Design of Topical Moxifloxacin Mucoadhesive Nanoemulsion for the Management of Ocular Bacterial Infections

Ocular bacterial infections can lead to serious visual disability without proper treatment. Moxifloxacin (MOX) has been approved by the US Food and Drug Administration as a monotherapy for ocular bacterial infections and is available commercially as an ophthalmic solution (0.5% w/v). However, precorneal retention, drainage, and low bioavailability remain the foremost challenges associated with current commercial eyedrops. With this study, we aimed to design a MOX-loaded nanoemulsion (NE; MOX-NE) with mucoadhesive agents (MOX-NEM) to sustain MOX release, as well as to overcome the potential drawbacks of the current commercial ophthalmic formulation. MOX-NE and MOX-NEM formulations were prepared by hot homogenization coupled with probe sonication technique and subsequently characterized. The lead formulations were further evaluated for in vitro release, ex vivo transcorneal permeation, sterilization, and antimicrobial efficacy studies. Commercial MOX ophthalmic solution was used as a control. The lead formulations showed the desired physicochemical properties and viscosity. All lead formulations showed sustained release profiles a period of more than 12 h. Filtered and autoclaved lead formulations were stable for one month (the last time point tested) under refrigeration and at room temperature. Ex vivo transcorneal permeation studies revealed a 2.1-fold improvement in MOX permeation of the lead MOX-NE formulation compared with Vigamox® eyedrops. However, MOX-NEM formulations showed similar flux and permeability coefficients to those of Vigamox® eyedrops. The lead formulations showed similar in vitro antibacterial activity as the commercial eyedrops and crude drug solution. Therefore, MOX-NE and MOX-NEM formulations could serve as effective delivery vehicles for MOX and could improve treatment outcomes in different ocular bacterial infections.


Introduction
Ocular bacterial infections present in many disease categories, including keratitis, conjunctivitis, blepharitis, and endophthalmitis [1,2]. Bacterial keratitis (BK) is the inflammation of the cornea due to bacteria, which should be considered an ophthalmic emergency. Bacterial invasion of the cornea can interfere with the pathway of light entering the eye globe and destroy the intact epithelial cells covering the cornea, which could result in irreversible visual impairment if left untreated [3]. The most predominant Gram-positive bacteria isolated in BK are Staphylococcus aureus, Streptococcus pneumoniae, Streptococcus viridans, and Enterococcus spp., whereas the most predominant Gram-negative bacteria include Moraxella lacunata, Pseudomonas aeruginosa, Haemophilus influenzae, Haemophilus parainfluenzae, Serratia process of NEs is easy and inexpensive, similar to conventional solutions, due to their submicron size [24].
Mucoadhesive polymers are known to prolong the residence time on the ocular surface after topical administration by adhering to the mucin layer and enabling a uniform distribution across the surface of the eye globe [25]. Hence, the incorporation of a mucoadhesive polymer in the NE formulation could result in improved ocular bioavailability and prolonged antibacterial activity [26]. Hypromellose (HPC; hydroxyl propyl methyl cellulose, 0.1-0.6% w/w) polymers, such as hypromellose 2906 (4000 mPa·s), hypromellose 2910 (4000 mPa·s), hypromellose 2910 (3.0 mPa·s), or hypromellose 2910 (5.0 mPa·s), are widely used in many FDA-approved ophthalmic formulations as a mucoadhesive agent due to their aqueous solubility, biocompatibility, transparency, and rheological properties [25,27]. Povidone (PVP; polyvinylpyrrolidone) is also a recognized polymer for ocular applications and is widely used in many FDA-approved ophthalmic products at concentrations ranging from 0.3 to 15% w/v [28]. PVP is an inert, non-toxic, biocompatible, biodegradable, mucoadhesive polymer [28].
The objective of the current investigation was to overcome the potential drawbacks of commercial MOX eyedrops by developing MOX-NEs by incorporating mucoadhesive agents (MOX-NEM) that could enhance ocular surface retention and penetration into ocular tissues and reduce the frequency of administration. Accordingly, MOX-NEs were prepared, optimized, and evaluated based on physicochemical characteristics and stability. The lead MOX-NE formulation was then converted into a mucoadhesive NE with the addition of HPMC and PVP polymers as thickening and mucoadhesive agents, respectively. MOX-NE and MOX-NEM formulations were evaluated for in vitro release, ex vivo permeation, and antimicrobial efficacy and compared against commercial MOX ophthalmic solution eyedrops (Vigamox ® , 0.5% as a base).

Biological Tissues and Samples
The whole eyes of mixed-gender albino New Zealand rabbits (weight, 4.75-5.75 lbs and; age, 8-12 weeks) were purchased from Pel-Freez Biologicals (Rogers, AR, USA). Microbial strains were obtained from the American Type Culture Collection (ATCC, Manassas, VA, USA).

