Neutrophil and emergency granulopoietic drivers of sepsis immune suppression and an extreme response to infection

The dysregulated host response to infection leading to organ dysfunction is highly heterogeneous. It is currently poorly delineated by sepsis as a clinical syndromic classification, thus confounding immunotherapy trials. Here we establish the pathophysiology and potential therapeutic targets of a specific extreme response to infection state (sepsis response signature SRS1), characterised by immune compromise and poor outcome. We first derive a whole blood single-cell multi-omic atlas of the sepsis response (2727,993 cells, n=39), finding an increase in IL1R2+ immature neutrophils in SRS1, which we confirmed by CyTOF and RNA-sequencing (n=53). We next uncovered high activity of neutrophil STAT3 gene expression programs in SRS1, which were shared across multiple infectioius disease settings (n=1044) irrespective of the clinical definition of the patient cohorts. We observed elevated plasma G-CSF and IL-6 in SRS1, suggesting heightened emergency granulopoiesis (EG). We therefore characterised patient and healthy control hematopoietic stem cells (HSCs) using single-cell RNA/chromatin accessibility multi-omics (29,366 cells, n=27), identifying SRS1-specific EG transcriptional skewing, together with STAT3 and EG master regulator CEBPB epigenetic signatures. Our findings establish a common cellular axis present across extreme responses to infection, reveal its hematopoietic origin, and nominate G-CSF and IL-6 as potential therapeutic targets for the SRS1 state.

The present data deposit includes processed and quality-controlled data tables for:

1. Whole blood leukocytes profiled with the BD Rhapsody platform in a cohort of 39 sepsis patients (RNA and protein count matrices, as well as their accompanying metadata table)

2. Circulating HSCs in blood profiled with the 10X multiomics platform in a cohort of 27 sepsis patients (RNA and ATAC-seq count matrices, as well as their accompanying metadata tables)

IMPORTANT NOTE: The files in Version 1.0 are incomplete (missing the cell barcode list and gene/feature names).
Please download Version 1.1 for the complete package (includes barcode and feature list files aligned to the matrices).