Deciphering the Molecular Basis: Integrative Multi‐Method Analysis of Bioactive Constituents and Their Anti‐Lung Cancer Mechanisms in Clinical Traditional Chinese Medicine Formulations

ABSTRACT

Despite variations in the composition of traditional Chinese medicine (TCM) formulas for lung cancer used in clinical practice, they show consistent clinical efficacy. This study employed an integrated approach combining network pharmacology, bioinformatics, and in vitro experiments to analyze 45 clinical prescriptions. From the 188 herbs identified, licorice and atractylodes rhizome were the most frequently used. Analysis revealed seven key active compounds, including quercetin and stigmasterol, acting on 149 shared targets, with significant enrichment in pathways such as PI3K/AKT. Molecular docking confirmed strong binding between core compounds and targets like PRKACA. Bioinformatics analysis of TCGA data further reveals that MMP9 may promote tumor invasion and metastasis by degrading the extracellular matrix. In vitro, quercetin and stigmasterol inhibited A549 cell proliferation and induced apoptosis, with quercetin suppressing PI3K/AKT activation. The findings indicate that licorice and atractylodes rhizome may play a synergistic role, and quercetin acts as a key anti‐cancer compound by inhibiting the PI3K/AKT pathway, elucidating a common pharmacological basis for these TCM formulas.

Based on clinical prescriptions of lung cancer patients from hospitals, this study systematically analyzes frequently used traditional Chinese medicines (TCMs). Network pharmacology is adopted to predict their active ingredients, core targets, and key signaling pathways. Furthermore, A549 cell experiments are performed to verify the anti‐lung cancer efficacy and molecular mechanisms of the major active ingredients.

1. Introduction

Lung cancer is the most common type of cancer globally. According to the 2024 statistics report by the International Agency for Research on Cancer, there were 2.48 million new cases of lung cancer in 2022, accounting for 12.4% of all cancer cases, with approximately 1.8 million deaths attributed to lung cancer [1]. Although targeted therapies and immunotherapies have significantly improved outcomes for lung cancer patients, further reducing the risks of recurrence, metastasis, and mortality remains a major clinical challenge [2]. Patients undergoing long‐term chemotherapy or targeted therapy often develop drug resistance and experience various adverse reactions such as liver and kidney dysfunction, bone marrow suppression, and peripheral neuropathy, posing significant challenges for clinical treatment [3, 4]. Therefore, there is an urgent need to explore more effective methods for the prevention and treatment of lung cancer.

Traditional Chinese medicine (TCM) compound formulas exhibit characteristics of multiple components, targets, and pathways in disease prevention and treatment. Due to their broad anti‐tumor effects, lower toxicity, and proven efficacy, TCM compound formulas have received considerable attention in the field of lung cancer prevention and treatment [5]. Research has shown that TCM compound formulas demonstrate significant advantages in improving symptoms, reducing tumor recurrence, controlling disease progression, extending survival periods, and enhancing the quality of life for lung cancer patients [6, 7, 8, 9]. These attributes make TCM compound formulas an important component of comprehensive lung cancer treatment strategies. However, due to the complex composition of compound traditional Chinese medicine, which includes multiple herbs and numerous active ingredients, it is very difficult to determine the main active ingredients [10]. Therefore, it is necessary to conduct in‐depth exploration based on clinical practice. Clinical research can not only more accurately evaluate the actual effectiveness and safety of compound traditional Chinese medicine, but also optimize drug formulations, identify key components and their interactions, thereby improving treatment efficacy and providing a scientific basis for its widespread application [11].

The aim of this study is to conduct an in‐depth analysis of the use of TCM among clinical lung cancer patients, meticulously sorting through various TCMs utilized by patients through data mining techniques. For drugs that appear with higher frequency, network pharmacology and bioinformatics methods are employed for in‐depth research, aiming to uncover the potential mechanisms and key targets of these TCMs and their active ingredients in the treatment of lung cancer. Through systematic analysis, this study hopes to screen out TCM ingredients with significant anticancer effects and preliminarily elucidate their mechanisms of action. Finally, in vitro experiments are conducted to further validate the mechanisms of these TCM ingredients, providing new ideas and strategies for the clinical treatment of lung cancer.

2. Materials and Methods

2.1. Clinical Screening and Evaluation of TCM

2.1.1. Data Collection

A retrospective study was conducted to systematically collect detailed data on the treatment of all lung cancer patients with traditional Chinese medicine in the oncology department of a certain hospital in 2023. These data were extracted strictly following the management procedures of the hospital's internal system, ensuring their accuracy and completeness. Additionally, the entire data collection process underwent rigorous review and registration by the hospital's ethics committee, fully respecting and protecting patients’ privacy rights and personal information security.

2.1.2. Inclusion and Exclusion Criteria for Prescriptions

A total of 122 patients with lung cancer who received TCM treatment at Integrated TCM & Western Medicine Department of Mianyang Central Hospital in 2023 were retrospectively enrolled in this study. The diagnosis of lung cancer was confirmed by histopathological or cytopathological examination according to the diagnostic criteria of the Chinese Guidelines for the Diagnosis and Treatment of Primary Lung Cancer [12]. TCM was used as an adjunctive therapy in combination with conventional treatments (including chemotherapy, radiotherapy, targeted therapy, or immunotherapy) or as a standalone treatment when patients were intolerant to or refused conventional therapies.

Inclusion Criteria: (1) Patients pathologically diagnosed with non‐small cell lung cancer or small cell lung cancer; (2) patients who received TCM treatment for a duration of no less than 4 weeks; (3) patients with available clinical and prescription records without critical missing data; (4) patients aged ≥ 18 years; (5) patients with an Eastern Cooperative Oncology Group performance status score of 0∼2 [13]; (6) patients who received TCM treatment as part of routine clinical care without additional research‐related interventions.

Exclusion Criteria: (1) Patients with a history of other active malignancies within the past 5 years; (2) patients receiving other investigational drugs or participating in another clinical trial during the treatment period; (3) patients with incomplete or illegible prescription records that prevent accurate interpretation; (4) patients under 18 years of age; (5) patients with lung cancer complicated by severe hepatic dysfunction (Child‐Pugh class C) or renal dysfunction (eGFR < 30 mL/min/1.73m²); 6) Pregnant with lung cancer.

