SMILE: Extraction‐free submicron‐resolution mapping of lipid chain length and unsaturation by stimulated Raman imaging

Abstract Lipid spatial heterogeneity is essential for linking metabolic regulation to cellular behavior. Yet, subcellular lipid profiling remains a challenge, as exogenous labels cannot detect lipid speciation and species diversity. Here, we developed submicron‐resolution mapping of intracellular lipid elements (SMILE), a vibrational spectrometry platform based on hyperspectral stimulated Raman scattering imaging with a spectral analysis pipeline for pixel‐resolved lipid profiling. SMILE enables quantitative mapping of lipid chain length and unsaturation at 300 nm resolution without the need for lipid extraction or labeling. Using SMILE, we revealed the spatial heterogeneity of lipid species and identified a distinct subset of highly aggregated lipid droplets (LDs) enriched in long‐chain polyunsaturated lipids in invasive cancer cells. Lipid elongation was found to promote LD aggregation and cancer migration/invasion, thereby defining this as a unique spatial lipid‐metabolic signature associated with cancer aggressiveness. Mechanistically, we identified linolenic acid as a key mediator driving aggressive lipid signatures and cancer invasion. Collectively, these findings demonstrate SMILE as an enabling spatial‐omics platform that expands the chemical resolution of vibrational spectroscopic imaging for fundamental and translational research.