Chromatin immunoprecipitation (ChIP) ES and FACS-sorted cells were fixed in % formaldehyde and sonicated until fragments of an average size of 5 bp were obtained. Soluble, sheared chromatin was diluted to a concentration of µg/ml in RIPA buffer ( mm Tris-HCl ph 7.5, mm EDTA,.5 mm EGTA, % Triton X-,.% SDS,.% Sodium Deoxycholate, 4 mm NaCl, protease inhibitors). To analyse Ringb binding 3 µg chromatin (3 µl) was added to 5 µl protein G Dynabeads (Invitrogen) saturated with anti-ringb antibody (MBL) or nonspecific IgG (DAKO) and incubated for at least 6 hours at 4 C on a rotating wheel. Nonimmunoprecipitated chromatin ( µl) was kept as input. Immune complexes were washed twice in RIPA buffer, twice in RIPA buffer with 5 mm NaCl, once in RIPA buffer with 5 mm LiCl, and twice in TE. Prior to elution, samples were transferred to a new tube. Elution was done in 5 µl elution buffer ( mm Tris-HCl ph 7.5, 5 mm EDTA, 5 mm NaCl, % SDS, 5 µg/ ml Proteinase K, 5 µg/ml RnaASE A) on a shaking thermal block at 68 C for at least 6 hours. Eluates were transferred to a new tube and beads incubated with additional 5 µl of elution buffer for 5 min. The two eluates were pooled and DNA was purified by phenol:chloroform extraction, precipitated and resuspended in 4-8 µl TE. µl were used in each PCR reaction. To analyse the abundance of modified histones the procedure was modified as follows. Each IP contained µg ( µl) chromatin which was added to 5 µl protein A Dynabeads (Invitrogen) saturated with either anti-h3k4me3 (Abcam), anti-h3k7me3 (Millipore), anti-total H3 (Abcam) or non-specific IgG (DAKO). Different anti-h3k4me3 (Millipore) and anti-h3k7me3 (Millipore) antibodies were used for the domain analysis shown in figure C. The reactions were done in duplicate in a 48- well format. Elution was done as described above or in µl modified elution buffer (5 mm Tris-HCl ph 7.5, mm EDTA, 5 µg/ml Proteinase K, 5 µg/ml RNase A, ph 9.8) by incubation for 3 min at 37 C, 6 hours at 68 C, min at 95 C. The eluate was used directly as template in quantitative PCR reactions (5 µl per sample). For sequential ChIP, µg chromatin was incubated with primary antibody (anti-h3k4me3) cross-linked with dimethyl pimelidate dihydrochloride (DMP, Sigma) to protein A-coated magnetic beads (Invitrogen). Following removal of unbound chromatin, the immunoprecipitated material was eluted in mm Tris-HCl ph 7.5, 5 mm EDTA, 5 mm NaCl, % SDS by incubating 3 min at room temperature and min at 68 C. The eluate was diluted : in RIPA
buffer without SDS and split into two aliquots receiving either the second antibody (anti- H3K7me3) or non-specific IgG (DAKO) pre-bound to protein A-coated magnetic beads. Immunoprecipitated DNA was quantitated by real-time PCR using primers amplifying the promoter regions of candidate genes (sequences are available upon request). Hematopoietic assay Assays for hematopoietic colonies were performed essentially as described (Fehling et al, Development. 3; 3:47-47). FACS sorted Flk+ (hemangioblast) Ringa -/- /Ringb fl/fl /Cre-ER T cells were reaggregated in either EB differentiation medium or blast medium containing 8 nm 4-OHT (tamoxifen) or vehicle for -3 days before plating in semi-solid hematopoietic medium. Colony scoring was done after 6 days. Figure S. Changes in gene expression and replication timing of candidate loci upon mesodermal differentiation. (A) Expression of marker (Sox, Oct4, Rex), mesoderm-associated (Bra/T, Flk, Ikaros, Myf5) and neuralassociated (Math, Nkx-, Nkx-9, Sox) genes or a ubiquitously expressed gene (β- Actin), in and FACS-enriched cell samples. Values were normalised to a housekeeping gene (Hmbs) and presented as fold change over positive control samples ( for β-actin, Oct4, Sox, Rex; FACS-sorted cells for Bra/T; FACS sorted cells for Flk; B3 pre-b cells for Ikaros, differentiated CC myoblasts for Myf5; embryonic head (E4.5) for Math, Nkx-; GFP+ FACS-sorted, differentiated Sox-GFP 46C for Sox; embryos (embryonic day 8) for Nkx-9). The average and standard deviation from three experiments are shown. (B) Summary of the replication timing profiles of individual loci during mesodermal commitment, colourcoded to facilitate comparisons as previously described (Azuara et al. Nat Cell Biol. 6; 8:53-538; Perry et al. Cell Cycle. 4; 3:645-65); early replication (peak in G and/or S) in green, middle-early (peak in S) in lime, middle (peak in S and S3) in yellow, middle-late (peak in S3) in orange and late replication (peak in S4 and/or G) in red. *Hem. Prog.: the adult hematopoietic precursor cell line FDCP mix (ref. 8). **Neural prog.: ES-derived neural precursors induced with retinoic acid (ref. 3). : Not analysed. Figure S. Overview of the replication timing assay. (A) Asynchronously dividing cell populations were pulsed with BrdU to label newly replicated
