Alginate/pectin dressing with niosomal mangosteen extract for enhanced wound healing: evaluating skin irritation by structure-activity relationship

Abstract

Most modern wound dressings assist the wound-healing process. In contrast, conventional wound dressings have limited antibacterial activity and promote sporadic fibroblast growth. Therefore, wound dressings with prolonged substance release must be improved. This research aimed to develop hydrogel films. These were synthesized from alginate and pectin, incorporated with mangosteen extract (ME), and encapsulated in niosomes (ME-loaded niosomes). Subsequently, we examined the in vitro release and physical characteristics of ME-loaded niosomes. These characteristics included particle pH, size, charge, polydispersity index (PDI), and drug loading properties. These properties included drug loading content (DLC), entrapment efficiency (EE), and yield (Y). Additionally, we examined the swelling ratio and biological characteristics of the hydrogel film. These characteristics included antibacterial activity, cytotoxicity (L929), cell attachment to the tested materials, cell migration, hemocompatibility, and in vivo irritation. Significant results were obtained using a 2:1 niosome preparation containing Span60 and cholesterol. Ratio influenced size, charge, PDI, DLC, EE, and Y. The results were 225.5 ± 5.83 nm, negatively charged, 0.38, 16.2 ± 0.87%, 64.8 ± 3.49%, and 87.3 ± 3.09%, respectively. Additionally, the release of encapsulated ME was pH sensitive because 85% of the ME can be released at a pH of 5.5 within seven days and decrease to 70% at a pH of 7.4. The maximum swelling ratios of patches with 0.5% and 1% Ca²⁺ crosslinking were 867 wt% and 1,025 wt%, respectively, after 30 min. These results suggested that a medium dose (15 mg) of niosomal ME incorporated in a hydrogel film provided better bacterial inhibition, cell migration, and cell adhesion in an in vitro model. Additionally, no toxicity was observed in the fibroblasts and red blood cells. Therefore, given the above-mentioned advantages, this product can be a promising candidate for wound dressing applications.

Hydrogel; Niosome; Mangosteen extract; Skin irritation.

1. Introduction

The skin is the largest organ in the body. It accounts for 15–20% of the whole-body mass. The skin tissue comprises three main layers: the epidermis, dermis, and hypodermis [1]. The dermal extracellular matrix (ECM) is an extensive molecular network comprising collagen, elastin, reticular fibers, and the dermis layer. Fibroblasts are primarily responsible for the production of collagen, proteoglycans, and glycosaminoglycans [2]. Skin defects, also known as wounds, are usually caused by thermal, physical, and chemical damage or physiological conditions [3]. Wound healing is a complex process chronologically comprising hemostasis, inflammation, proliferation, maturation, and differentiation. The prerequisites for an effective wound healing are a moist environment surrounding the wound [4], adequate oxygen circulation to promote cell regeneration [5] and absence of bacterial infection [6]. Products that accelerate wound healing have been widely used for wound dressing applications. Throughout the 1980s, numerous wound products have been developed by satisfying essential properties, such as biocompatibility, non-cytotoxicity, non-immunogenicity, and stimulation of fibroblast and endothelial cell growth [7]. A suitable wound dressing absorbs excess exudates effectively. Although various types of wound dressing products are already available in the market, hydrogel-based wound dressings are still of interest to researchers and the market [8, 9]. Hydrogels are hydrophilic polymer matrices that absorb water and exudates from 10 to 1000 times their dry weights [10]. A hydrogel-based wound dressing creates a cooling surface to reduce pain [9] and maintains a moist environment to facilitate autolytic debridement of necrotic tissues [11]. It is pain-free, particularly when wound dressings are removed, due to the adhesion-free coverage of sensitive underlying tissues [9, 12].

Alginate is a polyanionic, hydrophilic, and natural polymer extracted from marine brown algae. It is generally regarded as an inexpensive, non-toxic, and biocompatible biomaterial [13]. It has been widely used in medical applications, such as tissue engineering, dental implantation, and wound dressing [14]. However, it has relatively poor mechanical properties. To address this, alginate is combined with pectin [15]. Pectin is a polyanionic and hydrophilic polymer extracted from fruit and vegetable pomaces. It has been extensively used in wound dressing [16]. The use of hybrid hydrogels derived from alginate and pectin (AP) has been considered for wound dressing applications [17]. Although these polymers are non-toxic and biocompatible, their antibacterial properties are not sufficient to prevent wound infection [18]. Wound dressings must have satisfactory antibacterial activity. In this study, mangosteen extract (ME) was used for skin infection treatment and wound healing; the ingredient has been used for centuries [19]. Additionally, it is effective against numerous bacterial (both gram positive and negative bacteria) and fungal infections [20] and has various other characteristics, such as anti-inflammatory, antioxidant, and anti-allergy properties [20, 21]. However, owing to ME has a short duration of bacterial inhibition. Thus, developing antibacterial substances is imperative for long-term bacterial inhibition. Using polymers as drug carriers allows for the sustained and controlled release of antibiotics to prevent long-term infections [22]. Niosomes, a subcategory of drug delivery carriers, present promising properties, such as high drug encapsulation efficiency and non-toxicity [23]. Niosomes are nonionic surfactant vesicular systems. Substances are loaded into the vesicles, which are fabricated by nonionic surfactants, such as Span or Tween, and cholesterol (cho) through the sonication method [24]. Niosomes have been used for various applications, such as targeted, sustained, and controlled release of substances [25].

To the best of our knowledge, this study was the first to examine the sustained release, antibacterial activity, and wound healing properties of ME-loaded niosomes incorporated into A/P hydrogels. A study showed that hydrogels made from hyaluronic acid and chitosan lacked homogeneity because deacetylated chitosan is insoluble [26]; however, all wound dressings were applied to the skin at a pH of 5.5. Due to the constraints of combined hyaluronic acid and chitosan, natural polymers, such as sodium alginate (SA) and pectin, were used in this study.

2. Materials and methods

2.1. Materials

Garcinia mangostana was obtained from Asia & Pacific Quality Trade Co., Ltd (Bangkok, Thailand); Chloroform, from Thai Public Oil Co., Ltd (Bangkok, Thailand); SA (CAS:9005-38-3), pectin (CAS:9000-69-5), and dimethyl sulfoxide, from Sigma–Aldrich; Mueller–Hinton agar (MHA) and trypticase soya agar, from Difco™, Becton Dickinson Co., and Spark (MD, USA); Dulbecco's modified Eagle's medium (DMEM), fetal bovine serum (FBS), 0.25% trypsin–ethylenediaminetetraacetic acid, trypan blue, phosphate-buffered saline (PBS), penicillin, streptomycin, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT), glycerin, Span™ 60, and cho, from Gibco (CA, USA); and all solvents, from RCI Labscan Ltd (Bangkok, Thailand).

2.2. Preparation of ME-loaded niosomes

2.2.1. ME extraction

G. mangostana Linn pericarps were collected, cleaned with tap water, and air-dried at room temperature at the RECIHP laboratory., The pericarps were dried in a hot air oven at 45 °C for three days and ground into a fine powder in a stainless steel jar blender (model no. BL-Y66S-1; Guangdong, China). The dried powder (100 g) was extracted with 500 mL deionized (DI) water in an autoclave at 105 °C and 15 psi for 60 min. The aqueous phase was filtered through Whatman No. 1 filter paper (Whatman International Ltd., Kent, UK). The filtered crude aqueous extract was lyophilized under vacuum at—80 °C for 18 h. The crude extract yield was 4.62%. The crude extract was stored at −30 °C until use.

2.2.2. Compound analysis by liquid chromatography–quadrupole time-of-flight mass spectrometry (LC–Q-TOF-MS)

The constitution of ME was evaluated by ultra-high performance LC, using a Zorbax Eclipse Plus C18 Rapid Resolution HD column (2.1 × 150 mm, 1.8 μm) and an LC-Q-TOF-MS apparatus to detect positive and negative electrospray ionization (ESI) mass spectra. The temperature was maintained at 40 °C. Mobile phase A comprised 0.1% formic acid in water, and mobile phase B comprised 11.5% acetonitrile solution. The elution steps were as follows [1]: 70% A and 30% B for 2 min [2], 5% A and 95% B for 14 min [3], 70% A and 30% B for 14.20 min, and [4] 70% A and 30% B for 20 min. The sample volume was 2 μL, with a flow rate of 0.2 mL/min. The drying gas temperature was set to 350 °C at 10 L/min; sheath gas temperature, to 275 °C; and nebulizer pressure, to 60 psig. Using a dual Agilent Jet Stream ESI source, we conducted LC–MS/MS in positive and negative ion modes, with a scanning range of 100–1500 m/z.

