(A,B) Anti-PECAM1-stained yolk sacs from E8.5 littermate Brg1^fl/+ (A) and Brg1^fl/fl:Tie2-Cre^+/0 (B) embryos display a comparable vascular plexus. (C,D) Anti-PECAM1-stained yolk sacs from E9.5 littermate Brg1^fl/fl (C) and Brg1^fl/fl:Tie2-Cre^+/0 (D) embryos. (E) A higher magnification view of a different region of the Brg1^fl/fl:Tie2-Cre^+/0 yolk sac shown in D. Arrows indicate sprouting or regressing vessels; arrowheads indicate abnormally thin vessels. Scale bars: 500 μm.

(A,B) Histological sections of Brg1^fl/fl (A) and Brg1^fl/fl:Tie2-Cre^+/0 (B) yolk sac vessels filled with hematopoietic blood cell progenitors, from littermate E8.5 embryos. (C,D) Histological sections of a Brg1^fl/fl yolk sac vessel (C) and a Brg1^fl/fl:Tie2-Cre^+/0 yolk sac vessel (D). The mutant vessel is devoid of embryonic blood cells. Scale bars in A-D: 40 μm. (E,F) Transmission electron micrographs of E10.5 Brg1^fl/+ (E) and Brg1^fl/fl:Tie2-Cre^+/0 (F) yolk sac blood vessels. Arrowheads indicate abnormal embryonic blood cells and blood cell fragments. EB, embryonic blood cell; EC, endothelial cell; VE, visceral endoderm. Scale bars in E,F: 5 μm.

(A-D) TUNEL staining on histological sections of E9.5 littermate control Brg1^fl/fl (A) and mutant Brg1^fl/fl:Tie2-Cre^+/0 (B) embryos, and their corresponding control (C) and mutant (D) yolk sacs. The images were merged from separate DAPI (blue) and TUNEL (green) acquisitions. The arrow in A points to the lumen of an embryonic heart (outlined in white), filled with embryonic blood cells. The inset in B focuses on TUNEL-positive blood cells found within one of the paired dorsal aortae (arrows and outlined in white) of the mutant embryo. No TUNEL-positive blood cells are detected in the control yolk sac vessel (C; outlined in white), but TUNEL-positive blood cells are evident in the mutant yolk sac vessel (D, outlined in white; shown at higher magnification in the inset in D). Scale bars: 100 μm. (E) Mean percentages of TUNEL-positive blood cells from multiple serial sections of two control and two mutant embryos stained in three independent experiments. A total of 596 and 362 blood cells were counted from control and mutant sections, respectively. Errors were calculated as s.e.m.

Mutant embryos were crossed onto the ROSA26 reporter line and were stained with X-gal for detection of β-galactosidase, which marks Cre-mediated excision events. (A,B) Blood vessels from Brg1^fl/+;R26R^R/+:Tie2-Cre^+/0 control (A) and Brg1^fl/fl;R26R^R/+:Tie2-Cre^+/0 mutant (B) E8.5 yolk sacs contain a comparable number of blood cell precursors, the majority of which express Tie2-Cre. (C) The majority of circulating blood cell precursors still express Tie2-Cre at E9.5 in Brg1^fl/+;R26R^R/+:Tie2-Cre^+/0 control embryos. (D) By contrast, fewer circulating blood cell precursors are found in Brg1^fl/fl;R26R^R/+:Tie2-Cre^+/0 mutant embryos at E9.5, but, of those cells that persist, the proportion of cells that express Tie2-Cre and are presumably deficient for Brg1 expression is dramatically decreased when compared with the blood cells in the control embryo (C). The arrow in the inset points to an abnormally shaped mutant (Brg1^fl/fl:Tie2-Cre^+/0) blood cell. Scale bar: 40 μm. (E) Mean percentages of Tie2-Cre-positive (β-galactosidase-positive) blood cells from multiple serial sections of two control and two mutant embryos at E8.5, and two control and two mutant embryos at E9.5, carrying the ROSA26 reporter, as detected by X-gal staining. A total of 941 and 487 blood cells were counted from control and mutant sections at E8.5, respectively, while a total of 1,656 and 893 blood cells were counted from control and mutant sections at E9.5, respectively. Errors were calculated as s.e.m.

(A-H) Cryosections from an E9.5 control Brg1^fl/fl embryo (A,C,E,G) and a mutant Brg1^fl/fl:Tie2-Cre^+/0 embryo (B,D,F,H) were subjected to in situ hybridization with probes against embryonic and adult α-globins and embryonic β-globins. Almost all embryonic blood cells from the control embryo express embryonic α-globin ζ(A), adult α1/2-globins (C), embryonic β-globin ϵy (E) and embryonic βH1-globin (G). In the mutant embryo, adult α1/2-globin expression is normal (D), but many embryonic blood cells can be detected that express little or no embryonic ζ (B), ϵy (F) or βH1 (H), as indicated by the arrowheads in the respective insets. Scale bars: 40 μm. (I) Mean percentages of globin-expressing blood cells from multiple serial sections of two control and two mutant embryos at E9.5, as detected by in situ hybridization in three independent experiments. Total blood cells counted from control/mutant sections with each probe were: ζ, 1521/907; α1/2, 1292/635; ϵy; 977/721; and βH1, 492/818. Errors were calculated as s.e.m. (J) ChIP assay demonstrating that BRG1 is recruited to the β-globin LCR DNAse I hypersensitive site 3 (HS3) and the ζ promoter in primitive erythrocytes. No evidence of BRG1 recruitment is detected at the α1/2 promoter. MW, 100 bp molecular weight standard; Pos, wild-type genomic DNA served as a positive control for the PCR; Neg, no DNA was amplified as a negative control; Input, total chromatin, sheared but not immunoprecipitated; Ac-H3, ChIP material immunoprecipitated with an antibody against pan-acetyl histone 3 (H3), a mark of an open chromatin structure; BRG1, ChIP material immunoprecipitated with an antibody against BRG1; Mock, mock immunoprecipitation in which the sample was not treated with antibody but was otherwise handled identically to the ChIP samples.