HPLC
MOX concentration in all samples was quantified using HPLC, Alliance Waters e2695 separations module, and a Waters 2489 UV/Vis dual absorbance detector. A detection wavelength (λ max ) of 254 nm was set. The mobile phase consisted of a mixture of phos- phate buffer (18 mM) containing 0.1% v/v triethylamine (pH 2.8, adjusted with dilute phosphoric acid) and methanol (60:40 v/v) at a flow rate of 1.0 mL/min [29]. Chromatographic separation was achieved within 10 min using a Phenomenex Luna ® C 18 column (250 × 4.6 mm, 5 µ) as a stationary phase with a retention time of 7.1 ± 0.2 min. The samples were analyzed through a Waters chromatography data system and Empower software. The HPLC method was found to be linear over the MOX base concentration range of 1.0-100 µg/mL, with a limit of detection (LOD) and limit of quantitation (LOQ) of 0.8 and 2.4 µg/mL, respectively.

Screening of Oils
The solubility of MOX in various oils was determined by individually adding 10 mg of MOX to 200 mg of the oils in glass vials (3 mL). Then, the drug-oil mixture was vortexed (Vortex Genie ® 2, Scientific Industries, Inc., SI™, Bohemia, NY, USA) for 2 min. Next, the MOX-oil mixtures were heated at 80.0 ± 2.0 • C and kept under continuous magnetic stirring (2000 rpm) for 10 min. The mixtures were then allowed to cool to room temperature and examined visually for MOX precipitation. The oil that did not show any precipitation was chosen for formulation preparation.

Preparation of MOX-NE Formulations
Oil-in-water (O/W)-type NEs were prepared by hot homogenization followed by the probe sonication method [16]. The oily phase was prepared by dissolving an accurately weighed amount of MOX base within the selected oil by heating at 80.0 ± 2.0 • C to obtain a clear drug solution. An aqueous phase, prepared by adding Tween ® 80 and glycerin to Milli-Q water, was also heated at 80.0 ± 2.0 • C. Then, the hot aqueous phase was added to the heated oil phase dropwise under continuous magnetic stirring at 2000 rpm for 10 min to form a macroemulsion. This macroemulsion was then homogenized using a T25 digital Ultra-Turrax (IKA Works, Inc., Wilmington, NC, USA) at 11,000 rpm for 5 min at 65.0 ± 2.0 • C [16]. The macroemulsion was allowed to cool to room temperature before being subjected to probe sonication at 40% amplitude for 10 min with a 3 mm stepped microtip with a 500 watt power supply at 115 volts (Pulse on; 10 s and Pulse off; 15 s) using a Sonics Vibra-Cell™ sonicator (Newtown, CT, USA) to form NE [30].

Preparation of MOX-NEM Formulations
The NE preparation method, as described above, was followed for the preparation of the NEM formulations. However, the volume of Milli-Q water used to prepare NE was split into two equal parts: one part was used to prepare the aqueous solution of the mucoadhesive agent (HPMC K4M or PVP K29/32), and the other half was used to prepare the aqueous solution of surfactant and glycerin, as described above.

Control Formulation
Vigamox ® (Alcon Laboratories, Inc., Fort Worth, TX, USA) A commercial ophthalmic solution of moxifloxacin hydrochloride (0.5% w/v) was used as a control formulation for the in vitro release, ex vivo permeation, and antibacterial efficacy studies.
Crude Drug Solution (MOX-HCL-S) MOX hydrochloride powder (moxifloxacin hydrochloride 5.45 mg equivalent to 5.0 mg moxifloxacin base) was dissolved in 1 mL saline water to ensure that the concentrations of MOX base in the control and test formulations were equal for antimicrobial efficacy studies.
2.2.6. Measurement of Droplet Size (DS), Polydispersity Index (PDI), and Zeta Potential (ZP) The MOX-NE and MOX-NEM formulations were evaluated for their DS, PDI, and ZP using a Zetasizer instrument (Nano ZS Zen3600, Malvern Panalytical Inc., Westborough, MA, USA) at 25 • C in disposable, folded, clear, solvent-resistant micro cuvettes (ZEN0040) [31]. The cuvette was rinsed with 0.22 µm filtered Milli-Q water to ensure the absence of dust/particulates in each cuvette before performing the measurement. Prior to measurement, formulations were diluted (100 times) with Milli-Q water. The same diluted formulations were transferred to Zetasizer disposable folded capillary DTS1070 cells for ZP measurement after DS and PDI measurement. All DS, PDI, and ZP measurements were carried out in triplicate.

