EPYC functions as a novel prognostic biomarker for pancreatic cancer

Zhen Yang, Honglin Li, Jie Hao, Hanwei Mei, Minghan Qiu, Huaqing Wang, Ming GaoScientific 2024 · DOI

TL;DR

Researchers have identified Epiphycan (EPYC) as a novel prognostic biomarker for pancreatic cancer (PC). By analyzing large public datasets and conducting experimental validation, they found that EPYC is highly expressed in PC cells and promotes tumor growth and migration, suggesting its potential as a therapeutic target.

What Problem Does This Paper Address?

Pancreatic cancer (PC) is a devastating disease characterized by late diagnosis, rapid progression, and limited treatment options, leading to a very poor prognosis. The underlying molecular mechanisms driving PC are not fully understood, hindering the development of effective therapeutic strategies. This paper aims to address this by identifying novel biological signatures that can serve as clinically useful biomarkers and potential therapeutic targets for PC.

Methods

The study employed a multi-pronged approach:

  • Bioinformatics Analysis:
    • Data Acquisition: Publicly available PC data from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) databases were downloaded.
  • Differential Expression Analysis: Standard bioinformatics pipelines using R packages ("Limma", "edgeR", "DESeq2") were used to identify differentially expressed genes (DEGs) between tumor and normal tissues.
  • Functional Enrichment Analysis: Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses were performed to understand the biological processes associated with the identified DEGs.
  • Survival Analysis: Log-rank tests and univariate Cox regression were used to identify genes significantly associated with patient survival.
  • Model Building: Lasso regression was used to narrow down candidate genes, followed by multivariate Cox regression to identify independent risk factors.
  • Molecular Functional Experiments:
    • Gene Expression: RT-qPCR was used to measure mRNA levels of EPYC.
  • Cell Proliferation: CCK8 and EdU assays were employed to assess cell proliferation.
  • Colony Formation: Assessed the ability of cells to form colonies, indicating long-term proliferation potential.
  • Cell Viability: Live-dead staining was used to evaluate cell survival.
  • Cell Migration: Transwell assays were performed to measure cell migration.
  • Protein Expression: Western blotting was used to detect protein levels of EPYC and key signaling molecules.
  • In Vivo Studies: Subcutaneous tumor xenograft models in nude mice were used to evaluate the effect of EPYC on tumor growth.

Datasets

  • The Cancer Genome Atlas (TCGA): Used for initial data mining and model construction/validation.
  • Gene Expression Omnibus (GEO): Specifically, GSE28735 and GSE62452 datasets were used for data acquisition and validation.
  • Human Protein Atlas (HPA): Used for analyzing protein levels of EPYC via immunohistochemical staining.
  • GEPIA: Used to analyze the correlation between EPYC expression and the PI3K signaling pathway.

Key Insights and Ideas

  • EPYC as a Novel Biomarker: The study strongly suggests EPYC is a previously underappreciated player in pancreatic cancer, functioning as a significant prognostic marker.
  • EPYC's Oncogenic Role: Beyond being a marker, EPYC actively promotes tumor progression by enhancing cell proliferation and migration.
  • PI3K-AKT Pathway Involvement: The research points to the PI3K-AKT signaling pathway as a key mechanism through which EPYC exerts its pro-tumorigenic effects, offering a potential therapeutic target.
  • Integrated Bioinformatics and Experimental Validation: The study's strength lies in its rigorous approach, combining large-scale data analysis with targeted experimental validation to confirm findings.

Key Results

  • Identification of Candidate Genes: Initial analysis identified 1090 DEGs in TCGA and 387 DEGs in GEO. After several filtering steps, 8 candidate genes were identified through Lasso regression.
  • Multivariate Cox Regression: A multivariate Cox regression model identified EPYC and MET as statistically significant independent risk factors for PC.
  • EPYC Expression: EPYC was found to be significantly upregulated in PC cell lines compared to normal pancreatic cells, both at the mRNA and protein levels.
  • Functional Impact of EPYC Knockdown: Knocking down EPYC in PC cells (PANC-1 and CFPAC-1) led to:
  • Reduced cell proliferation (CCK8, EdU assays).
  • Decreased colony formation.
  • Increased cell death (live-dead staining).
  • Suppressed cell migration (Transwell assay).
  • In Vivo Tumor Growth: Knockdown of EPYC in a subcutaneous xenograft model resulted in smaller tumor sizes, confirming its role in promoting tumor growth in vivo.
  • PI3K-AKT Pathway Modulation: Reduced EPYC expression led to decreased phosphorylation of AKT, suggesting EPYC's influence on the PI3K-AKT pathway.
  • Model Validation: The developed multivariate Cox regression model showed good predictive performance, with a C-index of 0.72 for the training set and 0.70 for the test set, and was further validated using GEO data.

Limitations and Future Work

  • Upstream and Downstream Mechanisms: While the study implicates the PI3K-AKT pathway, the specific upstream regulators of EPYC and its downstream targets remain to be fully elucidated.
  • In Situ Tumor Models: The study acknowledges that further validation using in situ tumor transplantation models or transgenic mouse models would strengthen the findings.
  • Clinical Translation: The authors note that while EPYC shows promise as a prognostic biomarker, further extensive validation is needed before it can be considered a clinical therapeutic target.
  • Other Candidate Genes: The study identified other genes (e.g., MET) that were also significant but focused on EPYC due to its less-explored role in PC. Further research into the roles of these other genes could also be beneficial.

Summary updated

Abstract

Pancreatic cancer (PC) has become a worldwide challenge attributed to its difficult early diagnosis and rapid progression. Treatments continue to be limited besides surgical resection. Hence, we aimed to discover novel biological signatures as clinically effective therapeutic targets for PC via the mining of public tumor databases. We found that epiphycan (EPYC) could function as an independent risk factor to predict the poor prognosis in PC based on integrated bioinformatics analysis. We downloaded associated PC data profiles from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) online websites, then applied the software Rstudio to filter out genes under the strict criteria. After the batch survival analysis using Log-rank test and univariate cox regression, we obtained 39 candidate genes. Subsequently, we narrowed the scope to 8 genes by establishing a Lasso regression model. Eventually, we focused on 2 genes (EPYC and MET) by further building a multivariate cox regression model. Given that the role of EPYC in PC remains obscure, we then performed a series of molecular functional experiments, including RT-qPCR, CCK8, EdU, colony formation, Transwell, western blot, cell live-dead staining, subcutaneous tumor formation, to enhance our insight into its underlying molecular mechanisms. The above results demonstrated that EPYC was highly expressed in PC cell lines and could promote the proliferation of PCs via PI3K-AKT signaling pathway in vivo and in vitro. We arrived at a conclusion that EPYC was expected to be a biological neo-biomarker for PC followed by being a potential therapeutic target.
BibTeX
@inproceedings{yang2024epyc,
  author = {Zhen Yang and Honglin Li and Jie Hao and Hanwei Mei and Minghan Qiu and Huaqing Wang and Ming Gao},
  title = {{EPYC functions as a novel prognostic biomarker for pancreatic cancer}},
  booktitle = {Scientific},
  year = {2024},
  doi = {10.1038/s41598-024-51478-w},
}
Figure