Baicalin nano-micelles for dry eye disease: LC–MS/MS method validation, pharmacokinetics and pharmacodynamics in preclinical models

Abstract

This study aimed to establish a method for quantifying baicalin (BC) in rabbit ocular tissues and plasma, and evaluate the pharmacological efficacy and pharmacokinetic properties of BC and BC@HS15/DSPE-PEG2000-L-Val, a novel ocular formulation for dry eye treatment. BC@HS15/DSPE-PEG2000-L-Val or free BC was administered via eye drops to benzalkonium chloride (BAC)-induced dry eye mice. Corneal and conjunctival tissues were assessed for anti-dry eye efficacy. BC concentrations in cornea, conjunctiva, aqueous humor, and ocular plasma were quantified using LC–MS/MS. Noncompartmental pharmacokinetic parameters (AUC, Tmax) were calculated using DAS 2.0 software. The method demonstrated excellent linearity (0.50–500.00 ng/mL, r > 0.9905), precision (RSD < 10%), and accuracy (± 13%). Compared to free BC, BC@HS15/DSPE-PEG2000-L-Val significantly increased tear secretion, reduced MMP-3/MMP-9 expression, and preserved corneal epithelium integrity. In the micelle group, corneal and conjunctival Cmax values were 2.7- and 3.6-fold higher than the solution group, respectively. A sensitive and validated liquid chromatography-tandem mass spectrometry (LC–MS/MS) method was developed to measure baicalin concentrations in ocular plasma and tissues of rabbits. BC@HS15/DSPE-PEG2000-L-Val for treating dry eye demonstrated significantly superior outcomes. The nano-micelle notably enhanced BC concentration on the ocular surface and effectively prolonged its retention time.

Graphical abstract

Introduction

Dry eye disease (DED), also known as keratoconjunctivitis sicca or dysfunctional tear syndrome, is the most prevalent condition encountered in ophthalmic clinical practice (Inomata et al. 2020; Vehof et al. 2020). Patients with DED experience challenges such as impaired visual function, which negatively impacts their daily activities and overall quality of life (Tsubota et al. 2020). Consequently, there is a pressing need to develop effective treatments for this condition. Although DED is a complex disorder that is not yet fully understood, it is now recognized as being linked to ocular surface inflammation (Perez et al. 2020; Vereertbrugghen and Galletti 2022). T cells release inflammatory cytokines like IFN-γ and IL-17 on the ocular surface (Perez et al. 2016). These cytokines elevate the levels of pro-inflammatory mediators, such as chemokines and matrix metalloproteinases, on the corneal epithelium, resulting in ocular surface pain and further damage (Periman et al. 2020). Therefore, effective drug delivery to the deeper layers of the cornea and conjunctiva is essential for reducing ocular surface inflammation, which is crucial in treating dry eye syndrome (Awwad et al. 2017).

Baicalin (BC), a flavonoid compound extracted from Scutellaria baicalensis Georgi, possesses a variety of biological activities, including anti-inflammatory, antioxidant, anti-infection, and anti-tumor properties (Sun et al. 2016; Zhao et al. 2020). Research has highlighted that baicalin exhibits extensive biological effects in the eye, including anti-inflammatory, antibacterial, anti-cataract, antioxidant, and anti-angiogenesis activities. It has shown effectiveness in treating eye-related conditions like age-related macular degeneration (AMD) (Razavi et al. 2022), age-related macular edema (AME) (Sun et al. 2020), cataracts, diabetic retinopathy (DR) (Yingrui et al. 2022) and uveitis (Xiao et al. 2014). At present, some studies suggest that baicalin can relieve ocular surface inflammation and repair damaged corneal tissue, but the development of ophthalmic preparations still needs further research. The sugar groups present in baicalin contribute to its poor solubility and instability, resulting in low bioavailability, which severely limits its clinical application (Bao et al. 2022; Zhou et al. 2018). To address baicalin's limited bioavailability, nano-pharmaceutical technology has been extensively applied to enhance its performance (Li et al. 2018; Wang et al. 2023).

Although numerous studies have explored nano-formulations, the unique anatomical and physiological barriers of the ocular structure pose significant challenges in achieving effective drug concentrations in the retina and vitreous body following local or systemic administration (Chan et al. 2019; Tawfik et al. 2022). Compared to other nanoparticle formulations, nano-micelles have shown superior performance in enhancing the stability of baicalin (BC) and prolonging its residence time on the ocular surface (Begines et al. 2020; Tawfik et al. 2020). Consequently, a simple, efficient and safe mixed nano-micelle— HS15/DSPE-PEG2000-L-Val—was developed for the treatment of ocular surface diseases associated with inflammation, using BC as the therapeutic payload. The carrier system was designed using SPE-MPEG2000 and HS15, further modified with a targeting peptide transporter (Wu et al. 2022). It is well-known that various nanomaterials exert distinct effects on drug tissue distribution and metabolism, crucial factors that directly influence drug efficacy (Vaneev et al. 2021). Therefore, understanding the pharmacokinetics and pharmacological effects of BC@HS15/DSPE-PEG2000-L-Val is essential. In this study, a high-performance liquid chromatography-tandem mass spectrometry (LC–MS/MS) method was established to measure baicalin concentrations in rabbit ocular tissues and plasma. This approach was effectively applied to investigate the pharmacokinetics of baicalin following ocular administration in rabbits. The findings from this study offer a foundational reference for the detection and analysis of baicalin-containing biological samples.

Materials and methods

Materials and experimental animals

BC was procured from Aladdin Shanghai Biochemical Technology Co., Ltd. Nimodipine, fluorescein sodium, and Fmoc-L-valine were sourced from Shanghai Yuanye Bio-Technology Co., Ltd. The 15-hydroxystearic acid polyethylene glycol was obtained from BASF. DSPE-PEG2000-NH2 was acquired from A.V.T. (Shanghai) Pharmaceutical Co., Ltd. Phosphate buffer (powder) was purchased from Beijing Dingguo Changsheng Co., Ltd. Methanol and acetonitrile (LC/MS grade) came from Thermo Fisher Scientific. The 4% paraformaldehyde solution was supplied by Wuhan Baisheng Bioengineering Technology Co., Ltd., while the 3% BSA was from Wuhan Boster Biological Technology, Ltd. Phenol red thread was sourced from Tianjin Jingming New Technology Development Co., Ltd. Anti-MMP3 Rabbit pAb was produced from Wuhan Servicebio Technology CO., Ltd. and Anti-MMP9 Rabbit pAb was produced from Sawai Pharmaceutical Co. Ltd. Cy3-labeled goat anti-rabbit IgG antibody was produced from Wuhan Boster Biological Technology, Ltd. Isoflurane was purchased from RWD Life Science Co.,Ltd. The TUNEL staining kit was produced from Roche Diagnostics GmbH. All other chemicals used were of analytical reagent grade.

