Figure 3.

The −66/−61 GATA motif binds placental GATA-2 and is functionally important for cAMP inducibility of the Star promoter. A, Positioning of the −66/−61 (label in A) GATA site relative to a sequence previously defined as nonconsensus C/EBPβ-3 binding site (14), which overlaps two CRE half-sites originally described by Manna et al. (33). B, EMSAs were performed using ³²P-labeled probes corresponding to the WT −73/−48 sequence of the Star gene. Extracts were prepared from E9.5 TG cells (15 μg) and, where indicated, preincubated with specific antisera to the GATA proteins 2–4. Arrows denote the DNA-GATA complexes. Competition with ×50 excess of nonlabeled −73/−48 oligonucleotides included the WT sequence and a mutated GATA element mGATA (TTATCT to TTAagT, mGATA), as well as an AP-2/SF-1 binding element from the P450scc promoter [scc1 (15)]. C, PCR-based mutations in the −66/−61 GATA element (TTATCT to TTAagT; −96mutGATA) were created within the context of the −96/+6 sequence of the Star promoter and ligated to pCAT-Basic (14). The activities of the WT and mutated promoters were assayed in TG cells in the absence or presence of cAMP, as indicated. In addition, the −96/+6 WT promoter was coexpressed in the presence of a dominant-negative mutant protein of GATA (GATA DN). Transient transfection and CAT assays were conducted as described in Fig. 2. The results are presented as activity values relative to the activity of −96Star (mean ± sd). Activity levels were statistically significant when compared with the respective values obtained for the −96Star construct treated with cAMP: a, P < 0.05; b, P < 0.005.