Discovery and optimization of a novel non-nitrocatechol COMT inhibitor for modulating levodopa metabolism

Abstract

Human catechol-O-methyltransferase (COMT) is a key target for neuropsychiatric disorders. Inhibiting COMT to prevent levodopa (L-DOPA) metabolism is a crucial strategy for Parkinson's disease (PD) treatment. While clinically used COMT inhibitors are primarily nitrocatechol-based, they often cause adverse effects, prompting efforts to develop safer non-nitrocatechol alternatives. In this study, baicalein (BA) was identified as a potent lead compound for COMT inhibition after screening a series of natural flavonoids using a fluorescence-based visualization inhibitor screening method. Subsequent multi-dimensional structural optimizations addressed the druggable deficiencies of BA, resulting in compound BA24, which demonstrated a 26-fold increase in cellular COMT inhibition and approximately 10-fold improvements in metabolic stability, membrane permeability and oral bioavailability, respectively, compared to BA. Mechanistically, BA24 competitively inhibited COMT by binding to the catechol pocket with a K_i of 89.28 nM. Furthermore, BA24 exhibited favorable safety profiles and significantly modulated L-DOPA metabolism in rats. Additionally, the relationships between the structural properties, inhibitory activity and metabolic stability of flavonoids as COMT inhibitors were comprehensively investigated. Collectively, this work not only presents a novel non-nitrocatechol COMT inhibitor with favorable safety profiles and potent anti-COMT effects both in vitro and in vivo, but also provides valuable insights into optimizing the druggability of flavonoids as lead compounds.

Graphical abstract

Highlights

• •A fluorescence method was established to visualize and evaluate COMT inhibition in vitro and in living cells
• •BA24 showed 26-fold stronger cellular COMT inhibition than baicalein and 10-fold better stability, permeability and absorption, respectively.
• •BA24 competitively inhibited COMT by binding to the catechol pocket, with a Kᵢ of 89.28 nM and an IC₅₀ of 72.6 nM.
• •BA24 exhibited favorable safety profiles and significantly modulated levodopa metabolism in rats.

1. Introduction

Human catechol-O-methyltransferase (COMT, EC 2.1.1.6) is the primary enzyme responsible for deactivating catecholamine neurotransmitters, such as dopamine, norepinephrine, and epinephrine [1,2], as well as removing catechol steroids and xenobiotic catechols in humans by transferring a methyl group from S-adenosyl-L-methionine (SAM) to a single hydroxyl group in a catecholic motif [3,4]. Due to its critical involvement in dopamine metabolism and modulation of diverse brain functions, COMT has been identified as a therapeutic target for several neuropsychiatric disorders, including pain, schizophrenia, depression and Parkinson's disease (PD) [[5], [6], [7]]. Especially for PD, administration of levodopa (L-DOPA) is a primary first-line treatment approach, yet its effectiveness is limited as it is rapidly metabolized into 3-O-methyldopa (3-OMD) by COMT in peripheral tissues before reaching the brain [8,9]. Thus, COMT inhibitors are commonly used alongside L-DOPA in PD patients to increase L-DOPA bioavailability and alleviate wearing-off symptoms [[10], [11], [12]].

Given the critical role of COMT in neurotransmitter regulation, COMT inhibitors have attracted significant scholarly attention [13]. Several COMT inhibitors, including tolcapone [14], entacapone (EN) [15], and opicapone [16] are already in clinical use. However, these COMT inhibitors are of the nitrocatechol-type [17]. Chronic use of these inhibitors may lead to various toxic side effects, such as hepatotoxicity, diarrhea, headaches, excessive sweating, and abdominal discomfort [18]. Tolcapone, in particular, is restricted in many countries due to its severe hepatotoxicity [19]. Therefore, there is an urgent need to develop novel COMT inhibitors with an improved safety profile to advance the treatment of PD.

Traditional Chinese medicine (TCM) has represented a vast reservoir of potential bioactive compounds [[20], [21], [22]]. Over the past few decades, versatile synthetic and natural compounds including numerous food and herbal constituents have been found with anti-COMT effects [[23], [24], [25]]. More recently, we found that Scutellariae radix [26], vine tea [27], and Pu-erh tea [28] have inhibitory effects on COMT, with the primary inhibitory components identified as flavonoids, including baicalein (BA), oroxylin A, and scutellarein. Furthermore, certain herbal medicines and their active constituents, particularly flavonoids, exhibited neuroprotective effects, thereby preventing or slowing the onset and progression of age-related neurodegenerative diseases [29,30]. Therefore, flavonoid compounds have become a hot topic for the development of anti-Alzheimer's/anti-Parkinson's drugs both domestically and abroad [20]. While flavonoids have significant advantages, their low bioavailability, which results in minimal absorption into the circulatory system after oral administration, presents a challenge in formulating them as effective drugs [[31], [32], [33]]. In previous studies, we have found that the rapid metabolic clearance by human UDP-glucosyltransferase (UGTs) is one of the primary reasons for the poor bioavailability and short half-life of natural phenolic flavonoids [34]. Therefore, the rational development of flavonoid drugs requires not only systematic exploration of their structure-activity relationships but also careful optimization of their druggability.

This study aimed to discover more effective anti-COMT agents from natural and synthetic flavonoids, while also exploring the relationships between structural properties, inhibitory activity, and metabolic stability of flavonoids as COMT inhibitors and their underlying anti-COMT mechanisms. For these intents, a fluorescence-based screening method using the two-photon COMT substrate 3-(benzothiazol-2-yl)-7,8-dihydroxy-2H-chromen-2-one (3-BTD) was first developed to identify, characterize, and validate COMT inhibitors in vitro and in living cells [35]. Through preliminary screening of natural flavonoids, BA emerged as a promising lead compound. To optimize the druggability of BA, docking simulations were performed to elucidate its interactions with COMT, guiding a rational structural optimization strategy. A series of BA derivatives were then designed, synthesized, and analyzed to establish the relationships between structural properties, inhibitory activity and metabolic stability, leading to the identification of BA24 as a promising non-nitrocatechol COMT inhibitor. Finally, the anti-COMT mechanisms of BA24 were investigated through inhibition kinetic analyses and molecular dynamics simulations, while its efficacy and safety were evaluated at both cellular and animal levels. These findings provide valuable insights for medicinal chemists in developing more potent flavonoid-based COMT inhibitors.

