Sustained Ocular Delivery of Moxifloxacin–Ufasomes-Laden In Situ Gel for Keratitis Management

TL;DR

This study develops a new in situ gel containing moxifloxacin-loaded ufasomes to provide sustained drug delivery for treating keratitis, an eye infection.

Contribution

A novel ufasome-based in situ gel formulation is developed for sustained ocular delivery of moxifloxacin.

Findings

The optimized ufasomes had high entrapment efficiency (78.37%) and suitable particle size (203.13 nm) for ocular delivery.

The in situ gel retained moxifloxacin for over 6 hours and showed effective antimicrobial activity against Pseudomonas aeruginosa.

The formulation was well-tolerated in animal studies and improved drug bioavailability for keratitis treatment.

Abstract

Background/Objectives: Keratitis is an ocular disease caused by microbial infection or by non-infectious damage due to UV light exposure, chemical exposure, or eye injuries. Methods: Moxifloxacin-loaded ufasomes (MOX-UFAs) were optimized using a full factorial design (12.23) after being prepared by the vortex mixing method. The study evaluated the effects of the oleic acid amount, surface active agent (SAA) amount, and SAA type as independent factors on the entrapment efficiency percent (EE%), particle size (PS), polydispersity index (PDI), zeta potential (ZP), and the amount released after 6 h (Q6h%). Results: The optimized ufasomes (UFAs) formulation was spherical, with an EE% of 78.37 ± 3.91%, PS of 203.13 ± 20.31 nm, PDI of 0.334 ± 0.016, and ZP of −25.42 ± 1.27 mV. The in vitro release of moxifloxacin (MOX) from the UFAs was maintained for more than 6 h in the range of 40.0–75.0%. The optimum MOX-UFAs formulation was incorporated into an in situ gel (Pluronic F-127/HPMC K4M). The ex vivo studies (corneal permeation and confocal laser scanning microscopy) proved the successful retention of the MOX-UFAs-laden in situ gel. Furthermore, the in vitro and in vivo antimicrobial studies revealed their significant antimicrobial effect against Pseudomonas aeruginosa. In addition, the Draize test proved the tolerability of MOX-UFAs-laden in situ gel in animals. Conclusions: The incorporation of MOX-UFAs into Pluronic F-127/HPMC K4M in situ gel could successfully provide sustained ocular delivery and improve the bioavailability of MOX for the management of keratitis.

Figures

Full text

1. Introduction

Keratitis is an ocular disease caused by microbial infection or by non-infectious damage due to UV light exposure, chemical exposure, or eye injuries, with the reported incidence rates ranging from approximately 2.5 to 40.3 cases per 100,000 persons annually in developed regions, and at higher frequencies in parts of Africa, Asia, and the Middle East [1]. Microbial keratitis is a severe infection commonly caused by bacteria [2], and can lead to ulceration of the cornea, resulting in damage, neovascularization, and vision loss [3]. Pseudomonas aeruginosa is the primary cause responsible for microbial keratitis, and the clinical manifestation of this infection is generally more severe than that of other bacterial ulcers [4]. Topical antibiotics are the primary treatment for bacterial keratitis. Specialists consider various aspects when selecting an antibiotic regimen, including broad-spectrum efficacy, toxicity, availability, cost, regional epidemiology of infections, and resistance tendencies [1]. Furthermore, the blood–aqueous humor barrier and the avascular structure of the tissue make topical therapy more effective [5].

Moxifloxacin hydrochloride (MOX) (C_21_H_24_FN_3_O_4_·HCl) is one of the fourth-generation fluoroquinolones (Figure A1). It is a safe antimicrobial agent that has good effectiveness against Gram-positive and Gram-negative bacteria [6]. Its broad-spectrum activity can be attributed to the presence of a fluorine atom (at C-6) and a methoxy group (at C-8). It can be used for the treatment of bacterial keratitis either via topical application or by intracameral injection. However, intracameral injection may result in toxic anterior segment syndrome (TASS) and corneal endothelial toxicity [7]. MOX has good aqueous solubility and can mix with the lacrimal fluid when instilled into the eye as a solution. However, it remains in contact with the ocular mucous membrane for a few minutes as lacrimal fluid is continuously released. Additionally, blinking can cause fast drug loss due to drainage into the lacrimal sac via the lower and upper canaliculus and subsequently into the nasolacrimal duct [8]. Consequently, the topical administration of eye drops exhibits limited drug bioavailability because of the short duration of corneal residence, the quick turnover of tear fluid, and nasolacrimal leakage, which yields poor patient compliance due to the need for the frequent instillation of the medication—2–4 h for common infections and every 30 min for extreme corneal ulcers [7]. Also, hydrophilic drugs cannot flow through the aqueous humor because of the epithelium, one of the three layers of the human cornea and the main ocular barrier [9]. As effective methods require extended residence time and corneal penetration to address the short washout period and improve inadequate drug permeation [7], different ocular nanocolloid delivery systems have been reported, such as liposomes [10], polymeric nanoparticles [11], nanocapsules [12], and oleic acid-based vesicles [13]. Previous studies aiming to sustain the ocular availability of MOX have explored various delivery systems, including topical in situ gels [14], nanostructured lipid carriers [7], mucoadhesive polyelectrolyte complexes [15], mucoadhesive microspheres [16], terpesomes and leciplex formulations [17], nanoparticle-loaded in situ gels [18], nanoemulsions [19], and commercially available soft hydrogel contact lenses [20].

Ufasomes (UFAs) are colloidal carriers formed from fatty acids (e.g., oleic or linoleic acid) with their ionized species. Gebicki and Hicks started using UFAs in 1973 and described their stability [21]. Ufasomal vesicles are composed of sealed lipid bilayers formed in an alkaline medium (pH 7–9). This pH range is crucial for the formation of stable, closed lipid bilayer vesicles, as it allows for both the protonated and ionized forms of fatty acids to coexist. Outside this optimal range, the system tends to form micelles at a higher pH or oily emulsions at a lower pH, rather than stable vesicles [22]. The selection of fatty acids and buffers, the addition of cholesterol, the pH, and the divalent cations affect the stability of UFAs [22]. Structurally, UFAs are composed of single-chain amphiphiles arranged into flexible bilayers, with the hydrophobic hydrocarbon tails oriented inward, while the hydrophilic carboxylate head groups face outward. This single-chain architecture results in highly dynamic and deformable bilayers compared with conventional phospholipid liposomes, which exhibit relatively rigid membranes and limited epithelial penetration [23]. Importantly, UFAs offer distinct mechanistic advantages for ocular drug delivery due to the presence of oleic acid, a single-chain unsaturated fatty acid. At a physiological pH (7.4), oleic acid exists in a partially ionized state [24], leading to electrostatic repulsion between adjacent carboxylate head groups and increased bilayer fluidity. In addition, oleic acid is a well-recognized permeation enhancer that interacts with epithelial lipid domains and transiently disrupts membrane packing, thereby facilitating drug transport across biological barriers, including the corneal epithelium [25,26]. Although transfersomes exhibit deformability, their phospholipid-based composition lacks the intrinsic permeability-enhancing effect of oleic acid, making UFAs particularly suitable for improving ocular drug transport while maintaining formulation simplicity and stability [27]. Due to these advantages, UFAs have been utilized in various studies as colloid carriers for enhancing the topical penetration of drug molecules [28,29].

