Ocular and Systemic Pharmacokinetics of Baicalein and Baicalin After Intravitreal Injection and Oral Administration in Mice

Abstract

Background: Glaucoma requires therapies that extend beyond intraocular pressure (IOP)-lowering strategies, and baicalein (BA) offers dual IOP-lowering and neuroprotective potential. This study evaluated the pharmacokinetics of BA and its major metabolite baicalin (BG) in mouse eyes and serum after intravitreal (IVT) and oral administration to determine whether non-invasive oral dosing can achieve IVT-comparable ocular exposure. Methods: BA was administered via IVT injection (100 μM) or oral gavage (20 and 200 mg/kg) in mice, and concentrations of BA and BG in serum and ocular tissues were quantified using a validated ultra-performance liquid chromatography–mass spectrometry (UHPLC/MS) method. Results: After IVT, ocular BA peaked at 331.56 ± 17.75 ng/g at 5 min and declined to 7.13 ± 0.79 ng/g at 4 h, with minimal systemic exposure. Oral administration achieved comparable or higher peak ocular BA levels (380.43 ± 52.85 ng/g at 15 min for 20 mg/kg; 309.70 ± 24.75 ng/g at 5 min for 200 mg/kg), with markedly higher ocular area under the concentration–time curve (AUC: 2455.48 ± 667.83 h·ng/g for 200 mg/kg and 1224.88 ± 751.13 h·ng/g for 20 mg/kg) versus IVT (247.07 h·ng/g). Serum BA and BG peaked at 5 min after oral dosing, with systemic BG exposure substantially exceeding BA. Conclusions: Non-invasive oral BA dosing achieves ocular concentrations comparable to IVT injection, with significantly greater overall exposure and favorable pharmacokinetic profiles. This study provides the first demonstration in mice that non-invasive oral BA administration can replace invasive IVT delivery, establishing a strong rationale for its clinical development in glaucoma and retinal disease management.

1. Introduction

Glaucoma is a common yet serious optic neuropathy characterized by the progressive loss of retinal ganglion cells (RGCs) and visual function, representing the leading cause of irreversible blindness in aging populations [1,2]. Elevated intraocular pressure (IOP) is widely recognized as a key risk factor, and while glaucoma remains incurable, its progression can be retarded by IOP-lowering interventions, the only clinically effective treatments currently available [3]. However, these medications often result in drug resistance with prolonged use and cause undesirable side effects such as burning or stinging sensations, drowsiness, red eyes, and blurry vision, which impair patient compliance and treatment efficacy [4,5,6]. Most anti-glaucoma drugs rely on topical administration, which is hindered by corneal barriers and low bioavailability [7,8,9]. More importantly, existing therapies do not fully halt disease progression even with well-controlled IOP, underscoring additional pathogenic factors such as neuroinflammation [10,11]. With the world’s aging population rapidly expanding, safe and effective glaucoma treatments with novel mechanisms of action are urgently needed.

Baicalein (BA; C₁₅H₁₀O₅; MW 270.2) is a bioactive flavonoid extracted from the roots of Scutellaria baicalensis Georgi. It is the aglycone of baicalin (BG; C₂₁H₁₈O₁₁; MW 460.4). Both BA and BG have been reported to exhibit low cytotoxicity and process potent anti-inflammatory [12,13,14], antibacterial [15,16], antioxidant [17,18], and antitumor properties [19,20] in vitro and in vivo. BA has been suggested to display greater therapeutic potency than its glucuronide BG, likely due to its enhanced lipophilicity and superior barrier permeability in neuronal tissues [21]. In contrast, BG primarily functions as a circulating reservoir through enterohepatic recirculation, requiring enzymatic deconjugation to regenerate the bioactive aglycone BA [21,22].

Recent work has revealed BA’s therapeutic potential in glaucoma through dual IOP-lowering and neuroprotective effects [23,24,25,26]. In a retinal ischemia/reperfusion (I/R) mouse model, weekly intravitreal (IVT) injections preserved RGCs and visual function by inhibiting microglia activation and modulating immune responses [24]. Despite these encouraging findings, IVT delivery is invasive and less practical for clinical use. Most existing ocular therapeutics depend on topical drop administration, which is usually constrained by poor bioavailability. The efficacy of oral administration as an alternative remains to be established.

