2025 Volume 267 Issue 2 Pages 177-183
This study elucidated the effect and underlying mechanism of Curcuma aeruginosa Roxb. [C. zedoaria non Rosc.] (CAR) in human papillomavirus (HPV)-related cervical cancer treatment through a network pharmacology approach. Serum containing CAR was prepared using SD rats. The activities of CAR in HPV-related cervical cancer cell viability and migration/invasion were detected by using cell count kit and Transwell assays. Compounds in CAR and their targets were collected from TCMSP and SymMap. The cervical cancer-associated targets were searched from GeneCards and TTD databases. Genes targeted by HPV in human were collected from VISDB database. The networks were constructed using all the targets/compounds or HPV cervical cancer-related targets/compounds. The binding affinity of Furanodiene with Dipeptidyl peptidase IV (DPP4) was determined by the molecular docking method in CB-Dock2. Overexpression of DPP4 was used to discover the effects of dipeptidyl peptidase IV protein on anticancer activity of CAR. CAR-containing serum inhibited the cell viability and migration/invasion of SiHa and Ca Ski cells. Three CAR targets, DPP4, Nitric-oxide synthase, endothelial (NOS3), and Apoptosis regulator Bcl-2 (BCL2) were common with cervical cancer-related genes and HPV-targeted genes. NOS3 was targeted by Furanodiene, BCL2 was targeted by beta-elemene, and DPP4 was targeted by (-)-Epoxycaryophyllene, Zingiberene, Furanodiene, etc. Molecular docking of DPP4 with Furanodiene showed two positions with a Vina score of –6.8. Overexpression of DPP4 reversed the anticancer effects of CAR on HPV-related cervical cancer cells. CAR had inhibitory effects on HPV cervical cancer, possibly by downregulating the expression of DPP4.
Cervical cancer is a common medical burden faced by low- and middle-income countries and high-income countries (Vu et al. 2018). Not surprisingly, low- and middle-income countries are responsible for most of the total new cases worldwide, about 80%, because of poor compliance with screening, treatment, and post-treatment rehabilitation (Mahantshetty et al. 2021). Cervical cancer accounts for 7.5% of all female cancer-related deaths, approximately 90% of which occurred in low- and middle-income countries (Zhao et al. 2021). Cervical cancer can go through a long preclinical stage and can span decades without symptoms affecting women. Therefore, advanced cervical cancer is dominant in low- and middle-income countries, making treatment regimens also more costly (Vu et al. 2018). Progress in cost-effective treatment plans has expanded to traditional herbal medicine (Hsiao et al. 2019). Herbs or herb-based compounds have been verified to be less toxic compared to conventional synthetic compounds (Apolone et al. 2005). Given that herb-based drug formulations usually comprise several phytochemicals or even more than one plant. The main challenge in this direction would be to predict the effects of all phytochemicals or active compounds. Thus, a comprehensive method that can reflect most of the components in herbal drugs, as well as the targets of these components, is needed.
Curcuma aeruginosa Roxb. [C. zedoaria non Rosc.] (CAR), also known as Curcumae Rhizoma in China, is an herb listed in Pharmacopoeia of the People’s Republic of China. It has been widely prescribed for the therapy of cancer, either alone or in combination. Its terpenoids have been verified to have anticancer properties (Lu et al. 2012). Zedoary Turmeric Oil, the volatile oil from CAR, has been studied for its anti-cancer effect in cervical cancer (Jia et al. 2018). Clinically in China, the powder of CAR has been used for external application to the cancer site (Chen and Liu 2014). In addition, CAR has antiviral effects, such as the anti-human immunodeficiency virus effect (Sillapachaiyaporn et al. 2021). However, no herb-target interaction network approaches have been specifically explored for CAR in human papillomavirus (HPV)-related cervical cancer.
In this work, the effect of CAR on HPV-related cervical cancer cells was detected, and the active compounds and potential targets related to HPV cervical cancer were predicted.
After 3 days of acclimatization, the rats (Chengdu Dossy Experimental Animal Co., Ltd, China; SCXK Chuan 2020-030) were randomly divided into the control group and CAR group, with 5 rats in each group. After converting the human clinical dose according to the body surface area, each rat was administrated by gavage with 20 g/kg, twice daily for 8 doses. Two hours after the eighth dose, the rats were anesthetized and the blood was collected. Rat blood was centrifuged at room temperature for 15 min at 3,500 rpm after 30 min of rest. The supernatant was inactivated at 56℃ for 45 min, and filtered through a sterile microporous membrane. The serum was frozen at –20℃ for reserve.