2. Experimental

2.1. Chemicals and reagents

All solvents and reagents were purchased from commercial suppliers and used without further purification unless otherwise specified. Tetrahydrofuran, petroleum ether, dichloromethane, ethyl acetate, and other common solvents were obtained from Tansoole (Shanghai, China). 3,4,5-Trimethoxyphenol and potassium carbonate were purchased from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). p-Nitrocinnamoyl chloride, 1,4-dibromobutane, and 4-fluorobenzaldehyde were obtained from Aladdin (Shanghai, China). All the natural flavonoid compounds were purchased from Shanghai Standard Technology Co., Ltd. (Shanghai, China). The purity of each compound was verified to be >96% by the supplier's analytical reports. All the reactions were monitored by thin-layer chromatography (TLC) using silica gel GF254 plates, and crude products were purified by column chromatography using silica gel (200–300 mesh) purchased from Qingdao Haiyang Chemical Co. Ltd. (Qingdao, China). The ¹H nuclear magnetic resonance (¹H NMR) (600 MHz) and ¹³C nuclear magnetic resonance (¹³C NMR) (150 MHz) spectra of all synthetic compounds were recorded using tetramethylsilane (TMS) as the internal standard on a Varian INOVA spectrometer (Varian Inc., Palo Alto, CA, USA). LC grade acetonitrile and dimethyl sulfoxide (DMSO) for liquid chromatography–mass spectrometry (LC-MS) analysis were obtained from Sinopharm Chemical Regent Co., Ltd. (Shanghai, China). Purities of the tested compounds were determined by high performance liquid chromatography (HPLC) analysis, and it was found to be higher than 95% for all final compounds. In the supplemental materials, more information regarding the chemical synthesis and structural characterization were provided.

2.2. Enzyme inhibition assays

In this study, a fluorescence-based biochemical assay was performed to determine the residual activity of COMT by using a near-infrared COMT-specific fluorescent probe 3-BTD. The method has previously been reported in detail [26].

2.3. Molecular docking simulations

The crystal structure of COMT (PDB ID: 3BWY) was retrieved from a Protein Data Bank (http://www.rcsb.org/). To further explore the inhibition mechanism of BA and BA24 that have the most potent inhibitory effects against COMT-mediated 3-BTD-O-methylation, a series of inhibition kinetics analyses were performed. The study procedure has been detailed previously [26].

2.4. Stability in human liver microsomes

In phase II metabolism, the experimental mixture consisted of 190 μL of 0.05 M Tris-HCl (pH = 7.4), human liver microsomes (HLMs) (0.1 mg/mL), 5 mM MgCl₂, Brij 58 (5 μg/mL), and the compound (10 μM). Brij 58 (initial concentration of 1 mg/mL) was pre-mixed with human liver microsomes (initial concentration of 20 mg/mL) at a volume ratio of 1:1, incubated on ice for 20 min, and then added to the mixture system. The reaction mixture was preheated at 37 °C for 3 min, and the reaction was initiated by adding 10 μL of uridine diphosphate glucuronic acid (UDPGA) (2 mM). At different time points (including 0, 15, 30 min), 200 μL of ice-cold acetonitrile solution was added to terminate the reaction. After centrifugation at 14,400 rpm for 30 min, the supernatant was analyzed by liquid chromatography–ultraviolet detection (LC-UV) (Shimadzu, Kyoto, Japan). Statistical analysis was performed using GraphPad Prism 8.0 software, and the metabolic half-life (t_1/2) was obtained by linear fitting with the linear regression function in the software. HLMs (Batch No. X008070, 150-donor mixed gender pooled) were purchased from the Research Institute for Liver Diseases Co., Ltd. (Shanghai, China).

2.5. Cell membrane permeability

U-87-MG human glioblastoma cells were used as a model partially mimics the astrocyte-mediated blood-brain barrier (BBB) microenvironment to evaluate the central nervous system (CNS) penetration of COMT inhibitors [36]. NCM460 normal colon epithelial cells were selected to assess intestinal absorption potential, as they better reflect physiological features of normal intestinal epithelium compared with the tumor-derived Caco-2 line and are more suitable for preliminary toxicity evaluation. Together, these two cell models were applied to investigate both BBB permeability and oral bioavailability during lead optimization [37].

NCM-460/U87-MG cells were trypsinized and resuspended with the corresponding medium to density at 250,000 cells/mL. A 200 μL cell suspension was subjected to incubation with BA24, BA, and EN (all at a concentration of 10 μM) at 37 °C for a duration of 30 min. Following that, the suspension underwent centrifugation (at the speed of 1000 rpm) for 3 min, and subsequently, the supernatant was collected to analyzed the extracellular content of the compound, the cell pellet was washed using the phosphate buffered saline (PBS) buffer. The pellet was lazed by adding 200 μL of methanol that had been cooled on ice, and then centrifuged at a speed of 20,000 rpm for 30 min at 4 °C to eliminate the debris. The supernatant analyzed by high-performance liquid chromatography with ultraviolet detection (HPLC-UV) at 394 nm.

Instrument and column setup: use a reversed-phase HPLC column as the common choice (ShimeNex HE C18-EP, 150 mm × 2.1 mm, 5 μm, Shimadzu, Kyoto, Japan). Set the column temperature at 40 °C and injection volume to 10 μL. The mobile phase contained (A) 0.1 % formic acid and (B) acetonitrile. The flow rate is set to 0.4 mL/min. The following gradient was used: 0–2 min, 95% A; 2–10 min, 95% A to 5% A; 10–12 min, 5% A; 12–12.5 min, 5% A to 95% A; 12.5–15 min balance to 95% A. The injection volume was 10 μL. The data were collected by full-wavelength scanning, and the membrane permeability was calculated by comparing the peak area of compounds in blank matrix with that of processed cell supernatant.

2.6. COMT inhibition assay at cellular level

Stably transfected U87-MG cells in logarithmic growth phase were seeded into 96-well plates (Corning, Shanghai, China) at a density of 20,000 per well. When the cells in 96-well plates were about 60% confluent, different concentrations of BA24 and BA were added and incubated with the cells for 30 min under the condition of 37 °C, 5% CO₂. After that, the cells were treated with 3-BTD (5 μM, final concentration) for 30 min, then terminated by adding 100 μL ice-cold acetonitrile. The reaction mixture was centrifuged at 20,000 g for 30 min. The supernatant was subjected to liquid chromatography with fluorescence detection (LC-FD) analysis. Data analysis was performed using GraphPad Prism 8.0.

