and the log of remaining drug against time (first order, log [A]= log [A]₀ − K₁ t/2.303). Also, to further investigate the release mechanism, Korsmeyer-Peppas ([A] = [A]₀ − K_KP tⁿ) model was applied to the data. Coefficients of determination were then obtained to know the model that best fit with the data[48–50] where:

• [A] = the drug concentration at time t;

• [A]₀ = the total drug concentration at time zero;

• K = the kinetic rate constant; and

• N = the diffusion exponent.

3. RESULTS AND DISCUSSION

3.1. Molecular modeling of PRG interaction with mouse, rabbit, and human CACNA2D1

Molecular modeling stands as a powerful computational strategy extensively used to uncover vital insights into protein-ligand interaction relevant to glaucoma[19, 51, 52]. When integrated with systems genetics, molecular modelling becomes a tool for evaluating the impact of genetic variants present across different species on the structural integrity of glaucoma-associated proteins[53]. This approach illuminates disease mechanisms and potential therapeutic targets by unveiling the repercussions of genetic variations on protein structure. Such insights are pivotal for assessing drug efficacy, specificity, and binding affinities in different animal models.

In modeling studies, the sequence identity between mouse and human CACNA2D1 across the entire protein length was determined to be 93.7%, while that between rabbit and human CACNA2D1 was 95.7%. The following docking scores were measured upon Glide docking of pregabalin (PRG) against mouse, rabbit, and human CACNA2D1: −10.820, −11.138, and −11.027 kcal/mol, respectively (Fig. 1). Binding residues are the individual amino acids within a biomolecule that directly participate in ligand interactions, while binding sites refer to the specific spatial regions wherein these interactions occur. In the literature, 5 Å from the ligand is accepted as the distance cut-off for identify a protein residue as binding residue[54, 55]. In mouse CACNA2D1, the following binding residues were predicted to be within 5 Å of PRG: HIS167, TRP205, GLU206, VAL207, GLY209, ALA215, TYR217, TRP223, TYR236, VAL238, ARG241, PRO242, TRP243, TYR450, ASP452, ALA453, LEU454, VAL459, THR461, GLY489, VAL490, and ASP491 (Figs. 1A & 1B). Note that HIS167, GLU206, GLY209, PRO242, and VAL490 are absent in the diagram due to the limitations of Maestro in representing all residues within the 3D binding pocket of mouse CACNA2D1 in 2D images. All binding residues in the protein-ligand complex are presented in the corresponding video analysis file (Supplemental Video 1; mouse). In rabbit CACNA2D1, the binding residues predicted to be within 5 Å of PRG were HIS169, TRP207, VAL209, ALA217, TYR219, PRO221, TRP225, TYR238, VAL240, ARG243, TRP245, TYR452, ASP454, ALA455, LEU456, VAL461, THR463, and ASP493 (Figs. 1C & 1D). Note that HIS169, PRO221, VAL240, TRP245, VAL461, THR463, and ASP493 are absent from the diagram due to the limitations of Maestro in representing 3D structures of rabbit CACNA2D1 in 2D images. All binding residues in the protein-ligand complex are presented in the corresponding video analysis file (Supplemental Video 2; rabbit). In human CACNA2D1, the predicted binding residues within the same radius were TRP205, VAL207, ALA215, TYR217, PRO219, TRP223, TYR236, VAL238, ARG241, TRP243, TYR450, ASP452, ALA453, LEU454, VAL459, THR461, and ASP491 (Figs. 1E & 1F). Note that PRO219, THR461, and ASP491 are absent from the diagram due to the limitations of Maestro in representing 3D binding site of human CACNA2D1 in a 2D plane. As for mouse and rabbit, all the binding residues in the protein-ligand complex are presented in the corresponding video analysis file (Supplemental Video 3; human). Lastly, readers must be aware that the order of amino acids is shifted by +2 in the binding pocket of rabbit CACNA2D1 compared to those of both human and mouse CACNA2D1, according to their canonical sequences published by the National Centre for Biotechnology Information (NCBI)[56] with no affect on docking scores.

Fig. 1.

(A,B) PRG interacts with mouse CACNA2D1 via salt bridges (ARG241 and ASP491), cation-π exchange (TRP243), and hydrogen bond network (TRP243, ALA453, and ASP491) with a docking score of −10.820 kcal/mol which is indicative of high predicted binding affinity. (C,D) PRG interacts with rabbit CACNA2D1 via salt bridges (ASP454), cation-π exchange (TYR219), and hydrogen bond network (TYR238, ARG243, TRP245, TYR452, and ASP454, and a docking score of −11.138 kcal/mol which is also indicative of high predicted binding affinity. (E,F) PRG interacts with human CACNA2D1 via salt bridges (ARG241 and ASP452), cation-π exchange (TRP205, TYR217), and hydrogen bond network (ARG241, TRP243, ALA453, and ASP452). The corresponding docking score of −11.027 kcal/mol is also indicative of high predicted binding affinity.

Molecular docking analyses substantiated the robust binding affinity of PRG against the CACNA2D1 protein of mouse, rabbit, and human origins (Figs. 1A & B, 1C & 1D, and 1E & 1F, respectively), reinforcing its potential therapeutic relevance. Across these diverse species, the residues constituting the binding pocket (defined within 5 Å of PRG) displayed remarkable similarity, albeit with a minor offset of +2 residue numbers in rabbit CACNA2D1 (e.g., ARG241 in human and mouse CACNA2D1 corresponded spatially to ARG243 in rabbit CACNA2D1). Notably, a consistent set of crucial intermolecular interactions was discernible among all three species, encompassing a hydrogen bond between an arginine residue and the carboxylic acid moiety of PRG, as well as a corresponding hydrogen bond involving a tryptophan residue and the same carboxylic acid group. This observation is consistent with other experimental studies highlighting the importance of these arginine and tryptophan residues for ligand binding in CACNA2D1[19, 57, 58].

In addition to the mouse model, this investigation has ambitiously extended its computational scope to encompass other animal models, including human patients, to unravel the potential translational implications of PRG-mediated IOP reduction. This comparative analysis led to the discovery of a previously unidentified hydrogen bond between an alanine residue of CACNA2D1 and the carboxylic acid of PRG in mouse and human CACNA2D1, but not in rabbit CACNA2D1. Equally noteworthy is the identification of a species-specific intermolecular force, a π-cation bridge involving a tyrosine residue and the amine group of PRG, found in rabbit and human CACNA2D1, but not in mouse CACNA2D1. Together, the data demonstrate that PRG maintained a virtually identical 3D conformation within the binding pocket across all three species. The culmination of high sequence homologies and closely mirrored binding modes reinforces the plausibility that PRG-induced CACNA2D1 antagonism and subsequent IOP reduction demonstrated in rabbit and mouse models, may indeed manifest within humans. This is especially exciting within the clinical context of glaucoma because existing therapies for lowering IOP in primary open angle glaucoma patients face challenges, such as undesirable side effects and failure to address the underlying disease causes[59, 60]. Even FDA-approved topical prostaglandin analogue options, such as Vyzulta (latanoprostene bunod, 0.024%) and TRAVATAN Z (travoprost, 0.004%), can lead to adverse events including hyperemia and persistent pigmentation of ocular structures. Furthermore, frequent dosing of short-acting medications can escalate costs and jeopardize patient adherence[61]. Neglecting the complexity of glaucoma’s pathophysiology, current therapies overlook cases such as normal tension glaucoma (NTG), highlighting the necessity for identifying novel drugs with reduced side effects, prolonged efficacy, and diversified targets to better combat this condition[34]. Our modeling outcomes suggests that PRG could exhibit efficacy in human for reducing IOP. This raises the prospect of its transformation into an innovative therapeutic approach for primary open-angle glaucoma.

