Heterozygosity for Hemoglobin (Hb) S, sickle cell trait (SCT), affects over 40 million people and confers resistance to severe infection by Plasmodium falciparum. Homozygosity for HbS, or compound heterozygosity with certain other alleles of Hb, affects over 4 million individuals and causes sickle cell disease (SCD). Hemolytic anaemia is a prominent feature of SCD and is mainly extravascular, mediated by hepatic and splenic macrophages. No ligands for this process have been identified. As many macrophage phagocytic receptors recognise carbohydrates, we surveyed surface glycan expression by sickle cells using a panel of 8 lectins and flow cytometry. Most glycans were similar to those of healthy red blood cells (RBC), except much higher expression of terminal mannose.

We investigated the structural basis for these residues using glycomic mass spectroscopy, which showed them to be N-linked high (Man5-9GlcNAc2) mannoses, a surprising conclusion as these are usually intermediates in the formation of complex glycans and not displayed on cell surfaces. High resolution microscopy revealed the mannose residues to be carried in discrete microdomains on the surfaces of sickle cells. These structures were absent on the surfaces of healthy RBC, instead being present in the membrane skeleton under the cell surface. Lectin blots and immunoprecipitation showed the mannoses to co-migrate predominantly with spectrin.

We showed these mannose-bearing structures were able to stimulate phagocytosis of RBC by using a peripheral blood derived macrophage uptake assay. Sickle RBC were taken up at high rates compared to healthy RBC and this could be inhibited by congeners of mannose. We identified the importance of a cognate ligand (CD206: the mannose receptor) using blocking antibodies and knockdown of CD206 expression using siRNA.

The in vivo and pathogenic relevance of mannose exposure was investigated by taking advantage of the heterogeneity of hemolysis in SCD. RBC with SCD (n=94), SCT (n=57) and healthy individuals (n=54) were assayed for mannose exposure by flow cytometry. SCT and healthy RBC showed no mannose exposure but high levels were found on HbSS RBC (p<0.0001). Co-incident inheritance of HbSS with higher HbF values and alleles encoding alpha-thalassaemia resulted in lower surface mannose values. Overall, markers of hemolysis (RBC count, haemoglobin, reticulocyte count) correlated well with mannose exposure (Spearman correlation coefficients -0.68, -0.40, 0.37; p=0.0001, 0.0032, 0.0063 respectively). Plasma LDH is a marker of intravascular hemolysis and correlated with overall hemolysis within SCD (r=-0.25, p=0.016), but not mannose exposure (r=0.14, p=0.19). Thus mannose exposure correlated only with extravascular hemolysis.

Identification of a ligand pair mediating rapid clearance of sickle cells raised the possibility that they also mediate enhanced clearance of SCT RBC infected by malarial parasites. Indeed, P. falciparum cultures induced mannose expression at the pigmented trophozoite and schizont stages in infected HbAA RBCs, at levels corresponding to mild hemolysis in SCD. Mannose expression in infected HbAS RBCs was even higher, with levels corresponding to severe hemolysis in SCD.

Infection with P. falciparum and selection for HbS arose only recently in human evolution, raising the question of what the physiological triggers for this mechanism are. Infection with malarial parasites causes oxidative stress. We therefore subjected healthy RBC to copper sulphate, which resulted in surface mannose exposure as well as uptake by macrophages. Oxidized SCT RBC displayed more mannose than oxidized healthy RBC.

Thus, we have identified a new cell surface 'eat me' signalling mechanism that allows inspecting macrophages to engage with the rigid membrane skeleton and phagocytose the mannose displaying cell. The mechanism is stimulated by HbS: when present in high concentrations, the mechanism is activated constitutively, resulting in sickle cell anaemia. Heterozygosity for HbS is insufficient by itself to trigger mannose exposure. However, the mechanism is primed so that oxidative stress associated with infection by P. falciparum causes greater mannose display, increased parasitized cell clearance and protection against severe malaria. These findings should allow the design of inhibitors of sickle cell haemolysis and inducers of protection against malaria.

Disclosures

Cao:University of Aberdeen: Patents & Royalties. Barker:University of Aberdeen: Patents & Royalties. Vickers:University of Aberdeen: Patents & Royalties; GSK: Equity Ownership.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution