A computational approach for the study of ligand binding mode in exopolyphosphatase from E. coli

BOETSCH, C; LISA, AT; BEASSONI, PR

Cordoba

Congreso; Segundo Congreso Argentino de Bioinformática; 2011

A2B2C Asociación Argentina de Bioinformática y Biología Computacional

Background: Polyphosphates (PPn) have several functions, and besides of act as Pi reservoir, they were involved in pathogenesis and stress response. PPn are sinthetized by polyphosphate kinase (PPK) and hydrolyzed by exopolyphosphatase (PPX). E. coli PPX is a processive enzyme, Mg2+ and K+ dependent. It is a head to tail dimer and each monomer is divided in an N-terminal domain whth catalytic activity and the C-terminal domain involved in the binding of long chain PPn. A cleft was described in the dimeric interfase where the possitive electrostatic potential would be a key aspect for PPn and other ligans binding. In this work, we performed molecular simulation with PPX bound to Mg+2 and docking assays with small chain PPn (PP3 and PP5) and organic ligands (guanosine, adenosine, ATP, ADP, ppGpp and pppGpp) with NAMD software. PP5 molecule was optimized by parametrization with paratool of VMD and quantum chemistry calculations with Gaussian. Results: Docking assays show the preferential binding of phosphodiester ligands in regions with positive electrostatic potential. The presence of basic residues seems essential in this aspect. To predict the binding mode of long chains of PPn, docking with PP5 was performed on all solvent accessible surface of PPX, using 165 boxes of 15 Å on each side, large enough for a molecule of PP5 and allowing a good sampling in each trial. In Fig 1, the solvent accessible surface colored according PPX ligand binding energy is represented. Binding energies were favorable (Gb