3.2. Quantitative structure-property relationships (QSPR) analysis of PRG

QSPR (Quantitative structure-property relationships) is a computational method used in pharmaceutical research in which mathematical models link the structural attributes of a molecule to its physical or biological properties [62]. When applied to the topical application and corneal permeability of PRG, QSPR involves gathering its physiochemical data, such as its molecular weight and total surface area, to establish a quantitative relationship between these descriptors and corneal permeability (Table 1).

Table 1.

Quantitative structure-property relationships (QSPR) analysis of PRG and its physiochemical properties related to rule of thrumb for optimal ocular permeability and absorption for topical eye drops roposed by Karami et al.[14, 63]

QSPR Parameters	cLogDpH7.4a	cSpH7.4b (M)	ΔGo/wc (kJ/mol)	TPSAd (Å2)
Calculated value	−1.77	0.25	10.1	66.93
ROxe threshold	< 4.0	> 1 uM	< 20	< 250
Favorablility of RO x e	Yes	Yes	Yes	Yes

^acLogD_pH7.4: The distribution coefficient for a drug at pH 7.4, which is the pH of the tear film

^bcS_pH7.4: The aqueous solubility of a drug at pH of 7.4

^cΔG_o/w: The change in Gibbs free energy of distribution/partitioning of a drug, where higher the value the higher the hydrophilicity

^dTPSA: The topological surface area of a drug

^eRO_x: The rule of thrumb for optimal ocular permeability and absorption for topical eye drops proposed by Karami et al[14, 63]

cLogD_pH7.4 is the distribution coefficient for a drug at pH 7.4, which is the pH of the tear film. A cLogD_pH7.4 of −1.77 indicates that the drug is slightly more hydrophilic than hydrophobic in the context of tear film pH. At first glance, this may seem counter-intuitive because previous research has indicated that corneal permeability tends to increase as the distribution coefficient, represented by log D or hydrophobicity, increases[64]. However, Karami et al. conducted a study in which they discovered that out of 145 approved topical ophthalmic drugs, 135 had cLogD_pH7.4 values below 4.0. Their research team proposed an interesting perspective: while hydrophilicity may not be advantageous for transcellular transport, it could potentially enhance paracellular transport through diffusion across the corneal and non-corneal water-filled pores. Similarly, Karami et al. discovered that 141 out of the 145 FDA-approved topical ophthalmic drugs had cS_pH7.4, which is the aqueous solubility of a drug at pH of 7.4, greater than 1 μM. The team then derived ΔG_o/w, which is another solubility parameter describing the change in Gibbs free energy of distribution/partitioning of a drug, from the same list and discovered that 140 drugs had ΔG_o/w less than 20 kcal/mol, where lower the ΔG_o/w the higher the hydrophilicity. These three solubility parameters all suggest that paracellular diffusion driven by hydrophilicity of a drug may be just as important as a drug’s transcellular diffusion driven by hydrophobicity in the context of ocular permeability. PRG’s cLogD_pH7.4, cS_pH7.4, and ΔG_o/w all satisfy RoX proposed by Karami et al and are expected to be in favor of optimal absorption when delivered topically. Lastly, TPSA is the topological polar surface area. TPSA of a drug molecule has been observed to significantly influence their absorption across various biological cell membranes, including membranes of the cornea, brain, and nerve cells within the central nervous system[64]. Studies indicate that drugs with a TPSA of less than 60 Ų are typically fully absorbed, while those with a TPSA exceeding 140 Ų experience limited permeation. PRG’s TPSA of 66.93 Ų is less than the cut-off TPSA of 250 Ų proposed by Kamari et al. and is expected to be in favor of crossing the corneal membrane. All these quantitative qualifications support PRG as a promising IOP-lowering medication via topical delivery.

3.3. Preparation of PRG multilayered ME eye drops

The selected pseudo-ternary phase diagram that was used to prepare PRG ME is illustrated in Supplemental Fig. 1A.The ingredients of PRG ME were carefully selected to produce biocompatible eye drops [12]. The oil phase consisted of Labrafac Lipophile WL1349, which is a medium chain triglyceride ester of saturated fatty acids that decreases the possibility of rancidity (i.e., oil oxidation) upon storage due to the absence of unsaturated bonds which is considered a point of attack by free radicals. The presence of such oil helps to improve the physical stability and prolong the shelf life of the ME[65]. In addition, the lipophilic surfactants mixture that is responsible for the formation of the primary W/O ME consists of capryol 90 and soybean lecithin, (1:1) weight ratio. Capryol 90 is a propylene glycol ester of a medium chain fatty acid (caprylic acid) that has a similar chemical composition to the oil phase and resulted in a high level of compatibility between them. Soybean lecithin was selected as a surfactant because of its high biocompatibility due to its phospholipid nature that resembles the composition of the biological membranes[66]. The external aqueous phase of PRG ME consists of a hydrophilic surfactant mixture (Labrasol and Cremophor EL, 1:1), a cosurfactant (propylene glycol) and MilliQ water in which the bioadhesive polymer (Carbopol 981) is incorporated. Supplemental Fig. 1B presents a cartoon of PRG multilayered ME illustrating the three layers of the ME, water in oil in water. Supplemental Fig. 1C is an image of PRG in the W/O/W ME taken with a TEM microscope that illustrates the distinct multilayered structure of the PRG ME. In addition, magnified ME droplets consist of several nuclei of the internal aqueous phase entrapped in the intermediate oil phase, which is surrounded by the external aqueous layer (Supplemental Fig. 1D).

3.4. Pregabalin ME in vivo Evaluations

3.4.1. Dose finding and EC₅₀ determination in Dutch belted rabbits

Dutch belted rabbits are widely used as animal model to test the IOP lowering effect of different drugs because they spontaneously develop elevated IOP post-puberty[41, 67–69]. For this reason, Dutch belted rabbit strain was selected as an animal model to evaluated the IOP lowering ability of PRG ME. Our published data demonstrated that 0.6% PRG ME decreased IOP by greater than 40% in Dutch belted rabbits[12]. To elaborate on these findings, we sought to determine the minimal concentration of PRG required to produce the maximal reduction in IOP. To do so, we tested seven different PRG ME concentrations (0.1–0.7% w/w) in Dutch belted rabbits. Fig. 2A presents the % IOP reduction after a single drop application of PRG ME or blank ME formulations for both treated and control eyes of Dutch belted rabbits (n=5), respectively. These data demonstrate that after a single topical application, ME eye drops containing different PRG concentrations reduce IOP in rabbits in a dose-dependent manner. As illustrated in Table 2, the lowest tested PRG concentration (0.1% w/w) induced a percent IOP reduction of 24.0 ± 0.6 % while the highest tested concentration (0.7% w/w) induced a percent IOP reduction of 35.1 ± 2.5 % that returned to baseline at 6.0 ± 0.0 h, 33.2 ± 1.9 h after application respectively. Within that range, 0.6% PRG ME is the lowest concentration that produced the maximum IOP-lowering effect (the optimal PRG dose). The data also reveal that 0.6% PRG ME is the saturation concentration above which there is no further increase in the drug response. Statistical analysis of different pharmacodynamic parameters (Supplemental Table 1) illustrates that there is no significant difference in all the calculated PD parameters between 0.6 and 0.7% PRG ME including drug response (% IOP reduction, p > 0.99), the time of maximum response (T_max, p = 0.62), the time required for IOP to return to baseline value (T_end, p = 0.96) and total area under % IOP reduction versus time curve (AUC, p = 0.99). In contrast, there is a significant difference in both T_end (p < 0.0001) and AUC (p < 0.0004) between 0.5 and 0.6% PRG ME. However, there is no significant difference in either % IOP reduction at T_max (p = 0.96) or T_max (p > 1.0) between 0.5 and 0.6%. This nonsignificant difference in the % IOP reduction and T_max may be due to the drug which is incorporated in the external aqueous phase (70% of the drug load) that is available in both formulations in enough concentration to produce a rapid onset that reach its maximum after ~3 h (Table 2). Because the values of T_end and AUC depend mainly on the amount of PRG that is incorporated in the internal aqueous phase (30% of the drug load), the significant difference in both T_end and AUC between 0.5 and 0.6% is expected to be mainly due to the lower concentration of PRG in the internal aqueous phase of the 0.5% PRG ME that is insufficient to maintain the sustained drug release for longer period of time.