Drug Content
For drug content analysis, an accurate volume (50 µL) of each NE formulation was added to 950 µL methanol (extracting solvent). This extract was centrifuged (AccuSpin 17R centrifuge, Fisher Scientific, Hanover, IL, USA) at 13,000 rpm for 20 min. Then, the supernatant was diluted 10 times with the mobile phase and analyzed for MOX content using the HPLC method described above.
2.2.8. pH Measurement pH was measured using a Mettler Toledo pH meter (FiveEasy™, Columbus, OH, USA) equipped with an Inlab ® Micro Pro-ISM probe. Before measurement, the pH meter was calibrated using different standard buffers with known pH values of 4.01, 7.00, and 10.01. The pH of each prepared NE or NEM formulation was measured in triplicate.

Viscosity Measurement
A Brookfield cone and plate viscometer (LV-DV-II+ Pro Viscometer, Middleborough, MA, USA) was used to measure the viscosity (η) of all MOX-NE and MOX-NEM formulations using a CPE 52 spindle operated at 10 rpm. Each formulation (1.0 mL) was placed in the viscometer cup plate. The temperature of the cup was maintained at 25 • C using a circulating water bath. Data analysis was performed using Rheocalc software (version 3.3, build 49-1). The viscosity measurements of all formulations were carried out in triplicate. MOX-NE and MOX-NEM formulations were also subjected to autoclaving (121 • C under 15 psi for 15 min, 3850ELP-B/L-D Tuttnauer autoclave, Heidolph, Germany) in glass vials affixed with indicator tapes for the sterilization process. The sterilization cycle was confirmed by the color change of the indicator tapes attached to the glass vials. After the moist heat sterilization process, formulations were stored at RF and RT for one month (last time point tested) in the same containers and analyzed in the same way as the preautoclaved samples. STEM analysis was performed using a JSM-7200FLV scanning electron microscope (JOEL, Peabody, MA, USA) attached to a STEM detector with an accelerating voltage of 30 KV. STEM samples were examined according to a negative staining protocol with a solution of UranyLess. A carbon-plated copper grid was placed above one drop (20 µL) of the MOX-NE formulation for 60 s, and the excess sample was removed with the aid of a filter paper after grid removal from the formulation surface. The grid was then washed by dipping in distilled water, and the excess water was removed from the grid using filter paper. Next, the grid was placed above one drop (20 µL) of the staining solution for 60 s, and the excess stain was also drawn off the grid with the aid of a filter paper. The grid was allowed to air dry for a few minutes. The grid was examined under the scanning transmission electron microscope at 65 K times magnification power.

In Vitro Release Studies
Phosphate-buffered saline (PBS; pH 7.4) containing 5.0% w/v hydroxypropyl betacyclodextrin (HPβCD) was used as the receiver medium for in vitro release and ex vivo transcorneal permeation studies. The media were selected based on our earlier investigations [32]. In vitro release profiles of MOX from Vigamox ® , MOX-NE, and MOX-NEM formulations were evaluated using the dialysis method. Test and control formulations (200 µL) were added to 0.5 mL cup-like dialysis devices (donor compartment) and fitted on the top of scintillation glass vials (receiver compartment). The content of the receiver compartment was maintained under continuous magnetic stirring at 34.0 ± 2.0 • C on the top of a multi-stationed magnetic stirrer (IKA Works, Inc., Wilmington, NC, USA) at 500 rpm. Samples (1 mL) were collected from the receiver compartment and replaced with an equivalent volume of a freshly prepared release medium at scheduled time points (0.5, 1, 2, 3, 4, 6, 8, and 12 h). Samples were analyzed after collection using the HPLC method mentioned above. The time for 50% release of the loaded drug (T 50% ) was calculated. The release data were fitted to different mathematical release models using DDSolver software, a free add-on program for Microsoft Excel (Office365, 2016, Redmond, WA, USA), to elucidate the possible release mechanism.

Transcorneal Permeation Studies
The ex vivo transcorneal permeation of MOX from the test and control formulations was evaluated on corneas isolated from whole rabbit eyes that were shipped overnight over ice in Hanks' balanced salt solution. Ex vivo permeation testing was performed using a vertical Franz diffusion apparatus (PermeGear ® Inc., Hellertown, PA, USA). The corneas were excised carefully upon arrival and washed with IPBS solution (pH 7.4). Each excised cornea was crimped between the two chambers of each vertical Franz diffusion cell (spherical joint), with the corneal epithelium facing the donor chamber that contained the formulation. The receiver chamber contained a solution of HPβCD in PBS (5.0 mL, 5.0% w/v, pH 7.4) kept under continuous magnetic stirring at 34.0 ± 2.0 • C. Aliquots (0.5 mL) were withdrawn from the receiver compartment at scheduled time points and replenished with an equal volume of the receiver medium. The aliquots were quantified for MOX using the HPLC method described above. The cumulative amount of MOX-permeated (Q n ), steady-state flux (J ss ) and permeability coefficient (P eff ) across the excised rabbit corneas were calculated.
Q n was calculated based on the following formula: where n is the sampling time point, V r is the volume in the receiver compartment (mL), V S is the volume of the aliquot collected at the nth time point (mL), and C r(n) is the drug concentration in the receiver compartment at the nth time point (µg/mL). J ss was calculated by the following formula: where (dQ/dt) is the rate of transcorneal permeation calculated using the slope of Q n versus the time plot, and A is the effective area of transcorneal permeation (0.64 cm 2 ). The transcorneal permeability coefficient was calculated using the following formula: where C 0 is the initial donor concentration for MOX.

