Lung cancer metastasis, the leading cause of patient mortality, is driven by circulating tumor cells (CTCs), which act as direct mediators of metastatic spread. To elucidate the metabolic heterogeneity across lung cancer metastatic stages, a panoramic single-cell metabolomics study in a mouse lung cancer bone metastasis model was performed using a concentric hybrid nanoelectrospray ionization-atmospheric pressure chemical ionization source. This platform enables high-coverage detection of polar and nonpolar metabolites, overcoming limitations in sensitivity and metabolite diversity. Unsupervised clustering and dimensionality reduction (t-SNE) of single-cell metabolic profiles distinguished primary tumor cells, CTCs, and bone metastatic cells, revealing stage-specific metabolic reprogramming. Machine learning identified key metabolites (e.g., aminobenzoic acid, 2-methyl-3-ketovaleric acid, pantothenic acid) that robustly discriminated metastatic stages with high accuracy (AUC > 0.96). CTCs exhibited dynamic metabolic adaptions at different stages: during blood circulation, amino acid and glutamine metabolism dominated to counteract nutrient deprivation, while during bone colonization, the tricarboxylic acid cycle and one-carbon metabolism were upregulated to support proliferation. This study provides important data to shed light on the metabolic heterogeneity of tumor cells and the metastasis mechanism of lung cancer.
Hu et al. (Wed,) studied this question.