UFAs have been used to enhance the topical delivery of antifungal agents [23], dexamethasone [30], Glycyrrhiza glabra extract [31], and methotrexate [32]. Additionally, UFAs have been investigated for the pulmonary delivery of nintedanib [33], intranasal delivery of cinnarizine [34], and follicular delivery of minoxidil [29]. Moreover, UFAs have been employed to improve the intestinal absorption of a model dye [35]. To the best of our knowledge, no published studies have utilized UFAs to enhance the ocular permeation of drugs. Herein, we demonstrate the encapsulation of MOX in oleic acid-based nanocarriers, aiming to reduce the effective therapeutic dose through sustained release and improved ocular drug permeation.

Over the last decade, researchers have been continuously motivated to solve the problems of conventional ophthalmic formulations. Polymeric in situ gel systems that show rapid sol-to-gel transformation upon receiving biological stimuli, such as temperature, pH, and ion activation, could be designated as systems of choice for ocular drug delivery [36]. Ocular in situ gels have many advantages, such as providing higher bioavailability due to prolonged precorneal residence time [37] and enabling precise and controlled drug delivery [38], as the gelation of in situ gels acts as depots, providing a reservoir for controlled drug release over an extended period, which improve patient compliance [39]. Thus, loading nanocarriers in an in situ gel prepared with a naturally occurring biocompatible polymer can prolong their residence in the eye, as they will be removed slowly upon ocular administration [14].

Several attempts have been made to deliver ophthalmic drugs to the eye using different polymers, such as sodium alginate [40], chitosan [41], hydroxypropyl methyl cellulose (HPMC) [42], and poloxamers (such as Pluronic F-127). Poloxamer micellar systems have been shown to enhance the ocular delivery of antibiotics like ciprofloxacin and improve drug bioavailability [43]. They have demonstrated excellent ocular bioavailability, with two- to four-fold increases compared to standard formulations [44]. Because poloxamers are transparent, they do not interfere with normal vision and, therefore, are most suited for applications in ophthalmology [45]. Ocular barriers (anatomical and physiological) can be bypassed by the addition of colloidal nanocarriers into an in situ gel, which can significantly improve the bioavailability by targeting corneal penetration and augmenting the precorneal retention duration [46]. As far as we know, few investigations have been performed on the ocular delivery of MOX [19,47], and there is no published research about using a UFAs-laden in situ gel to improve the ocular delivery of MOX.

Therefore, we hypothesized that incorporating MOX into a UFAs-laden in situ gel would provide sustained ocular drug release and enhanced antimicrobial efficacy compared to conventional commercial eye drops (0.5%). Accordingly, the objective of this study was to develop, characterize, and evaluate this novel formulation both in vitro and in a rabbit keratitis model.

2. Results and Discussion

2.1. Statistical Analysis of Factorial Design

It is important to determine the factors that could influence the characteristics of a recently developed drug delivery system. In this regard, factorial designs are helpful since they allow for the simultaneous analysis of the effects of several factors. The measured dependent variables (responses) for different MOX-UFAs are represented in Table 1. To find the model with the greatest predicted and adjusted R^2^ to be applied in the analysis, each dependent variable (Y1: EE%; Y2: PS; Y3: PDI; Y4: ZP; and Y5: Q6h%) was evaluated for data fitting to various order models (linear, two-factor interaction, quadratic, and cubic models). All the responses exhibited an adequate high precision in the range of 29.64–40.07, which guaranteed that the model could assess the signal-to-noise ratio and could be used in navigating the design space [48].

2.1.1. Effect of Independent Variables on EE%

The entrapment of drugs in vesicles offers numerous advantages in drug delivery, including enhanced targeting, improved stability, increased solubility, and reduced toxicity. These benefits make vesicle-based drug delivery systems important tools in pharmaceutical research and development, with the potential to revolutionize the treatment of various diseases and conditions. Figure 1a and Table 2 demonstrate that MOX’s EE% varied from 40.33 ± 2.01 to 78.37 ± 3.91%. The best model was the two-factor interaction model, as it fit the data with high adequate precision (40.07), had a minor difference between the predicted and adjusted R^2^ (0.0112), and had a lack of fit that was non-significant (p = 1). The following equation displays the polynomial model for EE% (Y1):

$$[eqn]$$

The amount of oleic acid (X1) was found to have a significant positive influence on the EE% (p < 0.0001). It was seen that the drug-loading capacity of the oleic acid vesicles depends upon the ratio of the drug to oleic acid. Increasing the oleic acid content to 100 mg resulted in a drug:oil ratio of 1:2, which increased the EE% of UFAs. At lower oleic concentrations of 50 and 75 mg (drug:oil ratios of 1:1 and 1:1.5, respectively), the drug’s EE% was reduced due to drug saturation in the bilayer domain, which could have destabilized the vesicle membrane and led to the leakage of content [32]. This result agrees with Verma et al.’s conclusion that the entrapment of clotrimazole improved with a drug/oleic acid molar ratio of 4:6 [49].

Furthermore, the EE% significantly decreased as the amount of SAA (X2) increased (p < 0.0001). This outcome may be explained by enhanced drug diffusion from the vesicles caused by the increase in the SAA [50].

The EE% was strongly affected by the SAA type (X3) (p < 0.0001). Compared to those made with T80, the EE% values of UFAs containing T20 were higher, which might be attributed to the difference in the hydrophilicities of the two surfactants, as T20 is more hydrophilic (log P = 2.5, HLB = 16.7) than T80 (log P = 4.8, HLB = 14.9), and MOX is also hydrophilic. Also, using a hydrophilic SAA (CTAB) results in a higher EE% of MOX than a lipophilic SAA (DDAB), according to Albash et al. [17]. The same was stated by Ruckmani et al., as the EE% of zidovudine increased in the presence of T20 compared to T60 and T80, which are less hydrophilic [51].

2.1.2. Effect of Independent Variables on PS

The previous literature has stated that a small PS is significant for formulating safe and efficient ocular drug delivery without irritation. From a biopharmaceutical perspective, this can be ascribed to the fact that a small particle size enhances the rapid absorption of nanoparticles through the mucus layer of the tear film and promotes ocular uptake by epithelial cells [52].

The PS of the fabricated MOX-UFAs ranged from 194.37 ± 19.43 to 587.98 ± 58.79 nm, as presented in Table 2 and Figure 1b.

The following equation displays the polynomial model for PS (Y2):

$$[eqn]$$

The PS of the formed MOX-UFAs was shown to increase significantly (p < 0.0001) in response to an increase in the amount of oleic acid (X1), as larger bilayer vesicles formed to minimize the exposure of the chains of hydrocarbons to water [53,54]. This result is well correlated with those of Salimi et al., Brgles et al., and John et al. They described that vesicles with high lipid contents were more flexible and had a larger PS because they could hold more molecules [55,56,57].

