Supporting Information

Files in this Data Supplement:

SI Text

For generation of a tamoxifen-inducible HOXB4 protein C-terminally fused with the estrogen receptor (ER) ligand-binding domain, PCR-based modifications of the HoxB4 cDNA were performed by using FMEV-GFP2AHAHHOXB4wPRE as a template (1). In a first step, the GFP2AHAHHOXB4 expression cassette was amplified by using the oligodeoxynucleotides A (5'-GACCGTGTGCGGCCGCATGGTGAGCAAGGGCGAG-GAGCTGTTCA-3') and B (5'-TCTCATGTCTCCAGCAGATGGATCGAGCGC-GCGGGGGCCTCCATT-3') as primers. We thereby deleted the stop codon at the 3' end of the HOXB4 cDNA and fused it with the first 24 bp of the ER-ligand binding domain. The ER-ligand binding domain was amplified by using pERT2 (2) as a template and the oligodeoxynucleotides C (5'AATGGAGGCCCCCGCGCGCTCGATCCATCTG-CTGG-AGACATGAGA-3') and D (5'-ACGTTGTAAAACGACGGCCAGTGAA-3') as primers. The last 21 nt of 3'HOXB4 cDNA were added directly 5' in front of the start codon of the ER-ligand binding domain. After purification, the two amplification products were used for an "overlap PCR" serving as templates to amplify the complete GFP2AHAHOXB4ER fusion cDNA using the oligodeoxynucleotides A and D as primers. From the plasmid FMEV-GFP2AHAHHOXB4wPRE the GFP2AHAHHOXB4 cassette was finally excised with NotI and EcoRI and exchanged with the GFP2AHAHOXB4ER amplicon. Retroviral vectors, SF91-EGFP2AHAHOXB4-wPRE, expressing constitutively active HOXB4 (HOX^const) and SF11-tCD34, expressing a truncated form of human CD34 (tCD34) have been described (3, 4). Preparation of ecotropic-pseudotyped retroviral particles have been described (1).

For sorting of LSK subpopulations, freshly prepared BM cells or in vitro-cultivated progenitor cells were stained with PerCPCy5.5-labeled anti-ScaI, APC-labeled anti-c-kit, and a mixture of phycoerythrin (PE)-labeled mAbs directed against lineage markers (CD11b, Gr1, Ter119, CD3, CD4, CD8, B220), including anti-IL-7R. Alternatively, the cells were incubated with PE-conjugated anti-ScaI, allophycocyanin (APC)-conjugated anti-c-kit, and a biotin-conjugated mixture of lineage mAbs followed by staining with steptavidin-PerCPCy5.5. Cell aggregates were excluded with forward-scatter area versus pulse width. DAPI was used for exclusion of dead cells. Gates were set according to control samples stained according to the method called Fluorescence Minus One (FMO). FACSAria, LSRII, and FACS-Calibur instruments (Becton Dickinson) were used for cell sorting and analysis. Flow cytometry data were analyzed with FlowJo software (Tree Star). All antibodies were purchased either from BD Pharmigen or eBiocience (San Diego, CA).

Lin^-ScaI⁺ and sorted LSK cells were plated at density of 1 ´ 10⁵ cells/ml in serum-free medium (StemSpan, Stem Cell Technologies) containing rmSCF (50 ng/ml), rhFL (50 ng/ml), rhIL-11 (50 ng/ml), rmIL-3 (20 ng/ml), and 1% penicillin/streptomycin. Recombinant cytokines were obtained from Peprotech (Rocky Hill, NJ). After prestimulation for 36-40 h, cells were transduced by using plates preleoaded with RetroNectin (TaKaRa, Otsu, Japan) and retroviral vector particles at an multiplicity of infection (MOI) of 10. After two rounds of transduction on 2 consecutive days, cells were harvested, washed in PBS (PBS, Gibco) and resupended in the cytokine-supplemented serum-free medium described above. Proportions of transduced cells (GFP⁺ or tCD34⁺) were determined by flow cytometry.

For microarray studies, HOXB4ER- and HOXB4^const- transduced Lin^-ScaI⁺ cells were expanded for 14 days in cytokine-supplemented (mSCF, huFL, huIL-11, mIL-3), serum-free medium (StemSpan) in the presence of 10 nM TMX (Sigma). Subesquently, HOXB4ER⁺ and HOXB4^constLSK cells, respectively, were purified by FACS and further cultivated for 12 h in cytokine-supplemented serum-free medium in the presence of 10 nM TMX. Thereafter, both cell populations were split into two fractions and further incubated either with or without TMX. After 1 and 4 h of TMX withdrawal, the cells were lysed in Trizol (Invitrogen) and RNA was purified by using the RNeasy Kit (Qiagen, Hilden, Germany).

