Triple negative breast cancer (TNBC) is a particularly aggressive cancer subtype that is difficult to diagnose due to its discriminating epidemiology and obscure metabolome. For the first time, 3D spatial and chemometric analyses uncover the unique lipid metabolome of TNBC under the tandem modulation of two key metabolites - insulin and methionine - using non-invasive optical techniques. By conjugating heavy water (D₂O) probed Raman scattering with label-free two-photon fluorescence (TPF) microscopy, we observed altered de novo lipogenesis, 3D lipid droplet morphology, and lipid peroxidation under various methionine and insulin concentrations. Quantitative interrogation of both spatial and chemometric lipid metabolism under tandem metabolite modulation confirms significant interaction of insulin and methionine, which may prove to be critical therapeutic targets, and proposes a powerful optical imaging platform with subcellular resolution for metabolic and cancer research.

Lipids play crucial roles in many biological processes. Mapping spatial distributions and examining the metabolic dynamics of different lipid subtypes in cells and tissues are critical to better understanding their roles in aging and diseases. Commonly used imaging methods (such as mass spectrometry-based, fluorescence labeling, conventional optical imaging) can disrupt the native environment of cells/tissues, have limited spatial or spectral resolution, or cannot distinguish different lipid subtypes. Here we present a hyperspectral imaging platform that integrates a Penalized Reference Matching algorithm with Stimulated Raman Scattering (PRM-SRS) microscopy. Using this platform, we visualize and identify high density lipoprotein particles in human kidney, a high cholesterol to phosphatidylethanolamine ratio inside granule cells of mouse hippocampus, and subcellular distributions of sphingosine and cardiolipin in human brain. Our PRM-SRS displays unique advantages of enhanced chemical specificity, subcellular resolution, and fast data processing in distinguishing lipid subtypes in different organs and species.