2.7. 3D QSAR model building

3D quantitative structure-activity relationship (QSAR) (aims to develop a mathematical model by exploiting the 3D descriptors of molecules (i.e., geometric, physical characteristics or quantum chemical descriptors) [38,39] Based on the modelling results we can draw some useful conclusions. 3D QSAR models were built using the corresponding package of Drug Discovery Studio. Discovery Studio 2016 was used to build the partial least squares model, 2 probes for measuring electrostatic potential and steric effect were used for the energy lattice point calculation. We used a series of compounds with known activity values to construct a 3D QSAR model. Firstly, 41 compounds’ information was imported into the software and the experimentally obtained negative logarithm of the half maximal inhibitory concentration (pIC₅₀) values were added to the molecule set. The compounds were divided into a training set and a test set and stacked according to the parent kernel. The test set was imported into the molecule set, and the activity attribute was set to pIC₅₀, and the rest of the values were kept as default values to construct the 3D QSAR model.

2.8. Inhibition kinetic analyses

The inhibition kinetics of compound BA24 and BA against COMT-mediated 3-BTD-O-methylation was studied by performing a set of inhibition kinetics analyses. Briefly, in the presence of multiple concentrations of inhibitors, different concentrations of substrate were used to determine the reaction rate. After that, a Lineweaver-Burk plot was drawn to determine the inhibition pattern of the tested inhibitor, while the second plot of the slopes from the Lineweaver-Burk plot was used to calculate the corresponding inhibition constant (K_i) value. The data were fitted according to the parameters of different inhibition modes in Prism 8.0.

2.9. In vivo pharmacokinetic studies

All Sprague–Dawley (SD) rats were purchased from Shanghai Laboratory Animal Center, Chinese Academy of Sciences, aged 6–8 weeks and weighing 180–220 g. The SD rats were placed in a specific pathogen-free (SPF) environment, light/dark cycle for 12 h, stable temperature of 22 °C, relative humidity of 55%, and free diet. All experimental procedures and animal protocols were approved by the Shanghai University of Traditional Chinese Medicine Animal Research Ethics Committee (Approval No.: PZSHUTCM2403010017).

The test compound BA24 was dissolved in 10% dimethyl sulfoxide, 20% Solutol-15 and 70% 2-hydroxypropyl-β-cyclodextrin for both intravenous and oral administration. Ten SD rats (180–220 g) were randomly divided into 2 groups and fasted for 12 h before the test, with free access to water. The food was uniformly fed 2 h after drug administration. Animals were subjected to an intravenous injection of 2 mg/kg, and a gavage administration of 20 mg/kg, respectively. Blood samples were collected from the jugular vein 0.08, 0.25, 0.50, 1.00, 2.00, 4.00, 6.00, 8.00, and 12.00 h after intravenous administration and 0.08, 0.25, 0.50, 1.00, 2.00, 4.00, 6.00, 8.00, and 12.00 h after oral administration. Plasma samples were prepared by centrifugation of the blood samples. The concentration of the BA24 in each plasma sample was determined by liquid chromatography–tandem mass spectrometry (LC−MS/MS). The pharmacokinetic parameters were calculated by utilizing the WinNonlin software (version 6.1, Certara USA Inc., Cary, NC, USA).

2.10. In vivo pharmacodynamics studies

The SD rats were divided into four groups (n = 5 per group) and subjected to an overnight fast, before starting the experiment. The SD rats were then orally administered with (1) L-DOPA of 20 mg/kg, (2) L-DOPA plus carbidopa of 5 mg/kg, (3) L-DOPA and carbidopaplus plus BA24 of 50 mg/kg, (4) L-DOPA and carbidopa plus BA24 of 100 mg/kg. Different doses of BA24 were administered to rats by gavage, followed by a combination of L-DOPA and carbidopa 10 min later. The amount of L-DOPA and carbidopaplus administered per animal was fixed. Blood samples were collected at 30, 60, 90, and 120 min after administration of L-DOPA/carbidopa. The plasma sample was harvested by centrifuging blood at 4000 g for 10 min after sampling, and then stored at −80 °C until further preparation.

An AB Sciex LC–MS/MS system was used to detect the concentrations of L-DOPA, and 3-OMD in plasma and brain tissue samples. Chromatographic separation was performed using a Hypersil GOLD HILIC column (100 mm × 2.1 mm, 1.9 μm; Thermo Fisher Scientific., Pittsburgh, PA, USA) at 45 °C. Samples were held at 10 °C in an autosampler. The mobile phase contained (A) 0.1% formic acid and (B) acetonitrile. The flow rate was 0.3 mL/min with isocratic elution of 35% A and 65% B. The injection volume was 5 μL. MS/MS was performed using the Triple Quad™ 4500MD mass spectrometer (AB Sciex, Framingham, MA, USA) equipped with an electrospray ionization (ESI) source in the positive multiple reaction monitoring (MRM) scan mode. The dwell time of 100 ms per transition was used to analyze ion pairs at m/z 197.8 → 152.1 (L-DOPA), m/z 212.0 → 195.0 (3-OMD), and m/z 370.1 → 251.9 (lansoprazole, internal standard). The instrument dependent parameters were as follows: curtain gas, 35 psi; collision activated dissociation (CAD) gas, 8 psi; ion spray voltage, 5500 V; ion source temperature, 500 °C, nebulizer gas (GS1), 55 psi; and heater gas (GS2), 60 psi. The Analyst software (version 1.6.3) was used for the data acquisition and processing.

2.11. Acute toxicity study

Ten healthy mice were divided randomly into two groups for assessing the acute toxicity of BA24. One group was administered of BA24 at a dosage of 250 mg/kg, which was dissolved in 10% dimethyl sulfoxide, 20% Solutol-15 and 70% 2-hydroxypropyl-β-cyclodextrin. Following administration, the mice were monitored hourly for the first 6 h and then daily for 14 days for any signs of toxicity. Daily weight measurements were taken, and observations for any toxicity symptoms or mortality were recorded over the 14-day period. On the 14th day, cervical dislocation was used to euthanize the mice. Subsequently, vital organs such as the heart, liver, spleen, lungs, kidneys, brain, and gastrointestinal tract were collected and preserved in 10% buffered in preparation for further histological analysis. Plasma samples were prepared by centrifugation of the blood samples for subsequent biochemical analysis.

2.12. Statistical analysis

All kinetic and inhibition experiments were performed in triplicate and the data obtained from experiments are expressed as mean ± standard deviation. The half maximal inhibitory concentration (IC₅₀) and K_i values were calculated by using GraphPad Prism 8.0 software (GraphPad Software Inc., La Jolla, CA, USA). The statistical distinction between various groups was determined using one-way analysis of variance (ANOVA), with significance levels denoted as ^∗P < 0.05, ^∗∗P < 0.01, ^∗∗∗P < 0.001.

