HOXB4 Overexpression Promotes ... - Wiley Online Library

9 downloads 226 Views 964KB Size Report
colonies, representing the most primitive of hematopoietic myeloid progenitors .... Dr Mickie Bhatia (University of Western Ontario, London, ON,. Canada) and his ...
EMBRYONIC STEM CELLS HOXB4 Overexpression Promotes Hematopoietic Development by Human Embryonic Stem Cells KRISTIAN M. BOWLES, LUDOVIC VALLIER, JOSEPH R. SMITH, MORGAN R. J. ALEXANDER, ROGER A. PEDERSEN Department of Surgery, University of Cambridge, Cambridge Institute for Medical Research, Cambridge, United Kingdom Key Words. Human embryonic stem cells • Hematopoiesis • Homeobox genes

ABSTRACT Human embryonic stem cells (hESCs) are a potential source of hematopoietic cells for therapeutic transplantation and can provide a model for human hematopoiesis. Culture of hESCs on murine stromal layers or in stromal-free conditions as embryoid bodies results in low levels of hematopoietic cells. Here we demonstrate that overexpression of the transcription factor HOXB4 considerably augments hematopoietic development of hESCs. Stable HOXB4-expressing hESC clones were generated by lipofection and could be maintained in the undifferentiated state for prolonged passages. Moreover, differentiation of hESCs as embryoid bodies in serum-containing medium without the use of addi-

tional cytokines led to sequential expansion of first erythroid and then myeloid and monocytic progenitors from day 10 of culture. These cells retained the capacity to develop into formed blood elements during in vitro culture. Consistent with the development of committed hematopoietic cells, we observed the expression of transcription factors known to be critical for hematopoietic development. We thus demonstrate successful use of enforced gene expression to promote the differentiation of hESCs into a terminally differentiated tissue, thereby revealing an important role for HOXB4 in supporting their in vitro development along the hematopoietic pathway. STEM CELLS 2006;24:1359 –1369

INTRODUCTION

graftment are tempered by observations that cytokine-expanded, cord blood-derived human HSCs lose repopulating ability with prolonged cytokine co-culture [11, 12]. If hESCs are to be clinically useful in generating HSCs, then additional strategies are needed to achieve their development and to augment the selfrenewal of HSCs and their descendants, once formed. There is mounting evidence that the homeodomain transcription factor, HOXB4, has an important role in HSC selfrenewal. HOXB4 has been found to be expressed in a CD34positive human cell population enriched for HSCs and is subsequently downregulated on differentiation of these cells to committed hematopoietic progenitors [13]. Expression of HOXB4 in adult human bone marrow and cord blood has been shown to be effective in enhancing the repopulating effect of HSCs without impairing differentiation or inducing leukemia [14]. Importantly, expression of HOXB4 has been shown to confer definitive lympho-myeloid engraftment potential on murine bone marrow and embryonic stem cells [15, 16]. Human and murine HSC expansion has also been achieved with the use of recombinant HOXB4 protein [17, 18]. Consistent with such a role, HOXB4 knockout mice have reduced cellularity of

Human embryonic stem cells (hESCs) are pluripotent cells derived from the inner cell mass of embryos cultured from the blastocyst stage [1, 2]. The embryonic origin of hESCs makes them a valuable model for studying early human development, and their pluripotency makes them a potential future source of cells for therapeutic transplantation. Differentiation of hESCs in vitro as embryoid bodies (EBs) and coculture of hESCs with murine stromal cells have resulted in low levels of spontaneous generation of hematopoietic precursors. This process can be promoted to some extent by adding cytokines to the medium in which the cells are grown [3–9]. Hematopoietic stem cells (HSCs), the pluripotent progenitors of the entire hematopoietic lineage, have recently been shown capable of minimally repopulating hematopoiesis in primary and secondary recipient sheep models after derivation from hESCs (previously co-cultured with the stromal layer S17 in the absence of additional cytokines), although levels of chimerism were still very low [10]. The enduring goal is to achieve more robust and sustained levels of engraftment of hESC hematopoietic progeny in animal models. However, expectations that in vitro cytokine or stromal layer driven amplification of hemogenic precursors will lead to improved en-

Correspondence: K. M. Bowles, Department of Surgery, University of Cambridge, Cambridge Institute for Medical Research, Hills Road, Cambridge, CB2 2XY U.K. Telephone: ⫹44 (0)1223-763237; Fax: ⫹44 (0)1223-763350; e-mail: [email protected] Received May 9, 2005; accepted for publication January 1, 2006; first published online in STEM CELLS EXPRESS January 12, 2006. ©AlphaMed Press 1066-5099/2006/$20.00/0 doi: 10.1634/stemcells.2005-0210

STEM CELLS 2006;24:1359 –1369 www.StemCells.com

Expression of HOXB4 in hESCs

1360

hematopoietic organs and reduced numbers of hematopoietic progenitors [19]. Recently, one group has shown that progeny of cytokine-expanded hESCs were capable of short-term hematopoietic engraftment in SCID mice. Lentiviral-based ectopic expression of HOXB4 in such cells failed to enhance engraftment [20], leading the authors to hypothesize that in their approach gain of function of HOXB4 may not be sufficient for engraftment of hESC progeny. The very low level of engraftment in their work (⬍1% human chimerism at 8 weeks post-transplant) provides evidence that a number of technical improvements in both the derivation and selection of candidate HSCs and in the assay itself are required. Given compelling evidence from the murine system that HOXB4 gain of function is necessary for substantial long-term multilineage engraftment [16], we have chosen therefore to investigate further whether enforced stable HOXB4 expression in hESCs could promote HSC development, using recently developed techniques for transgene expression to enhance gene function in hESCs [21]. Our findings demonstrate that gain of function of this homeobox transcription factor can support development of hESCs into terminally differentiated blood elements apparently by promoting the expansion of hematopoietic progenitors and their descendant populations.

MATERIALS

AND

METHODS

hESC Culture The hESC line H9 was obtained from WiCell (WiCell Research Institute, Madison, WI, http://www.wicell.org). The cells were maintained in the undifferentiated state by culture on irradiated murine embryonic fibroblasts (MEFs) as described [1, 21, 22]. To avoid any consequences of chromosomal changes [23], only karyotypically normal cells at passage 45 to passage 65 were used in these experiments.

containing approximately 2,000 hESC colonies each were plated onto one six-well gelatin-coated plate containing 5 ⫻ 104 feeders. After 48 hours, the cells were transfected using Lipofectamine2000(Invitrogen,Paisley,U.K.,http://www.invitrogen. com), as described [21]. Three days after transfection, the cells were passed onto 60-mm gelatin-coated tissue-culture plates containing puromycin-resistant mouse fetal fibroblasts as feeders. After 3 additional days, puromycin (final concentration 1 ␮g/ml) was added. Puromycin-resistant colonies that appeared after approximately 12 days of selection were picked, dissociated, and plated onto 24-well gelatin-coated, feeder-containing plates, and expanded for further analysis (see below). Each clone was amplified from a single puromycin-resistant colony. Three individual HOXB4-expressing hESC clones were assessed, hereafter designated HOXB4.1, HOXB4.2, and HOXB4.3.

