Multiscale metabolic mapping of lung tissue via coregistered mass spectrometry and nonlinear optical imaging

The lung is a highly heterogeneous organ that is composed of numerous microanatomical units, each essential for maintaining intricate functions that work in concert. Disruptions in the molecular and cellular mechanisms can cause tissue fibrosis, inflammation, and severe breathing difficulties, which are common characteristics of the disease of prematurity, bronchopulmonary dysplasia (BPD). BPD's molecular changes are not well understood, and this has hindered effective diagnosis and treatment. Here, we present a multimodal imaging workflow for detailed molecular and metabolic characterization of lung tissue at multiple spatial scales. We also developed a hierarchical multimodal registration network for precise coregistration of the data from each modality. Our results show that this approach can reveal previously unknown metabolic changes in distinct functional tissue units affected by disease, including altered lipid distributions, reduced optical redox states, and collagen remodeling. This multimodal approach provided detailed maps of molecular shifts occurring in distinct microanatomical features that, when adopted to interrogate this and other tissue types, has the potential to enable the discovery of new therapeutics.