(1) Center for Biocrystallographic Research, Polish Academy of
Sciences, Poznan, Poland
(2) Department of Biochemistry and Molecular Biology, University
of Leeds, Leeds LS2 9JT, West Yorkshire, UK.
Type II L-asparaginase from E. coli (EcA), which hydrolyzes L-asparagine is used clinically for the treatment of leukaemia. Although its structure is known (Swain et al (1993), PNAS 90, 1474-8), its reaction mechanism and the reason for its antileukaemic activity are still unclear. There are two candidate nucleophiles in the active site, T12 and T89, which are conserved throughout all bacterial asparaginases. Mutating either threonine greatly reduces activity; mutating both destroys it completely. The reaction proceeds via a beta-acyl intermediate formed between the substrate and either of the active-site threonines. EcA crystallises in its active form, as a tetramer. The tetramer consists of two pairs of sub-units with an extensive dimer interface; residues from both subunits participate in each of the two active sites.
We performed molecular mechanics and preliminary molecular dynamics studies of dimers of wild type and T89V EcA complexed with Asp (the reaction product). Although the scaffold of the active site remains fairly constant during molecular dynamics, a loop adjacent to it (residues 16-30) is quite flexible. The dynamic behaviour of the ligands in the two equivalent active sites is not identical, indicating a certain degree of freedom. In the wild type, each aspartate can take up a large number of conformations; during dynamics H-bonds between the threonine -OH groups and the aspartate carboxyls are in rapid exchange. The aspartate carboxyls also form H-bonds to the main chain -NH groups of residues 12 and 89, and to a structurally important water molecule. In one active site, the bound aspartate "reversed" during a short dynamics run, so that its "main chain" carboxyl group adopted the position of the gamma-carboxyl group. This gives an indication of the size of the active site and the number of quasi-equivalent low energy positions which the substrate can adopt.
In the T89V mutant, we assume that the aspartate ligand is "forced" to form a covalent link with T12. The primary difference here is that no side chains in the active site form stable H-bonds with the ligand. If V89 is "mutated" back to threonine in the model, these H-bonds may re-form. We also examine the electrostatic potential within the active site, comparing different protonation states of key residues.