Reverse Transcription Polymerase Chain Reaction Analysis Cycle conditions were as follows: 94°C for 5 minutes, then 40 cycles of 94°C for 30 seconds, 60°C for 30 seconds, and 72°C for 30 seconds, followed by a final extension at 72°C for 5 minutes. The primers used were as follows: HOXB4-untranslated region (254 base pairs [bp]) 5⬘-ATC TGT CTT GTT TCC TCT GCC G-3⬘ and 5⬘-TGA ATG GGC ACG AAA GAT GAG G-3⬘, HOXB4-translated region (410 bp) 5⬘-TCT GTC CCC TCG GGC TCC TGC G-3⬘ and 5⬘-GGC AAC TTG TGG TCT TTT TTC C-3⬘, ␤-2-microglobulin (288 bp) 5⬘-ACT GAA AAA GAT GAG TA T GCC TGC CGT GTG AAC C-3⬘ and 5⬘-CCT GCT CAG ATA CAT CAA ACA TGG AGA CAG CAC T-3⬘.

Quantitative Reverse Transcription Polymerase Chain Reaction

The cells were incubated for 20 minutes in 1 mg/ml Collagenase IV (Gibco, Paisley, U.K., http://www.invitrogen.com) and scraped off the Petri dishes. The detached colonies were washed once in the differentiation medium (KO-D-MEM supplemented with 20% fetal bovine serum [FBS; HyClone, Logan, UT, http://www.hyclone. com], 2 mM L-glutamine, 0.1 mM ␤-mercaptoethanol, and 1% nonessential amino acids) and then resuspended in the differentiation medium. The cells were then transferred to sterile ultra-lowattachment six-well plates (Corning Life Sciences, Acton, MA: http://www. corning.com). The EBs were incubated at 37°C in 5% CO2. During differentiation, the medium was replaced every 2–3 days [4]. To assess the effect of additional cytokines, 300 ng/ml stem cell factor, 300 ng/ml Flt3 ligand, 10 ng/ml interleukin 3 (IL3), 10 ng/ml IL6, 50 ng/ml granulocyte-colony stimulating factor, and 50 ng/ml bone morphogenetic protein-4 (all from Peprotech EC, London, U.K., http://www.peprotech.com) were added to the differentiation medium [4].

Total RNA was extracted using the RNeasy Mini Kit (Qiagen) which included a RNase-free DNase digestion step. Following manufacturers instructions, 0.5 ␮g of RNA was reverse transcribed using 200 units of Superscript II (Invitrogen) and 0.5 ␮g of random primers (Promega). Real-time Taqman polymerase chain reaction (PCR) was performed using an ABI 7700 with 1⫻ Mastermix (Eurogentec, Seraing, Belgium, http://www. eurogentec.be) and 500 ng of cDNA. Primers and probes used in the real-time Taqman PCR were as follows: HOXB4, 1⫻ Assays-on-Demand (Hs 00256884_m1; Applied BioSystems, Foster City, CA, http://www.appliedbiosystems.com); stem cell leukemia gene (SCL), 1⫻ Assays-on-Demand (Hs00268434_m1, Applied BioSystems); GATA1, 1⫻ Assays-on-Demand (Hs00231112_m1; Applied BioSystems); hydroxymethylbilane synthase (HMBS), forward primer GGAGCCATGTCTGGTAACGG, reverse primer CCACGCGAATCACTCTCATCT, probe TTTCTTCCGCCGTTGCAGCCG. Cycle conditions were as recommended by Eurogentec. The comparative threshold cycle (CT) method was used to analyze data, with HOXB4, SCL, and GATA1 expression levels calibrated to the expression of the housekeeping gene HMBS.

Generation of HOXB4-Expressing hESC Lines

Southern Blot Analysis

Differentiation of hESCs as Embryoid Bodies

HOXB4-expressing and “human recombinant” green fluorescent protein (GFP)-expressing (Stratagene, La Jolla, CA, http:// www.stratagene.com) constructs were generated using the pTP6 expression vector [24]. For stable transfection with vectors encoding human HOXB4 or GFP, three confluent 60-mm plates

Genomic DNA (10 ␮g) from distinct HOXB4-transfected hESC clones was digested with EcoR1 and electrophoresed through 1% agarose gels followed by transfer to Hybond N nylon membrane (Amersham Biosciences, Chalfont St. Giles, U.K., http:// www4.amershambiosciences.com). The probe used for hybridiza-

Bowles, Vallier, Smith et al. tion to the Southern membrane was labeled with [32P]CTP using the High Prime DNA Labelling Kit (Roche Diagnostics, Basel, Switzerland, http://www.roche-applied-science.com). The probe used was a 0.3-kilobase pair fragment from the CMV promoter region of the pTP6 expression vector [24]. After hybridization, the membrane was exposed to X-ray film for 5 days.

Immunocytochemistry Colonies of wild-type and HOXB4-transfected hESCs cells were fixed in 4% (wt/vol) paraformaldehyde (Sigma-Aldrich, Gillingham, U.K., http://www.sigmaaldrich.com). Cells were stained with either HOXB4 (I12; Developmental Studies Hybridoma Bank, Iowa City, IA, http://www.uiowa.edu/⬃dshbwww), Oct-4 (C-10; Santa Cruz Biotechnology Inc., Santa Cruz, CA, http://www.scbt .com), or the appropriate isotype control antibodies (HOXB4, rat IgG2a [BD Biosciences, San Diego, http://www.bdbiosciences. com]; Oct-4, mouse IgG2b [DakoCytomation, Glostrup, Denmark, http://www.dakocytomation.com]), and subsequently, secondary antibody (HOXB4, fluorescein-isothiocyanate-conjugated rabbit anti-rat IgG [DakoCytomation]; Oct-4, Cy3-conjugated donkey anti-mouse IgG [Chemicon, Temecula, CA, http://www.chemicon. com]) was applied for 1 hour at room temperature. Finally cells were incubated for 90 seconds with Hoechst 33258 (0.1 mg/ml in phosphate-buffered saline [PBS]; Sigma-Aldrich). The cells were viewed in PBS by epifluorescence microscopy with the appropriate filter using a Zeiss Axiovert 200M (Carl Zeiss, Jena, Germany, http://www.zeiss.com).

Western Blot Analysis Fifty micrograms of total protein was run on a NuPage gel (Invitrogen). Gels were blotted onto a nitrocellulose membrane (Amersham Biosciences), which was then stained with rat anti-HOXB4 antibody (Developmental Studies Hybridoma Bank) followed by secondary anti-rat Ig antibody conjugated to horseradish peroxidase (HRP) (DakoCytomation). Membranes were developed using ECL Western blotting detection system (Amersham Biosciences) as per the manufacturer’s instructions. After stripping of the membrane, a mouse anti-actin antibody (Chemicon) followed by secondary goat anti-mouse antibody conjugated to HRP (DakoCytomation) was used as a positive control.

