Identification of Quality Markers in Viola kunawaresis Royel Using HPLC Fingerprints, Chemometric Analysis, and Network Pharmacology

TL;DR

This study identifies key chemical markers in a medicinal plant used in Uyghur medicine to improve quality control and understand its antiasthmatic effects.

Contribution

The study introduces a novel approach combining HPLC, chemometrics, and network pharmacology to identify quality markers in VTH.

Findings

Six representative quality markers were identified for VTH using HPLC and chemometric analysis.

Network pharmacology and molecular docking revealed potential antiasthmatic effects of the identified markers.

The study provides a foundation for the chemical basis and mechanism of action of VTH.

Abstract

Violae tianshanicae herba (VTH), a widely used crude drug in Uyghur medicine in China, is renowned for its immunomodulatory, anti‐inflammatory, antibacterial, antiviral, and antidiabetic activities, with notable efficacy in treating asthma. However, systematic research on its quality markers (Q‐markers) remains limited, underscoring the need to analyze its chemical composition to enable effective quality control. A high‐performance liquid chromatography (HPLC) fingerprint method was developed for 21 VTH batches in order to identify the common peaks of 23 flavonoids and esculetin, which were verified using liquid chromatography‐mass spectrometry (LC–MS) and reference standards. In this study, chemometric methods were employed to differentiate samples from four geographical origins and to screen for differential components. In addition, a thorough examination of virtual target prediction in network pharmacology, in conjunction with molecular docking validation, has identified Q‐markers with potential antiasthmatic effects. A group of six representative Q‐markers (including kaempferol 3‐O‐sophorosyl‐7‐O‐glucoside, esculetin, kaempferol 3‐O‐rutinosyl‐7‐O‐glucoside, nicotiflorin, narcissoside, and astragalin) was selected for the purpose of developing a method to quantitatively analyze these components simultaneously in VTH samples. This study provides important references for the comprehensive quality control of VTH and establishes the foundation for researchers to explore the chemical basis and mechanisms of action of VTH.

Figures

Comparison of HPLC chromatogram using different extraction solvents. (A) 60% ethanol; (B) 80% ethanol; (C) 100% ethanol; (D) 60% methanol; (E) 80% methanol; (F) 100% methanol.

Comparison of HPLC chromatograms with different mobile phases. (A) 0.2% formic acid–acetonitrile system; (B) 0.1% formic acid–acetonitrile system; (C) 0.2% phosphoric acid–acetonitrile system; (D) water–acetonitrile system.

Comparison of HPLC chromatogram with different chromatographic columns. (A) Shim‐pack Scepter C18 (4.6 × 150 mm, 5 μm); (B) Agilent TC‐C18 (4.6 × 250 mm, 5 μm); (C) Shim‐pack GIST C18 (4.6 × 150 mm, 5 μm); (D) InertSustain AQ‐C18 (4.6 × 150 mm, 5 μm); (E) Capcell Pak C18‐MGII (4.6 × 250 mm, 5 μm).

Comparison of HPLC chromatogram at different detection wavelengths. (A) 366 nm; (B) 342 nm; (C) 300 nm; (D) 272 nm; (E) 254 nm.

Superimposed fingerprint chromatograms of 21 VTH batches from different origins.

HPLC chromatogram of mixed standard solution (A) and VTH sample (B).

Hierarchical cluster analysis (HCA) of 21 batches of VTH.

Chemometrics analysis diagram. Principal component analysis (PCA) 2D (A) and score 3D (B) plots of 21 VTH batches; (C) score plot of orthogonal partial least squares‐discriminant analysis (OPLS‐DA) for 21 VTH batches; (D) variable importance in projection (VIP) values from the OPLS‐DA model.

Network pharmacology graph. (A) Key targets of VTH in asthma treatment; (B) active compound–target network diagram; (C) PPI network diagram; (D) GO enrichment analysis; (E) KEGG signaling pathway map; (F) construction of network diagram for mechanism prediction of VTH in treating asthma.

Heat map of molecular docking scores for VTH components and asthma‐related targets.

Molecular docking visualization between VTH active ingredients and key targets. (A) Compound 4–TNF; (B) Compound 10–AKT1; (C) Compound Q–PTGS2; (D) Compound 20–NF‐κB1; (E) Compound 21–GSK3β; (F) Compound 21–EGFR.

HPLC chromatogram of quantitative determination for VTH.

Stacked bar chart of the quantification results for major constituents in VTH.

Tables

TABLE 1. Information on VTH samples.

No.	Place of origin	Latitude and longitude	Harvest time
S1	Qira County, China	E 80°32′03″, N 36°17′28″	July 10, 2023
S2	Qira County, China	E 81°51′57″, N 36°14′15″	July 10, 2023
S3	Qira County, China	E 81°9′36″, N 36°10′09″	July 15, 2023
S4	Qira County, China	E 82°24′41″, N 36°9′25″	July 15, 2023
S5	Keriya County, China	E 81°25′48″, N 36°17′50″	July 10, 2023
S6	Tashkurgan County, China	E 75°27′14″, N 37°25′54″	July 24, 2023
S7	Qira County, China	E 81°49′34″, N 36°12′32″	July 22, 2023
S8	Qira County, China	E 81°1′2″, N 36°10′57″	July 12, 2023
S9	Hotan County, China	E 79°46′23″, N 36°46′53″	July 28, 2023
S10	Hotan County, China	E 79°57′50″, N 36°44′45″	July 28, 2023
S11	Hotan County, China	E 79°37′43″, N 36°51′39″	July 28, 2023
S12	Keriya County, China	E 81°45′09″, N 36°22′54″	July 15, 2023
S13	Keriya County, China	E 81°53′26″, N 36°22′15″	July 15, 2023
S14	Keriya County, China	E 82°18′16″, N 36°34′08″	July 15, 2023
S15	Hotan County, China	E 80°01′10″, N 36°51′39″	May 23, 2023
S16	Hotan County, China	E 79°52′01″, N 37°2′19″	June 8, 2023
S17	Hotan County, China	E 79°43′27″, N 36°07′15″	April 13, 2024
S18	Hotan County, China	E 80°17′11″, N 37°11′59″	May 14, 2024
S19	Hotan County, China	E 79°53′12″, N 36°13′36″	May 23, 2024
S20	Hotan County, China	E 79°38′42″, N 37°15′12″	June 4, 2024
S21	Hotan County, China	E 79°44′43″, N 36°59′31″	June 5, 2024

TABLE 2. Molecular docking score (kcal/mol).

