Integrated analysis of Dendrobium nobile extract Dendrobin A against pancreatic ductal adenocarcinoma based on network pharmacology, bioinformatics, and validation experiments

Background: Dendrobium nobile (D. nobile), a traditional Chinese medicine, has received attention as an anti-tumor drug, but its mechanism is still unclear. In this study, we applied network pharmacology, bioinformatics, and in vitro experiments to explore the effect and mechanism of Dendrobin A, the active ingredient of D. nobile, against pancreatic ductal adenocarcinoma (PDAC). Methods: The databases of SwissTargetPrediction and PharmMapper were used to obtain the potential targets of Dendrobin A, and the differentially expressed genes (DEGs) between PDAC and normal pancreatic tissues were obtained from The Cancer Genome Atlas and Genotype-Tissue Expression databases. The protein-protein interaction (PPI) network for Dendrobin A anti-PDAC targets was constructed based on the STRING database. Molecular docking was used to assess Dendrobin A anti-PDAC targets. PLAU, one of the key targets of Dendrobin A anti-PDAC, was immunohistochemically stained in clinical tissue arrays. Finally, in vitro experiments were used to validate the effects of Dendrobin A on PLAU expression and the proliferation, apoptosis, cell cycle, migration, and invasion of PDAC cells. Results: A total of 90 genes for Dendrobin A anti-PDAC were screened, and a PPI network for Dendrobin A anti-PDAC targets was constructed. Notably, a scale-free module with 19 genes in the PPI indicated that the PPI is highly credible. Among these 19 genes, PLAU was positively correlated with the cachexia status while negatively correlated with the overall survival of PDAC patients. Through molecular docking, Dendrobin A was found to bind to PLAU, and the Dendrobin A treatment led to an attenuated PLAU expression in PDAC cells. Based on clinical tissue arrays, PLAU protein was highly expressed in PDAC cells compared to normal controls, and PLAU protein levels were associated with the differentiation and lymph node metastatic status of PDAC. In vitro experiments further showed that Dendrobin A treatment significantly inhibited the proliferation, migration, and invasion, inducing apoptosis and arresting the cell cycle of PDAC cells at the G2/M phase. Conclusion: Dendrobin A, a representative active ingredient of D. nobile, can effectively fight against PDAC by targeting PLAU. Our results provide the foundation for future PDAC treatment based on D. nobile.


Introduction
Pancreatic ductal adenocarcinoma (PDAC) is a highly aggressive malignancy with poor clinical prognosis and treatment outcomes, and its incidence and mortality rates have been increasing (GBD, 2017Pancreatic Cancer Collaborators, 2019. Near 80%-90% of patients who suffered from PDAC had unresectable tumors or metastatic disease at the time of the first diagnosis, with a 5-year overall survival (OS) rate of 10%, despite the development of the radical surgical treatment. Although gemcitabine is a recognized first-line chemotherapy drug for PDAC, it is still unclear whether there is an advantage for survival (Zhang et al., 2022). Therefore, new cancer treatments are urgently needed for the treatment of metastatic or incurable PDAC (Von Hoff et al., 2013;Chen et al., 2020).
Cancer cachexia is a multifactorial combination of disease symptoms characterized by muscle wasting, which can lead to significant weight loss and affect patients' quality of life, treatment-tolerant response, and survival (Vaughan et al., 2013). Freire et al. (2020) reported that variants in cachexia-inducible factors and their associated regulatory genes had the highest frequency in pancreatic cancer patients, and the expression levels of key cachexia genes had predictive prognostic value in pancreatic cancer. Meanwhile, specifically targeting PLA2G7 can alleviate cachexia levels in tumor-bearing mice (Morigny et al., 2021). These findings suggest that identification of cachexia-related factors and targeted intervention of cachexia play an important role in the treatment of malignant tumors, such as pancreatic cancer.
The use of traditional Chinese medicine (TCM) in pancreatic cancer treatment is gradually attracting the attention of clinicians. Dendrobium nobile (D. nobile), as a typical representative of TCM, has recently been reported in the field of anti-tumor drugs. Reports have increasingly shown that bibenzyl compounds extracted from D. nobile displayed promising effects against cancers derived from the lung (Losuwannarak et al., 2020;Putri et al., 2021), skin (Cardile et al., 2020), bladder , liver (Chen et al., 2017), and breast (Yu et al., 2018). Dendrobin A (4,5-dihydroxy-3,3′dimetheoxybibenzyl) is a new bibenzyl, and there is a lack of studies related to its role in pancreatic cancer; therefore, further elucidation of the anti-pancreatic cancer effect of Dendrobin A and other Dendrobium species is of great significance for the development of TCM anti-tumor therapy.
Network pharmacology is a novel approach that integrates systems biology and bioinformatics methods to predict drug targets using biological networks and big data technologies (Hopkins, 2008;Nogales et al., 2022), which has transformed the drug mechanism of action from a single-target, single-drug model to a network-target, multi-component therapeutic model, and has shown exciting results, especially in the field of TCM target prediction . In this study, we integrated PDAC transcriptome expression data by network pharmacology technology and preferentially selected drug targets based on biological network characterization and survival analysis. Additionally, we performed in vitro experiments to validate the effects of Dendrobin A extracted from D. nobile on PDAC cells. Our results provide novel insights into Dendrobin A in treating PDAC and support the use of D. nobile in conventional medicine.