To analyze the repopulation ability of ex vivo expanded HOXB4-expressing cells, the progenies of HOXB4ER-transduced and tCD34 control vector transduced LSK cells were mixed in a 1:1 ratio after transduction, reflecting the expansion equivalents (EEs) of equal numbers of freshly isolated LSK cells. The cell mixes were then cultivated in serum-free medium ± 10 nM TMX without and with addition of either TNF-a (R&D, Wiesbaden, Germany) or 10% FCS (Stem Cell Technologies, Vancouver, Canada), respectively. TMX was added to the HOXB4ER/tCD34 cell samples 6 h before addition of the aforementioned substances to be tested. After 7 days of cocultivation, cohorts of mice were transplanted with suitable aliquots of the differently cultivated HOXB4ER/tCD34 cell mixes. Each mouse received in total the EEs of 4,000 freshly isolated LSK cells, half transduced with the HOXB4ER vector, the other transduced with the tCD34 vector. Donor cells were cotransplanted with freshly isolated 200,000 unfractioned CD45.2 competitor BM cells for radioprotection. Mice were bled from the tail vein and peripheral blood (PB) was analyzed 11 and 28 weeks after transplantation for multilineage reconstitution with HOXB4ER⁺(GFP⁺) and tCD34⁺ donor cells. PB samples were lysed with ammonium chloride before staining with antibodies against human CD34 and antibodies against anti-mouse CD3, CD11b, Gr-1, CD19, and Ter119.

Mice were purchased from Charles River and kept in microisolators in the pathogen-free animal facility of the Hannover Medical School. C57BL/6J (CD45.2) mice were used as transplant donors and recipients. Before transplantation (<24 h), female CD45.2 recipient mice of ³12 weeks of age were conditioned by lethal irradiation with a single exposure to a total body dose of 10 Gy. Irradiated mice received ciprofloxacin at 100 mg/ml in drinking water for the first 2 weeks. Irradiated recipients were transplanted i.v. (0.3 ml/mouse) by tail-vein injection. All animal experiments were approved by the local ethical committee and performed according to their guidelines.

The Hoxb4i ES cell line was cultured in DMEM supplemented with 0.1 mM nonessential amino cids, 2 mM glutamine, 15% FCS, 0.1 mM b-mercaptoethanol, and 10³ units/ml of leukemia inhibitory factor (LIF). EBs were made by trypsinizing ES cells and incubating them in IMDM supplemented with 15% FCS, 200 mg/ml transferrin, 4.5 mM monothioglycerol, 50 mg/ml ascorbic acid, and 2 mM glutamine at a concentration of 3,500 cells/ml in bacteriological plates. HOXB4i ES cells were treated from days 4-6 of EB formation with and without doxycycline (2 mg/ml) in the absence or presence of bFGF (50 ng/ml) or the Fgfr inhibitor SU5402 (Calbiochem) (30 mM) and/or the MEK1/2 inhibitor U0126 (Calbiochem) (30 mM). In experiments including these drugs, control cultures contained the corresponding amount of DMSO (the solvent for SU5402). In all cases EBs were collected on day 6 after their initial plating and were processed for methylcellulose-CFU assays or RNA extraction.

The CCE ES-cell line retrovirally expressing HOXB4 (5) was grown on gelatinized plates without feeder cells in DMEM (Gibco/BRL, Carlsbad, CA), supplemented with 2 mM L-glutamine (Gibco/BRL), 100 units/ml penicillin and 100 mg/ml streptomycin (Gibco/BRL), 15% (vol/vol) FCS (FCS, Euroclone, Milano, Italy), 1% (vol/vol) LIF-supernatant, and 1.5 ´ 10^-4 M monothioglycerol (MTG; Sigma, St. Louis, MO).

