Comprehensive Analysis of Changes in the Proteome during Maturation of Nascent Human Bone Marrow Reticulocytes into Red Cells

The final step during erythropoiesis is the maturation of nascent reticulocytes generated following nuclear extrusion by orthochromatic erythroblasts into mature red cells. Reticulocyte maturation is a highly dynamic process during which the multi-lobular, motile and mechanically fragile nascent reticulocyte transforms into a highly deformable and mechanically stable biconcave discoid red cell. This maturation process lasting 48 to 72 hours occurring in conjunction with intracellular organelle loss and extensive membrane remodeling. In the present study, we explored quantitative changes in the protein expression levels during maturation of nascent reticulocytes into mature red cells to define the mechanistic basis for distinct morphological and biophysical properties of reticulocytes and red cells.

We obtained de-identified bone marrow samples from the sternum of patients undergoing cardiothoracic surgeries and highly purified nascent reticulocyte population was isolated by FACS sorting following enrichment for reticulocytes using Percoll density gradient. The proteomes of sorted nascent bone marrow reticulocytes from four independent donors were analyzed by mass spectrometry (Gautier EF et al., Blood Adv, 2018). Copy numbers per cell of each quantified protein were calculated using hemoglobin content as an internal standard. Proteomes of nascent bone marrow reticulocyte (R1) were compared to that of circulating reticulocytes (R2) and erythrocytes (E).

More than 2000 proteins were quantified in the R1 population with a total of 1364 proteins as copy number per cell in all reticulocyte samples, with a high correlation of quantifications between proteomes (Pearson correlation > 0.96). Comparison of R1 proteomes with that of R2 and E showed a marked decrease in the total protein content in conjunction with the disappearance of mitochondrial and ribosomal proteins. a4 and b1 Integrins, involved in adhesion of erythroblasts to the central macrophage in the erythroblastic island were expressed in R1 but were undetectable in R2 and E populations. These findings were further validated by flow cytometry quantitation of surface expression at R1, R2 and E populations. Similar marked decreases in cytoskeletal proteins tubulin myosin’s and dynein’s were noted during reticulocyte maturation. In contrast, we noted that the expression levels of globin’s, membrane proteins such as spectrins, adducins, tropomyosin, Band3 and carbonic anhydrases remained unchanged during maturation. Similarly, majority of proteins of the two membrane skeletal protein complexes (ankyrin and protein 4.1) remained essentially unchanged, except for glycophorin A and C, which decreased between the R1 and R2 stages probably due to vesicle trafficking, as previously documented for glycophorin A. We noticed interesting patterns in terms of content of kinases such as NME1 and MAP2K3/MEK3. NME1, a histidine kinase is the most highly expressed protein kinase in R1, and its expression levels remained unchanged during maturation. MAP2K3/MEK3 kinase is the second most expressed protein kinase but contrary to NME1 its expression decreased markedly at the R2 stage. The heme-regulated inhibitor (HRI) kinase was only quantified at the R1 stage suggesting a potential regulation of protein and heme synthesis at this stage of reticulocyte development.

In summary, our findings provide new and comprehensive insights into the proteome of human reticulocytes at two distinct stages of maturation and has implications for defining the mechanistic basis for distinct morphological and biophysical properties of reticulocytes and red cells. These findings have implications for the understanding the pathobiology of red cells in erythroid disorders such as sickle cell disease.