3. Results and discussion

3.1. Discovery of a lead COMT inhibitor from natural flavonoids

Previous studies revealed that hydroxyflavones from TCMs such as Scutellariae radix and vine tea exhibit COMT inhibitory activity, inspiring us to further develop flavonoid-based COMT inhibitors. To this end, 67 naturally occurring flavonoids were collected, and their anti-COMT effects were first assessed using a fluorescence-based biochemical assay. As shown in Fig. S1 and Tables S1, 17 natural flavonoids exhibited significant COMT inhibitory activity with inhibition rates exceeding 50% at a concentration of 10 μM. Among them, BA15 and oroxylin A 28 exhibited the strongest COMT inhibitory activity with inhibition rates exceeding 90%. Additionally, almost all flavonoid glycosides exhibited negligible inhibitory effects on COMT, except for baicalin 16 and Oroxylin A 7-O-β-D-Glu 29. Subsequently, the dose-response curves were then plotted for flavonoids that exhibited less than 50% residual COMT activity at a concentration of 10 μM. As shown in Table S2 and Figs. S2–S4, we were surprised to find that BA and its derivatives, such as scutellarein, oroxylin A, jaceosidin, pectolinarigenin, and hispidulin exhibited potent COMT inhibitory activity, with IC₅₀ values of 165.4 nM, 151.6 nM, 105.8 nM, 92.3 nM, 96.6 nM, 99.6 nM, respectively. Moreover, hydroxylation or methoxylation at the B-ring did not adversely affect the COMT inhibitory activity of these polyphenolic flavonoids. These findings clearly demonstrate that BA derivatives have potent potential for inhibiting COMT. Considering the conditions for structural diversification, BA can serve as a promising lead compound for developing novel non-nitrocatechol COMT inhibitors.

3.2. Optimization strategies for the lead COMT inhibitor

To rationally modify the structure of BA, docking simulations were conducted to determine the preferred binding modes of BA within the catalytic cavity of COMT. As shown in Fig. 1A, BA was effectively positioned within the catalytic cavity of COMT, with ring A deeply penetrating the cavity and forming strong interactions with Glu199 via hydrogen bonding, Pro174 via π-alkyl interactions and Met40 via π-sulfur interactions. The ring B was oriented toward the outside of the binding pocket and created strong interactions with Trp143 via π-π interactions. Furthermore, BA was in contact with Lys144, Asp169, Asp141, Asn170, Lys46, Trp38, and Leu198 by van der Waals interactions. Since the ring B of BA was oriented towards the outer side of the binding pocket, it offers relatively more space for modification without affecting the molecule's ability to bind to COMT. Moreover, the metabolic sites of BA had been identified in our previous research [35]. The phenolic hydroxyl groups of BA are susceptible to glucuronidation by phase II metabolic enzyme UGT, with C-7 OH being the primary metabolic site and C-6 OH the secondary one.

Fig. 1.

Molecular docking of baicalein (BA) with catechol-O-methyltransferase (COMT) and the rationale for structural optimization. (A) The detailed view of the binding of BA in the catalytic cavity of COMT and its interaction, the structure of BA is yellow, Mg²⁺ is represented in rosy, conventional hydrogen bonds are green, Pi-sulfur is shown in orange, Pi-Pi stacking are magenta, Pi-Alkyl are shown in pink. (B) Profile overview of BA and considerations for the optimization strategy.

Taking these into consideration, optimization strategies were formulated with BA as the lead COMT inhibitor (Fig. 1B): (1) For ring B, introducing nitrogen-containing groups or lipophilic groups were employed to enhance solubility and permeability. (2) For the modification of ring A, enhancing metabolic stability can be achieved by directly blocking metabolic sites or by introducing hydrogen bond acceptors near the hydroxyl groups to form intramolecular hydrogen bonds. Following these strategies, a series of BA derivatives were synthesized by modifying the ring A and B of BA, and were subsequently evaluated for their anti-COMT activity and druggability.

3.3. Chemical synthesis of BA derivatives

A series of BA derivatives BA1−BA12 were synthesized via 2′-hydroxychalcone intermediates as shown in Scheme 1. The intermediates 2′-hydroxychalcone (71, 80–85) were synthesized by Friedel-Crafts acylation using the acylated p-nitrocinnamic acid reacts with 3,4,5-trimethoxyphenol or using 2-hydroxyacetophenone and various benzaldehyde under Claisen-Schmidt conditions. Subsequently, flavonoids were generated by oxidative cyclization in the I₂/DMSO system, and then the corresponding BA derivatives were obtained by removing methoxy using BBr₃ solution. Compounds BA13–BA30 were synthesized using BA as shown in Scheme 2. The phenolic hydroxyl group of BA was selectively protected using bromomethyl methyl ether protecting groups to obtain methylated products at different positions. Furthermore, halogen and cyanide groups were introduced at the C-8 position through free radical reactions. The detailed experimental procedures were described in Supporting Information.

Scheme 1.

Synthetic route for the preparation of compounds baicalein (BA)1−BA12. (A) Reagents and conditions: oxalyl chloride, dichloromethane (DCM), 0 °C (i); BF₃·Et₂O, Methylbenzene, 110 °C (ii);. I₂, dimethyl sulfoxide (DMSO), 140 °C (iii); BBr₃, DCM, 0 °C (iv); Na₂S₂O₅, EtOH: H₂O (2:1), 0 °C (v); and NaH, N,N-dimethylformamide (DMF), 0 °C (vi). (B) Reagents and conditions: acetic anhydride, K₂CO₃, tetrahydrofuran (THF), room temperature (RT) (i); BF₃·Et₂O, CHCl₃, 60 °C (ii); NaOH, MeOH, RT (iii); I₂, DMSO, 140 °C (iv); and BBr₃, DCM, 0 °C (v).

Scheme 2.

Synthetic route for the preparation of compounds baicalein (BA)13−BA30. Reagents and conditions: (i) MOMBr (2eq), K₂CO₃, DMF, room temperature (RT), 8 h; (ii) CH₃I (6eq), K₂CO₃, Acetone, RT, 6 h; (iii) idoxane hydrochloride solution, ethanol, RT, 2 h; (iv) MOMBr (1eq), K₂CO₃, DMF, 0 °C, 4 h; (v) CH₃I (16 eq), K₂CO₃, Acetone, RT, 20 h; (vi) HI/CH₃COOH (1:2), 110 °C, 2 h; (vii) a: X = NO₂, HNO₃, DCM, 3 h; b: X = Br, NBS, DMF, 0 °C, 2 h; c: X = Cl, NCS, methylbenzene, 100 °C→ RT, 6 h; (viii) BBr₃, DCM, 0 °C, 12 h.