Differential expression genes (DEGs) of PDAC patients
Due to the limitation of adjacent normal samples in The Cancer Genome Atlas (TCGA) datasets, we performed DEG analysis by integrating the transcriptome of PDAC tumor samples in TCGA and normal pancreas tissues in the Genotype-Tissue Expression (GTEx) datasets according to the algorithm proposed in a previous report (Vivian et al., 2017). Data were obtained from the "TCGA TARGET GTEx" cohort in UCSC Xena (http://xena.ucsc.edu/), including sequencing counts and Transcripts Per Million (TPM) normalized expression matrices. All data were normalized and batch-corrected; the datasets used in this study included 178 cancer and 167 normal samples. The R package edgeR 3.30.3 (Robinson et al., 2010) was applied for DEGs analysis based on the gene read counts matrix. Only genes with |log2FC| > 2 and FDR <0. 05 were considered significantly differentially expressed. An external dataset of PDAC was also downloaded from the GEO database (https://www.ncbi.nlm.nih.gov/geo/) with accession number GSE62452 (Yang et al., 2016), which included 69 pancreatic tumor and adjacent non-tumor tissues to validate the DEGs and clinical significance.

Construction and topological analysis of protein-protein interaction (PPI) network
The putative target genes of Dendrobin A and DEGs in PDAC samples were overlapped to identify the shared target genes for Dendrobin A in treating PDAC. These essential genes were next input into the STRING V11.5 (Szklarczyk et al., 2019) to construct the PPI network with default parameters. We performed a Pearson correlation analysis for each gene pair of the PPI network in PDAC samples to obtain the actual relationships in the tumor context. Correlation coefficients above 0.2 and p-value <0.05 were identified as significant. The PDAC-related PPI network was visualized using Cytoscape V3.7.2 (Shannon et al., 2003). The MCODE plugin was applied to identify the network modules. The PPI network's degree distribution, clustering coefficient, and characteristic path length were analyzed using R package igraph 1.2.6. To explore the potential biological function of the PPI network, we input the genes to Metascape (https://metascape.org/gp/index.html#/main/step1) with the setting of species ("Homo sapiens") (Zhou et al., 2019).

Survival prognosis analysis
The survival analysis of OS for essential genes of PDAC tumors in TCGA cohorts was performed through GEPIA2 (http://gepia2. cancer-pku.cn/) (Tang et al., 2019). The median expression of genes was used to divide the patients into high-expression and lowexpression groups. Then, the OS of these groups was compared by log-rank test.