2.2.3. Probability of bioactivity in G. mangostana L. Compounds

Using the PASS online server (http://way2drug.com/PassOnline/), we predicted the bioactivity of the top-selected G. mangostana L. compounds. The categories of antibacterial and wound healing activities were considered for prediction purposes. The software forecasts the activity spectrum of a substance as either “probable activity (Pa)” or “probable inactivity (Pi).” The predictions are based on structure-activity relationships (SARs) [27].

2.2.4. Calibration of ME concentration

ME solution (45 mg/mL) was prepared in DI water and subjected to wavelength scanning using a UV spectrometer at 200–600 nm. The λmax of ME was 285 nm (Figure 1A). The freeze-dried ME extract was dissolved in DI water at various concentrations (0, 9, 18, 45, 150, and 300 mg/mL). Using a microplate reader (Tecan, Switzerland), we determined the standard curve of ME solution by measuring the absorbance at 285 nm (Figure 1B).

Figure 1.

(A) UV absorption spectra of G. mangostana L. Pericarp extract at a pH of 5.5 (blue line) and 7.4 (red line). Standard curve of mangosteen extract at a pH of 5.5 (B) and 7.4 (C).

2.3. Antibacterial study of ME

The antibacterial activity of ME extracts was determined using the disk diffusion and broth dilution methods. The Kirby–Bauer disk diffusion method was used to determine the antibacterial activity of eight different concentrations of ME against S. epidermidis and S. aureus [28]. ME was briefly dissolved in normal saline solution (NSS) to obtain various concentrations (0.01, 0.03, 0.15, 0.3, 2.5, 5, 10, and 20 mg/mL). Four 6-mm blank disks containing 20 μL of each diluted ME extract solution were placed in plates seeded with different test bacteria. Vancomycin disks and disks impregnated with 20 μL of DI water were used as the positive (PC) and negative controls (NC), respectively. These were incubated for 24 h. Using ImageJ, we determined antibacterial activity by measuring the sizes of the zones of inhibition (ZOIs) and disk diameters in millimeters. The diameter of each disk was averaged from three ZOIs on different plates. A test extract with a ZOI ≥7 mm was considered positive [29], While a ZOI of 0 was considered negative. Activity indices (Ais) were calculated by dividing the ZOI of ME by that of vancomycin. An AI >0.5 was considered a significant antibacterial activity [30].

Using eight different concentrations, the minimum inhibitory concentration (MIC) of ME against S. aureus and S. epidermidis was determined through the Clinical and Laboratory Standards Institute (2010–2012) broth microdilution method [31]. Mueller-Hinton broth was briefly used to dilute bacteria. A UV spectrophotometer was used at a wavelength of 600 nm and an optical density (OD) of 0.08–0.1 to quantify bacterial concentration (approximately 10⁸ CFU/mL), and 10 μL of 10⁸ CFU/mL bacteria was added to each well. The bacterial concentration in the final inoculum was 10⁵ CFU/well. Eight two-fold serial dilutions of ME (100 μL) were added to the wells for the final high and low concentrations of 2.93 mg/mL and 0.02 mg/mL, respectively. Vancomycin and NSS were used as the PC and NC, respectively. After incubation at 37 °C for 24 h, pictures were captured before staining (Figure 2C, 1^st and 3^rd rows), and 10 μL of 0.005% resazurin was added to each well. To determine the MIC, antibacterial activity was examined at a λmax of 600 nm (blue) to determine the absence of viable bacteria [32, 33]. MEs with potentially therapeutic concentrations (changing from pink to blue) showed visible signs of antibacterial activity (MIC against the bacteria). All MIC values were averaged from six different plate readings. To determine the minimum bactericidal concentration (MBC), concentrations higher than an MIC of 10 μL were dripped and spread on Mueller–Hinton agar plates to determine the absence of bacterial growth.

Figure 2.

A) Percentage of bacterial viability. B) MIC₅₀ and MIC₉₀ of ME for S. aureus and S. epidermidis. C) Resazurin test to determine the MIC of ME.

2.3.1. ME encapsulation in niosomes

Niosomes were prepared through a thin-film hydration process employing Span60 (S60) (a non-ionic surfactant) and cho [34]. First, several ratios of S60 and cho were dissolved in chloroform (2:1, 3:1, 4:1, and 5:1) (Table 4). After 15 min of agitation, the chloroform evaporated and left a thin-film coating on a glass vial. The blank niosomes were hydrated with DI water, while, ME-loaded niosomes were submerged in the ME solution (1 mg/mL) and agitated for 30 min. The niosomes were constructed by probe sonication for 30 s at an amplitude of 80% amplitude and temperature of 4 °C. The ME-loaded nisomes were investigated by fourier-transform infrared (FTIR) microspectroscopy to define the encapsulation of ME in niosomes.

Table 4.

Blank niosomes.

S60: Cho	Size (nm) ± SD	PDI	Zeta potential (mV)	%DL	%EE	% Yield
2: 1	202.7 ± 2.46	0.46	-25.9	n/a	n/a	88.9 ± 6.94
3: 1	223.1 ± 2.43	0.38	-53.1	n/a	n/a	91.7 ± 3.09
4: 1	225.7 ± 3.66	0.39	-53.3	n/a	n/a	93.3 ± 2.45
5: 1	212.6 ± 3.77	0.28	-42.1	n/a	n/a	88.9 ± 4.99

Data of mean +SD of 3 independent experiments.

2.4. Physical characterization of ME-loaded niosomes

2.4.1. Size, polydispersity index, and zeta potential

The niosomes were diluted (1:100) with DI water to remove the effects of multiple scattering and immediately inserted into the module. Droplet size was determined immediately after sample processing. Using Zetasizer Nano ZS (Malvern Panalytical, Malvern, UK), we measured the mean particle size and size distribution of the nanoparticles through the integrated light scattering method. Each nanoparticle preparation was evaluated in duplicate, with 30 measurements taken for each nanoparticle suspended in DI water. The mean particle size and polydispersity index were determined. The polydispersity index reflects the dispersion size in a monodisperse system, with values < 0.3 suggesting accurate measurements and high-quality colloids [35]. Each experiment was performed in triplicate. The tables and figures show the mean values for the three batches. The surface charge of the nanoparticles was assessed using Zetasizer Nano ZS and converted to zeta potential using the Smoluchowski equation [36]. The measurements were performed in an aqueous solution at a temperature of 25 °C.

2.4.2. ME entrapment efficiency

UV-visible spectroscopy was used to assess the quantity of ME in the niosomes. The newly synthesized niosome solution (10 mL) was purified by centrifugation at 3,000 rpm for 10 min and filtered through a 0.45-μm syringe filter to eliminate niosome aggregation [37]. The unencapsulated ME was removed through a centrifugal filter (MilliporeSigma, MA, USA) with a molecular weight cut-off of 50 kDa. The filter separated the drug-loaded micelles above from the unencapsulated ME below. The unencapsulated ME was collected and lyophilized. The lyophilized particles were dissolved in chloroform to determine the amount of ME in the particles. The Lyophilized micelles were weighed for quantification.