Antimicrobial Efficacy
The antimicrobial activity of MOX-NE and MOX-NEM formulations was evaluated against methicillin-resistant Staphylococcus aureus ATCC 1708 (MRSA). Antimicrobial susceptibility testing was performed following a modified version of the Clinical and Laboratory Standards Institute (CLSI, 2012) methods by employing the most widely used susceptibility testing method, broth microdilution. MOX-NE and MOX-NEM formulations were serially diluted using cation-adjusted Mueller Hinton Broth 2 assay medium (pH 7.0). Then, the diluted samples (100 µL) were transferred to small, disposable, plastic "microdilution" trays (96 well). Inocula were prepared by correcting the OD630 of the bacterial suspensions in the incubation broth to provide recommended inocula as per CLSI protocol. Next, 5.0% Alamar Blue™ was added to MRSA microdilution trays. Crude MOX-HCL solution was used as a positive control in each assay. The optical density was measured for each of the panel wells by a Bio-Tek plate reader before and after incubation at 35 • C for 24 h. The minimum inhibitory concentration (MIC) was determined for all tested formulations and was defined as the lowest test MOX concentration that results in no visual growth. All experiments were carried out in triplicate.

Statistical Analysis
All results are presented as mean ± standard deviation. SPSS 28 software (IBM ® , Armonk, NY, USA) was used for statistical analysis of data. The significant difference between the data was compared at a p-value less than 0.05 (p < 0.05).

Screening of Oils
The solubility of the drug in the oily phase, which constitutes the dispersed phase in the O/W emulsion, is the most important criterion during the development of NE formulations for hydrophobic drugs. Drug loading in the NE is also dependent on the drug solubility within the surfactant micelles used to prepare NE formulations; however, drug precipitation could occur upon contact with biological fluids because the dilution of the NE formulation could affect the solubilizing power of the surfactant [16,33]. Therefore, the selection of oil is a critical step during the design of NE formulations. The solubility of MOX in various oils was evaluated based on visual examination (Table 1). MOX dissolved in oleic acid did not show any precipitation after cooling the drug-oil mixture to RT. Therefore, oleic acid was selected to prepare the NE and NEM formulations. Oleic acid is a long-chain unsaturated fatty acid that is widely used for ocular administration due to its biocompatibility and well-tolerated safety profile [34]. In addition, oleic acid has often been used in ophthalmic formulations, as it acts as a powerful penetration enhancer for both lipophilic and hydrophilic drugs, even at a concentration of 0.05% v/v, as it increases the fluidity of intercellular lipid barriers [35].

Preparation of MOX-NE Formulations
The compositions of the different NE formulations tested are presented in Table 2. MOX-NE formulations were prepared using oleic acid as the oily phase, selected based on the solubility of MOX in different oils. Tween ® 80 and glycerin represented the aqueous phase. According to the FDA inactive ingredient database, concentrations of Tween ® 80 and glycerin (tonicity-adjusting agent) of as much as 4.0 and 2.25% w/v, respectively, are present in FDA-approved topical ocular applications. Nonionic hydrophilic surfactants, such as Tween ® 80 with HLB > 10, have been reported for the preparation of uniform stable O/W NEs [36]. Drug loading was maintained at 0.5% w/v in the MOX-NE formulations to ensure that the concentrations of MOX base in the commercial (control) and test NE formulations were equal.
The MOX-NE formulations were optimized based on different oil concentrations (1, 2.5, and 5.0% w/v) and surfactant concentrations (0.75, 2.0, 3.0, and 4.0% w/v). The formulations were subjected to a one-month follow-up study at 4 • C based on visual examination. MOX-NE formulations (F1-F8) showed drug precipitations within 17 days after preparation. This could be due to an insufficient amount of oil to dissolve the drug. Formulation F9 showed phase separation, probably due to an insufficient amount of surfactant to stabilize the emulsion [37]. However, MOX-NE formulations F10-F12 were observed to be stable and did not show any precipitation or cracking, suggesting a sufficient amount of oil and surfactant. Therefore, formulations F10-F12 were selected for further evaluation and modifications.

