A conserved WGATAR motif with nearby E-boxes is insufficient for autonomous enhancer activity in chromatin. (A) Gata2 locus organization. Open and filled boxes depict noncoding and coding exons, respectively. The sequence conservation of the E-box- and GATA motif-containing core modules of the −77 and +9.5 kb sites is shown below. Arrows above the GATA motifs identify the orientation of each motif with respect to forward and reverse strands. (B) Representative photographs of whole-mount and transverse sections of two (left and right columns) E11.5 embryos harboring a transgene containing the Gata2 −77 kb site upstream of the Gata2 1S promoter fused to LacZ [(−77)1SLacZ]. For embryos containing (−77)1SLacZ, histological sections show complete lack of endothelial staining in the dorsal aorta (DA) and endocardium (EC) and also in the fetal liver (FL). −77(20) and −77(2) are two representative transgene-positive embryos. Note that the transgene lacks activity in these and 10 additional embryos tested. (C) Analysis of core module activities via generation of chimeric regulatory elements. HUVECs were transiently transfected with reporters derived from pGL3luc containing the Gata2 1S promoter cloned upstream of luciferase (1SLuc). The plot depicts the average luciferase activities of the cell lysates normalized by protein concentrations (at least three independent experiments). In each experiment, transfections were performed in triplicate. (D) G1E cells were transiently transfected with reporters derived from pGL3luc containing the Gata2 1S promoter cloned upstream of luciferase (1SLuc). The plot depicts the average luciferase activities of cell lysates normalized by the protein concentrations (at least three independent experiments). In each experiment, transfections were performed in triplicate. *, P < 0.05 with respect to (−77)1SLuc.

Strict architectural constraints for GATA factor-mediated combinatorial transcriptional control. (A) cis-element spacing requirements. Mutant plasmids were generated in which nucleotides between the E-box and WGATAR motifs were either deleted (−1, −2, and −3) or added (+1, +2, +3, +5, and +10). In the (+9.5 Δ4core)1SLuc plasmid, 4 bp between the E-box and WGATAR motifs were scrambled. G1E cells and HUVECs were transiently transfected with the indicated reporter plasmids. The plot depicts the average luciferase activities of the cell lysates normalized by protein concentrations (at least three independent experiments). In each experiment, transfections were performed in triplicate. *, P < 0.05 with respect to (+9.5)1SLuc. (B) cis element orientation requirements. G1E cells and HUVECs were transiently transfected with the indicated reporters. The plot depicts the average luciferase activities of the lysates normalized by protein concentrations (at least three independent experiments). In each experiment, transfections were performed in triplicate. *, P < 0.05 with respect to (+9.5)1SLuc.

Chromosome-wide GATA-2 occupancy at conserved WGATAR motifs and E-box-WGATAR composite motifs. (A) Quantitative ChIP analysis of GATA-2 occupancy at 63 conserved WGATAR motifs (within 3 to 50 kb of the corresponding conserved composite motifs of panel B) across mouse chromosome (Chr.) 6 in G1E cells (mean ± standard error from three independent experiments). The numbers on the x axis correspond to nearest-neighbor genes listed in Table S1 in the supplemental material. The numbers of GATA-2-occupied motifs are also shown at their specific chromosomal locations. (B to D) Quantitative ChIP analysis of GATA-2 occupancy at all conserved E-box-WGATAR composite motifs across mouse chromosome 6 (B), chromosome 1 (C), and chromosome 7 in G1E cells (mean ± standard error from three independent experiments). The numbers on the x axis correspond to nearest-neighbor genes listed in Table S1 in the supplemental material. The numbers of GATA-2-occupied motifs are also shown at their specific chromosomal location. (E) Sequence composition of 148 unoccupied (left) and 16 occupied composite elements (right). The x axis depicts the nucleotide position within the composite element, and the y axis represents the relative frequencies of the nucleotides.

Scl/TAL1 occupancy at conserved composite elements occurs exclusively at GATA-2-occupied elements. Quantitative ChIP analysis was conducted in G1E cells to measure Scl/TAL1 occupancy at GATA-2-occupied, conserved composite elements, GATA-2-unoccupied, conserved composite elements, and control sites lacking composite elements (mean ± standard error from three independent experiments).

GATA-2-Scl/TAL1-occupied and -unoccupied conserved composite elements have diagnostic epigenetic signatures. Quantitative ChIP analysis was conducted in G1E cells to measure the indicated epigenetic marks at GATA-2-occupied, conserved composite elements, GATA-2-unoccupied, conserved composite elements, and active (RPII215) and inactive (necdin) promoters lacking composite elements (mean ± standard deviation from two independent experiments).

GATA-2-Scl/TAL1-occupied composite elements function as enhancers in GATA-2- but not GATA-1-expressing cells. (A) G1E and MEL cells were transiently transfected with reporter constructs containing the indicated composite elements as well as 10 bp of upstream and downstream flanking sequence. The plot depicts the average luciferase activities of the cell lysates normalized by protein concentrations (mean ± standard error from at least three independent experiments). In each experiment, transfections were performed in triplicate. *, P < 0.05 with respect to 1SLuc. The horizontal gray bar delineates the 1.0 value of the 1SLuc construct. (B) G1E cells were transiently transfected with reporter constructs containing the wild type (sequence depicted on top of the graph), E-box mutant, WGATAR mutant, and E-box-WGATAR double mutant of the Riken 6530409C15 (Riken C15) composite element with 10 bp of upstream and downstream flanking sequence. The plot depicts the average luciferase activities of the cell lysates normalized by protein concentrations (mean ± standard error from three independent experiments). In each experiment, transfections were performed in triplicate. *, P < 0.05 with respect to (Rik. C15)1SLuc. (Rik. C15)1SLuc is the 1SLuc plasmid containing the Riken C15 composite element; (Rik. C15 mtE)1SLuc is the 1SLuc plasmid containing the Riken C15 composite element with a scrambled E-box (CATATG→GAATTC); (Rik. C15 mtG)1SLuc is the 1SLuc plasmid containing the Riken C15 composite element with a scrambled WGATAR motif (AGATAA→GAGCTC); (Rik. C15 mtE-mtG)1SLuc is the 1SLuc plasmid containing the Riken C15 composite element with a scrambled E-box and scrambled WGATAR motif.

Occupied composite elements reside at and near novel GATA factor target genes. (A) Diagrams of nearest-neighbor genes surrounding GATA-2-occupied E-box-GATA motifs on mouse chromosomes (Chr.) 1, 6, and 7. Asterisks denote the locations of conserved E-box-WGATAR motifs, arrows denote transcription start sites, and shaded boxes indicate the coding regions of the genes. (B) The table summarizes changes (fold) in the GAPDH-normalized expression of selected genes surrounding GATA-2-occupied E-box-WGATAR composite elements in day 3/4, 6, and 8 EBs derived from mouse ES cells following Dox-mediated GATA-2 induction (mean ± standard error from nine independent experiments for day 3/4 and 6 EBs and from six independent experiments for day 8 EBs) and also in G1E-ER-GATA-1 cells after estradiol-mediated activation of ER-GATA-1 (mean ± standard error from three independent experiments). mRNA levels were quantitated by real-time RT-PCR. (C) The graph depicts the GAPDH-normalized expression of genes surrounding GATA factor-occupied E-box-WGATAR composite elements in day 3 and 6 EBs derived from Gata2^−/− ES cells divided by their expression in day 3 and 6 EBs derived from wild-type ES cells, respectively (mean ± standard error from three independent experiments). mRNA levels were quantitated by real-time RT-PCR. (D) Model of BMP4-GATA-2-fibromodulin regulatory circuit.