Advertisement

Data supplements

  • Supplemental materials for: Tipping et al

    Electrophoretic mobility shift assay
    Bandshift experiments were essentially performed as previously described.26 Briefly, nuclear extracts were added to a mix of 32P-dCTP�labelled GATA oligonucleotide in binding buffer (20mM HEPES, 50mM NaCl, 1mM MgCl2, 4% Ficoll). Anti�GATA-2 antibody (sc9008, Santa Cruz) was added to these assembled mixes where appropriate. Mixes were run on 4% acrylamide (19:1) gels in 0.25 × TBE at 250V before drying and autoradiography.

    Transplantation assays
    All animal studies were approved by our institutional review board and animals maintained in accordance with local regulations. For 8-week engraftment analyses, sorted, transduced and washed CD34+CD38 cells were injected into the tail-vein of sublethally-irradiated (350 cGy) NOD/SCID mice. Each mouse received around 50,000 cells admixed with untransduced CD34+CD38+ cells and 1 × 106 irradiated CD34-depleted accessory CB cells. Aliquots of transduced cells were maintained in culture for 72hrs for assessment of GFP expression. Small bone marrow aspirates were taken at 4 weeks, and animals sacrificed at 8 weeks. For later experiments, transduced and washed CD34+ cells were injected directly into the bone cavity of anesthetized animals (~3.8 × 105 each). For all transplanted animals, both tibias and femurs were flushed with PBS + 5% FCS, red blood cells lysed with Gey�s solution, and cells stained with anti-human CD45-APC, anti�CD19-PeCy7 and anti�CD33-PE (Becton-Dickinson) for flow cytometry using gates based on unstained engrafted controls and stained unengrafted samples.

    Files in this Data Supplement:

    • Figure S1 (JPG, 876 KB) -
      (A) Growth suppression of CB CD34+ cells does not confer reduced Pyronin Y staining similarly to enforced GATA-2 expression. The first two panels re-present data shown in Fig. 2I, together with Hoechst/Pyronin Y staining patterns of CB CD34+ cells treated with 5ng/mL TGFβ1, 50 µM Olomoucine, 20 µM Roscovitine or 18Gy of γ-irradiation. All cells were treated for 18 hours in culture in the presence of SCF, TPO and Flt3L, then fixed in ethanol and stained with Hoechst and Pyronin Y. Only cells deprived of cytokines show a marked decrease in Pyronin Y staining similar to that observed with enforced expression of GATA-2. (B) Enforced GATA-2 expression confers increased quiescence on CD34+CD38+ cells. Hoechst 33342 and Pyronin Y staining profiles of CB CD34+CD38+ cells from a single representative experiment are shown, as shown in Fig. 2F. Briefly, samples were split and infected with lentiviral vectors, and sorted 3 days later on CD34 and CD38 before staining with Hoechst and Pyronin Y. Hoechst 33342 and Pyronin Y staining profiles are shown for the indicated populations from a representative sample. The mean percentage of quiescent CD34+CD38+ cells is shown in Fig. 2H. (C) Engraftment and in vivo reconstitution of NOD/SCID animals by CD34+CD38 CB cells is inhibited by enforced GATA-2 expression in a level-dependent manner. Staining for human CD45 and GFP signal are shown for all engrafted animals from the Vector-transduced and GATA-2��transduced groups at 4 and 8 weeks after transplantation (aspirate samples at 4 weeks are shown next to samples from the same animal sacrificed at 8 weeks).

Article Figures & Data

Figures

  • Figure 1

    Ligand-activated GATA-2/ER chimeric form reversibly induced quiescence in a manner reproduced with constitutive lentiviral expression of FLAG–GATA-2. (A) β-Estradiol (1 μM) induces G0/G1 accumulation of Ba/F3 cells expressing GATA-2/ER (top) concomitant with increased quiescence as measured by DNA staining with Hoechst 33342 and polysomal RNA staining with Pyronin Y (bottom). One of 3 representative experiments is shown, with the percentage of G0/G1 cells falling in the G0 gate indicated. (Bi) Withdrawal of ligand allows cells to progress into S phase and to exit quiescence, as assessed by reduced G0/G1 proportions indicated (top) and increased staining with Pyronin Y (bottom). (Bii) Proportion of G0/G1 cells falling in the G0 gate from panel Bi are plotted. One of 2 representative experiments is shown. (C) Cell growth, attenuated by GATA-2/ER activation, resumes on ligand withdrawal. (D) Return to normal cell growth lags behind reduction in the proportion of quiescent cells as measured by Hoechst/Pyronin Y. (E) Bicistronic lentiviral expression construct used to drive enforced expression of FLAG-tagged GATA-2 and GFP. (F) Constitutive expression of FLAG–GATA-2 also increases quiescence of Ba/F3 cells growing in IL-3. The mean of 2 representative experiments is shown. Error bars indicate SEM. (G) The GATA-2–transduced cells are outgrown by untransduced cells in a GATA-2–dependent manner; n = 2 experiments.