3.4. Structure-COMT inhibition relationship of BA derivatives

To enhance the water solubility and permeability of BA, nitrogen-containing hydrophilic groups and various halogen substituents were introduced onto ring B, resulting in the synthesis of compounds BA1−BA12. The IC₅₀ values of these compounds were shown in Table 1 and Figs. S5 and S6. Introducing a nitrogenous group (such as pyrrolidine, nitro group or cyano group) at the C-4′ position slightly enhanced the anti-COMT effects (BA4, IC₅₀ = 134.3 nM). However, the introduction of other substitutions, including amines, methylamines, piperidine, and cycloheximide, at this position resulted in a reduction of the anti-COMT effect. Additionally, introducing a halogen atom at the C-4′ position on ring B slightly diminished the anti-COMT activity, whereas the introduction of fluorine groups at both the C-4′ and C-2′ positions enhanced COMT activity (BA8, IC₅₀ = 106.5 nM).

Table 1.

The half maximal inhibitory concentration (IC₅₀) values of ring B-modified baicalein (BA) derivatives against catechol-O-methyltransferase (COMT).

Structure	Compound	R2’	R4’	IC50 (nM)	ClogP	TPSA
	BA	H	H	165.4 ± 18.9	3.00	90.9
	BA1	H	NO2	158.1 ± 37.0	2.74	134.0
	BA2	H	NH2	179.9 ± 29.5	1.78	116.9
	BA3	H	NHCH3	1661 ± 314.8	2.50	102.9
	BA4	H	Pyrrolidine	134.3 ± 21.9	3.28	94.1
	BA5	H	Piperidine	178.5 ± 58.2	3.84	94.1
	BA6	H	Azepane	276.3 ± 32.7	4.40	94.1
	BA7	H	F	282.1 ± 71.9	3.14	90.9
	BA8	F	F	106.5 ± 26.8	3.29	90.9
	BA9	H	CF3	668.2 ± 103.2	3.88	90.9
	BA10	H	Br	204.2 ± 52.0	3.86	90.9
	BA11	H	Cl	318.7 ± 86.8	3.89	90.9
	BA12	H	CN	138.3 ± 44.1	2.43	114.7

ClogP: calculated partition coefficient; TPSA: topological polar surface area.

To improve the metabolic stability of BA, smaller methyl groups were employed to block the hydroxyl metabolic sites, while hydrogen bond acceptors and halogen atoms were introduced at the C-8 position to reduce the metabolism rate, leading to the synthesis of compounds BA13−BA23. The results indicated that methylation of the hydroxyl group at the C-6 position enhanced the anti-COMT effect, whereas methylation at the C-7 or C-5 positions significantly reduced the inhibitory activity (Table 2 and Figs. S7 and S8). Introducing a halogen, nitro, or cyano group at the C-8 position slightly reduced the anti-COMT effect. Meanwhile, the addition of a hydrogen bond acceptor (hydroxypiperazine, dimethylamino-methyl) at the C-8 position significantly decreased the anti-COMT activity.

Table 2.

The half maximal inhibitory concentration (IC₅₀) values of ring A-modified baicalein (BA) derivatives against catechol-O-methyltransferase (COMT).

Structure	Compound	R5	R6	R7	R8	IC50 (nM)	ClogP	TPSA
	BA	OH	OH	OH	H	165.4 ± 18.9	3.00	90.9
	BA13	OMe	OH	OH	H	>10000	2.50	79.9
	BA14	OH	OMe	OH	H	105.8 ± 31.6	3.08	79.9
	BA15	OH	OH	OMe	H	>10000	3.33	79.9
	BA16	OH	OMe	OMe	H	>10000	3.52	68.9
	BA17	OMe	OMe	OMe	H	>10000	2.88	57.9
	BA18	OH	OH	OH	NO2	263.7 ± 37.8	1.83	134.0
	BA19	OH	OH	OH	Br	568.2 ± 141.3	3.73	90.9
	BA20	OH	OH	OH	Cl	185.3 ± 28.6	3.53	90.9
	BA21	OH	OH	OH	CN	396.9 ± 99.8	3.03	114.6
	BA22	OH	OH	OH	CH2N(CH3)2	>10000	2.78	94.1
	BA23	OH	OMe	OH		>10000	1.76	114.4
	BA4	OH	OH	OH	H	134.3 ± 21.9	3.28	94.1
	BA24	OH	OH	OMe	H	72.6 ± 6.8	2.82	83.1
	BA25	OH	OMe	OH	H	84.6 ± 11.0	2.94	83.1
	BA26	OMe	OMe	OMe	H	>10000	3.16	61.1
	BA27	OH	OH	OH	NO2	>10000	3.55	137.3
	BA8	OH	OH	OH	H	106.5 ± 26.8	3.29	90.9
	BA28	OH	OH	OMe	H	306.7 ± 91.1	2.82	79.9
	BA29	OH	OMe	OH	H	65.5 ± 17.6	2.94	79.9
	BA30	OMe	OMe	OMe	H	>10000	3.16	61.1
	BA31	OH	OH	OH	Cl	297.6 ± 38.9	3.81	86.9
	BA32	OH	OMe	OH	Cl	1778 ± 298.5	3.84	79.9
−	Entacapone	Positive COMT inhibitor				83.7 ± 15.4	−	−

ClogP: calculated partition coefficient; TPSA: topological polar surface area; −: no data.

Building on the modifications of rings A and B, two promising compounds, BA4 and BA8, were selected for further optimization, specifically enhancing metabolic stability through methylation on ring A. The results indicated that the methylation products of BA4 (BA24 and BA25), as well as the methylation product of BA8 (BA29), exhibited potent COMT inhibitory activity, with IC₅₀ values of 72.6 nM, 84.6 nM, and 65.5 nM, respectively (Table 2 and Fig. S7 and S8). Their COMT inhibitory activity was comparable to that of the clinically used EN (IC₅₀ = 83.7 nM). Additionally, the appropriate ClogP and topological polar surface area (TPSA) values of these three compounds suggest that they may have good oral bioavailability and membrane permeability.