In addition, increasing the amount of SAA (X_2_) significantly increased the PS (p < 0.0001). Kaur et al. previously reached a similar conclusion, finding that the PS of oxiconazole-loaded unsaturated fatty acid liposomes increased as the amount of SAA increased [50]. Aziz et al. suggested that high amounts of SAA would build up and form several layers around vesicles, which would increase the PS [58].

Furthermore, the SAA type (X_3_) significantly influenced the PS (p < 0.0009), as the PS of T80-based UFAs was greater than that of T20-based UFAs. This might be related to the variance in the length of alkyl chains, as T80 is composed of a longer alkyl chain than T20, thus producing UFAs with a larger PS [59].

2.1.3. Effect of Independent Variables on PDI

In terms of PDI, a dispersion that is fully heterogeneous is indicated by a value of 1, whereas a perfectly homogeneous one is indicated by a value of 0 [60]. The prepared UFAs showed PDI values between 0.3211 ± 0.04 and 0.5532 ± 0.02, which reflect the formation of uniform dispersion with a narrow size distribution (Table 2). The effects of different variables, such as the oleic acid amount (X1), SAA amount (X2), and SAA type (X3), on the PDI of UFAs are represented in Figure 2a. It was observed that the results for the PDI had a good correlation with the PS measurements [61]. Hence, the UFAs with the highest PS had the highest PDI value.

The following equation represents the polynomial model for PDI (Y3):

$$[eqn]$$

2.1.4. Effect of Independent Variables on ZP

An ideal charged surface can enhance physical stability by causing electrostatic repulsion among the vesicles, preventing them from aggregating or fusing [51]. The formulated UFAs demonstrated negatively charged values of ZP from −12.70 ± 0.56 to −26.8 ± 0.80 mV (Table 2). The effect of the independent variables (oleic acid amount, SAA amount, and SAA type) on the ZP is shown in Table 3 and Figure 2b. Although the UFAs exhibited moderate absolute ZP values, the presence of non-ionic SAAs (T20, T80) imparted steric stabilization. It was previously reported that the stability of ufasomes is not governed only by electrostatic repulsion. The presence of Tween in the formulation was expected to provide steric stabilization, arising from the adsorption of non-ionic surfactant molecules at the vesicle surface, which forms a hydrated polymeric layer that hinders vesicle–vesicle aggregation. Such steric effects can significantly enhance colloidal stability even at relatively low zeta potential values [62].

The following equation indicates the polynomial model for the ZP (Y4):

$$[eqn]$$

As the oleic acid amount (X_1_) increased, the ZP significantly increased, which agrees with Manca et al., who obtained comparable outcomes for rifampicin liposomes. In addition, studies performed by Ahmed et al. and Bolla et al. had similar findings for the preparation of propranolol hydrochloride-laden biconjugate UFAs and clotrimazole-loaded UFAs, respectively [59,63,64]. This may be attributed to the ionization of oleic acid that makes the dispersion anionic in nature [65].

Regarding the amount of SAA (X_2_), it was found that the ZP decreased significantly (p < 0.0001) as the amount of SAA increased, possibly due to the bulkiness of SAA molecules obscuring the surface charge of UFAs [66]. On the other hand, non-ionic SAAs promote expansion of the diffuse layer, leading to a lower zeta potential [67].

According to the ANOVA results, the ZP of UFAs dispersions was significantly (p < 0.0001) affected by changing the type of the SAA (X3). Its value shifted towards negative as the SAA’s HLB values increased. Smaller absolute values of ZP, between −15.44 ± 0.77 and −17.66 ± 0.88 mV, were obtained when T80 (HLB 15) was used to formulate the UFAs. However, when T20 (HLB 16.7) was used, the vesicles had higher absolute ZP values, ranging from −13.18 ± 0.65 to −25.42 ± 1.27 mV. This observation could be attributed to the surface energy, which is directly proportional to the HLB values. As the surface energy of the vesicles increases, the ZP values increase toward negative [68].

2.1.5. Effect of Independent Variables on Q6h (%)

Establishing the profiles of drug release from pharmaceutical dosage forms is valuable for predicting the in vivo action of a drug. The cumulative percentage of MOX released from the different UFAs is illustrated at definite time intervals in Figure 3. The MOX solution showed a 96% release in 5 h, which means that the drug and the dialysis membrane did not interact. The prepared MOX-UFAs formulations had a slower release rate than the free MOX solution. The cumulative percentage of MOX released after 6 h (Q6h%) was in the range of 40.0 ± 2.0 to 75.0 ± 3.75% for the different UFAs (Table 2 and Figure 4).

UFAs can maintain drug release for more than 6 h due to their oleic acid content, which can act as a reservoir for drug delivery. This finding has been reported in different studies to confirm that oleic acid vesicles can encapsulate drugs, regulate their release rate, and provide a mechanism for the slow and controlled release of drugs [69]. It is worth noting that UFAs with smaller PSs had a faster in vitro drug release rate due to the large surface area [70].

The effect of oleic acid amount (X_1_), SAA amount (X_2_), and SAA type (X_3_) on the Q6h (%) was significant (p < 0.0001) according to the ANONA (Table 3).

The following equation shows the polynomial model for Q6h% (Y5):

$$[eqn]$$

As the amount of oleic acid (X1) increases, there is a decrease in the Q6h%. Increasing the oleic acid content in UFAs dispersions leads to the formation of larger vesicles with reduced surface area that are exposed to the release media, consequently lowering MOX release. This observation is consistent with the findings reported by Ahmed et al., who noted that increasing the oleic acid content correlates with a diminished release profile of propranolol hydrochloride from UFAs [59].

Regarding the SAA amount (X2), it is clear that the Q6h% increased as the SAA amount increased. This result might be attributed to the micelle development in the bilayer, which could have improved the permeability of the membrane and the release of the drug [71].

From the ANOVA, the significant effect (p < 0.0001) of SAA type (X3) on the Q6h% was recognized. The T20-based UFAs had a higher Q6h% than those of T80-based formulations. This outcome agrees with Aboud et al. [72], who stated that SAA with a lower value of HLB and longer alkyl chains slow down release because SAAs that have longer alkyl chain lengths (T80) might produce more packed vesicles, which become less permeable when compared with other types of SAAs with shorter alkyl chain lengths (T20) [73].

2.1.6. Kinetics of Release

According to the calculated values of the regression coefficient, the Higuchi model, which suggests that drug release is controlled and proceeds via diffusion from the delivery system, best expressed the in vitro release profile of MOX from all UFAs formulations (Table A1).

