Exploring hypoxia- and cuproptosis-related biomarkers in periodontitis based on transcriptome and single-cell...

Periodontitis (PD) is a chronic, multifactorial inflammatory disorder characterized by the progressive destruction of periodontal tissues. Increasing evidence indicates that the dysregulation of hypoxia-related genes (HRGs) plays a pivotal role in inflammatory diseases, including PD. Recent studies have also implicated cuproptosis-a novel copper-dependent form of programmed cell death-in PD pathogenesis, suggesting a potential link with cuproptosis-related genes (CRGs). Despite these findings, the interaction between hypoxia, cuproptosis, and PD progression remains poorly understood. This study aims to identify and characterize key biomarkers associated with hypoxia and cuproptosis in PD, offering novel insights into its molecular mechanisms.

PD datasets were downloaded from the Gene Expression Omnibus (GEO) database. PD biomarkers were obtained through differential analysis, gene set variation analysis (GSVA), machine learning, diagnostic evaluation, and expression level validation.A predictive nomogram incorporating these biomarkers was constructed to evaluate their clinical utility. Additionally, functional enrichment analysis, immune cell infiltration profiling, and drug prediction were conducted to explore the biological roles and therapeutic implications of the biomarkers. Single-cell RNA sequencing data from GSE152042 were also analyzed to assess cell-type composition and examine biomarker expression at the single-cell level.

Differential analysis identified seven genes. The Boruta algorithm selected six feature genes (ALOX15B, FAM46C, IGHD, SAA1, SLC16A9, SPAG17), with SPAG17 showing the highest importance score. LASSO regression identified five genes (IGHD, FAM46C, SPAG17, SAA1, SLC16A9), while RFE pinpointed four genes (SPAG17, SLC16A9, SAA1, IGHD). In GSE16134 and GSE10334, SPAG17 and SLC16A9 exhibited significantly lower expression in PD compared to controls (Wilcoxon test, p < 0.05), whereas SAA1 and IGHD were significantly elevated. All four genes demonstrated AUC > 0.8. Unsupervised clustering identified eight cell types. SLC16A9 had the highest expression prevalence (82.1%) and mean expression (1.85 TPM) in epithelial cells; SPAG17 was predominantly expressed in perivascular cells (65.3%, 1.42 TPM), SAA1 in fibroblasts (73.6%, 2.01 TPM), and IGHD in B cells (89.2%, 2.37 TPM).

IGHD, SAA1, SLC16A9, and SPAG17 were identified as key biomarkers of PD. Pathway analysis and drug prediction provided insights into potential therapeutic targets, advancing the understanding of PD diagnosis and treatment.