Cell atlases can reveal cell heterogeneity within tissues, which are inherently complex biological samples. While high-throughput single-cell metabolomics has been reported, single-cell metabolic atlases are difficult to generate without accurate single-cell signal capture. Here, we describe CyMeta-ImaGating, metabolomic cytometry with an image-enabled gating strategy, for generating single-cell metabolic atlases with verified single-cell profiles from complex samples. These verified profiles are obtained by matching metabolomic profiles with brightfield single-cell images, a process termed image-enabled gating. In scraped HeLa cell suspensions, which contained abundant debris and cellular aggregates, the image-enabled gating strategy increased the proportion of single-cell profiles from 28% to 91-100%. Using CyMeta-ImaGating, we generated spleen cell atlases that distinguished major immune cell types (B, T, and NK cells) and revealed transitions of metabolic states after pathogen activation. We also characterized doxorubicin's effects using liver and cancer cell atlases to illustrate drug specificity: the liver cell atlas revealed specific metabolic states (e.g., arginine upregulation) during drug-induced apoptosis, while the cancer cell atlas showed that drug-responsive metabolic states remained robust despite reduced single-cell rate at high concentrations. These results demonstrate the single-cell metabolic atlas as a critical tool for resolving cell metabolic states and dynamics.
He et al. (2026) studied this question.