Estimation of the association between genes and cachexia
The information on 25 cachexia-inducing factors (CIFs) was collected from a previous study (Freire et al., 2020). The singlesample GSEA (ssGSEA) algorithm (Barbie et al., 2009) was applied to estimate the cachexia score for each sample. Wilcoxon's rank sum tests evaluated the cachexia score between PDAC in TCGA and normal samples in GTEx. Pearson correlation analysis was performed to estimate the association between the expression level of genes and cachexia score.

Molecular docking
In this study, the drug-protein binding patterns were predicted by molecular docking, which was performed using AutoDock Vina 1.1.21 software (The Scripps Research Institute, La Jolla, CA, United States) (Trott and Olson, 2010). Before the start of docking, the PLAU protein crystal structure was obtained from the PDB database with the protein search number 6AG7. The 3D structure of Dendrobin A was obtained from the PubChem database download, and energy minimization of the small molecule was performed using the MMFF94 force field. Before the start of formal docking, the protein was prepared using PyMol 2.5.2 software, including dehydrogenation, dehydration molecules, and non-liganded small molecules. The docking box was then defined to wrap the protein activity pocket. Then, the small molecules in PDB format and the receptor proteins were converted to PDBQT format using ADFRsuite 1.02 (Ravindranath et al., 2015). Finally, the docking work was performed, visualized, and analyzed using PyMol 2.5.2.

Assessment of Dendrobin A on cell proliferation by cell counting Kit-8 (CCK-8) assay
The human PDAC cell lines Panc-1 and Aspc-1, as well as the normal pancreatic cell line HPDE6-C7, were purchased from the Shanghai Chinese Academy of Sciences Cell bank. All cells were tested by short tandem repeat to confirm that there was no cross-contamination, and routine mycoplasma and chlamydia contamination tests were performed before the experiments. Cells were cultured in Dulbecco's modified eagle medium (HyClone, Beijing, China) supplemented with 10% fetal bovine serum (HyClone), 100 μg/mL streptomycin, and 100 U/mL penicillin (Beyotime, Nanjing, China), at 37°C in 5% CO 2 . Dendribin A was dissolved in dimethyl sulfoxide (DMSO), and the concentration of DMSO was <0.1%. The proliferation and cytotoxicity of three pancreatic cell lines were measured by a cell counting kit (CCK-8, Beyotime); the cells were seeded into a 96-well plate (2,000 cells per well) and incubated overnight. Cells were treated with 10,20,30,40,50,60,70,80,90, and 100 µM of Dendrobin A for 24 h at 37°C. After the CCK-8 reagent was added and incubated for 1 h, the optical density was detected at the 450 nm wavelength. Five replicate Frontiers in Pharmacology frontiersin.org wells were used for each group, and experiments were repeated three times.

Clone formation assay
PDAC cells at a density of 3,000 cells/well were seeded into 6-cm dishes overnight. The cells were treated with a medium containing Dendrobin A or DMSO for 24 h, and then the medium was replaced with Dendrobin A-free completed medium and cultured for 15 days. The culture medium was discarded, and then cells were washed twice with PBS, fixed in methanol for 20 min, and air dried after discarding methanol. The cells were stained with 0.3% crystal violet solution for 15 min and washed gently with ddH 2 O to remove the staining solution. The number of clones with >50 cells was counted after photographing. Statistical analysis was performed using Image J software (version 1.8.0). Experiments were repeated three times.
Wound healing assay PDAC cells were seeded on 12-well plates to reach the monolayer, and a linear wound was formed by scraping with a 200 μL micropipette tip, and the cells were treated with Dendrobin A for 24 h. The cell scratch area was observed with a ×20 microscope objective and photographed, and the wound distance was used to assess the cell migration rate. The degree of wound healing was calculated by analysis with Image J software (version 1.8.0). Experiments were repeated three times.