  • Figure 2

    High expression of endogenous GATA-2 correlates with quiescence of human cord blood (CB) cells. (A) Quiescent and cycling CD34+CD38 cells were sorted from freshly thawed CB on the basis of low and high Pyronin Y staining, respectively. (B) Quantitative RT-PCR analysis from 2 independent samples showed higher expression of GATA-2 relative to HPRT in the quiescent G0 cells. HPRT was expressed at similar levels in these subsets (data not shown). Error bars indicate SEM. (C) The bicistronic lentiviral expression construct in Figure 1E was used to drive enforced expression of FLAG-tagged GATA-2 and GFP. Transduced CD34+ cells selected on the basis of GFP expression (i) were used for quantitative RT-PCR for GATA-2 (ii) to show that expression of GATA-2 is higher in GFPhi cells. (D) DNA-binding competency and expression level of FLAG–GATA-2 expressed in sorted GFP+ total CD34+ CB cells as assessed by electrophoretic mobility shift assay using 32P-labeled GATA oligonucleotide (sequence, 5′ to 3′ TATTTTTATCTGATAGGAAGT). (E) Western blot showing expression level of lentivirally expressed GATA-2 in freshly isolated unstimulated CB CD34+ cells, or after transduction with vector or GATA-2. Expression relative to β-tubulin (normalized to ratio in vector-transduced cells) is shown below the blot. (F) CB CD34+ samples were split, infected with lentiviral vectors, and sorted 3 days later on CD34 and CD38 before staining with Hoechst and Pyronin Y. Hoechst 33342 and Pyronin Y staining profiles are shown for the indicated populations from a representative sample. The mean percentage of quiescent GFP+CD34+CD38 (G) and quiescent GFP+CD34+CD38+ cells (H) falling in the G0 gates is presented from 3 independent samples. Error bars indicate SEM. (I) CB CD34+ cells in culture in SCF, TPO, and Flt3L were washed to remove cytokines and left in culture overnight; a control culture had cytokines added back. Eighteen hours later the cells were fixed in ethanol and stained with Hoechst and Pyronin Y. (J) Enforced GATA-2 expression confers a profound growth defect. Transduced GFP+CD34+CD38 cells were put into culture in SCF, TPO, and Flt3L and counted at intervals. Cell density per milliliter is shown from 1 of 3 representative experiments. (K) Cells were sampled at day 2 and day 11 and stained for annexin V by flow cytometry. Cells with enforced expression of GATA-2 were not more apoptotic than those expressing empty vector. Similar data were obtained throughout 4 replicate in vitro proliferation experiments. (L) The proliferative defect is predominantly observed in cells expressing high levels of GATA-2. Samples of transduced CD34+ cells with the highest and lowest levels of GFP expression were sorted and cultured as before. Cell density per milliliter is shown from a representative experiment.

  • Figure 3

    In vitro stem and progenitor cell function is inhibited by constitutive expression of GATA-2. (A) Transduced CD34+CD38 cells show a GATA-2–specific loss of LTC-IC activity in vitro. Mean frequency (± SEM) of LTC-ICs in the transduced populations is shown for 5 independent experiments. (B) Data from 1 representative experiment shown in panel A is presented as the cell number plated per well versus the number of wells that did not contain an LTC-IC. (C) LTC-IC frequency in transduced CB CD34+CD38 cells sorted by GFP expression level. A single representative experiment is shown. (D) Same data as in panel C, plotted as the cell number plated per well versus the number of wells that did not contain an LTC-IC. (E) CFC frequency in transduced CB CD34+ cells. (n = 4; error bars indicate SEM). (F) Same data as in panel E, plotted for colony type distribution. Mean proportions of total colonies of myeloid (G/M/GM), erythroid (E), or mixed (GEMM) types are shown from 4 independent experiments. G indicates granulocyte; M, macrophage; GM, granulocyte/macrophage; GEMM, granulocyte/erythroid/monocyte/macrophage; n = 4; error bars indicate SEM. (G) Photomicrograph of representative erythroid colonies from transduced CB CD34+ cells (Nikon SMZ1500, 1× objective, Nikon DXM1200F camera using ACT-1 v.2.12; Nikon UK Limited, Kingston Upon Thames, United Kingdom). (H) Total CFC colony number in transduced CB CD34+ cells sorted by GFP expression level. A single representative experiment is shown. (I) Enforced expression of GATA-2 in CD34+CD38 cells similarly inhibits colony formation but (J) here leads to a modest increase in erythroid colony output; n = 4; error bars indicate SEM.