ME has an absorbance of 285 nm. The following equations were used to compute its drug attributes, such as DLC, EE, and Y [38, 39]:

$$\%\mathrm{D}\mathrm{L}\mathrm{C}=\frac{\mathrm{a}\mathrm{m}\mathrm{o}\mathrm{u}\mathrm{n}\mathrm{t}\mathrm{o}\mathrm{f}\mathrm{d}\mathrm{r}\mathrm{u}\mathrm{g}\mathrm{i}\mathrm{n}\mathrm{a}\mathrm{n}\mathrm{i}\mathrm{o}\mathrm{s}\mathrm{o}\mathrm{m}\mathrm{e}}{\mathrm{t}\mathrm{h}\mathrm{e}\mathrm{o}\mathrm{r}\mathrm{e}\mathrm{t}\mathrm{i}\mathrm{c}\mathrm{a}\mathrm{l}\mathrm{a}\mathrm{m}\mathrm{o}\mathrm{u}\mathrm{n}\mathrm{t}\mathrm{o}\mathrm{f}\mathrm{n}\mathrm{a}\mathrm{n}\mathrm{o}\mathrm{p}\mathrm{a}\mathrm{r}\mathrm{t}\mathrm{i}\mathrm{c}\mathrm{l}\mathrm{e}\mathrm{s}}\times 100$$ (1)

$$\%\mathrm{E}\mathrm{E}=\frac{\mathrm{a}\mathrm{m}\mathrm{o}\mathrm{u}\mathrm{n}\mathrm{t}\mathrm{o}\mathrm{f}\mathrm{d}\mathrm{r}\mathrm{u}\mathrm{g}\mathrm{i}\mathrm{n}\mathrm{a}\mathrm{n}\mathrm{i}\mathrm{o}\mathrm{s}\mathrm{o}\mathrm{m}\mathrm{e}}{\mathrm{i}\mathrm{n}\mathrm{i}\mathrm{t}\mathrm{i}\mathrm{a}\mathrm{l}\mathrm{a}\mathrm{m}\mathrm{o}\mathrm{u}\mathrm{n}\mathrm{t}\mathrm{o}\mathrm{f}\mathrm{d}\mathrm{r}\mathrm{u}\mathrm{g}\mathrm{i}\mathrm{n}\mathrm{t}\mathrm{h}\mathrm{e}\mathrm{s}\mathrm{y}\mathrm{s}\mathrm{t}\mathrm{e}\mathrm{m}}\times 100$$ (2)

$$\%\mathrm{Y}=\frac{\mathrm{p}\mathrm{r}\mathrm{a}\mathrm{c}\mathrm{t}\mathrm{i}\mathrm{c}\mathrm{a}\mathrm{l}\mathrm{t}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l}\mathrm{a}\mathrm{m}\mathrm{o}\mathrm{u}\mathrm{n}\mathrm{t}\mathrm{o}\mathrm{f}\mathrm{n}\mathrm{a}\mathrm{n}\mathrm{o}\mathrm{p}\mathrm{a}\mathrm{r}\mathrm{t}\mathrm{i}\mathrm{c}\mathrm{l}\mathrm{e}\mathrm{s}}{\mathrm{t}\mathrm{h}\mathrm{e}\mathrm{o}\mathrm{r}\mathrm{e}\mathrm{t}\mathrm{i}\mathrm{c}\mathrm{a}\mathrm{l}\mathrm{t}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l}\mathrm{a}\mathrm{m}\mathrm{o}\mathrm{u}\mathrm{n}\mathrm{t}\mathrm{o}\mathrm{f}\mathrm{n}\mathrm{a}\mathrm{n}\mathrm{o}\mathrm{p}\mathrm{a}\mathrm{r}\mathrm{t}\mathrm{i}\mathrm{c}\mathrm{l}\mathrm{e}\mathrm{s}}\times 100$$ (3)

2.4.3. Evaluation of the in vitro release of ME from niosomes

Niosomes containing ME were placed in a dialysis bag (MWCO; Spectrum Labs, CA, USA) with a molecular weight cutoff of 15 kDa [40]. The bag was immersed in 25 ml PBS (pH 5.5 and 7.4) and incubated in an incubator shaker at a temperature of 37 °C and speed of 90 rpm [41, 42]. The PBS was replaced at regular intervals to maintain a sink condition. Using SwissADME (http://www.swissadme.ch/index.php), we calculated the sink condition by selecting the main constituents of ME, as previously identified by LC–Q-TOF-MS. A microplate reader was used to determine the drug release rate in the collected samples.

2.5. Preparation of ME-loaded niosomes incorporated in AP hydrogels

2.5.1. Synthesis of AP hydrogels

AP films were produced using a solvent casting method. A 1:1 ratio of SA and pectin was obtained by completely dissolving the particles in DI water at 900 rpm. Based on the weight of the SAP powder, 15% (w/w) glycerol was added as a plasticizer agent during the production process [43]. The solution was placed in an ultrasound machine for 15 min and degassed for 12 h. To create a SAP film with glycerol, 10 mL of the solution was poured onto Petri dishes with a diameter of 10 cm. The films were completely dried for two days at 55 °C with controlled humidity (50%). To form thin hydrogels with better mechanical strength, dry film samples with acceptable dimensions were cut and submerged for 30, 60, and 120 s in a 6 mL aqueous solution of 0.5% and 1% (w/v) CaCl₂. The hydrogels were cleaned with distilled water and dried at 55 °C until a constant weight was attained before use.

2.5.2. Preparation of ME-loaded niosomal AP patch

For niosomes (S60 and cho), a 2:1 ratio results in a high percentage of drug loading, yield, and encapsulation. Therefore, the ratio was chosen for the wound dressing formulation. To create the wound dressing, we used S60, cho, SA, pectin, glycerin, crude ME extract, and purified water. To create the formulation, we used glycerin to dissolve the crude extract. Glycerin is a non-toxic and highly biocompatible substance. It is frequently used in food, cosmetics, and pharmaceutical products. To prepare the SAP/ME films, an appropriate amount of ME solution was progressively added to the alginate solution until the final SAP/ME ratio of 85:15 (v/v) was reached [44]. The mixture was agitated at 600 rpm for 30 min and 10 mL was poured onto each Petri dish. The films were completely dried for two days at a temperature of 55 °C with controlled humidity (50%). Dry film samples with acceptable dimensions were cut and submerged for 30 s in a 6 mL aqueous solution of 1% (w/v) CaCl₂. The hydrogels were washed with distilled water and dried at 55 °C until a stable weight was attained before use.

2.6. Hydrogel characterization

2.6.1. Swelling ratio analysis

The swelling behavior of the SAP films was investigated at room temperature. Each film was submerged in separate containers of 5 mL PBS with a pH of 7.4. The samples were collected and rapidly blotted twice on a Kimwipe paper at 5, 10, 15, and 30 min and then weighed. The initial wet weight (Wo) and the succeeding wet weights (W) were recorded at predefined time intervals throughout the immersion experiment [45]. The swelling ratio was determined using Eq [4]. Three samples were evaluated for each group. The data are presented as means ± standard deviations.

$$\%\mathrm{s}\mathrm{w}\mathrm{e}\mathrm{l}\mathrm{l}\mathrm{i}\mathrm{n}\mathrm{g}=\frac{\mathrm{W}-\mathrm{W}\mathrm{o}}{\mathrm{W}\mathrm{o}}\times 100$$ (4)

2.6.2. ME release study

Drug release was examined by immersing the ME-loaded SAP films (10 mg) in 25 mL PBS with a pH of 5.5 at 37 °C. To maintain the sink condition, the samples were removed at predefined time intervals and replaced with fresh PBS. Using SwissADME (http://www.swissadme.ch/index.php), we calculated the sink condition by selecting the main constituents of ME, as previously identified by LC–QTOF MS. A microplate reader was used to determine the amount of ME released at an absorbance of 286 nm. The trial was conducted in triplicate. Drug release was also tested in further experiments for the antibacterial study.

2.6.3. Antibacterial study

Antibacterial activity was evaluated using the broth dilution method. Mueller-Hinton broth was briefly used to dilute bacteria. A UV spectrophotometer was used at a wavelength of 600 nm and an OD of 0.08–0.1 to quantify bacterial concentration (approximately 10⁸ CFU/mL), and 10 μL of 10⁸ CFU/mL was added to each well. The bacterial concentration in the final inoculum was 10⁵ CFU/well [46]. Vancomycin and NSS were used as the PC and NC, respectively. After the release of ME (100 μL), the wells were incubated for 24 h at 37 °C. Subsequently, pictures were captured before staining (Figure 2C, 1^st and 3^rd rows), and the MIC values were determined at different time points. Specifically, 10 μL of 0.005% resazurin was added to each well, and antibacterial activity was detected at a λmax of 600 nm to indicate the absence of viable bacteria (blue) [32, 33]. MEs released from SAP films with potentially therapeutic concentrations (changing from pink to blue) showed visible signs of antibacterial activity. All values were averaged from six different plate readings. The values indicated the duration of the SAP/ME patches’ antibacterial effect.