New Zealand rabbits, weighing between 2.0 and 2.5 kg, were procured from Beijing Jinmuyang (Beijing, China). Healthy, female, specific pathogen-free (SPF) BALB/c mice, with an average weight of 20 ± 2 g, were obtained from S-Bio Beijing Biotechnology Co. Ltd. All animal studies adhered strictly to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research and were approved by the Laboratory Animal Ethics Committee of Beijing University of Chinese Medicine. (BUCM-4–2021072301-3027).

LC–MS/MS instrument and conditions

LC–MS/MS analysis was performed using an AB SCIEX QTRAP 4500 system, equipped with a triple quadrupole-linear ion trap mass spectrometer (SCIEX, USA). The mass spectrometry utilized an electrospray ionization source operating in negative ion mode. Chromatographic separation was achieved on an ACQUITY UPLC BEH C18 column (2.1 × 50 mm, 1.7 µm, Waters) Maintained at 35 °C, with the automatic injector set to a temperature of 10 °C. The mobile phase consisted of two components: Phase A was an acetonitrile solution containing 0.1% formic acid, and Phase B was an aqueous solution with 0.1% formic acid. The flow rate was Maintained at 0.2 mL·min⁻¹. For ocular tissue analysis, the following elution gradient was applied: 20%–60% A over 0–3 min, 60%–100% A over 3–7 min, followed by 100%–20% A over 7–10 min. For ocular plasma analysis, the gradient program was 20%–40% A over 0–3 min, 40%–100% A over 3–7 min, 100% A over 7–9 min, and finally 100%–20% A over 9–12 min. The sample injection volume was 10 μL, and the UV detector was set to a wavelength of 280 nm.

Multiple reaction monitoring (MRM) mode was employed for quantitative analysis. The ion source temperature was set at 500 °C, with the air curtain gas Maintained at 20 psi and the impact gas set to medium. The ion voltage was adjusted to −4.5 kV, while both the spray gas and auxiliary heating gas were Maintained at 50 psi. For the BC quantitative ion pair, the parameters were as follows: 445.1/174.9 for the ion pair, a declustering potential (DP) of −70 V, and a collision energy (CE) of −17 eV. The internal standard, Nimodipine, had a quantitative ion pair of 417.2/122.1, with a declustering potential (DP) of −100 V and a collision energy (CE) of −30 eV.

Preparation of drug solution, standard solution and quality-control samples

The preparation methods for DSPE-PEG2000-L-Val and baicalin micelles were based on a previous study (Wu et al. 2022). The baicalin solution was prepared by dissolving it in a phosphate buffer (pH 6). The drug content in the solution was consistent with that of the aforementioned micelles.

Internal standard solution: An accurately measured amount of nimodipine was weighed and initially dissolved in methanol to obtain a concentration of 5 mg/mL. This solution was subsequently diluted with methanol to achieve a concentration of 50 μg/mL and then further diluted using 50% methanol to prepare the final internal standard solution with a concentration of 200 ng/mL.

Standard solution: An exact amount of baicalin was accurately weighed and dissolved in DMSO to achieve a concentration of 5 mg/mL. This solution was subsequently diluted with methanol to prepare an intermediate standard solution at a concentration of 50 μg/mL. Finally, the intermediate solution was further diluted with 50% methanol to create standard solutions with concentrations approximately equal to 5, 2.5, and 1 μg/mL, as well as 500, 250, 100, 50, 25, 10, and 5 ng/mL.

Quality-control samples: The intermediate standard solution was carefully measured and serially diluted with 50% methanol to prepare a range of quality control standard solutions at the following concentrations for ocular tissues: 4 μg/mL (high concentration), 400 ng/mL (medium concentration), 15 ng/mL (low concentration), and 5 ng/mL (lower limit of quantification). Similarly, the quality-control samples for ocular plasma were prepared at concentrations of 4 μg/mL, 300 ng/mL, 15 ng/mL, and 5 ng/mL, respectively.

Sample preparation

Ocular Tissue Sample: A suitable amount of rabbit ocular tissue (The rabbit eye tissue used for the methodological study was a mixture of corneal and conjunctival aqueous humor) was placed in a 2 mL EP tube and processed using a Liquid nitrogen grinder at a frequency of 25 times per second for 5 min. Subsequently, 500 μL of extraction solvent (methanol/acetonitrile/water = 2:2:1, v/v/v) was added to the EP tube. The mixture was swirled for 2 min, sonicated for 20 min, swirled for an additional 1 min, and then centrifuged at 13,000 rpm at 4 °C for 10 min. Following centrifugation, 400 μL of the supernatant was collected for storage. The remaining residue was dissolved in 250 μL of extraction solvent, swirled for 1 min, and centrifuged again under the same conditions. Subsequently, 200 μL of the resulting supernatant was collected. Finally, the two supernatants were combined to form the ocular tissue sample.

The corneal and conjunctival tissue samples were processed as described earlier. Prior to testing, 100 μL of re-dissolution solvent was added. The mixture was vortexed for 2 min, sonicated for 10 min, vortexed again for 1 min, and then centrifuged at 13,000 rpm at 4 ℃ for 10 min. Subsequently, 80 μL of the supernatant was collected in a new EP tube and centrifuged twice. Finally, 70 μL of the resulting supernatant was transferred into a liquid-phase vial and injected into the LC–MS system for analysis.

Ocular plasma sample: Orbital blood (blood from ruptured fundus vessels after enucleation of the eyeball) from New Zealand rabbits was collected into EDTA anticoagulant tubes and centrifuged at 3,000 rpm for 15 min at room temperature. A 50 μL aliquot of the supernatant was extracted, followed by the addition of 445 μL of extraction solvent and 5 μL of internal standard solution. The mixture was vortexed for 2 min and then centrifuged for 10 min. Subsequently, 380 μL of the resulting supernatant was transferred into a new EP tube, concentrated for 2 h using a rotary concentrator, and stored at −80 °C for future use.