Flow Cytometry A single-cell suspension was produced from hESCs cultured on MEFs or from embryoid bodies as follows. Cells were washed once in PBS before incubation at 37°C in 1 mg/ml of Collagenase IV for 20 minutes. The cells were then washed once in PBS and incubated in enzyme-free cell dissociation buffer (Invitrogen) for 20 minutes at 37°C. Finally, cells were gently titurated before filtering through a 40-␮m mesh. For flow cytometry, cells were resuspended at approximately 0.1 ⫻ 105–1.0 ⫻ 105 cells per ml with PBS ⫹ 3% normal goat serum containing 0.1% azide (Serotec) and stained with fluorochrome-conjugated monoclonal antibodies, CD34-fluorescein isothiocyanate (FITC), CD34-phycoerythrin (PE), CD45-PE (all from BD Biosciences), and Glycophorin A-PE (DakoCytomation) or their corresponding isotype controls. For detection of stage-specific embryonic antigen-4, cells were stained with the primary SSEA-4 monoclonal antibody (MC-813–70; Developmental Studies Hybridoma Bank) or IgG3 isotype control (BD Biosciences) and then stained with a FITC-conjugated goat anti-mouse IgG antibody www.StemCells.com

1361

(Sigma-Aldrich). All cells were incubated with 20 ␮l/ml 7-aminoactinomycin D (7-AAD) viability dye (Immunotech, Luminy, France, http://www.immunotech.com) and live cells identified by 7-AAD exclusion were analyzed for surface-marker expression using a BD FACSCalibur flow cytometer with Cell Quest software (Becton, Dickinson and Company, Franklin Lakes, NJ, http://www. bd.com). For all dot plots, gates were set with at least 99.9% of the cells stained with the isotype control in the lower left (wild-type and HOXB4-transfected hESCs) or lower left and right (GFPtransfected hESCs) quadrants.

Clonogenic Assays A single-cell suspension was produced as described above. Human hematopoietic progenitor assays were performed by plating singlecell suspensions of undifferentiated hESCs and differentiated hESCs into Methocult GF⫹ medium (Stem Cell Technologies, Vancouver, BC, Canada, http://www.stemcell.com) consisting of 1% methylcellulose, 30% FBS, 1% BSA, 50 ng/ml stem cell factor, 20 ng/ml granulocyte-macrophage colony-stimulating factor, 20 ng/ml IL3, 20 ng/ml IL6, 20 ng/ml granulocyte colony-stimulating factor, and 3 units/ml erythropoietin. Cells were aliquoted in duplicate samples at between 1 ⫻ 104 and 1 ⫻ 105 cells per plate. After incubation at 37°C in 5% CO2 for 15 days in a humidified atmosphere, differential colony counts were performed based on morphological characteristics. To identify specific cell types, individual colonies were aspirated from the plates, washed once in PBS, and then resuspended in PBS-1% human albumin solution before cytospin and staining with modified Wright’s stain (Bayer, Newbury, U.K., http://www.bayer.co.uk) on a Hematek 1000 slide stainer (Miles Laboratories Inc.).

Statistical Analysis For flow cytometry and CFC data, results are expressed as mean ⫾ SEM. Statistical significance was assessed using the unpaired Student’s t test. Results were considered to be significantly different from control or each other if p ⬍ .05. Quantitative comparison of SCL and GATA1 expression between HOXB4-transfected cells and GFP-transfected cells by quantitative real-time PCR (Q-PCR) were assessed using factorial analysis of variance.

RESULTS HOXB4 Transfection of hESC We expressed the HOXB4 cDNA in hESCs (HOXB4-hESCs) using the pTP6 vector [24], as previously described [21]. As a control for the effect of transfection, we produced cell lines expressing the “human recombinant” green fluorescent protein gene (GFP-hESCs) in place of HOXB4 (Fig. 1A). These constructs link either HOXB4 or GFP with puromycin resistance genes using an internal ribosome entry site (IRES). The CAGG promoter in the pTP6 vector has been demonstrated to be capable of sustaining long-term, robust transgene expression in both mouse and human ESCs [21, 24, 25]. Lipofection using the pTP6 vector was effective in generating stable expression of the HOXB4 transgene. Following transfection, the cell lines were maintained in their undifferentiated state on

1362

Expression of HOXB4 in hESCs

Figure 1. HOXB4 expression in WT-hESCs, control GFP-hESCs, and HOXB4hESCs. (A): Map of the HOXB4-pTP6 and GFP-pTP6 vectors. pTP6 vector contains the CAGG promoter followed by the HOXB4 or GFP cDNA and an IRES, with puror allowing strong selection for transgene expression. (B): Semiquantitatively reverse transcription polymerase chain reaction (PCR) analysis of HOXB4UTR and HOXB4-TR of WT-hESCs, GFP-hESCs, and HOXB4-hESCs clones (HOXB4.1, HOXB4.2, and HOXB4.3). A sample of cDNA from the erythroleukemia cell line K562 was used as a positive control, and a sample containing no cDNA (H2O) was used as a negative control. The housekeeping gene B-2-m was used to normalize the PCR. (C): Real-time PCR analysis of HOXB4 expression in WT-hESCs, GFP-hESCs, and HOXB4-hESCs (clones HOXB4.1, HOXB4.2, and HOXB4.3) on day 0 (undifferentiated). Gene expression relative to K562 was calibrated by the comparative CT method (p ⬎ .18). Data points represent the mean expression of two independent experiments ⫾ SEM. (D): Southern blot analysis of HOXB4-hESCs (clones HOXB4.1, HOXB4.2, and HOXB4.3) demonstrating that each clone has a distinct integration site. The presence of two bands for HOXB4.1 suggests that two copies of the plasmid have integrated into the genome, and the presence of a single band for HOXB4.2 and HOXB4.3 suggests transfection of a single copy of the plasmid in these clones. (E): Immunocytochemistry of WT-hESCs with HOXB4 antibody and secondary fluorescein isothiocyanate (FITC)-conjugated antibody, HOXB4-hESC clone with HOXB4 antibody and secondary FITC-conjugated antibody, and HOXB4-hESCs with Oct4 antibody and secondary Cy3 antibody. Hoescht dye was used to stain the nuclei. Scale bars ⫽ 50 ␮m. (F): Western blot analysis of HOXB4 expression in WT-hESCs, GFP-hESCs, and HOXB4-hESCs (clones HOXB4.1, HOXB4.2, and HOXB4.3) on day 0 (undifferentiated). Protein extracted from the erythroleukemia cell line K562 was used as a positive control. (G): Analysis of the stem cell marker SSEA-4 by flow cytometry in WT-hESCs and HOXB4-hESCs as represented by the shaded histograms. The gated region is based on exclusion of 99.9% of cells of the isotype control (open histograms) and represents the mean ⫾ SEM (n ⫽ 5 for wild-type-hESCs; n ⫽ 9 for HOXB4-hESCs). There was no difference in SSEA-4 expression. (H): Western blot analysis of HOXB4 expression in WT-hESCs, GFP-hESCs, and HOXB4-hESCs (clones HOXB4.1, HOXB4.2, and HOXB4.3) after 20 days of differentiation as embryoid bodies. Protein extracted from the erythroleukemia cell line K562 was used as a positive control. Actin was used to normalize the Western blot reactions. Abbreviations: B-2-m, ␤-2-microglobulin; GFP, green fluorescent protein; GFP-hESC, green fluorescent protein-expressing hESC clone; hESC, human embryonic stem cell; HOXB4-hESC, HOXB4expressing hESC clone; HOXB4-TR, HOXB4 translated region; HOXB4-UTR, HOXB4 untranslated region; IRES, internal ribosome entry site; puror, puromycin resistance gene; WT, wild-type.