Physicochemical Characteristics of MOX-NE Formulations
The effect of Tween ® 80 concentration on the formulation was investigated. Results of varying Tween ® 80 concentrations in the MOX-NE formulations are presented in Table 3.
Increasing Tween ® 80 concentrations from 0.75% to 4.0% w/v decreased the DS significantly (p < 0.05) from 174.1 ± 5.9 to 124.7 ± 1.5 nm. No significant change was observed in PDI, ZP, viscosity, and drug content. The significant effect of increasing Tween ® 80 concentration on DS could be attributed to the significant reduction in surface tension and surface free energy that was generated during the homogenization and ultrasonication steps [38,39].
The DS of the nanocarriers is important for the adhesion and interaction with the ocular epithelial cells. Smaller particles/globules (100-200 nm) can be transported by receptor-mediated endocytosis uptake, whereas larger particles are internalized by phagocytosis [30,40,41]. PDI values reveal the width of DS distribution, and a low PDI value (<0.3) demonstrates that the nanoparticulate system is a uniform dispersion with narrow DS distribution [42]. ZP within ±30 mV (absolute value) provides good physical stability for NEs, and the stability becomes excellent when ZP values approach ±60 mV. However, a ZP value of >±20 mV provides only short-term NE stability, whereas the −5 to +5 mV range could suggest rapid aggregation of the dispersed oil globules [42]. Therefore, the successful NE formulations (F10-F12) showed good DS (<200 nm), PDI (<0.2), and ZP (>−30 mV) values.
Each NE formulation should have MOX content within the acceptance limits of the label's content to ensure the consistency of the dosage form units. The drug content of the three successful NE formulations was in the range of 99.1 ± 3.8-100.8 ± 7.3% of the theoretical value.

Addition of Mucoadhesive Agents to the MOX-NE Formulation
In our previous studies with ∆ 9 -tetrahydrocannabinol-valine-hemisuccinate NEs, we observed that increasing the Tween ® 80 concentration to 2.0% w/v enabled an autoclaving process [25]. Therefore, the MOX-NE F10 (2.0% w/v surfactant) formulation was selected for further development with the addition of mucoadhesive agents.
HPMC K 4 M is present in FDA-approved ophthalmic formulations to concentrations up to 0.5% w/v. HPMC is used as a viscosity enhancer, gelling agent, and mucoadhesive agent in eyedrops or as a hydrophilic polymeric matrix in ocular films and inserts [26]. HPMC K 4 M also exhibits a high swelling capacity, which can further aid in providing sustained drug delivery vehicles [43]. Such vehicles help to effectively transfer the target molecule to the site of action [44]. PVP is widely used to increase the residence time of many topical ophthalmic products by way of its viscosity-enhancing and mucoadhesive properties. It is also used as an effective lubricant for dry eye disease, as well as a suspending agent, wetting agent, and stabilizer in ophthalmic suspensions due to its amphiphilic properties [28]. Moreover, this amphiphilic nature offers good solubility in water and organic solvents [45]. HPMC K 4 M and PVP K29/32 were evaluated at a concentration of 0.4% w/v ( Table 2). The selected polymer concentration imparts viscosity while allowing for easy topical ocular application (≤50 cP) based on our earlier investigations [25]. Incidentally, initial trials with Carbopol ® 940 NF revealed that the NE formulations were physically unstable (data not presented).

Physicochemical Characteristics of MOX-NEM Formulations
The physicochemical characteristics of the MOX-NE formulations with and without the addition of the HPMC K4M and PVP to the NE formulations as mucoadhesive agents are illustrated in Table 4. Upon the addition of the PVP and HPMC to the MOX-NEs, there was no significant difference observed in DS, PDI, and ZP. Moreover, the drug content for all the formulations was found to be in the range of 98.3 ± 2.5 to 101.0 ± 1.3%.

pH and Viscosity
The normal human tear pH range is 6.5 to 7.6, with a mean value of 7.0 [46]. The eye can tolerate topically administered products with pH values in the range of 3.0 to 8.6, depending on the buffering capacity of the formulation (USP). We observed that the incorporation of the mucoadhesive agents did not significantly affect the pH of the MOX-NE formulation. All NE formulations had pH values ranging from 5.4 ± 0.1 to 5.5 ± 0.1, which is favorable for ocular application (Table 4).
Although viscosity evaluation is not a compendial test, it is an essential part of the manufacturer's specification of the ocular product because viscosity can affect product performance. A viscosity of up to 50 cP, allows for easy topical ocular application [47]. The viscosity of MOX-NE and MOX-NEM formulations was measured using a Brookfield cone and plate viscometer (Table 4); we observed that the viscosity of the NE formulation (5.6 ± 0.5 cP) was increased significantly (p < 0.05) after the inclusion of HPMC K 4 M (F13; 26.5 ± 0.9 cP) and PVP K29/32 (F14; 27.9 ± 0.3 cP) polymers within the formulation. The increase in viscosity of the NE formulation after the addition of the high-molecularweight mucoadhesive polymers could be due to the strong internal friction between the randomly coiled and swollen polymer chains and the surrounding water molecules [48]. There was no significant (p > 0.05) difference between the viscosity values obtained for the mucoadhesive formulations.

