(A) Schematic of a novel, sensitive and high-throughput miRNA labeling methodology for expression profiling. RT: reverse transcription; Bi: biotin; blue and green arrows: primers. (B) Multiple harvests of MEP (n=8), MEGA1 (n=4), MEGA2 (n=6), ERY1 (n=4), ERY2 (n=3), and ERY3 (n=2) populations were purified from human umbilical cord blood cells (Experimental Procedures) and profiled for miRNA expression. A heatmap is shown for log2-transformed data, with red color indicating higher expression and blue for lower expression. Data reflect the median expression of miRNAs within corresponding populations. Arrows point to miRNAs mentioned in text. (C) Median expression of miR-150 was plotted for each population, with the oval area proportional to the expression level. (D) miR-150 expression was measured using quantitative RT-PCR on multiple harvests of MEP (n=3), MEGA1 (n=4), MEGA2 (n=3), ERY1 (n=3), ERY2 (n=3) and ERY3 (n=3) populations. The ΔCt values (Ct (threshold cycle) of 18S minus Ct of miR-150) are shown for all samples. Note that ΔCt reflects log scale of expression. Samples indicated with black dots were also used in miRNA profiling in (B) and (C), whereas those with red circles were additional samples. (E) Megakaryocytes (CD41⁺CD61⁺) and erythrocytes (CD71⁺GlyA⁺) were FACS-sorted from the bone marrow of a healthy adult human donor. miR-150 expression was measured using quantitative RT-PCR. Data reflect 2 ^ΔCt in arbitrary units (AU) Error bars represent standard deviation.

(A) The function of miR-150 was assayed in an in vitro primary culture. CD34+ hematopoietic progenitors derived from human adult bone marrow cells were transduced with viral constructs expressing a control hairpin (shLuc), miR-150, a mutant miR-150 or miR-15b-16-2. The cells were cultured in the presence of erythropoietin (EPO) and thrombopoietin (TPO) to differentiate toward erythrocytes and megakaryocytes, and were analyzed after 10 days of differentiation, using flow cytometry with lineage markers CD41 (megakaryocytic) and GlyA (erythroid). (B) The effect of the expression constructs on the percentage of megakaryocytes (CD41⁺GlyA⁻) is shown. The data reflect a representative triplicate experiment. Error bars represent standard deviation. * P<0.03. (C) Representative flow cytometry plots are shown.

(A) The function of miR-150 was assayed in a murine transplant model. Stem/progenitor-cell-enriched bone marrow cells from donor mice were transduced with a control vector or a miR-150 expression construct. The vectors carry a GFP marker, thus labeling transduced cells and cells derived from them with green fluorescence. The mixture of transduced and non-transduced donor cells was transplanted into lethally irradiated recipients. Bone marrow and peripheral blood of recipients were analyzed 5 to 8 weeks post transplantation, when the hematopoietic system had largely recovered in the hosts. (B) The percentage of recipient bone marrow GFP⁺ or GFP⁻ megakaryocytes (CD41⁺Ter119⁻) is shown. Each dot represents data from one recipient mouse. n=7. (C) Bone marrow GFP⁺ or GFP⁻ erythrocyte (CD41⁻Ter119⁺) percentage is shown. n=7. (D) Representative flow cytometry plots on bone marrow cells with megakaryocytic (CD41) and erythroid (Ter119) markers. (E) Recipient bone marrow cells were FACS-sorted into GFP⁺ and GFP⁻ populations, and measured for PF4 expression using quantitative RT-PCR. Each pair of bars represents data from one recipient mouse. (F) The circulating platelets were analyzed in the peripheral blood of recipient animals. The ratio of GFP⁺ platelet percentage to the percentage of GFP⁺ bone marrow cells was plotted to reflect the thrombocytogenic potential of transduced bone marrow cells. n=5. (G) Representative flow cytometry plots of bone marrow cells assayed with CD71 and Ter119. The gates R1 to R4 represent immature to mature erythrocytes. (H) The percentage of R1 population among all erythrocytes (sum of R1 to R4) was used to derive a ratio between GFP⁺ and GFP⁻ population within the same recipient mouse. n=7. P<0.002. (I) Data for R4 population, similarly derived as in (I). P<2×10⁻⁴.

(A)Bone marrow cells from vector control or miR-150 recipient animals were sorted into GFP+ and GFP− populations. Megakaryocyte colony forming units (CFU-Mks) were quantitated from 100,000 sorted cells. n=4. (B) Bone marrow cells from 5FU treated mice were transduced with a control vector or miR-150 and 30,000 transduced bone marrow cells were assayed for erythroid colony forming units (CFU-E). (C) MEP cells sorted from wild-type C57BL/6J mice were assayed for CFU-Mk in the presence of PBS (vehicle), antagomir against miR-150 (anti-150) or a scrambled antagomir. 4000 cells were analyzed per assay. n=4. (D) Lin⁻Kit⁺Sca⁺ (LKS) stem cells were assayed as in (B). 1000 cells were analyzed per assay. n=4. (E) The expression of miR-150 in murine bone marrow cells was measured with quantitative RT-PCR after 3 days of treatment of PBS or antagomirs. Data reflect 2^ΔCt in arbitrary units (AU). Error bars represent standard deviation. (F) Representative erythroid colonies from control-vector- or miR-150-transduced bone marrow cells. Note that rare erythroid colonies formed from miR-150 transduced cells showed reduced cell number. (G) A representative megakaryocyte colony grown in the presence of antagomir-150 is shown. Brown color reflects megakaryocyte-specific acetylcholinesterase activity. (H) Mice were treated with PBS (control) or with phenylhydrazine to induce anemia. miR-150 expression was assayed by quantitative RT-PCR in lineage negative cells purified from bone marrow. n=3. P<0.001. Error bars represent standard deviation.

(A) K562 cells were transduced with constructs expressing GFP, mutant miR-150 or miR-150, and assayed for MYB and beta-Tubulin with western blot analysis. (B) Design of luciferase reporters for human MYB 3′ UTR, with green vertical bar indicating wild-type (wt) sites and red indicating mutant (mut) sites. The mutant miRNA site and mutant 3′UTR sites are complementary. (C) Normalized luciferase activities in 293T cells. Error bars represent standard deviation. n=8. (D) CD34+ human adult bone marrow cells were transduced with control vector (shLuc), miR-150 or two independent shRNAs against MYB in an in vitro culture as described in Figure 2. The percentage of megakaryocytes (CD41⁺GlyA⁻) was measured by flow cytometry. A representative triplicate experiment is shown. (E) The knockdown of MYB by shRNAs was measured in K562 cells, as determined by quantitative RT-PCR. The values reflect 2^ΔCt in arbitrary units (AU). (F) miR-150 and MYB expression constructs and their corresponding vector controls (shLuc, Vector) were assayed in combination in the human CD34⁺ cell culture. A representative triplicate experiment is shown. * P=0.04; ** P<0.01. (G) Representative flow cytometry plots as described in (F). Error bars in all panels represent standard deviation.