Zeliang Yang,* Yu Dong,* Yuanyuan Shang, Haoran Li, Weicong Ren, Shanshan Li, Yu Pang

Department of Bacteriology and Immunology, Beijing Chest Hospital, Capital Medical University/Beijing Tuberculosis and Thoracic Tumor Research Institute, Beijing, 101149, People’s Republic of China

*These authors contributed equally to this work

Correspondence: Yu Pang, Email pangyupound@163.com

Background: Accurate diagnosis of active tuberculosis (TB) remains challenging when facing with no clinical symptom and negative pathogen tests. Metabolic reprogramming is the main characteristic of Mycobacterium tuberculosis (Mtb) infection and has the potential to be used as a diagnostic biomarker for active TB.
Methods: Datasets including healthy donors (HCs) and active TB patients were obtained from the Gene Expression Omnibus database. Machine learning methods were used to identify the metabolism-related hub genes. Correlation analysis between gene expression and immune cell infiltration was performed using the CIBERSORT algorithm. Single-cell RNA-seq analysis was performed to explore the expression of hub genes in the different immune cells.
Results: In this study, we first obtained 41 differentially expressed metabolism-related genes in active TB patients compared to HCs through bulk transcriptomic analysis. Four metabolism-related hub genes (GCH1, GK, MTHFD2, and SLC7A6) were identified using machine learning algorithms for the diagnosis of active TB with high accuracy and verified using external datasets. A nomogram was constructed to comprehensively predict the risk of active TB. Mechanistically, protein–protein interactions and gene set enrichment analysis revealed that four hub genes affected pteridine and lipid metabolism and were associated with the innate immune pathways. Immune cell infiltration and single-cell sequencing analyses showed that GCH1, GK, and MTHFD2 were mainly expressed in M1 macrophages and were significantly upregulated after Mtb infection, suggesting that they might participate in macrophage-mediated anti-Mtb immune responses. Furthermore, the expression levels of GCH1, GK, and MTHFD2 in macrophages showed a strong correlation with the course and efficacy of antituberculosis therapy. Changes in the expression of these hub genes were validated in active TB samples and Mtb-infected mouse models.
Conclusion: Our results demonstrate that changes in immunometabolism-related genes are associated with TB pathogenesis and could serve as biomarkers for the evaluation of active TB.

Keywords: active tuberculosis, metabolism reprogramming, biomarkers, macrophages, single-cell transcriptomic analysis