Computational Assessment of Trimethoprim Resistance in Dihydrofolate Reductase

Abstract

We characterize structural and dynamical changes induced on dihydrofolate reductase(DHFR). We investigate the structural features of the mutant protein that allow it to survive natural selection. Single/double/triple mutants detected via the systematic experimental approach carried out in the morbidostat1 under the selection pressure induced by the antibiotic trimethoprim(TMP), a competitive inhibitor of dihydrofolate, are studied. We investigate the structural features of DHFR that lead to catalytic activity while simultaneously casting out TMP. Binding constant and catalytic activity measurements on different mutants provide conflicting results, implying alternative routes towards conferring drug resistance. We execute extensive molecular dynamics(MD) simulations for the mutants in their folate or TMP bound conformations. We evaluate the experimental findings regarding DHFR drug resistance through structural changes in the enzyme. To understand if the molecules displaying competitive binding to the same region are discriminated by free energy changes affecting binding probabilities, we perform thermodynamic analyses via alchemical free energy perturbation calculations. Another effective factor in DHFR function is the dynamics of the Met20 loop direct contacting folate. These dynamics include an isomerization that is too slow to be directly observed by conventional MD.2 We employ perturbation-response scanning method3 to quantify the interconversion propensity of the conformers. The competing(or synergistic) effects of dynamics and thermodynamics in the mutants are assessed through a combination of these computational approaches. We reveal a trade-off between stability and enzymatic efficiency: Those mutants that have increased affinity for TMP also have larger k_cat values, albeit always below that of the wild-type. Furthermore, mutants disrupt the correlated motions observed in the TMP-bound wild-type enzyme, while that of the DHF-bound remain unaltered.1. Toprak et al., Nature Genet. 44, 101(2012).2. Schnell et al., Annu. Rev. Biophys. Biomol. Struct. 33, 119(2004).3. Atilgan&Atilgan, PLoS Comput. Biol. 5, e1000544(2009).