Anti-HMGB1 monoclonal antibody rescues EPO signaling in murine model of sepsis. (A) RAGE expression levels measured by western blot in the bone marrow of WT vs Rage^−/− mice. (B) pSTAT5 levels measured by western blot in ex vivo EPO stimulation assays in the total nucleated bone marrow population from WT and Rage^−/− in the presence or absence of HMGB1. (C) Quantification of the levels of pSTAT5 in the indicated conditions expressed as fold change compared with pSTAT5 levels in WT in the absence of HMGB1. (D) Complete blood counts in WT and Rage^−/− mice 6 days after LPS treatment. (E) Seven days after HMGB1 treatment in vivo, pSTAT5 and pS6 levels measured by western blot in ex vivo EPO stimulation assays in the total nucleated bone marrow population from WT and Rage^−/− in the presence or absence of an additional dose of HMGB1. Sepsis was induced by cecal ligation and puncture (CLP) or not (eg, sham) in groups of mice. At 9 days after sepsis induction, mice were treated with isotype IgG or neutralizing anti-HMGB1 mAb (2G7) for 3 consecutive days. All data were collected at D21 after the CLP procedure. (F) Schematic of experimental design. (G) Representative flow cytograms with gating strategy of bone marrow terminal erythroid differentiation using Ter119 and CD71 surface markers. (H) Representative histograms of bone marrow ProEB pSTAT5 levels by phosphoflow. (I) Quantification of pSTAT5 levels in the ProEB, CD71^High and CD71^Low populations from bone marrow by phosphoflow expressed as fold change compared with pSTAT5 levels in mice treated with the isotype control. (J) Red cell and platelet indices of isotype and 2G7-treated CLP mice. Data presented as mean ± SEM. (C-D,I) One-way ANOVA with Tukey’s post hoc test: *P < .05; ns, not significant. (J) Unpaired, 2-tailed Student t test: **P < .01; ns, not significant.