COOPERATIVE INTERACTIONS ON THE INSULIN PROMOTER

Cornell University, Ithaca NY

Investigators

Linked publications, trials & patents

Abstract

This subproject is one of many research subprojects utilizing the resources provided by a Center grant funded by NIH/NCRR. Primary support for the subproject and the subproject's principal investigator may have been provided by other sources, including other NIH sources. The Total Cost listed for the subproject likely represents the estimated amount of Center infrastructure utilized by the subproject, not direct funding provided by the NCRR grant to the subproject or subproject staff. Cooperative interactions on the insulin promoter Summary: We are studying transcription factors that cooperatively regulate the insulin promoter: the homeodomain protein Pdx1 and the basic helix-loop-helix (bHLH) dimer between E47/NeuroD. Our goal is to determine the mechanism for their cooperativity. The immediate question we want to address is whether the DNA binding domains of these transcription factors interact on the DNA. We have determined the crystal structures of the individual protein/DNA complexes. We propose to use SAXS to characterize the Pdx1-E47/NeuroD-DNA complex using 6 different DNA sequences, with different spacing between the DNA binding sites. We propose to detect interactions between the proteins from changes in the pair distribution profile and the radius of gyration. If the data is good enough, we hope to model the complex by assembling the component crystal structures. Cooperative interactions between transcription factors are important for regulating transcription in many contexts. It is sometimes difficult to measure cooperative interactions on the DNA. If the current method is successful, it could be used as a general approach to detect cooperative interactions between transcription factors. Significance: The insulin gene is up-regulated and down-regulated in response to daily alterations in blood sugar levels and other long-range changes in demand, for example changes in body weight, pregnancy or abnormal blood sugar levels as a result of diabetes. Determining how transcription factors cooperate on the insulin promoter is key to understanding the regulation of this essential gene. In pancreatic beta cells several transcription factors contribute to the cell-specific expression of the insulin gene (1). In the effort to define such interactions we are studying two key transcription factors involved in the regulation of the insulin gene: Pdx1 and E47/NeuroD. The importance of these factors for beta cell function is demonstrated by the fact that mutations in Pdx1 and NeuroD1 cause a familial form of diabetes (2,3). While the E47-NeuroD 1and Pdx1 DNA binding sites, called E- and A-boxes respectively, are conserved in the human, mouse and rat insulin promoters, the spacing between the binding sites differs (4). This has led to the proposal that insulin is differently regulated in these species. We are using a combination of crystallography and small angle scattering in order to understand how the binding site spacing affects cooperative interactions between these factors. As pull-down assays suggest that the DNA binding domains interact directly (1), we have limited our studies to the DNA binding region of the proteins. We have determined the crystal structure of the Pdx1 and of the E47-NeuroD1 DNA binding domains bound to DNA (5,6). We also collected SANS data on a ternary complex between a 30 bp DNA fragment containing a E-box and a A-box, E47-NeuroD1 and Pdx1. We propose to use SAXS to further investigate the interactions between the DNA binding domains of Pdx1 and E47-NeuroD1 in the context of different promoter sequences. The pair distribution function from the SAXS data will indicate the proximity of the two protein domains with different promoter sequences, and whether these domains interact. The importance of the protein-protein interactions will be determined by correlating the SAXS data with assays measuring transcriptional activity. The SAXS data will also contribute to ab initio models of the Pdx1-E47-NeuroD1-DNA complex we are developing with the crystal structures and SANS data. Proposed studies: We have collected SANS contrast variation data at ORNL and at NIST. From this experience we have optimized the quality of our samples. Our recent data collection at HFIR suggested that the complex may be aggregating at high D2O. SAXS data will allow us to quickly investigate the conformation of a series of protein/DNA complexes with different DNA sequences. In order to model interactions of this complex on the DNA, we propose to collect SAXS data with the wild type rat insulin promoter sequence and five different modifications (Figure 1). To assure that the samples are not aggregating, we propose to collect a data series with 5 different concentrations of the protein/DNA complex. To determine whether the complex aggregates in high D2O, we will collect a data series with increasing concentrations of solvent D2O. Complex size: DNA - 30 bps, 19.4 kDa Pdx1 homeodomain - 9.3 kDa E47/NeuroD bHLH - 16.1 kDa From previous tests of these proteins in an X-ray beam, the proteins are prone to aggregation. The samples would be most stable if cooled to 4 C, and with the use of a flow cell. Data will be collected at a q-range from < 0.01/Angstrom to ~0.35/Angstrom. Proteins will be expressed and purified as described in (5,6). Complexes will be formed by mixing purified proteins with the DNA. Samples will be dialyzed against a known buffer, and the dialysate will be used as a blank during data collection.