Identification of critical PDX -1 binding sites in the duodenal -specific enhancer of the human adenosine deaminase gene

The human adenosine deaminase gene (ADA) provides a good model system for studying the molecular mechanism(s) involved in the regulation of enterocyte-specific gene expression in the intestinal epithelium. In the intestinal epithelium, enterocyte-specific genes are subject to cellular, regional and developmental control. The ADA gene, although ubiquitously expressed, is also subject to the same developmental and spatial regulation in the intestinal epithelium as genes expressed exclusively in enterocytes. To understand the molecular mechanisms orchestrating the precise developmental and spatial patterns of ADA in the intestine, we sought to identify specific DNA regions of the ADA gene that were capable of directing expression to the intestine. Initially, functional mapping studies were carried out in transgenic mice. In the first of these transgenic mouse studies, two transgenes were generated. They each contained a 13-kb fragment from the human ADA gene that had not been previously characterized in transgenic mice. When both fragments were tested for their ability to target reporter gene expression to the intestine in transgenic animals, only one of the two 13-kb fragments contained sufficient information to correctly target reporter gene expression to the intestine in a pattern that resembles the endogenous ADA gene. In these mice, reporter gene expression was highest in the duodenum and was limited to the villus epithelium.;To locate the duodenal enhancer residing within the 13-kb ADA gene segment, further mapping studies were carried out in Mode-K cells, a mouse intestinal cell line. Sequences at the 5' end of the 13-kb fragment demonstrated enhancer-like activity similar to that displayed by the intact 13-kb segment in this cell type. To validate the results of our cell culture studies, the 5' sequences were tested in transgenic mice. In vivo, these sequences were unable to target reporter gene expression to the duodenum. These results indicate that the DNA sequences at the 5' end of the 13-kb fragment did not contain the duodenal enhancer.;The duodenal enhancer was later identified within a 310-bp downstream segment of the 13-kb fragment by additional mapping studies in transgenic mice. Detailed characterization of the duodenal enhancer revealed putative transcription factor-binding sites for a number of proteins including a cluster of five potential binding sites for the homeodomain transcription factor known as pancreatic and duodenal homeobox-1 (PDX-1). Competitive gel-shift experiments and antibody studies confirmed that each of the five binding sites interacted with PDX-1 proteins. Co-transfection studies performed in CHO-K1 cells that lack endogenous PDX-1 demonstrate that PDX-1 can function to activate the duodenal enhancer through its interactions with the PDX-1 binding sites and the transcriptional coactivator p300. Site-specific mutations at one of the five PDX-1 binding sites drastically reduce this ability of the enhancer to function in CHO-K1 cells. These results demonstrate that duodenal enhancer function is dependent on PDX-1.