2.1.3. Content of Information Collected

The collected major contents include: (1) general basic information of lung cancer patients: gender, age, duration of illness, etc; (2) components and dosages of traditional Chinese medicine; (3) clinical manifestations of patients.

2.1.4. Analysis of TCM Components

To screen for high‐frequency drugs using statistical analysis, the first step is to collect and organize drug usage data, ensuring data accuracy. Subsequently, the usage frequency of each drug is calculated, and descriptive statistical analysis is conducted to visually present its distribution. Finally, based on the statistical results, TCMs with higher frequencies are selected for further network pharmacological analysis.

2.2. Network Pharmacology Analysis

2.2.1. Major TCMs and Disease Target Screening

Using the Traditional Chinese Medicine Systems Pharmacology Database and Analysis Platform (TCMSP), a search was conducted with the names of major TCMs as keywords, and the active ingredients of the retrieved compounds were screened. The screening criteria were set at Oral bioavailability (OB) >30% and drug similarity (DL) > 0.18, resulting in the identification of eligible active ingredients [14]. Based on these results, we selected the active ingredients with higher frequency for further analysis. After removing duplicate data and further screening, the UniProt protein database was used to standardize the targets. Searches were then conducted using “Lung Cancer” as a keyword through the GeneCards and OMIM databases. The first database extracted targets with a relevance score of 10 or higher, and the results from both databases were combined to eliminate duplicates, thereby obtaining the disease‐related targets. The intersection between the active ingredient targets of the major TCMs and the lung cancer‐related targets was identified using the Weishengxin online platform (http://www.bioinformatics.com.cn/). Specifically, the two target lists were separately uploaded to the platform, which automatically generated the overlapping targets. These overlapping targets were considered the potential key targets of the TCMs against lung cancer.

2.2.2. Construction of PPI and Active Ingredients‐Disease‐Target Network

To investigate the interactions among target proteins, the shared target of prominent TCMs and Lung Cancer is uploaded onto the STRING platform. Here, the species filter is set to “Homo Sapiens,” and no combined score screening is applied, leading to the generation of a protein‐protein interaction (PPI) network. Subsequently, the pertinent target information of significant TCMs and Lung Cancer is imported into Cytoscape software to establish an active ingredient‐disease‐target network.

2.2.3. GO Function and KEGG Pathway Enrichment Analysis

Using the DAVID database, we conducted Gene Ontology (GO) function analysis (encompassing three categories: biological processes, molecular functions, and cellular components) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis on the primary targets of key TCMs aimed at lung cancer prevention and treatment. The enrichment results were filtered using a significance threshold of p < 0.05. No additional multiple testing correction (e.g., Benjamini‐Hochberg FDR) or gene count criteria were applied, as the DAVID database does not require these parameters for basic enrichment analysis under the default settings. The obtained results were then saved and organized, and the most significant biological processes and signaling pathways were identified. For data visualization only, the enrichment results were imported into the Weishengxin online platform (http://www.bioinformatics.com.cn/) to generate the GO and KEGG enrichment diagrams. The platform was used solely as a graphing tool and did not involve any additional statistical calculations. We have clearly stated in the revised manuscript that a p‐value threshold of <0.05 was used for the enrichment analysis.

2.3. Bioinformatics Analysis

2.3.1. Molecular Docking

AutoDock was utilized to conduct the semi‐flexible molecular docking between the ligand and the receptor, and the free binding energy was determined using the scoring function of AutoDock Vina. A free binding energy of ≤ −5.0 kJ mol⁻¹ indicated successful docking between the receptor and the ligand. The binding ability improved as the value decreased [15]. For docking results with low free binding energy, PyMOL software was employed for visualization.

2.3.2. Differential Gene Expression in Lung Adenocarcinoma of Lung Cancer

To delve deeper into the molecular mechanisms of lung adenocarcinoma (LUAD), we obtained mRNA data from LUAD patients from TCGA, encompassing both cancerous and normal tissues. Using the DESeq2 package in R for analysis, we set the criteria of an absolute log2FoldChange ≥ 1 and a p value < 0.05 to screen for key differentially expressed genes. To visually represent their expression patterns, we utilized the ggplot R package to create volcano plots, which clearly display log2 Fold Change and −log10 (p value), facilitating the rapid identification of genes with significant changes and statistical significance, and providing crucial leads for subsequent functional analysis and experimental validation.

2.3.3. Kaplan–Meier (KM) Plotter Database

The Kaplan–Meier plotter, which can be accessed online at http://kmplot.com/analysis, utilizes a comprehensive online database specifically designed to assess the influence of genes on survival rates in patients with Lung Cancer [16]. An in‐depth online analysis was performed to evaluate the correlation between the expression levels of individual core genes and survival outcomes in Lung Cancer. The results of this analysis were presented in a clear and informative manner, including the hazard ratio, 95% confidence intervals, and a computed log‐rank p‐value, which together provide valuable insights into the prognostic significance of these core genes in Lung Cancer.

2.4. Experiments Part

2.4.1. Cell Culture and Drug Treatment

First, A549 lung cancer cells (RRID: CVCL_0023) were retrieved from liquid nitrogen, resuscitated, and cultured to the logarithmic growth phase. The cells were then seeded into 96‐well plates. After cell adherence, different concentrations of quercetin (CAS: 117‐39‐5), kaempferol (CAS: 520‐18‐3), stigmasterol (CAS: 83‐48‐7), and isorhamnetin (CAS: 480‐19‐3) were added (each concentration was set up in triplicate). For comparison of drug effects, a blank control group was also established.