2.1.7. Selection of the Optimum MOX-UFAs

Through a graphical and numerical analysis using Design Expert^®^ software, the appropriate levels of the variables were identified to formulate MOX-UFAs with the lowest PS, PDI, and Q6h%, as well as the highest EE% and ZP. F3 was selected as the optimum formula that maintained a desirability of 0.883. Therefore, F3 was used for further investigations. The DL% was calculated for F3 and was found to be 39.10 ± 0.15.

The f2 value was 22.62 (below 50), which indicates a significant difference between the two release profiles (MOX solution and optimum MOX-UFA). This lack of similarity confirms that the loading of MOX in ufasomes effectively modified its release behavior, resulting in sustained drug release compared to the immediate release observed with the free MOX solution.

2.2. Transmission Electron Microscopy (TEM)

The morphological investigation of the optimal MOX-UFAs (F3) was performed using a TEM. It exhibited sphere-shaped particles with a homogenous size distribution (Figure 5). Their size was consistent with the measurement recorded by the Zetasizer.

2.3. Effect of Storage on the Optimum UFAs

When the optimum UFAs formulation (F3) was stored at 4 °C, it retained its characteristics, showing values of 70.47 ± 4.81% for the EE%, 220.63 ± 19.31 nm for the PS, 0.344 ± 0.019 for the PDI, 22.52 ± 1.33 mV for the ZP, and 45.67 ± 3.08% for the Q6h% after 6 months. Similarly, at the end of the storage period at 25 °C, the values recorded for the aforementioned parameters were 76.83 ± 1.06%, 198.97 ± 6.45 nm, 0.339 ± 0.011, 26.22 ± 1.17 mV and 43.12 ± 3.22%, respectively. No statistically significant differences were observed between the stored and freshly prepared formulations (p > 0.05). Furthermore, the optimal formulation (F3) maintained its stability and showed no signs of physical alteration or aggregation throughout the storage period. This result was verified by calculating the similarity factors (f2) at 4 °C and 25 °C, which were above 50 (94.62 and 91.84, respectively), indicating that storage did not noticeably affect the release profile of the MOX from the UFAs. The stability of the vesicles is attributed to the strong hydrogen bond interactions between the protonated and deprotonated groups, specifically RCOO^−^⋯HOOCR [74].

2.4. Fourier-Transform Infrared Spectroscopy (FT-IR) Studies

An FTIR analysis was carried out to detect any chemical interaction between the MOX and the UFAs components. As displayed in Figure 6, the spectrum of MOX exhibited an N-H stretching band at 3450.65 cm^−1^ and an O-H carboxylic acid stretching band at 3161.33 cm^−1^. It also showed C=O stretching at 1708.15 cm^–1^ and 1624 cm^–1^; aromatic C=C stretching bands at 1618.13 cm^−1^, 1517.31 cm^–1^ and 1452.40 cm^–1^; C-N bending at 1354.03 cm^–1^; stretching of monofluorobenzene at 1184.29 cm^–1^; and C-H bending for the substituted benzene at 804.39 cm^–1^. The appearance of the distinctive peaks of MOX in the spectrum of the UFAs confirmed its successful loading and the absence of chemical interactions [75]. The labeled main characteristic peaks corresponding to the different functional groups are illustrated in Figure A2.

2.5. Characterization of MOX-UFAs-Laden In Situ Gels

The ocular application of UFAs as an aqueous dispersion may result in rapid drainage and low bioavailability due to their low viscosity. Hence, incorporating the prepared MOX-UFAs dispersion into an in situ gel may facilitate the topical application of UFAs and support their deposition in tissues as the viscosity increases [76]. A physical inspection of the prepared in situ gel showed it was translucent, with no grittiness on touch.

2.5.1. Evaluation of pH

The prepared MOX-UFAs in situ gel showed a pH value of 7.4 ± 0.241, which is near the pH of the eye, eliminating the possibility of irritation or inflammation [77].

2.5.2. Gelling Temperature (Tsol/gel)

The optimum MOX-UFAs-laden in situ gel showed a Tsol/gel of 34 ± 0.70 °C. An optimal ophthalmic thermosensitive in situ gel should be converted into a gel at the precorneal temperature (35 °C) and have a transition temperature (Tsol/gel) above room temperature, ideally 30 °C, when diluted by a few drops of tear fluid [78].

2.5.3. Gelation Time

When the system is exposed to its gelation temperature, it should be converted into gel instantly or in just a few minutes to avoid being rapidly removed by tear fluid. The MOX-UFAs-laden in situ gel recorded a quick gelation time. It was about 30 ± 1.23 s.

2.5.4. Rheology

By studying the rheological behavior of the MOX-UFAs-laden in situ gel at 34 °C, it was clear that as the shear rate increased, the viscosity decreased, which suggests a shear-thinning (pseudoplastic) flow. To confirm this behavior, Farrow’s constant (N) was obtained from Farrow’s equation, and it was 4.20, which surpassed one [79]. The association of polymeric molecules within the MOX-UFAs-laden in situ gel under relaxed conditions may explain this behavior, but when shear stress is applied, the molecules of polymers disentangle, expelling the captured solvent and lowering the apparent viscosity [79]. For ocular applications, pseudoplastic fluid is ideal due to its ability to minimize interference with the thin fluid layer over the corneal surface. The shear-thinning properties of the gel, as a result of blinking, enable an even distribution of the formulation over the eye’s surface. On the other hand, the low shear rates allow for the good adherence of the formulation to the corneal surface due to the high viscosity [80].

2.6. Ex Vivo Studies

2.6.1. Ex Vivo Permeation Studies

The distinct anatomy of the cornea makes drug transport across it a challenging prospect; corneal permeability is the major issue in the development of topical ocular formulations [81]. As shown in Figure 7, a comparative permeation study was conducted over 6 h to evaluate the quantity of MOX that permeated via the cornea from the MOX solution, the optimum MOX-UFAs (F3), and the MOX-UFAs-laden in situ gel. Table 4 presents the calculated permeation parameters of MOX from the studied formulations. The highest cumulative amounts of MOX permeation, steady-state flux (J_ss_), and permeation coefficient (KP) were for the aqueous solution, which may be attributed to the presence of MOX in its dissolved form. Also, the distinct structure and biphasic solubility of MOX could be the reason for this result [82]. It agrees with Polat et al. and Khalil et al., who found a higher drug transport level was obtained from an aqueous solution than the developed ocular insert formulations in [83,84]. Moreover, the presence of oleic acid might have decreased the permeability profile of MOX in the MOX-UFAs and MOX-UFAs-laden in situ gel in comparison to the MOX solution, as it can form depots in the cornea layers that sustain the permeation of MOX [85,86,87]. Both the MOX-UFAs and the MOX-UFAs-laden in situ gel showed typical permeation profiles but with lower values in the latter, as the Pluronic F-127 imparted viscosity to the gel form [78].

By applying a one-way ANOVA, it was found that the amount of MOX that permeated from the MOX solution was significantly higher than that from the MOX-UFAs (p < 0.01) and the MOX-UFAs-laden in situ gel (p < 0.001). In addition, the amount of MOX that permeated from the MOX-UFAs was found to be significantly higher than that from the MOX-UFAs-laden in situ gel (p < 0.05).