DNA. The cells were harvested and stained with propidium iodide (PI), sorted according to DNA content into six fractions representing G, four sequential S-phase stages (S-S4) and G/M, as indicated. As a control for equivalent handling of samples, an equal amount of BrdU-labelled Drosophila genomic DNA is spiked into each cell-cycle fraction. After DNA isolation and sonication, BrdU-labelled DNA is immunoprecipitated using anti-brdu antibodies. Quantitation by real-time PCR of the amount of locus-specific newly replicated DNA in each fraction is used to determine the timing of DNA replication as early (peak abundance in G and S), mid-early (peak in S), middle (peak in S and S3), mid-late (peak in S3) or late (peak in S4 and G/M). (B) Results showing equivalent recovery of spiked DNA (Drosophila Gbe gene), and early (α Globin locus) and late (X4 repeat) replicating loci are provided for reference. Figure S3. Analysis of PRC levels and function in Flk+ hemangioblast cells. (A) Quantitative RT-PCR analysis of the mrna levels of PRC components Mel8, Bmi, Ringa and Ringb during mesodermal differentiation of ES cells. Values were normalised to a housekeeping gene (Hmbs) and presented as fold change over. The average and standard deviation from three experiments are shown. (B) Western blot analysis of Ringb levels in sequential three-fold dilutions of whole cell extracts from ES and sorted cell populations. Anti-Tubulin antibody was used as a loading control. (C) FACS analysis of dissociated Ringa -/- /Ringb fl/fl /CreER T embryoid bodies stained with PE-conjugated anti-flk antibody, where the gate show a typical Flk- positive population. The profile is representative of several experiments (>5). (D) Western blot analysis of Ringb levels in whole cell extracts from untreated (-Tam) and tamoxifen treated (+Tam) FACS-sorted Flk+ Ringa -/- /Ringb fl/fl /CreER T cells. Anti-Lamin antibody was used as a loading control. (E) Quantitative RT-PCR analysis of the Ringb gene, hematopoietic Ringb target genes (Scl, Runx, Fli) or non-ringb target genes (Gata) in Ringa -/- / Ringb fl/fl /Cre-ER T (upper) or EB-derived Flk+ hemangioblast cells (lower) three days after addition of Tamoxifen (Tam) to delete Ringb (black bars), or in untreated controls (white bars). Values were normalized to a control gene (Hmbs) and expressed as fold change relative to untreated. The average and standard deviation from two-three experiments are shown. (F) Hematopoietic colony assay was performed in triplicate for untreated (-) and tamoxifen treated (+) samples after culturing FACS-sorted Flk+ Ringa -/- /Ringb fl/fl /CreER T cells for a single day in differentiation medium (with or without tamoxifen), before replating in semisolid hematopoietic medium. Asterisks indicate
significantly different numbers of colonies in treated compared to untreated samples (p<.5, student s t-test).
A mrna level relative to control/undifferentiated β-actin..8.6.4. Sox..8.6.4. Oct4..8.6.4. Rex..8.6.4...8.6.4...8.6.4. Bra/T Ikaros..8.6.4. Myf5..8.6.4. Flk Control Math..8.6.4...8.6.4...8.6.4...8.6.4. Nkx- Nkx-9 Sox GFP-/Flk- GFP-/Flk- GFP-/Flk- Control B Ctrl αglobin X4 ES Rex Sox Meso Bry Flk Ikaros Myf5 Neural Math Sox Nkx- Nkx-9 GFP-/Flk- GFP+/Flk- GFP+/Flk+ Hem. prog.* Neural prog.** G-S Early S Mid-early S-S3 Middle S3 Mid-late S4-G/M Late Figure S
A B Asynchronously growing cells 5 BrdU pulse to label newly replicated DNA 4 3 Quality control (Drosophila Gbe) PI staining and cell cycle sorting according to DNA content Relative cell number G S S S3 S4 G/M PI intensity (DNA content) Spike in Drosophila DNA Sonication and anti-brdu IP Newly replicated DNA (% of total) 6 5 4 3 5 4 3 G S S S3 S4 G/M Early control (α-globin) Late control (X4) Real time PCR Cell cycle fraction Figure S
A C D B mrna relative to Ringb Ringb Mel8 GFP-/Flk-.8.4 3 Ringa Bmi GFP-/Flk- GFP-/Flk- GFP+/Flk- GFP+/Flk+ E mrna level (fold change relative to -Tam) 8 6 4 3 PE-Flk 33 % Fl Flk+ (hemangioblast) Ringb F Lamin Hematopoietic colonies (per 5 cells) -Tam +Tam Day 4 * 8 6 4 -Tam +Tam Tubulin RingB Scl Runx Fli Gata Figure S3