Compound	TNF	AKT1	PTGS2	HSP90AA1	NF‐κB1	GSK3β	EGFR
1	−8.0	−8.5	−9.9	−8.2	−7.7	−8.7	−10.2
2	−7.2	−9.2	−7.9	−8.5	−8.3	−7.7	−8.7
3	−7.8	−8.5	−8.8	−7.6	−7.9	−9.0	−8.8
4	−7.9	−8.0	−9.0	−7.8	−8.7	−8.5	−8.9
5	−7.3	−8.9	−8.2	−7.4	−7.7	−7.7	−7.7
6	−8.2	−7.7	−8.1	−8.2	−8.1	−8.9	−10.0
7	−7.3	−7.8	−7.1	−7.0	−7.0	−7.0	−8.4
8	−7.3	−8.2	−7.9	−8.1	−7.6	−7.7	−9.2
9	−7.5	−7.9	−7.4	−7.0	−7.4	−6.8	−8.2
10	−7.3	−8.2	−8.8	−7.1	−8.1	−8.3	−7.7
11	−7.7	−7.7	−7.5	−6.7	−7.3	−7.2	−7.5
12	−7.1	−9.6	−9.6	−7.8	−8.0	−7.8	−10.2
13	−7.3	−8.0	−7.8	−7.0	−7.4	−7.4	−7.3
14	−6.0	−7.9	−7.3	−6.3	−6.9	−6.6	−8.5
15	−7.4	−8.1	−7.6	−7.6	−7.8	−7.7	−7.6
16	−6.2	−7.3	−7.9	−6.3	−7.0	−6.9	−7.2
17	−6.4	−7.1	−7.3	−8.3	−6.8	−7.1	−8.0
18	−6.5	−7.1	−7.8	−8.7	−7.3	−7.1	−7.3
19	−6.5	−7.0	−7.8	−8.0	−7.1	−6.8	−6.2
20	−6.6	−7.8	−6.9	−6.6	−7.2	−6.9	−6.5
21	−6.4	−6.1	−7.3	−7.7	−6.9	−6.0	−8.1
22	−7.0	−7.6	−7.4	−8.4	−7.2	−7.1	−8.9
23	−6.1	−7.4	−6.4	−6.9	−6.2	−6.2	−7.0
Q	−4.5	−4.4	−4.3	−5.0	−4.5	−4.4	−5.0

TABLE 3. Regression equation, correlation coefficient (r), and linear ranges of the standard curves.

Compound	Regression equation	r	Linea ranges/(ng/mL)	LOD	LOQ
				(ng/mL)	(ng/mL)
Kaempferol 3‐O‐sophorosyl‐7‐O‐glucoside	Y = 7.74X − 2334.2	0.9999	5260 ~ 266,390	90	260
Esculetin	Y = 48.7X − 2165.9	0.9999	100 ~ 10,010	20	60
Kaempferol 3‐O‐rutinosyl‐7‐O‐glucoside	Y = 11.8X − 2647.5	0.9999	4070 ~ 199,630	40	110
Nicotiflorin	Y = 23.2X − 791.2	0.9999	870 ~ 45,680	40	130
Narcissoside	Y = 23.5X − 294.6	0.9999	900 ~ 46,660	40	110
Astragalin	Y = 31.3X − 1952.1	0.9999	330 ~ 20,250	30	90

TABLE 4. Quantitative results of major components in VTH (ng/g).

No.	Kaempferol 3‐O‐sophorosyl‐7‐O‐glucoside	Esculetin	Kaempferol 3‐O‐rutinosyl‐7‐O‐glucoside	Nicotiflorin	Narcissoside	Astragalin
S1	1460	220	638	285	208	231
S2	4128	710	2418	503	531	106
S3	1669	250	837	337	249	070
S4	4160	710	2671	539	573	101
S5	1468	220	827	315	237	211
S6	2317	740	1148	334	302	144
S7	2106	510	1380	629	528	202
S8	3697	760	1788	794	542	273
S9	2558	430	1013	430	352	197
S10	3042	510	1269	497	356	188
S11	2598	450	1264	436	314	165
S12	2623	470	1150	482	337	192
S13	2581	470	1228	490	355	193
S14	2504	440	1155	463	328	188
S15	3231	490	1068	605	298	203
S16	2863	520	1247	646	353	183
S17	3230	560	1258	633	323	200
S18	2736	1084	1109	542	275	172
S19	2450	1435	0992	479	264	161
S20	3592	670	1430	682	368	215
S21	3373	450	1952	816	392	166

Full text

Introduction

1

Violae tianshanicae herba, the dried whole plant of Viola kunawarensis Royel, is a widely used herbal remedy in Uyghur medicine in China. The plant is a perennial herb of the family Violaceae and is endemic to subalpine meadows and shrublands situated at elevations in excess of 2900 m within regions such as the Himalayas, the Pamirs, and the Tianshan Mountains (Marcussen et al. 2022; Zhu et al. 2016). VTH is a monarch medicine in a novel Uyghur medicine formulation “Compound Binafuxi Granules,” which was approved by the China National Medical Products Administration in 2025. According to Uyghur medical theory, VTH is believed to expel abnormal black bile and exert diverse therapeutic effects, including detoxification and relief from cough. Contemporary pharmacological studies have demonstrated that VTH exhibits immunomodulatory, anti‐inflammatory, antibacterial, antiviral, and antitussive properties (da Silva, Sganzerla, et al. 2023; Dong and Lin 2021; Qin et al. 2015; Wang et al. 2017). A phytochemical analysis of VTH has led to the identification of various chemical constituents, including flavonoids, phenylpropanoids, tetracyclic and pentacyclic triterpenoids, alkaloids, megastigmanes, and cyclic peptides. Flavonoids and phenylpropanoids are recognized as the primary active components of the plant (Chen and Aisa 2017; Fernández‐Bobey et al. 2023; Qin et al. 2013; Yamamoto et al. 2008). In order to ensure the consistency and efficacy of VTH‐based medicinal preparations, it is imperative to develop an HPLC fingerprint method. This approach will facilitate a comprehensive evaluation of the VTH quality, thereby enabling the identification of Q‐markers. Such identification will establish a correlation between the chemical constituents and their pharmacological effects, thus laying the foundation for a robust quality assessment system for this traditional medicinal herb.

Network pharmacology is an emerging interdisciplinary field that integrates systems biology, omics technologies, and network analysis to elucidate drug actions within complex biological systems (Li et al. 2023; Zhai et al. 2025). The utilization of artificial intelligence and big data analytics in this approach enables the identification of active compounds in traditional medicines and the elucidation of their potential mechanisms of action (Nogales et al. 2022; Zhang et al. 2023). The methodological framework of network pharmacology is typicaly involves three key stages: first, the mining of disease‐associated biological systems; second, the identification of potential drug targets; and third, the network‐based navigation of compound–target interactions. This facilitates the systematic prediction of relationships between diseases or syndromes and relevant gene/protein targets, as well as the interactions between chemical constituents and their pharmacodynamic targets (Zhang et al. 2025).