To elucidate the structure-COMT inhibitory activity relationship of flavonoids, QSAR studies were conducted using Drug Discovery Studio, which has been extensively applied to explore the correlations between biological activities and molecular descriptors of compounds [38]. In our study, 3D QSAR modeling was based on a training set of 32 structurally different compounds with a test set of 8 structurally different compounds with activities (IC₅₀ values) spanning 3 orders of magnitude. The modeling results showed that the experimental activity values of the training set molecules were compared to the predicted values, yielding an R² of 0.9575, where values closer to 1 indicate more accurate predictions. The test set had an R² of 0.9291, indicating that the model was reasonably reliable.

The flavonoids aligned with the isosurfaces of the 3D QSAR model coefficients on van der Waals and electrostatic potential grids were shown in Figs. 2A and B. In the electrostatic map, red contours around regions where high electron density (negative charge) was favorable for COMT inhibition, and blue contours represent areas where low electron density (partial positive charge) was expected to increase activity (Fig. 2A). Introducing an electron-withdrawing group at the C-6,7 positions of the compound is beneficial for enhancing its activity, which is consistent with the experimental result that introducing a methoxy group at this position is beneficial for inhibiting COMT. Introducing an electron-withdrawing group at the C-4′ position of the B ring is also beneficial for enhancing the compound's inhibitory activity. Conversely, introducing an electron-donating group at the C-8 position is beneficial for enhancing the compound. Therefore, the introduction of nitro, halogen, and N-alkyl groups weakens the compound's inhibitory activity against COMT, which is consistent with the experimental results. Similarly, the green-yellow steric contours depicted in Fig. 2B illustrate areas where steric bulk was predicted to increase (green) or decrease (yellow) activity. The introduction of bulky groups at the C-5,6 positions enhances the compound's activity, while the introduction of smaller bulky groups at the C-3,8 positions also favors activity. Additionally, bulky groups at the C-4′ position on the B ring enhance the compound's activity. The resulting mappings suggested that both 3D steric and electronic interactions could strongly affect the inhibitory effects of flavonoids on COMT.

Fig. 2.

Structure-activity relationships (SARs) analysis of flavonoids as catechol-O-methyltransferase (COMT) inhibitors. (A) The flavonoids aligned with the isosurfaces of the 3D qantitative structure-activity relationship (QSAR) model coefficients on van der Waals grids. (B) Electrostatic potential grids. (C) Summary of the SARs of flavonoids as COMT inhibitors.

Combining all the results, the structure-COMT inhibition relationship of flavonoids was summarized in Fig. 2C. Introducing a pyrogallol group on ring A significantly enhances anti-COMT effects. Single methylation of the C6 phenolic group is beneficial for COMT inhibition, whereas methylation at the C5 or C7 positions, or multiple phenolic methylations, leads to a loss of COMT inhibition. Introducing a nitro or halogen group at the C-8 site of ring A slightly reduces COMT inhibition, while introducing N-alkyl groups significantly reduces anti-COMT effects. Introducing hydroxyl, methoxy, or nitrogenous groups at the C4’ site of the ring B did not adversely affect COMT inhibitory activity. Additionally, introducing a hydroxyl group at the C-3 site is detrimental to COMT inhibition. These findings offer new insights into how the steric-electrostatic properties and various substituents of flavonoids affect COMT inhibitory activity, which could aid in the rational design of novel flavonoids as potent COMT inhibitors.

3.5. The structure-metabolism stability relationships of BA derivatives

The phase II metabolic clearance mediated by UGTs is the primary factor contributing to the poor metabolic stability of polyphenolic flavonoids [40,41]. Therefore, we conducted Phase II metabolic stability assays on a series of synthesized BA derivatives with the aim of finding compounds with strong activity and favorable metabolic stability. The half-lives of phase II metabolism of BA derivatives were shown in Table 3. It was found that methylation at the C5−7 position hydroxyl of BA was observed to enhance its metabolic stability. Specifically, methylation at the 7-OH position resulted in the greatest improvement of metabolic stability (BA15, t_1/2 = 36.4 min vs. BA, t_1/2 = 15.7 min; BA24, t_1/2 = 141.5 min vs. BA4, t_1/2 = 26.4 min), which due to the 7-OH is the major metabolic site of BA. It was also found that the introduction of a halogen atom or nitro group at the C-8 position could enhance the metabolic stability, as demonstrated by the half-life values of BA18 (t_1/2 = 73.6 min) and BA20 (t_1/2 = 35.5 min) compared to BA (t_1/2 = 15.7 min), BA27 (t_1/2 = 221.4 min) vs. BA4 (t_1/2 = 26.4 min), BA31 (t_1/2 = 130.4 min) vs. BA8 (t_1/2 = 16.7 min), BA32 (t_1/2 = 225.9 min) vs. BA29 (t_1/2 = 14.8 min). Furthermore, introducing a Mannich base at the C-8 position of BA could strongly improve the metabolic stability, which may be attributed to the formation of an intramolecular hydrogen bond between the C-8 Mannich base substituents and C-7 phenolic groups, which played a predominant role in reducing glucuronidation. Additionally, modifications to the ring B of BA had a relatively small impact on its metabolic stability, likely due to their distance from the metabolic sites.

Table 3.

The half-lives of phase II metabolism of baicalein (BA) derivatives.

No.	t1/2 (min)	No.	t1/2 (min)
BA	15.68 ± 1.26	BA17	44.51 ± 2.09
BA1	18.85 ± 0.54	BA18	73.68 ± 4.03
BA2	12.71 ± 0.93	BA19	11.83 ± 0.29
BA3	49.75 ± 1.75	BA20	35.58 ± 2.11
BA4	26.44 ± 1.01	BA21	19.26 ± 0.69
BA5	25.26 ± 0.98	BA22	243.04 ± 8.46
BA6	21.64 ± 1.22	BA23	90.31 ± 4.62
BA7	16.12 ± 0.48	BA24	141.46 ± 3.30
BA8	16.67 ± 0.66	BA25	27.24 ± 1.22
BA9	20.11 ± 0.86	BA26	32.09 ± 2.31
BA10	16.67 ± 0.29	BA27	221.43 ± 7.10
BA11	17.23 ± 0.33	BA28	19.05 ± 0.50
BA12	25.76 ± 0.91	BA29	14.75 ± 0.29
BA13	21.00 ± 1.06	BA30	34.76 ± 2.76
BA14	29.13 ± 1.53	BA31	130.41 ± 3.94
BA15	36.41 ± 2.80	BA32	225.89 ± 6.90
BA16	37.14 ± 1.92	−	−

t_1/2: metabolic half-life; −: no data.

