Associate Professor University of California San Diego La Jolla, California, United States
Introduction: : The lung is composed of various functional units, each essential for maintaining non-labored breathing and respiration. Disruptions in the lung’s molecular and cellular mechanisms result in inflammation, fibrosis and remodeling, which are characteristic of lung disorders such as the lung disease following premature birth, bronchopulmonary dysplasia (BPD). Molecular derangements in human BPD have been difficult to study due to rare access to sufficient human lung samples and high-resolution technologies, thus hindering identification of effective diagnosis and treatment. Here, we present a new multimodal imaging workflow for detailed molecular and metabolic characterization of human tissues at spatial scales. For analysis, we developed a hierarchical multimodal registration network (HiMReg). Using a combination of techniques, we precisely co-registered matrix-assisted laser desorption/ionization mass spectrometry imaging (MALDI-MSI) with label-free, tissue-sparing, multimodal microscopy including two-photon fluorescence (TPF), second harmonic generation (SHG), and stimulated Raman scattering (SRS). This approach revealed previously unknown metabolic changes in distinct functional tissue units affected by BPD when compared to corrected age matched, full term born, healthy lung, including altered lipid distributions, reduced optical redox states, and specific collagen remodeling in bronchioles. Our findings demonstrate spatially defined alterations in lipid composition and metabolism in BPD-affected alveoli, providing novel insights into disease pathophysiology to translate into therapy. By providing detailed maps of the metabolic shifts occurring in distinct tissue microanatomical features, the multi-modal methods as developed here enable the discovery of new therapeutic avenues making the integrated approach highly attractive for the field of biomedical research in lung and other organs
Materials and
Methods: : Lung tissue was procured through the BioRepository for Investigation of Diseases of the Lung (BRINDL)57. Agarose inflated, CMC embedded fresh-frozen four lung tissue blocks were chosen from the left-upper lobe of a donor with bronchopulmonary dysplasia and an age matched control. Serial tissue sections were cut at 20 µm thickness and thaw mounted on two Superfrost™ Plus Microscope Slides (Fisher Scientific) and two ITO-coated glass slides (Delta Technology). Additional tissue sections (200 µm) were collected in a 1.7 ml Sorenson tube for MPLEx sample preparation and lipid extraction and liquid-chromatography mass spectrometry (LC-MS) analysis. MALDI-MSI and Label-free multimodal imaging data were co-registered with Hierarchical Multimodal Registration Network (HiMReg) for spatial correlation analysis.
Results, Conclusions, and Discussions:: We demonstrate the advantages of employing a multimodal imaging workflow to investigate metabolic differences between lung FTUs from healthy and diseased tissue. Using MALDI-MSI, we identified lipids enriched in each FTU of healthy lung tissue, and then compared the relative abundances of these lipids to those found in a BPD lung, where we measured significant changes in some lipidomic composition. Following MALDI-MSI, we employed label-free multimodal microscopy to explore broad metabolic changes at higher resolution. This analysis revealed increased lipid unsaturation and optical redox in the vasculature of healthy lung tissue and an overall increase in optical redox in BPD lung tissue. To accurately align serial MALDI-MSI and microscopy images, we implemented a novel co-registration strategy called HiMReg. This co-registration allowed us to perform qualitative analysis of the images and further segment the data based on the optical redox ratio images. We discovered that multiple SLs were upregulated in high-redox regions, while a few PCs were upregulated in low-redox regions. Overall, our workflow provides a method for analyzing broad metabolic changes at high spatial resolution, and it permitted us to link these metabolic shifts to specific lipidomic alterations within tissue sections. Continued advancements in multimodal workflows will facilitate the mapping of specific FTUs and their disease-related changes, enhancing our understanding of pathological processes.
.jpg "Lingyan Shi, PhD photo")