Introduction

Depression represents a significant global mental health concern, characterized by profound detrimental effects on both psychological well-being and physiological functioning. Current epidemiological data indicate that this disorder impacts approximately 264 million individuals worldwide and is associated with nearly 60% of global suicide-related mortality.1 First-line pharmacological interventions, including first-generation tricyclic antidepressants (TCAs) and second-generation agents such as selective serotonin reuptake inhibitors (SSRIs), dopamine reuptake inhibitors (DRIs), and norepinephrine reuptake inhibitors (NRIs), remain the cornerstone of clinical management.2 However, these therapeutic approaches are frequently limited by suboptimal efficacy, variable patient tolerance, and adverse side effects, which may compromise treatment adherence and long-term outcomes.3 Consequently, there is an urgent need to advance the development of novel, safer, and more effective pharmacotherapeutic strategies to address the multifaceted challenges of depression management and improve patient quality of life.

Duloxetine (DLX), a second-generation antidepressant classified as a serotonin-norepinephrine reuptake inhibitor (SNRI), is widely prescribed as a first-line pharmacotherapy for major depressive disorder. Its therapeutic mechanism involves dual inhibition of presynaptic serotonin (5-HT) and norepinephrine (NE) transporters, thereby enhancing extracellular concentrations of these monoamines to ameliorate depressive symptomatology.4 Compared to conventional antidepressants, such as tricyclic antidepressants (eg, amitriptyline, clomipramine, doxepin) and selective serotonin reuptake inhibitors (eg, citalopram), DLX demonstrates a distinct pharmacodynamic profile characterized by balanced monoaminergic reuptake inhibition, enhanced therapeutic efficacy, and improved tolerability.3 Clinical evidence highlights its advantages, including a favorable safety profile, accelerated symptom remission, reduced adverse effects (eg, anticholinergic or cardiotoxic reactions), and minimal off-target receptor binding, collectively enhancing treatment adherence and patient outcomes.5 These attributes underscore DLX’s clinical utility in depression management.

DLX, despite exhibiting favorable oral absorption kinetics, demonstrates suboptimal systemic bioavailability (approximately 40%) owing to presystemic degradation in the acidic gastrointestinal environment and extensive first-pass metabolism mediated by hepatic cytochrome P450 1A2 (CYP1A2), reducing its therapeutic effectiveness and leading to variable patient responses.4

Lipid carriers (LCs) have emerged as a promising nanoplatform for enhancing the delivery of lipophilic drugs like DLX. LCs demonstrated significant efficacy in improving drug solubility and bioavailability, offering distinct advantages in stability, controlled release kinetics, and biocompatibility.6 These LC systems are typically formulated using a matrix comprising biocompatible solid lipids classified as generally recognized as safe (GRAS) for oral, topical, and parenteral administration. Liquid lipids are often incorporated into the solid lipid matrix to optimize drug loading and release properties, creating a hybrid architecture combining both phases’ benefits.7

Incorporating natural oils, such as sage oil, into LC formulations (SLCs) may further enhance their therapeutic potential. Sage oil possesses antioxidant and anti-inflammatory properties, which could synergistically augment the antidepressant effects of DLX.8 In acute preclinical studies, linalool, a primary constituent of the essential oil derived from Salvia species (sage), has demonstrated potential antidepressant-like effects.9 Emerging evidence suggests that its pharmacological activity may be mediated by the modulation of monoaminergic pathways, particularly through agonist-like interactions at serotonin 5-HT1A receptors and α2-adrenergic receptors.10 These receptor systems are critically implicated in mood regulation, and their activation aligns with proposed mechanisms underlying linalool’s acute neurobehavioral effects. Such findings highlight its potential as a phytochemical candidate for further investigation in depression-related therapeutics, though rigorous clinical validation remains necessary to elucidate its translational relevance.11

The therapeutic efficacy of antidepressants is contingent upon sustained drug concentration at the central nervous system (CNS) target site, particularly the brain. However, the blood-brain barrier (BBB), a highly selective interface characterized by tight endothelial junctions, efflux transporters (eg, P-glycoprotein), and metabolic enzymes, restricts systemic access to many orally administered antidepressants.12 This limitation necessitates higher doses to achieve therapeutic CNS levels, increasing the risk of systemic adverse effects. Targeted drug delivery strategies that enhance BBB permeability or bypass systemic circulation can elevate cerebrospinal fluid (CSF) drug concentrations, enabling lower therapeutic doses while minimizing off-target toxicity. By optimizing brain-specific delivery, such approaches mitigate dose-dependent side effects and improve treatment adherence, underscoring the importance of advanced CNS-targeted formulations in depression management.13

As a BCS class-II agent characterized by low aqueous solubility and high membrane permeability, DLX presents a compelling candidate for alternative delivery strategies, such as buccal administration. Its physicochemical attributes, including moderate lipophilicity (log P = 4.2), molecular weight of 330 g/mol, and inherent permeability, align with the prerequisites for effective transmucosal drug delivery.14 The buccal route offers potential advantages in bypassing hepatic metabolism and gastric degradation, enhancing efficiency and optimizing therapeutic outcomes. These properties position DLX as a viable candidate for innovative formulation approaches to overcome limitations.

Mucoadhesive buccal films can facilitate the buccal administration of DLX-loaded SLCs. Buccal drug delivery offers several benefits, including bypassing the hepatic first-pass effect, providing a rapid onset of action, and improving patient compliance due to its non-invasive nature.15,16 Chitosan, a natural polysaccharide, is widely recognized for its mucoadhesive properties and biocompatibility, making it an ideal candidate for buccal film formulations.17–19 However, native chitosan’s limited solubility at physiological pH can hinder its effectiveness in buccal applications.20 These constraints have prompted the development of novel chitosan derivatives engineered to enhance solubility across a broader pH range while improving mucoadhesive and permeation properties. Such derivatives are tailored to specific administration routes and dosage forms, with recent advancements focusing on structural modifications of free amine groups, thereby optimizing mucoadhesive performance.21

For instance, chitosan derivatives conjugated with fatty acids, including myristic, capric, azelaic, and stearic acids, have stabilized emulsion systems.22 Studies indicate that increasing the carbon chain length of the fatty acid elevates polymer hydrophobicity, fostering robust interfacial network structures that improve emulsion stability compared to unmodified chitosan.23 Furthermore, while native chitosan salts often exhibit rapid drug release, fatty acid-modified derivatives address this limitation by reducing polymer buccal solubility and delaying erosion, enabling prolonged and controlled drug release.24,25

For the first time, a novel chitosan derivative will be synthesized to enhance stability and mucoadhesive characteristics at buccal pH. One such derivative is chitosan pimelate, synthesized by conjugating chitosan with pimelic acid, a seven-carbon dicarboxylic acid. Pimelic acid’s structure, comprising a hydrophobic hydrocarbon chain and a hydrophilic carboxylic acid group, coupled with its low aqueous solubility, enhances buccal pH stability.26 This modification aims to improve the polymer’s stability in the buccal environment and strengthen its interaction with the mucosal surface, thereby prolonging the residence time of the drug delivery system and enhancing drug absorption.