2.6.2. Corneal Hydration

This serves as a significant in vitro indicator of potential corneal irritation, as values ranging from 76% to 83% are considered normal. Values of more than 83% imply that the cornea has various levels of damage. The corneal hydration of the freshly excised cornea in this study was 80.15 ± 3.5, 79.54 ± 2.15, and 81.5 ± 4.1 for the MOX solution, MOX-UFAs, and MOX-UFAs-laden in situ gel, respectively, showing that all hydration values for the intact cornea were within the normal range [88].

2.6.3. Confocal Laser Scanning Microscopy (CLSM) Study

According to the images obtained with CLSM (Figure 8), the capacity of the fluor-labeled UFAs and fluor-labeled UFAs-laden in situ gel that could be deposited into the ocular tissues was evaluated [89]. They demonstrated a high fluorescence concentration in the various corneal tissues. The photographs of the longitudinal section provide a detailed representation of the distribution in the corneal layers. When the maximal fluorescent light intensity was calculated, the FDA solution (Figure 8a), the optimum UFAs (Figure 8b), and the UFAs-laden in situ gel (Figure 8c) showed significant differences (p < 0.05), with average intensities of 145.83 ± 4.96, 223.06 ± 5.65, and 243.06 ± 7.41, respectively. Notably, the CLSM results and the ex vivo permeation data agree well. This result verifies the effective permeation of the UFAs and their successful deposition into the corneal tissues by their incorporation into an in situ gel form.

2.7. In Vitro Antibacterial Studies

According to the findings of the agar diffusion test for antimicrobial susceptibility, the results reveal clear zones of inhibition, with diameters of 45 ± 0.50 mm, 53 ± 0.91 mm, and 55 ± 0.45 mm for the MOX solution, the optimum MOX-UFAs (F3), and the MOX-UFAs-laden in situ gel, respectively, after 24 h. It is clear that the optimum UFAs and the MOX-UFAs-laden in situ gel exhibited larger zones of inhibition compared to the MOX solution. This result can be attributed to the fact that the MOX can continuously penetrate the bacteria due to the adsorption and fusion of UFAs with the cell walls, in addition to the mucoadhesive and viscosity-inducing properties of the in situ gel, and considering the aim of increasing the antibacterial activity by attaining sustainable transcorneal drug permeation rates [90,91].

2.8. In Vivo Studies

2.8.1. In Vivo Studies

Bacterial keratitis was induced in rabbit eyes to evaluate the effectiveness of the different formulations. Keratitis was confirmed in all the experimental animals 48 h post-infection. Clinical scores of the four groups were assessed before and after treatment, and the results are presented in Table 5. The scores are expressed as the mean ± SD (n = 3). The infected eyes displayed extensive clinical inflammatory signs, as shown in Figure 9a, with no significant differences among the groups based on a one-way ANOVA of total clinical scores (p > 0.05). The clinical condition of the eyes after 10 days of treatment is shown in Figure 9b.

After 10 days, the control group (GP IV) had a total score of 16.3 ± 0.58, which was significantly higher than the scores of the other three groups that received the MOX treatment in the different formulations (p < 0.0001). When comparing the total clinical scores of GP I (receiving commercial MOX eye drops) and GP II (receiving MOX-UFAs), no statistically significant difference was observed (p > 0.05). However, the total score of GP III (receiving MOX-UFAs-laden in situ gel) was significantly lower than that of both GP I and GP II (p < 0.05).

These results indicate that MOX is effective at treating bacterial keratitis caused by Pseudomonas aeruginosa, which is consistent with the findings from previous studies [15]. However, some inflammatory signs persisted in the animals in GP I and GP II after 10 days of treatment, with average total scores of 3.3 ± 1.16 and 2.67 ± 1.53, respectively. This may be attributed to the rapid and extensive precorneal drug loss due to drainage and the high tear fluid turnover. Consequently, the ocular residence time of commercial eye drops and the MOX-UFAs was limited, requiring more frequent administration than the twice-daily dosing used in this in vivo study. In contrast, the eyes of animals in GP III, which received the MOX-UFAs-laden in situ gel, showed a complete recovery and normal appearance, confirmed by a total clinical score of 0. This improved outcome may be due to the increased contact time between the dosage form and the corneal surface caused by the in situ gel, which most likely allowed the formation of MOX depot in the corneal tissues, aided by the UFAs and the oleic acid content. Thus, the combination of UFAs and in situ gel enabled the administration of MOX in the form of eye drops, which transformed into a gel due to the ocular temperature. This transformation was facilitated by the synergistic effect of Pluronic F-127 and HPMC K4M, resulting in a gel with enhanced viscosity and prolonged ocular residence time, and caused a sustained MOX therapeutic effect. Similar results were observed by Sanjay et al. in [15], where animals treated with MOX-loaded cationized gelatin alginate nanoparticles showed significantly lower CFU/mL and clinical scores than the animals treated with a MOX solution, which was attributed to the penetration efficiency of the nanoparticles.

2.8.2. Microbiological Evaluation

The microbiological study evaluated the bacterial load in the ocular tissue infected with Pseudomonas aeruginosa without and with antibacterial therapy. Under sterile conditions, the diseased and treated eyes of rabbits were sectioned into the anterior and posterior parts. It was found that the untreated eyes had the highest (p < 0.05) CFU, with 99.5 ± 0.408 × 10^5^ and 89.00 ± 0.816 × 10^5^ for the anterior and posterior parts, respectively. In contrast, the MOX-UFAs-laden in situ gel showed the highest efficacy, with the lowest CFU of 0.15 ± 0.05 × 10^5^ in the anterior chamber and no detectable CFU in the posterior one (Table A2 and Figure 10) (significance was determined through a one-way ANOVA by SPSSVR software 22.0). A post hoc test was performed employing Tukey’s honest significant difference (HSD) test. This suggested that the MOX-UFAs-laden in situ gel sustained the release of MOX and enhanced its activity. This effect was likely due to two factors: first, the synergistic antibacterial properties of oleic acid, which enhanced the effectiveness of the nanocarrier-loaded antibacterials [92], and second, the bioadhesive properties of the UFAs-laden gel, which are attributed to the presence of Pluronic F-127, which improves corneal retention and sustains drug release [93].

2.8.3. Ocular Irritation Study

The conjunctiva and other parts of the eye were examined through external observation under appropriate lighting following the administration of MOX-UFAs-laden in situ gel into the rabbits’ eyes. The total Draize score was recorded at each time point to assess any potential ocular irritation caused by the formulation. The findings showed no changes in the conjunctiva or other ocular structures. Additionally, there were no noticeable signs of conjunctival discharge, redness, or chemosis in any of the rabbits after the gel was applied. This confirms that the MOX-UFAs-laden in situ gel showed non-irritant properties to the rabbits’ eyes.