Traditional medicine is inherently characterized by multicomponent, multitarget, and holistic therapeutic strategies (Li et al. 2024). Network pharmacology is therefore of pivotal importance in the identification of Q‐markers, defined as bioactive constituents that are both chemically measurable and pharmacologically relevant (Mishra et al. 2023; Zhao et al. 2023). By constructing comprehensive multidimensional networks, this approach enables a systematic understanding of synergistic effects and the underlying regulatory mechanisms (Qin et al. 2024; Xuan et al. 2020). Unlike conventional reductionist methods that are oriented toward the investigation of individual components, network pharmacology accentuates critical pathways and pivotal molecular targets. Furthermore, the integration of molecular docking provides a means to validate predicted interactions at the atomic level, thereby enhancing the reliability of Q‐markers identification (Lyu et al. 2022; Xia et al. 2025). In consideration of the intricate and multifaceted phytochemical composition of VTH, which is distinguished by a wide array of compound types, fluctuating concentrations, and synergistic interactions, conventional analytical techniques frequently prove to be inadequate for the purpose of correlating specific constituents with biological activity. Therefore, a network pharmacology‐based approach is essential for the systematic exploration of VTH pharmacological mechanisms and the identification of reliable Q‐markers for its quality control and clinical application.

Materials and Methods

2

Plant Materials, Chemicals, and Reagents

2.1

Samples of V. kunawarensis were collected in Xinjiang, China, during the years 2023 and 2024. The samples were authenticated by the authors as the dried whole plant of V. kunawarensis Royle (VTH), classified within the genus Viola of the family Violaceae. A voucher specimen (No. 20230715TSJC) has been deposited at the Xinjiang Uyghur Autonomous Region Institute for Drug Control, China. Detailed information is presented in Table 1.

Esculetin (batch number 110741‐201708, with a purity of ≥ 99.9%), Kaempferol 3‐O‐glucoside (batch number 112007‐201602, with a purity of ≥ 90.8%), and isorhamnetin 3‐O‐rutinoside (batch number 111997‐201501, with a purity of ≥ 93.1%) were purchased from the National Institute for Food and Drug Control, China. Kaempferol 3‐O‐rhamnosyl‐7‐O‐glucoside (batch number PS013100, with a purity of ≥ 98.0%) was obtained from Chengdu Lemeitian Pharmaceutical Technology Co. Ltd. Astragaloside (batch number F18HB175834, with a purity of ≥ 98.0%) was sourced from Shanghai Yuanye Biotechnology Co. Ltd. Additionally, Kaempferol 3‐O‐rutinosyl‐7‐O‐glucoside (batch number PS011797, with a purity of ≥ 99.67%) was acquired from Shanghai Life Sciences Co. Ltd. HPLC‐grade solvents ethanol (batch number F24O3M202, GR) and acetonitrile (batch number F24O1O201, GR) were purchased from Thermo Fisher Scientific, USA. Other reference substances were isolated within the laboratory, and the remaining reagents utilized are of analytical grade.

Instruments

2.2

Chromatographic analysis was performed using Shimadzu Prominence LC‐20AT LC systems (with SPD‐M10A VP photodiode array detector, SIL‐20A autosampler, and CTO‐10AS VP column oven) and Nexera LC‐40D XR HPLC systems (Shimadzu Corporation, Japan), equipped with SIL‐40 autosampler, CTO‐40C column oven, and SPD‐M40 photodiode array detector, and operated via the LabSolutions data processing system. Mass spectrometric analysis was carried out using an API 4000 LC/MS–MS system (AB Sciex, USA) with an electrospray ionization (ESI) source, controlled via the Analyst software. Additional equipment included an MS204S electronic balance (Mettler Toledo, Switzerland), CPA225D electronic balance (Sartorius, Germany), SK8210LHC ultrasonic cleaner (Shanghai Keda Ultrasonic Instrument Co. Ltd., China), and FJY1002UVE ultrapure water system (Qingdao Fullem Technology Co. Ltd., China).

Conditions of Mass Spectrometry

2.3

Qualitative detection was performed using the ESI source in both positive and negative ion modes. Source‐dependent parameters were optimized as follows: nebulizer gas flow of 30 psi, curtain gas flow of 25 psi, ion spray voltage of 4.5 kV (positive ion mode) and 3.5 kV (negative ion mode), and temperature set at 350°C. Compound‐dependent parameters were optimized during tuning and included the declustering potential (DP, 40 eV), focusing potential (FP, 40 eV), entrance potential (EP, 15 eV), collision energy (ce, 12 eV), and collision cell exit potential (CXP, 12 eV).

Preparation of the Solutions

2.4

Preparation of Mixed Standard Solution

2.4.1

Standard substances were accurately weighed and dissolved in methanol to prepare a stock solution containing: 400 μg/mL of kaempferol 3‐O‐sophoroside 7‐O‐glucoside, 200 μg/mL of esculetin, 800 μg/mL of kaempferol 3‐O‐rutinosyl‐7‐O‐glucoside, 500 μg/mL of nictoflorin, 500 μg/mL of narcissoside, and 400 μg/mL of astragaloside. The stock solution was diluted to obtain working solutions with respective concentrations of 270, 10, 200, 50, 50, and 20 μg/mL. All solutions were stored at 4°C prior to use.

Preparation of Sample Solution

2.4.2

The VTH sample (S) was ground into a fine powder and sieved through a 40‐mesh screen. The powder (1.00 g) was placed in an Erlenmeyer flask with 20 mL of methanol, establishing a solid‐to‐liquid ratio of 1:20 (g/v). The mixture was allowed to stand for 30 min, followed by ultrasonic extraction for 30 min at room temperature. After the extraction process, the solution was cooled and adjusted for volume loss with methanol. The resulting mixture was then homogenized and filtered through a 0.22‐μm organic microporous membrane to yield the final test solution.

Method of HPLC Fingerprints

2.5

HPLC Chromatographic Conditions

2.5.1

HPLC analysis was performed on a Shim‐pack Scepter C18 column (4.6 × 250 mm, 5 μm) maintained at 30°C and an injection volume of 10 μL. The mobile phase consists of acetonitrile (A) and 0.1% formic acid aqueous solution (B) with a gradient elution as follows: 0–3 min, 2% A; 3–8 min, 2%–6% A; 8–14 min, 6%–8% A; 14–16 min, 8%–10% A; 16–39 min, 10%–12% A; 39–59 min, 12%–17% A; and 59–100 min, 21%–24% A. The detection wavelength was 342 nm, and the flow rate was 0.8 mL/min.

Method Validation

2.5.2

Precision was assessed using six replicate injections of Sample 3. Stability was evaluated at intervals of 0, 2, 4, 8, 12, 16, 20, and 24 h. Repeatability was tested with six independently prepared solutions derived from the same sample. The relative retention time and the relative standard deviations (RSDs) of peak area for common peaks were determined using nicotiflorin as the internal reference.