Invasive and migration assay
The ability of PDAC cells to invade and migrate was assayed using the Transwell method. Matrigel gel (BD, China, Shanghai) and serumfree medium were diluted at a ratio of 1:8, and 100 μL per well was added in the upper chamber of the Transwell (Corning, NC, United States) and incubated at 37°C for 3 h. The migration assay was performed without Matrigel gel, and the procedure was the same. 4 × 10 4 Aspc-1 cells and 1 × 10 5 Panc-1 cells were loaded into the upper chamber with 200 µL of serum-free medium, while medium containing 20% serum was added to the lower chamber after 24 h of incubation. The upper chamber was fixed in 4% paraformaldehyde and stained with 0.1% crystal violet. Invaded or migrated cells were counted and photographed, and differences in the number of cells that penetrated the membrane represented the altered cell motility. Statistical analysis was performed using Image J software (version 1.8.0). Experiments were repeated three times.
Flow cytometry analysis of cell cycle and apoptosis PDAC cells were added in 6-well plates allowing recovery overnight, then Dendrobin A was added for 24 h, and the cells were harvested for cell cycle and apoptosis analysis. PDAC cells were stained with propidium iodide (PI) for cell cycle analysis. Cell Cycle Assay Kit-PI/RNase staining (BD, China) was followed by the detection of cell ratios in the G0/G1, S, and G2/M phases. For apoptosis, PDAC cells were stained with Annexin V fluorescein isothiocyanate (FITC) and PI using the Annexin V-FITC Apoptosis Detection Kit (BD, China). Three wells were used for each group. Experiments were repeated three times.

Tissue microarrays and immunohistochemical staining
Human tissue microarrays of PDAC (#HPanA180Su03, Shanghai Outdo Biotech Company, China) were used to detect the expression of PLAU protein. Tissue arrays include 81 cancer tissues and 81 paracancerous tissues from surviving

No.
δ H (J in Hz) δ C , type  (Bai et al., 2022). The scores of PLAU were calculated briefly as staining intensity: 0, no staining; 1, light yellow; 2, brown; and 3, dark brown; staining positive rate: 0%-100%. The total score was calculated as follows: "staining intensity score" × "staining positive rate" ×10. The mean value of a total score of 15.75 for all PDAC samples was used as the grouping criterion, and samples with a score of ≥15.75 belonged to the high-expression group, while samples with a score of <15.75 were included in the low-expression group.

Statistical analysis
For network pharmacology and bioinformatics, the statistical analyses were performed using R software. The associations between clinical features and PLAU expression were evaluated using the Wilcoxon signed-rank test and the chi-square test. Clinical features related to OS in PDAC patients were identified using the Kaplan-Meier method. For in vitro experiments, data were shown as mean ± SD. Statistical analysis was performed with GraphPad Prism 8.0 software, and the significance was analyzed with Student's t-test or ANOVA. p < 0.05 was considered statistically significant.

Results
Extraction, isolation, and identification of Dendrobin A According to our previously described protocols (Cao et al., 2021), we successfully obtained 548.4 mg Dendrobin A from 13.0 kg D. nobile. The structure of Dendrobin A is shown in Figure 1. In addition, the 1 H NMR (500 MHz) and 13 C NMR (125 MHz) data of Dendrobin A in CDCl3 were included in Table 1.

Identification and analysis of Dendrobin A potential targets in PDAC
Based on SwissTargetPrediction and PharmMapper databases, 353 pharmacological targets of Dendrobin A were identified. To explore the expression pattern of these putative drug targets in a tumor context, we performed DEG analysis by integrating the transcriptome of pancreas tumor in TCGA and normal pancreatic samples in GTEx datasets. We identified 3,131 DEGs, comprising 2,358 upregulated and 773 downregulated genes in PDAC samples (Figure 2A). By overlapping 3,131 DEGs and 353 drug targets, we found 90 genes related to Dendrobin A treatment and expressed differently in PDAC samples. Among them, 76 were upregulated, and 14 were downregulated ( Figure 2B). Further functional enrichment analysis indicated that these genes were significantly enriched in several metabolism-related biological processes like collagen catabolic, organic hydroxyl compound metabolic process, and response to the drug ( Figure 2C). For KEGG pathways, cancer-related pathways like transcriptional misregulation, PPAR signaling, and drug metabolism were significantly enriched ( Figure 2D). Next, we obtained 504 genes related to PDAC from MalaCards and intersected with the 90 essential genes mentioned above. We identified 13 shared genes, namely, CCNA2, AURKA, ALB, LCN2, NQO1, CDA, PLAU, MMP7, REG1A, MMP9, LGALS3, MMP2, and PIK3CG ( Figure 2E).