2.6.4. Cell and sample preparation

A mouse fibroblast cell line (NCTC clone L929, IFO50409) was obtained from the Japanese Collection of Research Bioresources Cell Bank (Osaka, Japan). The cells were maintained in DMEM (Gibco, CA, USA) supplemented with 10% FBS (Gibco, CA, USA) and 1% (v/v) antibiotic–antimycotic solution (Gibco, CA, USA) at 37 °C under a 5% CO₂ atmosphere.

All samples (blank, 10, 15, and 20 mg ME-loaded membranes) were cut into 1 × 1 cm squares, sterilized under UV light for 1 h, washed with DI water, extracted according to ISO 10993 Part 12, and submerged in DMEM at 37 °C for 72 h. The collected (extracted) medium from each sample was filtered (Minisart® 0.2 μm cellulose acetate syringe filters; Sartorius, Göttingen, Germany) and stored at –20 °C until use.

2.6.5. Viability assays

The test was performed according to ISO 10993–5 [47]. L929 cells were seeded at a density of 10,000 cells/well in a 96-well plate and incubated for 24 h to allow for cell attachment. The culture medium was diluted with the extract to 100, 50, 25, and 10%. The cells were further maintained for 24 h before a quantitative MTT assay was performed [48]. To evaluate cell viability, the cells were incubated with 0.5 mg/mL of MTT solution (VWR Life Science, AMRESCO LLC, OH, USA) for 1 h at 37 °C. The insoluble formazan crystals were dissolved and measured using a microplate photometer (HiPo MPP-60; Biosan, Rīga, Latvia) at 568 nm. The cells that remained in the growth medium were used as the NC, whereas cells treated with 1% Triton X-100 in PBS (pH 7.4) for 15 min were used as the PC.

2.6.6. Cell morphology study

Sterilized membranes were used as substrates for cell attachment. L929 cells were directly seeded (10,000 cells/mL) onto the membranes in a 6-well plate. Sterilized glass coverslips were used as controls. After a 24-h r incubation, all samples were fixed with 3% glutaraldehyde (Sigma–Aldrich, MO, USA) for 15 min, dehydrated with an ethanol series [49, 50], and treated with hexamethyldisilazane (Sigma–Aldrich, MO, USA) for 5 min [51]. The samples were sputter-coated with gold and examined using a scanning electron microscope (SEM; JCM-7000; JEOL, Tokyo, Japan).

2.6.7. Cell migration assays

Cell migration was assessed in vitro through a scratch test [52, 53]. Cells were seeded at a density of 20,000 cells/well in 24-well plates and maintained in a standard growth media for 24 h. A scratch was created using a sterile pipette tip, and the cells were exposed to extracted mediums from distinct membranes (100%). Images were captured at 24 and 48 h, and ImageJ was used to quantify cell migration.

2.6.8. Hemocompatibility

Hemocompatibility was assessed through a complete blood count. A normal red blood cell (RBC) scattergram flag was detected by the HA3 Hematology Analyzer (BioSystems S.A., Barcelona, Spain) (Figure 12). RBC hemolysis can be quantified from a normal RBC count. RBCs were adjusted to equal concentrations before being tested with NSS. DI water was used as the PC. Patches containing different doses of ME (10, 15, and 20 mg) were extracted according to ISO10993-12 [54]. The extracted substances were incubated with the prepared RBCs. Previously, WUEC-22-158-01 was approved by the Internal Review Board (IRB) of the Center of Ethic Reinforcement for Human Research at Walailak University.

Figure 12.

RBC histogram from the hemolytic activity test of (A) the NC (normal saline) (B) PC (distilled water), and wound dressings containing (C) 10 mg (D) 15 mg, and (E) 20 mg ME.

2.6.9. In vivo skin irritation study

The in vivo study was conducted at an ISO 17025-certified National Institute of Health Laboratory in Thailand. The study protocol was approved by the Animal Ethics Committee of Walailak University (protocol no. WU-ACUC-65044). The protocol complied with ISO 10993–10 [55]. We used three healthy, young female New Zealand white rabbits. Their body weights ranged from 2,000 to 3,000 g, in compliance with the OECD Guidelines for the Testing of Chemicals (Test no. 404) [56]. Animals from production sites undergo veterinary health assessments in conventional sanitary systems. At least 24 h before the test, the rabbits' fur on the back was shaved with a clipper, from the shoulder blades down to the line parallel to the lower spine (approximately 2.5 cm). The approximate width and length were 15 cm and 10 cm, respectively. Hydrogel films containing ME were cut to a size of 2.5 × 2.5 cm² and a thickness of no more than 0.5 mm, and their surfaces were smoothened. The rabbits had normal skin conditions and no scars, and they had not undergone other tests. After preparing a test animal, four test patches were taken and covered with adhesive tape. These were tightly wrapped around the body with a stretchable fabric and then wrapped with a transparent plaster to prevent sliding. After 4 h, all test patches were removed. The sample that remained on the test skin was gently wiped with distilled water several times and then allowed to dry. Abnormalities were observed in the area covered by the test patches at 1, 24, 48, and 72 h. The results and scores were recorded, and the primary irritation index was calculated. The scoring and reading criteria were based on skin redness (erythema and eschar formation) and edema formation.

2.7. Statistical analysis

All experiments were conducted in triplicate, and the results are presented as means ± SDs. Analysis of variance was performed using SPSS Statistics 17.0. Statistical significance was set at P < 0.05.

3. Results and discussion

3.1. Analysis of the chemical constituents and biological activities of G. Mangostana L

We performed LC–Q-TOF-MS in the positive and negative modes to qualitatively and quantitatively investigate the chemical components of ME. The corresponding SARs revealed the chemical constituents and bioactivity of ME. The area under the curve identified phenolic compounds, quinones, sugars, and xanthones as the principal ingredients of the top five ME extracts (Tables 1 and 2). The most significant −ESI compounds were 1) threo-isocitric acid (retention time [RT] = 2.092 min), which accounted for 7.48% of all ESI compounds; 2) epicatechin and 2-methoxy-1,4-benzoquinone (RT = 10.675 min), 6.85%; 3) procyanidin B2 (RT = 10.073 min), 5.86%; 4) quinic acid (RT = 2.305 min), 3.14%; and 5) catechin (RT = 9.684 min), 2.45%. Meanwhile, the most significant + ESI compound was palmitic amide (RT = 20.696 min) at 11.46%. Using the SAR method, we evaluated the structural formulae of these compounds and compared with their antibacterial and wound healing efficacies (Table 3). Given a Pa value > 0.7, threo-isocitric acid and palmitic amide reduced bacterial growth by inhibiting pseudolysin, as in previous studies [57, 58]. This showed that threo-isocitric acid and palmitic amide suppressed antimicrobial resistance and biofilms. Meanwhile, 2-methoxy-1,4-benzoquinone and quinic acid promoted wound healing through anti-inflammatory effects. Numerous studies [59, 60] have reported that quinones have antiviral, anti-ulcer, antitumor, and anti-inflammatory properties. Due to these, quinones have been used as scaffolds [61] to promote cell growth. Additionally, procyanidin B2 exhibited antihemorrhagic properties. It promoted the angiogenic activity of endothelial progenitor cells and wound healing in diabetic mice [62]. Specifically, procyanidin B2 protected the angiogenic activity, survival, and migration of endothelial progenitor cells by eliminating high glucose-induced oxidative stress and its corresponding damage. These mechanistic results depended on Nrf2 activation. Lastly, catechin and epicatechin exhibited antioxidant and antihistamine properties, respectively. Using an in vivo model, a previous study [63] evaluated the effects of catechin and epicatechin on scar formation in full-thickness incision wound healing. Both chemicals significantly improved the quality of scar formation in terms of collagen fiber growth and arrangement and increased the number of new blood vessels.

Table 1.