In this study, we propose developing a novel buccal film system comprising DLX-loaded SLCs incorporated into a chitosan pimelate matrix. This integrated approach seeks to leverage the benefits of SLCs for improved drug encapsulation and release, the therapeutic properties of sage oil, and the enhanced mucoadhesive performance of chitosan pimelate. The formulation aims to provide a sustained release of DLX, improved effectiveness, and enhanced antidepressant efficacy, potentially offering a more effective treatment modality for patients with depression.

By integrating advanced nanotechnology with novel polymer chemistry, this research aims to develop an innovative buccal drug delivery system for DLX, potentially improving therapeutic outcomes for patients suffering from depression.

Materials and Methods

Cutina® HR powder was purchased from BASF, Germany. Chitosan (medium MWT) and pimelic acid were bought from Sigma-Aldrich (MO, USA). Duloxetine was gifted to us by EVA Pharmaceuticals (Cairo, Egypt). Sage oil was purchased from Harraz (Cairo, Egypt). Cremophor® RH 40 was gifted from Amoun Pharmaceuticals (Cairo, Egypt). Tween® 80 and glycerin were gifted from Sigma Quesna (Cairo, Egypt).

Preparation of Duloxetine Sage Lipid Carrier Nanoparticles

Sage lipid carrier (SLCs) nanoparticles were formulated via a hot homogenization and ultrasonication technique (Table 1).27 Cutina HR, sage oil, and Cremophor RH 40 were combined in a 50 mL beaker and heated to 65°C until fully molten. DLX, 10 mg, was dissolved in the lipid phase, followed by adding 20 mL of an aqueous surfactant solution containing 100 mg of Tween 80, preheated to 65°C. The aqueous phase was introduced into the molten lipid under high-speed homogenization (12,000 rpm, 10 min; Homogenizer Model 302, Mechanika Precyzyjna, Warszawa, Poland) to generate an oil-in-water emulsion. The emulsion was cooled to ambient temperature and further processed by probe ultrasonication (5 min, 35% amplitude; Branson Sonifier 250 W/102C, Danbury, CT, USA) to reduce droplet size. The formulation parameters were systematically investigated to optimize nanoparticle characteristics, including Cutina HR: Sage oil mass ratios (80:20, 60:40, and 50:50%) and Cremophor RH 40 surfactant concentrations (2%, 3%, 4%).

Table 1 DLX-SLCs and Their Size, PI, ζ Potential, and EE% Results

Photon Correlation Spectroscopy (PCS) Investigation

Particle size analysis was conducted utilizing photon correlation spectroscopy (PCS) to evaluate the hydrodynamic diameter and size distribution of the nanoparticles. Measurements were performed using a Zetasizer Nano ZS90 (Malvern Instruments, UK), which provided quantitative metrics, including the Z-average (mean hydrodynamic diameter) and polydispersity index (PI), the latter reflecting the homogeneity of the particle population. Samples were diluted in double-distilled water at a concentration compliant with the manufacturer’s specifications to ensure optimal light scattering intensity. Surface charge characterization was concurrently performed via zeta potential analysis using the same instrument, offering insights into the colloidal stability of the formulations.

Encapsulation Efficiency (EE%) Estimation

The encapsulation efficiency (EE%), representing the proportion of DLX effectively retained within SLCs compared to the primary DLX input, was determined via a dialysis-based method. Briefly, 1 mL of the DLX-SLCs was enclosed in a cellulose dialysis tube (12–14 kDa molecular weight cutoff) and submerged in 100 mL of phosphate-buffered saline (PBS, pH 7.4, 37°C), ensuring sink conditions. The assembly was subjected to constant agitation (100 rpm, 4 h) to promote the passive diffusion of unencapsulated DLX into the surrounding medium. The unbound DLX concentration in the dialysate was measured using a UV-Vis spectrophotometer (UV-1601PC, Shimadzu, Japan) at a maximum absorbance wavelength (λmax) of 272 nm using the first derivative technique. Entrapment efficiency was derived using the following equation:

Optimization of Duloxetine Sage Lipid Carrier Nanoparticles

The formulation of DLX-SLCs was optimized through a comprehensive factorial experimental design (32) using Design-Expert® software (v.11, Stat-Ease Inc., USA). Two key independent variables, Cutina HR: Sage oil ratio (X1: 80:20, 60:40, and 50:50%) and Cremophor RH 40 concentration (X2: 2, 3, and 4%) were systematically evaluated for their impact on critical response variables: particle size (Y1, nm), polydispersity index (Y2), ζ potential (Y3, mV), and EE (Y4, %). A total of nine experimental runs, randomized and acted in triplicate, were conducted to mitigate batch-dependent variability. Response surface methodology (RSM) and analysis of variance (ANOVA) were applied to generate quadratic regression models, elucidate factor interactions, and identify optimal formulation parameters. The optimization criteria prioritized minimizing particle size and PI (<0.3) while maximizing EE, and ζ potential magnitude as modulus, ensuring colloidal stability. This data-driven approach facilitated the derivation of robust, statistically validated conditions for synthesizing DLX-SLCs with enhanced physicochemical performance.