Although the serum pharmacokinetics of BA and BG have been reported in humans [27], monkeys [28], rats [29,30,31], and mice [32], their ocular disposition in mice, a widely used animal model for ocular disease research, has not yet been investigated. This study investigates whether oral BA achieves effective ocular concentrations comparable to those obtained via IVT, as adopted in previous studies [23,24]. This was addressed by quantifying the pharmacokinetics of BA and its metabolite BG in mouse eyes and serum following IVT or oral administration.

2. Materials and Methods

2.1. Materials

Analytical standards of baicalein (BA, ≥95%), baicalin (BG, ≥98%), and formononetin (≥95%), were purchased from Cayman Chemical (Ann Arbor, MI, USA) and Aladdin (Shanghai, China). (2-Hydroxypropyl)-β-cyclodextrin and ascorbic acid were obtained from Sigma-Aldrich (St. Louis, MO, USA) and Macklin (Shanghai, China), respectively. Tropicamide and 0.5% proparacaine hydrochloride were sourced from Dirui Pharmaceutical (Changchun, China) and Alcon Laboratories (Fort Worth, TX, USA), respectively. Water was purified using an ELGA system (ELGA VEOLIA, High Wycombe, UK). UHPLC-grade acetonitrile, methanol, and all other reagents were purchased from Thermo Fisher Scientific (Waltham, MA, USA) unless otherwise specified.

2.2. Methods

2.2.1. Animals and Experimental Groups

Adult male C57BL/6J mice (8 weeks old; n = 192) were obtained from Vital River Laboratory Animal Technology (Beijing, China) and housed under standard 12 h light/dark cycles with free access to food and water. All animal procedures adhered to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research and were approved by the Animal Subjects Ethics Sub-Committee of the Hong Kong Polytechnic University (ASESC Case No. 23-24/896-SO-R-HMRF).

The mice were randomly assigned to four groups for the pharmacokinetic study:

Group 1: Vehicle control (n = 6)

Group 2: IVT injection of 100 μM BA (9 time points; n = 6 per time point)

Group 3: Oral administration of 20 mg/kg BA (12 time points; n = 6 per time point)

Group 4: Oral administration of 200 mg/kg BA (12 time points; n = 6 per time point).

2.2.2. Drug Administration

For IVT administration, BA was freshly prepared by dissolving it in 20% (2-hydroxypropyl)-β-cyclodextrin immediately before use. A concentration of 100 μM BA was selected for pharmacokinetic evaluation, as this dose had previously demonstrated potent anti-inflammatory and neuroprotective effects without toxicity in mouse studies [24]. The mice were anesthetized by intraperitoneal injection of a weight-based mixture of ketamine (120 mg/kg) and xylazine (20 mg/kg) in sterile saline (1:1:4 ratio). One drop each of 0.5% proparacaine hydrochloride and 1% tropicamide was applied topically to both eyes. A limbal puncture was made with a 30-gauge needle, and 2 μL of 100 μM BA (MW 270.2 g/mol, equating to a dose of ~54 ng BA per eye) was injected into the vitreous of the eye using a glass micropipette under a stereomicroscope. Untreated mice and mice treated with 20% (2-hydroxypropyl)-β-cyclodextrin served as blank and vehicle controls, respectively. Care was taken to avoid damage to the lens and retina. After injection, the micropipette was slowly withdrawn from the eye to minimize reflux of the injected solution. Mice were euthanized by CO₂ overdose, and eyeballs and blood (500–700 μL) were collected at 0.083, 0.25, 0.5, 1, 2, 4, 8, 24, and 96 h post-injection for quantification of BA and its metabolite BG.

Similarly, for oral administration, a single oral dose of BA (20 mg/kg or 200 mg/kg) was administered by oral gavage, based on dosing regimens used in recent studies in other disease models [33,34]. Eyeballs and blood were collected at 0.083, 0.25, 0.5, 1, 2, 4, 6, 8, 10, 24, and 96 h after BA or vehicle treatment. Pharmacokinetic parameters were determined and compared between oral and IVT administration.

2.2.3. Sample Preparation

In contrast to commonly reported pharmacokinetic studies that use rabbit ocular tissues [35], the present study employed whole mouse eyes, leveraging the mouse model’s suitability for genetic manipulation [36] and the current lack of mouse ocular pharmacokinetic data for comparison. Given the small size of the mouse eye, dissection of individual ocular tissues or compartments poses a high risk of cross-contamination and unreliable results. Therefore, whole-eye homogenates and serum were used for the pharmacokinetic analysis.

