Spatiotemporal remodeling of human milk fat globule compositional mapping revealed by plasmonic chemical imaging

The structure and composition of the human milk fat globules and their surrounding membrane (MFGM) are key determinants of their biological functions, yet how these architectures remodel across lactation has remained largely uncharted. Traditional MFGM analysis methods are hampered by the MFGM nanoscale dimensions and inherent heterogeneity, and the spatial information loss intrinsic to bulk approaches. Here, we present a plasmonic chemical imaging strategy using a bimetallic Ag@Au nanopillar substrate coupled to surface-enhanced Raman scattering (SERS) imaging for spatially resolved mapping of individual MFGs in minimally processed human milk. Across samples collected at different lactation stages, we observed reproducible stage-associated spatial partitioning of nutrient-related spectral features. Oligosaccharide-associated signals were found enriched at the globule periphery, whereas protein-associated signals were concentrated toward the globule core. Complementary compositional analyses showed lactation stage-dependent remodeling of the milk matrix. Computational classification further indicated that spatially resolved spectral signatures contained sufficient information for distinguishing between different lactation stages. This study presents a microstructural framework for studying a dynamic food system, such as human milk, offering a foundation for future benchmarking of infant formula design.