Before the test, 50 μL of re-dissolution solvent was added, followed by vortex mixing for 2 min, sonication for 10 min, and an additional 1-min vortex. The sample was then centrifuged for 10 min. Subsequently, 45 μL of the supernatant was collected and centrifuged twice. Finally, 40 μL of the resulting supernatant was transferred to a liquid-phase vial and injected into the LC–MS system for analysis. The aqueous humor was processed similarly prior for detection.

Method validation

Specificity and linearity

A 600 μL sample of blank ocular tissue solution was taken and evaporated to dryness. Subsequently, 50 μL of re-dissolution solvent (50% methanol) was added. The mixture was swirled for 2 min, sonicated for 10 min, swirled again for 1 min, and then centrifuged at 13,000 rpm at 4 °C for 10 min. The resulting supernatant was collected and centrifuged twice more to obtain the final supernatant, which served as the blank control. For the preparation of the spiked blank Matrices, 600 μL of blank ocular tissue solution was combined with 10 μL of internal standard solution and 10 μL of a standard solution at a concentration of 200 ng/mL. The mixture was evaporated to dryness and processed using the same protocol described above. Following the administration of self-made baicalin micelle eye drops, 600 μL of ocular tissue solution and 10 μL of internal standard solution were mixed, evaporated to dryness, and processed using the same method to prepare the sample dosed with baicalin micelles. For plasma samples, 50 μL of plasma was used to prepare the blank control, the spiked blank matrices and the sample dosed with baicalin micelles, all following the same procedure outlined above. Finally, the blank control, spiked blank matrices, and biological samples dosed with baicalin micelles for both ocular tissue and plasma were individually analyzed to evaluate the specificity of the method.

An appropriate amount of blank ocular tissue extract and blank plasma was measured separately. Subsequently, precise volumes of internal standard solution and standard solution were added. Following the established sample processing method, the concentrations of the prepared sample solutions were set to 0.50, 1.00, 2.50, 5.00, 10.00, 25.00, 50.00, 100.00, 250.00, and 500.00 ng/mL. These sample solutions were then analyzed using LC–MS/MS. The peak area ratio of baicalin to the internal standard was recorded as the ordinate (y), while the baicalin concentration served as the abscissa (x). Linear regression analysis was performed, resulting in the generation of a standard curve. The linear range and the regression coefficient (r) were calculated accordingly.

Precision, accuracy, stability, recovery and matrix effect

An appropriate amount of blank ocular tissue was placed into a 2 mL EP tube. Internal standard solution and quality control sample solution were then added. Following the sample preparation method described above, solutions were prepared at concentrations of 0.50, 1.50, 40.00, and 400.00 ng/mL. These concentrations were designated as the lower limit of quantitation, low, medium, and high levels, respectively. The prepared solutions were subsequently analyzed using LC–MS/MS. For plasma samples, the prepared concentrations were 0.50, 1.50, 30.00, and 400.00 ng/mL. Six replicates were performed for each concentration, and three separate batches were processed continuously over three days.

Different conditions were established to evaluate the stability of the quality control sample, including: long-term freeze–thaw stability, where the sample was stored in a −40 ℃ environment for 14 days; freeze–thaw stability under three cycles at −20 ℃; and short-term stability, where the sample, after treatment, was placed in an automatic sampler set at 4 ℃ for 24 h.

The ratio of the baicalin blank substrate's peak area to that of the solvent was calculated, as was the ratio of the internal standard substrate's peak area to the solvent. These ratios were divided to determine the matrix effect, using the internal standard normalization method. The extraction recovery was assessed by comparing the mean peak area of the standard substance added before extraction to the mean peak area of the same standard added after extraction through the medium.

Studies on pharmacokinetics of BC@DSPE-PEG2000-L-Val in rabbits

Forty-eight New Zealand rabbits, each weighing between 2.0 and 2.5 kg, were evenly divided by gender and randomly assigned to two groups of 24. Prior to the experiment, all rabbits were acclimatized in the laboratory for one week. The experimental group was administered 40 μL (the dose was administered by eye drops in each eye for one time) of BC@DSPE-PEG2000-L-Val (at a concentration of 1.2 mg/mL, equivalent to 48 μg of baicalin), while the control group received free BC. At intervals of 5, 15, 30, 60, 90, 120, 180 and 240 min post-treatment, the rabbits were euthanized using the air plug method. Briefly, we first anesthetized the animal deeply with isoflurane (The initial concentration was 3% to 5%, and the flow rate was 1 to 2 L/min for 2 to 5 min), and subsequently 10 to 20 mL of air was infused into the ear vein into the heart to form a foam embolism that occluded the pulmonary vessels and resulted in rapid death from circulatory failure while minimizing suffering (National Research Council Committee for the Update of the Guide for the and Use of Laboratory, 2011). Samples of the cornea, conjunctiva, aqueous humor, and intraocular plasma of six eyes from three animals at each time point were collected for analysis via LC–MS/MS.

Pharmacodynamic studies on dry eye mice

Sixty BALB/c mice with an average weight of 20 ± 2 g were selected and consecutively numbered. 0.2% BAC solution was applied superficially to the ocular surface twice daily for 14 days to establish a murine model of DED (Chaudhari et al. 2023). On the 15th day, tear secretion levels and corneal fluorescein sodium staining were evaluated in all mice. The mice were then randomly divided into four groups (n = 15 per group) as follows: (1) model group, model mice treated with saline; (2) normal group, healthy, untreated mice treated with saline; (3) solution group, model mice treated with BC solution; (4) micelle group, model mice treated with BC@DSPE-PEG2000-L-Val.

Eye drops (5 μL, concentration = 1.2 mg/mL, equivalent to 6 μg baicalin) were applied superficially to the ocular surface four times daily for 7 days. At 3rd and 7th day post-administration, 6 mice from each group were randomly selected to measure tear production and assess corneal fluorescein sodium staining (n = 12 per timepoint). Additionally, 3 mice from each group were randomly chosen for tissue collection to perform the following analyses: (1) Lacrimal gland: HE staining; (2) Cornea: HE staining, TUNEL staining, and corneal immunofluorescence staining.

Corneal fluorescein staining

An appropriate amount of 1% fluorescein sodium solution was applied to the ocular surface of the mice. After one minute, the eyes were rapidly rinsed with normal saline and subsequently observed under a slit lamp. The corneal staining was graded by dividing the cornea into four quadrants, with each quadrant assessed individually (Skiadaresi et al. 2016). A score of 4 indicated a positive fluorescein plaque, while a score of 3 represented very dense dot-like fluorescence. A score of 2 signified a dense dot-like pattern, and a score of 1 indicated sparse, faintly fluorescent dots. A score of 0 denoted the complete absence of fluorescence. The individual quadrant scores were summed, yielding a total score for each eye, ranging from a minimum of 0 to a Maximum of 16.