After enucleation of the eyeball at each time point, the eye was promptly rinsed thoroughly with PBS, and adherent tissue was carefully removed before tissue homogenization. After centrifuging at 13,000 rpm for 10 min at 4 °C, 0.5% ascorbic acid was added immediately to the samples to stabilize BA and its metabolites BG [37,38]. The samples were stored at −80 °C until further processing. For homogenization, 100 μL of water containing 0.5% ascorbic acid and four steel balls were added to an Eppendorf tube containing one eyeball, kept at 4 °C, and homogenized using a Bertin Precellys Evolution (Paris, France) at 5800 rpm for 3.5 min. Then, 10 μL of formononetin (25 ng/mL) as internal standard (IS) and 400 μL of acetonitrile (pre-cooled at 4 °C) were added to each eye homogenate or serum sample (100 μL). The mixture was vortexed for 3 min and then ultra-centrifuged at 13,000 rpm for 15 min at 4 °C. The supernatant was dried under a stream of nitrogen, and the residue was reconstituted in 100 μL of methanol: water (50:50, v/v). Finally, 2 μL of the reconstituted solution was injected into the column-switching UHPLC-MS/MS system.

2.2.4. Pharmacokinetic Analysis via UHPLC-MS/MS System

A Shimadzu UHPLC system (Kyoto, Japan) coupled to a SCIEX 7500 Triple Quad mass spectrometer (Framingham, MA, USA) was used for the high-throughput quantification of BA and the metabolite BG. Chromatographic separation was performed on a Waters ACQUITY Premier BEH C18 column with VanGuard FIT (2.1 × 100 mm, 1.7 µm) using a mobile phase consisting of 0.1% formic acid in water (solvent A) and acetonitrile (solvent B) at a flow rate of 0.3 mL/min. The gradient elution was as follows: 0.0–3.0 min, 30–40% B; 3.0–5.0 min, 40–45% B; 5.0–5.5 min, 45–90% B; 5.5–7.5 min, 90% B; 7.5–7.6 min, 90–30% B; and 7.6–10.0 min, 30% B. The column temperature was maintained at 30 °C. Mass spectrometric detection was carried out using an ESI source in positive ion mode with multiple reaction monitoring (MRM). The optimized MS parameters were as follows: spray voltage 4500 V; temperature 450 °C; ion source gas one 55 psi; ion source gas two 70 psi; and curtain gas 46 psi. The transitions, associated entrance potential (V), and collision energies (V) were as follows: BA, 271.1 → 123.0 and 253.0, 10, 43; BG, 447.2 → 271.1 and 123.0, 10, 35; and formononetin, 269.1 → 197.0 and 253.0, 10, 40.

BA and BG standards (10 μg/mL) were prepared in methanol and further diluted with water/methanol to obtain the calibration samples. Quality control (QC) samples were also prepared in a similar manner as the calibration standards, including 8 ng/mL for BA and 4 ng/mL for BG. A comparison to a linear regression of the standard curve was used to calculate sample concentrations. The results were calculated in ng/g eye weight.

2.3. Statistical Analysis

Pharmacokinetic data were analyzed using Phoenix WinNonlin 8.1 software (Certara, Radnor, PA, USA). For IVT administration, a one-compartment model for bolus injection with first-order kinetics was applied. Given the practical difficulties of separating the vitreous cavity from the anterior chamber in mice, the whole eye was treated as a single, joint compartment [39,40,41]. The following equation was used [42]:

$$C\left(t\right)={\displaystyle \frac{Dose}{V_d}}\times {\exp }^{-kt}$$

where C(t) denotes the concentration of BA at time t, V_d is the volume of distribution, and K(t⁻¹) denotes the elimination rate constant. For reference, IVT data were also analyzed using a non-compartmental model (NCA).

For oral administration, pharmacokinetic analysis of BA and BG in the eye [43,44,45] and serum [27,29,30] was performed using non-compartmental analysis with the Phoenix NCA model. Graphs and statistical analyses were carried out using GraphPad Prism 10.4.1 software (San Diego, CA, USA). Statistical differences were assessed using Student’s unpaired t-test or two-way ANOVA, followed by Tukey’s multiple comparisons test, as appropriate.