2.4.2. Cell Proliferation Inhibition Experiment

After 24 h of drug treatment, CCK‐8 reagent (Cat. No. C0037) was added to each well, followed by incubation in a 37°C incubator for 90 min. The absorbance at 450 nm was then measured using a microplate reader to calculate cell viability. Based on these viability data, the IC₅₀ values of the drugs were determined to evaluate their proliferation inhibitory effects. Before curve fitting, the drug concentrations were converted to logarithmic (Log) form. Subsequently, a non‐linear regression analysis was performed using GraphPad Prism software against a variable‐slope four‐parameter logistic model. This dose‐response curve fitting allowed for the calculation of the half‐maximal inhibitory concentration together with its 95% confidence interval.

2.4.3. Cell Apoptosis Detection

According to the IC₅₀ results, the two optimal drug concentrations for each compound were selected to treat the cells for 24 h. The cells were then collected and stained according to the instructions of the Annexin V‐FITC/PI apoptosis kit (Cat. No. C1062M). Subsequently, the proportion of apoptotic cells was recorded using flow cytometry. Statistical analysis was performed on these proportions to assess the effect of the drugs on cell apoptosis.

2.4.4. Detection of Major Protein Expression by Western Blotting

Based on the IC₅₀ and apoptosis results, the most effective drug (at two concentrations) was selected for a 24 h treatment. First, the cells were collected, and total protein was extracted. The protein samples were then subjected to SDS‐PAGE electrophoresis and transferred onto a PVDF membrane. Next, the membrane was incubated overnight with primary antibodies (PI3K, Cell signaling TECHNOLOGY, CAS: 4249S; Phospho‐PI3K, Cell signaling TECHNOLOGY, CAS: 4228S; pAKT, Cell signaling TECHNOLOGY, CAS: 4060S; Bcl‐2, Cell signaling TECHNOLOGY, CAS: 15071S; Caspase‐3, Cell signaling TECHNOLOGY, CAS: 9662S; Caspase‐9, Cell signaling TECHNOLOGY, CAS: 9502T; GAPDH, Beyotime, CAS: AF0006; β‐Tubulin, Beyotime, CAS: AF2835) corresponding to the core targets, followed by a colorimetric reaction using secondary antibodies (Goat anti‐Rabbit IgG (H+L), Beyotime, CAS: A0208; Goat anti‐Mouse IgG (H+L), Beyotime, CAS: A0216). Finally, the gray values of the protein bands were observed and recorded using a high‐sensitivity chemiluminescence imaging system (RRID: SCR_019037) to evaluate the drug's impact on the expression levels of key proteins, thereby validating the network pharmacology pathway results in this study.

2.4.5. Detection of Major Protein mRNA Expression by qRT‐PCR

Total protein was extracted from A549 lung cancer cells (RRID: CVCL_0023) using SparkZol reagent after drug treatment. Cells were divided into six groups: control, low‐dose quercetin, high‐dose quercetin, IGF‐1 alone, IGF‐1 + quercetin low‐dose, and IGF‐1 + quercetin high‐dose. For each group, 1 µg of RNA was quantified, and cDNA was synthesized using a reverse transcription kit and amplified by PCR according to the manufacturer's protocols. GAPDH was used as an internal reference to calculate the relative expression of the target genes. The primer sequences employed in this analysis are listed in Table 1 [17].

TABLE 1.

List of primer.

Species	Primer name	Sequence (5’‐3’)	Alkaline base
Mus musculus	PI3K(F)	GTCAGTTCCCAAGTTATTTCACAG	24
	PI3K(R)	AAGGCAGAAGGCACAGGTC	19
	Akt(F)	TCAGGATGTGGATCAGCGAGAGTC	24
	Akt(R)	AGGCAGCGGATGATAAAGGTGTTG	24
	GAPDH(F)	GGTGGAGCCAAAAGGGTCAT	20
	GAPDH(R)	GGTCATGAGCCCTTCCACAA	20

2.5. Statistical Analysis

The results are reported as the average value ± standard deviation (SD) based on repeated measurements. Statistical analysis was conducted using the Student's t‐test through the SPSS 20.0 software (IBM, New York, USA). A p value of less than 0.05 was considered statistically significant, with a p‐value of less than 0.01 indicating a higher level of significance.

3. Results

3.1. Clinical Data Analysis Results

3.1.1. Basic Information of Lung Cancer Patients

This study collected a total of 45 cases of lung cancer patients, including 28 males and 17 females. Sixteen patients were under 60 years old, and 29 were aged 60 years or above. All patients received TCM treatment for at least 4 weeks. Regarding tumor stage, there were 10 cases at stage I, 12 at stage II, 15 at stage III, and 8 at stage IV. Histologically, adenocarcinoma was identified in 20 cases, squamous cell carcinoma in 15 cases, small cell carcinoma in 8 cases, and other types in 2 cases (both confirmed as non‐small cell lung cancer). Six patients had a disease duration of 1 year or more, and 39 patients had a duration of less than 1 year, with one patient experiencing relapse for 6 months. Regarding treatment status, 25 patients were treatment‐naïve, 12 were post‐surgery, and the remaining patients did not receive chemotherapy or targeted therapy but were treated with TCM alone.

A total of 45 individualized TCM formulas were collected from lung cancer patients, each consisting of a unique combination of herbs with specified dosages. Across all prescriptions, 188 distinct herb species were identified (cumulative unique herbs after removing duplicates). Based on these 45 clinical prescriptions, we further selected one representative prescription that best reflects the overall medication rules for analysis. This prescription is composed of 21 TCM, namely prepared ephedra, pinellia tuber, prepared white mulberry root‐bark, prepared stemona root, prepared aster root, prepared coltsfoot flower, peucedanum root, stir‐fried apricot kernel, platycodon root, dwarf lilyturf root, honeysuckle flower, thunberg fritillary bulb, snakegourd fruit, Chinese white olive, fragrant solomonseal rhizome, figwort root, astragalus root, codonopsis root, largehead atractylodes rhizome, licorice root and tendrilled fritillary bulb. In this prescription, prepared ephedra and pinellia tuber serve as the monarch drugs to diffuse the lung and resolve phlegm; prepared white mulberry root‐bark, prepared stemona root, prepared aster root, prepared coltsfoot flower, peucedanum root, stir‐fried apricot kernel, platycodon root, dwarf lilyturf root, honeysuckle flower, thunberg fritillary bulb, snakegourd fruit, Chinese white olive, fragrant solomonseal rhizome and figwort root act as the minister drugs to moisten the lung to relieve cough, clear heat and remove toxins; astragalus root, codonopsis root and largehead atractylodes rhizome function as the assistant drugs to reinforce qi and strengthen the body; licorice root and tendrilled fritillary bulb serve as the guide drugs to harmonize all the ingredients, moisten the lung and resolve phlegm, which collectively embodies the core therapeutic principles of “diffusing the lung and resolving phlegm, moistening the lung to relieve cough, and reinforcing qi and nourishing yin”.