Tear production

After administering an injection of 0.005 mL/g of a 1% pentobarbital sodium solution into the peritoneum of each mouse and securing the mouse's head, the lower eyelid was gently pulled down with surgical forceps to reveal the conjunctival sac. A phenol red thread was then placed on the palpebral conjunctiva of the lower eyelid, precisely one-third of the distance from the lateral canthus, for a duration of 10 s. The length of the moistened section, which transformed from yellow to red, was measured using a vernier caliper.

Immunofluorescence staining

Tissue cryosections were fixed with 4% paraformaldehyde for 24 h at −20 °C, then blocked with 3% BSA for 30 min. Next, the sections were incubated overnight at 4 °C with primary antibodies, including rabbit anti-MMP-3 (1:200) and rabbit anti-MMP-9 (1:100). Following this, the samples were treated with a Cy3-labeled goat anti-rabbit IgG antibody (1:400) for 50 min in the dark. Finally, the section images were analyzed and captured using a fluorescence microscope.

TUNEL staining

To assess the apoptosis of the corneal epithelium, an in situ TUNEL assay was conducted on frozen eye sections using a TUNEL assay kit, following the manufacturer’s instructions. The corneal sections were counterstained with DAPI, and images were captured using fluorescence microscopy. The experiment was repeated three times, and the average values were calculated.

H&E staining of corneal tissue and lacrimal gland

Paraffin-embedded sections of the eye and lacrimal gland were stained with hematoxylin and eosin (H&E) for histological analysis. The morphology and structural integrity of the cornea and lacrimal gland tissue were assessed using a light microscope, and representative images were captured for documentation. The thickness of the central corneal epithelium layer was measured using Case Viewer software. For each section, measurements were performed five times at the central region of the cornea, and the average value was calculated and recorded.

Data and statistical analysis

GraphPad Prism 9.5 was used for mapping and spss20.0 was used for data analysis, repeated measures analysis of variance (RM-ANOVA) and linear mixed models (LMM) were used with animal as a random effect and eye as a nested factor. The pharmacokinetic data were analyzed using DAS 2.0 software, with baicalin concentration plotted on the y-axis and sampling time on the x-axis. Pharmacokinetic parameters, such as time to peak (T_max), peak plasma concentration (C_max), mean residence time (MRT), and area under the curve (AUC), were calculated using the non-compartmental model method.

Results

Validation of methodology

Specificity

The specificity results for the baicalin detection method in ocular tissues and ocular plasma are presented separately in Fig. 1. The analysis demonstrated that both baicalin and the internal standard, nimodipine, exhibited robust responses under the established LC–MS/MS analysis and detection conditions. Furthermore, baicalin and nimodipine were effectively separated, with no interference from endogenous components observed during the determination process. These findings confirm that the method's specificity satisfies the established requirements.

Fig. 1.

Representative LC–MS/MS chromatograms of baicalin in ocular tissues (a-c) and plasma (d-f): blank samples (a, d); spiked blank matrices (blank + standard) (b, e); biological samples from animals dosed with baicalin micelles (c, f)

Linearity

The standard curve equations and correlation coefficients for baicalin detection in ocular tissues and ocular plasma were as follows: y = 0.155x + 0.000598 (r = 0.9905) and y = 0.00288x + 0.000928 (r = 0.9943). Both equations satisfied the criteria of the biological sample analysis method. They demonstrated strong Linearity within the range of 0.50–500.00 ng/mL, where the lowest points for each served as the lower Limit of quantification, established at 0.50 ng/mL.

Precision and accuracy

The results indicated that the accuracy of the lower quantitation concentration sample (0.50 ng/mL) in ocular tissues was 96.93%, with an RSD of 6.51% and a measured value of 0.48 ± 0.03 ng/mL. In ocular plasma, the accuracy for the same concentration sample was 107.32%, with an RSD of 8.26% and a measured value of 0.54 ± 0.04 ng/mL. Both results adhered to the established methodological standards. Additionally, the precision and accuracy outcomes for the quality control sample solutions at low, medium, and high concentrations are presented in Table 1. The intra-day and inter-day precision values showed RSDs of less than 15%, aligning with the required methodological criteria.

Table 1.

Accuracy and precision for the determination of baicalin in ocular tissues and ocular plasma

Samples	Concentration (ng/mL)	Intra-day (n = 6)			Inter-day (n = 18)
		Mean ± SD (ng/mL)	RSD%	Accuracy%	Mean ± SD (ng/mL)	RSD%	Accuracy%
Ocular tissues	1.50	1.49 ± 0.10	6.36	99.08	1.57 ± 0.10	6.57	104.59
	40.00	39.05 ± 1.78	4.58	97.62	41.11 ± 2.19	5.32	102.78
	400.00	348.00 ± 6.37	1.77	87.07	377.39 ± 34.35	9.07	94.36
Ocular plasma	1.50	1.40 ± 0.03	2.12	92.93	1.37 ± 0.03	3.93	91.14
	30.00	27.90 ± 0.64	1.10	93.25	28.03 ± 1.54	3.92	96.25
	400.00	356.83 ± 9.93	2.76	89.23	364.56 ± 10.43	2.84	91.15

Recovery and matrix effect

The results presented in Table 2 demonstrate that the recovery and matrix effects for baicalin in both ocular tissues and ocular plasma adhered to the methodological requirements. The relative standard deviations (RSD) for extraction recovery and Matrix effects were all below 15%, meeting the established criteria.

Table 2.

Recovery and matrix effect for the determination of baicalin in ocular tissues and ocular plasma (n = 6)

Samples	Concentration (ng/mL)	Recovery%	RSD%	Matrix effect%	RSD%
Ocular tissues	1.50	95.94 ± 0.04	3.74	103.78 ± 0.05	4.82
	40.00	91.11 ± 0.05	5.61	97.36 ± 0.03	3.16
	400.00	93.82 ± 0.03	3.16	99.31 ± 0.04	3.75
Ocular plasma	1.50	77.82 ± 2.34	3.01	102.36 ± 4.33	4.23
	30.00	67.24 ± 2.33	3.47	84.63 ± 2.05	2.42
	400.00	74.80 ± 1.54	2.06	94.66 ± 1.08	1.14

Stability

After the sample solutions of ocular tissues and ocular plasma were stored at −40 ℃ for 14 days, subjected to three freeze–thaw cycles at −20 ℃, and kept in an automatic sampler at 4 ℃ for 24 h following treatment, the stability results are presented in Table 3. The measured values closely aligned with the theoretical values, with relative standard deviations (RSDs) remaining below 15%, thereby meeting the methodological requirements.