Fig. 2.

(A) Percentage IOP reduction/time profiles after topical application of a single dose of PRG-loaded MEs containing different concentrations of the drug ranging from 0.1 to 0.7% w/w to the right eyes and the blank formulations to the left eyes of Dutch belted rabbits (mean ± SEM; n = 5). Drops containing PRG (0.1–0.7% w/w) reduced IOP in rabbits in a dose dependent manner after a single topical application. There is no significant difference in all the calculated PD parameters between 0.6 and 0.7% PRG ME. Therefore, the PRG minimal concentration required to produce the maximal reduction in IOP is selected to be 0.6%. (B) AUC (%. h) versus PRG concentrations plot. The calculated EC₅₀ of PRG is 0.5% w/w which is equivalent to 31.4 mM. (C) Percentage IOP reduction/time profiles after topical application of a single dose of 0.6% PRG ME to the right eyes and the blank formulations to the left eyes of three mouse strains, B6, BXD14 and BXD44 (mean ± SEM; n = 6). A single dose of 0.6% PRG ME eye drops induce a markedly reduced IOP (> 37%) in the three mice strains—B6, BXD14 and BXD44—which does not return to baseline until >31 h post-application. (D) Percentage IOP reduction/time profiles after topical application of a single dose (30 μL) of 0.6% PRG ME or latanoprost (LTP) eyedrops 0.005% to the right eyes and the blank formulation of each to the left eyes of Dutch belted rabbits (mean ± SEM; n = 5 each). There is no significant difference in the IOP lowering ability of the drugs (% IOP reduction, p=0.2705) and the time of maximum response (T_max , p= 0.6491) between the two formulations. However, there is a significant difference between the two formulations in both T_end (p < 0.0001) and AUC (p = 0.0013).

Table 2.

Pharmacodynamic parameters after topical application of single dose of seven different concentrations of PRG^a ME^b eye drops to Dutch belted rabbits^c

Pharmacodynamic parameters	Concentration of PRG in ME Eye Drops (% w/w)
	0.1%	0.2%	0.3%	0.4%	0.5%	0.6%	0.7%
Baseline IOP d	19.7 ± 0.3	20.2 ± 0.5	19.9 ± 0.6	19.8 ± 1.0	19.8 ± 0.7	20.5 ± 0.7	20.3 ± 0.7
IOP at T max e	15.0 ± 0.3	14.5 ± 0.4	15.0 ± 0.0	14.0 ± 0.4	13.2 ± 0.2	13.1 ± 0.1	13.1 ± 0.4
ΔIOP	−4.7 ± 0.1	−5.7 ± 0.4	−4.9 ± 0.6	−5.8 ± 0.9	−6.6 ± 0.8	−7.4 ± 0.6	−7.2 ± 0.7
% IOP reduction at T max	24.0 ± 0.6	28.0 ± 1.8	24.6 ± 2.0	28.6 ± 3.2	33.0 ± 2. 7	36.0 ± 1.8	35.1 ± 2.5
Tmax (h)	1.6 ± 0.3	3.2 ± 0.5	2.6 ± 0.3	4.8 ± 0.9	3.2 ± 0.4	3.4 ± 0.4	5.0 ± 1.3
Tendf (h)	6.0 ± 0.0	10.0 ± 0.6	10.0 ± 0.6	19.2 ± 2.0	23.6 ± 1.0	34.8 ± 1.0	33.2 ± 1.9
AUCg (%. h)	69.5 ± 3.1	160.8 ± 14.8	154.0 ± 16.2	289.2 ± 44.8	390.1 ± 49.3	720.3 ± 53.5	673.2 ± 82.1

^aPRG: pregabalin

^bME: microemulsion

^cData are expressed as mean ± SEM; n = 5

^dIOP: intraocular pressure

^eT_max: time to maximum response in hours

^fT_end: time to end of response in hours

^gAUC (%. h): total area under % IOP reduction versus time curve

Fig. 2B represents the AUC versus PRG concentrations values for all the tested formulations. The EC₅₀ was calculated from AUC versus PRG concentration curve. Because AUC represents both the amplitude of IOP reduction and duration of action, AUC is the best representation of PRG IOP-lowering efficacy. The EC₅₀ represents the concentration at which the drug exerts half of its maximal response. The calculated EC₅₀ of PRG as an IOP-lowering medication is 0.5% w/w, which is equivalent to 31.4mM of PRG.

3.4.2. Efficacy study on different mice strains

Based on the results of dose finding study, 0.6% PRG ME (PRG optimal dose) was used to test PRG IOP-lowering efficacy on different mice strains [C57BL/6J (B6), BXD14 and BXD44], all strains are from the BXD family of recombinant mice, the largest murine genetic reference panel available[11]. Fig. 2C demonstrates that a single dose of 0.6% PRG ME eye drops induced a markedly reduced IOP (> 37%) in all three mice strains evaluated—B6, BXD14 and BXD44—which does not return to baseline until >31 h post-application. Table 3 lists the calculated PD parameters including % IOP reduction, T_max, T_end and AUC. Statistical analysis demonstrated that, there is no significant difference in drug response (% IOP reduction) between B6 and BXD44 (p = 0.94) and between BXD14 & BXD44 (p = 0.06) (Supplemental Table 2). In contrast, there is a significant difference in % IOP reduction values between B6 and BXD14 (p = 0.03). However, there is no significant difference in all other calculated PD parameters between B6, BXD14 and BXD44 including T_max (p = 0.48), T_end (p = 0.20) and AUC values (p = 0.21) (Supplemental Table 2). Although there is some variability in the amplitude of IOP reduction between the mice strains, they all respond very well to PRG. From these data we can conclude that PRG ME is an effective IOP-lowering medication in different mice strains with the ability to maintain its IOP-lowering effect for more than one day after a single topical application. Collectively, IOP data from both Dutch belted rabbits and different mice strains strongly support the use of the ME platform as a once daily eye drops for glaucoma therapy because of its ability to extend the duration of action of PRG.

Table 3.

Pharmacodynamic parameters after topical application of single dose of 0.6% PRG^a ME^b eye drops to different mice strains^c

Pharmacodynamic parameters	Mouse strain
	B6	BXD14	BXD44
Baseline IOP d	16.5±0.4	18.0±0.6	16.7±0.6
IOP at T max e	10.3±0.4	9.8±0.3	10.2±0.3
ΔIOP	−6.3±0.3	−8.2±0.5	−6.5±0.6
% IOP reduction at T max	37.9±1.3	45.2±1.7	38.8±2.2
Tmax (h)	4.2±0.8	3.2±0.3	4.3±0.9
Tendf (h)	33.3±0.4	33.0±0.7	31.7±0.8
AUCg (%. h)	844.8±54.1	973.1±64.5	812.1±73.1