In-vitro Release Study

The in vitro release of the optimized DLX-SLCs (F8) and pure DLX was assessed using a dialysis method. Samples containing (2 mg DLX) were sealed in a dialysis membrane (cutoff 12–14 KDa), submerged in vessels charged with 100 mL of phosphate buffer (pH 6.8) and maintained at 37 ± 0.5°C under continuous paddle rotation (50 rpm). Aliquots (1 mL) were periodically withdrawn through a 0.45 μm syringe filter (Merck Millipore, Germany), with immediate replenishment of fresh medium to preserve sink conditions. DLX concentration was quantified spectrophotometrically (UV-160A, Shimadzu, Japan) at λmax = 272 nm using the first derivative technique, with triplicate measurements ensuring methodological reproducibility. Release data were analyzed via DDSolver software, employing nonlinear regression to fit various kinetic models.28  

Surface Morphology of Optimized Duloxetine Sage Lipid Carrier Nanoparticles

The morphological characteristics of the optimized DLX-SLCs (F8) were analyzed using transmission electron microscopy (TEM; JEM-2100, JEOL, Japan). Samples were prepared by depositing a diluted nanoparticle suspension onto a carbon-coated copper grid (300 mesh) via desiccation at ambient conditions. To enhance electron contrast, the grid was negatively stained with 2% (w/v) uranyl acetate solution for 60 seconds, followed by air-drying. TEM micrographs were acquired under high vacuum conditions at an accelerating voltage of 200 kV, enabling visualization of nanoparticle size, shape, and structural homogeneity.29

Synthesis of Chitosan Pimelate (CS-Pim) Polymer

Chitosan pimelate (CS-Pim) polymer was synthesized through a carbodiimide-mediated coupling reaction between chitosan (medium or low molecular weight, MMW or LMW) and pimelic acid at a 6:1 mass ratio (Table 2). Initially, chitosan (CS) was dissolved in 1% (v/v) acetic acid, while pimelic acid was solubilized in ethanol. Ethyl-3-(3-dimethylaminopropyl) carbodiimide (EDC), acting as a crosslinker, was incorporated into the pimelic acid solution at chitosan-to-EDC mass ratios of 1:0.001 and 1:0.1. The mixture was stirred at ambient temperature for 25 min to activate the carboxylic acid moieties of pimelic acid. Subsequently, the activated solution was combined with the chitosan solution and agitated for 24 hours at room temperature to facilitate covalent conjugation. The chemical mechanism of the coupling reaction was illustrated in Supplementary 1. The resultant polymer was isolated by precipitation in 25% (v/v) ammonia solution, followed by centrifugation (5,000 rpm, 5 min) and repeated washing with purified water until a neutral pH (7.4) was attained. The purified CS-Pim was freeze-dried (Martin Christ GmbH, Germany) at −80°C for 84 hours to yield a dry, stable powder.30

Table 2 Composition of Tested Buccal Films

Chitosan Pimelate Characterization via Fourier-Transform Infrared (FT-IR) Spectroscopy

Fourier-transform infrared (FT-IR) spectroscopy was employed to characterize structural distinctions among pimelic acid, CS and synthesized CS-Pim polymer variants. Spectral analyses were conducted using a Bruker ALPHA II spectrometer (Bruker AXS GmbH, Karlsruhe, Germany) with a spectral resolution of 4 cm−1 across the mid-infrared region (4000–400 cm−1). Samples were prepared by homogenously dispersing the polymers in potassium bromide (KBr) pellets at a 1:200 (w/w) polymer-to-KBr ratio. This methodology enabled the identification of functional group alterations and molecular interactions, such as covalent bond formation between chitosan amino groups and pimelic acid carboxyl moieties, across the tested polymer derivatives.

Formulation of the Chitosan Pimelate Buccal Films

Buccal film preparations were developed using a solvent casting technique, employing different chitosan pimelate (CS-Pim) derivatives (Table 2) as bioadhesive matrices and glycerin as a plasticizing agent. For preliminary optimization, blank films (without DLX) were prepared by homogenously dispersing CS-Pim (3% w/w) in purified water under continuous stirring for 12 h at ambient temperature, followed by the incorporation of glycerin (3% w/w). The resultant polymeric dispersion was subjected to ultrasonication to eliminate entrapped air, cast into a 6 cm diameter plastic Petri dish, and thermally dried in an oven (40°C, 24 h), then at room temperature for 3 days. Post-drying, the patches were delicately peeled and encapsulated in a sealed jar to prevent moisture absorption.

Drug-loaded buccal films were fabricated by incorporating DLX-SLCs into chitosan pimelate (CS-Pim) matrices selected for their optimal buccal delivery performance, as determined by prior physicochemical characterization studies. The CS-Pim polymer, identified as the ideal formulation through rigorous evaluation parameters, was combined with DLX-SLCs at a targeted loading density of 6 mg per cm2 of patch surface area. The mixture was homogenized under continuous magnetic stirring (12 h) to ensure uniform dispersion of DLX-SLCs within the polymeric matrix. Subsequent steps, including plasticizer addition, solvent casting, controlled drying, and precise sectioning into 1 cm2 units, were executed per the standardized protocol established for blank film preparation.

Characterization of the Buccal Film Formulations Physicochemical Properties Assessment of the Buccal Films

The physicochemical properties of buccal films (1 cm2) were systematically evaluated. Film mass was determined using an analytical balance, with mean weight and standard deviation calculated from five replicates. Thickness uniformity was assessed via a digital micrometer (Guanglu, China) across five randomly selected films. For pH analysis, films were equilibrated in 5 mL PBS (pH 6.8) for 2 hours in sealed Petri dishes to prevent atmospheric interference, followed by pH measurement using a calibrated meter (Jenway 3510, Staffordshire, UK). Drug-containing films underwent identical pH testing to evaluate the influence of DLX on formulation acidity. Moisture loss was quantified by storing pre-weighed films (M1) in desiccators containing anhydrous calcium chloride (3 days), with post-desiccation mass (M2) used to calculate percentage moisture loss via the following Eq.:

Mechanical properties were characterized using a Dynamic-Mechanical-Analysis (DMA) (DMA Q800 V21.1 Build 51, TA instruments, UK). Film strips were clamped and elongated at a constant rate of 0.1000 N/min to 18.0000 N until fracture. Tensile strength and elongation at break were derived from force-extension curves, with triplicate measurements ensuring statistical reliability.

The tensile strength (σ) and elongation at break (ε) % of the films were calculated using the following Eq:

Where F is the force at failure (N), and A is the cross-sectional area of the film (cm2).

where Linitial is the initial length (mm), and Lbreak is the extended length at fracture (mm).

