Microarray images are analysed by using Feature extraction software version 10.7.3.1 from Agilent technologies and Agilent normalization protocol GE2_107_Sep09 with 028004_D_F_20100430 as design. Default settings are used. All processing methods used for gene expression analysis were performed on the median signal from Agilent Feature Extraction raw data files using functions and packages collected in the R Bioconductor project (Gentleman et al,Ê 2004, Genome Biology, 5:R80) as well as custom written routines. Flagged spots as well as control spot were removed. Normalization was then performed using the normalizeWithinArrays and the normalizeInterArrays functions from R package LIMMA following by an InterArrays (Smyth, 2004, Statistical applications in Genetics and molecular biology, vol3, N¡1, article3). The Array Quality Metrics package from Bioinconductor (AQM, Kauffmann A., Huber W. was used to validate the quality of the arrays. All intensity profiles (2 profiles for each array) were then imported into BrB Array Tools software (R. Simon et al, http://linus.nci.nih.gov/brb) for statistical analysis.Ê The data are filtered at an intensity level of 10 with Genes whose expression has a minimum value of 10 and differed by at least 1.5 fold from the median in at least 20% of the arrays were retained. We identified genes that were differentially expressed among the two or moreÊ classes using the Class Comparison method, a random-variance t-test (of F-test). The random-variance t-test is an improvement over the standard separate t-test as it permits sharing information among genes about within-class variation without assuming that all genes have the same variance (3). Genes were considered statistically significant if their p value was less than 0.001.
Scanning perform with a Agilent G2505C DNA Microarray scanner using default parameters: 20 bits mode, 3 um resolution, at 20¡C in low ozone concentration environment.
Agilent Hybridization Protocol (Gene expression Hyb kit Large (ref 5188-5280), Chamber type: Agilent SureHyb Chamber; Quantity of labelled extract: 825ng; duration: 17 hours; volume: 110 ul ; Temperature in ¡C: 65). Add the following components in clean 1.5 ml tubes: 400 ng linearly amplified cRNA labeled with Cy5 for each sample, and 400 ng linearly amplified cRNA labeled with Cy3 for the control. Conversely, for dye swapped arrays, mix 400 ng linearly amplified cRNA labelled with Cy3 for each sample, and 400 ng linearly amplified cRNA labeled with Cy5 for the control. In each case, add 5 ul 2X Blocking Agent, 1 ul of Fragmentation buffer 25X and nuclease free water up to 19 ul. Mix by vortexing. Incubate at 60¡C for 30 minutes in dark. Spin briefly, add 25 ul of 2X hybridization buffer. Mix gently. Assembly the Sure-Hyb hybridization chamber from Agilent. Place a backing side with the plastic inner on upper side. Gently add 50 ul of 1x hybridization solution and cover with the Agilent 60k array properly oriented (active surface in contact with liquid). Finish to assembly the chamber and tight the screw. Hybridization carry out for 17 hours at 65¡C in a rotating oven (Robbins Scientific, Mountain View, CA) at 10 rpm. The arrays are disassembled at room temperature in SSPE Wash 1 Buffer (ref 5188-5327), then wash 1 minute in a glass dish (Wheathon) at room temperature in Wash 1 Buffer, then 1 minute in SSPE Wash 2 Buffer (ref 5188-5327) at 37¡C and finally acetonitryl wash (sigma aldrich), then 1 minute at room temperature.
(100 ng, Amplification=RNA polymerases). Agilent oligo Cy5 or Cy3 probes labeling protocol. Kit used for probe labeling: Agilent Low Input QuickAmp labelling kit (ref 5190-2306) adapted for small amount of total RNA (100 ng total RNA per reaction). In 1.5 ml tubes add 100 ng total RNA from sample. Add 0.8 ul T7 promoter primer, 1 ul Spike In dilution 1/3200 (ref 5188-5279), use Spike In A to label Cy3 target and Spike In B to label Cy5 target, then add nuclease free water (Invitrogen ref:10977-015) to bring the volume up to 7.1 ul. Denature by incubating at 65¡C for 10 minutes. Place the reactions on ice and incubate 5 min. Spin briefly. Prepare a Reverse Transcription master mix, adding for one reaction 2 ul First strand Buffer 5X, 1 ul DTT 0.1M, 0.5 ul dNTP 10 mM mix, 1.2 ul AffinityScript Rnase Block mix,. Master mix is prepared in batch for all the samples included in the study (vol per one reaction multiplied by number of samples. In each reaction tube containing denaturated RNA and T7 promoter primer in a volume of 7.1 ul, add 4.7 ul of Reverse Transcriptase master mix, and mix by gently pipetting. Incubate at 40¡C in a circulating water bath for 2 hours. Move samples to 70 ¡C for 15 minutes to inactivate affinityScript Rnase block mix, and incubate on ice for 5 minutes. Spin briefly. Prepare a in vitro transcription master mix, adding for one reaction : 0.75 ul Nuclease free water, 3.2 ul Transcription buffer 5X, 0.6 ul DTT 0.1 M, 1 ul NTP mix, 0.21 ul T7 RNA polymerase Blend. Transcription master mix is prepared in batch for all the samples included in the study: (vol per one reaction multiplied by number of samples). Master mix is splited in two aliquots, one for Cy5 and one for Cy3. Add 0.24 ul (multiplied by reactions number) CTP-Cy5 25 mM or 0.24 ul CTP-Cy3 (multiplied by reactions number) to a total volume of 6 ul/ reaction. To each RT reaction, add 6 ul Transcription master mix. Incubate at 40¡C for 2 hours. Freeze at -20¡C. Labelled probes are purified using Qiagen Rneasy mini kit and protocol provided by Agilent. For each probe, add 16ul of nuclease free water, 350 ul RLT buffer, and 250 ul ethanol 100 ¡. Mix by gently pipetting. Apply 700 ul on Rneasy columns and spin at 13,000 g for 30 s at 4¡C. Discard flow-through. Wash twice with RPE buffer. Dry the column and elute in 30 ul nuclease free water. Measure concentration and Cy5/3 incorporation using a Nanodrop spectrophotometer. Adjust concentration at 100 ng/ul. Freeze at -20¡C until hybridisation.
Total cellular RNA was isolated using Norgen extraction kit (Biotech Corp.) following recomendation of the manufacturer
CD34+ cells were isolated using an immunomagnetic beads technique (Miltenyi Biotech, Paris, France) and grown in serum-free medium supplemented with human recombinant thrombopoietin (TPO) and stem cell factor (SCF) as previously described (Debili et al 1993).hESC were mechanically harvested, and dissociated as small clumps and cells were seeded on a sub-confluent OP9 cell layer in the presence of 20 ng/mL Vascular Endothelial Growth Factor VEGF (Miltenyi Biotech, Paris France) and 20 ng/mL TPO. At day 7, 25 ng/mL SCF was added. At day 10-12 sorting of CD34+/CD41+ cells that were replated on OP9 cells monolayer in the presence of TPO and SCF was performed. Before RNA extraction cells were sorted on the positivity for CD41 and CD42 and on the negativity for CD14 and CD15 markers.