Dendrobin A-related PPI network
The 90 essential Dendrobin A-related DEGs were input into STRING to obtain the interactions of proteins. As a result, a PPI network containing 81 nodes and 299 edges was constructed. Through co-expression analysis and based on the threshold of | Pearson r| > 0.2 & p < 0.05, we found 57.86% (173/299) positive correlation edges and 7.02% (21/299) negative correlation edges in PDAC samples ( Figure 3A). We utilized the hierarchical model that linked genes, biological processes, and previously described cancer hallmarks to determine the association between the PPI network and tumorigenesis (Poulia et al., 2020). Many genes in the PPI network were associated with multiple cancer hallmarks ( Figure 3B). These results suggest that Dendrobin A-related genes play essential roles in cancer processes. Through topological feature analysis, we found that the PPI network displayed scale-free distribution with R 2 = 0.682 (Supplementary Figure S1A), suggesting that the network showed scale-free characteristics. Moreover, the PPI network's clustering coefficient and characteristic path length were  Frontiers in Pharmacology frontiersin.org significantly increased when compared with random networks, as expected for reduced global efficiency and module characteristics (p-value <0.001, Supplementary Figures S1B, C). Therefore, a critical PPI network module with 19 genes was extracted through the MCODE plugin in Cytoscape software (score: 5.556, Figure 3C). The functional enrichment results showed that this module was strongly associated with histone phosphorylation, collagen catabolic process, and pathways in cancer ( Figure 3D).

Expression and survival significance of Dendrobin A-related genes in PDAC
We next explored the significance of 19 Dendrobin A-related genes in the survival of PDAC patients. We found that patients with higher expression of AURKA, CCNA2, CDK1, KIF11, LGALS3, MMP1, and PLAU had worse OS in PDAC patients (HR > 1 and log-rank p < 0.05). At the same time, CCNA1 indicated a favorable prognosis in PDAC (HR < 1 and log-rank p < 0.05, Figure 4A), and the remaining 11 genes showed no significance with OS of PDAC (data not shown). Furthermore, the expression levels of these eight OS-related genes were displayed ( Figure 4B). Therefore, these eight OS-related genes, except CCNA1, may play oncogenic roles in PDAC.

Association of Dendrobin A-related genes with cachexia in PDAC
Since cachexia is a classic feature of PDAC patients (Freire et al., 2020), we next explored whether the dysregulation of Dendrobin A-related genes was associated with cachexia in PDAC patients. DEG analysis showed that 18 of 25 cachexia-induced factors were upregulated in PDAC samples ( Figure 5A). In addition, the cachexia score of cancerous pancreatic tissues was significantly higher than that of normal pancreatic tissues ( Figure 5B). We thus evaluated the associations between risky prognostic genes and cachexia in PDAC patients. As a result, the expression level of PLAU was significantly positively correlated with the cachexia score ( Figure 5C). Consistent with this finding, a similar upregulated expression pattern, prognostic risk, and the positive correlation with cachexia of PLAU levels in another PDAC dataset were observed (GSE62452, Figures 5D-G). We analyzed the association between PLAU mRNA levels and PDAC patients' clinical parameters, including age, gender, stage, and nodal metastasis status. The results showed that PLAU was differentially expressed between stage 1 and stage 2 (Supplementary Figure S2).

Molecular docking
Molecular docking is a convenient and effective means to explore the interaction of small molecules with their targets (Trott and Olson, 2010;Ravindranath et al., 2015); we used Vina 1.1.21 software to perform a molecular docking study of Dendrobin A with PLAU proteins. Interestingly, Dendrobin A is bound within the PLAU active pocket (score: −6.9 kcal/mol) ( Figure 6A). In addition, western blot results showed that Dendrobin A treatment reduced the expression level of PLAU protein in PDAC cells ( Figure 6B).