^aPRG: pregabalin

^bME: microemulsion

^cData are expressed as mean ± SEM; n = 6

^dIOP: Intraocular pressure

^eT_max (h): time to maximum response in hours

^fT_end (h): time to end of response in hours

^gAUC (%. h): total area under % IOP reduction versus time curve

3.4.3. Efficacy study in comparison to market leader commercial glaucoma medication

To compare PRG ME IOP-lowering efficacy to one of the current glaucoma eye drops—latanoprost eye drops (0.005%)—two groups of Dutch belted rabbits were dosed in their right eyes with 30 μL of each formulation while the left eyes received the blank formulation and served as a control. Fig. 2D presents the % IOP reduction after a single drop application (30 μL) of ME containing the optimal PRG dose (0.6%) or latanoprost eye drops (0.005%) for Dutch belted rabbits. The figure demonstrates that after a single topical application, 0.6% PRG ME eye drops induced a percent IOP reduction of 28.1 ± 1.73 % while latanoprost eye drops (0.005%) induced a percent IOP reduction of 24.9 ± 1.46 % that returned to baseline at 31.0 ± 0.683 h and 12.67 ± 0.67 h after application, respectively. In addition, the time of maximum response after topical application of 0.6% PRG ME eye drops and latanoprost eye drops (0.005%) were 3.5 ± 0.67 h and 3.0 ± 0.58 h, respectively. As illustrated in Fig. 2D, total area under % IOP reduction versus time curve was 588.5 ± 51.1 %. h and 192.5 ± 21.8 %. h for 0.6% PRG ME eye drops and latanoprost eye drops (0.005%), respectively. Statistical analysis of pharmacodynamic parameters demonstrates that there is no significant difference in both drug responses (% IOP reduction, p=0.2705) and T_max (p= 0.6491) between 0.6% PRG ME eye drops and latanoprost eye drops (0.005%). However, there is a significant difference between the two formulations in both T_end (p < 0.0001) and AUC (p = 0.0013). These results illustrate that 0.6% PRG ME has a comparable IOP lowering effect as latanoprost eye drops (0.005%), while it is superior to latanoprost with regard to its long duration of action due to the bioadhesive behavior of the formulation [70].

3.4.4. Bioadhesion and corneal residence time assessment

We selected a bioadhesive polymer as the outermost aqueous phase to enhance the ability of the ME to adhere to the eye [70]. To demonstrate the corneal bioadhesive property of the ME system and its ability to increase the drug corneal contact time with extended release properties we engineered acridine orange (AO) ME. We chose AO as a surrogate for PRG because it has roughly equivalent physiochemical characteristics to PRG (both PRG and AO are BCS class I with high aqueous solubility and good permeability and have similar molecular weights); hence, AO ME data will reflect that of the PRG ME [71]. To assist in the ability to quantify bioadhesion, AO has an added feature in that it fluoresces when illuminated with blue light, making it readily visible. AO ME or PRG ME were topically dosed to one eye of Dutch belted rabbits (n=5) while the fellow eye received either AO in PBS or PRG in water as a control formulation. As is evident by the images of Dutch belted rabbits eye surface presented in Fig. 3A, the fluorescent ME remains on the surface of the eye for at least 24 h after a single dose. In contrast, the fluorescent dye dosed in PBS nearly disappeared from the control eye within one hour. Fig. 3B presents the quantitative estimation of fluorescence intensity from the photos of rabbit eyes. Eyes treated with AO ME maintain a higher level of fluorescent intensity for at least 24h (6.1 ± 0.42 A.U.) compared to the dye in PBS treated eye (0.4 ± 0.02 A.U.). In addition, Fig. 3C demonstrates that after 24 h, AO measured concentration in tears was 28.1 ± 10.9 ng/μl in the eye received AO ME, while there was no detected dye in the tears of the eye dosed with AO in PBS. In parallel studies we measured PRG in tears after dosing with PRG ME formulation. After 24 h, the PRG tears level was significantly higher in the eye that received PRG ME (57.7 ± 12.9 ng/μl) compared to the fellow eye that received PRG in water (0.45 ± 0.27 ng/μl; Fig. 3D). To visualize the ability of AO to enter the eye, we imaged for fluorescence in histological sections taken from the rabbits after topical dosing with AO ME. Figs. 3E & 3F demonstrate that the fluorescent AO was observed in all layers of the cornea. Similarly, AO was also able to reach anterior segment tissues that are associated with IOP modulation [ciliary body (CB) and trabecular meshwork (TM)] (Figs. 3G & 3H), both of which express the PRG-target protein CACNA2D1, as we previously demonstrated in mice, Dutch belted rabbits and human donor eyes[11, 12]. These data demonstrate that the ME is able to maintain the inherent high permeability of AO, similar to PRG[72, 73], faciliting access to target intraocular structures. In conclusion, the ME controls AO entry into the eye and maintains the formulation on the eye surface for up to 24 h with continuous drug penetration through the cornea to the anterior segment structures that modulate IOP.

Fig. 3.

Bioadhesion and corneal residence time assessment of acridine orange (AO) ME and AO in PBS (A-C) or PRG ME and PRG in water (D) in Dutch belted rabbits (mean ± SEM; n = 5). (A) Fluorescence images of Dutch belted rabbits eye surface at different time intervals after single application of AO ME or AO in PBS. Acridine orange loaded into the ME remains on the ocular surface for at least 24 h. In contrast, the dye in PBS, is washed away within the first hour after dosing. (B) Quantitative measurement of the fluorescence intensity on the eye surface using ImageJ software prove that ME (green) maintains an elevated amount of dye for at least 24 h which is consistent with the measured concentration of AO in tears (C). (D) Similarly, after 24 h, PRG concentration in tears is significantly high in the eye received PRG ME (red) compared to the eye received PRG in water (black). (E,F) Confocal microscope fluorescent immunohistochemistry (FIHC) images of the cornea, (G,H) ciliary body (CB) and trabecular meshwork (TM) for AO in PBS and AO ME treated Dutch belted rabbit eye, respectively. Acridine orange is labeled in green, CACNA2D1 in red, beta-actin in blue and DAPI in white. FIHC images (E-H) demonstrate the ability of the ME system to maintain and control the inherent high permeability of AO molecule as a BCS class I drug and its ability to deliver it to different eye tissues (cornea, CB and TM).

These data demonstrate that the ME processes a strong bioadhesion force that enables it to remain on the eye surface for longer duration, greatly improving its corneal residence time. The improved bioadhesion of the ME formulation is likely due to the careful selection of the ME excipients, especially Carbopol 981. Incorporation of Carbopol 981 as a bioadhesive and film forming polymer[74] in the external aqueous phase of the W/O/W ME allows the formulation to adhere to the eye surface (cornea and conjunctiva) and acts as a drug reservoir that continuously releases PRG for more than 24 h. This excellent bioadhesion of Carbopol 981 is due to its ability to form a strong H-bonding with the mucin layer on the eye surface. Among the ME excipients that help to improve the bioadhesion and subsequently the corneal residence time, are the nonionic surfactants that help Carbopol 981 to better interact with mucin layer by improving the wetting, swelling and flexibility of its polymer chains resulting in exposure of more adhesive sites and facilitating better bioadhesion[75]. In addition to its bioadhesiveness, Carbopol 981 also has a pH-triggered in situ gelling behavior that allows it to undergo a sol-to-gel transition in aqueous solution as the pH raised above its pK_a of about 5.5[76]. Because the PRG ME has a pH of 5.3 ± 0.1, Carbopol 981 will be in its solution state (low viscosity) that will enable it to be easily dispensed from a dropper onto the eye. Upon installation on the eye surface and due to the tears buffering capacity, the pH of the ME will be raised to the normal physiological pH (7.0)[77, 78] that results in the conversion of the ME into its gel state (higher viscosity) on the eye surface. This sol-to-gel transition occurs immediately after ME application to the eye surface and it is highly advantageous in preventing the rapid drainage of the ME from the eye surface and allowing enough time for a strong bioadhesion between Carbopol 981 and mucin to occur[79]. These data are consistent with our earlier findings in which we demonstrated that PRG Carbopol-loaded ME has significantly higher in vitro bioadhesive force (p < 0.01) when mixed with gastric mucin type II dispersion compared to the corresponding control formulation that lack the ME[12]. These in vivo bioadhesion data clearly demonstrate the ability of the ME system to support the once daily use of the PRG ME as a new IOP-lowering eye drops by extending the precorneal drug residence time up to 24 h with continuous diffusion of drug molecules through the cornea and subsequent extended duration of IOP-lowering effect.