Similarity Analysis

2.5.3

A total of 21 batches of VTH from different sources were analyzed by HPLC. Fingerprints were imported into the Chinese Herbal Medicine Chromatographic Fingerprint Similarity Evaluation System (2012 edition). A reference fingerprint was generated via the median method. Similarity values between each batch and the reference were computed. Characteristic peaks were identified using standard references (Feng et al. 2023).

Peak Identification

2.5.4

The selected test samples were subjected to liquid chromatography mass spectrometry (LC–MS) analysis. The peaks in the fingerprint chromatograms were identified by comparing retention times, ultraviolet (UV) spectra, and mass spectrometry (MS) fragmentation patterns with those of reference standards, allowing identification of flavonoids and coumarins.

Network Pharmacology Analysis

2.6

Based on reports of in vivo and in vitro activities of common components in fingerprint chromatogram and queries from the traditional Chinese medicine system pharmacology (TCMSP) database (Campos‐Vidal et al. 2022; Chen et al. 2023; Choi et al. 2019; El Gendy et al. 2025; Fang et al. 2022; Guijarro‐Díez et al. 2015; Ha et al. 2022; Hai‐Ming et al. 2014; Lee et al. 2011; Lee et al. 2023; Li et al. 2006; Li et al. 2021; Manach et al. 1997; Mohamed et al. 2024; Nugraha et al. 2015; Saybel et al. 2025; Siewek et al. 1985; Silva, Azevedo, et al. 2023; Tao et al. 2022; Yokozawa et al. 2002; Zahoránszky‐Khalmi et al. 2020), active components of VTH were screened according to the ADME attribute values (oral bioavailability [OB] ≥ 30% and drug‐likeness [DL] ≥ 0.18). The chemical structures of these compounds were drawn in ChemDraw 20.0 and converted to SMILES format using NovoPro online tools. Target prediction was performed using the TCMSP, PharmMapper, SwissTargetPrediction, SEA, and SuperPred databases (Wang et al. 2025). Additionally, antiasthmatic targets were retrieved using the keyword “bronchial asthma” from GeneCards, DisGeNET, and OMIM databases, with the biological species set as “ Homo sapiens ” (Xinhe Wang et al. 2024).

Common targets between compounds and asthma were identified using Venny 2.1.0. These targets were submitted to the STRING database (with a threshold > 0.40 and the species as H. sapiens ) to construct a PPI network. The resulting data, formatted in TSV, were analyzed using Cytoscape 3.9.1, where degree centrality (DC), closeness centrality (CC), and betweenness centrality (bc) were calculated to identify key targets (B. Zhao et al. 2024).

GO and KEGG analyses were conducted via the DAVID database. Targets from the PPI network were submitted, with a significance threshold set at p < 0.05. A comprehensive summary of the biological processes (BPs) and enriched pathways related to the antiasthmatic effects of VTH was systematically compiled.

Molecular Docking Analysis

2.7

Molecular docking was performed using Molecular Operating Environment (MOE) software version 2019.0102. The selection of key targets was based on their degree values, and these targets were sourced from the UniProt and Protein Data Bank (PDB), with a resolution threshold of less than 3 Å. Ligands that were cocrystallized in the structures served as positive controls. The preparation of proteins included the elimination of water molecules, hydrogenation, and active site identification. The structures of candidate compounds were optimized prior to docking. The conformation with the lowest docking score was considered optimal for subsequent interaction analysis (Sun et al. 2024).

Quantitative Analysis of Q‐Markers

2.8

Quantification was carried out using an InertSustain AQ‐C18 column (250 × 4.6 mm, 5 μm) maintained at 30°C and an injection volume of 10 μL. The mobile phase consisted of acetonitrile (A) and 0.1% formic acid (B) with the following gradient: 0–2 min, 9% A; 2–25 min, 9%–12% A; 25–45 min, 12%–17% A; 45–80 min, 17%–21% A; 80–85 min, 21% A. The detection wavelength was 350 nm, and the flow rate was 0.8 mL/min. In this context, Peak 4 exhibits a resolution greater than 1.25, with a peak purity of 98.76% and an overlap of 0.62% with the adjacent Peak 3, demonstrating significant analytical relevance in the quality control and analysis of traditional Chinese medicinal materials.

Mixed standard solutions of 100 μL, 500 μL, 1 mL, 2 mL, 3 mL, 4 mL, and 5 mL were prepared in a 5‐mL volumetric flask, followed by the addition of methanol. These solutions were then injected into the HPLC system for analysis. A standard curve was established by plotting the concentration of the injected samples on the X‐axis against the corresponding peak area on the Y‐axis, from which a regression equation was derived. In addition, six parallel preparations of VTH samples were conducted, with each sample being injected consecutively six times. The retention times and RSD values of the peak areas for kaempferol 3‐O‐sophorosyl‐7‐O‐glucoside, esculetin, kaempferol 3‐O‐rutinosyl‐7‐O‐glucoside, nicotiflorin, narcissoside, and astragalin were calculated to evaluate precision. The samples were analyzed at intervals of 0, 2, 4, 8, 12, 16, 24, and 48 h to assess their stability over a 48‐h period. Furthermore, repeatability was evaluated through continuous injections of six samples from the same batch.

Sample 3 of VTH was ground and divided into nine aliquots, each weighing 0.5 g. A mixed solution containing three different concentrations of reference compounds was added to each aliquot, and the samples were prepared following established protocols. The analysis was conducted using chromatographic techniques, and the resulting chromatograms were documented. The peak areas obtained were utilized in regression equations to ascertain the concentrations of kaempferol 3‐O‐sophorosyl‐7‐O‐glucoside, esculetin, kaempferol 3‐O‐rutinosyl‐7‐O‐glucoside, nicotiflorin, narcissoside, and astragalin, thereby enabling the evaluation of recovery rates for each individual component.

Results

3

Method Validation for Fingerprint Analysis

3.1

To acquire chromatographic fingerprints with rich, high‐resolution, and comprehensive chemical data, a comparative analysis was carried out using an array of mobile phases, encompassing 0.1% phosphoric acid–acetonitrile, 0.1% and 0.2% formic acid–acetonitrile, and water–acetonitrile. A variety of extraction solvents were evaluated simultaneously, including ethanol and methanol at concentrations of 60%, 80% and absolute. The performance of different chromatographic column brands and detection wavelengths was also assessed (Figures 1, 2, 3, and 4).

Comparison of HPLC chromatogram using different extraction solvents. (A) 60% ethanol; (B) 80% ethanol; (C) 100% ethanol; (D) 60% methanol; (E) 80% methanol; (F) 100% methanol.

Comparison of HPLC chromatograms with different mobile phases. (A) 0.2% formic acid–acetonitrile system; (B) 0.1% formic acid–acetonitrile system; (C) 0.2% phosphoric acid–acetonitrile system; (D) water–acetonitrile system.