PLAU expression and association with clinical parameters in PDAC patients
We further examined PLAU protein expression in clinical PDAC samples by IHC. Commercial tissue arrays with 81 cases of PDAC and paired paraneoplastic tissues were used. The results showed that PDAC PLAU protein levels were significantly enhanced compared to paraneoplastic tissues ( Figure 7A). Further analysis showed that higher PLAU protein expression in PDAC was positively correlated with the differentiation and lymph node metastasis status of PDAC ( Figures 7B, C). Interestingly, PLAU protein expression in PDAC was associated with gender but not age, T classification, M classification, TNM stage, tumor size, and vascular and nerve invasion (Table 2).

In vitro validation of Dendrobin A on the proliferation, cycle, apoptosis, and migration in PDAC cells
Results of the CCK-8 assay indicated that Dendrobin A effectively inhibited the proliferation potentials of PDAC cells in a concentration-dependent manner but not for normal pancreatic cells ( Figure 8A). Subsequent experiments were performed using Dendrobin A at 60 µM according to its IC 50 values. The colony formation assay results further confirmed the inhibitory effects of Dendrobin A on PDAC cells ( Figure 8B). Considering that cell cycle and apoptosis are closely related to cell proliferation, we applied to flow cytometry to examine the effect of Dendrobin A on PDAC cell cycle and apoptosis. The results showed that the values of G 2 /M phases in PDAC cell lines Aspc-1 and Panc-1 have increased by 88.5% and 256% post-drug treatment, respectively ( Figure 8C), and the total number of apoptotic cells in Dendrobin A-treated Aspc-1 and Panc-1 cells were increased to 77% and 79.5%, respectively ( Figure 8D). We also tested the effects of Dendrobin A on PDAC cell migration and invasion. The wound healing and Transwell migration assay showed that Dendrobin A impaired PDAC cell migration and invasion abilities ( Figures 8E, F).