Sterilization by Filtration
All ophthalmic dosage forms must be sterile and comply with the pharmacopeial tests for sterility. Filtration is widely adopted as a sterilization approach for NE formulations due to their small DS, which is less than the maximum nominal pore size (0.22 µm) of membrane filters utilized in this sterilizing process [16]. Thus, filtration of the three lead formulations was tested using different types of filter membranes. We observed that all three formulations could be easily filtered through Durapore™, nylon, and PES membrane filters; however, they faced resistance through the Fluoropore™ membrane filter.
Durapore™, nylon, and PES are hydrophilic membrane filters, whereas Fluoropore™ is a hydrophobic membrane filter. This could be the reason that all formulations passed through the hydrophilic filters but faced resistance through the hydrophobic filter materials because water constitutes the major composition of O/W NEs. Filtration could affect the particle size and distribution, along with the drug content, during the sterilization process. Therefore, the physicochemical characterization of all formulations was investigated before and after the filtration process. No significant (p > 0.05) difference was observed in terms of DS, PDI, ZP, and MOX content between the pre-and post-filtration formulations (Figure 1). passed through the hydrophilic filters but faced resistance through the hydrophobic filter materials because water constitutes the major composition of O/W NEs. Filtration could affect the particle size and distribution, along with the drug content, during the sterilization process. Therefore, the physicochemical characterization of all formulations was investigated before and after the filtration process. No significant (p > 0.05) difference was observed in terms of DS, PDI, ZP, and MOX content between the pre-and post-filtration formulations (Figure 1). Steam sterilization is probably the most widely used of all sterilization methods. As opposed to aseptic manufacture, terminal sterilization in the final dosage form container is the preferred approach from a regulatory point of view. Owing to the advantage as a terminal sterilization process, moist-heat sterilization was also investigated for the NE and NEM formulations. F10, F13, and F14 formulations were selected to evaluate moistheat sterilization process stability based on a three-month stability study of the unsterilized formulations. The pre-and post-moist-heat sterilization physicochemical characteristics are presented in Table 5. The autoclaved formulations remained stable at RF and RT; DS, PDI, ZP, pH, and drug content did not show a significant (p < 0.05) change in comparison to the pre-autoclaved formulation for one month. The sterilized formulations maintained high negative ZP values (≥30 mV), reflecting a high negative surface charge for adequate droplet-droplet-repulsive forces, improving NE and NEM stability. Steam sterilization is probably the most widely used of all sterilization methods. As opposed to aseptic manufacture, terminal sterilization in the final dosage form container is the preferred approach from a regulatory point of view. Owing to the advantage as a terminal sterilization process, moist-heat sterilization was also investigated for the NE and NEM formulations. F10, F13, and F14 formulations were selected to evaluate moist-heat sterilization process stability based on a three-month stability study of the unsterilized formulations. The pre-and post-moist-heat sterilization physicochemical characteristics are presented in Table 5. The autoclaved formulations remained stable at RF and RT; DS, PDI, ZP, pH, and drug content did not show a significant (p < 0.05) change in comparison to the pre-autoclaved formulation for one month. The sterilized formulations maintained high negative ZP values (≥30 mV), reflecting a high negative surface charge for adequate droplet-droplet-repulsive forces, improving NE and NEM stability. Table 5. Effect of autoclaving on the physicochemical characteristics of F10, F13, and F14 formulations before and after sterilization over one month of storage at 4 • C and 25 • C (mean ± SD, n = 3).

STEM
STEM images obtained for the F10 formulation exhibited spherical oil droplets with a diameter of less than 200 nm, as shown in Figure 2. The DS distribution profile was in close agreement with Zetasizer results.

Code
Day Storage at 25 ± 2 °C

STEM
STEM images obtained for the F10 formulation exhibited spherical oil droplets with a diameter of less than 200 nm, as shown in Figure 2. The DS distribution profile was in close agreement with Zetasizer results.