2.9. Histopathological Studies

The four groups, including GP I (MOX solution), GP II (optimal MOX-UFAs), GP III (MOX-UFAs-laden in situ gel), and GP IV (positive untreated control), in addition to the normal left eyes, all had their tissues examined. Figure 11a shows the normal histological structure of the cornea of the normal left eye. The cornea of GP I, treated with the MOX solution, shows stromal infiltration by low numbers of mononuclear inflammatory cells (arrow), with neovascularization of various sizes (Figure 11b). Regarding GP II, the optimum MOX-UFA-treated group, the photomicrograph shows a normal corneal histology (Figure 11c). For GP III, the optimum MOX-UFAs-laden in situ gel-treated group, the photomicrograph shows a normal histological structure of the cornea (Figure 11d). The positive control group (GP IV) shows inflammation of the cornea, in which there is stromal infiltration by high numbers of inflammatory cells, mainly neutrophils, eosinophils, and lymphocytes with necrotic debris (star) (Figure 11e). The data from the in vivo studies do not clearly show corneal intolerability and suggest that both the optimal MOX-UFAs and MOX-UFAs-laden in situ gel are safe and not suspected of inducing corneal irritation in clinical investigations.

3. Materials and Methods

3.1. Materials

Moxifloxacin hydrochloride (MOX) was donated by Al Kahira Pharmaceutical Co. (Cairo, Egypt). Oleic acid, Tween 80 (T80), Tween 20 (T20), methanol, and ethanol (95%) analytical grades were purchased from El Nasr Pharmaceutical Chemical Company (Cairo, Egypt). Pluronic F-127 was obtained from BASF (Ludwigshafen, Germany). Hydroxypropyl methylcellulose (HPMC K4M, 4000 cps) was provided as gift from Colorcon (Dartford, UK). Spectra Por semipermeable membrane (molecular weight cutoff 12,000–14,000) was obtained from Spectrum Laboratories (USA). Fluorescein diacetate (FDA) was purchased from Sigma Aldrich (St. Louis, MO, USA).

3.2. Preparation of MOX-UFAs

Different MOX-UFAs were prepared by the vortex mixing technique as described by Kaur et al. [50]. Briefly, MOX (50 mg) and varying amounts of oleic acid (50–100 mg) and SAA (Tween 80 or Tween 20) in a range between 25 and 75 mg were dissolved in 2 mL ethanol, as presented in Table 1. The resultant solution was mixed with 10 mL phosphate buffer saline (PBS, pH 7.4) using a vortex (Boekel, 270100 Tap Dancer-Vortex Mixer, Feasterville-Trevose, PA, USA). The formed dispersion was then subjected to 15 min of sonication (Ultra Sonicator, Model LC 60/H Elma, Singen, Germany) to obtain UFAs dispersions with uniformly reduced particle size.

3.3. Optimization of MOX-UFAs

A 1^2^.2^3^ full factorial design was used to optimize the MOX-UFAs by Design-Expert^®^ Software (Version 13, Stat-Ease Inc., Minneapolis, MN, USA). While the type of SAA (X3) was set as an independent categoric variable and was evaluated for changing between two different types, namely, T80 and T20, the impact of the amount of oleic acid (X1) and SAA (X2), as numerical independent variables, on the properties of UFAs was studied at three levels (−1, 0, +1). The other formulation variables remained unchanged. The mathematical correlations between the five dependent variables—the entrapment efficiency (EE; Y1), particle size (PS; Y2), polydispersity index (PDI; Y3), zeta potential (ZP; Y4), and amount of drug released after 6 h (Q6h%; Y5)—and the previously identified independent variables were obtained using the design presented in Table 1. There were 18 runs in the design, as shown in Table 2.

3.4. Characterization of MOX-UFAs

3.4.1. Determination of EE%

Based on the indirect method for the determination of EE%, 1 mL of MOX-UFAs dispersion was centrifuged at 20,000 rpm (cooling ultracentrifuge 3–30 K; Sigma, Humburg, Germany) at 4 °C for two hours. The amount of free MOX in the supernatant of each UFAs formulation was measured spectrophotometrically (UV spectrophotometer; Shimadzu, Kyoto, Japan) at 288 nm, and then the EE% of MOX was calculated according to the following equation [80]:

$$[eqn]$$

where A represents the total MOX content in the UFAs and B represents the free MOX content in the supernatant. The measurements were done in triplicate, and the results are presented as the mean ± SD.

3.4.2. Determination of PS, PDI, and ZP

The PS, PDI, and ZP of the MOX-UFAs were determined after dilution of each sample with distilled water by a Malvern Zetasizer (Malvern Zetasizer Nano ZS, Ver. 6.20, Malvern, UK), which is based on dynamic light scattering (DLS) at 25 °C. The light scattering angle was fixed at 90°. Each measurement was made three times, and the results are presented as the mean ± SD.

3.4.3. In Vitro Release of MOX-UFAs

A USP dissolution apparatus (Hanson RS-8 Plus, Chatsworth, CA, USA) was used to study the release behavior of MOX from the different UFAs through a cellulose membrane that was previously soaked in PBS (pH 7.4). To the rotating shaft (50 rpm) of the USP dissolution equipment, a glass tube that contained formulations (MOX solution or MOX-UFAs) containing 5 mg of MOX was fixed. One end of the tube was closed by a cellulose membrane and was then submerged in 50 mL of PBS (pH 7.4) at 35 °C ± 0.5. At definite time intervals (0.25, 0.5, 1, 1.5, 2, 3, 4, 5, and 6 h), the withdrawn samples were evaluated spectrophotometrically at a λ_max_ of 288 nm for the released amount of MOX. The experiment was performed in triplicate. To keep the dissolution medium’s volume constant, 1 mL of it was removed and replaced with the same volume of fresh PBS.

3.4.4. Kinetic Analysis of the Release Data

The data from the release experiment were fitted to various kinetic models, including zero, first-order, Higuchi, Korsmeyer–Peppas, and Hixson–Crowell models, to determine the drug release mechanism of MOX from the UFAs according to the best-fit model with the highest R^2^.

3.4.5. Selection of the Optimum Formula

The best formula was determined by considering the desirability function of Design-Expert^®^ Software, which estimates the ideal levels of the components under study. The criteria for choosing the best formula were to achieve the maximum EE% and ZP (as absolute values), but the lowest PS, PDI, and Q6h%.

The release profiles of MOX solution and optimum MOX-UFA were compared using a similarity factor (f2) based on Moore and Flanner’s model-independent approach from the following equation [94]:

$$[eqn]$$

where n represents the number of sampling time points, and R_t_ and T_t_ are the percentages of the drug released from the MOX solution and from the optimum MOX-UFA (F3), respectively, at time t. The profiles are identical when the f2 value is 100. They are considered similar if the value is 50–100 and not similar when F2 < 50.

3.5. Drug Loading

Drug loading was calculated for the optimum selected MOX-UFAs formulation using the following equation:

$$[eqn]$$

3.6. Morphological Examination

On top of a grid coated with carbon, a drop of 1% phosphotungstic acid was used for negative staining of the optimum MOX-UFAs formula after dilution with ten-fold distilled water [95].