3.4.5. Efficacy, safety and drug-tissue biodistribution studies after 60 days of daily application

Because glaucoma is a chronic disease, the ME formulation must be safe during prolonged use. Therefore, we evaluated the safety and efficacy of the PRG ME formulation after 60 days of daily dosing in Dutch belted rabbits. Fig. 4A presents the IOP profile of the rabbits over 60 days of the study. IOP was measured at baseline, at 3 h post-dosing (T_max) and prior to each subsequent dose at 24 h. Three hours after the initial dose of PRG ME formulation, the IOP of rabbits was decreased from its baseline of 21.8 ± 0.6 mmHg to 14.6 ± 0.4 mmHg. After the first day and with continued dosing with PRG ME, IOP gradually continued to decrease until reaching its minimum (saturation state) at 30 days (Fig. 4A; p < 0.0001 at T_max compared to baseline IOP of 21.8 ± 0.6 mmHg; n = 5). Subsequently, the IOP versus time curve demonstrated a reduction in IOP that ranged between 35.3% to 41.3%, which indicates that IOP is well controlled by the ME throughout the day. In contrast, the IOP of control eyes that received the blank ME remained elevated with minor fluctuations due to normal diurnal variation of the IOP[80]. This observed gradual decrease in IOP between days 0 and 30 in eyes treated with PRG ME may be due to the gradual accumulation of PRG at its target tissue until reaching the saturation state of CACNA2D1[11]. Alteratively, it could indicate a remodeling of an ocular structure involved in IOP regulation. Collectively, these data demonstrate that the formulation maintains its IOP-lowering ability using a single daily dose without the development of any observed side effects, irritation, or decrease in the drug response upon prolonged use (tachyphylaxis). Corneal histopathological examination was used to confirm the safety of the ME formulations (Figs. 4B–E). These data demonstrate that corneas were healthy with normal histological structure and all corneal layers maintained their full thickness and architecture. Slit-lamp biomicroscopical examinations after 60 days of dosing (Figs. 4F & 4G) demonstrated that corneas of the rabbits appeared transparent with a smooth epithelial surface, clear lenses and no cells or flare were detected in the aqueous humor, all indications that the ME was well tolerated by the eye. Figs. 4H & 4I present the funduscopic examination of rabbits eyes after 60 days of daily dosing with the ME formulation. The fundus images present with healthy optic nerve heads and normal vasculature. Also, the vitreous humor was within normal limits and didn’t present with opacities or any suspended particles. This excellent safety profile of the ME formulation is mainly attributed to the careful selection of the ME excipients with previously demonstrated safety and biocompatibility[66, 81–84].

Fig. 4.

(A) IOP profiles of Dutch belted rabbits for 60 days of daily dosing with blank ME and PRG ME formulations. The graph demonstrates that after the first daily dose of PRG ME, IOP remained in the physiological range for 60 days without any decrease in the drug response during prolonged use (mean ± SEM; n = 5). In contrast, the eye dosed with blank ME demonstrates that IOP remained at its high baseline value (22.0 ± 0.45) with minor fluctuations. (B, C) Images of histological examination of rabbit corneas after 60 days of daily dosing of blank ME and (D, E) of PRG ME demonstrate normal corneal architecture, including the epithelium (Epi). (F, G) Slit-lamp biomicroscopic examination of rabbit eyes after 60 days of daily dosing with blank ME and PRG ME formulations, respectively. Both photos illustrate that the corneal epithelium is smooth, the aqueous humor is clear, and the lens appears normal, which collectively demonstrates the safety of the PRG ME formulation with prolonged use. (H, I) Fundus images of rabbit eyes after 60 days of daily dosing with blank ME and PRG ME, respectively. Images demonstrate that the optic nerve head looks healthy with normal vasculature and the vitreous humor demonstrates no opacities.

Figs. 5A & 5B present the drug ocular tissue biodistribution after 60 consecutive days of topical single daily dosing with the blank ME and PRG ME, respectively. Due to the expected very low PRG tissues levels, in the current study we used a highly sensitive LC-MS/MS assay method with limit of quantification of 2.5pg/mg PRG assay in various eye tissues, the frontal lobe, peripheral organs and plasma [29]. It is evident that the control eye which dosed with blank ME accumulated a miniscule amount of the drug, while the treated eye that dosed with PRG ME accumulated higher concentrations of PRG in all parts of the eyeball. This enhancement of PRG entry into the eye was likely attributed to the bioadhesion property of the ME, which localized the drug in the treated eye and limited its crossing to the fellow eye. PRG was also accumulated in higher concentrations in eye tissues compared to the eye fluids (aqueous humor and vitreous humor). The data demonstrated that there are higher PRG accumulations in some eye tissues over the others which may be due to the high expression of PRG-target protein, CACNA2D1 in those tissues [11]. Similar to the control eye, miniscule drug concentrations were detected in frontal lobe of the brain and also in peripheral organs such as lungs, heart, and spleen. However, no PRG is detected in plasma and other peripheral organs such as kidneys and liver (Fig. 5C). These outcomes may be also due to the bioadhesion characteristics of the ME that enhanced PRG entry into the eye, which decreased its nasolacrimal drainage and subsequent systemic absorption. We can conclude from the 60 day dosing study that PRG ME was able to maintain its IOP-lowering effect without any decrease in its drug response upon prolonged use and that it had an excellent safety profile.

Fig. 5.

(A, B) Drug/eye tissue biodistribution after 60 days of daily dosing of blank ME and PRG ME formulations, respectively, demonstrate that the control eye dosed with the blank ME only accumulates a miniscule amount of PRG, while PRG is distributed through various tissues of the treated eye. (C) PRG level in brain, peripheral organs and plasma after 60 days of daily dosing. There is a miniscule drug concentration within brain and peripheral organs such as lungs, heart and spleen. In contrast, no drug was detected in other peripheral organs such as kidneys, liver and plasma.

3.4.6. Pilot pharmacokinetics study after ocular application

This pilot PK study was performed as an exploratory evaluation to measure the concentration of PRG in various eye tissues after a single topical dose. Fig. 6 presents both the anterior and posterior segment tissues concentration-time curves of PRG after a single ocular application of the tested formulations (PRG in water and PRG ME). The data demonstrated that both PRG formulations are able to deliver PRG to various eye tissues, which is likely due to the inherent high permeability of PRG molecule as a BCS class I drug[85, 86]. However, the data illustrate that PRG ME formulation is capable of improving significantly PRG bioavailability in all eye tissues except cornea, aqueous humor and lens (data table not shown) compared to the control formulation, which is evidenced by the significant higher PRG level in all eye tissues (except cornea, aqueous humor and lens) that received PRG ME. The superiority of the ME ability to deliver significant higher level of PRG to both anterior and posterior tissues may be attributed to the high concentration of the nonionic surfactants contents in the ME, which act as strong permeation enhancers and allow for more PRG to penetrate the eyeball and reach higher concentrations in almost all eye tissues[87]. The results also demonstrate that the systemic absorption of PRG after the topical application of either PRG ME or PRG in water formulations is nearly negligible and is not expected to cause any systemic side effects. Fig. 7 demonstrates that PRG concentration in frontal lobe, occipital lobe, optic chiasm and peripheral organs such as heart, lung, liver, spleen and kidney is below 0.24 ng/mg after 1, 2, 3, 4 and 6 h post application. This is because of the very low dose of the formulations used in this study (30μl/eye, equivalent to 0.36 mg of PRG for both eyes) compared to the commercially available oral dosage form of PRG (Lyrica^®, dose 150–600 mg/day) which is 400–1600 times greater than the ophthalmic daily dose of the PRG ME[88].