Bowles, Vallier, Smith et al. MEFs. Neither untransfected wild-type hESCs (WT-hESCs) nor control GFP-hESCs expressed HOXB4. Reverse transcriptionPCR of WT-hESCs, GFP-hESCs, and HOXB4-hESCs showed that gene expression was confined to the HOXB4-transfected clones (Fig. 1B). As only the translated region of the HOXB4 gene was transfected in the expression vector, we were able to determine that the HOXB4 transcription in the HOXB4-hESCs was a result of transgene expression and not endogenous expression of HOXB4 by using two sets of primer pairs, one within the translated region and another in the untranslated region of the HOXB4 gene (Fig. 1B). Real-time quantitative PCR using human HOXB4-specific primers demonstrated levels of HOXB4 transcript in the HOXB4-transfected hESCs that were generally similar to those of the erythroleukemia cell line K562 (Fig. 1C). Southern blot analysis confirmed that transgene integration sites in the selected clones were distinct from each other (Fig. 1D). Immunocytochemistry and Western blot analysis confirmed that HOXB4 expression was confined to the HOXB4-hESCs (Fig. 1E, 1F). Furthermore, immunocytochemistry confirmed that the HOXB4 protein was localized in the nucleus and that all cells within each of the colonies expressed similar levels of the transgene. Both undifferentiated WT-hESCs (data not shown) and HOXB4-hESCs similarly expressed the hESC markers Oct-4 [2] by immunocytochemistry (Fig. 1E) and SSEA-4 [26] by flow cytometry (Fig. 1G). Expression of HOXB4, Oct-4, and SSEA-4 in HOXB4-hESCs was stable for at least 15 passages (the maximum stage studied here) following transfection. To establish whether HOXB4 expression was maintained upon differentiation of HOXB4-hESC clones, Western blot analysis of EBs at day 20 of culture in serum-containing medium was carried out, confirming that HOXB4 expression was sustained and not downregulated significantly (Fig. 1H). Therefore, we considered HOXB4 transfection and gene expression to be compatible both with the undifferentiated state of hESCs and with their subsequent differentiation.

Control hESCs Exhibit a Low Level of Hematopoietic Differentiation upon Differentiation as EBs The experimental approach used here to examine the initial phenotypic consequences of HOXB4 gain of function was to differentiate hESCs as EBs in serum-containing medium (without additional hemogenic cytokines) during a 20-day period, analyzing the samples at 5-day intervals using flow cytometry. With this approach, differentiation of WT-hESCs and GFPhESCs generated low levels of cells expressing hematopoietic markers. Flow cytometry identified the development of a small population of CD45-positive cells that peaked on day 20 of the experiment in both untransfected (WT-hESCs 0.6% ⫾ 0.1%, both CD34-positive and CD34-negative combined; five independent experiments; Fig. 2A) and GFP-transfected (0.7% ⫾ 0.1%; six independent experiments; Fig. 2B) hESCs. In WThESCs, CD34/CD45 dual-positive cells were less than or equal to 0.1% on all days of the experiment. Glycophorin A, a surface marker specific to erythroid cells, was not greater than 0.1% at any time point in non-HOXB4-transfected hESCs. There was not a significant difference in expression of CD34, CD45, and Glycophorin A between WT-hESCs and GFP-hESCs, both of which generated very modest numbers of hematopoietic cells. www.StemCells.com

1363

HOXB4 Augments the Generation of Hematopoietic Cells from hESCs Three individual HOXB4-hESC clones were examined extensively (HOXB4.1-hESCs, three independent experiments; HOXB4.2-hESCs, six independent experiments; HOXB4.3hESCs, six independent experiments). Differentiation of HOXB4-hESCs over 15 days resulted in a similar number of cells per well compared with WT-hESCs and GFP-hESCs (data not shown). Flow cytometry showed substantial increases in CD45, CD34, and Glycophorin A expression in all three clones, compared with GFP-transfected and untransfected hESC controls (Fig. 2C–2E). On day 20, CD45 was expressed by a total of 4.2% ⫾ 0.3% of HOXB4.1-hESCs (p ⬍ .001), 21.1% ⫾ 3.3% of HOXB4.2-hESCs (p ⬍ .001), and 9.1% ⫾ 1.9% of HOXB4.3-hESCs (p ⫽ .001), compared with 0.7% ⫾ 0.1% of GFP-hESCs. Glycophorin A on day 10 was expressed by 1.0% ⫾ 0.1% of HOXB4.1-hESCs (p ⬍ .001), 8.3% ⫾ 1.2% of HOXB4.2-hESCs (p ⬍ .001), and 5.2% ⫾ 1.2% of HOXB4.3-hESCs (p ⬍ .001), compared with 0.1% ⫾ 0.1% of GFP-hESCs. Interestingly, HOXB4-hESC clones showed peaks of Glycophorin A expression at day 10 or 15, followed by a subsequent decline. To test whether enforced HOXB4 expression is additive or simply a substitute for the effects that can be achieved with cytokines, factors at concentrations previously shown to support hematopoietic development from hESCs were added to the differentiation medium as described [4]. After 15 days of further development, there was a 15-fold increase in CD45⫹ cells from HOXB4-expressing hESCs as compared with control hESCs. Addition of cytokines resulted in a further threefold increase in the yield of CD45⫹ over HOXB4-transfected hESCs cultured in the absence of additional cytokines (p ⬍ .05) (Fig. 3A). After 20 days of differentiation, there was a 12-fold increase in CD45⫹ cells from HOXB4-expressing hESCs compared with control hESCs. Addition of cytokines resulted in a further fivefold increase in the yield of CD45⫹ over HOXB4-transfected hESCs cultured in the absence of additional cytokines (p ⬍ .01) (Fig. 3B). Thus, HOXB4 appeared to augment hematopoietic development additively to the previously reported effects of cytokines [4]. In parallel to flow cytometry, we also analyzed samples by Q-PCR for SCL/Tal-1 (hereafter called SCL) and GATA-1 expression in the absence of added cytokines. SCL plays a crucial role in hematopoietic development, as evidenced by both the SCL knockout mouse, which is an early lethal as a result of failure of embryonic blood development [27, 28], and by studies of murine embryonic stem cells in vitro, which revealed SCL to be necessary for hemagioblast development from mesoderm [29, 30]. GATA-1 is also required for normal hematopoiesis and is a key regulator of development of the erythroid lineage [31]. Time course analysis by Q-PCR of control GFP-hESCs suggested a trend of slightly increased expression of SCL and GATA-1 (Fig. 4A and 4B, solid lines). Compared with GFPhESCs, however, HOXB4-hESCs showed highly significant increases in SCL (p ⫽ .016) and GATA-1 (p ⫽ .0005) expression when they were differentiated as EBs (Fig. 4A and 4B, interrupted lines). These pronounced increases in expression of SCL and GATA1 by HOXB4-hESCs progressed over time (SCL, p ⫽ .0011; GATA1, p ⫽ .0005), first becoming apparent on day 10. The maximum increase of expression for both genes (observed on day 20 of the experiment) was approximately

