Deep spatial proteomic exploration of severe COVID-19-related pulmonary injury in post-mortem specimens

Abstract

The lung, as a primary target of SARS-CoV-2, exhibits heterogeneous microenvironment accompanied by various histopathological changes following virus infection. However, comprehensive insight into the protein basis of COVID-19-related pulmonary injury with spatial resolution is currently deficient. Here, we generated a region-resolved quantitative proteomic atlas of seven major pathological structures within the lungs of COVID-19 victims by integrating histological examination, laser microdissection, and ultrasensitive proteomic technologies. Over 10,000 proteins were quantified across 71 dissected FFPE post-mortem specimens. By comparison with control samples, we identified a spectrum of COVID-19-induced protein and pathway dysregulations in alveolar epithelium, bronchial epithelium, and pulmonary blood vessels, providing evidence for the proliferation of transitional-state pneumocytes. Additionally, we profiled the region-specific proteomes of hallmark COVID-19 pulmonary injuries, including bronchiole mucus plug, pulmonary fibrosis, airspace inflammation, and hyperplastic alveolar type 2 cells. Bioinformatic analysis revealed the enrichment of cell-type and functional markers in these regions (e.g. enriched TGFBI in fibrotic region). Furthermore, we identified the up-regulation of proteins associated with viral entry, host restriction, and inflammatory response in COVID-19 lungs, such as FURIN and HGF. Collectively, this study provides spatial proteomic insights for understanding COVID-19-caused pulmonary injury, and may serve as a valuable reference for improving therapeutic intervention for severe pneumonia.

Multimodal single cell-resolved spatial proteomics reveals pancreatic tumor heterogeneity

“…Starting with either fresh frozen or formalin-fixed, paraffin-embedded (FFPE) tissue slice samples, LMD-based spatial proteomics has been widely applied to ovarian cancer 11 , colon cancer 12,13 , tuberculosis 14 , pancreatic cancer 15 , COVID-19 16,17 , etc. However, the spatial resolution of these studies largely depends on experienced pathological examination, subjective cell typing, and time-consuming manual selection of tissue regions of interest (ROI).…”

“…Guided by hematoxylin-eosin (H&E)- or immunohistochemical (IHC)-stained images, laser microdissection (LMD)-based spatial proteomics has been successfully demonstrated in providing unprecedented insights into the heterogeneous tissue context of lethal disease tissue samples with cell-type resolution. Starting with either fresh frozen or formalin-fixed, paraffin-embedded (FFPE) tissue slice samples, LMD-based spatial proteomics has been widely applied to ovarian cancer 11 , colon cancer 12,13 , tuberculosis 14 , pancreatic cancer 15 , COVID-19 16,17 , etc. However, the spatial resolution of these studies largely depends on experienced pathological examination, subjective cell typing, and time-consuming manual selection of tissue regions of interest (ROI).…”