Abstract
Growing evidence defines macrophages (Mφ) as plastic cells with wide-ranging states of activation and expression of different markers that are time and location dependent. Distinct from the simple M1/M2 dichotomy initially proposed, extensive diversity of macrophage phenotypes have been extensively demonstrated as characteristic features of monocyte-macrophage differentiation, highlighting the difficulty of defining complex profiles by a limited number of genes. Since the description of macrophage activation is currently contentious and confusing, the generation of a simple and reliable framework to categorize major Mφ phenotypes in the context of complex clinical conditions would be extremely relevant to unravel different roles played by these cells in pathophysiological scenarios. In the current study, we integrated transcriptome data using bioinformatics tools to generate two macrophage molecular signatures. We validated our signatures in in vitro experiments and in clinical samples. More importantly, we were able to attribute prognostic and predictive values to components of our signatures. Our study provides a framework to guide the interrogation of macrophage phenotypes in the context of health and disease. The approach described here could be used to propose new biomarkers for diagnosis in diverse clinical settings including dengue infections, asthma and sepsis resolution.