Accurate quantification and spatial localization of small-molecule metabolites at the single-cell level remains a major challenge in cancer research. Conventional techniques such as LC/MS or isotope tracing offer limited spatial resolution, provide only relative quantification, or are cost-prohibitive for exploratory imaging—especially when targeting low-abundance metabolites in complex biological systems.
To address this gap, we present Null-Deflection Infrared Scanning Probe Microscopy (NDIR)—a novel, label-free imaging platform that integrates infrared spectroscopy with atomic force microscopy. NDIR enables high-resolution, sub-cellular topographic and chemical imaging with nanometer precision, without the need for labels or destructive sample preparation. Unlike traditional infrared or mass spectrometry-based techniques, NDIR provides direct, quantitative visualization of proteins and metabolites in situ.
Thus far, we have established a robust and reproducible sample preparation workflow optimized for NDIR imaging. Mammalian cells were cultured in 3D Matrigel matrices to preserve native tissue architecture, then fixed, dehydrated, resin-embedded, and ultramicrotome-sectioned for nanoscale imaging
NDIR imaging at 1660 cm⁻¹ (Amide I band) allowed visualization of protein structures, while imaging at 1025 cm⁻¹ (C-O stretching) enabled mapping of glucose uptake in single cells and spheroids—demonstrating successful resolution of small-molecule peaks in intact biological systems. To support metabolite quantification, we built an extensive spectral calibration library of key cancer metabolites using a 3D-printed microfluidic system for precise and reproducible standard deposition. This library, combined with NDIR’s spatial resolution, forms a basis for ongoing development of machine learning–based deconvolution tools.
Here, we will present the latest results of combining NDIR with machine learning for quantitative nanoscale mapping of metabolites in single cells.