ENT1-mediated adenosine uptake induced cyclic nucleotide accumulation and enhanced in vitro erythropoiesis. (A) CD34⁺ progenitors from peripheral blood of healthy donors were stimulated with rEPO in the absence or presence of adenosine (15 µM) at day 0 of the differentiation phase. Terminal differentiation was monitored as a function of the α4-integrin/Band3 profiles of GPA^pos cells. Differentiation steps can be followed as α4-integrin^high/Band3^low (quadrant 1 [Q1]), α4-integrin^med/Band3^med (Q2), and α4-integrin^low/Band3^high (Q3). Expression levels of α4-integrin and Band3 levels inside GPA⁺ cells are shown at days 5, 9, and 12 for control and adenosine-treated cells. (B) The percentages of α4-integrin^high/Band3^low, α4-integrin^med/Band3^med, and α4-integrin^low/Band3^high in control (black circles) and adenosine-treated (red triangles) cells are plotted in histograms. Results are expressed as the mean cell % ± standard error of the mean (SEM). Multiple comparisons of 2-way analysis of variance statistical test were performed and corresponding significant differences are shown. *P < .05; **P < .005; ***P < .001. (C) Representative images of MGG-stained Cytospins of control and adenosine-treated cells (+ADO) at day 7 of differentiation. Scale bars, 10 µm. (D) Surface expression levels of GPA and Band3 erythroid markers and CD11a and CD33 myeloid markers were evaluated by flow cytometry at day 7 of rEPO-induced erythroid differentiation. Representative histograms showing GFP^pos gated cells for shControl (filled black histograms), shENT1 (empty black histograms), and shENT1 treated with adenosine (empty red histograms) transduced cells are shown (left); gray shaded histograms correspond to nonspecific IgG staining and were used to evaluate positive staining. Quantification of 3 independent experiments is presented for each marker (right). ns, nonsignificant, *P < .05, **P < .01, ***P < .001, ****P < .0001. (E) shENT1-and shControl transduced CD34⁺ cells were stimulated with EPO and cultured in absence or presence of adenosine (15 µM). Surface expression levels of GPA were monitored at indicated days and plotted in scatter dot plot histograms as mean cell % ± SEM. Two-way analysis of variance statistical test was performed and statistical significance is indicated (n = 3 independent experiments). **P < .01; ***P < .001; ****P < .0001. (F) Intracellular concentrations of cAMP, cGMP, and adenine nucleotides inside erythroid precursors at day 3 of erythroid differentiation were measured by a liquid chromatography–mass spectrometry approach, and representative bar histograms are shown for control (black) and adenosine-treated cells (red). Results are expressed as mean ± SEM, n = 3. *P < .05; **P < .01; ***P < .001 (paired t test). (G) The phosphorylation state of CREB (Ser133) was evaluated by intracellular staining using an antiphosphoCREB antibody and analyzed by flow cytometry throughout erythroid terminal differentiation. GMFI (left) and percentages of P-CREB⁺ cells (right) are presented at the indicated days as means ± SEM of 4 independent experiments. **P < .01, ***P < .001. (H) Phosphorylation of PKA RIIβ (Ser114) was monitored at the indicated time points. GMFI (left) and the percentages of P-PKA⁺ cells (right) are shown as means ± SEM of 4 independent experiments. *P < .05, **P < .01.