Cell culture and transfectionHPV-related cervical cancer cell lines, including SiHa and Ca Ski, were purchased from ATCC (USA). The cells were cultured in the appropriate recommended medium, including Eagle’s Minimum Essential Medium and RPMI-1640 medium (both Millipore Sigma, USA). A final concentration of 10% fetal bovine serum (Millipore Sigma, USA) was added to the base medium. The plasmid pcDNA human DPP4 (ov-DPP4) and the pcDNA vector were prepared by Amyjet Scientific (China). HPV-associated cervical cancer SiHa and Ca Ski cells were transfected with 0.5 mg of DNA using the Lipofectamine 2000 reagent (Life Technologies, USA).
Cell treatmentSiHa and Ca Ski cells were inoculated in 96-well plates and incubated with 5%, 10%, and 20% CAR-containing serum for 24 hours. The control group was incubated with 10% serum from control rats.
Cell viability assayCell Counting Kit-8 (CCK-8, TargetMol, USA) was used. Treated SiHa and Ca Ski cells were made into a suspension and then 100 µL of it was added to a 96-well plate. Then, 10 µL of CCK8 detection reagent was added to each well of the plate, and incubation was continued for 2 hours. After incubation, the plates were shaken on a shaker for about 1 min. The cell activity was calculated by reading the 450 nm absorbance value.
Cell viability = [(experimental group - blank group) / (control group - blank group)] × 100%
Cell migration assaySiHa and Ca Ski cell migration was assayed using 8 μm pore, 6.5 mm-transwell inserts in a 24-well plate (Corning, USA). Treated SiHa and Ca Ski cells (4 × 104 cells per well) were subjected to serum starvation for 12 hours. Starved cells were placed in the upper chamber, while a medium containing 15% fetal bovine serum was added to the lower chamber. After 24 hours, Transwell inserts were stained and analyzed by manual cell counts from captured images (Olympus BX41 microscope with an Olympus DP70 camera attachment).
Cell invasion assayThe Transwell inserts were coated using Matrigel (Becton-dickinson, USA). Then, the procedures were the same as those in the cell migration assay.
Collection of chemicals and targets for CARThe chemicals from CAR were collected from Traditional Chinese Medicine Systems Pharmacology Database and Analysis Platform (TCMSP) (https://www.tcmsp-e.com/#/home) and SymMap (http://www.symmap.org). All these chemicals were subjected to target prediction using TCMSP database and TTD database (https://idrblab.net/ttd/).
Collection of HPV cervical cancer genesThe genes related to cervical cancer were collected from TTD database and GeneCards database (https://www.genecards.org/). The genes related to HPV were collected from Viral Integration Site DataBase (VISDB) database (https://bioinfo.uth.edu/VISDB).
Construction of networksTo facilitate scientific interpretation of the relationship between compounds involved in CAR, compounds-target network analysis was performed. The procedure for these network constructions was conducted in two perspectives as follows: (1) the vivo “all compound-all target network’’ was constructed by linking all compounds in CAR and all their candidate targets. (2) the HPV cervical cancer-related “compounds-candidate target network’’ was constructed by linking the HPV cervical cancer-related genes to the corresponding compounds. These two networks were generated by Cytoscape 3.7.1 version which is an open-source software project for network analysis and visualization.
Molecular dockingThe binding affinity of furanodiene with DPP4 was determined by the molecular docking method, which was performed using the blind docking method in CB-Dock2 (Liu et al. 2022). For molecular docking preparation, the high-resolution three-dimensional X-ray crystal structure of DPP4 (6B1E) was retrieved from protein data bank (PDB) (http://www.rcsb.org/), and furanodiene 3D structure was downloaded from PubChem database (https://pubchem.ncbi.nlm.nih.gov/). PyMol was used to remove ligands and heteroatoms in DPP4 (6B1E).
Analysis of DPP4 mRNA and protein expressionThe differences in DPP4 mRNA expression in different cells were detected using real-time PCR. Total RNA was extracted from cells using RNeasy mini-Kits (Qiagen in Japan, Japan) and reverse-transcribed using a Prime Script RT reagent Kit (Takara Bio Inc., Japan). Thermal cycle was run on a Roche LC 480II qRT-PCR (Roche, Germany) using a SYBR Premix Ex Taq Kit (Takara). The ct values were calculated by 2−ΔΔCt algorithm by normalization to the expression of β-actin.
The protein levels of DPP4 were detected using a Western blot. Proteins in cells were extracted after cell wash and cell lysis. Equivalent proteins (15 μL) were loaded into a NuPAGE Bis-Tris gels (Thermo Fisher, USA). The protein separation was performed in an SDS running buffer (Thermo Fisher, USA) in electrophoresis. The protein bands were then transferred to a nitrocellulose membrane. The nitrocellulose membrane was incubated with the human DPPIV/CD26 antibody (1:000, R&D systems #AF1180, USA) or the human β-actin (1:10,000; Sigma, USA) antibody. After three washes, the protein bands were incubated using secondary antibodies. The proteins were visualized using an enhanced chemiluminescence detection system (Amersham, UK) and analyzed using ImageJ software.