3. Results

3.1. LC-MS Method Development and Validation

In this assay, a liquid-liquid extraction method suitable for BA yielded unsatisfactory recovery for its metabolite BG; therefore, protein precipitation was adopted for BG. As shown in

Table 1. Extraction recoveries of baicalein and baicalin in the eye and serum (n = 4).

Compound	Spiked Conc. (ng/mL)	Extraction Recovery (%)		Matrix Effects (%)
		Eye	Serum	Eye	Serum
Baicalein	2	98.99 ± 1.10	94.01 ± 1.00	102.88 ±1.97	95.76 ± 0.31
	128	95.17 ± 1.04	94.91 ± 0.77	94.33 ± 1.92	101.23 ± 0.58
	768	95.76 ± 0.31	99.59 ± 1.22	100.07 ± 2.11	103.72 ± 1.02
Baicalin	2	94.95 ± 1.97	94.05 ± 0.80	99.76 ± 1.92	94.21 ± 1.55
	128	96.07 ± 1.72	95.34 ± 0.61	84.94 ± 0.80	90.95 ± 0.66
	768	93.24 ± 1.26	97.76 ± 0.71	90.57 ± 0.76	94.07 ± 1.40
Formononetin (IS)	25	92.24 ± 0.99	95.74 ± 0.93	102.40 ± 1.78	85.45 ± 0.86

Data are presented as mean ± SEM.

, extraction recoveries for both BA and BG were high and consistent across the tested concentration ranges.

A selective UHPLC–MS/MS method was developed for the simultaneous quantification of BA and BG (Figure 1), with analytical times of 1.85 min and 4.13 min, respectively. The method was fully validated and demonstrated good linearity in both eye tissue and serum, with correlation coefficients ≥ 0.99 for BA and BG (

Table 2. Linear ranges, regression equations, and correlation coefficients (R²) of the calibration curves for baicalein and baicalin in the eye and serum.

Compound	Tissue	Linear Range (ng/mL)	Linear Regression Equation	R2
Baicalein	Eye	0.25–512	y = 0.12963x + 0.01034	0.99506
		0.25–512	y = 0.10783x + 0.19104	0.99620
	Serum	0.25–256	y = 0.22202x + 0.01908	0.99630
		2–1024	y = 0.01159x + 0.00614	0.99507
Baicalin	Eye	0.25–256	y = 0.31980x − 0.02202	0.99029
		0.25–512	y = 0.23888x + 0.03587	0.99826
	Serum	0.25–64	y = 0.71231x + 0.02398	0.99637
		16–16,384	y = 0.01103x + 0.01527	0.99116

). Extraction recoveries of BA from eye and serum ranged from 94.01% to 99.59%, while BG recoveries ranged from 93.24% to 97.76%, with a standard error of the mean (SEM) of ≤1.97%. Matrix effects for BA ranged from 94.33% to 103.72%, and for BG from 84.94% to 99.76% (SEM ≤ 1.97% across QC levels). The internal standard (formononetin) exhibited 102.40% recovery in eye homogenates and 85.45% in serum. Precision and accuracy assessments for BA and BG in both ocular homogenates and serum were performed as part of the method validation. The results demonstrated satisfactory performance, with intra- and inter-day precision (RSD) < 15% and accuracy (bias) within ±15% of nominal concentrations for quality control samples at low, medium, and high levels (Table S1). These results indicated good reproducibility and satisfactory extraction efficiency, supporting the suitability of the method for ocular and systemic pharmacokinetic studies in mice.

3.2. Ocular Pharmacokinetics of BA and BG After IVT Administration

BA concentrations in the eye were quantified using appropriate calibration curves in SCIEX OS 3.4.5 software (Framingham, MA, USA). The linear calibration curves for BA and BG in eye tissue covered concentration ranges of 0.25–512 ng/mL and 0.25–256 ng/mL, with correlation coefficient of 0.99506 and 0.99029, respectively (Table 2). The lower limit of quantification was estimated at approximately 0.125 ng/mL (signal-to-noise ratio > 3). The AIC values for the one-compartment model of BA were 58.85 and 77.20 for the 0–4 h and 0–24 h datasets, respectively. The residual plots for the non-compartmental analysis were summarized in Supplementary Figures S1–S5.