The integrated analysis of COMT inhibitory activity and in vitro metabolic stability of BA derivatives was depicted in Fig. 3A. Among them, compound BA24 exhibited excellent COMT inhibitory activity and favorable in vitro metabolic stability. Specifically, BA24 demonstrated potent COMT inhibition with an IC₅₀ of 72.6 nM, surpassing the marketed drug EN (Fig. 3B), and exhibited a nine-fold improvement in metabolic stability compared to BA (Fig. 3C).

Fig. 3.

Integrated evaluation of inhibitory activity and metabolic stability. (A) The integrated analysis of catechol-O-methyltransferase (COMT) inhibitory activity and in vitro metabolic stability of baicalein (BA) derivatives. (B) The dose-inhibition curves of entacapone (EN), BA and BA24 against COMT-catalyzed 3-(benzothiazol-2-yl)-7,8-dihydroxy-2H-chromen-2-one (3-BTD)-O-methylation. (C) The in vitro metabolic half-life (t_1/2) of BA and BA24.

3.6. The inhibitory mechanism of BA24 against COMT

To deeply understand the inhibitory mechanism of BA24 against COMT, a set of inhibition kinetic assays were performed. As illustrated in Fig. 4A, increasing concentrations of BA24 did not create significant changes in the maximal velocity of COMT-catalyzed 3-BTD-O-methylation. These data combined with Lineweaver-Burk plot (Fig. 4B) clearly showed that BA24 potently inhibited COMT in a competitive inhibition manner, showing an extremely low K_i value at 89.28 nM. Furthermore, docking simulations were performed to decipher the interaction patterns between BA24 and the COMT crystal structure (PDB ID: 3BWY). As depicted in Fig. 4C, BA24 could tightly bind to COMT by interacting with the amino acid residues surrounding the catalytic cavity, with a binding energy of −7.31 kcal/mol. The C-5 and C-6 phenolic hydroxyl groups of BA24 form strong hydrogen bonding interactions with Glu199, Asn170. The ring A and C ring forms a π-alkyl interaction with Pro174 and Met40, while the ring B and Trp143 form a π-π stack (Figs. 4D and E). Notably, the pyrrolidine at the C-4′ position of BA24 have generated hydrophobic interactions with Trp143 compared to BA. These results indicated that BA24 could be well trapped by the catechol pocket of COMT against 3-BTD-O-methylation in a competitive inhibition manner.

Fig. 4.

Inhibition kinetics and molecular docking analysis of baicalein 24 (BA24) with catechol-O-methyltransferase (COMT). (A) Inhibition kinetic plots of BA24 against COMT-catalyzed 3-(benzothiazol-2-yl)-7,8-dihydroxy-2H-chromen-2-one (3-BTD)-O-methylation. (B) The corresponding Lineweaver-Burk plots of BA24. (C) The overview of BA24 docked with COMT. (D) Binding conformation and the detailed view of BA24 (cyan) and COMT. (E) The 2D interactions between BA24 and the residues surrounding the catalytic cavity of COMT.

3.7. COMT inhibition activity of BA24 in living cells

Since COMT is an intracellular enzyme, its optimal inhibitors should possess the ability to penetrate the cell membrane. Therefore, the membrane permeability of BA24, BA, and EN was assessed in U87-MG cells (Fig. 5A) and NCM-460 cells (Fig. 5B). Notably, after 30 min incubation, a considerable amount of BA24 was observed within cells, whereas the BA and EN demonstrated relatively low membrane permeability. Quantitative analysis revealed that the permeability of BA24 was approximately 10-fold and 18-fold higher than that of BA and EN, respectively. It is worth noting that the literature has confirmed that EN is classified as a Class IV compound in the Bioequivalence Classification System (BCS) [42,43]. This is characterised by low solubility and poor permeability, which is highly consistent with our experimental results, further validating the low permeability phenomenon of EN observed in this study.

Fig. 5.

Cellular membrane permeability and intracellular catechol-O-methyltransferase (COMT) inhibition of baicalein 24 (BA24). (A) Relative cellular uptake of BA24, BA, and entacapone (EN) in U87-MG cells. (B) Relative cellular uptake of BA24, BA, and EN in NCM-460 cells. (C) Fluorescence images of U87-MG cells showing intracellular COMT activity after treatment with BA24 at different doses. (D) Dose-response curve of intracellular COMT inhibition by BA24 derived from fluorescence data. (E) Fluorescence images of U87-MG cells showing intracellular COMT activity after treatment with BA at different doses. (F) Dose-response curve of intracellular COMT inhibition by BA derived from fluorescence data. Data are presented as mean ± standard deviation (n = 3). ^∗∗∗P < 0.001.

Inspired by these encouraging findings, the inhibitory effects of BA24 on intracellular COMT were further investigated. CCK-8 assays revealed that BA24 exhibited negligible cytotoxicity and favorable biocompatibility in U87-MG cells (Fig. S9). Subsequently, a COMT-specific fluorogenic substrate (3-BTD) was utilized to visualize the real-time activity of intracellular COMT in a non-invasive manner. The dose-dependent attenuation in green fluorescence derived from the COMT-mediated 3-BTD methylation was pronounced with increasing BA24 concentrations (Fig. 5C). This effect was attributed to the potent inhibition of intracellular COMT by BA24, with an IC₅₀ value at the cellular level as low as 0.86 μM (Fig. 5D). Meanwhile, the inhibitory effects of BA on intracellular COMT were also tested under identical conditions (Fig. 5E). BA could also dose-dependently inhibit intracellular COMT in U87-MG cells, with an IC₅₀ value of 22.65 μM (Fig. 5F). These results clearly demonstrated that BA24 exhibited a significantly more potent anti-COMT effect than BA (approximately 26-fold enhancement) in living cells, which was primarily attributed to its superior cell membrane permeability.

3.8. Pharmacokinetics of BA24 by oral and intravenous administration in rats

Having evaluated the effect of BA24 on the anti-COMT activity, metabolic stability and membrane permeability, the in vivo pharmacokinetics of BA24 were further investigated after intravenous (iv, 2 mg/kg, n = 5) and oral administration (p.o, 20 mg/kg, n = 5) in rats to demonstrate its suitability for extensive in vivo. As shown in Table S3, the area under the concentration-time curve (AUC_{0–12 h}) of compound BA24 is about 682.27 ng/mL·h. The peak plasma concentration was 248.37 ng/mL at 1.3 h. Besides, the mean residence time (MRT) and t_1/2 were 5.40 and 3.14 h, respectively. The oral bioavailability of BA24 was approximately 10-fold higher than that of the reported BA [44].