Comparison of HPLC chromatogram with different chromatographic columns. (A) Shim‐pack Scepter C18 (4.6 × 150 mm, 5 μm); (B) Agilent TC‐C18 (4.6 × 250 mm, 5 μm); (C) Shim‐pack GIST C18 (4.6 × 150 mm, 5 μm); (D) InertSustain AQ‐C18 (4.6 × 150 mm, 5 μm); (E) Capcell Pak C18‐MGII (4.6 × 250 mm, 5 μm).

Comparison of HPLC chromatogram at different detection wavelengths. (A) 366 nm; (B) 342 nm; (C) 300 nm; (D) 272 nm; (E) 254 nm.

Of the solvents tested, methanol demonstrated superior extraction efficiency, producing a greater number of peaks, better separation, and more symmetrical peak shapes. In contrast, mobile phases containing water or 0.1% phosphoric acid resulted in poorer separation and lower peak intensities. Optimal chromatographic performance was achieved at a detection wavelength of 342 nm, providing a stable baseline and comprehensive peak resolution. Based on these results, methanol was selected as the extraction solvent, and a mobile phase consisting of 0.1% formic acid and acetonitrile was adopted for subsequent HPLC analyses of VTH at 342 nm.

Furthermore, the resolution, retention time, and peak morphology of VTH components varied depending on different chromatographic columns used, likely due to the distinct chemical and physical properties of the stationary phase materials. Therefore, careful selection of a suitable column is recommended to ensure consistent, high‐quality results.

The method validation demonstrated excellent performance. The precision and stability evaluations showed a chromatographic similarity score of 1.000 across multiple analyses. In repeatability assessments of six independent VTH sample injections, similarity values were 0.998, 0.998, 0.999, 0.999, 1.000, and 0.998, respectively. The RSDs for the main peaks were all below 3.0%, indicating good method precision and repeatability. The sample solutions also remained stable at room temperature with no observable precipitation during the testing period.

Establishment of the HPLC Fingerprint

3.2

Identification of Common Peaks

3.2.1

A total of 21 batches of VTH samples were analyzed in accordance with established protocols. Each batch was processed to prepare sample solutions, and HPLC was employed to generate corresponding chromatographic fingerprints. These data were analyzed using the Chinese Medicine Chromatographic Fingerprint Similarity Evaluation System (Version 2012A). Sample S1 was designated as the reference chromatogram, and alignment was achieved using the median method with a time window of 0.5.

An overlay chromatogram of the 21 VTH samples was constructed (Figure 5), and peaks with a separation factor greater than 1.5 were selected as reference peaks. A total of 23 characteristic peaks were identified, representing over 92% of the total peak area, and were thus designated as common peaks in the VTH's HPLC fingerprint.

Superimposed fingerprint chromatograms of 21 VTH batches from different origins.

Peak identification was achieved by comparing retention times with those of authenticated chemical reference standards and interpreting mass spectrometry fragmentation patterns, and consulting relevant literature. The following compounds were successfully identified: quercetin 3‐O‐rhamnosyl(1′′′ → 6′′)‐galactosyl‐7‐O‐sophoroside (1), quercetin‐3‐O‐sophorosyl‐7‐O‐glucoside (2), kaempferol 3‐O‐(2′′‐glucosyl)rutinosyl‐7‐O‐glucoside (3), kaempferol 3‐O‐sophorosyl‐7‐O‐glucoside (4), quercetin 3,7‐di‐O‐glucoside (5), isorhamnetin 3‐O‐sophorosyl‐7‐O‐glucoside (6), kaempferol 3,7‐di‐O‐glucoside (7), quercetin 3‐O‐rutinosyl‐7‐*O‐*glucoside (8), kaempferol 3‐O‐(2′′‐O‐pentosy)‐D‐rutinosyl‐7‐O‐glucoside (9), kaempferol 3‐O‐rutinosyl‐7‐O‐glucoside (10), isorhamnetin 3,7‐di‐O‐glucoside (11), isorhamnetin 3‐O‐rutinosyl‐7‐O‐glucoside (12), quercetin 3‐O‐(2′′‐glucosyl) rutinoside (13), apigenin 6‐C‐arabinosyl‐8‐C‐glucoside (14), kaempferol 3‐O‐(2′′‐glucosyl) rutinoside (15), kaempferol 3‐O‐sophoroside (16), apigenin 6,8‐di‐C‐arabinoside (17), rutin (18), kaempferol 3‐O‐(2,6‐di‐rhamnosyl)‐glucoside (19), nicotiflorin (20), narcissoside (21), astragalin (22), isorhamnetin 3‐O‐glucoside (23), and esculetin (Q). These compounds were confirmed by comparison with authentic standards and are consistent with previously reported in the literatures, as illustrated in Figure 6.

HPLC chromatogram of mixed standard solution (A) and VTH sample (B).

Similarity Analysis of the HPLC Fingerprints

3.2.2

The HPLC chromatograms of 21 batches of VTH were analyzed using the Chinese Medicine Chromatographic Fingerprint Similarity Evaluation System (version 2012A). A reference fingerprint chromatogram was established and designated as “R,” and the similarity of each batch was calculated in comparison to this reference. The resulting similarity scores were 0.999, 0.966, 0.999, 0.961, 0.998, 0.977, 0.999, 0.980, 0.982, 0.967, 0.984, 0.984, 0.983, 0.983, 0.984, 0.984, 0.984, 0.983, 0.926, 0.963, and 0.984. These values indicate a generally high degree of similarity between the batches, suggesting consistent chemical composition across most samples.

Chemometric Analysis of VTH

3.3

Hierarchical cluster analysis (HCA) was performed on Samples 1–21, with the peak areas of 23 common peaks used as the variables. IBM SPSS Statistics 25 software was used to conduct the analysis, employing the between‐groups linkage method with cosine similarity as the distance measure (Lv et al. 2022). The resulting dendrogram (Figure 7) shows that the samples were grouped into five distinct clusters at a classification distance of 6: Cluster I (Samples 18 and 19), Cluster II (Sample 21), Cluster III (Samples 2 and 4), Cluster IV (Sample 9), and Cluster V (the remaining 15 samples).

Hierarchical cluster analysis (HCA) of 21 batches of VTH.

Cluster V comprises samples from Keriya County (KC), Qira County (QC), Tashkurgan County (TC), and Hotan County (HC). The samples are of a relatively consistent quality, likely due to the similar altitude and climatic conditions across the Kunlun Mountains region. However, notable classification differences were observed among samples from HC (Samples 9, 18, 19, and 21) and QC (Samples 2 and 4), which can be attributed to variations in their chemical profiles. These results emphasize the chemical diversity even within the same geographical origin.