The ocular BA content, estimated as the product of the measured tissue concentration (ng/g) and the mean eye weight (0.027 g), was found to be within the range of the administered dose. BA was detectable in eye samples as early as 5 min post-injection. Subsequent time points showed a gradual decline in BA concentration, indicating progressive elimination from the vitreous after IVT administration. The maximum concentration (C_max) was reached within 5 min. The mean ocular BA concentration over time was fitted by a one-compartment model (Figure 2 and Figure 3a). For the 100 µM dose, the ocular half-life (T_1/2) of BA was 0.49 h, the projected C_max was 351.65 ng/g, and the mean residence time was 0.70 h (

Table 3. Pharmacokinetic parameters of baicalein in the eye.

Parameter	T1/2 (h)	Cmax (ng/g of Eye)	Tmax (h)	AUCall (h·ng/g)	MRT (h)	CLss (g/h)	Vss (g)
One-compartment	0.49	351.65	0	247.07	0.70	8.10	5.69
Non-compartment	0.72	331.56	0.083	295.40	0.89	6.61	6.82

T_1/2: Half-life; C_max: Observed or predicted maximum concentration; T_max: Time to C_max; AUC_all: Area under the curve across all available time points; MRT: Mean residence time; CL_ss: Clearance at steady-state concentration; V_ss: Apparent volume of distribution at steady-state concentration. Mean wet weight of mouse eye = 27 mg.

). In addition, the BA metabolite BG was increasingly detected in mouse eyes after BA injection (Figure 3b), suggesting local metabolic conversion of BA to BG, albeit at low levels.

3.3. Serum Pharmacokinetics of BA and BG After IVT Administration

Given the substantial differences in expected analyte concentrations across matrices (ocular tissue vs. serum), analytes (BA vs. BG), and administration routes (IVT vs. Oral), separate calibration curves were prepared for each condition to ensure accuracy across the study’s concentration ranges. A total of eight calibration curves were evaluated and validated during method development. The serum calibration curves for BA and BG, constructed from 9 and 11 concentration levels, were linear over the ranges of 0.25 to 64 ng/mL and 0.25 to 256 ng/mL, with correlation coefficients of 0.99630 and 0.99637, respectively (Table 2). The reported correlation coefficients represented values from individual calibration curves constructed for each matrix-analyte combination. At 5 and 15 min post-IVT injection, low but detectable levels of BA and BG were observed in serum (

Table 4. Concentrations of baicalein and baicalin in serum.

Time (h)	Concentration (ng/mL)
	Baicalein	Baicalin
0.083	1.79 ± 0.24	0.06 ± 0.007
0.25	<LLOQ	0.03 ± 0.005
0.5, 1, 2, 4, 8, 24, 96	<LLOQ	<LLOQ

Data are expressed as mean ± SEM (n = 6). LLOQ: Lower Limit of Quantification.

); however, neither compound was detectable at later time points.

3.4. Ocular Pharmacokinetics of BA and BG After Oral Administration

Ocular BA levels were assessed following oral administration. Mean BA concentrations in the eye at different time points are presented in

Table 5. Mean baicalein concentrations in the eye after IVT (2 μL of 100 μM) or oral (20 or 200 mg/kg) administration.

Time (h)	2 μL of 100 μM (IVT, ng/g)	20 mg/kg (Oral, ng/g)	200 mg/kg (Oral, ng/g)
0.083	331.56 ± 17.75	49.70 ± 2.07 ##	309.70 ± 24.75 *
0.25	236.02 ± 17.60	380.43 ± 52.85	203.67 ± 6.89
0.5	132.77 ± 15.03	65.47 ± 3.23	216.49 ± 20.19
1	124.77 ± 16.00	96.49 ± 13.15 #	324.18 ± 25.19 *
2	25.30 ± 2.56	194.63 ± 34.99	200.50 ± 8.52
4	7.13 ± 0.79	64.81 ± 3.19	191.96 ± 13.24
6	/	34.22 ± 1.37	81.14 ± 2.06
8	7.79 ± 0.99	191.02 ± 20.17	121.00 ± 11.77
10	/	17.48 ± 1.67	104.03 ± 5.33
24	4.30 ± 0.60	14.01 ± 2.32	27.56 ± 1.18
96	0.62 ± 0.12	6.50 ± 1.25	52.57 ± 3.16

Values are presented as mean ± SEM (n = 6); /: not determined. * Significantly different from the 20 mg/kg oral group at the same time point (* p < 0.05). # Significantly different from the 2 μL of 100 μM IVT group at the same time point (# p < 0.05; ## p < 0.01).