Table 3.

Results of stability test for the determination of baicalin in ocular tissues and ocular plasma (n = 6)

samples	Concentration (ng/mL)	short-term stability (ng/mL)	Precision%	RSD%	long-term stability (ng/mL)	Precision%	RSD%	freeze–thaw stability (ng/mL)	Precision%	RSD%
ocular	1.50	1.65 ± 0.11	110.00	6.80	1.56 ± 0.13	103.85	8.07	1.49 ± 0.11	99.05	7.32
tissues	40.00	41.03 ± 1.13	102.70	2.94	38.78 ± 2.72	96.95	6.98	45.40 ± 1.13	113.30	2.54
	400.00	351.83 ± 9.37	87.97	2.72	364.33 ± 7.76	91.10	2.11	423.50 ± 28.68	105.92	6.82
ocular	1.50	1.38 ± 0.13	91.92	9.35	1.45 ± 0.09	97.12	6.04	1.30 ± 0.02	87.06	6.31
plasma	30.00	27.36 ± 0.09	91.14	0.32	30.88 ± 0.18	102.80	0.44	28.14 ± 0.46	93.40	1.78
	400.00	391.80 ± 2.77	97.94	0.66	378.40 ± 3.78	94.64	1.06	343.00 ± 2.00	85.68	1.27

Pharmacokinetic studies

Pharmacokinetic studies in rabbit cornea and conjunctiva

The pharmacokinetic curves of BC solution and micelle in the cornea are illustrated in Fig. 2. Following sample preparation, LC–MS was employed to obtain the pharmacokinetic data, which were subsequently processed using DAS2.0 software. The pharmacokinetic parameters were determined utilizing a non-compartmental model method. Both the nano-micelle and solution showed a noticeable decrease in BC levels over time. However, the rate of reduction in BC levels was significantly lower in nano-micelle-treated rabbits. After ocular administration, the concentration of BC in the cornea and conjunctiva was notably higher in rabbits administered with nano-micelle compared to those given free BC at most time points, underscoring the formulation benefits of nano-micelles (Table 4). Pharmacokinetic parameters, including C_max and Mean Retention Time (MRT), were markedly enhanced in the nano-micelle compared to free BC. Upon administration of BC@HS15/DSPE-PEG2000-L-Val, the concentration of BC in the cornea and conjunctiva reached approximately 27.60 ± 7.98 ng/g and 270.65 ± 181.42 ng/g respectively (Table 5). In contrast, these values were 10.11 ± 2.10 ng/g and 75.51 ± 36.79 ng/g for free BC in the cornea and conjunctiva, respectively. In both the cornea and conjunctiva, the BC content in the solution group was significantly lower compared to the micelle group and was rapidly metabolized. This indicates that nano-micelle formulations prolong the residence time of the drug on the ocular surface, in line with results from our previous nano-micelle ocular retention studies (Wu et al. 2022).

Fig. 2.

Concentration–time curves in conjunctiva (a), cornea (b), aqueous humor (c) and ocular plasma (d); (n = 6)

Table 4.

Temporal changes of baicalin concentration in cornea and conjunctiva of rabbits (n = 6, $\overline{x}\pm SD$)

	Cornea		conjunctiva
Time (min)	Baicalin micelle (ng/mL)	Baicalin solution (ng/mL)	Baicalin micelle (ng/mL)	Baicalin solution (ng/mL)
5	6.42 ± 2.37	1.30 ± 0.25	20.42 ± 20.37	21.08 ± 8.70
15	27.59 ± 7.29	4.33 ± 1.00	120.65 ± 109.15	47.54 ± 28.59
30	18.31 ± 3.37	8.74 ± 2.69	250.99 ± 183.37	47.54 ± 31.48
60	20.21 ± 6.56	7.95 ± 2.28	40.94 ± 32.18	27.36 ± 38.96
90	20.34 ± 6.57	5.05 ± 0.98	17.07 ± 10.27	22.88 ± 18.31
120	6.88 ± 2.09	2.32 ± 0.71	24.83 ± 21.25	6.89 ± 5.30
180	10.32 ± 4.98	3.30 ± 1.36	65.79 ± 64.06	24.11 ± 28.71
240	11.53 ± 7.76	4.68 ± 1.35	16.31 ± 15.31	17.65 ± 19.47

Table 5.

Main pharmacokinetic parameters in cornea and conjunctiva of rabbits (n = 6, $\overline{x}\pm SD$)

	Cornea		conjunctiva
pharmacokinetic parameters	Baicalin micelle	Baicalin solution	Baicalin micelle	Baicalin solution
Cmax (ng/mL)	27.60 ± 7.98	10.11 ± 2.10	270.65 ± 181.42	75.51 ± 36.79
Tmax (h)	0.25	0.50	0.50	0.38
AUC0-t (ng•h/mL)	55.28 ± 11.85	18.27 ± 3.71*	243.92 ± 160.64	94.50 ± 39.72*
AUC0-∞ (ng•h/mL)	84.06 ± 25.69	22.14 ± 4.51	454.30 ± 380.38	160.47 ± 105.68
MRT0-t (h)	1.69 ± 0.26	1.77 ± 0.05	1.34 ± 0.36	1.68 ± 0.53
MRT0-∞ (h)	3.25 ± 0.37	3.91 ± 0.85	1.24 ± 0.84	3.77 ± 4.06

1. Compared with baicalin solution, *P < 0.05

2. T_max data were presented as median

The nano-micelle formulation significantly increased the AUC_0-t and AUC_0-∞ of BC compared to free BC. In the cornea and conjunctiva, the AUC0-t of BC for nano-micelle was 3.03 and 1.52 times higher than that of free BC, respectively.