3.1.2. Analysis of Medication Use

In terms of medication administration to patients, a total of 188 types of TCMs were involved among 45 patients. Among them, 21 TCMs were used more than 10 times. The top 10 most frequently used herbs were licorice root, atractylodes rhizome, codonopsis root, white peony root, astragalus root, Cremastra seemannii rhizome, Poria cocos, Ophiopogon japonicus tuber, fried malt, and Scutellaria root. The dosage of licorice was concentrated between 10∼20 g, that of Atractylodes rhizome between 10∼20 g, Codonopsis pilosula between 10∼20 g, white Paeonia lactiflora between 10∼20 g, astragalus root mainly at 10 or 30 g, Shanci mushroom at 10 or 15 g, P. cocos mainly at 15 or 20 g, O. japonicus at 10 or 15 g, fried malt at 15 or 20 g, and S. baicalensis at 10 or 15 g. The research findings are presented in Table 2. Furthermore, the TCM herbs with a usage frequency of over 10 will be included in the network pharmacological analysis for further investigation.

TABLE 2.

The Chinese herbal medicines with a usage frequency of 10 or more.

Ranking	TCMs	Number of occurrences	Dosage (frequency)	Total proportion
1	Licorice	33	10 (16), 20 (8), 15 (6), 20 (1), 6 (1), and 9 g (1)	4.1%
2	Atractylodes rhizome	25	20 (11), 10 (6), 15 (3), 30 (3), 15 (1), and 25 g (1)	3.1%
3	C. pilosula	24	20 (8), 10 (7), 12 (3), 15 (3), 30 (2), and 40 g (1)	3.0%
4	P. lactiflora	23	20 (5), 10 (4), 15 (4), 30 (4), 25 (3), 40 (2), and 12 g (1)	2.8%
5	Astragalus root	23	30 (8), 10 (6), 20 (4), 40 (3), and 50 g (2)	2.8%
6	Shanci mushroom	20	15 (10), 10 (6), 15 (1), 12 (1), 18 (1), and 20 g (1)	2.5%
7	P. cocos	17	20 (8), 15 (7), 10 (1), and 25 g (1)	2.1%
8	O. japonicus	17	10 (7), 15 (5), 25 (3), 20 (1), and 20 g (1)	2.1%
9	Fried malt	15	15 (4), 20 (4), 30 (3), 25 (2), 25 (1), and 10 g (1)	1.8%
10	S. baicalensis	15	15 (8), 10 (6), and 9 g (1)	1.8%
11	Turtle shell	14	20 (8), 25 (4), 10 (1), and 15 g (1)	1.7%
12	Tortoise plastron	14	30 (6), 25 (5), and 20 g (3)	1.7%
13	C. cassia	14	10 (5), 20 (3), 25 (3), 15 (1), 30 (1), and 9 g (1)	1.7%
14	Bee nests	13	15 (7), 10 (3), 12 (1), 8 (1), and 9 g (1)	1.6%
15	Radix bupleuri	12	12 (6), 10 (2), 15 (2), and 25 g (2)	1.5%
16	Chenpi	12	10 (6), 15 (3), 12 (1), 25 (1), and 6 g (1)	1.5%
17	Balloonflower	12	10 (8), 15 (2), and 9 g (2)	1.5%
18	Chinese angelica	11	15 (3), 20 (3), 10 (2), 12 (2), and 30 g (1)	1.3%
19	Fructus Aurantii Immaturus	11	15 (7), 10 (2), and 12 g (2)	1.3%
20	Oyster	11	25 (4), 15 (3), 20 (3), and 30 g (1)	1.3%
21	Magnolia officinalis	10	15 (5), 20 (3), and 25 g (2)	1.2%

3.2. Network Pharmacology Analysis Results

3.2.1. Major TCMs and Disease Target Screening

We included the TCM herbs with a usage frequency of more than 10 times (identified in the previous section) for further in‐depth analysis of their active components and potential drug targets. Based on the stringent screening criteria of OB > 30% and DL > 0.18, we meticulously screened the potential active ingredients of these traditional Chinese medicines. As a result, the top seven active ingredients with the highest proportion were beta‐sitosterol, stigmasterol, sitosterol, kaempferol, isorhamnetin, mairin and quercetin (Table 3). Among these, quercetin exhibited the highest number of corresponding targets, whereas sitosterol showed the fewest (only 3 targets). Subsequently, we further analyzed the relevant targets of these seven key active ingredients. The detailed target information for each active ingredient is summarized in Table 3.

TABLE 3.

Information of active ingredients and related targets from several TCM (Fufang CM).

