HeptaTB Dx: a diagnostic model leveraging cuproptosis-ferroptosis crosstalk for distinguishing latent from active...

Distinguishing latent tuberculosis infection (LTBI) from active tuberculosis (ATB) remains challenging. The roles of cuproptosis-ferroptosis crosstalk in TB immunopathology and diagnostic potential are unexplored. Transcriptomic data from Gene Expression Omnibus data sets (GSE37250/GSE28623) were analyzed to identify cuproptosis-/ferroptosis-related differentially expressed genes. Bioinformatics (limma, weighted gene co-expression network analysis) and machine learning (LASSO, SVM-RFE) screened key biomarkers. A logistic regression model (HeptaTB Dx Model) was developed and validated in independent cohorts. Real-world validation included RNA-seq (n = 20) and reverse transcription quantitative polymerase chain reaction (RT-qPCR) cohorts (HC/LTBI/ATB, n = 111). Immune cell infiltration (ssGSEA/CIBERSORT) and consensus clustering elucidated mechanisms. We identified seven core biomarkers (MT1G, SCO2, CREB5, MGST1, PARP9, ATF3, and MUC1) regulating cuproptosis-ferroptosis interplay. HeptaTB Dx achieved exceptional performance: training, area under the curve (AUC) = 0.963 (sensitivity 0.928, specificity 0.897) and validation, AUC = 0.930 (sensitivity 0.920, specificity 0.870). Real-world RT-qPCR validation confirmed significant differential expression for 5/7 genes (CREB5, ATF3, MT1G, PARP9, and MGST1; P < 0.05) and model AUC = 0.778. Mechanistically, these genes formed a cooperative network linking immune regulation (ATF3/PARP9), ferroptosis suppression (MT1G/MGST1), barrier function (MUC1), and cuproptosis-metabolism (SCO2). CREB5 correlated with neutrophil infiltration (R = 0.83, P < 0.001), validating ferroptosis-immune crosstalk. LTBI subtypes exhibited divergent lipid metabolism-ferroptosis coupling and antiviral pathway enrichment. The HeptaTB Dx Model is the first diagnostic tool leveraging cuproptosis-ferroptosis crosstalk, offering high accuracy and mechanistic insights for LTBI management.

The differentiation between latent tuberculosis infection (LTBI) and active tuberculosis (TB) is a persistent challenge in global TB control, with current diagnostics failing to reliably distinguish these states or predict progression. This study introduces the HeptaTB Dx Model, the first diagnostic signature derived from the crosstalk between cuproptosis and ferroptosis-two metal-dependent regulated cell death pathways with emerging roles in Mycobacterium tuberculosis pathogenesis. By integrating seven key genes (MT1G, SCO2, CREB5, MGST1, PARP9, ATF3, and MUC1), the model achieves high diagnostic accuracy (area under the curve up to 0.963) and provides mechanistic insights into immune-metabolic dysregulation during TB infection. Validated in both public datasets and prospective clinical cohorts, HeptaTB Dx offers a scalable, transcriptome-based tool that outperforms existing single-pathway models and protein-based assays. This work not only advances TB diagnostics but also illuminates novel pathogenic mechanisms involving copper-iron interplay, with potential implications for therapeutic targeting.