Fig. 6.

Pilot ocular pharmacokinetic study after a single dose of PRG ME formulation and the control (PRG in water) for 6h (mean ± SEM; n = 4). PRG in anterior and posterior eye tissues concentration-time curves (A) cornea, (B) bulbar conjunctiva, (C) palpebral conjunctiva, (D) aqueous humor, (E) iris, (F) ciliary body, (G) trabecular meshwork, (H) lens, (I) vitreous humor, (J) retina, (K) eye cup, and (L) optic nerve. PRG ME formulation improved PRG bioavailability in all eye tissues compared with the control (PRG in water).

Fig. 7.

Pilot pharmacokinetic study after a single dose of PRG ME formulation and the control (PRG in water) for 6h (mean ± SEM; n = 2). PRG in central and peripheral organs concentration-time curves (A) brain, (B) occipital lobe, (C) optic chiasm, (D) heart, (E) lung, (F) liver, (G) spleen, (H) kidney, and (I) plasma. Trace concentration of PRG is detected in all central and peripheral organs (< 0.24ng/mg). PRG level in the central organs that are directly connected to the eyeball through the optic nerve such as brain, occipital lobe and optic chiasm accumulated more PRG in rabbits dosed with PRG ME compared to the rabbits received PRG in water. In contrast, other peripheral organs such as heart, lung, liver, spleen, kidney & plasma retain relatively higher PRG levels in rabbits received PRG in water compared to the rabbits dosed with PRG ME.

3.5. Stability studies of PRG ME

Because the engineered PRG ME eye drops would potentially be a new drug product and its optimal storage conditions are unknown, we evaluated its stability following the guidelines put forth by the International Conference on Harmonization (ICH)[42]. The selected conditions included the expected range of temperatures that cover the long-term, intermediate and accelerated stability conditions for formulations intended to be stored either in ambient temperature or in refrigerator[42]. Per the ICH guidelines, at least three batches of the new pharmaceutical product are required to conduct the stability study[42]. Furthermore, if the drug product is intended to be stored in a refrigerator, the long-term stability study should be assessed at 5° ± 3°C for 12 months, while the accelerated stability study should be conducted at 25° ± 2°C for 6 months. However, when there is no specific temperature established for storage, in that general case the long-term stability study should be conducted at 25° ± 2°C for 12 months, while the accelerated stability study should be conducted at 40° ± 2°C for 6 months. Also, in the general case, an intermediate stability study that should be conducted at 30° ± 2°C for 6 months[42]. Therefore, because optimal storage conditions have yet to be established, physical and chemical stability of the formulation were evaluated at 5°, 25°, 30°, and 40°C. Furthermore, the physical stability of the ME formulation during shipping and transportation was gauged using different physical stress tests such as repeated freeze-thaw cycles and ultracentrifugation at different speeds.

3.5.1. Physical Stability Tests

3.5.1.1. Physical stress tests

PRG ME system was subjected to several freeze-thaw cycles and ultracentrifugation tests which demonstrates its high physical stability. The ME successfully passed all these physical stress tests without the appearance of any sign of instability such as creaming, turbidity, phase separation or precipitation. This high stability of PRG ME against physical stress may be due to its thermodynamic stability as a microemulsion system[89].

3.5.1.2. Physical appearance

After 1 year of storage at 5° & 25°C and 6 months at 30° & 40°C, no signs of physical instability were observed except the appearance of a slight yellow color in the formulation stored at 40°C after four months of storage (not shown). This is likely due to oxidative degradation of some ME lipid contents at higher temperature which led to color change[90]. These findings confirmed the high physical stability of the ME formulation especially at 5°, 25°, and 30°C.

3.5.1.3. pH evaluation

The freshly prepared PRG ME formulation had a pH value of 5.3 ± 0.1. Fig. 8A presents pH values at various time points during the stability study (1, 2, 3, 6 months at all the evaluated temperatures, in addition to 9 and 12 months at 5° and 25°C). There were no significant changes in the pH values during the entire study except at 40°C (p = 0.01) after 6 months. This significant pH change at 40°C after 6 months may be due to oxidative degradation of some ME lipid contents such as soybean lecithin or Labrafac lipophile WL1349, which led to the release of some free acids which decreased the formulation pH [91]. Although there is change in pH at 40°C after 6 months, the overall pH values results confirmed the high physical stability of the ME formulation even at high temperatures such as 30°C.

Fig. 8.

ME evaluation after storage for 6 months at 30° and 40°C & 12 months at 5° and 25°C. (A) pH, (B) Average droplet size, (C) Polydispersity index, (D) Zeta potential, (E) Drug content and (F) Cumulative amount of PRG released (%) from ME formulation after storage for 1, 2, 3, 6, 9 and 12 months at 5°C (mean ± SEM; n = 3). More than 90% of PRG concentration remains in the formulation for a period of 3–6 months at both 5° and 25°C.

3.5.1.4. Average droplet size, polydispersity index (PDI) and zeta potential

Fig. 8B presents the mean droplet size of PRG ME formulation at various time points of storage at the four studied temperatures. The average droplet size of the fresh ME was 15.6 ± 0.1 nm. After 2 months of storage, there was no significant change in the measured droplet size (p > 0.05) except at 40°C (p = 0.02). However, after three months there were no significant changes at either 5°C (p = 0.73) or 25°C (p = 0.75), with a significant increase in the measured droplet size values at both 30° and 40°C. After 6 months of storage, there was a significant increase in the average droplet size at all the studied temperatures (p < 0.0001). Regarding the droplet size distribution (PDI), Fig. 8C illustrates that the PDI value of the fresh ME was 0.256 ± 0.011. There were no significant changes in the PDI values after 6 months of storage at 5°, 25° and 30°C (p > 0.05). However, a significant increase in PDI was observed at 40°C after 6 months of storage (p = 0.04). This observed increase in the droplet size and droplet size distribution may be attributed to the possible fusion of the ME droplets that led to the formation of bigger droplets, especially after long term storage at higher temperature. Although there were significant increases in droplet size and PDI values at some time points, the highest measured values of droplet size and PDI through the 1 year stability study were 20.6 ± 0.11 nm and 0.358 ± 0.01, respectively, which remains in the acceptable range for a stable microemulsion formulation. In conclusion, the ME formulation has a high physical stability because it maintained its small droplet size (< 21 nm) and PDI (<0.4) after 1 year of storage at different temperatures. It has been reported that ME with smaller droplet size and PDI are more stable than MEs with larger droplets and PDI values, therefore, the high physical stability of the ME is likely due to the very small droplet size and PDI values[46, 92]. Similarly, Nordiyana et al. 2016 reported that a smaller PDI value indicates a higher homogeneous distribution and a highly stable ME[92]. In support of this, Nayak et al.,[46] presented that triamcinolone acetonide loaded-PEGylated ME did not show significant alteration in either size or PDI, which indicated higher ME stability, after storage at 4°C for three months[46].