Mol ID	Mol name	Mol structure	Related targets	OB(%)	DL
MOL000358	Beta‐sitosterol		PGR、NCOA2、PTGS1、PTGS2、HSP90AA1、KCNH2、PRKACA、DRD1、CHRM3、CHRM1、SCN5A、CHRM4、PDE3A、ADRA1A、CHRM2、ADRA1B、ADRB2、CHRNA2、SLC6A4、OPRM1、CHRNA7、BCL2、BAX、CASP9、JUN、CASP3、CASP8、PRKCA、PON1、MAP2	36.91	0.75
MOL000449	Stigmasterol		PGR、NR3C2、NCOA2、RXRA、NCOA1、PTGS1、PTGS2、ADRA2A、SLC6A2、SLC6A3、ADRB2、AKR1B1、PLAU、LTA4H、MAOB、MAOA、PRKACA、CTRB1、CHRM3、CHRM1、ADRB1、SCN5A、ADRA1A、CHRM2、ADRA1B、CHRNA7	43.83	0.75
MOL000359	Sitosterol		PGR、PNRC2、NR3C2	36.91	0.75
MOL000422	Kaempferol		NOS2、PTGS1、AR、SCN5A、PTGS2、ESR2、DPP4、HSP90AA1、CHEK1、PRSS1、NCOA2	50.83	0.29
MOL000354	Isorhamnetin		NOS2、PTGS1、ESR1、AR、PPARG、PTGS2、PTPN1、ESR2、DPP4、MAPK14、GSK3B、HSP90AA1、PRKACA、PRSS1、CCNA2、NCOA2、PYGM、PPARD、CHEK1、AKR1B1、NCOA1、F7、F2、ACHE、MAOB、GRIA2、RELA、NCF1、OLR1	49.60	0.31
MOL000211	Mairin		PGR	55.38	0.78
MOL000098	Quercetin		PTGS1、AR、PPARG、PTGS2、F2、HSP90AA1、NCOA2、DPP4、AKR1B1、PRSS1、TOP2A、KCNH2、SCN5A、F10、ADRB2、MMP3、PRKACA、F7、RXRA、ACHE、MAOB、RELA、EGFR、AKT1、CCND1、BCL2、BCL2L1、FOS、CDKN1A、EIF6、BAX、CASP9、PLAU、MMP2、MMP9、MAPK1、IL10、RB1、TNFSF15、JUN、IL6、AHSA1、CASP3、TP63、ELK1、NFKBIA、ODC1、CASP8、TOP1、RAF1、SOD1、PRKCA、MMP1、HIF1A、STAT1、RUNX1T1、ERBB2、PPARG、ACACA、HMOX1、CYP3A4、CAV1、MYC、F3、	46.43	0.28
			GJA1、CYP1A1、ICAM1、IL1B、SELE、VCAM1、CXCL8、PRKCB、BIRC5、DUOX2、NOS3、HSPB1、MGAM、IL2、NR1I2、CYP1B1、CCNB1、PLAT、THBD、SERPINE1、IFNG、ALOX5、IL1A、MPO、TOP2A、NCF1、ABCG2、HAS2、NFE2L2、NQO1、PARP1、AHR、PSMD3、SLC2A4、COL3A1、CXCL11、CXCL2、DCAF5、NR1I3、CHEK2、INSR、CLDN4、PPARA、PPARD、HSF1、CRP、CXCL10、CHUK、SPP1、RUNX2、RASSF1、E2F1、E2F2、ACPP、CTSD、IGFBP3、IGF2、CD40LG、IRF1、ERBB3、PON1、DIO1、PCOLCE、NPEPPS、HK2、RASA1、GSTM1、GSTM2

3.2.2. Construction of PPI and Active Ingredients‐Disease‐Target Network

We compared 200 TCM‐related targets with 4716 lung cancer targets, finding 149 common targets, accounting for a substantial 3.1% of the total by using the Venny online service. These shared targets were recognized as potential candidates for TCM‐based therapeutic strategies for lung cancer, as shown in Figure 1. To delve deeper into the complex PPI among these potential targets, we uploaded them to the STRING platform, generating a PPI network depicted in Figure 2. These core targets were incorporated into the STRING platform, revealing 149 nodes and 323 edges, where nodes represent proteins and edges indicate their interactions, with colors (yellow to blue) reflecting interaction strength. The top‐ranked targets with high relevance include AKT1, TNF, IL6, TP53, PTGS2, and others. To better understand target interactions and identify core targets based on their connectivity, we imported these potential shared targets and their associated data into Cytoscape software. This enabled us to create an active component‐disease‐target network, providing a comprehensive view in Figure 3. Network Analyzer identified four key active ingredients (Isorhamnetin, Kaempferol, Quercetin, and Stigmasterol) and six key targets (AR, FGR, PPARG, PRKACA, PTGS1, and PTGS2), demonstrating the potential of TCM in lung cancer therapy.

FIGURE 1.

Venn diagram of Fufang CM and lung cancer.

FIGURE 2.

PPI network common targets for Fufang CM and lung cancer.Note: the bluer the color, the closer the connection between the target and other targets.

FIGURE 3.

Drug‐ingredient‐target‐disease network diagram.Note: triangles represent gene targets, diamonds represent active ingredients, circles represent drugs, and hexagons represent diseases.

3.2.3. GO and KEGG Enrichment Analysis Results

The potential mechanisms of the 149 candidate targets for lung cancer treatment were explored, identifying 76 related biological processes, molecular functions, and cellular components with a significance threshold of p < 0.05 (Figure 4). These targets were primarily enriched in processes such as positive regulation of transcription by RNA polymerase II, positive regulation of DNA‐templated transcription, positive regulation of gene expression, plasma membrane, cytoplasm, cytosol, protein binding, identical protein binding, and protein homodimerization activity. Besides, KEGG enrichment analysis was conducted to investigate the potential mechanisms of the 175 candidate targets in the Fufang CM treatment protocol for lung cancer, yielding 175 significant items (p < 0.05). The top 23 enriched items are illustrated in a bubble chart and histogram (Figure 5). The results revealed that the co‐action targets were predominantly enriched in diseases such as Pathways in cancer, Chemical carcinogenesis‐receptor activation, PI3K‐Akt signaling pathway, MAPK signaling pathway, AGE‐RAGE signaling pathway in diabetic complications, TNF signaling pathway, Chemical carcinogenesis‐reactive oxygen species, MicroRNAs in cancer, Pathways of neurodegeneration‐multiple diseases. These findings provide valuable insights into the therapeutic mechanisms of Fufang CM in addressing lung cancer.

FIGURE 4.

GO functional enrichment analysis.

FIGURE 5.

KEGG pathway enrichment analysis.