, and the concentration-time profiles of BA and BG in the eye are shown in Figure 4, with NCA pharmacokinetic parameters listed in

Table 6. Pharmacokinetic parameters of baicalein and baicalin in the eye after a single oral administration of 20 or 200 mg/kg baicalein.

Concentration	Compound	T1/2 (h)	Cmax (ng/g)	Tmax (h)	AUClast (h·ng/g)	MRT (h)	CL/F (g/h/kg)	Vz/F (g/kg)
20 mg/kg (Oral)	Baicalein	8.56 ± 1.91 *	425.20 ± 109.90	1.58 ± 0.52	1224.88 ± 306.65	10.69 ± 1.63 **	29,974.07 ± 10,597.90	247,394.32 ± 65,390.70 *
	Baicalin	5.13 ± 0.57	172.76 ± 15.39	0.083 ± 0	699.63 ± 71.22	9.06 ± 0.72	/	/
200 mg/kg (Oral)	Baicalein	9.35 ± 0.58 ***	370.40 ± 47.97	0.76 ± 0.12 *	2455.48 ± 272.64 **	11.86 ± 0.69 ***	101,059.02 ± 11,721.36 *	1,391,602.60 ± 178,176.54 *
	Baicalin	4.85 ± 0.12	558.39 ± 46.66	1.19 ± 0.16	5558.40 ± 467.38	7.86 ± 0.03	/	/
2 μL 100 μM (IVT)	Baicalein	0.75 ± 0.01	495.27 ± 63.95	0.083 ± 0	298.62 ± 65.72	0.80 ± 0.03	0.37 ± 0.06	0.40 ± 0.07

T_1/2: Half-life; C_max: Observed or predicted maximum concentration; T_max: Time to C_max; AUC_last: Area under the curve from time zero to the last measurable time point; MRT: Mean residence time; CL/F: Apparent total clearance; V_z/F: Apparent volume of distribution. Mean wet weight of mouse eye = 27 mg. * Significantly different from 2 μL of 100 μM (IVT) for baicalein in the non-compartment model (* p < 0.05; ** p < 0.01; *** p < 0.001).

. After oral administration of 20 mg/kg and 200 mg/kg BA, the total BA amount in the eyeball peaked at 15 min and 1 h, with C_max values of 425.20 ± 109.90 ng/g and 370.40 ± 47.97 ng/g, respectively. In addition, ocular BG concentrations at both doses peaked at 5 min, with C_max values of 172.76 ± 15.39 ng/g and 558.39 ± 46.66 ng/g, respectively.

3.5. Serum Pharmacokinetics of BA and BG After Oral Administration

Because BA was administered in two doses, the resulting serum concentrations of BA and BG spanned a wide range, exceeding the linear range of a single calibration curve. Therefore, two separate calibration ranges were used to ensure accurate quantification. The high-concentration calibration curve for serum BA and BG was linear from 2 to 1024 ng/mL and 16 to 16,384 ng/mL, with correlation coefficients of 0.99507 and 0.99116, respectively (Table 2).

Complete serum concentration-time profiles for BA and BG after oral administration are shown in Figure 5, with the corresponding pharmacokinetic parameters summarized in

Table 7. The pharmacokinetic parameters of baicalein and baicalin in serum after a single dose of 20 or 200 mg/kg of baicalein.

Concentration	Compound	T1/2 (h)	Cmax (ng/g)	Tmax (h)	AUClast (h·ng/g)	MRT (h)	CL/F (g/h/kg)	Vz/F (g/kg)
20 mg/kg	Baicalein	6.29 ± 0.33	133.02 ± 13.88	0.083 ± 0	103.62 ± 11.15	7.76 ± 0.25	236,436.68 ± 19,452.83	1,919,849.9 ± 108,728.07
	Baicalin	2.11 ± 0.04	5920.99 ± 309.29	0.083 ± 0	14,995.30 ± 1359.83	5.77 ± 0.07	/	/
200 mg/kg	Baicalein	4.34 ± 0.07	269.16 ± 55.17	1.42 ± 0.33	769.31 ± 80.30	6.58 ± 0.15	336,943.77 ± 30,431.13	2,159,422.7 ± 210,417.33
	Baicalin	3.32 ± 0.06	11,214.76 ± 764.57	0.72 ± 0.16	110,646.81 ± 8573.94	7.29 ± 0.07	/	/

T_1/2: Half-life; C_max: Observed or predicted maximum concentration; T_max: Time to C_max; AUC_last: Area under the curve from time zero to the last measurable time point; MRT: Mean residence time; CL/F: Apparent total clearance; V_z/F: Apparent volume of distribution.