Drug Content Consistency Quantification

Drug-loaded films were fabricated by incorporating DLX at a standardized concentration of 6 mg/cm2. For quantification, individual films (1 cm2 surface area) were fully dissolved in 500 mL of phosphate-buffered saline (PBS, pH 6.8) under sonication (30 min). Post-dissolution, 1 mL aliquots were withdrawn, filtered through a 0.2 μm syringe filter (Nylon syringe filter, PRC) to remove particulate matter, and subjected to spectrophotometric analysis. DLX content was quantified using a UV-Vis spectrophotometer (UV-1601PC, Shimadzu, Japan) operating in first derivative mode, with absorbance measured at the λmax of 272 nm to enhance selectivity and minimize matrix interference.

Swelling Studies

This method provided insights into the fluid uptake and matrix expansion, critical for evaluating the film’s ability to adhere to and hydrate the buccal mucosa. The swelling behavior of buccal films was quantified gravimetrically to assess their hydration capacity and mucoadhesive potential. Pre-weighed films (M1) were immersed in 5 mL of PBS (pH 6.8) under controlled conditions (37 ± 1°C) for 2 h. At predetermined intervals, films were carefully removed, superficially blotted with filter paper to eliminate unabsorbed surface moisture, and reweighed (M2). The experiment was conducted in triplicate to ensure reproducibility. The swelling index (%) was calculated using the following equation:

Ex-vivo Mucoadhesive Strength Evaluation

The ex vivo mucoadhesive strength of the buccal films was evaluated by a quantitative approach providing a direct measure of bioadhesion, reflecting the interfacial binding capacity between the film and mucosal membrane under simulated in vivo conditions, utilizing an adapted balance method. Freshly dissected rabbit buccal mucosal tissue was equilibrated with PBS (pH 6.8) to simulate physiological hydration. A film was affixed to the mucosal surface under constant pressure for 5 min to establish adhesive contact. Incremental weights of distilled water were then added until the patch detached from the mucosal substrate. The total mass of water (g) required to induce detachment was recorded.29

Surface Morphology of the Buccal Films

The surface morphology of buccal films was characterized using scanning electron microscopy (SEM). Beforehand imaging, samples were mounted on copper stubs and sputter-coated with a gold layer to enhance surface conductivity and mitigate electron charging artifacts. High-resolution micrographs were acquired using an SEM system (Oxford Instruments, UK) operated at an accelerating voltage of 20 kV in secondary electron detection mode.

Contact Angle Measurement of the Buccal Films

Contact angle measurements were conducted to evaluate the wettability of buccal films, utilizing Optical tensiometers (Theta flow, Biolin Scientific, UK). Films were securely affixed to a stage, and ultrapure water was vertically dispensed from a definite height onto the film surface at 25 ± 1°C. To account for surface heterogeneity, seven high-resolution images per droplet were captured 10 seconds post-dispensing under standard illumination, focal planes, and shading conditions. ImageJ software (v.1.53, NIH, USA) was employed to analyze droplet morphology, while contact angles were calculated via the sessile drop method. Triplicate measurements were performed across distinct film regions to ensure statistical robustness, and final values were derived from averaged data.31

Differential Scanning Calorimetry (DSC) Investigation

The thermal performance and stability of DLX, excipients (Cutina HR, sage oil, Cremophor RH 40, Tween 80), CS-Pim, and lipid-based formulations (plain SLCs, DLX-SLCs, and selected DLX-SLCs buccal film) were investigated via differential scanning calorimetry (DSC Q200, TA Instruments, USA). Before the assessment, the instrument was calibrated for temperature and enthalpy using high-purity indium (melting point: 156.6°C) and zinc (melting point: 419.5°C) standards. Samples (2–5 mg) were hermetically sealed in aluminum crucibles, with an empty crucible serving as a reference. Thermograms were recorded under a dynamic nitrogen atmosphere (50 mL/min) across a temperature range of 25–300°C, employing a heating rate of 10°C/min. This protocol enabled the identification of phase transitions, including melting, crystallization, and glass transitions, as well as incompatibilities between formulation components.32

Fourier-Transform Infrared (FTIR) Spectroscopy Inspection

FTIR spectroscopy was employed to investigate molecular interactions and compatibility among formulation components, including pure DLX, Cutina HR, sage oil, Cremophor RH 40, Tween 80, CS-Pim, plain SLCs, DLX-SLCs, and the selected DLX-SLCs buccal film. Spectra were acquired using a Bruker Alpha II FTIR spectrophotometer (Ettlingen, Germany) equipped with a diamond attenuated total reflectance (ATR) accessory. Before analysis, samples were uniformly compressed under hydraulic compression to ensure optimal contact with the ATR crystal. Scans were conducted across the mid-infrared region (4000–400 cm−1) at a resolution of 4 cm−1, with 32 co-added scans per sample to enhance signal-to-noise ratio.33,34

Ex-vivo Skin Permeation Studies of the Optimized Buccal Film

Ex vivo skin permeation studies were conducted using full-thickness abdominal skin harvested from male Wistar rats (200 ± 20 g) that were humanely euthanized using ketamine 80 mg/kg and xylazine 10 mg/kg intraperitoneally. The skin was surgically excised, treated with 0.3 N ammonium hydroxide solution to depilate hair and remove subcutaneous adipose tissue, and rinsed thoroughly with saline to eliminate residual alkali. Tissue integrity was verified visually, and samples with uniform thickness were selected. Afore experimentation, the skin was equilibrated in phosphate buffer (pH 7.4, 1 h) and air-dried. A 1 cm2 section of the optimized DLX film was applied to the epidermal surface under gentle pressure and mounted in a Franz diffusion cell, with the stratum corneum facing the donor chamber and the dermal layer contacting the receptor compartment (300 mL PBS, pH 6.8, maintained at 37 ± 0.5°C under sink conditions). Aliquots (1 mL) were periodically withdrawn from the receptor medium, replaced with fresh PBS, and analyzed via UV spectrophotometry (λmax = 272 nm). Steady-state flux (J, µg/cm2/h) was derived from the slope of the linear region of cumulative drug permeation (Q) versus time (t) profiles, providing quantitative insights into transdermal delivery kinetics.

