A transcriptional-switch model for Slr1738-controlled gene expression in the cyanobacterium Synechocystis

In this study, using a method combining sequence homology, structural analogy modeling and biochemical data, we first build the 3D structure of the complex between the poorly-characterized PerR-like regulator Slr1738 and its target DNA, which controls the defences against metal and oxidative stresses in Synechocystis. In a second step, we propose an expanded version of the Slr1738-DNA structure, which accommodates the DNA binding of Slr1738 multimers, a feature likely operating in the complex Slr1738-mediated regulation of stress responses. Finally, in agreement with experimental data we present a 3D-structure of the Slr1738-DNA complex resulting from the binding of multimers of the FUR-like regulator onto its target DNA that possesses internal repeats.Using a combination of different types of data, we build and validate a relevant model of the tridimensional structure of a biologically important protein-DNA complex. Then, based on published observations, we propose more elaborated multimeric models that may be biologically important to understand molecular mechanisms.DNA-binding proteins play a crucial role in many fundamental biological processes including transcription, regulation, as well as DNA replication and repair. Thus, a better understanding of DNA-protein interactions has both a fundamental research interest and an applied importance in medicine (development of drugs interfering with oncogene expression) and biotechnology (genetic engineering of microbial organisms).In the past, a lot of effort has been made to understand the basic principles that govern the specificity of protein-DNA interactions. It appeared that there is no simple recognition code linking the DNA interacting amino acids of a protein with their target DNA nucleotides [1]. Furthermore, there are currently no standard methods to build a 3D-structure model for the representation of a DNA-protein complex, unlike what occurs for protein-protein interactions [2]. All current methods for predi