Antibacterial Activity of MOX-NE and MOX-NEM
Staphylococci are the most common causative agents of ocular bacterial infections, and MRSA is associated with persistent and complicated ocular infections [49]. The goal of this study was to evaluate the antimicrobial efficacy of all developed formulations in comparison to commercial Vigamox ® eyedrops and the crude MOX solution against MRSA. The MIC 90 value (6.25 μg/mL) obtained for all three formulations was the same as that of Vigamox ® eyedrops and crude MOX-HCL solution, as shown in Table 6. This implies

Antibacterial Activity of MOX-NE and MOX-NEM
Staphylococci are the most common causative agents of ocular bacterial infections, and MRSA is associated with persistent and complicated ocular infections [49]. The goal of this study was to evaluate the antimicrobial efficacy of all developed formulations in comparison to commercial Vigamox ® eyedrops and the crude MOX solution against MRSA. The MIC 90 value (6.25 µg/mL) obtained for all three formulations was the same as that of Vigamox ® eyedrops and crude MOX-HCL solution, as shown in Table 6. This implies that all lead NE formulations were as effective as the commercial formulation in terms of their antimicrobial efficacy against this vision-threatening bacterium.

In Vitro Release Studies
In vitro release testing of MOX from F10, F13, F14, and Vigamox ® formulations was performed by employing the dialysis method; the in vitro drug release profiles are graphically illustrated in Figure 3. The T 50% values of the release profiles of the control, F10, F13, and F14 formulations were 1.2, 3.6, 12.0, and 6.4 h, respectively. Thus, F10, F13, and F14 formulations sustained the release of the loaded drug compared to the commercial formulation. However, drug release from the PVP-containing NE formulation (F14) was slightly faster than that from the HPMC-containing formulation (F13). The entrapment of drug molecules within the NE formulation oil globules and slow drug diffusion from the oil phase to the aqueous phase may have contributed to the sustained release behavior from the NE and NEM formulations [48]. Moreover, the retarded drug release rates from F13 and F14 compared to the F10 formulation could be attributed to the increased viscosity imparted by the addition of the mucoadhesive polymers to the formulation [50]. These observations are consistent with many previously reported investigations [51,52].
Although the inclusion of a mucoadhesive agent prolongs the contact time of the formulation with the ocular surface, it could inhibit the diffusion of the drug out of the formulation [25]. However, the release data demonstrate that all test formulations achieved MOX concentrations greater than the obtained MIC 90 (6.25 µg/mL) within the first hour-10.6 µg/mL for F10, 6.8 µg/mL for F13, and 8.2 µg/mL for F14-indicating that the MIC against MRSA was attained across at release time points.
Four conventional release models, including zero-order, first-order, Higuchi, and Korsmeyer-Peppas, were fitted to each of the three data sets, and a regression analysis was performed in each case to verify the goodness of fit. The release model that showed the highest value for the coefficient of determination (R 2 ) was regarded as the best model to describe release kinetics, as illustrated in Table 7. Mathematical model fitting revealed that the dissolution profile of all developed formulations followed the Korsmeyer-Peppas model. All calculated slope values (n) of the Korsmeyer-Peppas model (0.5 < n < 1.0) indicated non-Fickian drug release profiles controlled by a diffusion mechanism. The general sustained release behavior of O/W NE is due to the fact that the release of hydrophobic drugs from this NE type involves many consecutive steps, starting from partitioning (diffusion) of the hydrophobic drug from oil into surfactant and then into the aqueous phase [16].

Ex Vivo Transcorneal Permeation
The rabbit eye is considered a reference animal model for transcorneal permeation studies because of the high similarity between rabbit and human eyes [53]. Flux and permeability of MOX across rabbit corneas from the commercial solution, F10, F13, and F14 formulations were studied, and the data are shown in Figure 4. The flux of MOX from the mucoadhesive NE formulations-1.54 ± 0.19 µg/min/cm 2 for F13 and 1.65 ± 0.17 µg/min/cm 2 for F14-did not show a significant difference (p > 0.05) when compared to the MOX flux from the control eyedrops (1.57 ± 0.04 µg/min/cm 2 ). However, the transcorneal flux of MOX from F10 formulations was approximately 2.0-fold higher compared to the control eyedrops and the mucoadhesive NE formulations. Consequently, the corneal permeation of MOX was approximately 2.1-fold greater from NE as compared to that of the commercial and mucoadhesive NE formulations. The observed improved drug flux from NE, as compared to the mucoadhesive NE formulations, could be due to the fact that the addition of the mucoadhesive polymer could slow the diffusion of the drug out of the formulation to the cornea; however, this phenomenon could be beneficial for prolonging ocular surface retention and sustaining penetration into the intraocular tissues [25].  A possible reasons for the improved corneal penetration of drugs loaded in the NE formulation could be that the presence of surfactants increases the membrane permeability, thereby increasing drug uptake [23]. Moreover, nanoparticles with a particle size of less than 200 nm could be internalized by a receptor-mediated endocytosis uptake mechanism through the corneal tissue [36]. In addition, oleic acid has been reported to improve the ocular drug delivery of both lipophilic and hydrophilic compounds [35]. Until now, no clear mechanisms have been reported for the ocular penetration-enhancing effect of A possible reasons for the improved corneal penetration of drugs loaded in the NE formulation could be that the presence of surfactants increases the membrane permeability, thereby increasing drug uptake [23]. Moreover, nanoparticles with a particle size of less than 200 nm could be internalized by a receptor-mediated endocytosis uptake mechanism through the corneal tissue [36]. In addition, oleic acid has been reported to improve the ocular drug delivery of both lipophilic and hydrophilic compounds [35]. Until now, no clear mechanisms have been reported for the ocular penetration-enhancing effect of oleic acid; however, oleic acid could produce transient ultrastructure changes in corneal epithelium, offering pathways for actives by perturbation of the highly ordered lipid bilayer.
Some previous studies evaluated MOX-loaded nanoparticle formulations. Shah et al. prepared NE formulations with 12% w/v Tween ® 80, which is in excess of the FDA inactive ingredient database concentration for topical ocular applications by three times, which could lead to irritation and toxicity [54]. Gade et al. prepared MOX (0.2% w/v)-loaded nanostructured lipid carriers (NLC; MOX-NLC). The low drug loading, which lacks a comparison with commercial eyedrops, makes evaluation of any improvement in permeation difficult [55]. In another study, MOX ocular inserts were prepared by Sebastián et al. [56]. However, these ocular inserts were prepared by the solvent cast approach, which raises concerns about residual solvents within the final dosage form, and the prepared inserts were non-biodegradable, therefore requiring removal after each use, which could reduce patient compliance.

