Integrated bioinformatics and pharmacology reveal esculin's mechanism against pancreatic adenocarcinoma drug resistance.

BACKGROUND

Pancreatic adenocarcinoma (PAAD) remains a highly aggressive cancer with limited treatment options and a poor prognosis. A lack of reliable biomarkers and insufficient understanding of tumor heterogeneity impede early PAAD detection and personalized treatment.

PURPOSE: This study aimed to construct a gene-based risk model and explore targeted approaches to overcome tumor drug resistance in PAAD.

METHODS

Single-cell RNA sequencing (scRNA-seq) bioinformatics identified key driver genes, stratifying PAAD patients into two subtypes. A nine-gene risk model was constructed using Lasso Cox regression and validated via immunohistochemistry (IHC) and RT-qPCR. The natural compound esculin was evaluated for its ability to target proteins encoded by hub genes. Its effects on drug resistance were assessed using in vitro and in vivo models.

RESULTS

In the study, scRNA-seq data identified 20 hub genes, which were used to classify PAAD patients into two subtypes via consensus clustering. These subtypes exhibited significant differences in immune infiltration patterns and overall survival (OS) outcomes. Building on this, a nine-gene prognostic model was constructed and the model demonstrated strong predictive capability for OS, and its accuracy was substantiated by external cohorts, IHC staining and RT-qPCR. Furthermore, esculin emerged as a promising therapeutic candidate targeting hub gene proteins TP53, EP300, and PIAS1. Both in vitro and in vivo drug-resistant tumor model experiments confirmed that esculin exhibits favorable biosafety and reduces tumor drug resistance by downregulating TP53, EP300, and PIAS1 proteins.

CONCLUSIONS

Collectively, our study not only provides a powerful prognostic tool but also identifies a promising multi-target natural compound, offering a novel strategy to overcome chemoresistance in PAAD.