Ineffective Erythropoiesis Negatively Impacts the Erythroblastic Island in Sickle Cell Disease

Sickle cell disease (SCD) is a genetic recessive disorder, characterized by painful episodes of vaso-occlusion, chronic hemolytic anemia, and progressive organ failure. It is one of the most common and severe monogenic diseases worldwide. SCD is caused by a point mutation in the β-globin gene, leading to the expression of an abnormal hemoglobin (HbS) that polymerizes under hypoxic conditions, driving red blood cells (RBC) sickling.

Peripheral RBCs are extensively studied in SCD, as hemolysis is the major contributor to anemia, but little is known about erythroid differentiation in this pathology. We recently demonstrated the occurrence of ineffective erythropoiesis in the bone marrow (BM) of SCD patients, characterized by the death of a significant proportion of erythroblasts between the polychromatic and the orthochromatic stages. We hypothesize that this increased cell death could negatively impact the central macrophages of the erythroblastic islands (EBIs) as they would engulf high numbers of hemoglobin-rich apoptotic erythroblasts and undergo ferroptosis because of excess iron.

To investigate our hypothesis, we used an in vitro system whereby CD14 ⁺ monocytes were isolated from peripheral blood and cultured in media supplemented with M-CSF and dexamethasone. This led them to differentiate towards anti-inflammatory tissue macrophages, with an “EBI-like” phenotype, expressing markers such as CD16, CD163, CD169, CD206.

Phagocytosis of high numbers of RBCs by “EBI-like” macrophages led to increased cell death, reactive oxygen species, and lipid peroxidation, and to the upregulation of genes involved in ferroptosis pathways such as HO-1 and SLC7A11. This was circumvented by the ferroptosis inhibitors ferrostatin-1 and liproxtatin-1. Furthermore, phagocytosis of apoptotic in vitro-generated SCD erythroblasts also induced ferroptosis in EBI-like macrophages, with increased cell death and lipid peroxidation in a time-dependent manner. Interestingly, while this was the case with hemoglobin-rich, terminally differentiating erythroblasts, phagocytosis of cells without hemoglobin, such as K562 or GPA ^- progenitor cells did not affect cell viability, indicating that the observed ferroptosis was due to the iron content of the internalized cells. We investigated the effect of increased RBC phagocytosis on the phenotype of surviving macrophages and found an M2 to M1 shift indicating a switch to a pro-inflammatory phenotype under the condition of elevated levels of intracellular iron and oxidative stress. We challenged these findings in vivo using the Townes humanized SCD mouse model. Data on 4-month-old mice showed the expected switch in erythroid production from steady-state erythropoiesis in the BM to a stress erythropoiesis in the spleen. This was associated with a decrease in the number of EBI macrophages (F4/80 ^HighVCAM+CD169+Ly6G-) in the BM of SS mice when compared to AS and AA mice, and an expansion of red pulp macrophages (RPMs) F4/80 ^HighCD169+VCAM+. Using a flow cytometry-based assay we found high lipid peroxidation levels in BM EBI macrophages of SS mice as compared to AA, suggesting occurrence of ferroptosis in these cells.

Our study suggests that IE in SCD may contribute to anemia, directly by cell death of differentiating erythroblasts, and indirectly by disrupting the erythroblastic island or by inducing its collapse following ferroptosis of the central macrophage. Our work with the Townes mouse model and BM samples from SCD patients will help to elucidate this hypothesis.