Single Cell Analysis and Functional Features of Circulating CD34+ Cells from Patients with Sickle Cell Anaemia Unveil a Defect in the Myeloid Lineage

Polymorphonuclear neutrophils (PMN) have long been suspected to be involved in the pathophysiology of Sickle cell anemia (SCA) since early studies have reported a higher neutrophil count in patients at steady-state and its positive correlation with SCA severity. This high leukocyte count has long be considered as a peripheral phenomenon, driven by RBC hemolysis causing systemic inflammation and neutrophils recruitment. Nevertheless, several lines of evidence suggest that hematopoietic stem/progenitor cell (HSPC) function might be perturbed in SCA. Given that recent report showed an increased apoptosis during terminal erythroid maturation of SCA cells and that erythroid and myeloid differentiation derive from the same common myeloid progenitor (CMP), we investigated a possible enhanced myelopoiesis in SCA and its causes.

First, to better understand when and how hyperleukocytosis and chronic inflammation begin in these patients, we analyzed longitudinal biological data from 1250 infants with SCA aged 1-24 months. We observed that the increase in neutrophils count began as early as the first months of life, and was statistically correlated with the drop in HbF rate and with the increase in reticulocytes count, much more than with the increase of the LDH level. These results seem to contradict the hypothesis of chronic inflammation driven by intravascular hemolysis.

To assess whether the abnormal hematopoiesis is associated with a defect on lineage commitment of Hematopoietic Stem/Progenitor Cells (HSPCs), we evaluated the phenotype of myeloid and erythroid progenitors isolated from peripheral blood (PB) of untreated patients with SCA. While the portion of the common myeloid progenitor (CMP) was similar among both groups, we observed a significant increase of granulocyte-monocyte progenitors (GMP) in SCA patients (n=6) compared to health donors (HDs) (n=9), as well as a significant decrease of erythroid progenitors (MEP). Interestingly, HSPCs from SCA PB produced more myeloid colonies in vitro than those in HD HSPCs. In addition, colony-forming assay of sorted SCA CMP displayed an increase of granulo-monocytes colonies CFU-GM and decrease of erythroid colonies BFU-E compared to HDs, indicating an increased myeloid commitment of SCA HSPCs.

To comprehensively characterize HSPC composition, we performed 10´ scRNA-seq on CD34+ HSPCs isolated from HDs (n=3) and patients with SCA (n=4). We confirmed that SCA HSPCs have a high proportion of neutrophil progenitors. More interestingly, MPP/HSC populations from SCA patients highly express several genes involved in myeloid differentiation like SELL (L-Selectin) and CSF3R (G-CSF receptor) genes, indicating an enhanced myeloid commitment of early progenitor in SCA.

To better characterize the myeloid features of these progenitors, we investigated ex-vivo granulopoiesis of PB HSPCs from SCA patients (n=10) and HDs (n=9) in presence of G-CSF as CSF3R is highly expressed in early SCA progenitors. While cell proliferation was similar among both groups, unexpectedly SCA myeloid cells presented a significant delayed expression of CD11b (p<0.01). Interestingly, we also observed an overexpression of monocytic markers HLA-DR and CD36 in a subset of differentiated cells, suggesting a potential commitment of part of the myeloid progenitors to a monocyte lineage, confirmed by MGG staining.

To further understand this defect, we performed RNAseq analysis of the GMP isolated from PB of SCA patients (n=4) and HD (n=4). We identified a total of 420 differentially expressed genes (DEGs). Among these DEGs, significantly downregulated genes in SCA included granulocyte differentiation program such as GFI1, C/EBPa, and JAK3. In return, significantly upregulated genes included KLF4, supporting the monocytic bias observed in the early myeloid progenitors.

In conclusion, these data reveal the myeloid-primed signature in PB HSPCs of patients with SCA, likely responsible for the elevated PMN and monocyte count in these patients. This initial breakthrough in the understanding of the development of hyperleukocytosis and chronic inflammation in these patients will open up new early therapeutic options. Moreover, since CD34+ cells of the peripheral blood are the main source of HSPCs for the development of gene therapy in SCA, deciphering their functional feature and myelo-monocytic skewing could be of interest to guide the gene editing strategy.