. After oral dosing at 20 and 200 mg/kg, both BA and BG were detectable in serum at the first sampling time (5 min). BA reached 103.24 ± 5.41 ng/mL and 263.31 ± 26.67 ng/mL, and BG reached 4810.27 ± 156.40 ng/mL and 10,596.93 ± 404.70 ng/mL, respectively, with peak concentrations at 5 min exceeding those at later time points.

4. Discussion

This study systematically evaluated the feasibility of oral BA administration by comparing its ocular and systemic pharmacokinetics with those of IVT BA in mice, aiming to determine whether oral dosing can achieve therapeutically relevant ocular concentrations. To the best of our knowledge, this is the first report to characterize the pharmacokinetic profile of BA and BG specifically in mouse eyes following IVT and oral administration, thereby providing new insights into the ocular disposition and metabolism of this flavonoid.

Considering the advantages of mouse genetic models [36], including knockout, transgenic, and CRISPR-edited strains, for studying ocular diseases, ocular pharmacokinetic studies in mice are warranted and valuable. Such studies are particularly crucial for drug development, providing essential inter-species pharmacokinetic data and dosing information. While most existing ocular drugs rely on topical drops (suitable for anterior segment diseases but limited by poor bioavailability) or invasive IVT injections (primarily for posterior segment diseases), our study evaluates oral administration as a promising non-invasive alternative. This addresses a critical unmet need: systemic oral therapies that can effectively reach the eye to treat ocular disorders while avoiding the risks of repeated intraocular injections, thereby enhancing patient compliance and adherence.

Previous work in rabbits showed that after IVT injection of 50 μL of 10 mg/mL BA, the highest concentration in aqueous humor observed at 30 min (C_max of 4.11 ± 0.75 μg/g, T_1/2 of 39.9 min), while BA reached the cornea within 5 min with a C_max of 56.53 ± 17.02 μg/g and a T_1/2 of 14.8 min [35]. In the mouse model, the ocular T_1/2 of BA after IVT administration was 29.4 min, which is in good agreement with the findings in rabbits. Previous mouse ocular pharmacokinetic studies have treated the eye as a single compartment [46]. Our one-compartment model fit (R² > 0.98) supports the validity of this assumption. In addition, whole-eye homogenates have been employed to assess the pharmacokinetics of other drugs following IVT injection in mice [42], supporting its utility for obtaining reliable global exposure data in mouse ocular studies.

Our results indicated that only a small fraction of IVT-administered BA enters the systemic circulation (Table 4), and this fraction is rapidly converted to BG within 15 min. BA and BG were not detectable in serum at later time points. Despite this short T_1/2, our prior work demonstrated sustained inhibition of microglial activation and neuroprotection after weekly IVT BA in a retinal I/R injury model [24], suggesting that BA can exert prolonged biological effects even with transient exposure.

Species-dependent differences in oral absorption and metabolism have been reported [27,29,30,31,47,48]. In humans, after a 200 mg oral dose of BA, C_max is reached at 3 h with a long T_1/2 of 8.22 h [27]. BA is then hydrolyzed to BG by β-glucuronidase in the gut, with BA reaching T_max at 2.92 h and a T_1/2 of 10 h. In rats, oral administration of 18 mg/kg BA results in plasma T_max at 65.6 min, while BG reaches T_max at 20.6 min [30]. In our study, the concentration-time profiles of BA and BG after 200 mg/kg oral BA in mice showed multiple peaks, with BG maintained at relatively high systemic levels for about 10 h in the high-dose group, in line with previous reports [30,31,48]. These findings support the notion that orally administered BA undergoes glucuronidation by UDP-glucuronosyltransferase (UGT) enzymes in the intestinal mucosa and liver to form BG, which then undergoes enterohepatic circulation, where BG is deconjugated by gut microbial enzymes to regenerate BA, which is reabsorbed and further glucuronidated, leading to secondary peaks of BG in plasma [30,31,48].