Characterization of G. mangostana L. compounds by LC-QTOF-MS in the −ESI mode.

No.	Name	Formula	RT	m/z
1	Galactaric acid	C6 H10O8	1.916	209.0303
2	D-Glucose 6-sulfate	C6H12O9S	1.953	259.0129
3	L-Altruronic acid	C6 H10O7	2.041	193.0353
4	Galactonic acid	C6H12O7	2.066	195.0513
5	Threo-Isocitric acid	C6H8O7	2.092	191.0199
6	Quinic acid	C7H12O6	2.305	191.0563
7	Oxalosuccinic acid	C6H6O7	2.468	189.0042
8	Homoisocitrate	C7H10O7	2.618	205.0353
9	Sucrose	C12H22O11	2.618	341.1089
10	Coriose	C7H14O7	2.643	209.0666
11	Maleic acid	C4H4O4	2.819	115.0037
12	Malic acid	C4H6O5	2.869	133.0143
13	Arabinopyranobiose	C10H18O9	2.995	281.0876
14	Shikimic acid	C7H10O5	3.02	173.0455
15	D-Galactopyranosyl-(1->3)-D-galactopyranosyl-(1->3)-L-arabinose	C17H30O15	3.12	473.1506
16	3-Furoic acid	C5H4O3	3.597	111.0088
17	Citric acid	C6H8O7	4.626	191.02
18	Ascladiol	C7H8O4	4.727	155.035
19	2,5-Dimethyl-3(2H)-furanone	C6H8O2	5.003	111.0452
20	Phthalate 3,4-cis-dihydrodiol	C8H8O6	5.053	199.0246
21	Methyl-2-alpha-L-fucopyranosyl-beta-D-galactoside	C13H24O10	5.429	339.1295
22	Isopropyl apiosylglucoside	C14H26O10	6.032	353.145
23	3-Glucosyl-2,3′,4,4′,6-pentahydroxybenzophenone	C19H20O11	7.914	423.0936
24	Pyrocatechol	C6H6O2	8.015	109.0296
25	2,6-dihydroxybenzoic acid	C7H6O4	8.04	153.0192
26	Cynaroside A	C21H32O10	8.441	443.1919
27	(-)-epicatechin-3′-O-glucuronide	C21H24O11	8.742	451.1235
28	Chlorogenic Acid	C16H18O9	9.345	353.0875
29	3,4-Dihydroxybenzaldehyde	C7H6O3	9.521	137.024
30	(±)-Catechin	C15H14O6	9.684	289.0718
31	Procyanidin B2	C30H26O12	10.073	577.1349
32	Verbasoside	C20H30O12	10.223	461.1659
33	Khellol glucoside	C19H20O10	10.424	407.0981
34	Coriandrone C	C13H10O5	10.449	245.0456
35	2-Methoxy-1,4-benzoquinone	C7H6O3	10.675	137.0242
36	Epicatechin	C15H14O6	10.675	289.0721
37	Marmesin	C14H14O4	10.7	245.0818
38	Vanillin 4-sulfate	C9H10O6S	10.851	245.0125
39	2-Acetyl-5,8-dihydroxy-3-methoxy-1,4-naphthoquinone	C13H10O6	11.127	261.0406
40	Epifisetinidol-(4beta->8)-catechin	C30H26O11	11.227	561.1399
41	4-O-8′,5′-5″-Dehydrotriferulic acid	C30H26O12	11.629	577.1342
42	Garcimangosone D	C19H20O9	11.83	391.1031
43	Amaroswerin	C29H30O14	12.381	601.1585
44	O-Demethylfonsecin	C14H12O6	12.733	275.056
45	Miscanthoside	C21H22O11	12.846	449.1086
46	Streptonigrin	C25H22N4O8	12.997	505.1381
47	Citrusin B	C27H36O13	13.235	567.2076
48	(R)-1-O-[b-D-Glucopyranosyl-(1->6)-b-D-glucopyranoside]-1,3-octanediol	C20H38O12	14.038	469.2286
49	Urolithin D	C13H8O6	16.046	259.0245
50	Coriandrin	C13H10O4	18.054	229.0508
51	Muscomin	C18H18O7	19.008	345.0978
52	(2S,4S,6S)-2-[2-(4-Hydroxy-3-meyhoxyphenyl)ethyl]tetrahydro-6-(4,5-dihydroxy-3-methoxyphenyl)-2H-pyran-4-yl 4-acetate	C23H28O8	20.163	431.1705
53	Cubebininolide	C24H30O8	21.417	445.1861
54	Ugaxanthone	C18H16O6	24.781	327.0874
55	Garcinone C	C23H26O7	26.989	413.1602
56	1,4,6-Trihydroxy-5-methoxy-7-prenylxanthone	C19H18O6	27.441	341.1025
57	1-Isomangostin hydrate	C24H28O7	27.993	427.1759
58	Garcinone A	C23H24 O5	33.364	379.1549
59	Dulciol C	C28H34O7	33.515	481.2223
60	BR-Xanthone A	C23H24O6	34.669	395.1499
61	α-Mangostin	C24H26 O6	35.121	409.1657
62	Dulciol B	C28H32O6	35.987	463.2124

Table 2.

Characterization of G. mangostana L. compounds by LC-QTOF-MS in the +ESI mode.

No.	Name	Formula	RT	m/z
1	Choline	C5H14NO	1.898	104.1073
2	6-(alpha-D-Glucosaminyl)-1D-myo-inositol	C12H23NO10	1.947	342.1394
3	Trimethylammonioacetate	C5H12 NO2	2.023	118.0863
4	α-D-Glucose	C6H12O6	2.023	203.0525
5	Maltose	C12H22O11	2.199	365.1058
6	Trigonelline	C7H8NO2	2.199	138.0552
7	Valiolone	C7H12O6	2.299	215.0527
8	Firocoxib	C17H20O5S	2.324	337.1103
9	Malonylcarnitine	C10H18NO6	2.449	248.113
10	Sucrose	C12H22O11	2.5	365.1056
11	2-(beta-D-Glucosyl)-sn-glycerol	C9H18O8	2.751	277.0894
12	(-)-Dioxibrassinin	C11H12N2O2S2	2.801	306.9967
13	Galactose-beta-1,4-xylose	C11H20O10	3.102	335.0946
14	Adenine	C5H5N5	3.504	136.0615
15	4-Guanidinobutanoic acid	C5H11N3O2	3.529	146.0923
16	D-Galactopyranosyl-(1->3)-D-galactopyranosyl-(1->3)-L-arabinose	C17H30O15	4.508	497.1468
17	DL-o-Tyrosine	C9H11NO3	4.633	182.0808
18	Cystine	C6H12N2O4S2	4.658	241.0317
19	Ethyl beta-D-glucopyranoside	C8H16O6	4.658	231.0839
20	(R)-N-Methylsalsolinol	C11H15NO2	4.934	194.1175
21	2′-Aminoacetophenone	C8H9NO	4.959	136.0754
22	Kojic Acid	C6H6O4	5.135	143.0338
23	2,3-Butanediol glucoside	C10H20O7	5.211	275.1102
24	Methyl-2-alpha-L-fucopyranosyl-beta-D-galactoside	C13H24O10	5.361	363.1256
25	Isopropyl beta-D-glucoside	C9H18O6	5.713	245.0994
26	Isopropyl apiosylglucoside	C14H26O10	5.988	377.1414
27	Leonuriside A	C14H20O9	6.34	355.0997
28	Cynaroside A	C21H32O10	8.473	467.188
29	3-Hydroxycoumarin	C9H6O3	9.327	163.039
30	2-[4-(3-Hydroxypropyl)-2-methoxyphenoxy]-1,3-propanediol 1-glucoside	C19H30O10	9.351	441.1727
31	Procyanidin B2	C30H26O12	10.029	579.1502
32	Procyanidin B3	C30H26O12	10.054	601.1311
33	Phenylethyl primeveroside	C19H28O10	10.23	439.1569
34	Epicatechin	C15H14O6	10.657	291.0866
35	cis-3,4-Leucopelargonidin	C15H14O6	10.681	313.0682
36	2-Methoxy-1,4-benzoquinone	C7H6O3	10.682	139.039
37	Epifisetinidol-(4beta->8)-catechin	C30H26O11	11.234	563.1539
38	Unshuoside A	C16H28O7	13.041	355.1724
39	Isolariciresinol 9-O-beta-D-glucoside	C26H34O11	13.242	545.1987
40	2,6 dimethylheptanoyl carnitine	C16H32NO4	13.668	302.2326
41	S-Furanopetasitin	C24 H32O5S	13.945	433.204
42	Hulupinic acid	C15H20O4	16.379	287.1254
43	Palmitic amide	C16H33NO	20.696	256.2633
44	C16 Sphinganine	C16H35NO2	20.721	274.2754
45	Alpha-CEHC	C16H22O4	32.742	301.141
46	α-Mangostin	C24H26O6	35.127	411.1804
47	Isoartocarpesin	C20H18O6	35.177	355.1176

Table 3.

