Abstract
BACKGROUND: Cellular metabolic irregularities are intricately associated with the initiation and progression of tumors. Emerging evidence suggests that interactions between intratumoral microbiomes and host mediate this process. However, a comprehensive understanding of the role of metabolism-related intratumoral microbes (MRIMs) in lung adenocarcinoma (LUAD) is still lacking. This study aimed to investigate the characteristics and prognostic significance of MRIMs, as well as elucidate their potential implications in relation to the microenvironment in LUAD. METHODS: Integrated analyses were conducted using accessible datasets of the microbiome, bulk and single-cell transcriptomes. Spearman's coefficient between metabolic activity score and microbial abundance was used to identify MRIMs. An unsupervised clustering approach was utilized to distinguish the MRIMs-featured subtypes in LUAD samples. The Scissor algorithm was executed to select the cell subpopulations featured by MRIMs, and the underlying regulatory network in MRIMs-featured cells was explored. Additionally, a prognostic signature based on the microbial abundance of MRIMs was developed, and comprehensive analyses were subsequently carried out to reveal the correlation between MRIMs and LUAD microenvironment. RESULTS: Ten microbial species were identified as MRIMs, enabling the classification of LUAD samples into two distinct subtypes that showed significantly associated with clinical features and survival outcomes. The scRNA-seq analysis revealed notable differences in T cells, ciliated cells, mast cells, endothelial cells, and fibroblasts between MRIM+ and MRIM- subpopulations. BCL3, KLF3, and NFKB2 were the regulons in the regulatory network of MRIM-featured cells. Additionally, a microbial prognostic-predictive signature was established comprising Succinimonas, Collimonas, and Marichromatium, which also exhibited potential for indicating immunotherapeutic benefit and predicting drug sensitivity to cisplatin, cytarabine, pyrimethamine, olaparib, bicalutamide and vorinostat in LUAD treatment. CONCLUSIONS: This study identified intratumoral microbes associated with metabolism, revealed distinct subtypes and their roles in LUAD, and established a predictive signature for the prognosis and therapeutic responsiveness of LUAD.