A Roadmap from Single-Cell Transcriptome to Patient Classification for the Immune Response to Trauma

Immune dysfunction is an important factor driving mortality and adverse outcomes after trauma but remains poorly understood, especially at cellular level. To deconvolute the trauma-induced immune response, we applied single-cell RNA sequencing to circulating and bone marrow mononuclear cells in injured mice and circulating mononuclear cells in trauma patients. In mice, the greatest changes in gene expression were seen in monocytes across both compartments. After systemic injury, the gene expression pattern of monocytes markedly deviated from steady state with corresponding changes in critical transcription factors (TFs), which can be traced back to myeloid progenitors. These changes were largely recapitulated in the human single-cell analysis. We generalized the major changes in human CD14+ monocytes into six signatures, which further defined two trauma patient subtypes (SG1 vs. SG2) identified in the whole blood leukocyte transcriptome in the initial 12h after injury. Compared with SG2, SG1 patients exhibited delayed recovery, more severe organ dysfunction and a higher incidence of infection and non-infectious complications. The two patient subtypes were also recapitulated in burn and sepsis patients, revealing a shared pattern of immune response across critical illness. Our data will be broadly useful to further explore the immune response to inflammatory diseases and critical illness. Overall design: We longitudinally single-cell RNA sequenced circulating and bone marrow mononuclear cells in injured mice (3hrs, 6hrs and 24hrs representing escalating, peak and recovery phase of inflammation) and circulating mononuclear cells in trauma patients (<4hrs, 24hrs and 72hrs mainly representing the early changes after systematic injury) along with paired healthy controls.