Hypoxia leads to both secondary polycythemia and activation of complement cascade (CC) in bone marrow (BM) (

) by classical pathway (
) and important CC proteins (e.g., C3 and C5) are produced by BM macrophages in erythropoietic islands. These observations prompted us to study the potential role of CC cleavage fragments in erythropoiesis. We noticed that human and murine cells from the erythroid lineage express functional G-protein coupled C3aR and C5aR and that C3 and C5 cleavage fragments - anaphylatoxins C3a, desArgC3a, C5a and desArgC5a enhance SDF-1-mediated chemotaxis of erythroid progenitors and induce in normal erythroblasts

  • adhesion to fibronectin,

  • calcium flux and

  • phosphorylation of MAPKp42/44 and AKT.

To further explore the potential role of the C3a–C3aR and C5a–C5aR axes in erythropoiesis, we focused on C3aR−/− and C5−/− mice. We noticed that C3aR−/− animals compared to wild type (wt) littermates have

  • lower hematocrits and

  • display delayed recovery of hematocrit values after sublethal irradiation.

More importantly, we noticed that the number of clonogeneic BFU-E progenitors in BM of C3aR−/− mice, but not CFU-GM and CFU-Meg, is reduced by ∼65%. Similarly, however, less pronounced changes were observed in C5−/− mice. Thus, in order to learn more about the role of the C3a–C3aR and C5a–C5aR axes in erythropoiesis we moved to an in vitro model and co-stimulated normal human CD34+ or murine SKL cells with CC anaphylatoxins in the presence of suboptimal doses of EpO and KL, but no effect on BFU-E colony formation was observed. We found, however, that CC anaphylatoxins increase in early erythroid cells expression of mRNA for VEGF and MMP-2 (C3a > desArgC3a > C5a > desArgC5a) - which are important for maturation/enucleation and egress of erythrocytes from the BM into peripheral blood. These data were subsequently confirmed by ELISA (VEGF) and zymography (MMP-2). Furthermore, we noticed that at the morphological level C3aR−/− and C5−/− mice as compared to wt littermates had a significantly reduced number of erytropoietic islands in BM. This suggests that in the absence of the CC anaphylatoxins, maturing erythroblasts do not traffic properly in the BM microenvironment and do not associate with C3 and C5 producing central macrophages in erythropoietic islands. In conclusion, the C3a–C3aR and C5a–C5aR axes modulate the interaction of erythroid progenitors with the hematopoietic environment and play an important and underappreciated role in developmental trafficking of erythroid cells in BM. Thus, CC anaplylatoxins may enhance erythropoietic effects of erythropoietin in polycythemias related to hypoxia. Furthermore, since hypoxia activates CC, we postulate that CC anaphylatoxins could be involved in the pathogenesis of hypoxia-related polycythemias and C3aR or C5aR antagonists could find a potential supportive role in controlling erythropoiesis.

Author notes

Disclosure: No relevant conflicts of interest to declare.