1364

Expression of HOXB4 in hESCs

Figure 2. HOXB4-transfected hESCs generate greater numbers of cells expressing hematopoietic cell surface markers than do wild-type and control hESCs. Embryoid bodies were formed and differentiated in a serum-containing medium without the addition of further cytokines. Samples were analyzed by flow cytometry 10, 15, and 20 days after the generation of the EBs for cell surface marker expression of CD34, CD45, and Glycophorin A (arrows). (A, B): Untransfected (WT-hESCs) (A) and GFP-transfected (GFP-hESCs) (B) hESCs rarely generate cells expressing hematopoietic cell surface markers, showing less than 1% CD45- or Glycophorin A-expressing cells. (C–E): Three HOXB4transfected hESC clones (HOXB4.1-hESC, HOXB4.2-hESC, and HOXB4.3hESC) show a significant increase in the number of hematopoietic cells generated. Quadrant percentages represent the mean ⫾ SEM for five (WT-hESCs), six (GFP-hESCs), three (HOXB4.1-hESCs), six (HOXB4.2-hESCs), and six (HOXB4.3-hESCs) independent experiments. Abbreviations: hESC, human embryonic stem cell; HOXB4 hESC, HOXB4-transfected hESC clone; GFP, green fluorescent protein.

Bowles, Vallier, Smith et al.

Figure 3. Cytokines augment hematopoietic development from HOXB4hESCs. Proportion of CD45⫹ cells following 15 days (A) and 20 days (B) of in vitro differentiation as EBs from HOXB4-hESCs without (HOXB4/no cytokines) and with (HOXB4/plus cytokines) additional cytokines in the medium relative to control EBs not treated with cytokines (control/no cytokines). On day 15, HOXB4 resulted in a 15-fold increase in CD45⫹ cells compared with control, with an additional threefold further increase on addition of cytokines (ⴱ, t test comparing HOXB4 without additional cytokines and HOXB4 with additional cytokines, p ⬍ .05, n ⫽ 2). On day 20, HOXB4 generated a 12-fold increase in CD45⫹ cells, and additional cytokines resulted in a further fivefold increase in CD45⫹ cells (ⴱ, p ⬍ .01, n ⫽ 2). Error bars depict SEM.

90-fold for SCL and 350-fold for GATA1. Consequently, HOXB4 expression is capable of amplifying a subset of cells that emerges between day 5 and day 10 and results in increased expression of SCL and GATA-1, two representative genes critical to the pathway of hematopoietic development.

HOXB4 Expression Enhances Clonogenic Hematopoietic Progenitor Development from hESCs To assess the consequences of HOXB4 expression for in vitro hematopoietic development, we carried out a clonogenic progenitor assay. This was accomplished by determining the number of hematopoietic colonies formed by dissociating and subculturing EBs at different stages of their development (0 –20 days; Materials and Methods). At the earliest time points (0 and 5 days of culture), there was no significant increase in the total number of colonies (including all hematopoietic cell lineages) formed from the HOXB4-hESC clones, resembling the paucity of such colonies in cultures of wild-type and GFP-hESCs. On day 10 of differentiation, however, there were significant increases in the mean numbers of total colonies from HOXB4transfected hESCs compared with GFP-hESCs as follows (Fig. 5A–5C): HOXB4.1-hESCs, 83 ⫾ 15 colonies per 106 cells plated (p ⬍ .001); HOXB4.2, 233 ⫾ 40 colonies per 106 cells plated (p ⬍ .001); and HOXB4.3, 460 ⫾ 89 colonies per 106 cells plated (p ⬍ .001). On day 15 of differentiation, the mean www.StemCells.com

1365

Figure 4. HOXB4-expressing human embryonic stem cells (hESCs) show upregulation of genes associated with hematopoiesis upon differentiation. Quantitative real-time PCR analysis of SCL (A) and GATA1 (B) expression by GFP-hESCs (solid lines) and HOXB4.2-hESCs (interrupted lines) differentiated in serum containing medium without the addition of further cytokines over 20 days. Gene expression relative to undifferentiated hESCs was calibrated by the comparative CT method. Data points represent the mean expression of two independent experiments ⫾ SEM.

numbers of total colonies from HOXB4-transfected hESCs were as follows: HOXB4.1-hESCs, 167 ⫾ 22 colonies per 106 cells plated (p ⬍ .001); HOXB4.2, 1083 ⫾ 131 colonies per 106 cells plated (p ⬍ .001); and HOXB4.3, 795 ⫾ 84 colonies per 106 cells plated (p ⬍ .001). On day 20 of differentiation, the mean numbers of total colonies from HOXB4-transfected hESCs were as follows: HOXB4.1-hESCs, 206 ⫾ 24 colonies per 106 cells plated (p ⬍ .01); HOXB4.2, 592 ⫾ 77 colonies per 106 cells plated (p ⬍ .001); and HOXB4.3, 498 ⫾ 42 colonies per 106 cells plated (p ⬍ .001). Formation of hematopoietic colonies from WT-hESCs and GFP-hESCs was a rare event at any stage of EB development under our experimental conditions. Colonies were only observed on day 15 and day 20 of control EB development (Fig. 5A–5C). The mean total number of colonies formed from WT-hESCs was 24 ⫾ 10 colonies per 106 cells plated on day 15 and 68 ⫾ 34 colonies per 106 cells plated on day 20. The mean total number of colonies formed from GFPhESCs was 32 ⫾ 10 colonies per 106 cells plated on day 15 and 83 ⫾ 17 colonies per 106 cells plated on day 20. Thus, the total number of all hematopoietic colonies generated by HOXB4expressing hESCs was substantially increased at all stages between day 10 and day 20 of EB development. Interestingly, the composition of the hematopoietic colonies formed from HOXB4-expressing hESCs varied with the stage of EB development, as well as between hESC sublines. In all three HOXB4-expressing clones, erythroid colony (burst forming unit-erythroid [BFU-E]) development peaked on day 10, whereas myeloid (colony forming unit-granulocytic, monocytic [CFU-GM]) and monocytic (colony forming