Stability Studies of NE and Mucoadhesive NE Formulations
The physicochemical stability of the non-sterilized MOX-NE and MOX-NEM formulations was evaluated at RF and RT, as well as under accelerated (40 • C) storage conditions for 90 days (last time point tested). The effect of storage conditions on PS, PDI, ZP, pH, and drug content is shown in Figures 5-7. None of the three formulations showed any precipitation or cracking upon visual examination. Moreover, there was no significant difference (p > 0.05) observed in DS, PDI, pH, ZP, and drug content after three months of storage.

Conclusions
MOX-loaded NEs and NEM formulations were successfully prepared using HPMC K4M and PVP K29/32 as mucoadhesive agents. MOX-NE and MOX-NEM formulations were sterilized by filtration and moist heat, and the autoclaved formulations were found to be stable for at least one month (last time point tested) after sterilization. Moreover, in vitro release testing showed sustained MOX release profiles from both MOX-NE and MOX-NEM formulations. Ex vivo transcorneal permeation studies revealed a 2.0-fold improvement in permeability and flux from MOX-NE when compared to commercial eyedrops. Both MOX-NEM formulations showed MOX permeability and flux coefficients similar to those of commercial eye drops. Overall, the NE and NEM formulations developed in this investigation appear to be suitable for ocular permeation enhancement of the drug due to the prolonged contact time with the ocular surface and/or extended drug release from the prepared formulation. MOX-NEM formulations could decrease the need for frequent dosing, improve therapeutic outcomes, and increase patient compliance compared to conventional commercial eyedrops. No apparent difference in the functionalities of the two mucoadhesive polymers was observed based on the experiments conducted, although MOX release from the PVP-containing NE formulation was slightly faster than that from the HPMC-containing formulation. Future in vivo studies could elucidate differences in ocular surface retention and delivery based on the mucoadhesive characteris-

Conclusions
MOX-loaded NEs and NEM formulations were successfully prepared using HPMC K4M and PVP K29/32 as mucoadhesive agents. MOX-NE and MOX-NEM formulations were sterilized by filtration and moist heat, and the autoclaved formulations were found to be stable for at least one month (last time point tested) after sterilization. Moreover, in vitro release testing showed sustained MOX release profiles from both MOX-NE and MOX-NEM formulations. Ex vivo transcorneal permeation studies revealed a 2.0-fold improvement in permeability and flux from MOX-NE when compared to commercial eyedrops. Both MOX-NEM formulations showed MOX permeability and flux coefficients similar to those of commercial eye drops. Overall, the NE and NEM formulations developed in this investigation appear to be suitable for ocular permeation enhancement of the drug due to the prolonged contact time with the ocular surface and/or extended drug release from the prepared formulation. MOX-NEM formulations could decrease the need for frequent dosing, improve therapeutic outcomes, and increase patient compliance compared to conventional commercial eyedrops. No apparent difference in the functionalities of the two mucoadhesive polymers was observed based on the experiments conducted, although MOX release from the PVP-containing NE formulation was slightly faster than that from the HPMC-containing formulation. Future in vivo studies could elucidate differences in ocular surface retention and delivery based on the mucoadhesive characteristics of the two polymers. In summary, the NE formulations represent a promising MOX delivery platform for the treatment of ocular bacterial infections.

Data Availability Statement:
The data presented in this study are available upon request from the corresponding author.