The isolation of reticulocyte-free human red blood cells

We depleted reticulocytes from erythrocytes of both sickle cell disease (SCD) subjects and healthy controls by four methods: fluorescence-activated cell sorting (FACS), Miltenyi immunomagnetic depletion (MACS), a combination of these methods (FACS + MACS) and Percoll density separation. The efficiency of these methods was assessed by new methylene blue staining and manual enumeration of the reticulocytes. FACS sorted erythrocytes from reticulocytes based on size and granularity, as well as the absence of dsDNA staining. MACS depleted reticulocytes from erythrocytes based on the immunoaffinity to CD36 and CD71. Reticulocytes from healthy controls were depleted to <= 0.1% using either the FACS or MACS method (alpha = 0.1). Reticulocytes from SCD subjects were depleted from 13.6% +/- 0.52% to 5.45% +/- 0.33% using MACS (n = 2), and from 10.9% +/- 0.47% to 2.0% +/- 0.2% using FACS (n = 4, alpha = 0.05). When combining FACS with MACS (n = 3), the percentage of reticulocytes was decreased in SCD samples from 13.0% +/- 0.51% down to 1.5% +/- 0.17% (alpha = 0.1). Sedimentation through 75% percoll resulted in control and SCD samples being reduced from 0.27% +/- 0.6 (control) and 6.93% +/- 0.8 (SCD) reticulocytes to <4.8 reticulocytes per million control RBCs and <2.5 per million SCD RBCs. This same method results in <2.1 leukocytes per million control RBCs and <3.7 per million SCD RBCs. We conclude that the percoll density method described here is the most effective method for isolating RBCs for proteomic analysis.