3.7. Effect of Storage

The optimum MOX-UFAs formulation was stored in a refrigerator at 4 °C for 6 months. Additionally, F3 was stored at 25 °C for 2 weeks to assess the short-term stability of the optimum MOX-UFAs at room temperature. During the storage period at both temperatures, visual examinations were conducted at specific intervals of 0, 1, 3 and 6 months for storage at 4 °C, and at 0, 5, and 10 days for storage at 25 °C to detect any signs of instability. At the end of the storage period, the EE%, PS, PDI, ZP, and Q6h% were measured and compared with those of a freshly prepared formulation [96]. In addition, the release profiles of the stored and fresh MOX-UFAs were compared using the calculated similarity factor (f2).

3.8. Fourier-Transform Infrared Spectroscopy (FT-IR) Studies

After two hours of freezing at −20 °C, the optimum formula was lyophilized for 48 h. The spectra of MOX and the lyophilized sample of UFAs were obtained using the potassium bromide disc procedure on an FT-IR spectrophotometer (Genesis II Mattson, Madison, WI, USA) between 4000 cm^−1^ and 400 cm^−1^ with a resolution of 4 cm^−1^ [97].

3.9. Preparation of MOX-UFAs-Laden In Situ Gel

The MOX-UFAs-laden in situ gel was prepared using HPMC K4M and Pluronic F-127 polymers. Briefly, a calculated amount of HPMC K4M was dispersed in water and stirred at 70 °C until a uniform dispersion was obtained. The resulting solution was then refrigerated at 4 °C. Subsequently, Pluronic F-127 was added to the cooled HPMC solution and stored at 4 °C for 24 h to ensure complete dissolution. Finally, the optimized MOX-UFAs dispersion (prepared in half the required volume of buffer) was mixed with an equal volume of the polymer solution under continuous stirring to yield the final MOX-UFAs-laden in situ gel containing 0.5% w/v MOX, 1.5% w/v HPMC K4M, and 14% w/v Pluronic F-127 [98,99].

3.10. Characterization of MOX-UFAs-Laden In Situ Gel

3.10.1. Visual Appearance

The appearance of the formulations (color and clarity) was examined visually both before and after the formation of the gel.

3.10.2. Determination of pH

The pH of MOX-UFAs-laden in situ gel was obtained from the average of three readings using a calibrated pH meter (Hanna, type 211, Cluj-Napoca, Romania).

3.10.3. Determination of Gelation Time

The gelation time of the MOX-UFAs-laden in situ gel was assessed through the tube inversion method by placing 2 mL of the refrigerated gel, at 4 °C, in a test tube that was then immersed in a water bath (34 °C). The gelation progress was monitored by periodically inverting the tube. The time when no flow was observed during inversion was the gelation time [100,101].

3.10.4. Determination of Gelation Temperature (Tsol/gel)

To determine the gelation temperature, a vial with a magnetic bar was filled with 10 g of MOX-UFAs-laden in situ gel, which was then heated gradually by 1 °C/min starting from 20 °C with constant stirring at 100 rpm. The average gelation temperature of triplet readings was identified from the temperature at which gelation occurred and hindered the rotation of the magnetic bar [102,103].

3.10.5. Rheology

The viscosity assessment of MOX-UFAs-laden in situ gel was performed using a viscometer (Brookfield cone and plate, Middleboro, MA, USA), using spindle 52 at 25 °C and 34 °C to determine the impact of temperature variation on viscosity [104]. The freshly prepared gel was placed in the cup of the viscometer, and then continuous shearing at various rates provided measurements for the shear stress (dyne/cm^2^), shear rate (s^−1^), and viscosity in centipoises (cP). Measurements were taken across the entire speed range from 50 to 175 rpm and then from 175 to 50 rpm. The intervals between each two consecutive speeds were set to be 10 s. The shear rate (y-axis) was plotted against the shear stress (x-axis) to illustrate the rheogram. The following equation (Equation (9)) was applied to obtain Farrow’s constant (N) to identify the rheological behavior [79].

$$[eqn]$$

G and F are the shear rate (s^−1^) and shearing stress (dynes/cm^2^), respectively, while N expresses Farrow’s constant and η is the viscosity (cP). The value of N is nearly 1 in the case of Newtonian systems, but less than 1 in shear-thickening systems, and more than 1 when a system exhibits shear-thinning behavior.

3.11. Ex Vivo Permeation Studies

3.11.1. Permeation Parameters

The permeation experiment was conducted using excised bovine corneas in a USP dissolution apparatus by immersing the formulae in a simulated tear fluid as the medium for diffusion at 35 ± 0.5 °C. A total of 1 mL of MOX solution, optimum MOX-UFAs dispersion, and MOX-UFAs-laden in situ gel having the same MOX amount (5 mg) were placed in an open-ended glass cylinder with a 0.78 cm^2^ area of diffusion. At different time intervals, we calculated the permeated amount of MOX per unit area spectrophotometrically at λ_max_ of 288 nm, which was plotted against the time to obtain the slope. The steady-state flux (J_ss_) and permeation coefficient (KP) were calculated as follows [65]:

$$[eqn]$$ $$[eqn]$$

where dQ/dt (μg/h) indicates the slope of the straight-line section of the permeation graph, A is the corneal surface area (cm^2^) from which diffusion occurs, and C_0_ is the starting concentration of MOX in the tested formulations (μg/cm^2^).

3.11.2. Corneal Hydration

The gravimetric analysis method was used to calculate the corneal hydration level. Each cornea was weighed after the permeation experiment, placed in 1 mL of methanol, allowed to dry overnight, and then reweighed. According to Equation (12), the corneal hydration percentage level (HL%) was determined.

$$[eqn]$$

where “HL%” is the percentage of corneal hydration, and “W_t_” and “W_d_” are the wet and dry corneal weights, respectively [80].

3.11.3. Confocal Laser Scanning Microscopy (CLSM)

Fluro-labeled optimum UFAs, fluro-labeled UFAs-laden in situ gel, and FDA solution were formulated in the same previous way, except for substituting 1% (w/v) fluorescein diacetate (FDA) in the aqueous phase for the MOX. These fluro-labeled formulae were deposited in the cornea and tracked using confocal microscopy by fixing the bovine corneas with the formulae placed on them for 6 h, as previously explained in the ex vivo permeation study. Detection of the fluorescence in the tissues of the cornea was performed by cutting sections of the cornea longitudinally, placing them in paraffin wax, and using a microtome (Rotary Leica RM2245; Leica Biosystems, Germany) to cut them into sections. For the FDA, the excitation and emission wavelengths (λ_max_) were 497 and 516 nm, respectively. LSM Image Browser software, version 4.2 (Carl Zeiss Microimaging, Germany) was used to obtain the confocal figures with an analysis of light intensity [105,106]. SPSS^®^ software 22.0 was used to apply an ANOVA test to determine the statistical significance.