1366

Expression of HOXB4 in hESCs

Figure 5. HOXB4 expression in human embryonic stem cells induces increased hematopoietic progenitors. (A–C): Total number of hematopoietic colony-forming progenitor cells formed from WT-hESCs, GFP-hESCs, and HOXB4-transfected (HOXB4.1-hESCs, HOXB4.2-hESCs, and HOXB4.3hESCs) human embryonic stem cells differentiated as embryoid bodies in serum-containing medium after 10, 15, and 20 days of embryoid body differentiation. Error bars represent SEM for three to five independent experiments (ⴱ, p ⬍ .001; ⴱⴱ, p ⬍ .01). (D–H): Quantification of colonies subdivided by morphological subtype (BFU-E, CFU-GM, CFU-M, and CFU-GEM) formed from WT-hESCs, GFP-hESCs, HOXB4.1-hESCs, HOXB4.2-hESCs, and HOXB4.3-hESCs on days 0, 5, 10, 15, and 20 of EB differentiation. Shown are phase contrast images of an erythroid colony (BFU-E) (I), a myeloid colony (CFU-GM) (K), and a monocytic colony (CFU-M) (M). Photomicrograph of modified Wright’s-stained cytospin preparations of representative erythroid (J), myeloid (L), and monocytic (N) colonies picked from the Methocult medium. Scale bars ⫽ 100 ␮m. Abbreviations: BFU-E, burst-forming unit-erythroid; CFU-GEM, colony forming unit-granulocytic, erythroid, monocytic; CFU-GM, colony forming unit-granulocytic, monocytic; CFU-M, colony forming unit-monocytic; GFP, green fluorescent protein-transfected; hESC, human embryonic stem cell; HOXB4-hESC, HOXB4-transfected human embryonic stem cell; WT, untransfected wild-type.

unit-monocytic [CFU-M]) colony development peaked on day 15 or 20. Although all three HOXB4-hESCs clones produced significantly more erythroid, myeloid, and monocytic colonies than either WT-hESCs or GFP-hESCs, there was a noticeable difference in overall colony-forming efficiency between the three HOXB4-hESCs clones analyzed (Fig. 5A–5H). There were, however, general similarities in the patterns of the different blood lineages formed, with a

preponderance of myeloid (CFU-GM) and monocytic (CFU-M) colony formation, particularly at the later stages of EB culture (days 15 and 20). An exception was that colony forming unit-granulocytic, erythroid, monocytic (CFU-GEM) colonies, representing the most primitive of hematopoietic myeloid progenitors, were only observed at a frequency of ⱖ10 colonies per 106 cells plated using HOXB4.3-hESCs. No erythroid colonies (BFU-E) and no multipotent colonies

Bowles, Vallier, Smith et al. (CFU-GEM) were detected in any of 11 individual experiments on WT-hESCs and GFP-hESCs. Finally, hematopoietic cells arising from HOXB4-hESCs were capable of developing into fully formed blood elements. Colonies scored as BFU-E (Fig. 5I) contained cells of morphologically maturing erythroid lineage with characteristic nuclei, eosinophilic cytoplasm, and occasional anucleate cells (Fig. 5J). Colonies scored as CFU-GM (Fig. 5K) contained cells comprising all morphological stages of myeloid maturation; blast, promyelocyte, myelocyte, and metamyelocyte through to segmented forms with three or more nuclear segments. Within the CFU-GM, eosinophils and their precursors, as well as the occasional monocytoid cell, were seen (Fig. 5L). Colonies scored as CFU-M (Fig. 5M) contained exclusively lipid-laden monocytoid cells (Fig. 5N). Colonies scored as CFU-GEM contained cells from all lineages. Thus, HOXB4 expression not only increased the total number of hematopoietic colonies but was also consistent with their development in vitro into terminally differentiated erythroid or myeloid cells.

DISCUSSION This report demonstrates the use of transcription factor overexpression to promote the development of a clinically useful tissue from hESCs. Stable HOXB4 expression in hESCs was compatible with the undifferentiated state (as evidenced by Oct4 and SSEA-4 expression), thus permitting prolonged passaging of individual hESC clones. Moreover, differentiation of these hESC clones in a serumcontaining medium without additional cytokines resulted in a significant increase in hematopoietic development. Addition of cytokines further increased the yield of hematopoietic cells by HOXB4expressing hESCs, suggesting that HOXB4 may act independently to enhance such development. Our findings thus extend to hESCs the utility of homeodomain transcription factors as a means of augmenting in vitro hematopoietic development, as previously demonstrated for mouse ESCs [32]. HOXB4 overexpression in hESCs resulted in sequential increases in erythroid cells and leucocytes, as evidenced by increased expression of Glycophorin A [33, 34] and CD45 [35]. In vitro development of differentiated HOXB4-expressing hESCs in a methylcellulose colony-forming assay resulted in ⬎10-fold overall increase in the combined total of erythroid, myeloid, and monocytic colonies. Furthermore, expression of SCL and GATA-1, two transcription factors critical for hematopoiesis, was significantly greater in HOXB4-transfected cells compared with control GFPtransfected hESCs differentiated as EBs. The overall extent of increased hematopoietic development may in part reflect the low baseline levels in our experimental conditions, in which wild-type and control GFP-transfected hESCs infrequently undergo differentiation to HSCs. However, other studies in which cytokines were omitted report similarly low baseline levels of CD45⫹ cells at comparable periods (0.1%–2% at days 8 –15 of differentiation) [3, 4, 6 –9]. Thus, HOXB4 expression clearly promotes the hematopoietic development in the descendants of hESCs undergoing differentiation as EBs even in the absence of animal stromal cells and further cytokine additives. In a recent study, HSCs derived from hESCs have been shown to be capable of repopulating hematopoiesis in primary and secondary recipient sheep [10]. The transplanted cells had been previously cocultured with the stromal layer S17 in the www.StemCells.com