Statistical analysisp values were computed with a two-way Analysis of Variance, using GraphPad Prism.
To screen the in vitro inhibitory activity of CAR on HPV-related cervical cancer, CAR-containing serum were tested at the concentration of 5%, 10%, and 20%. Compared the control serum group, the viability of SiHa and Ca Ski cells was significantly reduced (p < 0.01 or p < 0.001) after incubation with CAR-containing serum (5%, 10%, and 20%) (Fig. 1A). Compared with the control group serum culture, the serum containing CAR (15%, 10%, and 20%) inhibited the migration (p < 0.001) (Fig. 1B) and invasion ability (p < 0.01 or p < 0.001) of SiHa and Ca Ski cells (Fig. 1C).

The anticancer effect of Curcuma aeruginosa Roxb. on human papillomavirus (HPV)-related cervical cancer.
Inhibitory activities of Curcuma aeruginosa Roxb. on SiHa and Ca Ski cell proliferation (A), migration (B), and invasion (C) were detected by using Cell Counting Kit-8 and Transwell assays. **p < 0.01, ***p < 0.001.
After TCMSP screening, 62 unduplicated chemical constituents and 87 unduplicated targets were obtained, which included 583 compound-target relationships (Fig. 2A). Among the targets, three targets, Dipeptidyl peptidase IV (DPP4), Nitric-oxide synthase, endothelial (NOS3), and Apoptosis regulator Bcl-2 (BCL2) were common with cervical cancer-related genes and HPV-targeted genes (Fig. 2B). Among the three targets, NOS3 was targeted by Furanodiene, BCL2 was targeted by beta-elemene, and DPP4 was targeted by (-)-Epoxycaryophyllene, Zingiberene, Furanodiene, (5R,6R)-5-isopropenyl-3,6-dimethyl-6-vinyl-5,7-dihydrobenzofuran-4-one, Terpilene, (1S,3E,7E,11S)-1,5,5,8-tetramethyl-12-oxabicyclo[9.1.0]dodeca-3,7-diene, (6R)-2-methyl-6-(4-methylphenyl) hept-2-en-4-one (Fig. 2C). After analysis via NetworkAnalyzer plugin of Cytoscpe software, degree of Furanodiene ranked first in compounds, while degree of DPP4 ranked first in targets (Fig. 2C, Table 1). Then, molecular docking of DPP4 with Furanodiene showed two positions with Vina score of –6.8 (Fig. 3).

Compound-target-pathway networks.
(A) Compound-target-pathway networks of the key targets of all compounds in Curcuma aeruginosa Roxb. (B) The Venn diagram showed numbers of targets found in TCMSP, GeneCards/TTD, and VISDB. (C) Herb-compound-target-disease network of Curcuma aeruginosa Roxb. against human papillomavirus-related cervical cancer (HPV-CSCC).

The degree of network factors based on NetworkAnalyzer plugin of Cytoscpe software.
(6R)-MMPEO, (6R)-2-methyl-6-(4-methylphenyl) hept-2-en-4-one; (1S,3E,7E,11S)-TMOCDD, (1S,3E,7E,11S)-1,5,5,8-tetramethyl-12-oxabicyclo [9.1.0] dodeca-3,7-diene; (5R,6R)-IPDMVDHO, (5R,6R)-5-isopropenyl-3,6-dimethyl-6-vinyl-5,7-dihydrobenzofuran-4-one; HPV-CSCC, human papillomavirus-related cervical cancer.

Molecular docking positions of Furanodiene with DPP4 using blind docking method in CB-Dock2.
The DPP4 displayed binding affinity for Furanodiene at two different pockets (Vina Scores: -6.8 kcal/mol, respectively).
Expression level of DPP4 mRNA was increased in cervical cancer SiHa and Ca Ski cells (Fig. 4A). Considering the important role of DPP4 in our network, the expression change of DPP4 in CAR-treated cells was investigated. It appeared that the expression of DPP4 was reduced in CAR-treated SiHa and Ca Ski cells, whereas DPP4 pcDNA increased the DPP4 expression again (p < 0.001) (Fig. 4B). The same trends were found in the levels of DPP4 protein (Fig. 4C). Regarding cell viability, SiHa and Ca Ski cell viability was decreased upon CAR treatment, while DPP4 overexpression increased their viability (p < 0.001) (Fig. 4D). Moreover, DPP4 overexpression rescued the reduced migration (p < 0.001) (Fig. 4E) and invasion (p < 0.05 or p < 0.001) (Fig. 4F) caused by CAR treatment.

DPP4 was a target of Curcuma aeruginosa Roxb. (CAR) against human papillomavirus (HPV)-related cervical cancer.