Regarding the ME zeta potential (Fig. 8D), there were no significant changes in the measured zeta potential after 1 month of storage at the four studied temperatures (p > 0.05). In contrast, there was a significant increase in the measured negative zeta potential values after 2 months of storage at 30° and 40°C (p = 0.0003 & p < 0.0001, respectively). In addition, after 6 months of storage, the highest measured negative zeta potential was observed at 40°C (−47.8 ± 3.0 mV) compared to the initial zeta potential value (−31.4 ± 0.7 mV). After 9 months, there were no significant changes in the measured zeta potential (p = 0.11) at 5°C. In contrast, there was a significant increase after 12 months of storage at the same temperature (p = 0.001). In general, for any colloidal system, the surface charge is very important factor to achieve a long-term stability as it causes repulsion between droplets and prevention of their coalescence upon standing. In addition, the higher the surface charge—either positive or negative—the higher the system stability[93, 94]. Thus, the high negative charge of the ME formulation supports its high physical stability during 1 year of storage. This increase in the negativity during the stability study, especially at higher temperature after long term storage, may be attributed to the oxidative degradation of some ME lipid contents, such as soybean lecithin or Labrafac lipophile WL1349, which led to the release of free acids and exposure of more negatively charged carboxylic groups, that resulted in increasing in the overall negative zeta potential of the ME system[95].

3.5.2. Chemical Stability

3.5.2.1. Drug content of PRG ME formulation

The drug content as an indicator of the ME formulation chemical stability was calculated monthly as PRG-percentage amount that remained in the prepared formulation during the 6 months storage at 30° and 40°C & 1 year storage at 5° and 25°C (Fig. 8E). There were no significant changes in the measured drug content (p > 0.05) during the first 3 months at all temperatures except for the third month at 40°C (p = 0.04). In contrast, after 6 months, there were significant decrease in the drug contents at all temperatures. Generally, drug storage in aqueous media greatly potentiates its degradation. It is known that PRG can undergo degradation at either alkaline pH or by the oxidative pathway[43]. Because the ME has a weak acidic pH (5.3 ± 0.1), the alkaline degradation might not be a possible pathway of PRG loss during storage. Therefore, PRG is likely subjected to oxidative degradation in the ME system, which could negatively affect the chemical stability of the ME formulation. Another explanation for the significant decrease in the drug contents after 6 months at all temperatures is the instability associated with lipid-based formulations such as MEs. Subsequently, lipids that are used during the preparation of lipid-based formulations may be considered as a main source of instability for the drug products in which it is incorporated. Lipids are easily attacked by oxygen, light or heavy metal traces in the medium which could in turn initiate the process of lipid oxidation and result in the formation of reactive species, such as peroxides and free radicals. These reactive species could then attack the drug molecules and increase their degradation rate, especially if it is already susceptible to oxidation, as is PRG[90]. Collectively, it could be recommended that PRG ME formulation should be stored below 25°C to prolong its chemical stablity for a period of up to 6 months[48, 96, 97].

3.5.2.2. Degradation kinetics

After screening the data of the formulation drug contents using different kinetics order equations (zero, first and second) we determined that the data was best fitted with second order degradation kinetics at different temperatures (Table 4). The second order degradation rate constant (k) was calculated from the slope of the reciprocal of the remaining drug concentration vs. time curve. Half-lives (t_1/2) and shelf-lives (expiration date, t_90%) for PRG in PRG ME were calculated using the following equations:

$${\mathbf{\text{t}}}_{\mathbf{1}\mathbf{/}\mathbf{2}}=(1/{\text{k[A]}}_0)$$ (1)

$${\mathbf{\text{t}}}_{\mathbf{90}\mathbf{\%}}=(0.111/{\text{k[A]}}_0)$$ (2)

where k is the second order rate constant, [A]₀ is the total drug concentration at time zero.

Table 4.

PRG degradation kinetics after storage of PRG ME formulation for a period of 6 months at 30° and 40°C & 12 months at 5° and 25°C

Coefficient of determination (R2)
Degradation order	5°C	25°C	30°C	40°C
Zero	0.848±0.02	0.879±0.04	0.793±0.01	0.885±0.06
First	0.861±0.01	0.893±0.04	0.807±0.01	0.892±0.05
Second	0.873±0.01	0.900±0.03	0.821±0.01	0.898±0.04

The results of the predicted shelf-lives calculated using the second order degradation equation (Table 5) are in support of the measured drug content at various temperatures (Fig. 8E). The results (Table 5) suggest that PRG ME formulations should be stored at temperature below 25°C.

Table 5.

The predicted half-lives and shelf-lives of PRG ME formulation from second order degradation equations

Temperature	Slope (1/ka)	K (M−1 Month−1)	t1/2b (months) (1/k [A]0c)	t90%d (months) (0.111/k [A]0)
5°C	0.0002±0.0	5714.8±338.6	33.9±1.9	3.76±0.2
25°C	0.0001±0.0	8121.9±1415.2	25.4±4.9	2.82±0.5
30°C	0.00009±0.0	11483.2±1206.0	17.16±2.0	1.91±0.2
40°C	0.00004±0.0	25946.6±4505.2	7.81 ±1.2	0.867±0.1

^aK: the second order rate constant

^bt₁/2: the time required for the total drug concentration to fall to half of its initial value

^c[A]_0: the molar initial drug concentration

^dt₉₀%: the time required for the total drug concentration to fall to 90% of its initial value

Although the ME formulation contains only one drug (PRG), it follows second order degradation kinetics which required the presence of two reactants. This could be explained by the inclusion of oxygen as a reactant in the degradation reaction. As mentioned previously, PRG is susceptible to oxidative degradation[90]; therefore, the presence of oxygen may act as a stressor and/or a second reactant in the degradation reaction which promoted the second order degradation of PRG especially at high storage temperature and in the presence of water (one component of the ME)[98]. Similar reports previously documented the tendency of drugs incorporated in ME systems to follow a second order degradation reaction due to various stressors such as oxygen or ME components, such as water, which can contribute as a reactant in the degradation reaction[98, 99].

3.5.2.3. In vitro drug release

During the stability study, drug release behavior of PRG from ME formulation was evaluated to determine the ability of the ME system to maintain its sustained release properties of PRG after storage for a period of 6 months at 30° and 40°C and 12 months at 5° and 25°C. Fig. 8F illustrates the cumulative amount of released PRG up to 24 h after 1, 2, 3, 6, 9 and 12 months of storage at 5°C compared to the fresh sample. Fig. 8F demonstrates that PRG release pattern didn’t change by time during storage at 5°C, as all the release profiles showed the plateau at 8 h followed by continuous slow release for up to 24 h. Statistical analysis of the percentage cumulative amount released of PRG after 24 h presented that, after 1 year of storage at 5°C, there was no significant difference (p > 0.05) in the percentage cumulative amount released of PRG from the ME formulation compared to the fresh sample (Supplemental Table 3). In addition, Supplemental Fig. 2A illustrates the cumulative amount released of PRG up to 24 h after 1, 2, 3, 6, 9 and 12 months of storage at 25°C, while Supplemental Figs. 2B and 2C present the cumulative amount released of PRG up to 24 h after 1, 2, 3, and 6 months of storage at 30° and 40°C, respectively. Similarly, Supplemental Figs. 2A–C and Supplemental Table 3, illustrate no significant difference in the percentage cumulative amount released of PRG from the ME formulation after 24 h compared to the fresh sample after storage for 6 months at both 30°& 40°C and 1 year at 25°C (p > 0.05). These results confirm the stability of the ME system and its ability to maintain the sustained release behavior of the incorporated drug even after storage at high temperatures for several months.

