Spatially visualized single-cell pathology of highly multiplexed protein profiles in health and disease

• Authors: Mayar Allam, Emory University; Thomas Hu, Emory University; Shuangyi Cai, Emory University; Krishnan Laxminarayanan, Georgia Institute of Technology, Atlanta; Robert B Hughley, Georgia Institute of Technology, Atlanta; Ahmet F Coskun, Emory University
• Journal Title: Communications Biology
• Volume: Volume 4
• Publisher: Nature Research (part of Springer Nature) | 2021-05-27
• Type of Work: Article | Final Publisher PDF
• Publisher DOI: http://doi.org/10.1038/s42003-021-02166-2
• Subjects: Engineering, Biomedical; Computer Science
• Keywords: Single-cell imaging; Image processing; Imaging the immune system

Abstract

Deep molecular profiling of biological tissues is an indicator of health and disease. We used imaging mass cytometry (IMC) to acquire spatially resolved 20-plex protein data in tissue sections from normal and chronic tonsillitis cases. We present SpatialViz, a suite of algorithms to explore spatial relationships in multiplexed tissue images by visualizing and quantifying single-cell granularity and anatomical complexity in diverse multiplexed tissue imaging data. Single-cell and spatial maps confirmed that CD68+ cells were correlated with the enhanced Granzyme B expression and CD3+ cells exhibited enrichment of CD4+ phenotype in chronic tonsillitis. SpatialViz revealed morphological distributions of cellular organizations in distinct anatomical areas, spatially resolved single-cell associations across anatomical categories, and distance maps between the markers. Spatial topographic maps showed the unique organization of different tissue layers. The spatial reference framework generated network-based comparisons of multiplex data from healthy and diseased tonsils. SpatialViz is broadly applicable to multiplexed tissue biology.