Pa and Pi values of phytochemical G. mangostana L. for antibacterial and wound healing predicted using the PASS online server.

No.	Compounds	Pa	Pi	Activities
1	Threo-isocitric acid	0.800	0.006	Pseudolysin inhibitor
2	Epicatechin	0.791	0.003	Histamine release inhibitor
3	2-methoxy-1,4-benzoquinone	0.743	0.011	Anti-inflammatory
4	Procyanidin B2	0.884	0.002	Antihemorrhagic
5	Quinic acid	0.705	0.015	Anti-inflammatory
6	Catechin	0.810	0.003	Antioxidant
7	Palmitic amide	0.802	0.006	Pseudolysin inhibitor

3.2. ME concentration standard curve

We used a scanning spectrogram with a step length of 1 nm to determine the detection wavelength of the ME in a universal microplate reader (Figure 1A). The absorbance varied with the pH of the surrounding environment. At a pH of 7.4, ME exhibited absorbance at 285 nm. At a pH of 5.5, ME exhibited absorbance at 286 nm. Thus, the detection wavelengths of ME were 285 and 286 nm, respectively [64, 65]. The universal microplate reader has the analytical power to simultaneously detect multiple samples. For example, an ultraviolet spectrophotometer can simultaneously detect a maximum of three samples, making it unsuitable for large samples. Large samples cause random and systematic errors that result in poor reproducibility and stability of experimental data [66].

In contrast, a universal microplate reader can simultaneously detect hundreds of samples. It utilizes a vertical rather than a horizontal optical path, creating a shorter light path than that of the UV spectrophotometer. This helps reduce detection errors [67, 68]. A calibration curve was created by plotting a graph of absorbance versus ME concentration (0.05–0.5 mg/mL) (Figures 1B and 1C). Six concentrations were used, and each concentration was analyzed in triplicate. The regression equation for ME was constructed using the least squares method (Figures 1B and 1C), where y represents the absorbance value of ME, while x represents the concentration of ME in mg/L. The linear correlation coefficient (R²) was 1.0000 (n = 6) [69].

3.3. Antibacterial study of ME

ME was effective against S. aureus and S. epidermidis, as shown by the MIC and MBC values. These investigations employed various ME concentrations from 0.02 to 2.93 mg/mL (Figure 2A). The MIC₅₀ and MIC₉₀ concentrations of ME extract against S. aureus were 0.54 ± 0.01 and 1.27 ± 0.03 mg/mL, respectively (Figure 2B). The MIC₅₀ and MIC₉₀ values of ME against S. epidermidis were 0.17 ± 0.01 and 0.61 ± 0.02 mg/mL, respectively. These results showed that ME can inhibit S. epidermidis better than S. aureus because a lower ME concentration was used against S. epidermidis. However, the patch requires a sufficient ME concentration (a minimum concentration of 1.27 mg/mL) to suppress both the growth of both bacteria.

In addition, the antibacterial activity of ME was examined using the agar diffusion method. Figure 3 summarizes the results obtained from the corresponding ZOIs. The agar diffusion method measures the width of a substance that can inhibit bacteria. Generally, the size of the ZOI increased with antibacterial property. The NCs included the same components as the tested samples, except for ME. MEs with doses <0.15 and 0.3 mg did not inhibit S. aureus and S. epidermidis, respectively. Meanwhile, 20 mg ME inhibited S. aureus and S. epidermidis, with ZOIs reaching 17 ± 1.1 mm and 16 ± 1.2 mm, respectively. These findings demonstrated that ME concentration increased with antibacterial activity. However, excessive levels of ME have detrimental effects on fibroblasts (L929). Lower concentrations of ME were optimal for bacterial growth inhibition. The optimal AI value was [30] >0.5. S. aureus (AIs = 0.64) and S. epidermidis (AIs = 0.52) were suppressed by 2.5 mg ME.

Figure 3.

Disk diffusion (ZOI) results for S. epidermidis (left) and S. aureus (right): Vancomycin as the PC, NSS as the NC, and different concentrations of ME.

3.4. Characterization of ME-loaded niosomes

We investigated the effects of different S60 and cho ratios (Table 4). The ME-loaded niosomes are presented in Table 5. FTIR was performed to examine the possible interactions between ME and niosomal components (Figure 4). Previous studies that used 2 to 5 parts S60 demonstrated that the proportion of S60 increased with size and % Y. A 4:1 S60 and cho ratio showed the highest % Y. However, % Y values decreased with the addition of S60 to ME-loaded niosomes. A 2:1 ratio of S60 and cho was selected because it showed the highest % Y while preserving ME content and a low PDI This indicated that the particle sizes of ME in this study were similar to those reported in another study [70]. These results concluded that a high percentage of S60 increased % Y and DLC values; however, if S60 exceeds 4 parts, the excess niosomes will produce the lowest % Y [71]. This may be explained by the presence of multiple bonds, which result in chain bending and alteration in molecular packing conditions [72]. In this study, the niosomal membrane exhibited increased permeability [73], which explained why excessive S60 in niosomes resulted in the lowest % Y. Therefore a lower percentage of S60 also resulted in a higher % Y and % DLC (Table 5).

Table 5.

ME-loaded niosomes.

S60: Cho: ME	Size (nm) ± SD	PDI	Zeta potential (mV)	%DL	%EE	% Yield
2: 1: 1	214.2 ± 9.0	0.23	-43.0	16.2 ± 0.87	64.8 ± 3.49	87.3 ± 3.09
3: 1: 1	358.5 ± 14.3	0.44	-35.0	13.7 ± 0.05	68.5 ± 0.26	53.0 ± 2.45
4: 1: 1	282.8 ± 10.2	0.44	-32.0	11.7 ± 0.25	70.24 ± 1.50	52.7 ± 2.49
5: 1: 1	353.6 ± 9.9	0.46	-30.1	9.8 ± 0.19	68.91 ± 1.33	50.0 ± 8.16

Data of mean ± SD of 3 independent experiments.

Figure 4.

ATR-FTIR spectra of niosomes (red line), ME (green line), and ME-loaded niosomes (blue line).

3.5. Characterization by FTIR

FTIR spectroscopy was performed to examine the chemical composition of ME, niosomal surface, and entrapment of ME in niosomes. Figure 4 shows the FTIR spectra of the ME extracts (green line). Numerous stretching vibration bands linked to organic functional groups were readily apparent in all observed spectra. At 3,274 cm⁻¹, 1,608 cm⁻¹, 1,442 cm⁻¹, 1,060 cm⁻¹, and 675 cm⁻¹, the round-shaped bands correspond to the vibrations of the O–H of alcohol, alkanes, C=O and C–O stretching, C–C, and the C=C of alkene [74]. As in the previous UV-visible spectrometry, the presence of phenolic arenes and polycyclic aromatic hydrocarbons with two or more rings were confirmed. This indicated the presence of cyanidin-3-sophoroside [74]. Many other compounds, such as quinones and xanthones, were identified (Tables 1 and 2).

