Multi-targeted Mechanisms of Erianin against Pancreatic Ductal Adenocarcinoma: Integrative Network Pharmacology, Bioinformatics, andExperimental Validation

Introduction: Erianin is a natural benzyl compound extracted from Dendrobium drumbium. However, the mechanism of action of Erianin against pancreatic cancer (PC) is still unclear. The aim of this study was to systematically elucidate the therapeutic potential of Erianin in PC and its molecular mechanisms. materials and methods: Screening of Potential Targets of Erianin Against Pancreatic Cancer Potential therapeutic targets of Erianin were predicted using three pharmacological databases (Swiss Target Prediction(http://www.swisstargetprediction.ch/), PharmMapper（https://www.lilab-ecust.cn/pharmmapper/submitfile.html）, and SuperPred(https://prediction.charite.de/)), resulting in 237 unique targets after deduplication. In parallel, 2,249 pancreatic cancer-related targets were systematically curated from four disease-specific databases: GeneCards(https://www.genecards.org/), OMIM(https://www.omim.org/), TTD(http://db.idrblab.net/ttd/), and DisGeNet(https://www.disgenet.org/). Gene symbol normalization for all protein entries was performed via the UniProt(https://www.uniprot.org/) database to ensure nomenclature consistency. Subsequent Venn diagram analysis identified 110 overlapping targets between Erianin and pancreatic cancer, revealing putative therapeutic candidates for further investigation. Construction of Protein-Protein Interaction Network and Screening of Key Targets The protein-protein interaction (PPI) network for Erianin-pancreatic cancer was constructed based on the STRING database, with a confidence threshold set at ≥0.7 and restricted to human genes. Network data were imported into Cytoscape 3.9.1 for visual analysis. Using the CytoNCA plugin, parameters including betweenness centrality (BC), closeness centrality (CC), degree centrality (DC), eigenvector centrality (EC), and local average connectivity (LAC) were analyzed for relevant target genes. This process ultimately identified the top 10 key targets: HIF1A, ESR1, HSP90AA1, ALB, ERBB2, SRC, CCND1, NFKB1, PPARG, and PIK3CA. Gene Ontology and Pathway Enrichment Analysis Functional enrichment analysis was performed using the DAVID database(https://david.ncifcrf.gov/). Gene Ontology (GO) analysis covered three dimensions: biological processes (BP), cellular components (CC), and molecular functions (MF), with a significance screening threshold of P≤0.05 and false discovery rate (FDR)≤0.05. After screening, the top 15 GO enrichment terms and 15 Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways were selected.Following the organization of GO and KEGG data, visual plotting was conducted via an online bioinformatics platform. Finally, data integrating Erianin, core targets, and the top 15 pathway relationships were used to construct a "compound-target-pathway" (C-T-P) network in Cytoscape. Expression Analysis of Key Targets in Pancreatic Cancer The pancreatic cancer tissue and healthy tissue gene expression samples (GSE28735, GSE62452) were downloaded through the Clinical Bioinformatics Home platform. The R package was used to analyze the expression differences of 10 key target genes between tumor tissues and normal tissues. Using the GEPIA2 database (http://gepia2.cancer-pku.cn/#index), with the median expression value (50%) as the grouping threshold, the hazard ratio (HR) and 95% confidence interval were calculated. Taking P 0.05 as the significance criterion, the associations between the key target genes and the overall survival (OS) and disease-free survival (DFS) of the patients were evaluated. Molecular Docking The three-dimensional (3D) structures of proteins were retrieved from the Protein Data Bank (PDB, https://www.rcsb.org/) based on their corresponding PDB IDs, following retrieval of relevant information from the UniProt website (https://www.uniprot.org/). Specifically, the structures of GRB2 (PDB ID: 1GRI) and HRAS (PDB ID: 8CNJ) were downloaded. Amino acid residues missing in the structures were supplemented using the SWISS-MODEL online platform (https://swissmodel.expasy.org/). For the protein IRS1, which lacked a 3D structure, its amino acid sequence was downloaded, and the 3D structures of both proteins were