Blood monocytes serve as the first line of host defense and are equipped to recognize and respond to infection by triggering an immune-inflammatory response. While most information on these cells comes from in vitro studies in humans or in vivo studies in mice, little is known about monocytes under human disease conditions. We investigated the role of monocytes during sepsis and its resolution in humans. A transcriptomal and functional analysis of blood monocytes from patients during gram negative sepsis and at recovery was performed. Monocytes from sepsis patients showed upregulation of a large number of pro-inflammatory genes and cytokines/chemokines, consistent with an ongoing systemic inflammation. However, these cells showed impairment to ex vivo endotoxin (LPS) challenge, displaying a quantitative decrease in the number of LPS-inducible genes. Moreover, they downregulated the expression of several pro-inflammatory cytokine/chemokine genes, activation marker genes and transcription factors associated with monocyte/macrophage activation, upon ex vivo LPS stimulation. Functionally, they downregulated expression of inflammatory cytokines/chemokines and antigen presentation-related molecules and functions. In contrast, genes and functions related to phagocytosis, anti-microbial activity and tissue remodeling where remained unaffected or even enhanced . Collectively, our observations suggest a genetic and functional re-programming of these cells during human sepsis progression. Understanding the molecular mechanisms which regulate this re-programming will allow to devise strategies which could modulate the response of these cells and hence, disease progression.

Blood monocytes from gram-negative sepsis patients during sepsis (Sepsis) and following their recovery (Recovery/Basal) as well as healthy donor (control) were isolated. Thereafter, these cells were treated ex vivo with or without LPS for 3h and analysed for transcriptomic study.

Lipopolysaccharide exposure (2021 ICD-10-CM code* = Not Applicable)

Array data were extracted at the probe level with background correction using Illumina's BeadStudio software. These raw data were then quantile normalized in R v2.15.2