Figure 4 illustrates the FTIR spectrum of freeze-dried niosome powder (red). All peaks associated with cho and S60 were present in the niosome spectra. Additional peaks that were not apparent in the cho or S60 spectra were also observed. The cho spectra revealed C–O stretching (1,060 cm⁻¹), C–H bond stretching (2,952 cm⁻¹), C–H bond bending (1,373 cm⁻¹), and –OH stretching (a wide peak at 3000–3700 cm⁻¹). The S60 spectra revealed C=O stretching (1,731 cm⁻¹); -C–CO–O- (1,180 cm⁻¹); aliphatic C–H stretching; and asymmetric, symmetric, and aliphatic (2,914 cm⁻¹, 2,852 cm⁻¹, and 723 cm⁻¹, respectively) –CH2- rocking (723 cm⁻¹). ME-loaded niosomes were used to illustrate the peaks of ME. Lacking peaks in ME upon being loaded into niosomes indicated that ME was successfully loaded into niosomes. Significant hydrogen bonding across formulation components was attributed to the extremely wide peak at 3,000–3,700 cm⁻¹. According to previous investigations, hydrogen bonds are more likely to form among ME, cho, and S60 [75,76].

3.6. Release of ME from niosomes

In studying the solubility of drug constituents, drug release is essential to sink condition. The maximum water solubility should not be reached during drug release. The sink condition of each substance was calculated using SwissADME (http://www.swissadme.ch/index.php) (Table 6). The least soluble substance was procyanidin B2, with a solubility of 4.15 × 10⁻³ mg/mL in water. In this study, we used 25 mL PBS. In this regard, the maximum amount of procyanidin B2 in the total solution was 0.104 mg, which was 5.86% of the total substance. Thus, the maximum amount of ME that remained in a sink condition was 1.775 mg. In its in vitro release, ME did not exceed 1.775 mg at any interval. The highest amount of ME released was 1.567 mg, indicating that ME remained in the sink condition.

Table 6.

Main compounds of ME and their water solubility properties.

No.	Compounds	Percent containing	Solubility (mg/mL)
1	Palmitic amide	11.46%	6.31 × 10−3
2	Threo-isocitric acid	7.48%	5.20 × 102
3	Epicatechin	6.85 %	1.74
4	2-methoxy-1,4-benzoquinone	6.85 %	35.3
5	Procyanidin B2	5.86%	4.15 × 10−3
6	Quinic acid	3.14%	6.48 × 102
7	Catechin	2.45%	1.74

Using two different pH values (5.5 and 7.4), we investigated the release of ME from niosomes with various ratios of S60 and cho (Figure 5). The amount of ME released from niosomes varied with pH. This revealed that ME-loaded niosomes loaded were pH sensitive. The amount of ME released decreased when the pH increased from 5.5 to 7.4. Thus, 85% of the ME released at a pH of 5.5 decreased to 70% at a pH of 7.4 within seven days. When administered to the wound, the minimal amount of ME released at a pH of 7.4, which corresponded to the pH of the blood, helped reduce side effects. Thus, the pH sensitivity of niosomes increased ME release on the skin surface (pH: 5.5). Compared with other ratios, the 2:1 ratio of S60 and cho (red line) allowed for a quicker ME release (Figure 5A) potentially because it had the lowest % EE (Figure 5B), resulting in more unencapsulated ME [77]. Additionally, increased surfactant structures in S60 delayed the release of ME from niosomes [78].

Figure 5.

In vitro drug release profile of ME-loaded nanoparticles prepared using different ratios of S60 and cholesterol; 2:1 (red), 3:1 (blue), 4:1 (green), and 5:1 (orange) at a pH of 5.5 (square) and 2:1 (red), 3:1 (blue), 4:1 (green), and 5:1 (orange) at a pH of 7.4 (triangle) shown mg (A), and percent (B).

3.7. Swelling ratio characterization

When a dry film comes into contact with a moist wound surface, wound exudates are absorbed into the film matrix. This hydrate expands and transforms the film into a gel on the wound surface. This film expansion study aimed to assess the rate of wound dressing expansion (or hydration). A gelatin model was utilized to recreate the environment of a wet wound surface [79]. Figure 6A shows the percentage of film enlargement over time. The square-shaped films expanded in all directions and were finally converted into gels after a delayed hydration process on the gelatin surface. Hydration occurred more slowly in 1% SAP crosslinking than in 0.5% SAP crosslinking. After 30 min, SAP crosslinked with 0.5% and 1% Ca²⁺ progressively expanded to 867 and 1,025% of their starting weights, respectively. These films expanded more quickly potentially due to their poor mechanical strength [80]. As such, long-term crosslinking with high Ca²⁺ concentrations changed the morphology of these patches. Fig. 6B-H shows that 0.5% Ca²⁺ crosslinking should last no more than 60 s, while 1% Ca²⁺ crosslinking should last no more than 30 s at a time. These results showed that crosslinking SAP with 1% Ca²⁺ for 30 s may preserve the form of the patches during wound dressing application, while undergoing a hydration process to maintain a moist wound bed environment. Extended use of a wound dressing that retains its structure and form is ideal for extremely exudative wounds [81].

Figure 6.

(A) Swelling curves of wound dressings containing a 2:1 ratio of alginate and pectin crosslinked with 0.5% (red) and 1% (green) CaCl₂ in PBS with a pH of 7.4. ∗P < 0.05, ∗∗P < 0.01, and ∗∗∗P < 0.005 compared at the same time points. Physical appearance of AP hydrogel films at CaCl₂ concentrations of 0.5% (C, D, and E) and 1% (F, G, and H) and time intervals of 30 (C and F), 60 (D and G), and 90 (E and H) min compared with that of control without crosslinking (B).

3.8. Antibacterial study of ME-loaded niosomes incorporated in AP hydrogels

Bacterial infections are widely known to have increased wound exudates, which delay the wound-healing process [82]. Thus, an ideal wound dressing material should have potent antibacterial capabilities to accelerate wound healing by lowering the number of microorganisms and decreasing the inflammatory response of wounds [83]. In this study, we conducted a test to determine the antibacterial activity of ME released from niosomes in SAP patches (Figure 7A and 7B) against S. aureus and S. epidermidis. Antibacterial activity was determined using 10⁵ CFU/mL bacterial suspensions. Based on these findings, the most effective patches were those that comprised 1 part ME and were placed in niosomes with a 2:1 ratio of S60 and cho and then crosslinked with 1% AP for 30 s. A 1:1 ratio of ME and culture media inhibited (>90 percent) S. aureus and S. epidermidis for up to 4.5 days (Figures 7C and 7D). This indicated that the antibacterial action of the ME released significantly decreased over time. Nevertheless, the ME concentration increased, enabling the effective suppression of both bacteria. Additionally, while an excessive amount of ME inhibits bacteria, it may have a detrimental impact on normal cells as well.

Figure 7.

In vitro drug release of ME (A) Concentration of ME release in mg/mL and B) percentage of cumulative drug release (C) Percentage of bacterial viability after exposure according to ME release at various time points (D) Resazurin test after the exposure according to drug release at specified time intervals.

The antibacterial activity of the SAP dressings incorporated with ME-loaded niosomes (Figure 8A) was determined by calculating the ZOIs around the patches. On culture plates inoculated with S. aureus and S. epidermidis, wound dressing patches with medium (15 mg) and high doses (20 mg) showed consistent bacterial growth inhibition (Figure 8B) and thus sustained antibacterial activities. After one day of incubation, the ZOIs surrounding the wound dressing patches were observed. As demonstrated by narrower ZOIs, a medium dose of ME-loaded niosomes incorporated into SAP patches showed a less effective antibacterial activity than that those with a high dose.

Figure 8.

(A) Physical appearance of a wound dressing (B) Disk diffusion (ZOI) results of a wound dressing containing medium- (15 mg) and high- (20 mg) doses of ME for bacterial inhibition.