Acute Toxicity Test

The OECD Guideline No. 423 was followed in toxicity studies.35 Rats were given oral doses of 500, 1000, and 2000 mg/kg body weight of DLX-SLCs buccal film after being split into groups of three unisexual animals each. Following 48 hours of close monitoring, the animals were assessed for behavioural abnormalities such as paw licking, writhing, exhaustion, and decreased hunger, as well as any indications of fatality. Additionally, observations were made to verify regular activity and ensure there were no negative impacts on the animals’ overall health.

In-vivo Studies Using Lipopolysaccharide (LPS)-Induced Depression Rat Model

Wistar rats (180–220 g) from the Sultan Qaboos University animal house were kept in stainless steel cages with a 12:12-hour light-dark cycle and controlled temperatures (22 ± 1°C). They were given unlimited access to water and food pellets. The recommendations of the National Institutes of Health Handbook for the Care and Use of Laboratory Animals were followed when conducting the research. The Sultan Qaboos University Standing Ethics Committee for Animal Use in Research gave the study ethical approval (Approval Code: SQU/EC-AUR/2024-2025/5).

The assessment of the enhanced antidepressant action of the drug and optimized formulation was achieved using a lipopolysaccharide (LPS)-induced rat model of depression-like behaviour. Five experimental groups, I, II, III, IV, and V (n = 6), were used in the study: negative control, positive control (LPS), Pure DLX, marketed DLX, and DLX-SLCs buccal film. The negative control was given saline orally and then IP saline after 30 min, while the positive control was given oral saline and then, after 30 min., administered with LPS intraperitoneally (LPS; Sigma-Aldrich, St. Louis, MO, USA) at a dose of 0.1 mg/kg once daily for 14 days.36 Group III was given pure-DLX, 30mg/kg, after 30 min. of IP injection of LPS. Group IV was given Marketed-DLX, 30mg/kg, after 30 min. of IP injection of LPS. Then Group V was given DLX-SLCs Buccal Film 30 mg/kg after 30 min. of IP injection of LPS.

The tail suspension test (TST) and elevated plus maze (EPM) were used for behavioural evaluations. Ketamine 80 mg/kg and xylazine 10 mg/kg (given intraperitoneally) were used to establish profound anaesthesia in the rats before they were slaughtered. After being carefully removed, the brains were separated into two parts and cleaned with regular saline solution. While the first portion was maintained in 10% neutral buffered formalin for histopathological analysis, the second piece was snap-frozen in liquid nitrogen and kept at −80°C for a subsequent biochemical analysis.

Behavioural Analysis

To evaluate DLX-SLCs buccal film’s antidepressant efficacy, behavioural despair models such as the Elevated Plus maze (EPM) and Tail suspension test (TST) were employed.

EPM consists of 4 arms in a cross shape with a central zone in the middle, placed approximately 45 cm above the ground. Two opposing standing arms have walls that are open at the top and do not interfere with the central zone. The test usually takes 10 min., enough to start the habituation process.37 Rodents often evade open and brightly lit areas, although concurrently, they have a propensity to investigate novel environments. Consequently, the ratio of these conflicting stimuli was assessed.38 The frequency of entrances into the open and closed arms and the central zone, and the total duration spent in these areas, were documented. Additional assessed indicators comprise raising, sniffing, grooming, and defecating. Prolonged duration in open arms signifies a diminished level of “anxiety” in the animal.36

The TST elicits behavior analogous to that observed in the Porsolt test. The advantage of this test against the Porsolt test was to eliminate the risk of hypothermia caused by water, as well as the possibility of assessing the strength and energy of the movement of the animal.39 TST is primarily utilized in rodents, where the rat is suspended by its tail, with its body dangling in the air. The examination lasts around 6 min. and may be administered multiple times.36 TST posits that the animal will attempt to evade stressful circumstances. After a while, the animal stops its struggle, resulting in immobility; prolonged stages of immobility indicate depressive behavior. The immobility phase is shortened following the introduction of antidepressants. Various strains of rats have distinct reactions to specific categories of antidepressants.

Sucrose Preference Test (SPT)

The SPT assessed hedonic response by providing animals with concurrent access to two bottles: one containing a 1% sucrose solution (1% w/v) and the other containing plain tap water. The percentage of sucrose preference, an indicator of anhedonia, was derived from sucrose solution intake and expressed as a proportion of total liquid consumption recorded over the last four days of the experimental period.

Measurement of ACTH, TNF-α, IL-1β, GABA, and Cortisol

Following careful collection into Vacutainer® Tubes containing EDTA, the blood extracted from the tail vein was centrifuged for 15 min at 4°C at 6000 rpm. Before analysis, the plasma was separated and kept at −80°C. Following the manufacturer’s instructions, an ELISA test kit (Elabscience, Houston, TX, USA, Cat. No. E-EL-0160, E-EL-H0109, Cat. No. E-EL-H0149, and E-BC-K852-M, respectively) was used to measure the levels of ACTH, tumor necrosis factor-alpha (TNF-α), interleukin-1 beta (IL-1β), and Gamma-aminobutyric acid (GABA). Triplicate assays were made, results were represented as pg/mL, and the mean value for each sample was determined. Similarly, the cortisol ELISA kit (UNEB0007) was used to measure cortisol levels.

Measurement of Serotonin

Serum samples were obtained by centrifuging blood at 3000 × g for 10 min. and preserved at −20°C until analysis. Serotonin levels were quantified utilizing a competitive ELISA kit (Abcam, Cambridge, UK, Cat. No. ab133053) in accordance with the manufacturer’s guidelines. Samples and standards were introduced in duplicate to antibody-coated microplate wells, treated with alkaline phosphatase-conjugated serotonin antigen and anti-serotonin antibody, and subsequently developed using p-nitrophenyl phosphate substrate. Absorbance was measured at 405 nm, and concentrations were ascertained using a standard curve.

Measurement of Antioxidant Enzymes

Superoxide dismutase (SOD) and malondialdehyde (MDA) kits from My BioSource, Inc. (San Diego, CA, USA) were measured by the colorimetric method, using the Bio Vision kit (Milpitas, CA, USA).

Histopathological Investigation

Following the rat’s brain tissue samples’ extraction, they were preserved in 10% neutral buffered formalin, gradually dried using a range of ethanol concentrations, cleaned in xylene, and then embedded in paraffin wax. After that, the tissue was divided into sections that were 5 μm thick and stained with hematoxylin and eosin. A light microscope was used to look for histopathological alterations in the stained sections.

