Abstract 10241: Single-Cell RNA-Sequencing Reveals Hyperacute NK Cell and Monocyte Cell States Correlating with Poor Neurological Outcomes After Cardiac Arrest
Tomoyoshi Tamura,Changde Cheng,Alexandra Weissman,Mayra Pinilla-Vera,Louis Merriam,Jack Varon,Sam Saliba,Peter Hou,Raghu Seethala,Patrick R Lawler,William M Oldham,Jiyoung Choi,Kohei Ikeda,Kei Hayashida,Yohannes Tesfaigzi,Katherine Berg,Erin Bohula,David Morrow,Rebecca Baron,Michael Donnino,Fumito Ichinose,Xiang Chen,Edy Kim,
DOI: https://doi.org/10.1161/circ.144.suppl_1.10241
IF: 37.8
2021-11-16
Circulation
Abstract:Background: Lymphopenia and elevated cytokine levels suggest a role for the systemic immune response after cardiac arrest (CA). However, little is known about the phenotypes and interactions of immune cells after CA. This study aimed to investigate the immunological network after CA and identify cell states correlating with poor neurological outcomes. Methods: Peripheral mononuclear blood cells of 11 post-out-of-hospital CA patients and 3 healthy subjects were analyzed by single-cell RNA-sequencing (scRNA-seq), with validation by flow cytometry, bulk RNA-seq of sorted cell subsets, plasma levels of cytokines, and the murine model of CA. Good and poor neurological outcomes at hospital discharge were defined by Cerebral Performance Category 1-2 and 3-5, respectively. Results: The scRNA-seq analysis of 96,179 cells revealed 6 major cell lineages with several subclusters. At 6h post-CA, cells of patients with good outcomes co-clustered with healthy subjects, while patients with poor outcomes formed distinct transcriptomic clusters. This distinction was lost by 48h post-CA. At 6h post-CA, we identified pro-inflammatory and hypoxia-responsive Tim-3 + NK cell and Nectin-2 + monocyte cell states specific to poor outcomes. In a validation cohort of post-CA patients, Tim-3 + NK cells and Nectin-2 + monocyte cell states were validated by flow cytometry. Bulk RNA-seq analysis of these sorted cells reproduced the transcriptomic profile determined by scRNA-seq. Interactome analysis identified NK cell-monocyte axes in poor outcomes, including IFNG-IFNGR2. Measurement of plasma cytokine levels revealed increased IFNγ levels at 6h post-CA that were associated with poor outcomes, but both poor and good outcomes had similarIFNγ levels by 48h. A murine model of CA demonstrated that IFNγ drives poor outcomes and mortality after CA. Conclusions: In clinical CA, transcriptomic analysis at single-cell resolution revealed hyperacute cell states of NK cells and monocytes associated with poor neurological outcomes. These hyperacute cell states interact by IFNγ-IFNγR2 axis, which is pathogenic in the murine model of CA. The immunophenotypes that distinguish patients with poor or good neurological outcomes are present at 6h but are no longer detectable by 48h post-CA.
cardiac & cardiovascular systems,peripheral vascular disease