Fig. 4

In vitro binding of NeuroD and Pax6b to P^E specific sequence motives. (A) Schematic representation of potential binding sites for the transcription factors NeuroD (E-box in red) and Pax6 (blue) within the hb9 pancreas enhancer as revealed by bioinformatic sequence analyses. Pax6 sites with a core match between 0.76 and 0.9 are shown in light blue, and sites with a core match between 0.9 and 1 are shown in dark blue (core = TCACG). Orientation of Pax6 sites is indicated by position relative to the line (above: positive DNA strand; below: negative DNA). (B) EMSA studies show binding of in vitro synthesized NeuroD/E47 dimers (Prot) to oligonucleotides (DNA) spanning the E-boxes (E¹, E²) but not to corresponding oligonucleotides with mutated E boxes (E^1m and E^2m). Competitor assays with unlabelled oligonucleotides (comp, as indicated) revealed efficient competition by E¹ and E² but not by E^1m and E^2m. Notably, unlabelled E^2m reduced binding of labelled E² to NeuroD/E47, suggesting that E² binds NeuroD/E47 with lower affinity as compared to E¹. Bands indicated by the asterisk presumably correspond to monomeric DNA–protein complexes. (C) EMSA studies with in vitro synthesized Pax6b protein (Prot) and oligonucleotides spanning individual or multiple potential Pax6b sites (as indicated). EMSA signals could be detected for P^{1 + 2 + 3}, P^{4 + 5 + 6}, P^7 + 8, P^9 + 10, P¹⁴ but not for P¹¹, P¹² and P¹³ (positions of protein-bound DNAs are indicated by arrowhead).