View Abstract

A2303 - Cell-Type Resolved Proteome and Lipidome of Human Lung at Population and Single Cell Levels
Author Block: G. Clair1, J. E. Kyle1, Y. Zhu1, G. Bandyopadhyay2, E. Zink1, R. Misra2, R. Kelly1, G. S. Pryhuber2, J. Carson3, C. Ansong1; 1Pacific Northwest National Laboratory, Richland, WA, United States, 2University of Rochester Medical Center, Rochester, NY, United States, 3Texas Advanced Computing Center, Austin, TX, United States.
Rationale. The lung is a complex organ comprised of more than 40 cell types each playing overlapping and niche roles in facilitating normal lung development and function. While cell-type specific transcriptome analyses of the murine lung have been reported, the cellular and molecular pathways that regulate lung development process are still poorly understood. Proteins and lipids perform the biochemical activities required for biological functions and are only moderately correlated with mRNA abundance. Thus characterization of the proteome and lipidome of individual lung cell types during development will help to better understand the specification mechanisms that drive normal lung formation and function. Methods. Unbiased mass spectrometry-based proteomics and lipidomics was used to characterize the proteome and lipidome of four major lung cell types (mesenchymal, epithelial, endothelial and mixed immune) isolated from three 20 month donors. Proteins and lipids were extracted using a modified Folch extraction method (metabolite, protein, and lipid extraction, MPLEx) and then introduced into a LC-MS/MS platform for population proteomics and lipidomics analyses. For single cell analyses, single epithelial and fibroblast cells were prepared using a novel nanowell platform and introduced into a LC-MS/MS platform for proteomics analyses. Results. Our integrative multi-omics approaches identified ~5,000 proteins and ~300 lipids from the cell-type resolved proteome and lipidome analyses at the population level. Principal component analysis (PCA) of the proteome data partitioned the four cell types into four well defined populations. The populations differentially retained expression of well-known markers typically used to discriminate lung cell type (SFTPC, NKX2.1, EPCAM, ABCA3, CD11c, CD288, PDGFRb, PECAM/CD31, etc.) and identified several potential “signature proteins” enriched in specific cell types. Cell-type lipidomics supported the presence of lipofibroblasts in the human lung and revealed cellular cooperation in lung function. Finally our nanowell-LC-MS/MS platform was able to identify ~500 proteins from single lung cells in an unbiased label-free manner, and PCA analysis partitioned the single cells based on protein expression alone. Conclusion. Understanding the proteomic and lipidomic constituents of human lung cell populations is highly informative and improves our knowledge of lung function needed to develop insights into the treatment of lung diseases. Furthermore, our study demonstrated the feasibility of unbiased label-free single-cell proteomics and was able to differentiate single lung cells based on proteome expression alone.