Statistical Analysis

Minitab and Design-Expert tools were used for optimization analysis. Optimization analysis was conducted using one-way ANOVA at a significance threshold of p < 0.05. One-way ANOVA and Tukey’s post hoc test were used for in vivo statistical analysis (GraphPad Software 8, Inc., San Diego, CA, USA). To quantify the precision of the findings, a 95% CI was employed. Using G-Power software version 3.1.9.4 (Fraz Faul, Germany), the sample size was computed to find the smallest number needed to test the study hypothesis.

Results and Discussion

Particle dimension is a vital quality parameter in nanopharmaceutical development, directly affecting formulation performance by altering pharmaceutical release, cellular uptake efficiency, dispersion stability, and medicinal bioactivity. Empirical evidence demonstrates that below 300 nm particle dimensions possess increased efficiency and greater cellular internalization due to improved mucosal penetration and decreased macrophage clearance, as supported by previous investigations.40 The present study revealed that DLX-SLCs had a monomodal size distribution between 56.8 ± 2.9 nm and 207.1 ± 5.8 nm, as determined by DLS. A multivariate regression study revealed that the two independent variables (concentration of Cremophore RH 40 surfactant (X1, % w/v) and Cutina HR: Sage oil ratio (X2)) have significantly influenced (p < 0.05) particle diameter (Table 1). The interactive effects of these variables on particle size were further elucidated through three-dimensional response surface and contour plots (Figure 1A and B). Maximum particle diameters were observed at suboptimal surfactant concentrations (2% w/v) and a lipid phase ratio of 80:20 (Cutina HR: Sage oil), whereas incremental increases in Cremophore RH 40 to 4% w/v correlated with a pronounced reduction in hydrodynamic size. This inverse relationship aligns with the surfactant’s role in lowering interfacial energy between the lipid and aqueous phases, enhancing emulsification efficiency during high-shear homogenization. At elevated concentrations, Cremophore RH 40 facilitates the formation of a stabilized monolayer at the lipid-water interface, promoting finer droplet subdivision and preventing coalescence.41 Conversely, insufficient surfactant availability results in incomplete surface coverage, favoring particle aggregation and larger colloidal assemblies. Thus, optimum surfactant-to-lipid ratios can achieve nanoscale particle dimensions, a prerequisite for enhanced mucosal permeation and controlled drug release.

Figure 1 3D response surface analysis and Contour plots depicting the influence of Cutina HR: Sage oil ratio and surfactant (Cremophore RH 40) concentration on critical quality attributes of DLX-SLCs: (A and B) particle size (PS), (C and D) PI, (E and F) zeta potential (ZP), and (G and H) EE%.

The ratio of solid lipid (Cutina HR) to liquid lipid (sage oil) exerted a statistically significant influence (p < 0.05) on the particle size of DLX-SLCs. Elevated proportions of Cutina HR correlated with increased particle diameters, attributed to the rigid crystalline structure of Cutina HR solid lipids, which promotes coalescence during the cooling and solidification stages of nanoparticle synthesis. Empirical evidence indicates that formulations with higher solid lipid content (eg, 80:20 solid-to-liquid lipid ratios) yield particles in the range of 150–281 nm, determined by the precise ratio employed.42

Conversely, incremental incorporation of sage oil (liquid lipid) induced a marked reduction in particle size, mediated by its capacity to modulate the lipid matrix’s thermodynamic behavior. Liquid lipids lower the melting point and enhance the fluidity of the lipid phase, enabling finer droplet subdivision during high-energy homogenization. This improved emulsification facilitates the formation of smaller monodisperse colloidal structures, which are stabilized during solidification.42

PI is an important measure for evaluating the assembly and stability of NPs. It assesses the size diversity; a low PI shows a consistent size, whereas a high PI denotes size diversity. A PI value of below 0.3 is excellent because it implies homogenous particle sizes, which promote stability, eliminate aggregation, and assure reliable quality and efficacy of the NPs. The PI of the DLX-SLCs (Table 1) ranged between 0.142 and 0.461. PI exhibited no statistically significant correlation (p > 0.05) with the independent variables evaluated in the experimental design (Figure 1C and D). However, elevated liquid lipid (sage oil) content relative to the total lipid phase, combined with an optimized surfactant (Cremophore RH 40) concentration, significantly enhanced particle size uniformity, as evidenced by reduced PI values. This phenomenon is attributed to the liquid lipid’s capacity to lower interfacial tension and improve emulsification efficiency, fostering monodisperse droplet formation during homogenization.43 Concurrently, surfactant concentrations near CMC stabilized the lipid-aqueous interface, minimizing coalescence. These findings align with prior studies demonstrating that liquid lipid incorporation and surfactant optimization synergistically enhance colloidal homogeneity.27

The ζ potential of NPs is an additional significant measure of their stability, as it represents the possibility of clustering or distribution. It offers information about the surface potential and any changes performed to the NPs. Furthermore, it influences the intake of drugs by cells and plays an important role in drug delivery. The ζ potential of DLX-SLCs (Table 1) varied from −15.3 ±2.6 to −31.8 ±1.2 mV. The ζ potential of SLCs was significantly modulated by both independent variables, as depicted in the contour plots and response surface analysis (Figure 1E and F). Incremental increases in Cremophore RH 40 concentration (2% to 4% w/v) induced a marked reduction in zeta potential magnitude, attributable to the nonionic surfactant’s capacity to neutralize surface charges via steric stabilization. Unlike ionic surfactants, Cremophore RH 40 lacks charged functional groups, instead forming a hydrated polymeric layer around particles that mitigates aggregation through spatial hindrance rather than electrostatic repulsion. At elevated concentrations, the dominance of steric stabilization over electrostatic interactions alters the colloidal equilibrium, potentially compromising long-term dispersion stability despite reduced surface charge.44 The zeta potential of SLCs exhibited a direct relationship with the proportion of sage oil incorporated into the lipid matrix. This trend is likely mediated by ionizable acidic moieties within sage oil, which confer an enhanced negative surface charge to the nanoparticles. Comparative studies on lipid-based systems, including those utilizing black seed oil and linseed oil, confirm this phenomenon, demonstrating that liquid lipids with polar functional groups generate pronounced negative zeta potentials (typically ranging from −30 to −50 mV). Such elevated surface charges amplify electrostatic repulsion between particles, thereby enhancing colloidal stability by mitigating aggregation.45,46 In contrast, formulations with higher solid lipid content displayed attenuated negative zeta potentials, indicative of reduced electrostatic stabilization. This decline arises from the nonpolar nature of solid lipids, which limits the availability of ionizable groups to contribute to surface charge. Consequently, SLCs dominated by solid lipids exhibit diminished stability compared to their liquid lipid-enriched counterparts, underscoring the critical role of lipid hydrophilicity in modulating interfacial charge dynamics and colloidal behavior.47

