Identification of hypoxia-related gene signatures and molecular subtypes in chronic rhinosinusitis with nasal polyps

TL;DR

This study finds four hypoxia-related genes that could help diagnose nasal polyps and identifies two distinct disease subtypes for better treatment.

Contribution

The study introduces four novel hypoxia-related genes as potential biomarkers and reveals two molecular subtypes of CRSwNP.

Findings

Four hypoxia-related genes (CXCR4, HMOX1, DTNA, FBP1) were identified as potential biomarkers for CRSwNP.

Consensus clustering analysis revealed two distinct molecular subtypes of CRSwNP with different hypoxia characteristics.

The identified genes showed high diagnostic accuracy with AUC values exceeding 0.8.

Abstract

•This study identifies four genes as potential biomarkers for CRSwNP.•Machine learning techniques were used to effectively identify key genes.•Clustering analysis reveals two distinct molecular subtypes of CRSwNP.•This study identified sensitive drugs with therapeutic potential for two clusters.

This study identifies four genes as potential biomarkers for CRSwNP.

Machine learning techniques were used to effectively identify key genes.

Clustering analysis reveals two distinct molecular subtypes of CRSwNP.

This study identified sensitive drugs with therapeutic potential for two clusters.

Chronic Rhinosinusitis with Nasal Polyps (CRSwNP) is a common and severe subtype of Chronic Rhinosinusitis (CRS), characterized by persistent inflammation and tissue remodeling in the nasal cavity. Hypoxia, resulting from chronic inflammation, is believed to play a pivotal role in CRSwNP pathogenesis. This study aims to identify Hypoxia-Related Genes (HRGs) involved in CRSwNP and explore their potential as diagnostic biomarkers and therapeutic targets.

We retrieved transcriptomic datasets (GSE136825 and GSE179265) from the Gene Expression Omnibus (GEO) database. Differential expression analysis was conducted to identify Hypoxia-Related Differentially Expressed Genes (HRG-DEGs) in CRSwNP. Machine learning techniques, including Least Absolute Shrinkage and Selection Operator (LASSO) regression and random forest, were employed to identify key HRGs. Validation of these genes was performed using qRT-PCR on clinical CRSwNP and control tissue samples. Additionally, consensus clustering analysis was applied to categorize CRSwNP patients into different molecular subtypes based on HRG expression, followed by gene ontology (GO) enrichment and immune infiltration analysis.

A total of 19 hypoxia-related Differentially Expressed Genes (DEGs) were identified. Through LASSO regression and Random Forest analysis, followed by validation using an external dataset and qRT-PCR of clinical samples, four hypoxia-related hub genes ‒ CXCR4, HMOX1, DTNA, and FBP1 ‒ were identified. Receiver Operating Characteristic (ROC) curve analysis demonstrated that all four genes had Area Under the Curve (AUC) values exceeding 0.8. Additionally, consensus clustering revealed two distinct clusters with different hypoxia characteristics.

This study identifies four key hypoxia-related genes (CXCR4, HMOX1, DTNA, and FBP1) that may serve as novel diagnostic biomarkers for CRSwNP. The molecular subtypes identified through clustering provide further insights into CRSwNP heterogeneity, suggesting the need for personalized treatment approaches targeting hypoxia-related pathways.

Figures