1367

absence of additional cytokines, but the level of chimerism was low [10] (typically 0.1% CD34⫹ and 0.1% CD45⫹ cells in primary recipients). Chadwick et al. have shown that the addition of cytokines to the differentiation medium increased the yield of cells capable of forming hematopoietic colonies in vitro [4]. Furthermore, Wang et al. identified a cell derived from hESCs with both in vitro blood and endothelial potential [36], and they have recently demonstrated that intra-bone marrow injection of hESC-derived cells into SCID mice can result in a low level of human hematopoietic engraftment 8 weeks posttransplantation (⬍1% human chimerism) [20]. In those experiments, a subpopulation of cytokine-expanded cells that were exposed to lentivirus-encoding HOXB4 and selected for transgene expression approximately 1 week later had no increased repopulating capacity. This is in contrast with ample evidence for HOXB4 amplified engraftment into mice using HSCs derived from mouse yolk sac [16], bone marrow [15], and human cord blood cells [37]. Therefore, our goal was to achieve constitutive HOXB4 expression by enforcing transgene expression using an IRES vector driven by a promoter previously demonstrated to sustain robust transgene expression in hESCs [21]. Such HOXB4-expressing cells may ultimately be capable of significantly increasing reconstitution of multilineage hematopoiesis in animal models. Similar experiments overexpressing HOXB4 in murine ES cells resulted in increases only in erythroid and mixed erythroid/ myeloid colonies and not in granulocytic or monocytic colonies [32]. This is in contrast to the effect of HOXB4 in hESCs, in which myeloid and monocytic colonies were increased significantly by HOXB4 expression. This difference in the observed effect of HOXB4 on murine and human ES cells may occur as a result of differences in transgene expression. The effect of HOXB4 on human bone marrow HSCs has been shown to vary depending on the level of expression, with higher transient levels of expression associated with increased myeloid differentiation and reduced proliferation [37–39]. Furthermore, the effects of HOXB4 expression have been shown to differ with various cell types [40 – 42]. Therefore, the timing of expression, level of expression, and cellular context may all affect the outcome of HOXB4 expression and may thus account for different hematopoietic phenotypes observed when overexpressing HOXB4 using different cells or methods. A similar explanation might account for the marked differences in overall colony-forming efficiency between HOXB4-expressing hESC lines. The HOXB4 hESC cell lines were initially selected for their ability to maintain transgene expression in the undifferentiated state. These HOXB4-expressing hESCs clones were shown by real-time PCR, Western blotting, and immunocytochemistry to have transgene expression levels generally similar to each other and to the erythroleukemia cell line K562. Nevertheless, the HOXB4hESC clones differed significantly from each other in hematopoietic output, suggesting that other variables could also affect ultimate phenotypic consequences during the course of in vitro differentiation to hemogenic progenitors. Even with constitutive expression of HOXB4 mRNA and protein for multiple generations in the undifferentiated state and for up to 20 days of EB development, only a fraction of such transgenic hESCs undergo hematopoietic differentiation. This raises the question of what limits the incidence of hematopoietic development in such cultures. It seems unlikely that HOXB4 acts directly to promote the differentiation of hESCs into HSCs.

Expression of HOXB4 in hESCs

1368

A more probable mechanism of HOXB4 action in hESCs is that it affects hematopoietic cells that have spontaneously differentiated from hESCs to mesoderm and thence to HSCs. Such a mechanism would be consistent with HOXB4 effects in mouse HSCs derived from ESCs and other sources, where it is thought to act by amplifying the number of HSCs and their descendants through an increase in HSC self-renewal [14, 15, 43, 44]. If HOXB4 functions to promote the expansion of hematopoietic progenitors, then two conditions must be met for it to act in descendants of hESCs: first, hESCs will need to differentiate into cells that are receptive to the effect of the HOXB4 protein; and second, HOXB4 expression will need to be maintained in the differentiated cell type in which it will act. The paucity of blood lineages formed in wild-type and control EBs suggests that spontaneous differentiation from hESCs to mesoderm and onwards to HSCs is a rare event. This may reflect an underlying tendency for hESCs to differentiate instead along the neuro-ectodermal pathway [22, 45]. In addition, a means of inducing mesoderm formation with high efficiency has not yet been defined for hESCs [22, 45] and is subject to sources of variation in culture conditions [46]. In any case, hESC-based generation of hematopoietic cells for clinical application would require a substantial improvement in the efficiency of the differentiation steps leading to HSC formation.

CONCLUSION By demonstrating that HOXB4 can promote hematopoietic development from hESCs, this work provides a model for amplification of differentiated cell types from hESCs with potential therapeutic application. It thereby addresses a problem that must be solved in

REFERENCES

any attempt to provide clinically useful tissues, namely self-renewal of progenitor and intermediate (transit amplifying) stem cell populations to generate sufficient numbers of terminally differentiated cells. The remaining challenge of establishing substantial long-term, multilineage human hematopoiesis in an animal HSC transplant model should be facilitated by such an approach.

ACKNOWLEDGMENTS We are grateful for the help received from members of the Departments of Surgery, Pathology and Haematology at the University of Cambridge, in particular Professors Andrew Bradley, Tony Green, and Charles ffrench-Constant for support during this work and Dr. Isabelle Bouhon for encouragement during its inception. We thank Dr. T. Pratt (University of Edinburgh, Edinburgh, U.K.) for the pTP6 vector, Dr. R.K. Humphries (British Columbia Cancer Agency, Vancouver, BC, Canada) for the HOXB4 cDNA, and Dr. K. Vintersten (Samuel Lunenfeld Research Institute, Toronto, ON, Canada) for the puromycin resistant feeders. We are also grateful to Dr Mickie Bhatia (University of Western Ontario, London, ON, Canada) and his team for advice with respect to analysis of hESCderived hematopoietic cells. This study was supported by a Wellcome Trust Clinical Research Training Fellowship (K.M.B.), a Raymond and Beverly Sackler Studentship (K.M.B.), a Medical Research Council International Appointments Initiative grant (L.V., M.R.J.A., R.A.P.), a Human Frontiers Science Program grant (RG0110/2000-M; R.A.P., L.V.), and a Cambridge-MIT Institute grant (J.R.S.).

DISCLOSURES R.A.P. has a financial interest in Stemnion, LLC.

10 Narayan AD, Chase JL, Lewis RL et al. Human embryonic stem cellderived hematopoietic cells are capable of engrafting primary as well as secondary fetal sheep recipients. Blood 2006;107:2180 –2183.

1

Thomson JA, Itskovitz-Eldor J, Shapiro SS et al. Embryonic stem cell lines derived from human blastocysts. Science 1998;282:1145–1147.

2

Reubinoff BE, Pera MF, Fong CY et al. Embryonic stem cell lines from human blastocysts: Somatic differentiation in vitro. Nat Biotechnol 2000; 18:399 – 404.

11 Bhatia M, Bonnet D, Kapp U et al. Quantitative analysis reveals expansion of human hematopoietic repopulating cells after short-term ex vivo culture. J Exp Med 1997;186:619 – 624.

3

Kaufman DS, Hanson ET, Lewis RL et al. Hematopoietic colonyforming cells derived from human embryonic stem cells. Proc Natl Acad Sci U S A 2001;98:10716 –10721.

12 Glimm H, Eaves CJ. Direct evidence for multiple self-renewal divisions of human in vivo repopulating hematopoietic cells in short-term culture. Blood 1999;94:2161–2168.

4

Chadwick K, Wang L, Li L et al. Cytokines and BMP-4 promote hematopoietic differentiation of human embryonic stem cells. Blood 2003;102:906 –915.

5

Cerdan C, Rouleau A, Bhatia M. VEGF-A165 augments erythropoietic development from human embryonic stem cells. Blood 2004;103:2504 – 2512.

13 Sauvageau G, Lansdorp PM, Eaves CJ et al. Differential expression of homeobox genes in functionally distinct CD34⫹ subpopulations of human bone marrow cells. Proc Natl Acad Sci U S A 1994;91:12223– 12227.

6

Zhan X, Dravid G, Ye Z et al. Functional antigen-presenting leucocytes derived from human embryonic stem cells in vitro. Lancet 2004;364: 163–171.

7

Tian X, Morris JK, Linehan JL et al. Cytokine requirements differ for stroma and embryoid body-mediated hematopoiesis from human embryonic stem cells. Exp Hematol 2004;32:1000 –1009.