3.12. In Vitro Antibacterial Studies

An in vitro experiment was conducted using a cell suspension of Pseudomonas aeruginosa (strain ATCC 27,853) in sterile saline (0.5 McFarland units). The suspension was inoculated to reach a concentration of 10^5^ CFU/mL [107]. The antibacterial efficacy of recently prepared formulae (MOX solution, optimum MOX-UFAs, and MOX-UFAs-laden in situ gel) against Pseudomonas aeruginosa was assessed by the agar cup method. After seeding 0.2 mL of the test microorganism in a Petri plate containing 20 mL of nutritional agar, the plates were left to harden to allow for the formation of cups (10 mm in diameter) in the plates under aseptic conditions. The cups were filled with 100 µL of the following: MOX solution, optimum MOX-UFAs, MOX-UFAs-laden in situ gel, and plain UFAs. The plates were maintained at 25 °C for 4 h and then incubated at 37 °C for 24 h to determine the inhibition zones [108].

3.13. In Vivo Studies

3.13.1. Animals

Adult male New Zealand albino rabbits weighing between 2.8 and 3.5 kg were housed and provided with water and food at 55% relative humidity and 22 °C according to the guidelines of the Care and Use of Laboratory Animals under the supervision of practicing veterinarians. Before the experiments, the rabbits were maintained in a 12 h light/dark cycle for a week to adapt and check for clinical abnormalities. The approval of the experiment design (PH13-2022, approved on 6 December 2022) was obtained from the Ethical Committee of the College of Pharmaceutical Sciences and Drug Manufacturing, Misr University of Science and Technology, Egypt.

3.13.2. In Vivo Antimicrobial Studies

Induction of bacterial keratitis was performed in the right eyes of 12 New Zealand albino rabbits (2.8–3.5 kg), while the left eyes were used as the negative controls. A total of 1 mL of Ketalar^®^ (ketamine hydrochloride, 50 mg/mL) was used intramuscularly to anesthetize the rabbits [107], Isopto^®^ Fenicol eye drops (chloramphenicol, 0.5%) were used to wash any possible bacteria from the conjunctiva, and the superficial epithelium was removed by scraping. Each rabbit’s eye was infected separately using the isolate of Pseudomonas aeruginosa (2 × 10^8^ CFU/mL) by intrastromal injection of 0.1 mL of the bacterial suspension into the center of each rabbit’s cornea to a depth of around half the thickness of the cornea by a 30-gauge needle.

Evaluation of the infected eyes was performed based on the clinical scoring method described by Wenxiang Lin et al. [109], which includes four clinical parameters: discharge, conjunctival hyperemia/edema, hypopyon, and corneal infiltration/edema. Each parameter was graded on a scale from 0 (for normal) to 4 (maximally severe), with a corneal score of 5 assigned in cases of corneal perforation (Table 6). Infection was confirmed 48 h post-inoculation by the presence of a 3 ≥ corneal score < 5.

The experiment was performed after dividing the animals with a validated infection into four groups. GP I, II, and III received 50 μL aliquots of MOX solution (fortymox, 0.5%), optimum MOX-UFAs (0.5%), and MOX-UFAs-laden in situ gel (0.5%), respectively. GP IV was left untreated (positive control). The topical treatment started 48 h post-infection, involving twice-daily application (every 12 h) of the formulations. The eyes were examined daily for the severity and symptoms of infection, and evaluated on day 10 as shown in Table 6 by two blinded ophthalmologists.

For sterilization, all the materials used in the preparation of the optimized MOX-UFAs formulation and the MOX-UFAs-laden in situ gel were sterilized by autoclaving at 121 °C for 20 min. The UFAs and in situ gel formulations were then prepared under aseptic conditions in a laminar airflow cabinet to ensure sterility without exposing the final vesicular systems to heat-induced degradation [110].

3.13.3. In Vivo Microbiological Evaluation

An estimate of the CFU of the bacterial load was determined after separating the posterior and the anterior chambers of each cornea. They were then homogenized and suspended in 5 mL of saline, and the number of colonies were counted using the Miles and Misra approach [111]. Briefly, 20 µL of each homogenate sample was transferred to 180 µL of 0.9% saline to start a serial dilution to 10^−5^. Subsequently, 10 µL of each dilution was transferred to six equally divided sectors on nutrient agar plates. The plates were incubated at 37 °C for 24 h after being allowed to dry. The sector with the most well-defined isolated colonies was used for the enumeration. The colonies were enumerated using the following equation:

$$[eqn]$$

3.13.4. Ocular Irritation Potential (Modified Draize Test)

The ocular irritation potential of MOX-UFAs-laden in situ gel was assessed using a modified Draize test in adult albino rabbits weighing between 2 and 3 kg, as outlined by Baeyens et al. [112] and Mishra & Jain [113]. Six New Zealand rabbits were randomly assigned to two groups, each consisting of three animals. The test group received the MOX-UFAs-laden in situ gel, while the control group was treated with normal saline. A single drop (50 µL) was instilled into the right eye of each rabbit. The ocular responses, redness, discharge, and conjunctival swelling (chemosis) were monitored at intervals of 5, 10, 15, and 30 min, and subsequently at 1, 2, 3, 6, 9, 12, and 24 h after administration.

The degree of ocular irritation was assessed using a clinical evaluation scale adapted from Mishra & Jain [113] and is presented in Table 7. The overall ocular irritation index was calculated based on the scores recorded for each parameter at the respective time points. A score of 2 or 3 for any individual parameter, or a total irritation index exceeding 4, was considered indicative of significant irritation.

3.14. Histopathological Study

Following the autopsy, the samples were dried and kept in 10% formalin. The samples were cleaned with xylene and preserved in paraffin before being sliced at 4 mm by a microtome (Leica Microsystems SM2400, Cambridge, UK). The samples were examined under a light microscope after being deparaffinized, stained, and evaluated histopathologically.

4. Conclusions

MOX-loaded UFAs were successfully formulated and optimized using a full factorial design. The optimized formulation demonstrated favorable physicochemical properties, including a small PS, low PDI, acceptable EE, good stability, and sustained drug release. To further enhance the ocular delivery, the optimized MOX-UFAs were incorporated into an in situ gelling system, combining the advantages of eye drops, such as ease of application, dose adjustability, and patient compliance, with those of gels, including prolonged ocular residence time and sustained therapeutic efficacy. Although the MOX-UFAs-laden in situ gel exhibited lower ex vivo permeation parameters compared to the MOX solution, the in vivo studies revealed superior therapeutic efficacy, achieving complete resolution of bacterial keratitis with twice-daily application. The formulation was also well tolerated, showing no signs of ocular irritation in the modified Draize test. The promising results of the UFAs-laden in situ gel formulation suggest strong translational potential for the treatment of bacterial keratitis, offering improved ocular bioavailability, sustained drug release, and enhanced therapeutic efficacy compared to conventional eye drops. The formulation’s components are generally recognized as safe, supporting its potential for clinical application. However, limitations remain, including the need for long-term safety data and comprehensive evaluation in human clinical trials.