Discussion
TCM plays a special role in human life and health. Network pharmacology is an emerging technology to explore and identify the key targets of drugs, which provides an alternative way of exploring the application of TCM in the field of tumor treatment (Tang and Aittokallio, 2014;Asakawa et al., 2020). As one of the effective extracts of D. nobile, bibenzyl has various pharmacological activities such as anti-tumor, -inflammatory, and -diabetic (Zhao et al., 2018;Asakawa et al., 2020). Dendrobin A is a recently identified type of bibenzyl extracted from D. nobile (He et al., 2020), which has anti-cancer activities. However, its effects and mechanism of action on specified tumors remain unsolved. The goal of this work was to uncover the effects and mechanisms of Dendrobin A against PDAC. Frontiers in Pharmacology frontiersin.org Using SwissTargetPrediction and PharmMapper platform, 353 targets for Dendrobin A were identified. Subsequently, based on TCGA and GTXs databases, 3,131 DEGs between PDAC and normal pancreatic tissues were obtained. Finally, 90 potential targets of Dendrobin A against PDAC were identified. GO annotation indicated that these 90 genes were involved in metabolismrelated biological processes and responses to drugs, while KEGG signals enrichment showed that cancer-related pathways and drug metabolism were enriched. To further explore the signatures of targets for Dendrobin A against PDAC, we overlapped these 90 genes with targets related to PDAC from the MalaCards database, and 13 key genes previously reported in PDAC were found, namely, AURKA, CCNA2, ALB, LCN2, NQO1, CDA, PLAU, LGALS3, MMP2, REG1A, MMP7, MMP9, and PIK3CG. Therefore, these 90 genes indicated that the role of Dendrobin A on PDAC is multifaceted.
Next, we established a PPI for these 90 genes using the STRING database. As a result, a close link among the PPIs was identified. We also found that a specific module consisted of 19 key genes, namely, AURKA, KIF11, TK1, CDK1, CCNA1, AURKB, CCNA2, PLAU, SERPINA1, STSB, CTSS, REN, SELP, PPARG, LGALS3, MMP1, MMP9, NOS2, and LCN2. The functional analysis further indicated that these 19 genes were indeed involved in pathways in cancer. Thus, these 19 genes provided a cue for Dendrobin A against PDAC.
Given that these 19 genes play a critical role in PDAC carcinogenesis, we analyzed the significance of these 19 genes in the OS of PDAC patients. However, eight genes displayed the OS-predicated values in PDAC. Among them, high expression of AURKA, CCNA2, CDK1, KIF11, LGALS3, MMP1, and PLAU is the risk factor, while high expression of CCNA1 is a protective factor for PDAC. Indeed, AURKA (Gomes-Filho et al., 2020), CCNA2 , CDK1 (Piao et al., 2019), KIF11 (Gu et al., 2022), LGALS3 (Yi et al., 2020), MMP1 , and PLAU (Kisling et al., 2021;Wang et al., 2022) are significantly associated with the prognosis of PDAC patients. Our current study provides focus targets of Dendrobin A against PDAC.
Cachexia, a typical characteristic of malignant tumors, especially for pancreatic cancer patients, is significantly associated with PDAC prognosis, treatment, and quality of life. Studies have reported the positive role of TCM in improving the cachexia status of tumor patients (Xu et al., 2021). In this study, we found a high association of Dendrobin A targets with PDAC cachexia. Among these genes, PLAU highlights the association with PDAC cachexia. Increasing reports showed that high expression of PLAU is a risk factor in various tumors, including PDAC (Ai et al., 2020;Fang et al., 2021;Li et al., 2021;Wang et al., 2022). Additionally, molecular docking found that Dendrobin A could be bound within the PLAU active pocket. In addition, after treating the PDAC cells with Dendrobin A, PLAU expression in PDAC cell lines was also dramatically changed. Thus, PLAU may have important molecular targeting value in anti-PDAC treatment. Indeed, an early report has proposed that PLAU is a potent target for the anti-PDAC using TCM  and a potential prognostic marker in PDAC (Kisling et al., 2021;Wang et al., 2022). Since PLAU was one of the representative key targets identified in this study, the expression level of PLAU protein in PDAC and control paraneoplastic tissues in tissue microarrays was detected by immunohistochemistry. The results showed that the expression of PLAU was significantly higher in PDAC tissues, and its expression level correlated with the differentiation and lymph node metastasis status of PDAC. We also observed that PLAU protein levels were higher in female PDAC patients than in male PDAC patients, though the reason for this is unknown. Together, these results highlight the pathophysiological role of PLAU in PDAC.
Finally, we performed several in vitro experiments to test the feasibility of anti-PDAC using Dendrobin A. Similar to the expected results, we found that administration of PDAC cells with Dendrobin A led to substantially anti-cancer effects, as evidenced by the attenuated cell viability, migration, invasion, and enhanced apoptosis, and changed the cell cycle. Thus, the anti-PDAC properties of Dendrobin A are versatile.
In summary, our current study integrated network pharmacology, bioinformatics, and validation experiments, and found that Dendrobin A may exert its anti-PDAC action through multiple targets, and that PLAU may be a potential target for Dendrobin A against PDAC. Our study provides alternative evidence for the use of TCM to treat PDAC. However, a larger cohort and animal experiments are needed to verify these results.

Data availability statement
The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding authors.

Ethics statement
Ethical review and approval was not required for the study on human participants in accordance with the local legislation and institutional requirements. Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements.
Author contributions SZ, WJ, and WM developed the concept and designed the study. LY and HD extracted and isolated Dendrobin A from D. nobile. XX, YY, BW, BR, and YF performed most of the research. WJ and SZ supervised the research. XX and WJ wrote the manuscript. BR, BW, and XZ helped in the investigation and data analysis. All authors proofread and approved the final manuscript.

Funding
This work was supported by Hainan Province Science and Technology special fund (ZDYF2020132, ZDYF2022SHFZ065 to SZ; ZDYF2021SHFZ098 to XZ); the National Natural Science Foundation of China (81960528 to SZ); the specific research fund of the Innovation Platform for Academicians of Hainan Province Frontiers in Pharmacology frontiersin.org