3.3. Bioinformatics Analysis Results

3.3.1. Molecular Docking

To investigate the interactions occurring between receptors and ligands, molecular docking was carried out to assess the top six target proteins and the top four active components within the drug‐ingredient‐target‐disease network. When the binding energy value is below zero, it implies that ligand molecules and receptor proteins bind spontaneously. Conversely, an affinity value lower than −5.0 kJ mol⁻¹ indicates a strong binding affinity between them. In general, a lower binding affinity value signifies a more favorable molecular docking process. The results revealed that all the active ingredients demonstrated robust binding to the target proteins, with their affinities uniformly falling below −5.0 kJ mol^{− 1} (as shown in Table 4). Notably, the top four active ingredients (Isorhamnetin, Kaempferol, Quercetin, and Stigmasterol) displayed exceptionally strong binding affinities to the target proteins. All four exhibited outstanding binding affinity to PRKACA, with affinities not exceeding −9.2 kJ mol⁻¹. This strong binding was mainly facilitated by hydrogen bond interactions with the active site of PRKACA, which greatly enhanced its binding capacity. The heat map illustrating the molecular docking results is presented in Figure 6. These findings underscore the fact that these four primary active ingredients possess strong binding affinities to the target proteins, thereby promoting effective intermolecular interactions.

TABLE 4.

Information of active ingredients and related targets from Fufang CM.

Active Ingredients	Key targets	PDB ID	Affinity (kJ mol−1)	Best‐docked complex (3D) and (2D)
Isorhamnetin	AR	1T65	−8.9
	FGR	7UY0	−8.0
	PPARG	1FM9	−8.2
	PRKACA	3AMA	−9.2
	PTGS1	6Y3C	−8.6
	PTGS2	5F19	−9.0
Kaempferol	AR	1T65	−8.3
	FGR	7UY0	−8.3
	PPARG	1FM9	−9.2
	PRKACA	3AMA	−9.3
	PTGS1	6Y3C	−8.3
	PTGS2	5F19	−8.4
Quercetin	AR	1T65	−8.4
	FGR	7UY0	−8.3
	PPARG	1FM9	−9.2
	PRKACA	3AMA	−9.3
	PTGS1	6Y3C	−8.1
	PTGS2	5F19	−9.0
Stigmasterol	AR	1T65	−1.0
	FGR	7UY0	−9.1
	PPARG	1FM9	−8.2
	PRKACA	3AMA	−10.2
	PTGS1	6Y3C	−7.6
	PTGS2	5F19	−8.7

FIGURE 6.

Heat map of molecular docking results.Note: the redder the color, the smaller the binding energy, and the better the binding; The greener the color, the greater the bonding energy, and the worse the bonding.

3.3.2. Gene Expression Differences Results in Lung Adenocarcinoma

LUAD, as the main histological subtype of non‐small cell lung cancer, accounts for about 40%∼50% of all lung cancer cases [18]. Due to its representative significance in molecular pathogenesis, clinical features, and treatment strategies, in‐depth research on LUAD largely reflects the core content and development direction of overall research on lung cancer. We obtained LUAD mRNA data from TCGA and used the R package DESeq2 to identify DEGs between cancerous and normal tissues (|log₂FoldChange| = 1, p value = 0.05). A volcano plot via ggplot2 showed that core genes “IL6, PGR, PPARG, and IL1B” were downregulated in tumor tissues, possibly implying their protective or regulatory roles in normal lung tissue are disrupted in LUAD, affecting immune response, cell growth/differentiation, and the inflammatory microenvironment. Conversely, “MMP9” was upregulated in tumor tissues, suggesting it may promote tumor invasion and metastasis by degrading the extracellular matrix. These findings offer insights into LUAD's molecular mechanisms and potential therapeutic and diagnostic targets. The heatmap (Figure 7A) clearly illustrates the distinct expression clustering between tumor and normal samples, with the differentially expressed genes effectively separating the two groups. The heatmap and volcano map are shown in Figure 7A,B.

FIGURE 7.

Results of bioinformatics analysis.Note: (A) LUAD differential gene heatmap: gene expression patterns between tumor and normal tissues. (B) LUAD differential gene volcano map: changes in gene expression between tumor and normal tissues. (C) Survival analysis performed on six major genes implicated in cancer prognosis via the KM Plotter database.

3.3.3. Survival Analysis

All lung cancer cases were stratified into high‐ and low‐expression groups based on the expression levels of the five core genes. The prognostic significance of each gene was evaluated through the KM Plotter database. The results revealed that lower expression levels of these genes were generally associated with poorer overall survival in lung cancer patients. Except for PPARG, which did not show a statistically significant association with overall survival, the remaining five genes exhibited a strong correlation with patient prognosis (p < 0.01). These findings suggest that these differentially expressed genes may play critical regulatory roles in lung cancer progression, potentially serving as tumor suppressors. Their downregulation could contribute to tumor aggressiveness and worse clinical outcomes, highlighting their potential as prognostic biomarkers or therapeutic targets in lung cancer management. The details of results were shown in Figure 7C.

3.4. Experiments Results

3.4.1. Results of Cell Proliferation Inhibition Experiment

This study used the CCK‐8 method to examine the inhibitory effects of quercetin, kaempferol, stigmasterol, and isorhamnetin on the proliferation of A549 cells after 24 h of treatment at concentrations ranging from 0.05 to 200 µM. The results showed that all four compounds exhibited concentration‐dependent inhibition of cell viability, with stigmasterol and quercetin demonstrating the most significant inhibitory effects. Stigmasterol showed statistically significant differences compared to the control group at concentrations of 12.5 and 50 µM, with an IC₅₀ value of 16.47 µM. Quercetin also exhibited similarly significant inhibitory effects at concentrations of 50 and 200 µM, with an IC₅₀ value of 115 µM. In contrast, the IC₅₀ values of kaempferol and isorhamnetin were greater than 200 µM, indicating that they did not achieve a 50% inhibition rate within this concentration range, and their inhibitory effects were relatively weaker. In summary, stigmasterol and quercetin may be potential candidate compounds for inhibiting lung cancer cell proliferation. The results are shown in Figure 8A.

FIGURE 8.

