Runx1-VE+ Runx1+VE+CD41-Runx1+VE+CD41+Runx1+VE-CD41+ Supplementary Figure 1 Supplementary Figure 1: Validation of cell populations for gene expression assays. Marker gene expression was analysed in populations 1, 2, 3 and 4 using quantitative RT-PCR. The following genes were tested and displayed expected expression patterns: Runx1, Gata1, b Major, B H1, Gfi1b and Cldn5. Runx1-VE+ Runx1+VE+CD41-Runx1+VE+CD41+Runx1+VE-CD41+

Peak NumbersSequence motifs Population 267Max, JunD, CRE, Runx Population 351Runx Population 4747Runx, Gata, ETS,JunD GO overrepresentation (p value < 0.001) Population 2Chitin binding, membrane, ABC transporters Population 3Chitin binding, membrane, ABC transporters, cytosceleton Population 4Phosphoprotein, acetylation, transcription regulation, myeloid cell differentiation, endomembrane system Supplementary Figure 2 A B Supplementary figure 2: Motif and Gene Ontology analysis of Runx1 ChIP-Seq results for individual populations. A) Overrepresented motifs in Runx1 ChIP-Seq peaks from populations 2, 3 and 4. Shown is the number of peaks found for each population as well as the names of overrepresented motifs. B) Gene ontology functional enrichment analysis for genes next to Runx1 ChIP-Seq peaks from populations 2, 3 and 4. Shown are all Gene Ontology categories enriched with a p value <

Supplementary Figure 3 Supplementary figure 3: Haematopoietic potential of sorted populations. In vitro hematopoietic differentiation potential of flow-sorted populations was assayed by culture on OP9 feeder cells. 500 cells from each subsets were seeded on OP9 feeder cells in the presence of SCF, Flt3L and IL-7. After 14 days of culture, the cells were harvested and analyzed by FACS for the expression of Mac1 and CD19 to check the production of myeloid and lymphoid lineages. Myeloid potential Lymphoid potential Runx1-VE-cadherin+ -- Runx1+VE-cadherin+CD Runx1+VE-cadherin+CD Runx1+VE-cadherin-CD41+ ++

Supplementary Figure 4 Supplementary Figure 4: Validation of Runx1 ChIP assays. Runx1 ChIP assays in populations 2, 3 and 4 were validated using quantitative PCR with primer pairs for a region known to be bound by Runx1 (+23 kb, blue bars) and a negative control region (+30 kb, green bars). Primer pairs are the same as in Wilson et al (Blood (22): ) Runx1+ VE-cadherin+/CD41- Runx1+ VE-cadherin+/CD41+ Runx1+ VE-cadherin-/CD41+ Runx1 +23kb (positive region) Runx1 +30kb (negative region) Fold Enrichment

Supplementary Figure 5 Supplementary Figure 5: Gene ontology (GO) analysis for Runx-1 bound correlated (A) and anti-correlated (B) genes. Shown are the full GO trees generated by the FATIGO program as part of the BABELOMICS suite of analysis tools (Medina et al, Nucleic Acids Res (Web Server issue):W210-3). The degree of shading corresponds to the significance of overrepresentation. AB

Supplementary Figure 6 Supplementary Figure 6: The CD41 promoter is bound by both the Scl/Gata2/Fli1 triad and Runx1 in haematopoietic progenitor cells. Raw ChIP-Seq read data were transformed into density plots and displayed as custom tracks within the UCSC genome browser above the UCSC track for gene structure (data obtained from Wilson et al Cell Stem Cell (4):532-44). All four transcription factors show very significant enrichment over the CD41 promoter region.

Supplementary Figure 7 ~LS NP EHF LHF ES(YS) 7sp(YS) ~4sp(YS) VE-cadherin CD41 Supplementary Figure 7: Timecourse analysis of CD41 expression in early mouse embryos. CD41 and VE-cadherin expression were analysed by flow cytometry between E7.0 and E8.25 of mouse embryonic development. Seven distinct developmental stages between E7.0 and E8.5 were analysed. A distinct CD41 bright population only becomes clearly separable by FACS at the ES stage. LS = late streak stage, NP = neural plate stage, EHF = early head fold stage, LHF = late head fold stage, ES = early somite pair stage (0-1 smites pairs), 4sp = approximately 3-4 somite pair stage, 7sp = 7 somite pair stage. The Runx1 +/- and Runx1 -/- embryos analyzed in Figure 5 are approximately equivalent to the NP stage. Each FACS plot of LS-LHF stage embryos is from 8-10 pooled stage-matched whole embryos. Each FACS plot from somite pair stage embryos is from 4-6 pooled stage- matched yolk sacs. All embryos are from ICR females crossed with GFP-ICR males to avoid contamination of maternal cells.

Unknown3 Runx1 Gata2 Unknown1 Unknown2 Supplementary Figure 8 Supplementary figure 8: De novo motif discovery on Runx1 peaks from all 3 populations analysed by ChIP-Seq. In addition to Runx1 and Gata consensus motifs, a novel motif (unknown 1) was identified both as a shorter 15 bp motif and also embedded in a longer 27 bp motif (both the motif and its reverse complement were found by MEME). The same motif was also recently found overrepresented in regions bound by GABPα in human blood stem/progenitor cells (Yu et al BLOOD 2011). Three other motifs (Unknown 2-4) were also found over-represented, with motif Unknown 4 representing the typical false-positive long motifs found by MEME, and caused by repetitive DNA sequences. Unknown4