Multimodal Single-Cell Characterization of the Human Granulocyte Lineage

Abstract

AbstractHigh throughput single cell transcriptomics (scRNA-seq) has been successfully applied to characterize immune cell heterogeneity across a diverse range of settings; however, analysis of human granulocytes remains a significant challenge due to their low gene expression transcript detection. Consequently, granulocytes are typically either absent or highly under-represented and inaccurately enumerated in most human scRNA-seq datasets. Here, we apply multi-modal CITE-seq profiling to characterize granulocytes in human whole blood and bone marrow, and we show that these populations can be accurately detected and analyzed using the antibody-based modality, and that their frequencies and phenotype align well with antibody-based characterization of the same samples using CyTOF. These analyses also clearly highlight extremely low gene transcript detection across the entire granulocyte lineage including the earliest neutrophil progenitor populations when using the 10X Genomics platform. By contrast, when performing parallel analyses of the same samples using the BD Rhapsody platform, we recovered a much higher proportion of granulocyte gene transcripts, enabling true multi-modal characterization of human granulocyte heterogeneity.