Pharmacokinetic studies in rabbit aqueous humor and ocular plasma

Drug concentrations in the aqueous humor following eye drop administration were typically low, nearing the lower limit of quantification for this method (Table 6). The aqueous humor contains high concentrations of glucose and ascorbic acid, which supply essential nutrients to the cornea and lens, facilitate the removal of their metabolic waste, support normal physiological functions, and help maintain transparency (Benagiano et al. 2024; Liu et al. 2003). As a relatively closed system, the aqueous humor presents significant barriers to drug penetration. Even when drugs do enter, they are rapidly cleared due to the aqueous humor’s dynamic cycling (Basson et al. 2022). According to the test results, the micelle group achieved an average peak concentration of 1.01 ng/mL, with an AUC₀–t of 2.47 ng•h/mL. In comparison, the solution group exhibited an average peak concentration of 0.70 ng/mL and an AUC_0-t of 1.96 ng•h/mL (Table 7).

Table 6.

Temporal changes of baicalin concentration in aqueous humor and ocular plasma (n = 6, $\overline{x}\pm SD$)

	Aqueous humor		Ocular plasma
Time (min)	Baicalin micelle (ng/mL)	Baicalin solution (ng/mL)	Baicalin micelle (ng/mL)	Baicalin solution (ng/mL)
5	0.42 ± 0.23	0.46 ± 0.30	-	-
15	0.59 ± 0.29	0.49 ± 0.27	14.90 ± 0.60	14.25 ± 6.62
30	0.83 ± 0.36	0.54 ± 0.33	3.03 ± 0.57	16.20 ± 13.45
60	0.67 ± 0.46	0.54 ± 0.31	52.33 ± 1.27	6.26 ± 2.96
90	0.69 ± 0.53	0.59 ± 0.44	210.75 ± 22.25	70.98 ± 35.37
120	0.60 ± 0.55	0.56 ± 0.33	29.53 ± 1.26	1.16 ± 0.31
180	0.59 ± 0.60	0.44 ± 0.28	21.58 ± 0.58	2.05 ± 0.07
240	0.51 ± 0.48	0.32 ± 0.13	24.95 ± 1.32	11.97 ± 4.70

Table 7.

Main pharmacokinetic parameters in aqueous humor and ocular plasma (n = 6, $\overline{x}\pm SD$)

	Aqueous humor		Ocular plasma
pharmacokinetic parameters	Baicalin micelle	Baicalin solution	Baicalin micelle	Baicalin solution
Cmax (ng/mL)	1.01 ± 0.45	0.70 ± 0.45	210.75 ± 25.70	70.98 ± 38.75
Tmax (h)	1.00	1.75	1.50	1.50
AUC0-t (ng•h/mL)	2.47 ± 1.77	1.96 ± 1.30*	193.04 ± 12.79	59.98 ± 30.72***
AUC0-∞ (ng•h/mL)	7.14 ± 9.40	3.47 ± 1.73	223.14 ± 15.84	68.29 ± 32.68
MRT0-t (h)	1.82 ± 0.36	1.93 ± 0.25	1.79 ± 0.01	1.58 ± 0.11
MRT0-∞ (h)	4.30 ± 2.78	4.18 ± 1.21	3.13 ± 0.10	2.15 ± 0.18

1. Compared with baicalin solution, *P < 0.05; ***P < 0.001

2. T_max data were presented as median

After administering baicalin micelle eye drops and baicalin solution eye drops, the drug concentration in plasma peaked at 90 min. The peak concentration reached 210.75 ng/mL for the baicalin micelle and 70.98 ng/mL for the baicalin solution. Similarly, the AUC₀–ₜ values were 193.04 ng•h/mL and 59.98 ng•h/mL, respectively. Notably, the relative bioavailability of baicalin micelle was approximately 322% compared to that of the solution group.

Due to the physiological structure of the eye, drugs administered ocularly first distribute across the anterior surface (cornea) with the tear film. They then rapidly elute from the ocular surface into the conjunctival sac and drain into the nasal cavity via the nasolacrimal duct or enter the digestive tract through the nasopharynx (Sun et al. 2020). Additionally, the blink reflex and the natural flushing action of tears accelerate the clearance of drugs from the ocular surface. For eye drop formulations, drug absorption into the bloodstream primarily occurs as the preparation traverses these pathways. Compared to other formulations, micelles exhibit superior biofilm affinity and transmembrane permeability, allowing them to be more readily absorbed into the bloodstream through the vascular mucosal surface (Yingrui et al. 2022). This enhanced absorption may explain why the micelle group consistently outperforms the solution group in experimental results.

Induction of the dry eye

The secretion levels of the lacrimal gland in the normal group and the model group were 6.43 ± 1.32 mm (n = 30) and 4.69 ± 1.22 mm (n = 90), respectively (Table 8). In comparison to the normal group, the lacrimal gland secretion in the model group was significantly reduced (P < 0.001). The fluorescein staining score was 0.83 prior to the application of BAC to the ocular surface, whereas it rose to 6.44 in the model group. Relative to the normal group, the corneal fluorescein sodium staining score was markedly elevated in the model group (Fig. 3). Following the administration of 0.2% BAC solution to the eyes of mice for two weeks, a successful model was established.

Table 8.

Tear secretion volume in dry eye model mice after administration (mm, n = 12, $\overline{x}\pm SD$))

Time	Normal group	Model group	Solution group	Micelle group
3 days	6.95 ± 0.46	5.09 ± 0.89	5.30 ± 0.86	6.33 ± 0.95
7 days	6.99 ± 0.75	6.06 ± 0.88	6.58 ± 0.59	6.89 ± 0.39

Fig. 3.

Corneal fluorescein staining results in different groups for 3 and 7 days

BC@DSPE-PEG2000-L-Val reduced corneal fluorescein staining scores

Representative fluorescein staining images are displayed in Fig. 3, facilitating the assessment and comparison of corneal surface damage across the treatment groups based on the intensity of green fluorescence. Positive plaque staining was distinctly observed on the corneal surface in three groups: the model group, micelle group, and solution group, indicating that the experimental model was successfully established in each of these groups. In contrast, only punctate staining was detected in the micelle and solution groups. By day 7, micro punctate staining was barely observable in the micelle group. When compared to the model group, both the micelle and solution groups showed significant reductions in fluorescein staining scores. Furthermore, the micelle group exhibited weaker fluorescence and lower scores compared to the solution group. These results suggest that BC@HS15/DSPE-PEG2000-L-Val MMs have strong therapeutic potential as a treatment option for DED.