predicted using the AlphaFold3 online tool (https://alphafoldserver.com/). The predicted protein structures were inspected with PyMOL , and the conformation with the highest prediction rank was selected as the structural file. Hydrogen atoms and charges were added using AutoDock Tools 1.5.6 , and .pdbqt files were generated for docking. The small-molecule ligand structures were downloaded from the PubChem database (https://pubchem.ncbi.nlm.nih.gov/), and their molecular structures were optimized under the MMFF94 force field using ChemOffice 2019 for subsequent molecular docking. Ligand structures were processed with AutoDock Tools 1.5.6 to calculate charges and generate .pdbqt files for docking. Molecular docking was performed using AutoDock Vina . The docking results were visualized using PyMOL and LigPlot+. Functional Association Network Analysis via GeneMANIA A functional association network was established using the GeneMANIA platform (http://genemania.org/) by systematically analyzing ten core target genes. The network was constructed with co-expression, genetic interaction, and physical interaction as primary association criteria. For each key target, the ten most strongly associated genes were identified and integrated. After deduplication and consolidation, an expanded gene set comprising 109 targets (including the original ten) was generated. Functional annotation and pathway enrichment analyses were subsequently performed on this expanded set through the DAVID database, including Gene Ontology (GO) classification and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway mapping. Experimental Validation of Erianin's Anti-Pancreatic Cancer Effects Erianin (purity 99.63%, CAS No. 95041-90-0) was purchased from MedChemExpress (MCE, New Jersey, USA). Human pancreatic cancer cell lines PANC-1 and SW1990 were obtained from Procell (Wuhan, China) and Bidepharm (Chongqing, China), respectively. Cells were cultured in Dulbecco's Modified Eagle Medium (DMEM) or Roswell Park Memorial Institute (RPMI-1640) medium supplemented with 10 Methods: The potential therapeutic targets of Erianin in the treatment of PC were identified through online databases. The GeneMANIA database was used to expand the network pharmacology (NP) analysis results. The "Clinical Home" platform and GEPIA2 database were used to clarify the clinical significance of the key targets. Furthermore, protein-protein interaction (PPI), molecular docking, kinetic simulation, and enrichment analysis were used to predict the core targets and potential mechanisms of Erianin against PC. Finally, NP prediction results were confirmed by in vitro cell experiments. Results: A total of 10 hub genes, such as ESR1, HSP90AA1, HIF1A, ERBB2, ALB, NFKB1, SRC, PIK3CA, PPARG, and HARS, were identified by functional association of NP using the gene- MANIA database (GMFA). These targets were significantly enriched in PI3K-Akt and MAPK signaling pathways, and Erianin had a strong binding ability with key targets. In vitro experiments showed that Erianin significantly inhibited the proliferation, migration, and invasion of PC cells. The expression of hub genes was downregulated by Erianin, and Erianin inhibited the activation of the PI3K/AKT/MAPK pathway. Discussion: In the PI3K-Akt signaling pathway, in addition to core targets, related genes such as IRS1, GRB2, and HRAS were additionally identified, further refining the target network through which erianin exerts its antitumor effects via this pathway. Cell-level experimental results strongly validated the predictions from network pharmacology and bioinformatics. CCK8 proliferation assays, colony formation assays, wound healing assays, and Transwell invasion assays showed that Erianin significantly inhibited the proliferation, migration, and invasion abilities of pancreatic cancer cells in a dose-dependent manner. qRT-PCR results revealed that Erianin could downregulate the expression levels of multiple key genes, such as HSP90AA1, CCND1, and ERBB2, which are closely associated with the previously predicted targets and signaling pathways. Conclusions: Erianin exerted a multi-target effect against PC through the PI3K/Akt/MAPK signaling pathway and the key genes PIK3CA, HIF1A, and ERBB2, providing a theoretical basis for further clinical research on Erianin.