Three days before differentiation in vitro, ES cells were transferred to IMDM (Gibco/BRL) supplemented with the same components as DMEM. For EB differentiation, ES cells were plated on a nontreated Petri dish at a concentration of 1,000 cells/ml in EB differentiation medium containing IMDM (Gibco/BRL), 2 mM L-glutamine, 100 units/ml penicillin and 100 mg/ml streptomycin, 15% (vol/vol) FCS, 50% (wt/vol) methylcellulose (Sigma, St. Louis, MO), 50 mg/ml ascorbic acid (Sigma), 300 mg/ml iron-saturated human transferrin (Sigma), 5% (vol/vol) protein-free hybridoma medium (PFHM-II; Gibco/BRL), and 4 ´ 10^-4 M MTG. EB differentiation medium was replaced at day 5 of EB differentiation. For dissociation, day-6 EBs were incubated at 37°C in 1x trypsin for 3-5 min, FCS was added, and EBs were dissociated by repeated pipetting.

Dissociated EB cells were seeded into SCM, a serum-free medium (StemPro34 plus nutrient supplement; Life Technologies, Gibco/BRL) supplemented with 100 ng/ml murine SCF (R&D Systems, Minneapolis, MN), 10^-6 M dexamethasone (Sigma), 40 ng/ml IGF-1 (Promega, Madison, WI), 2 ng/ml mIL-3, 5 ng/ml IL-6, and 10 ng/ml mflt-3L. At day 1 of cultivation, cells growing in suspension were removed from adherent cells and plated onto a new Petri dish. Cell number and cell size was determined by using an electronic cell counter (CASY-TTC; Schärfe-System, Reutlingen, Germany). Cell density was maintained between 2 and 4 ´ 10⁶ cells/ml.

For FGF-signaling studies, HOXB4-ER expressing ES-HCs (day 21 after EB dissociation) were grown for 2 weeks in SCM containing either recombinant human FGF2 (rh-FGF2, basic FGF) at a concentration of 50 ng/ml + Heparin (Liquemin N5000; Hoffmann LaRoche, Grenzach-Wyhlen; f.c. 1:1,000) or the FGF-signaling inhibitor SU5402 (Calbiochem) at final concentration of 10 mM. Cells were counted every day, pelleted, and resuspended in fresh SCM +/-FGF2 or SU5402 at a concentration of 2 ´ 10⁶/ml.

Hematopoietic colony assays were performed in M3434 complete MethoCult Media (StemCell Technologies) according to the manufacturer's protocol.

ES-HCs were transduced 7 days after dissociation of EBs (ES-HC, d7) with the ecotropic HOXB4ER retroviral expression vector by two rounds of spin inoculation on two consecutive days, as described (1). A multiplicity of infection (MOI) of 10 was used. GFP-expressing cells were flowcytometrically sorted 48 h later and grown in SCM containing 100 nM TMX (Sigma).

For HOXB4 target gene expression analysis, HOXB4-ER ES-HCs were washed twice and cultured over night without TMX. The next day, the cells were counted, their viability was determined (CASY-TTC; Schärfe System, Reutlingen, Germany), split into equal fractions, and incubated with (or without) cycloheximid (f.c. 5 mg/ml) to block protein synthesis for 30 min. Subsequently, the cells were incubated ± 100 nM TMX to activate the HOXB4-ER fusion. Samples containing 10⁶ cells each were withdrawn at 0, 2, and 4 h after induction of HOXB4 and lysed in Trizol (Invitrogen) and total RNA was purified by using the RNeasy Kit (Qiagen, Hilden, Germany). Fifty nanograms of each RNA was used for cDNA synthesis and amplication with the QuantiTect Whole Transcriptome Kit (Qiagen, Hilden, Germany), according to the manufacturer´s protocol. Amplified cDNA was column purified via the QIAmp DNA Mini kit (Qiagen).

For analysis of HOXB4 target gene expression in ES-HCs, real-time PCRs were performed in triplicate with 50 ng amplified cDNA per reaction using the QuantiFast SYBR Green PCR System in combination with QuantiTect Primers (Qiagen). Amplification was carried out in 96-well PCR plates (Applied Biosystems), in a 7300 Real-Time-PCR machine (Applied Biosystems). Expression of mouse actin was used as the endogenous control. Relative gene expression was calculated by the comparative 2^-DDCT method.

FACS-sorted LSK cells as indicated were initially lysed in Trizol (Invitrogen). After recovery of the aqueous phase, RNA was purified by using the RNEasy Mini Elute Kit (Qiagen). Ten to 50 ng of each RNA was used for cDNA synthesis and amplified once by using the WT-Ovation RNA Amplification System (Nugen, San Carlos, CA) or the QuantiTect Whole Transcriptome Kit (Qiagen), according to the manufacturer´s protocol.