BC@HS15/DSPE-PEG2000-L-Val MMs increased tear secretion

Tear volume is a critical parameter in the clinical evaluation of dry eye (DE) (Khanna et al. 2022). To assess the impact of BC@HS15/DSPE-PEG2000-L-Val treatment on tear volume in mouse models, a comparative analysis was conducted. The tear volume in the model group was measured at 6.06 ± 0.88 mm. In contrast, the micelle group and solution group displayed tear volumes of 6.89 ± 0.39 mm and 6.58 ± 0.59 mm, respectively. Notably, the micelle group demonstrated an increase in tear volume compared to the model group. Furthermore, there were significant differences in tear volume observed between the micelle group and the solution group (Table 8; Fig. 4a)

Fig. 4.

The results of tear secretion, apoptotic rate of central corneal epithelial cells, central corneal epithelial layer thickness in mice after treatment with DED model. ****P < 0.0001, ***P < 0.001, **P < 0.01, *P < 0.05

BC@HS15/DSPE-PEG2000-L-Val decreased MMP-3 and MMP-9 expressions in corneal epithelium

Studies have shown that disruption of the corneal epithelial barrier in DED is associated with an increased production of matrix metalloproteinases (MMPs), specifically MMP-3 and MMP-9 (Lu et al. 2022). To investigate this phenomenon, immunofluorescence staining was employed to assess the expression levels of MMP-3 and MMP-9 in the corneal epithelium. Representative fluorescein-staining images are presented in Fig. 5. The fluorescence intensity enabled a clear evaluation and comparison of MMP-3 and MMP-9 expression levels among the treatment groups. Positive staining was evident in three groups: the model group, the micelle group, and the solution group. This finding highlights the role of MMP-3 and MMP-9 as crucial factors in the pathogenesis of DED. Compared to the model group, the solution and micelle groups exhibited slightly weaker fluorescence after three days of treatment. By the seventh day, plaque-like staining was observed in the model and solution groups. Notably, the micelle group displayed even weaker fluorescence intensity, indicating its superior effect in reducing MMP-3 and MMP-9 expression.

Fig. 5.

Immunofluorescence staining of the cornea on DED mice model after 3 and 7 days of treatment

BC@HS15/DSPE-PEG2000-L-Val suppressed apoptosis of ocular surface epithelium

Corneal epithelial cell apoptosis has been observed in BAC-induced dry eye models. In this study, in situ TUNEL staining was employed to further investigate the effect of BC@HS15/DSPE-PEG2000-L-Val on corneal epithelial cell apoptosis. As depicted in Fig. 6, TUNEL-positive cells were evident in the model group, micelle group, and solution group. However, after three days of treatment, a significant reduction in the corneal epithelial cell apoptosis rate was observed in mice. This reduction became even more pronounced following seven days of treatment. Quantitative analysis revealed that both BC@HS15/DSPE-PEG2000-L-Val micelles and free BC effectively decreased corneal epithelial cell apoptosis. Notably, compared to the solution group, the micelle group exhibited significantly greater inhibition of corneal epithelial cell apoptosis (Fig. 4b).

Fig. 6.

TUNEL staining of the cornea of DED model mice treated for 3 and 7 days

BC@HS15/DSPE-PEG2000-L-Val alleviated damage to lacrimal gland and cornea

Upon topical application of BAC, the DED murine models exhibited alterations in ocular characteristics. Notably, there was a significant reduction in corneal epithelial thickness, along with visible damage. The corneal stroma was considerably thinner and more lax compared to the control group, with disorganized layers. The acinar Lumens of the lacrimal gland were enlarged, accompanied by increased intercellular spaces. After 7 days of treatment with BC@HS15/DSPE-PEG2000-L-Val and free BC, the damaged corneal epithelial layers began to heal. The corneal morphology smoothed out, presenting a similar epithelial thickness and stromal structure, with an increase in central corneal epithelial thickness (Fig. 4c). Meanwhile, the lacrimal gland cells started to compact, and the acini gradually returned to normal. However, neither the corneal morphology nor the lacrimal gland fully reverted to their normal states, even with BC@HS15/DSPE-PEG2000-L-Val administration. No differences were observed in improving corneal physiological status between the micelle and solution groups. Nevertheless, as illustrated in Fig. 7, the lacrimal gland treated with BC@HS15/DSPE-PEG2000-L-Val showed greater recovery than the one treated with free BC, with nearly no visible intercellular spaces. This underscores the formulation advantages of nano-micelles.

Fig. 7.

HE staining of the lacrimal gland (a) and cornea (b) in DED model mice after 3 and 7 days of drug treatment

Discussion

DED is a complex disease, the precise etiology of which remains unclear (Meloni et al. 2011; Zemanová, 2021). Numerous factors contribute to the disease's development and progression (Mohamed et al. 2022). The incidence of dry eye has been steadily increasing year by year, highlighting a significant unmet medical need for new treatment options (Cai et al. 2022). Baicalin (BC), extracted from the roots of Scutellaria baicalensis, has shown inhibitory effects on ocular surface inflammation and oxidative stress induced by dry eye (Ma et al. 2021; Xiao et al. 2014). However, its poor water solubility and pH sensitivity present challenges for effective ocular delivery, thereby limiting its clinical applications (Ashraf et al. 2018; Huang et al. 2019; Xu et al. 2022). To address these challenges, various nanocarriers based on nanomaterials, such as ZIF-8 (Yang et al. 2022) and nano-capsules (Rebibo et al. 2021), have been developed to improve BC's bioavailability and stability. In a previous study, we developed and characterized a novel nano-micelle that showed significant advantages, including high drug loading, good encapsulation efficiency and remarkable stability (Wu et al. 2022). It can enhance the retention time of the ocular surface, which may be related to the nature of the micelles themselves, covalently bind to the mucin on the corneal surface, resulting in stable and sustained drug release and improved bioavailability. This has also been verified in several clinical studies (Cai et al. 2023). However, modifications to the carrier can influence BC's pharmacokinetics, thereby impacting clinical efficacy. Consequently, pharmacokinetic and pharmacological studies are essential to provide vital references for clinical applications (Yu et al. 2018).