(A) Expression of DPP4 mRNA was detected in cervical cancer cells (SiHa and Ca Ski) and cervical canal cells (End1/E6E7 and Ect1/E6E7). (B) Expression of DPP4 mRNA was detected in cervical cancer cells (SiHa and Ca Ski) after CAR treatment. (C) Protein levels of DPP4 were detected in cervical cancer cells (SiHa and Ca Ski) after CAR treatment by Western Blot. DPP4 overexpression reversed the inhibitory activities of Curcuma aeruginosa Roxb. on SiHa and Ca Ski cell proliferation (D), migration (E), and invasion (F) were detected by using Cell Counting Kit-8 and Transwell assays. *p < 0.05, ***p < 0.001.
Cervical cancer is one of the leading malignancies among women. Chemotherapy is the mainstream for cancer of the cervix before or after surgery. However, chemoresistance is a huge bottleneck in the treatment of cervical cancer. Given this bottleneck, drug research and development has gradually changed from the “single target and single drug” mode to the “network-target, multiple-component-therapeutics” mode (Zhang et al. 2019). In terms of treating diseases, the overall philosophy of Traditional Chinese Medicine and the multi-component and multi-target characteristics of herbal medicine share key concepts with new modes of drug discovery. Therefore, fully excavating the therapeutic mechanism of traditional Chinese medicine is of great significance to drug research. In China, CAR can be prescribed for cervical cancer in various forms, including powder of crude herb, suppository, injection, or decoction combined with others (Wang and Liu 2014). Because of this, each compound of CAR would be pharmacologically active. Therefore, this study explored the therapy targets of CAR based on all compounds and drug-like compounds respectively, and then screened the targets of CAR against HPV cervical cancer.
Modern trends in traditional medicine informatics supply opportunities for studying anticancer plant products by using in silico and systems pharmacology tools (Feng et al. 2024). Systems biology, such as network pharmacology (Zhang et al. 2019; Zhou et al. 2020), is a comprehensive method that could reflect the variation of herbs and constituents in crude drugs. As information technology and bioinformatics developed, a lot of resources and databases that report herbal formulations, active components of the herb, and related target gene information have been built. There are several efforts in TCMSP (https://old.tcmsp-e.com/load_intro.php?id=27) (Ru et al. 2014) and SymMap (Wu et al. 2019). Several herbs have been evaluated using a network pharmacology tool to explore their anticancer compounds (Zhang et al. 2020; Yuan et al. 2021). The identification of compound-target interactions is crucial to better understand the mechanism of herb action at the molecular level. However, the huge number of pairwise interactions in an herb is the main obstacle to experimentally screening all the possible interactions. Network approaches are faster and can accurately discriminate potential compound-target interactions. Therefore, the network pharmacology methods appear to be an interesting alternative to provide supporting evidence for experimental studies (Luo et al. 2020). This study searched from TCMSP, obtaining the whole and the drug-like compounds, as well as their corresponding targets. It is known that elemenes, including β-elemene (MOL000908), γ-elemene (MOL000037), and δ-elemene (MOL000925), have been developed to be injection and oral emulsion to treat cancers (Zhi et al. 2009; Jiang et al. 2017). Among the compounds of CAR, hederagenin (MOL000296) has been verified to inhibit proliferation but induce apoptosis of cervical cancer cells by inhibiting STAT3 pathway (Fang et al. 2019). Demethoxycurrcumin (MOL000887) can suppress cervical cancer metastasis via NF-κB pathway (Lin et al. 2018). Isocurcumenol (MOL000889) and curcumol (MOL000902) have antitumor effects on human cancer cells (Lakshmi et al. 2011; Hashem et al. 2021). Therefore, the chemical constituents of CAR have potent anticancer potential.
The identification of the targets would envision an understanding of the mechanism for the function and behavior of herbs. It has been reported that CAR exhibits cytotoxicity in HeLa cells by inducing apoptosis (Zohmachhuana et al. 2022). In this study, after overlapping with the HPV cervical cancer-related genes, the target genes of CAR against HPV cervical cancer included NOS3, BCL2, and DPP4. The association of NOS3 with HPV-infected cervical cancer is perceived with the 4a allele and shows a protective effect (Inácio et al. 2023). HPV16 E2, an important early protein of HPV16, can increase the expression of Bcl-2, leading to a decrease in E2-induced apoptosis (Prabhavathy et al. 2014). Overexpression of DPP4 can promote cell migration and proliferation in cervical cancer cells (Niazmand et al. 2024). Here, we found that CAR can target DPP4 in cervical cancer.
In summary, this study revealed that CAR had inhibitory effects on HPV cervical cancer in a multi-target manner. DPP4 may be the hub target of CAR against HPV cervical cancer.
The authors declare no conflict of interest.