The Results of in vitro experiments. Note: (A) Cell viability assay. (B) Flow cytometry results of cell apoptosis. (C) Bar chart showing apoptosis rates. (D) Western blotting results of key proteins. (E) mRNA expression results of PI3K and AKT. ^* p < 0.05, ^** p < 0.01, ^*** p < 0.001, ^**** p < 0.0001.

3.4.2. Results of Cell Apoptosis

The apoptotic effects of quercetin, kaempferol, stigmasterol, and isorhamnetin on A549 cells after 24 h of treatment at low and high concentrations were detected using the Annexin V‐FITC/PI double‐staining method. Flow cytometry analysis revealed that, except for kaempferol, the other three compounds induced apoptosis in a concentration‐dependent manner, with quercetin showing the most significant effect. The apoptosis rate in the low‐concentration group was 26.98% ± 0.96%, and in the high‐concentration group it was 58.35% ± 0.49%, both of which showed statistically significant differences compared with the control group (p < 0.001). Stigmasterol and isorhamnetin exhibited similar trends in inducing apoptosis, with apoptosis rates of 11.85% ± 0.53% and 9.74% ± 0.07%, respectively, in the high‐concentration groups. To further verify whether quercetin‐induced apoptosis is mediated through the PI3K/AKT pathway, we additionally included quercetin combined with IGF‐1 treatment groups. The results showed that, compared with quercetin alone at both low and high concentrations, the apoptosis rates in the combined treatment groups were significantly reduced, and these groups also showed statistical differences compared with the control group, indicating that IGF‐1 partially reversed the pro‐apoptotic effect of quercetin. These findings are consistent with the proliferation inhibition data obtained from the CCK‐8 assay, further supporting the possibility that quercetin suppresses lung cancer cell growth by inducing apoptosis through inhibition of the PI3K/AKT pathway. The results are shown in Figure 8B,C. To further validate this mechanism, we next examined the effects of quercetin on the expression of key proteins and mRNA levels involved in the PI3K/AKT pathway.

3.4.3. Results of Major Protein Expression by Western Blotting and qRT‐PCR

Based on the preliminary network pharmacology KEGG pathway enrichment analysis, the PI3K/AKT signaling pathway was identified as a key mechanism potentially involved in inhibiting lung cancer cell proliferation. To further investigate this, we examined the effects of quercetin on the PI3K/AKT pathway and its downstream apoptosis‐related proteins using Western blotting, in conjunction with the results from CCK‐8 and apoptosis assays. Three experimental groups were included: control, low‐dose quercetin, and high‐dose quercetin. The results showed that quercetin had no significant effect on the total protein expression of PI3K, but dose‐dependently reduced the phosphorylation levels of p‐PI3K and p‐AKT, with a more pronounced inhibitory effect observed in the high‐dose group. Meanwhile, quercetin treatment dose‐dependently upregulated the expression of the anti‐apoptotic protein Bcl‐2, while downregulating the expression of cleaved caspase‐9 and cleaved caspase‐3. These findings suggest that quercetin may inhibit lung cancer cell proliferation by suppressing the activation of the PI3K/AKT signaling pathway, thereby modulating Bcl‐2 family proteins and the caspase cascade. The results The results are shown in Figure 8D.

To determine whether the regulatory effect of quercetin on the PI3K/AKT pathway occurs at the transcriptional level, we performed qRT‐PCR to detect the mRNA expression of PI3K and AKT. Compared with the control group, both quercetin low‐ and high‐dose groups showed significantly reduced mRNA levels of PI3K and AKT in a dose‐dependent manner. Treatment with IGF‐1 alone, an activator of the PI3K/AKT pathway, increased the mRNA expression of PI3K and AKT to approximately 2‐fold that of the control group. In the IGF‐1 + quercetin low‐dose group, the mRNA expression of PI3K and AKT decreased to approximately 0.8∼1, while in the IGF‐1 + high‐dose quercetin group, it rebounded to approximately 1.3. These results indicate that quercetin suppresses PI3K and AKT expression at the transcriptional level, and that IGF‐1 can partially reverse this inhibitory effect, further confirming that quercetin exerts its anti‐lung cancer activity through targeting the PI3K/AKT signaling axis. Collectively, the Western blot and qRT‐PCR results demonstrate that quercetin inhibits the phosphorylation and activation of the PI3K/AKT pathway, upregulates Bcl‐2 expression, and suppresses the activation of caspase‐9 and caspase‐3, thereby inhibiting lung cancer cell proliferation (Figure 8E). Additionally, studies have also indicated that quercetin may exert potential regulatory effects on the PI3K/AKT pathway [19].

4. Discussion

Lung cancer remains one of the most prevalent and lethal malignancies worldwide, and conventional therapies are often limited by drug resistance and severe adverse effects [20, 21]. With its characteristic multi‐component and multi‐target therapeutic approach, TCM has demonstrated promising clinical efficacy in lung cancer treatment and is increasingly regarded as a potential complementary or synergistic strategy. However, it remains unclear whether different TCM formulas share a common pharmacological basis, and the key active constituents responsible for their therapeutic effects—along with their underlying mechanisms—have not been systematically elucidated [22, 23]. To further clarify the shared therapeutic foundation among various clinically used anti‐lung cancer TCM formulas and precisely identify the key bioactive components driving their efficacy, this study employed an integrated approach combining network pharmacology, bioinformatics, and in vitro experimental validation to systematically investigate their core active constituents and potential mechanisms. Specifically, we first analyzed clinical prescription data from lung cancer patients and selected herbal medicines with a usage frequency of >10 times, resulting in the identification of the 21 most commonly used herbs for further analysis. Based on this screening, we then extracted the putative active compounds of these herbs using the TCMSP database, thereby establishing a foundation for subsequent multidimensional mechanistic investigations. This analysis revealed the highest‐frequency bioactive compounds: β‐sitosterol, mairin, sitosterol, quercetin, kaempferol, isorhamnetin, and stigmasterol. Network pharmacology analysis identified the PI3K/AKT signaling pathway as a core regulatory mechanism underlying their anti‐lung cancer effects, consistent with multiple reports linking this pathway to lung cancer treatment [24, 25, 26].