As shown in Figure 5, both BA and BG were detected at 5 min, indicating rapid absorption of BA and its prompt conversion to BG in mice. BG exhibited relatively high systemic exposure, as reflected by its C_max and AUC_last in serum. Furthermore, the ocular concentration of BA at 15 min after a 20 mg/kg oral dose is comparable to the initial concentration achieved following IVT injection (Figure 6). Similarly, at 5 min after a 200 mg/kg oral dose, the ocular BA concentration is similar to that after IVT, with a higher C_max in ocular tissues at 1 h. Notably, the AUC_last for the 20 mg/kg and 200 mg/kg oral doses is approximately 4-fold and 8-fold greater, respectively, than that observed with IVT administration, indicating substantially greater overall ocular exposure with oral dosing.

Comparisons with previously reported pharmacokinetic data also reveals notable interspecies differences. While the AUC_last of BA in humans after a 200 mg oral dose (1340 ± 539 h·ng/mL) [27] and in rhesus monkeys after a 150 mg/kg oral BA (1156.1 ± 205.2 h·ng/mL) [28] are comparable, mice given 200 mg/kg BA in our study demonstrated a lower AUC_last of 769.31 ± 80.30 h·ng/mL, suggesting reduced oral bioavailability in small animal species like rodents. Moreover, the T_1/2 of BA in mice (4.34 ± 0.07 h) was shorter than that in rhesus monkeys (6.4 ± 3.6 h) and humans (8.22 ± 2.63 h), indicating more rapid metabolic clearance in smaller animals. In the non-human species studied (mice, rats, and monkeys), oral administration of BA results in extensive glucuronidation to BG, whereas this conversion appears less pronounced in humans. Following oral dosing (20 and 200 mg/kg), BA exhibits less-than-dose-proportional exposure, as reflected by C_max and AUC_last values (Figure 4 and Table 5). This observation aligns with previously reported nonlinear pharmacokinetics and absorption saturation in monkey and human studies [28,49]. The deviation from dose proportionality may be attributed to saturable gastrointestinal absorption, extensive first-pass glucuronidation to BG, and enterohepatic recirculation, as suggested by the multiple serum concentration peaks, which limits systemic and ocular bioavailability at higher doses [28,49].

Previous reports demonstrate that both BA and BG are detectable in rat brain tissue following oral administration of 300 mg/kg Scutellariae Radix extract, suggesting, although not definitively proving, blood-brain barrier penetration [47]. Direct permeability measurements across this barrier were not performed in that study. We hypothesize that BA and BG may potentially permeate ocular barriers such as the blood-retinal barrier and blood-aqueous barrier, enabling effective delivery to the eye after systemic administration. However, in this study, whole-eye homogenates were used due to analytical sensitivity limits and the minimal tissue mass of mouse eyes, which precludes reliable dissection of compartments without cross-contamination risk. While oral administration achieved ocular BA concentrations comparable to IVT injection, these represent global averages from whole-eye homogenates rather than site-specific levels (e.g., retina or optic nerve head) relative to intended therapeutic targets. This approach, necessitated by mouse eye size and assay sensitivity, limits precise assessment of intraocular distribution and ocular barrier penetration. Future studies in larger species (e.g., rabbits) should examine compartmental pharmacokinetics to better correlate exposure with therapeutic efficacy.

Following oral administration, serum BG concentrations were several times higher than those of BA, whereas ocular levels of the two compounds were more comparable. This discrepancy may arise from differences in the physicochemical properties of BA and BG, including molecular weight, size, structures, membrane permeability, and affinity [50,51]. Collectively, our data indicate that oral BA administration is pharmacokinetically favorable compared to IVT injection, with longer T_1/2 and higher AUC. The higher C_max and AUC_last suggest non-invasive systemic oral administration can maintain more sustained BA and BG concentrations in the eye over time than local, invasive IVT injection. Despite this, the higher systemic exposure and elevated serum BG levels observed post-oral dosing underscore the need for careful compound selection to ensure a safe, non-toxic regimen.

5. Conclusions

Glaucoma remains a major cause of irreversible blindness, and although IOP-lowering therapies are the mainstay of treatment, they do not fully prevent disease progression and are often limited by side effects and poor long-term compliance. BA has emerged as a promising candidate, demonstrating both IOP-lowering and neuroprotective effects in preclinical models [23,24]. These pharmacokinetic data demonstrate that oral administration of BA can achieve ocular concentrations comparable to those obtained via IVT injection, with significantly higher overall exposure over time. The findings support the potential of oral BA as a practical and non-invasive approach and provide a pharmacokinetic basis for further development of BA as an orally administered therapy for glaucoma and other ocular diseases.