Tear volume is a critical index for evaluating DED progression in clinical settings (Narang et al. 2023). The results of the study indicate that tear volume increased following BC treatment, with the micelle group showing a more significant increase compared to the solution group. A larger tear volume typically correlates with a more stable tear film (Recchioni et al. 2022), and the findings suggest that BC@HS15/DSPE-PEG2000-L-Val facilitates tear film recovery. Fluorescein staining, an effective method to assess ocular surface damage (Herbaut et al. 2019), showed that the micelle group achieved a significant reduction in corneal fluorescein staining scores compared to both the solution and model groups. These results suggest that BC@HS15/DSPE-PEG2000-L-Val is more effective than free BC in promoting the repair of corneal tissue damage. Normalizing corneal epithelial morphology is another critical aspect of DED treatment. In the model group, corneal epithelial thickness was reduced, while lacrimal gland cells displayed enlarged acini and increased intercellular spaces. However, following a 7-day treatment period, HE staining results indicated that the cornea was almost completely restored, and the lacrimal gland cells were tightly arranged. These findings further confirm that BC@HS15/DSPE-PEG2000-L-Val facilitates the formation of a stable tear film. The study also revealed that the overexpression of MMP-3 and MMP-9 is associated with increased corneal epithelial cell apoptosis. Immunofluorescence staining confirmed that BC@HS15/DSPE-PEG2000-L-Val effectively inhibits the excessive expression of MMP-3 and MMP-9, thereby safeguarding corneal epithelial cells. TUNEL staining results further demonstrated that BC@HS15/DSPE-PEG2000-L-Val is more effective than free BC in suppressing corneal epithelial cell apoptosis. Consistent with the immunofluorescence staining results, the TUNEL staining findings reinforce that BC@HS15/DSPE-PEG2000-L-Val promotes the restoration of damaged corneal tissue and supports the development of a stable tear film.

The detection of drug content in eye tissue is an indispensable part of the study of ocular pharmacokinetics, but the detection of drug content in biological samples is still a challenge (Bajek-Bil et al. 2023). Most of them use HPLC method, or combined with ultraviolet spectrophotometry (Li et al. 2013), zinc ion complex method, etc., and some use HPLC–MS/MS method (Feng et al. 2018; Wu et al. 2020). Using this technique, we have established an accurate, rapid and reliable method for the determination of baicalin in ocular tissues. Due to the significant difference in matrix composition between ocular tissue and plasma, ocular tissue contains complex components such as lipids, proteins and cell debris, while plasma is rich in proteins, electrolytes and metabolites. These matrix components may bind to the drug and affect its extraction efficiency and chromatographic behavior, so different elution procedures are designed to optimize the separation effect in our study. The pharmacokinetic results corroborated this observation, demonstrating that BC@HS15/DSPE-PEG2000-L-Val exhibited rapid absorption through the cornea and conjunctiva following ocular administration. Peak concentrations were achieved quickly and sustained at high levels, with an average corneal concentration of approximately 27.60 ng/g and an average peak concentration of 270.65 ng/g in the conjunctiva.

Micelle is a nano-scale core–shell structure formed by the self-assembly of amphiphilic molecules. Its nano-size helps to penetrate the corneal barrier. Some studies have suggested that micelle significantly enhances drug corneal penetration by mechanisms such as transcellular transport, tight junction regulation, and prolonged corneal retention time. It may utilize specific transporters to enhance the corneal penetration of drugs. Notably, the corneal and conjunctival concentrations of BC were significantly higher in the micelle group compared to the solution group. This enhanced absorption can be attributed to various interactions between the nano-micelle and corneal epithelial proteins, including endocytic proteins (clathrin and caveolin), tight junction proteins (claudin-1 and ZO-1), and PepT1 transporter (Sun et al. 2022; Thompson et al. 2020; Zhang et al. 2008). These interactions promoted the uptake of BC into the corneal epithelium, thereby enhancing its bioavailability. Furthermore, compared to the solution, the nano-micelle formulation improved the stability of BC, boosting its activity and therapeutic effectiveness.

In summary, this study highlights the considerable potential of nano-micelles for drug delivery, as verified through robust analytical methods. The nano-micelle sustained higher retention and increased concentrations of BC in the cornea and conjunctiva, thereby enhancing its absorption and capacity for deep tissue delivery. Pharmacological validation underscored the significant efficacy of the nano-micelle in treating DED in mice. These findings provide strong support for the use of nano-micelles as a drug delivery system and open new avenues for related research and therapeutic applications.

Conclusion

The innovation of our study is to combine the methodological validation of LC–MS/MS with the in vivo pharmacokinetics study of the micelles. We have developed an accurate and highly sensitive LC–MS/MS method with a Linear range of 0.50–500.00 ng/mL and a lower Limit of quantification of 0.50 ng/mL. This method proved to be specific, exhibiting no interference from endogenous components. RSDs for precision, accuracy, Matrix effect, extraction recovery, and stability were all below 15%. Building upon prior research conducted in our laboratory, we successfully developed BC@HS15/DSPE-PEG2000-L-Val, which significantly improved the concentration of BC in biological samples. Using the established assay, pharmacokinetic experiments demonstrated that the nano-micelle formulation outperformed free BC in terms of enhanced drug delivery to the deeper layers of the cornea. Moreover, pharmacodynamic studies confirmed that the nano-micelle achieved superior therapeutic efficacy in treating BAC-induced dry eye model in mice. Additionally, the nano-micelle offered several advantages, including high bioavailability and ease of synthesis. These findings suggest that BC@HS15/DSPE-PEG2000-L-Val holds great promise as an effective carrier for ocular drug delivery and controlled release in the treatment of DED. In the future work, we will continue to investigate its long-term toxicity in order to carry out clinical trials and continue to add new evidence for the development and application of ocular nano-micelle.

Author contributions

All authors contributed to the study conception and design. Material preparation, data collection and analysis were performed by [Yuchen Xu], [Ziheng Wang] and [Manting Liu]. Animal experiments were designed by [Yuchen Xu], [Huimin Wu] and [Changhai Qu]. The experiment was carried out with the assistance of [Yuchang Yang], [Juan Liu], [Qian Zhu] and [Wenjuan Shi]. The project was cosupervised by [Xiaoxu Dong], [Xingbin Yin], [Jian Ni] and [Changhai Qu], they were responsible for the overall project management and experimental design, review of the manuscript, etc. The first draft of the manuscript was written by [Ziheng Wang] and all authors commented on previous versions of the manuscript.

Data Availability

The authors confirm that the data supporting the findings of this study are available within the article.

Declarations

Competing interests

The authors have no relevant financial or non-financial interests to disclose.

Ethical statement

Ethical approval was obtained for the conduct of this work, which is listed in the batch number in the article, all authors contributed to this work, which is listed in the author's contribution at the end of the article, all authors have approved the article for publication, and there is no conflict of interest between the authors. Funding for supporting this work is also noted in the acknowledgement. The authors vouch for the authenticity of the data presented in the content and that the data have not been falsified.