3.9. Cell viability assay

Biocompatibility testing of medical products and devices is critical to ensure their safe use, particularly in patients. The biocompatibility of the patches was in terms of cytotoxicity (in vitro cell culture model), hemocompatibility, and skin irritation (animal model). Cytotoxicity testing is critical for wound-dressing patches [84]. In vitro cytotoxicity may illustrate the presence of changes inside cells, ranging from cell death to minor changes in some cellular activities. An MTT assay is often used to determine in vitro cytotoxicity because it is a rapid and efficient approach to determine mitochondrial dysfunction that correlates well with cell growth. It is based on the use of tetrazolium salt MTT, which may be transformed into an insoluble blue formazan product by mitochondrial enzymes in live cells [85]. The L929 cell line is frequently used to assess the cytotoxicity of medical devices according to ISO 10993–5 [86]. Therefore, we used the L929 cell line as an in vitro model for cytotoxicity assays. Figure 9A shows the corresponding cytotoxic effects on cell viability and morphology. Both low- and medium-dose patches had no cytotoxic effects, exhibiting good cell viability (>80%) for all extract concentrations. Additionally, Figures 9B-9S, except for 9C, shows no significant alteration in the morphology of L929 cells after a 24-h culture in DMEM supplemented with patch extracts. While high-dose patches (20 mg) influenced L929 survival, cell survival was <80% at 100% and 50% DMEM extraction. Therefore, low- (10 mg) and medium-dose (15 mg) wound dressings are optimal because these do not inhibit fibroblast growth.

Figure 9.

(A) Histograms representing the percentage, with respect to the NC (100%), of viable cells after exposure to wound dressings containing 0 mg (MC), 10 mg (L), 15 mg (M), and 20 mg (H) ME. ∗P < 0.05, ∗∗P < 0.01, and ∗∗∗P < 0.005 compared with the NC. Morphology of L929 cells treated with DMEM as the NC (B), 1% Triton X-100 as the PC (C), and after exposure to various extractions from 100% (P, Q, R, and S) at the bottom to 10% at the top (D, E, F, and G) and different samples at 0 mg (MC) (D, H, L and P), 10 mg (E, I, M, and Q), 15 mg (F, J, N, and R), and 20 mg (G, K, O, and S). Scale bars = 100 μm.

3.10. Cell morphology study

SEM images were taken to study cell adherence to the patches Cell proliferation and filopodia were observed. For the low- (Figure 10C) and medium-dose (Figure 10D) patches, the cells adhered to the patches and were distributed compared to glass coverslip (Figure 10A) and blank hydrogel (Figure 10B). However, due to their large size, fibroblasts cannot enter the pore cavity and attached to the patches' surface. Fibroblasts produce and arrange the ECM, which is critical for lesion healing, and prevent the development of hypertrophic scars and keloids [87, 88]. For the high-dose patches (Figure 10E), fibroblast development was inhibited. Additionally, the cells remained spherical and lacked filopodia, indicating the patches' non-cytocompatibility [89]. This was due to ME's toxicity at higher concentrations. Therefore, we recommend the use of medium-dose (15 mg) patches.

Figure 10.

SEM image of L929 cells adhered to a glass coverslip (A), hydrogels with 0 mg (B), 10 mg (C), 15 mg (D), and 20 mg ME (E). Scale bars = 10 μm.

3.11. Cell migration

Figure 11 illustrates the results of the in vitro cell migration test. After a 24-hour incubation period, approximately 20.63 ± 0.08% of the damaged area healed using the bare patch due to the migration and proliferation of L929 cells into the scratched region. Compared to those treated with a bare patch, cells treated with low- (26.68 ± 1.02), medium- (49.99 ± 0.13), or high-dose patches (43.07 ± 1.61) had a statistically significant (p < 0.05) improvement in wound closure. A significant difference in wound contraction was observed in the low-dose patch compared with the bare patch. Further, compared to the low-dose patch, medium- and high-dose patches have a higher therapeutic efficacy. However, at 48 h, L929 cells treated with a medium-dose patch demonstrated a significantly greater wound healing (p < 0.05) than other types of patches. Interestingly, scratched regions treated with a medium-dose patch demonstrated a faster wound closure rate (97.79 ± 1.53).

Figure 11.

Effect of wound dressing containing ME on L929 cell migration. Scale bars = 100 μm.

3.12. Hemocompatibility

A hemolysis experiment was performed to determine the blood compatibility of SAP patches with various ME dosages. Blood compatibility is a critical for the development of substances for biomedical use. When RBCs come into contact with the extract, they swelled and burst due to the subsequent release of adenosine diphosphate. Adenosine diphosphate synthesis promotes platelet aggregation and accelerates blood clotting and thrombus formation [90]. Hemolysis, platelet adhesion, and blood clotting are distinct stages of the same process. Figures 12A–12D shows that all patches, except for the high-dose patch (Figure 12 E), with harmful RBCs were non-hemolytic. We used NSS as the NC. Due of its non-hemolytic action, a medium-dose patch has a potential in wound dressing application.

3.13. In vivo study

Using albino rabbits, we evaluated skin irritation in SAP patches incorporated with niosomes loaded with medium-dose ME [91]; the results are summarized in Table 6. Skin irritation is a localized inflammatory response that occurs immediately after stimulation. It is characterized by the development of acute inflammatory reactions clinically identified as erythema (redness), edema (swelling), itching, and discomfort [92]. Erythema and edema grading scales were used to rate the severity of skin irritation. The scores were as follows: 0 = none, 1 = slight erythema/slight edema (hardly visible), 2 = clearly visible/slight edema, 3 = moderate erythema/1-mm raised skin, 4 = serious erythema (dark red)/swelling even beyond the patch area. Table 7 shows no evidence of erythema or edema 72 h after applying medium-dose patches to the skin of the albino rabbits. All data revealed a favorable reaction in terms of skin irritation. Thus, the animal model experiments demonstrated that the SAP patches were biocompatible and had no harmful effects on skin cells.

Table 7.

Responsive scores of skin irritation.

Skin responses	Time (Hrs)	Score
		Rabbit 1	Rabbit 2	Rabbit 3
Erythema	1	1	1	1
	24	0	1	1
	48	0	0	0
	72	0	0	0
Oedema	1	1	1	0
	24	0	0	0
	48	0	0	0
	72	0	0	0
Primary irritation index (PII)		PII = 2/8 = 0.25	PII = 3/8 = 0.38	PII = 2/8 = 0.25
Average Primary irritation index = (0.25 + 0.38+0.25)/3 = 0.29

Data of mean ± SD of 3 independent experiments.

4. Conclusion

Due to the high percentage of drug loading, yield, and encapsulation in niosomes with a 2:1 ratio of S60 and cho, the ratio was selected to encapsulate ME for wound dressing. The SAP patch incorporated with niosomes loaded with a medium-dose of ME demonstrated a high water absorption rate, which served as an effective barrier against bacterial penetration. Additionally, these patches generally exhibited a negative influence on cytotoxicity and skin irritation in vitro cell culture and in vivo models. Thus, we concluded that the patches are human/animal-friendly biomaterials with a significant biomedical application potential. It may be useful for burn skin therapy and as wound dressing materials. Although these patches are effective, a more detailed research on wound healing is needed before moving on to clinical trials. Nevertheless, this patch-fabrication technology involved loading active compounds into niosomes, which were then added into AP patches. Thus, this may be a key model for the creation of hydrogel patches. In the future, newly discovered novel compounds can be incorporated into this technology.

Declarations

Author contribution statement

Komgrit Eawsakul: Conceived and designed the experiments; Performed the experiments; Analysed and interpreted; Contributed reagents, materials, analysis tools or data; Wrote the paper.

Tassanee Ongtanasup: Conceived and designed the experiments; Performed the experiments; Analysed and interpreted the data; Wrote the paper.

Jitbanjong Tangpong: Conceived and designed the experiments; Wrote the paper.

Philaslak Pooprommin: Performed the experiments; Analysed and interpreted the data; Wrote the paper.

Chawan Manaspon: Performed the experiments; Analysed and interpreted the data; Contributed reagents, materials, analysis tools or data; Wrote the paper.

Smith Wanmasae: Performed the experiments; Analysed and interpreted the data; Wrote the paper.

Surat Sangkaew: Contributed reagents, materials, analysis tools or data.

Anisha Mazumder: Contributed reagents, materials, analysis tools or data; Wrote the paper.

Anupma Dwivedi: Contributed reagents, materials, analysis tools or data; Wrote the paper.

Funding statement

Komgrit Eawsakul was supported by Walailak University, Thailand [WU-IRG-64-034].

Data availability statement

Data will be made available on request.

Declaration of interest's statement

The authors declare no conflict of interest.

Additional information

No additional information is available for this paper.