Comprehensive bioinformatics validation included molecular docking between the top four candidate compounds (selected by degree value) and five key target proteins, which demonstrated robust binding interactions. Notably, quercetin, kaempferol, isorhamnetin, and stigmasterol all exhibited strong binding affinities with PRKACA. As the primary catalytic subunit of protein kinase A, PRKACA plays crucial roles in cellular signaling by phosphorylating downstream targets to regulate metabolism, gene expression, and cell proliferation [27]. Previous studies report that PRKACA can modulate PI3K/AKT pathway activity through AKT phosphorylation, with bidirectional crosstalk between these pathways collectively regulating cellular growth and metabolic processes [28, 29]. TCGA database analysis revealed significantly differentially expressed genes in LUAD tissues (including downregulated IL6, PGR, PPARG, and IL1B, and upregulated MMP9), which may influence tumor progression by modulating immune responses, cell differentiation, and extracellular matrix degradation. However, it should be acknowledged that these molecular docking and TCGA‐based predictions have not yet been experimentally validated. In future work, emerging target discovery technologies such as PROTAC (proteolysis‐targeting chimera) probe technology and quantitative proteomic techniques could be employed to further investigate the functional roles of PRKACA and MMP9, as well as the downstream effector proteins modulated by the active compounds [30, 31]. Survival analysis further demonstrated the prognostic significance of these genes in lung cancer patients. Particularly, AR, FGR, PRKACA, PTGS1, and PTGS2 showed strong correlations with patient prognosis [32, 33, 34]. Notably, several of these prognosis‐associated genes overlap with the key targets identified from network pharmacology (e.g., PRKACA, PTGS1, PTGS2). This overlap suggests that the active compounds targeting these genes may have clinical relevance in LUAD. Thus, the network pharmacology predictions are not only supported by differential expression evidence but also aligned with survival outcomes, reinforcing the potential therapeutic value of the identified compounds.

Subsequent in vitro experiments with A549 cells validated the network pharmacology predictions for the four primary active compounds. CCK‐8 and apoptosis assays identified stigmasterol and quercetin as potential anti‐proliferative candidates, with quercetin showing particularly potent inhibitory effects on lung cancer cell growth. Western blot analysis revealed that quercetin exerts its anti‐cancer effects primarily through selective inhibition of PI3K/AKT pathway activation, significantly suppressing PI3K and AKT phosphorylation without altering total protein expression levels. This dose‐dependent inhibition was most pronounced at higher concentrations, providing compelling evidence for quercetin's potential as a targeted therapy. Notably, accumulating preclinical evidence across multiple epithelial cell models demonstrates that quercetin produces minimal cytotoxicity against normal epithelial cells at concentrations sufficient to suppress malignant cell proliferation. Compared with various lung, breast, prostate, and cervical tumor cell lines, matched normal epithelial cells exhibit much higher tolerance to quercetin treatment, which endows this flavonoid with a favorable therapeutic selectivity window [35, 36, 37, 38]. Such differential cytotoxicity between normal and cancer cells has been mechanistically attributed to the distinct baseline activation status of PI3K/AKT and other stress‐related pathways in healthy versus transformed cells. These findings not only validate our initial network pharmacology predictions but also align with existing literature on quercetin's modulation of oncogenic signaling pathways [39, 40, 41]. Although these research advances have expanded our understanding of the mechanisms through which TCM combats cancer, several methodological and conceptual limitations warrant careful consideration. The exclusive focus on A549 cells may not fully capture lung cancer heterogeneity, and the relatively weaker effects observed for kaempferol and isorhamnetin warrant further investigation. Of note, tissue‐specific selectivity and pulmonary biosafety of quercetin still require further verification via parallel cytotoxicity assessments using normal bronchial epithelial cell lines (e.g., BEAS‐2B) under pulmonary microenvironment‐mimicking culture conditions. Additionally, the absence of in vivo pharmacodynamic validation represents a critical gap. Future studies should incorporate broader lung cancer cell lines, normal bronchial epithelial cell controls, clinical pharmacodynamic analyses in patients, and more comprehensive mechanistic investigations to strengthen the translational potential of these findings.

5. Conclusions

To further elucidate the shared therapeutic basis of multiple clinically used TCM formulas for lung cancer and precisely identify the key active components responsible for their efficacy, this study employed an integrated strategy combining network pharmacology, bioinformatics, and in vitro experimental validation to systematically investigate the core bioactive constituents and underlying mechanisms of these formulas. First, clinical prescription data were analyzed to identify frequently used herbs—such as licorice and atractylodes rhizome—and to focus on their common active ingredients. Subsequently, network pharmacology and pathway enrichment analyses pinpointed key compounds (notably quercetin) and their molecular targets, highlighting the central role of signaling pathways such as PI3K/AKT. In vitro experiments confirmed that quercetin significantly inhibits the proliferation of A549 lung cancer cells and effectively suppresses PI3K/AKT pathway activity. Collectively, this study not only clarifies the potential common pharmacological basis and synergistic mechanisms of diverse TCM formulas in lung cancer treatment from a multi‐component, multi‐target perspective, but also provides a scientific foundation for the future development of precision TCM interventions based on key active constituents.

Author Contributions

Bin Yu: conceptualization, methodology, formal analysis, funding, writing – original draft, review and editing, visualization. Jiankun Zhang: conceptualization, visualization, formal analysis, data curation. Lin Ren: methodology, formal analysis, validation, investigation. Kexin Xu: validation. Huan Su: visualization. Yuemeng Zou: formal analysis. Hui Xu: project administration, funding. Guangyao Lv: project administration, supervision, funding, writing – review and editing. Jiangping Yu: supervision, funding, resources, writing – review and editing. All authors have read and agreed to the published version of the manuscript.

Ethics Statement

The studies involving humans were approved by the Biomedical Ethics Committee of Mianyang Central Hospital (approval No. S202603100‐01). The studies were conducted in accordance with the local legislation and institutional requirements. Written informed consent for participation was not required from the participants or the participants' legal guardians/next of kin in accordance with the national legislation and institutional requirements

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.