Principal component analysis (PCA) was performed using the SIMCA‐P^+^ 14.1 software package. The peak areas of the 23 common peaks were standardized across all 21 VTH samples (Figure 8A,B). Four principal components (PCs) were extracted, cumulatively explaining 91.25% of the total variance. PC1 reflected differences in most of the compounds, excluding Components 1 and 22; PC2 emphasized variations in Components 1, 6, 11–13, 15, and 17–20; PC3 captured variability mainly in Components 1, 14, and 22–23; and PC4 contributed minimal variance, focusing on Components 1, 3, 5, 14, 16, 19, and 23. Components 1, 14, 22, and 23 exhibited minor variations across batches, making them reliable markers for batch‐to‐batch comparison. In contrast, Components 6, 11–13, 15, and 17–20 showed moderate variability, reflecting differences in specific quantitative attributes. Components 2–5, 7–10, and 16 demonstrated more substantial variations, indicating that they are major contributors to chemical differentiation. Overall quality was assessed using PCA, which revealed that the cumulative contribution of PC1 to PC3 reached 84.27%. This suggests strong chemical similarity across VTH samples from different regions. However, notable dispersion was observed within groups: Samples from QC (excluding Samples 1 and 3) displayed the greatest variability, whereas HC samples (excluding Samples 9–11) formed a separate group. KC samples, meanwhile, were generally more tightly clustered. Interestingly, QC Samples 1 and 3, as well as HC Samples 9–11, were closely aligned with KC samples, indicating a similarity in their chemical compositions. Conversely, QC Samples 2 and 4, as well as HC Sample 9, showed substantial divergence, corroborating the HCA findings.

Chemometrics analysis diagram. Principal component analysis (PCA) 2D (A) and score 3D (B) plots of 21 VTH batches; (C) score plot of orthogonal partial least squares‐discriminant analysis (OPLS‐DA) for 21 VTH batches; (D) variable importance in projection (VIP) values from the OPLS‐DA model.

Overall, PCA provided a detailed model for distinguishing regional sources, assessing intragroup dispersion and variability, and evaluating consistency in quantitative composition. This enhanced the interpretability of fingerprint‐based quality control.

To further explore interbatch differences, an orthogonal partial least squares discriminant analysis (OPLS‐DA) was performed on the 21 VTH samples, using the peak areas of 23 common peaks as the variables. The model demonstrated good fitness, with R ^2^ X and R ^2^ Y values of 0.98 and 0.68, respectively. The score plot (Figure 8C) showed that samples from TC were more similar to those from KC. Samples from QC, however, were clearly distinguishable from both Hotan and Keriya Counties, except for Sample 5. These results were consistent with the findings from PCA. Furthermore, variable importance in projection (VIP) analysis (Figure 8D) identified 11 peaks with VIP values exceeding 1.0, indicating their substantial contribution to sample differentiation. The top‐ranked markers were Components 19, 20, 16, 2, 14, 21, 13, 17, 22, 6, and 15. These variables serve as key indicators of quality differentiation among VTH samples and may aid in tracing their geographical origins.

Network Pharmacology Results

3.4

The active components identified from the HPLC fingerprints of VTH were queried in the TCMSP database, using the results of in vivo and in vitro assays. This yielded 15 exposed components (compounds 7–8 and 11–23) and eight predicted active components (Compounds 1–6 and 9–10), which were used for target identification. A total of 201 potential targets were associated with these 23 candidate compounds using SwissTargetPrediction and SuperPred. Meanwhile, 2014 asthma‐related targets were retrieved from the GeneCards, DisGeNET, and OMIM databases. The intersection of these two datasets identified 88 overlapping targets (Figure 9A), which were then used to create a protein–protein interaction (PPI) network comprising 382 edges, with an average node degree of 16.6 (Figure 9A–B). The top 20 nodes exhibited an average degree of 19.1, indicating strong multitarget activity. The three most connected targets had degrees of 50, 43, and 39, respectively, suggesting a pivotal role for them in the antiasthmatic mechanism of VTH. Subsequent network analysis using Cytoscape 3.9.1 and the CentiScaPe 2.2 plugin identified 19 core targets and 150 interactions with an average degree of 7.9 (Figure 9C). In the visual network, nodes with darker colors and larger sizes denote higher centrality. Key targets included serine/threonine‐protein kinase 1 (AKT1), tumor necrosis factor (TNF), heat shock protein 90 alpha family class A member 1 (HSP90AA1), epidermal growth factor receptor (EGFR), prostaglandin‐endoperoxide synthase 2 (PTGS2), and nuclear factor of kappa light polypeptide gene enhancer in B‐cells 1 (NF‐κB1), all of which may be crucial in mediating VTH's antiasthmatic effects.

Network pharmacology graph. (A) Key targets of VTH in asthma treatment; (B) active compound–target network diagram; (C) PPI network diagram; (D) GO enrichment analysis; (E) KEGG signaling pathway map; (F) construction of network diagram for mechanism prediction of VTH in treating asthma.

To elucidate the biological relevance of VTH's asthma‐related targets, Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses were conducted on the 51 potential antiasthmatic targets. GO analysis revealed 1862 enriched terms, including 1702 BPs, 113 molecular functions (MFs), and 47 cellular components (CCs). Prominent BPs included the regulation of cellular reactive oxygen species metabolism (e.g., AKT signaling), the positive regulation of nitric oxide biosynthesis, and epithelial cell proliferation. The enriched CCs involved the basement membrane, lipid rafts, and phagocytic cups (Figure 9D). KEGG pathway enrichment analysis identified 143 significant pathways (p < 0.05), of which 38 remained after filtering (p < 1 × 10^−5^). Of these, 22 pathways were disease‐related, mainly involving microbial infections such as human cytomegalovirus and Kaposi sarcoma‐associated herpesvirus infection, as well as cancer‐related pathways, including proteoglycans in cancer, prostate cancer, and resistance to FGFR tyrosine kinase inhibitors. The remaining 16 non‐disease‐related pathways highlighted the PI3K‐Akt signaling pathway, lipid metabolism and atherosclerosis in the bronchus, and chemokine signaling pathway in inflammation (Figure 9E), suggesting that these pathways could be potential therapeutic targets in the treatment of asthma with VTH.