EE% serves as a critical determinant in the design of lipid-based nanocarriers, directly governing the therapeutic dose delivered by SLCs and ensuring clinical efficacy. The encapsulation efficiency of DLX within the SLCs is summarized in Table 1, demonstrating high values across all tested formulations (73.8% ±4.1 to 79.9% ±3.8). To optimize EE and mitigate premature release, the effects of two formulation variables, solid lipid-liquid lipid (Cutina HR: Sage oil) ratio and surfactant (Cremophore RH 40) concentration, on EE% were systematically investigated. The EE% of DLX was predominantly governed by both variables, with Cutina HR: Sage oil ratio exerting a more pronounced influence than Cremophore RH 40 concentration (Figure 1G and H). Elevated lipid content enhanced EE% due to DLX’s higher partition coefficient in the lipid matrix, favoring thermodynamic affinity for the hydrophobic core over the aqueous phase. A significant dependence on the amount of sage oil was observed; formulations with reduced sage oil content exhibited significantly lower EE% than those enriched with liquid lipid. This trend underscores the pivotal role of sage oil in enhancing DLX entrapment, attributable to the drug’s higher solubility in the liquid lipid phase relative to the solid matrix. Furthermore, formulations with elevated liquid lipid proportions yielded smaller nanoparticle diameters, as previously reported, which amplifies the interfacial surface area available for DLX partitioning during nanocarrier assembly. The synergistic interplay between lipid solubility and reduced particle size optimizes drug incorporation, aligning with established principles of lipid nanoparticle design where liquid lipids enhance drug loading and colloidal stability.48 Conversely, a nonlinear relationship was observed between Cremophore RH 40 concentration and DLX encapsulation efficiency in the SLCs (Table 1). Incremental surfactant incorporation (from 2 to 3% w/v) enhanced EE%, likely due to improved interfacial stabilization and DLX partitioning into the lipid phase during nanoparticle assembly.47 However, a paradoxical reduction in EE% occurred at the highest surfactant concentration (4% w/v), deviating from the anticipated trend. This attenuation may arise from micellar solubilization of liquid lipids by excess surfactant, which exceeds the critical micelle concentration (CMC). Such micellar entrapment reduces the available lipid phase for drug incorporation, diminishing overall EE%.49 These findings revealed the surfactant’s dual role as a stabilizer and a potential competitor in drug-loading dynamics, underscoring the necessity of optimizing surfactant-to-lipid ratios to balance emulsification efficacy and entrapment capacity.41,50,51 It was proposed that 3% of Cremophore RH 40 is the optimal level of surfactant.

The desirability aspect is frequently employed in optimizing with multiple goals to determine the ideal formulation for various processes. The system consolidates several response factors into a singular combined desirability score, facilitating the simultaneous optimization of all aspects. It measures the appealing qualities of the preparations for every response factor. The adoption of the F8 formulation was determined by its optimal desirability factor of 0.821, warranting further assessment.

The in vitro release profiles of DLX from pure-DLX and optimized DLX-SLCs in phosphate-buffered saline (PBS, pH 6.8, 37°C) are presented in Figure 2A. Pure DLX exhibited rapid release, with 59.46% of the drug liberated within 1 h and near-complete release achieved by 24 h. This rapid dissolution of pure-DLX in PBS, pH 6.8, is attributable to its pH-dependent solubility profile. DLX, a weakly basic compound (pKa ≈ 9.7), undergoes amine protonation in neutral-to-acidic media, significantly enhancing its aqueous solubility at pH 6.8.52 Empirical studies confirm near-complete solubility (>99%) in pH 6.8 PBS due to a favorable ionization equilibrium.7 This high solubility accelerates drug dissolution and diffusion, explaining the observed burst release (59.46% within 1 h). In contrast, optimized DLX-SLCs demonstrated sustained release behavior with only 10.3% of DLX released within the initial hour, followed by prolonged, near-linear release reaching 61.8% at 24 h. The diminished burst release (≤15%) from SLCs confirms homogeneous dispersion of DLX within the lipid matrix, minimizing surface-associated drug fractions. This modulated release profile aligns with the lipid matrix-controlled liberation behavior, which is characteristic of lipid-based nanocarriers, where drug partitioning into the lipid core retards aqueous dissolution. A similar result was reported.7

Figure 2 (A) In vitro DLX release profile from Pure-DLX and optimized DLX-SLCs and (B) TEM representations of optimized DLX-SLCs (scale bar = 500 nm) (i) and a histogram representing the presumed average hydrodynamic particle sizes derived from TEM (ii).

The drug release mechanisms for pure-DLX and optimized DLX-SLCs were elucidated through kinetic modeling of dissolution data. Release profiles were fitted to established mathematical models, such as zero-order, first-order, Higuchi diffusion, and Korsmeyer-Peppas, using DDsolver software (Table 3). This quantitative approach identifies the governing release kinetics by determining the model that best describes the drug liberation process. Such mathematical formalization provides critical insights into formulation performance while optimizing resource utilization: it discriminates between diffusion-controlled, erosion-mediated, and anomalous transport mechanisms.