8

9

Vodyanik MA, Bork JA, Thomson JA et al. Human embryonic stem cell-derived CD34⫹ cells: Efficient production in the coculture with OP9 stromal cells and analysis of lymphohematopoietic potential. Blood 2005;105:617– 626. Zambidis ET, Peault B, Park TS et al. Hematopoietic differentiation of human embryonic stem cells progresses through sequential hematoendothelial, primitive, and definitive stages resembling human yolk sac development. Blood 2005;106:867– 870.

14 Sauvageau G, Thorsteinsdottir U, Eaves CJ et al. Overexpression of HOXB4 in hematopoietic cells causes the selective expansion of more primitive populations in vitro and in vivo. Genes Dev 1995;9:1753–1765. 15 Antonchuk J, Sauvageau G, Humphries RK. HOXB4-induced expansion of adult hematopoietic stem cells ex vivo. Cell 2002;109:39 – 45. 16 Kyba M, Perlingeiro RC, Daley GQ. HoxB4 confers definitive lymphoid-myeloid engraftment potential on embryonic stem cell and yolk sac hematopoietic progenitors. Cell 2002;109:29 –37. 17 Amsellem S, Pflumio F, Bardinet D et al. Ex vivo expansion of human hematopoietic stem cells by direct delivery of the HOXB4 homeoprotein. Nat Med 2003;9:1423–1427. 18 Krosl J, Austin P, Beslu N et al. In vitro expansion of hematopoietic stem cells by recombinant TAT-HOXB4 protein. Nat Med 2003;9:1428 – 1432.

Bowles, Vallier, Smith et al.

1369

19 Brun AC, Bjornsson JM, Magnusson M et al. Hoxb4-deficient mice undergo normal hematopoietic development but exhibit a mild proliferation defect in hematopoietic stem cells. Blood 2004;103:4126 – 4133.

33 Nakahata T, Okumura N. Cell surface antigen expression in human erythroid progenitors: Erythroid and megakaryocytic markers. Leuk Lymphoma 1994;13:401– 409.

20 Wang L, Menendez P, Shojaei F et al. Generation of hematopoietic repopulating cells from human embryonic stem cells independent of ectopic HOXB4 expression. J Exp Med 2005;201:1603–1614.

34 Rogers CE, Bradley MS, Palsson BO et al. Flow cytometric analysis of human bone marrow perfusion cultures: Erythroid development and relationship with burst-forming units-erythroid. Exp Hematol 1996;24: 597– 604.

21 Vallier L, Rugg-Gunn PJ, Bouhon IA et al. Enhancing and diminishing gene function in human embryonic stem cells. STEM CELLS 2004;22: 2–11. 22 Vallier L, Reynolds D, Pedersen RA. Nodal inhibits differentiation of human embryonic stem cells along the neuroectodermal default pathway. Dev Biol 2004;275:403– 421. 23 Draper JS, Smith K, Gokhale P et al. Recurrent gain of chromosomes 17q and 12 in cultured human embryonic stem cells. Nat Biotechnol 2004; 22:53–54. 24 Pratt T, Sharp L, Nichols J et al. Embryonic stem cells and transgenic mice ubiquitously expressing a tau-tagged green fluorescent protein. Dev Biol 2000;228:19 –28. 25 Niwa H, Yamamura K, Miyazaki J. Efficient selection for high-expression transfectants with a novel eukaryotic vector. Gene 1991;108:193– 199. 26 Draper JS, Pigott C, Thomson JA et al. Surface antigens of human embryonic stem cells: Changes upon differentiation in culture. J Anat 2002;200:249 –258. 27 Porcher C, Swat W, Rockwell K et al. The T cell leukemia oncoprotein SCL/tal-1 is essential for development of all hematopoietic lineages. Cell 1996;86:47–57. 28 Robb L, Elwood NJ, Elefanty AG et al. The scl gene product is required for the generation of all hematopoietic lineages in the adult mouse. EMBO J 1996;15:4123– 4129. 29 Gering M, Rodaway AR, Gottgens B et al. The SCL gene specifies haemangioblast development from early mesoderm. EMBO J 1998;17: 4029 – 4045. 30 Robertson SM, Kennedy M, Shannon JM et al. A transitional stage in the commitment of mesoderm to hematopoiesis requiring the transcription factor SCL/tal-1. Development 2000;127:2447–2459. 31 Crispino JD. GATA1 in normal and malignant hematopoiesis. Semin Cell Dev Biol 2005;16:137–147. 32 Helgason CD, Sauvageau G, Lawrence HJ et al. Overexpression of HOXB4 enhances the hematopoietic potential of embryonic stem cells differentiated in vitro. Blood 1996;87:2740 –2749.

www.StemCells.com

35 Woodford-Thomas T, Thomas ML. The leukocyte common antigen, CD45 and other protein tyrosine phosphatases in hematopoietic cells. Semin Cell Biol 1993;4:409 – 418. 36 Wang L, Li L, Shojaei F et al. Endothelial and hematopoietic cell fate of human embryonic stem cells originates from primitive endothelium with hemangioblastic properties. Immunity 2004;21:31– 41. 37 Buske C, Feuring-Buske M, Abramovich C et al. Deregulated expression of HOXB4 enhances the primitive growth activity of human hematopoietic cells. Blood 2002;100:862– 868. 38 Brun AC, Fan X, Bjornsson JM et al. Enforced adenoviral vectormediated expression of HOXB4 in human umbilical cord blood CD34⫹ cells promotes myeloid differentiation but not proliferation. Mol Ther 2003;8:618 – 628. 39 Schiedlmeier B, Klump H, Will E et al. High-level ectopic HOXB4 expression confers a profound in vivo competitive growth advantage on human cord blood CD34⫹ cells, but impairs lymphomyeloid differentiation. Blood 2003;101:1759 –1768. 40 Krosl J, Baban S, Krosl G et al. Cellular proliferation and transformation induced by HOXB4 and HOXB3 proteins involves cooperation with PBX1. Oncogene 1998;16:3403–3412. 41 Krosl J, Sauvageau G. AP-1 complex is effector of Hox-induced cellular proliferation and transformation. Oncogene 2000;19:5134 –5141. 42 Schmittwolf C, Porsch M, Greiner A et al. HOXB4 confers a constant rate of in vitro proliferation to transduced bone marrow cells. Oncogene 2005;20:561–572. 43 Thorsteinsdottir U, Sauvageau G, Humphries RK. Enhanced in vivo regenerative potential of HOXB4-transduced hematopoietic stem cells with regulation of their pool size. Blood 1999;94:2605–2612. 44 Antonchuk J, Sauvageau G, Humphries RK. HOXB4 overexpression mediates very rapid stem cell regeneration and competitive hematopoietic repopulation. Exp Hematol 2001;29:1125–1134. 45 Munoz-Sanjuan I, Brivanlou AH. Neural induction, the default model and embryonic stem cells. Nat Rev Neurosci 2002;3:271–280. 46 Wang L, Menendez P, Cerdan C et al. Hematopoietic development from human embryonic stem cell lines. Exp Hematol 2005;33:987–996.