Inhibition of Human Erythropoiesis during Inflammation Is Mediated By High Mobility Group Box Protein 1 (HMGB1) through Decreased Commitment of Hematopoietic Stem Cells to the Erythroid Lineage and By Increased Apoptosis of Terminally Differentiating Erythroblasts

Anemia plays a significant role in the morbidity and mortality of chronic inflammatory diseases and post-sepsis recovery. Inflammation-associated anemia is further complicated as these conditions are typically refractory to erythropoietin (EPO) administration, which suggests that circulating inflammatory molecules prevent effective erythropoiesis. High mobility group box protein 1 (HMGB1), a pro-inflammatory molecule involved in sepsis has been shown to mediate anemia in a murine model of sepsis survivors. However, the effects of soluble HMGB1 on human erythropoiesis remain poorly understood. To address this issue, we studied the effect of increasing concentrations of HMGB1 on human erythropoiesis in vitro using CD34+ hematopoietic stem and progenitor cells (CB HSPCs) and a previously described 3-phase erythroid differentiation culture system to asses the effects on erythroid cell proliferation as well as the semisolid methylcellulose assay to study effects on erythroid progenitors. HMGB1 markedly reduced erythroid cell proliferation in the 3-phase culture system in a dose-dependent manner. We evaluated the contributions of early- and late-stage erythropoiesis to decreased erythroid proliferation using using flow cytometric analyses of cell surface markers interleukin-3 receptor (IL-3R), glycophorin A (GPA), CD36, CD34 (BFU-E and CFU-E) and GPA, α4-integrin and band 3 (ProEB-orthoEB). HMGB1 increased the percentage of IL-3R^pos cells (vehicle v HMGB1: 8.6% vs 15.5%), and reduced the percentage of GPA^pos cells (vehicle v HMGB1: 32.2% v 12.8%) implying decreased commitment of CB HSPCs to erythroid lineage in the presence of HMGB1. Flow cytometric analysis confirmed an increased commitment to myeloid lineage and Giemsa-stained cytospins revealed increased numbers of myelo-monocytic cells in HMGB1 cultures. Furthermore, the decreased number of erythroid progenitors generated in culture was blocked in terminal differentiation leading to increased apoptosis of erythroblasts.

Since we found increased numbers of myelo-monocytic cells and increased erythroblast apoptosis in HMGB1 treated cultures, we performed additional studies to investigate the role of soluble HMGB1 on HSPC lineage commitment and erythroblast survival signaling (i.e. EPO signaling). CB CD34+ HSPCs were seeded in fully enriched methylcellulose ± HMGB1, and hematopoietic colony numbers were quantified based on morphologic characteristics. HMGB1 increased the number of CFU-GM colonies and decreased the number of erythroid colonies (CFU-E, BFU-E). Further, erythroblasts treated with 3IU/mL EPO ± HMGB1 revealed that HMGB1 significantly reduced EPO-dependent STAT5 phosphorylation in a dose-dependent manner.

In order to begin to identify the HMGB1 receptors responsible for HSPC lineage skewing and erythroblast apoptosis we used transcriptome-wide RNA-sequencing data to examine the expression patterns of various HMGB1 receptors during differentiation of HSPCs in our culture system. Toll-like receptor 2 (TLR2), TLR4 and CXCR4 are expressed during the early-stages of erythroid differentiation while the receptor for advanced glycation end products (RAGE) is expressed only during late stages of erythroid differentiation. These findings suggest that two discrete HMGB1-receptor signaling axes may underlie the noted changes in early- and late-stage cellular differentiation in our culture system.

Taken together, our results identify HMGB1 as a potent inhibitor of human erythropoiesis. HMGB1 exerts this effect at two levels of erythroid differentiation, the first during the commitment of HSPC to erythroid differentiation and subsequently during terminal erythroid differentiation. The erythroid differentiation defect appears to be partially dependent on EPO signaling while it has no effect on the HSPC defect. We proposed that the elevated HMGB1 during inflammation could explain the persistent erythroid defect observed in patients with chronic inflammatory diseases and sepsis survivors.