Table 3 Release Kinetics of DLX from Pure-DLX and Optimized DLX-SLCs

The drug release mechanism from DLX-SLCs was diffusion-dominated, as evidenced by exceptional fit to the Korsmeyer-Peppas model (R2 = 0.997), with a release exponent (n = 0.588) indicating anomalous (non-Fickian) transport, proposing an extended drug release mechanism that encompasses many mechanisms, including diffusion, swelling, and erosion. The documented studies imply that n < 0.43 signifies diffusion-dependent liberation from nanosystems, but n > 0.43 suggests anomalous techniques with an escalating non-Fickian involvement as the n number rises. Our findings indicate that DLX-SLCs adhere to anomalous mechanisms characterized by a significant non-Fickian (case II transport) involvement. The observed behavior aligns with lipid-based carriers undergoing structural reorganization during dissolution, where interfacial hydration triggers progressive softening and erosion of the lipid matrix, thereby augmenting drug diffusion through aqueous pathways. Erosion kinetics modeling revealed that drug release from DLX-SLCs follows homogeneous, diffusion-controlled erosion mechanisms, as demonstrated by the superior fit to the Baker-Lonsdale model (R2 = 0.955) compared to the Hopfenberg model.17 This high correlation signifies uniform bulk erosion from spherical lipid matrices, consistent with the monodisperse morphology. The Baker-Lonsdale fit further confirms three governing mechanisms, constant diffusivity due to homogeneous drug distribution within the lipid core, stable spherical geometry during dissolution, and boundary layer-controlled aqueous penetration enabling gradual matrix erosion.27,53

Transmission electron microscopy (TEM) analysis of the optimized DLX-SLCs revealed the formation of monodisperse spherical nanoparticles (Figure 2Bi), as evidenced by the corresponding hydrodynamic size distribution histogram (Figure 2Bii). Particle dimensions obtained via TEM corroborated dynamic light scattering (DLS) measurements, confirming the precision of size characterization across orthogonal analytical techniques. Notably, the absence of particle aggregation or coalescence in TEM micrographs underscores the colloidal stability of the optimized DLX-SLCs, consistent with the low polydispersity index (PI < 0.3) derived from DLS analysis. The congruence between TEM and DLS data validates the robustness of the formulation process in achieving uniform nanoparticle morphology.40

Comparative FT-IR analysis of pure chitosan (CS), pure pimelic acid (Pim), and the CS-Pim conjugate polymer confirmed successful covalent modification through distinct spectral alterations (Figure 3A). Pure-Pim Figure 3Ai showed diagnostic peaks at 1681 cm−1 (carboxylic C=O stretch) and 1266 cm−1 (C-O stretch),54,55 while pure-CS (LMW and MMW) (Figure 3Aii and iii, respectively) exhibited characteristic bands at 3332 cm−1 (O-H/N-H stretch), 1652 cm−1 (amide I), and 1587 cm−1 (amide II).56 The different CS-Pim spectra (Figure 3Aiv, v, vi and vii) demonstrated critical changes including the disappearance of Pim’s carboxylic C=O peak (1681 cm−1), emergence of a new amide I band at 1644 cm−1, shift in amide II to 1550 cm−1, broadened N-H/O-H stretch (3309–3285 cm−1), collectively evidencing amide bond formation between CS amine groups and Pim carboxyl moieties. Preservation of saccharide backbone vibrations (1024–1149 cm−1) confirmed structural integrity post-modification. These spectral transitions, particularly the carbonyl frequency reduction, verify covalent conjugation while excluding physical mixture artifacts, establishing a robust foundation for functional polymer design. A similar result was reported for different CS derivatives using other dicarboxylic acids.20

Figure 3 (A) FTIR spectra of (i) Pimelic acid, (ii) LMWT CS, (iii) MMWT CS, (iv) CS-Pim (B1), (v) CS-Pim (B2), (vi) CS-Pim (B3), and (vii) CS-Pim; (B) Swelling percent of buccal mucoadhesive film formulations and (C) Surface morphology of buccal mucoadhesive film formulations. (i) B1, (ii) B2, (iii) B3, (iv) B4.

The films were assessed for various parameters, including thickness, weight uniformity, pH, moisture loss, tensile strength, and elongation at break, to determine their appropriateness for buccal delivery. Buccal films must be sufficiently thin to avoid discomfort, as they will reside in the oral cavity for a designated duration. The film thicknesses were measured between 0.04 and 0.08 mm using an electronic micrometer. The weights of the films ranged from 11.1 to 28.7 mg. In accordance with reports, the augmented polymer molecular weight enabled the film to retain more water during the drying process.57

The weight and thickness of the films (Table 4) demonstrated concentration-dependent responses to EDC crosslinking, where higher EDC levels (0.1 M) typically increased both parameters due to enhanced polymer network density with more pimelic acid and restricted chain contraction during drying. Thus, formulations B2 and B4 (0.001 M EDC) exhibited reduced thickness (0.04 mm) compared to B1/B3 (0.1 M EDC) despite equivalent polymer concentration. This anomaly arises from EDC-mediated amide bond formation with the carboxylic group of pimelic acid, enabling denser polymer networks. Consequently, targeted low-level EDC crosslinking (0.001 M) achieves patient-preferred thinness (<0.05 mm) while preserving free amine groups for mucoadhesion.23

Table 4 Physicochemical Properties, Drug Content, and Mucoadhesive Strength of Buccal Films

The surface pH of formulated buccal films was quantified to evaluate mucosal irritation potential, yielding values between 6.83 ± 0.04 and 7.08 ± 0.03 across all formulations (Table 4), with no statistically significant inter-group differences. This near-neutral range demonstrates physiological compatibility with the buccal mucosal environment (typical pH 6.3–7.4), thereby minimizing risks of epithelial irritation, salivary buffering disruption, or mucin denaturation.58

The moisture loss of buccal films ranged from 13.19 ± 0.69% to 16.38 ± 0.59%, demonstrating a positive correlation with increasing polymer molecular weight (MW). MMW-CS exhibited greater moisture loss, attributed to greater free hydroxyl/amine groups per chain, increasing water capacity within the polymer, amplifying percentage loss due to higher initial hydration.59

Dynamic mechanical analysis (DMA) for tensile strength (Table 4) revealed that films incorporating MMW-CS exhibited significantly enhanced viscoelastic properties compared to LMW-CS at identical polymer concentrations (3% w/v). Specifically, MMW-CS film B4 demonstrated a higher tensile strength (10.07±0.34 N/cm2 vs 3.08±0.24 N/cm2 for LMW-CS (B2)) and greater elongation at break (109.9±7.3% vs 55.22±4.6%), attributable to the entanglement network density of MMW-CS chains forming more topological entanglements, increasing resistance to deformation.60

The DLX content consistency across buccal film formulations complied with pharmacopeial standards, as evidenced by analysis of 10 randomly selected films exhibiting DLX content within 85–115% of the labeled claim (mean: 98.19% ± 3.2% RSD). This narrow variability (≤6% RSD) satisfies regulatory requirements for dosage unit uniformity, ensuring consistent therapeutic dosing and batch-