3.9. BA24 improved L-DOPA metabolism in rats with a favorable safety profile

To validate the inhibitory effect of BA24 on COMT in vivo, the concentrations of L-DOPA and 3-OMD in rat plasma were measured. The protocol of the animal experimental was illustrated in Fig. 6A. When L-DOPA was administered alone, its concentration was very low due to its rapid metabolism and clearance in the peripheral system by dopa decarboxylase (DDC) and COMT. After the addition of carbidopa (a DDC inhibitor), the plasma concentration of L-DOPA increased approximately 3.2 fold, indicating that carbidopa prevented the peripheral conversion of L-DOPA to dopamine by inhibiting DDC. With the further addition of BA24 (100 mg/kg), the plasma L-DOPA level increased significantly in a dose-dependent manner. The peak plasma concentration (c_max) and the area under the curve (AUC_{0–120 min}) of L-DOPA increased by approximately 4.3 fold and 3.5 fold, respectively, compared to the group receiving L-DOPA and carbidopa (Fig. 6B). Notably, 3-OMD, a COMT metabolite of L-DOPA, was significantly elevated in the plasma when L-DOPA was co-administered with carbidopa, as the inhibition of DDC metabolism promoted COMT-mediated metabolism to become the primary pathway for L-DOPA. With the further addition of BA24 treatment, the c_max and AUC_{0–120 min} of 3-OMD levels were markedly reduced by 4.6-fold and 3.6-fold, respectively (Fig. 6C). These results clearly demonstrated that BA24 significantly inhibited COMT-mediated methylation of L-DOPA in vivo, showing comparable activity to current clinically used COMT inhibitors [[45], [46], [47], [48]].

Fig. 6.

In vivo pharmacodynamic evaluation and safety profile of baicalein 24 (BA24). (A) The animal experimental protocols for pharmacodynamic study of BA24 in rats. (B) Plasma levodopa (L-DOPA) levels over time under various treatments. (C) Plasma 3-OMD levels over time under various treatments (n = 5). (D) Body weight changes of mice after oral administration of 250 mg/kg BA24, the control group was administered only the 0.5% sodium carboxymethyl cellulose (n = 8). (E) Histopathological studies of the mouse organs following oral administration of BA24 (250 mg/kg) for 14 consecutive days. (F) The serum biochemical indicators in control mice and BA24-treated mice. Statistical differences are indicated by P values (one-way analysis of variance (ANOVA)), ^∗P < 0.05, ^∗∗P < 0.01, ^∗∗∗P < 0.001 vs. L-DOPA + carbidopa group.

Finally, the safety profile of BA24 was evaluated in mice following oral administration of 250 mg/kg/day for 14 days. Throughout the study, no significant differences in body weight were observed between the BA24-treated group and the control group (Fig. 6D). Hematoxylin and eosin (H&E) staining results showed no apparent injuries in the heart, liver, spleen, lungs, kidneys, brain, or gastrointestinal tissues in either group (Fig. 6E). Additionally, serum biochemical markers, including liver function indicators (albumin (ALB), alkaline phosphatase (ALP), alanine aminotransferase (ALT), aspartate aminotransferase (AST), total protein (TP)), renal function markers (creatinine (CREA), urea (UREA)), blood glucose and lipid levels (total cholesterol (TC), triglycerides (TG), glucose (GLU)), and myocardial injury markers (creatine kinase (CK)), showed no significant differences between the BA24-treated and control groups (Fig. 6F). These findings collectively indicate that BA24 possesses a favorable safety profile. This characteristic is attributed to the fact that our derivative retains the core flavonoid structure while optimizing functional groups, avoiding structural motifs (e.g., certain electrophilic moieties or reactive metabolites) reported to be associated with hepatotoxicity or off-target interactions in some marketed COMT inhibitors [49]. Furthermore, the structural similarity between BA24 and endogenously occurring polyphenols may also contribute to a more favorable metabolic fate, reducing the likelihood of toxic intermediate accumulation. Taken together, these results suggest that BA24 effectively increases the peripheral concentration of L-DOPA by inhibiting COMT-mediated metabolism in vivo, while demonstrating a favorable safety profile, potentially enhancing its therapeutic efficacy in treating diseases like PD.

4. Conclusion

In summary, a fluorescence-based visualization inhibitor screening method was employed to screen, characterize and validate the bioactive properties of COMT inhibitor in vitro and in living cells. Following the preliminary screening of natural flavonoids against COMT, BA emerged as a promising lead compound. Based on docking simulations that revealed the interaction patterns between BA and COMT, as well as BA's druggability limitations, a multi-dimensional structural optimization strategy was developed, leading to the design and synthesis of 32 BA derivatives. Among them, compound BA24 exhibited a 26-fold increase in cellular COMT inhibition and approximately 10-fold enhancements in metabolic stability, membrane permeability and oral bioavailability, respectively, compared to BA, achieving a simultaneous improvement in both the activity and druggability of BA. Mechanistically, BA24 competitively inhibited COMT by binding to the catechol pocket with a K_i of 89.28 nM. Furthermore, BA24 exhibited favorable safety profiles and significantly modulated L-DOPA metabolism in rats. These findings suggested that BA24 could serve as a promising non-nitrocatechol COMT inhibitor with favorable safety profiles and potent anti-COMT effects both in vitro and in vivo, making it a potential candidate for the treatment of PD. Additionally, the relationships between structural properties, inhibitory activity, and metabolic stability of flavonoids as COMT inhibitors were discussed and summarized, which providing valuable insights for the optimization of flavonoids as lead compounds for drug development.

CRediT authorship contribution statement

Rong Zhu: Writing – original draft, Methodology, Data curation. Pu Wang: Validation, Methodology. Sheng-Lan Qi: Validation, Methodology, Data curation. Yu Zhang: Investigation, Funding acquisition, Data curation. Chun-Lan Xie: Validation, Investigation. Ya Yang: Methodology, Data curation. Cong Hu: Validation, Methodology. Zi-Qiong Zhou: Validation, Methodology. Dong-Fang Zhao: Methodology, Investigation. Chao Yang: Methodology, Investigation. Jie Sun: Methodology, Data curation. Xian-Wen Yang: Writing – review & editing, Resources, Conceptualization. Guang-Bo Ge: Writing – review & editing, Supervision, Funding acquisition. Ping Wang: Writing – review & editing, Investigation, Funding acquisition, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following is the Supplementary data to this article.