  • Figure 4

    Enforced GATA-2 expression alters expression of genes involved in cell-cycle control but does not confer quiescence by MEF repression or CKI induction. (A) Quantitative RT-PCR of C/EBPα, Gfi-1, Egr2, Pu.1, and MEF in CB CD34+CD38 cells on enforced expression of GATA-2. Expression relative to HPRT (n ≥ 4; error bars indicate SEM). P < .01 for all. (B) Enforced expression of MEF from an LNGFR-expressing bicistronic construct does not prevent the conferral of quiescence by enforced expression of GATA-2 in CD34+ cells. Cells cotransduced with all combinations of GFP- and LNGFR-marked expression vector, empty or encoding GATA-2 or MEF, respectively, were stained with Hoechst and Pyronin Y. Enforced MEF expression has positive effects on cell cycle independently of GATA-2, but the effect of enforced GATA-2 expression is dominant to this. (C) CKIs p16INK4A, p19INK4D, p21CIP1, p27KIP1, and p57KIP2 are not induced by enforced GATA-2 expression in CB CD34+CD38 cells by quantitative RT-PCR relative to HPRT (n ≥ 4l error bars indicate SEM). (D) Enforced expression of GATA-2 in CB CD34+ cells does not cause induction of p27KIP1 as measured by intracellular flow cytometry. (E) Enforced expression of GATA-2 induces quiescence in p21−/− mouse LinKit+Sca+ cells as measured by Hoechst/Pyronin staining, but (F) p27KIP1 is not induced in these cells. (G) Expression of CCND3, CDK4, CDK6, and MCM5 is reduced by enforced GATA-2 expression in CB CD34+CD38 cells, whereas HES1 expression is induced (quantitative RT-PCR relative to HPRT). (n ≥ 4 except n = 1 for HES1; P < .05 for all except MCM5 and HES1. MCM5 expression is reduced in 4 of 5 independent experiments and as a result P > .05). (H) Expression of CDK4, CDK6, and β-tubulin by Western blot in CD34+ cell extracts prepared from freshly isolated unstimulated cells and cells transduced with empty vector and GATA-2 after 3 days in SCF, TPO, and Flt3L. Densitometric analysis of expression relative to β-tubulin, normalized to the ratio in vector-transduced cells, is shown below each panel.

  • Figure 5

    Engraftment and in vivo reconstitution of NOD-SCID animals by CD34+CD38 CB cells is inhibited by enforced GATA-2 expression in a level-dependent manner. (A) GFP expression level after 3 days in vitro culture, showing GFP intensity and transduction efficiency of transplanted GATA-2–transduced populations. Staining for human CD45 and GFP signal are shown for 1 representative animal from each group at 4 and 8 weeks (same animal sampled at 4 weeks is shown at 8 weeks). All animals are represented in Figure S1C. (B) High GATA-2–mediated expansion defect in NOD-SCID animals. Mean proportions of engrafted GFP+ cells falling in the upper gate shown in panel A, with high expression of GFP, are plotted for all engrafted animals at 8 weeks (n = 4 for each transgene). Error bars indicate SEM. (C) Expression level-dependent inhibition of hematopoiesis in NOD-SCID mice is not due to alterations in homing. GFP expression level after 3 days in vitro culture and in vivo 4 weeks after intraosseous injection. Engrafted cells expressing high levels of GATA-2 are not observed at the same frequency as in the initial transplant material. Cells were injected directly into the bone marrow cavity after only 5 hours of exposure to lentivirus. Staining for human CD45 and GFP signal are shown. (D) Proportions of engrafted human GFP+ cells falling in the upper GFP gate shown in panel C are plotted for the populations given as a transplant (dark gray) and from all engrafted animals at 4 weeks after intraosseous injection (black; n = 2 for vector-transduced cells; n = 1 for GATA-2). (E) Similar to panel C, cells were injected directly into the bone marrow cavity after 5 hours of transduction, and engraftment was assessed 12 weeks later. (F) Similar to panel D, proportions of engrafted human GFP+ cells falling in the upper GFP gate shown in panel E are plotted for the transplant (dark gray) and from all engrafted animals at 12 weeks after intraosseous injection; n = 2 for both groups. (G) Relative lymphoid and myeloid reconstitution is affected by enforced GATA-2 expression. For each animal, the proportion of human CD45+GFP and human CD45+ GFP+ cells falling in CD19+ lymphoid and CD33+ myeloid gates are plotted; n = 4 animals in each group. Vector-transduced and GFP cells show no significant difference in lineage distribution, whereas GATA-2–transduced GFP+ cells read out predominantly in the myeloid lineage. (H) Ki-67 expression of engrafted GFP-positive cells sorted from the bone marrow of NOD-SCID mice. Many more vector-transduced cells express the proliferation marker Ki-67 than GATA-2–transduced cells. Plots are shown from 1 of 2 representative experiments, where 1 of 2 vector-engrafted animals is compared with a single GATA-2–engrafted animal. (I) Plots showing proportions of quiescent cells lacking expression of Ki-67 from all animals in the experiment represented in panel H.