Rac1^Δ/Δ;Rac2^−/− erythroblasts do not demonstrate decreased enucleation. Total inhibition of Rac-GTPases is required to inhibit enucleation of erythroblasts in long-term in vitro erythropoiesis cultures. Representative flow cytograms per Syto-16 and FSC (A) and per Syto-16 and Ter119 expression (B) demonstrating that enucleation of Rac1^Δ/Δ;Rac2^−/− erythroblasts was more efficient than that of WT erythroblasts in parallel ex vivo erythropoiesis cultures. Percentages of SYTO16^low/− cells (red) out of the Ter119⁺ cells are shown as mean ± SEM; n = 6. P < .05. (C) Up-regulation of Rac3 in Rac1^Δ/Δ;Rac2^−/− erythrocytes (representative example of 3 biologic repeats). Rac1^Δ/Δ;Rac2^−/− erythrocytes show residual Rac1 expression from Rac1^flox/flox hematopoiesis with competitive advantage.^18,27  (D) Densitometry of Rac1 and Rac3 expression in the above blot, expressed as percentage of the corresponding protein in WT red blood cells. (E) Pharmacologic inhibition of all Rac GTPases by NSC23766 at a dose of 50 and 100μM inhibited enucleation in WT (n = 8, P < .01 between 0 and 50μM, 0 and 100μM, and 50 and 100μM) and Rac1^Δ/Δ;Rac2^−/− erythroblasts (n = 5, P < .05 between 0 and 50μM, 0 and 100μM, not statistically significant between 50 and 100μM). (F) Rac1, Rac2, and Rac3 activity (GTP-bound isoforms) of WT and Rac1^Δ/Δ;Rac2^−/− erythroid cells at the enucleation stage of a long-term in vitro erythropoiesis culture after treatment with increasing concentrations of NSC23766 (representative blots of at least 3 different samples for each effector domain pull-down assay). Total Rac protein for each sample before performing pull-down is detected with polyclonal anti-Rac1, Rac2, Rac3 antibody. GAPDH from the same samples (before pull-down) is shown as loading control. (G) Densitometry of the above blot, demonstrating total Rac protein and Rac1, Rac2, and Rac3 activity normalized for GAPDH loading control for each sample and expressed as percentage of the corresponding protein expression or activity in the sample not treated with NSC23766. The faint bands in the GTP-Rac1 and Rac2 lanes of the Rac1^Δ/Δ;Rac2^−/− sample are not plotted because they represent either minimal residual activity (possible for Rac1) or a nonspecific antibody reaction (especially true for the Rac2 lane because the sample is Rac2-null).