The release data was screened against various release kinetics models (zero, first and Higuchi). The data demonstrate that, regardless of the storage temperature, the release kinetics of PRG from the ME formulation is a biphasic process, and both of the phases follow different release kinetics models. The first release phase (0–4 h) follows zero order release kinetics, while the second phase (started after 4 h), obeys the Higuchi release kinetics model[100]. Table 6 presents the values of the coefficients of determination obtained by fitting the first release phase data to zero order kinetics model and the second phase to Higuchi kinetics model. This biphasic release behavior of PRG from the ME may be attributed to the presence of PRG in the two aqueous phases of the ME (internal and external aqueous phases), which is separated by an intermediate oil phase. Therefore, the presence of PRG in the external aqueous phase of the ME could be the reason for the first release phase that is governed by zero order kinetics. This is because the PRG in the external aqueous phase is ready for direct release by simple diffusion through the viscous polymeric external aqueous phase. In the first release phase, the released amount of PRG is immediately replenished by the PRG in the internal aqueous phase, which acts as a reservoir for PRG. Subsequently, PRG concentration in the external aqueous phase remains nearly constant and the release rate is not dependent on the drug concentration (i.e., zero order kinetics). However, at the second release phase, after all PRG in the internal aqueous phase is depleted and no additional PRG is available to replenish the released drug, the release kinetics are converted to another model (i.e., Higuchi, diffusion release mechanism) in which the rate of release is depending on the drug concentration in the external aqueous phase. Because PRG release is controlled mainly by diffusion, the Korsmeyer-Peppas model is applied to further investigate the type of diffusion by which the drug is released. These data demonstrate that the values of Korsmeyer-Peppas release exponents (n) were < 0.5 (data not shown). Based on these findings, we can conclude that the mechanism of PRG diffusion from the ME vehicle could be described as a quasi-Fickian diffusion or pure diffusion mechanism through the viscous polymeric external aqueous phase[49, 101]. Advantageously, the drug release mechanism and kinetics behavior of the PRG ME don’t change by storage for 1 year at 5° and 25°C or for 6 months at higher temperatures at 30° and 40°C.

Table 6.

Biphasic release kinetics of PRG^a during storage at different temperatures

Time (months)	Coefficient of determination (R2)
	5°C		25°C		30°C		40°C
	1st phase	2nd phase	1st phase	2nd phase	1st phase	2nd phase	1st phase	2nd phase
1	0.978±0.01	0.955±0.02	0.927±0.01	0.988±0.01	0.956±0.01	0.984±0.01	0.953±0.01	0.999±0.00
2	0.973±0.01	0.988±0.01	0.962±0.02	1.00±0.00	0.951±0.01	0.885±0.06	0.973±0.01	0.995±0.00
3	0.918±0.02	0.999±0.00	0.950±0.00	0.997±0.00	0.941±0.01	0.998±0.00	0.875±0.02	0.998±0.00
6	0.919±0.01	0.990±0.00	0.950±0.02	0.995±0.00	0.922±0.01	0.990±0.00	0.913±0.02	0.996±0.00
9	0.913±0.02	0.992±0.01	0.901±0.02	1.00±0.00	—	—	—	—
12	0.927±0.01	0.986±0.01	0.916±0.02	0.995±0.00	—	—	—	—

^aPRG: pregabalin. 1^st phase obeys zero order kinetics model, 2^nd phase obeys Higuchi kinetics model.

4. CONCLUSION

Molecular docking analyses predicted a robust binding affinity of PRG to CACNA2D1 receptors across mouse, rabbit, and human species. Given the shared binding affinities to CACNA2D1, the observed reduction in IOP by PRG in mouse and rabbit models serves as pre-clinical evidence to the possibility of similar clinical effects in humans, further supporting PRG as a promising candidate for therapeutic intervention in glaucoma. The ME was designed to overcome the drawbacks associated with traditional eye drops that include rapid drainage, short corneal contact time and minimal corneal penetration; all of which led to reduced efficacy and poor patient adherence. We accomplished our goal by engineering a multilayered ME using highly biocompatible components with in situ gelling properties that improves bioadhesion, enhances corneal penetration and provides continuous pharmacological action for >30 h in two different animal models (Dutch belted rabbits and various mouse strains). Comparison of PRG ME with latanoprost eye drops demonstrated that both eye drops have a equivalent IOP lowering ability, however PRG ME is superior regarding its significant longer duration of action. The in vivo bioadhesion study demonstrates that the ME system remains adhered to the eye surface for up to 24 h. Also, we determined that repeated daily dosing of spontaneously elevated IOP Dutch belted rabbits with PRG ME for 60 days neither resulted in a decrease in efficacy of the formulation nor produced any side effect. Throughout the 60 days study, IOP remained at a reduced level in the treated eye, while the control eye remained near its elevated baseline IOP. In addition, PRG reached effective levels in target tissues within the treated eye; however, the control eye that dosed with blank ME contained only miniscule levels of PRG, as did plasma and all peripheral organs. In-life ocular examinations and post-mortem histopathological exams after 60 days of daily dosing determined that the ME is safe and well tolerated by the eye. The ME has a good physical stability for one year of storage at different temperatures while the chemical stability remain within the acceptable pharmacopial limit for 3–6 months at ambient temperatures. Additional studies are needed to improve the chemical stability of PRG ME by selecting of grades of ME excipient with low level or no impurities such as heavy metals, peroxides and solvents. The outcomes of our investigations have provided ample evidence to support PRG ME as a once daily topical treatment for IOP reduction in glaucoma.

Supplementary Material

1

Supplemental Fig. 1: (A) Pseudo-ternary phase diagram of the primary W/O ME. The primary W/O ME was formed of Labrafac oil/Milli-Q water and capryol 90/lecithin [1:1]. The shaded regions represent the W/O ME area. The selected W/O ME point consisted of 20% Milli-Q water +30% Labrafac oil +50% surfactant mixture (capryol 90 and soybean lecithin, 1:1). (B) Cartoon of PRG multilayered ME illustrating the three layers of the ME, water in oil in water. Internal surfactants were used during the formation of the primary water in oil ME and external surfactants were used during the emulsification of the primary W/O ME into the external aqueous phase (C) TEM of PRG water in oil in water (W/O/W) ME showing the distinct multilayered structure of the PRG ME. (D) Magnified ME droplets consist of several nuclei of the internal aqueous phase entrapped in the intermediate oil phase, which is surrounded by the external aqueous layer.

2

Supplemental Fig. 2: (A) Cumulative amount of PRG released (%) from ME formulation after storage for 1, 2, 3, 6, 9 and 12 months at 25°C (mean ± SEM; n = 3). (B & C) Cumulative amount of PRG released (%) from ME formulation after storage for 1, 2, 3 and 6 months at 30° and 40°C (mean ± SEM; n = 3). The Figs. illustrate no significant difference in the percentage cumulative amount released of PRG from theME formulation after 24 h compared to the fresh sample after storage for 1 year at 25°C and 6 months at both 30°& 40°C (p > 0.05).

6

Supplemental Video 1: Video analysis of S-PRG in complex with mouse CACNA2D1, predicted by Glide with a docking score of −10.820 kcal/mol. S-PRG is represented as stick Fig. in green, amino acid residues within 5 Å of S-PRG as stick Fig. in gray, and CACNA2D1 secondary structures as ribbon diagrams.

7

Supplemental Video 2: Video analysis of S-PRG in complex with rabbit CACNA2D1, predicted by Glide with a docking score of −11.138 kcal/mol. S-PRG is represented as stick Fig. in green, amino acid residues within 5 Å of S-PRG as stick Fig. in gray, and CACNA2D1 secondary structures as ribbon diagrams.

8

Supplemental Video 3: Video analysis of S-PRG in complex with human CACNA2D1, predicted by Glide with a docking score of −11.027 kcal/mol. S-PRG is represented as stick Fig. in green, amino acid residues within 5 Å of S-PRG as stick Fig. in gray, and CACNA2D1 secondary structures as ribbon diagrams.

• Molecular docking analyses estimate strong binding affinity of pregabalin, a novel therapy for open angle glaucoma to CACNA2D1 in diverse species including mice, Dutch belted rabbits and humans

• Pregabalin microemulsion eye drops supports once daily use by extending the precorneal drug residence time up to 24h

• Pregabalin microemulsion eye drops didn’t have any tachyphylactic effect upon repeated dosing for 2 months

• Pregabalin microemulsion eye drops is physically stable for one year at different temperatures and chemically stable for 3–6 month at temperature below 25°C.