A comprehensive pharmacological network was constructed using Cytoscape 3.9.1 to link active components, targets, and pathways (Figure 9F). This network highlighted the central roles of TNF, AKT1, PTGS2, NF‐κB1, HSP90AA1, glycogen synthase kinase 3‐β (GSK3β), and EGFR in the pharmacological activity of VTH. TNF is known to be involved in chronic inflammation (Mangova et al. 2020) and modulates dendritic cell differentiation toward Th2/Th17 lineages in asthma models (Vroman et al. 2018). Inhibiting TNF reduces ST2 expression, which can potentially elevate free IL‐33 levels and trigger airway inflammation (Kaur et al. 2020). PTGS2 (COX‐2) regulates oxidative stress and inflammation; its overexpression in bronchial smooth muscle can result in prostaglandin overproduction and airway hyperresponsiveness (AHR) (Chiba et al. 2018; Rumzhum and Ammit 2016). HSP90AA1 influences the duration and intensity of inflammation via the NF‐κB pathway (H. Huang et al. 2024; Zuehlke et al. 2015). Under hypoxic conditions, the expression of vascular endothelial growth factor (VEGF) is mediated by hypoxia‐inducible factor 1‐alpha (HIF‐1α), thereby exacerbating bronchial inflammation (Jiang et al. 2021; Laddha and Kulkarni 2019). GSK3β, a serine/threonine kinase, plays a key role in signal transduction and airway inflammation (Huang et al. 2022; Liu et al. 2025). Furthermore, nuclear factor erythroid 2‐related factor 2 (NFE2L2) has been identified as a critical regulator of oxidative stress and airway inflammation in asthma (Córdova et al. 2011; Luo et al. 2025). Collectively, these findings reveal the multicomponent, multitarget, and multipathway mechanisms underlying the therapeutic effects of VTH in asthma treatment.

Molecular Docking Verification

3.5

Molecular docking was employed to predict and refine the binding conformations of VTH components (ligands) with five key protein targets (receptors): TNF, AKT1, PTGS2, NF‐κB1, HSP90AA1, GSK3β, and EGFR. Docking simulations were performed for 23 compounds, and the crystal structures of the target proteins, together with the docking scores for their optimal ligands, are shown in Figure 10 and Table 2. A total of 161 docking scores were below −6.0 kcal/mol, indicating strong binding affinities between most VTH components and the selected targets.

Heat map of molecular docking scores for VTH components and asthma‐related targets.

The docking results indicate that the active components exhibit favorable affinity patterns with various targets. The study identified a positive correlation between the number of glycosides in flavonols and their potential protein targets, accompanied by an improvement in binding affinity as evidenced by lower docking scores. Furthermore, the monoside, diside, and triside compounds of flavonols exhibit conformational similarity and display comparable binding effects with proteins. For instance, Compound 4 binds to TNF, and it forms multiple hydrogen bonds between the hydroxyl groups of the glycoside and residues such as Gln25, Gln27, and Asp45, while also establishing hydrogen bonds with residues such as Asn19, Asn46, Gln47, Lys90, Glu135, and Asn137. Additionally, this compound demonstrates hydrophobic interactions with Ala22, Leu26, Ile136, and Pro139 (see Figure 11A). Stable hydrogen bonds and hydrophobic interactions, including π–H, π–π, and π–cation interactions, are evident between flavonol glycosides and protein residues, as the corresponding hydroxyl or phenolic hydroxyl groups can act as hydrogen bond donors or acceptors. These interactions enhance the stable binding of flavonol glycoside molecules to their active sites, thereby inhibiting protein activity. Figure 11B–F illustrates the interactions between AKT1, PTGS2, NF‐κB1, HSP90AA1, GSK3β, and EGFR with representative compounds.

Molecular docking visualization between VTH active ingredients and key targets. (A) Compound 4–TNF; (B) Compound 10–AKT1; (C) Compound Q–PTGS2; (D) Compound 20–NF‐κB1; (E) Compound 21–GSK3β; (F) Compound 21–EGFR.

Quantification of Q‐Markers in VTH

3.6

Selection of Q‐Maker

3.6.1

A total of 23 flavonoid compounds were identified in the HPLC fingerprint of VTH, primarily as quercetin, kaempferol, isorhamnetin, and apigenin derivatives glycosylated via C–O or C–C bonds with mono‐ to tetra‐saccharides. Network pharmacology analysis revealed strong correlations between these compounds and seven key asthma‐related targets (TNF, AKT1, PTGS2, NF‐κB1, HSP90AA1, GSK3β, and EGFR), which were supported by favorable molecular docking results. Furthermore, PCA and OPLS‐DA analyses revealed 11 differential components. In animal experiments, esculetin has been shown to reduce the infiltration of inflammatory cells in the lungs of mice with LPS‐induced acute lung injury (ALI) (H. C. Lee et al. 2020). This effect is mediated through the inhibition of the Akt/ERK/NF‐κB and RORγt/IL‐17 signaling pathways, which are pivotal in inflammation. Similarly, kaempferol 3‐sophoroside‐7‐glucoside mitigated lipopolysaccharide (LPS)–induced ALI by inhibiting the activation of nuclear factor‐κB (NF‐κB) and significantly downregulating the expression of its downstream inflammatory cytokine, tumor necrosis factor‐α (TNF‐α) (Abdulaal et al. 2024). Astragalin was found to inhibit the levels of TNF‐α and matrix metallopeptidase 9 (MMP‐9), as well as the oxidative stress markers malondialdehyde (MDA) and reactive oxygen species (ROS), while promoting the activity of heme oxygenase‐1 (HO‐1), thereby alleviating lung injury (Zheng et al. 2019). In cellular experiments, kaempferol 3‐O‐rutinoside‐7‐O‐glucoside exhibited a concentration‐dependent inhibitory effect (1.25, 2.5, and 5 μM) on the secretion of nitric oxide (NO), prostaglandin E2 (PGE2), TNF‐α, interleukin‐1β (IL‐1β), and interleukin‐6 (IL‐6) in LPS‐stimulated RAW 264.7 cells. Additionally, nicotiflorin and narcissoside were shown to reduce the release of NO, an inflammatory mediator following endothelial cell damage. These findings collectively indicate their potential anti‐inflammatory efficacy (Fang et al. 2022; Yang et al. 2025; Yu et al. 2021). Therefore, these six components were selected for further analysis.

Based on the five criteria for Q‐markers in traditional Chinese medicine—specificity, measurability, traceability, relevance to traditional theories, and efficacy—five representative compounds (Peaks 4, 10, 20, 21, and 22) were selected as preliminary Q‐markers (Figure 12). These were then quantified using validated analytical methods.

HPLC chromatogram of quantitative determination for VTH.

Method Validation for Quantitative Analysis

3.6.2

The quantitative methods for six compounds (kaempferol 3‐O‐sophorosyl‐7‐O‐glucoside, esculetin, kaempferol 3‐O‐rutinosyl‐7‐O‐glucoside, nicotiflorin, narcissoside, and astragalin) were validated for precision, repeatability, stability, and accuracy. The RSDs for precision ranged from 0.36% to 0.63%, and those for repeatability ranged from 0.42% to 2.14%. Stability testing over a short period yielded RSDs between 2.03% and 3.30%.

Average recoveries ranged from 92.83% to 103.49%, with associated RSDs below 3.00%, confirming the method's accuracy. Regression analysis demonstrated excellent linearity for all compounds (r > 0.9999), and their linear ranges, detection limits, and quantification limits are summarized in Table 3.

Quantitative Results of Q‐Markers

3.6.3

Twenty‐one batches of VTH samples were analyzed under optimized chromatographic conditions. The chromatographic peak areas were used to quantify six compounds via established regression equations (Table 4). The concentration distributions were visualized using cumulative sum plots (Figure 13), which revealed significant regional differences in Q‐marker content. Samples from the QC region exhibited the highest levels of Q‐markers, followed by HC and KC regions. These findings were consistent with those from HCA and PCA. Esculetin showed substantial variation in concentration, ranging from 1084 ng/g in Batch S18 to 1435 ng/g in Batch S19, whereas concentrations in other batches remained relatively stable, suggesting its potential as a distinctive marker.

Stacked bar chart of the quantification results for major constituents in VTH.

Kaempferol 3‐O‐rhamnoside‐7‐O‐glucoside also showed pronounced variation (factor of 4.19), ranging from 837 ng/g in Batch S3 to 2418 ng/g in Batch S2. The remaining four compounds—astragalin, hyperoside, kaempferol 3‐O‐rutinosyl‐7‐O‐glucoside, and narcissoside—demonstrated variation factors of 3.90, 2.86, 2.85, and 2.75, respectively. Standard deviation analysis revealed that astragalin, hyperoside, and narcissoside exhibited relatively minor deviations (SD from 0.05 to 0.15), suggesting their reliability as stable Q‐markers for VTH.

Discussion

4

To improve the scientific robustness of the HPLC fingerprinting method for VTH, extensive optimization experiments were conducted. These included evaluating different extraction solvents, mobile phase compositions, and chromatographic columns. Methanol was found to be the most effective extraction solvent. Consequently, a mobile phase consisting of 0.1% formic acid and acetonitrile was chosen for the subsequent analysis of VTH samples.

HPLC fingerprint profiles were successfully established for 21 batches of VTH, revealing 23 constituents present in all of them. Chemometric techniques, including HCA, principal component analysis (PCA), and orthogonal partial least squares discriminant analysis (OPLS‐DA), were employed to classify and differentiate these samples. These analyses supported the specificity and potential of certain compounds as Q‐markers. PCA effectively distinguished samples from different regions and provided insights into component consistency and distribution. Meanwhile, OPLS‐DA offered robust discrimination of sample origins based on chemical composition, thereby contributing to the scientific basis for Q‐marker screening and traceability.

Network pharmacology and molecular docking were employed to investigate the potential mechanisms of VTH in asthma treatment. Network pharmacology is a contemporary paradigm in biomedical research that utilizes high‐throughput data, computational modeling, and network databases to explore complex biological interactions. Leveraging this approach identified key targets associated with asthma, including TNF, AKT1, PTGS2, NF‐κB1, HSP90AA1, GSK3β, and EGFR. These targets are closely related to bronchial inflammation and airway remodeling, which are hallmarks of asthma pathophysiology. TNF‐α, a pro‐inflammatory cytokine, plays a central role in the development of AHR by directly affecting airway smooth muscle (ASM) and contributing to the neutrophilic inflammation observed in severe asthma (Aldhalmi et al. 2022). Furthermore, EGFR and its signaling partner GSK3β play a major role in regulating Mucin 5AC (MUC5AC) in bronchial epithelial cells, with increased EGFR activation leading to elevated MUC5AC expression (Memon et al. 2020; Zhen et al. 2007). This pathway is activated by environmental pollutants such as cigarette smoke, resulting in excessive mucus secretion and epithelial barrier dysfunction (Luo et al. 2025). Pathway analysis via KEGG highlighted the importance of the PI3K‐Akt signaling pathways in this context. The PI3K/AKT pathway and its key downstream mediators (such as mTOR, GSK3β, and NF‐κB) play a pivotal role in modulating autophagy, inflammation, and cell proliferation (Cheng et al. 2021; Ma et al. 2021; S. Wang et al. 2022). Inhibiting this pathway has been shown to downregulate inflammatory mediators such as IL‐4, IL‐5, IgE, and eosinophils, as well as reducing ASM proliferation and oxidative stress. GSK3β also participates in feedback regulation within the PI3K/AKT/mTOR network (Hermida et al. 2017). Collectively, these findings suggest that modulation of the PI3K/Akt axis is a key mechanism by which VTH may exert antiasthmatic effects.

Molecular docking further supported these hypotheses by revealing stable interactions between VTH components and the aforementioned protein targets. Notably, glycosylated flavonoids exhibited enhanced binding affinities due to an increased potential for hydrogen bonding and steric interactions. This suggests that glycosylation can enhance the pharmacological activity of VTH components, potentially improving therapeutic outcomes. These results collectively indicate that VTH components may reduce asthma‐related airway inflammation and remodeling by affecting phosphorylation events within the PI3K/AKT pathway.

However, it is important to acknowledge the limitations of this study. The mechanistic insights presented are based primarily on computational predictions. Although these findings offer a promising foundation, they lack experimental validation. Further in vitro and in vivo studies are essential to verify the proposed target interactions and biological activities and to clarify the precise molecular mechanisms by which VTH exerts its therapeutic effects against asthma.

Conclusions

5

In this study, an HPLC fingerprint chromatogram was established for VTH, identifying 23 common peaks. Chemometric methods, including HCA, PCA, and OPLS‐DA, were employed to distinguish samples from different geographical origins and to assess variations in the quality of the medicinal materials. In the evaluation of Q‐markers for VTH, specific and differential compounds were identified to enable accurate quantitative analysis.

Furthermore, network pharmacology combined with molecular docking techniques was applied to investigate the therapeutic potential of the active constituents of VTH in the treatment of asthma. The results suggest that these components may alleviate airway inflammation in asthma by modulating key targets (including AKT1, NF‐κB1, HSP90AA1, and GSK3β) within the PI3K/AKT signaling pathways, as well as other signaling pathways (such as PTGS2 and EGFR). These findings highlight the potential of VTH as a promising treatment for bronchial asthma and provide a foundation for future pharmacological and clinical investigations.

Representative Q‐markers were further screened, and the contents of several bioactive compounds—such as kaempferol 3‐O‐sophorosyl‐7‐O‐glucoside, esculetin, kaempferol 3‐O‐rutinosyl‐7‐O‐glucoside, nicotiflorin, narcissoside, and astragalin—were quantitatively determined. This enabled a comparative analysis of medicinal quality across samples from different regions and offers a scientific basis for the standardization and quality control of VTH.

Author Contributions

Haifeng Liu, Rongmei Zhao: writing – original draft, data curation, investigation, visualization, conceptualization. Tabusi Manaer, Le Pan, Hong Xu: formal analysis, methodology, resources, validation, writing – review and editing. Lu Jin, Amatjan Ayupbek: conceptualization, funding acquisition, supervision, validation, writing – review and editing. All authors have read and agreed to the published version of the manuscript.

Funding

This work was supported by the Major Science and Technology Projects in the Xinjiang Uyghur Autonomous Region (grant number: 2022A03008‐3).

Conflicts of Interest

The authors declare no conflicts of interest.