HuPON1 (Human Paraoxonase 1) has been targeted as a potential ‘bioscavenger’ enzyme to hydrolyze organophosphorus (OP) nerve agent compounds. With little structural data and an unknown active site mechanism, a collaborative project involving the Weizmann Institute, the USAMRICD (United States Army Medical Research Institute of Chemical Defense), NIH, and other universities was developed with the goal of engineering HuPON1 for catalytic and broad substrate hydrolysis of OP nerve agents. Along with directed evolution and mutagenesis techniques, computational modeling has proved effective in finding important active site residues for binding. Directed evolution studies by the Weizmann Institute of Science developed highly expressible rePON1 variant G3C9, and later stereospecific coumarin-OP mutants 2B4, 3B3, and 8C8. Computational models of these variants were relaxed through molecular dynamics, and used in docking studies with coumarin-OPs, V-agents, and G-agents.
The ability to turn over the more toxic (-) isomer of nerve agents exclusively has been sought in recent directed evolution and mutagenesis studies. In order to rationalize the method of producing (-) isomer stereospecificity, mutants of negligible specificity (G3C9), (+) isomer specificity (2B4/3B3), and (-) isomer specificity (8C8), were computationally studied and compared by molecular dynamics results and docking conformations. Unique binding modes were seen amongst the models for the same coumarin-OP ligands, which may be due to spatial changes in the active sites. A movement of the catalytic calcium was observed in each mutant, with an attempted rationalization by the way active site mutations affect the calcium coordination environment. The combination of calcium movement and spatial changes in active sites has been considered important in rationalizing stereospecificity.
The G3C9 variant displayed two predominant binding modes for low energy coumarin-OP and V-agent compounds. No consistent theme was observed between (R) and (S) isomer binding conformations. Kinetic data from G3C9 single mutants with coumarin-OP substrates was obtained to rationalize docking results. The 3B3 and 2B4 mutants display a lowered calcium environment likely due to alternative conformations of active site mutations L240S and T332A. Stereospecificity for (+) isomer coumarin-OPs was investigated through binding pose MM-PB/SA energy and orientation relative to a proposed hydrolysis mechanism involving His285, Asp269, and a water nucleophile. The VR / VX ratio for V-agents in the 2B4/3B3 variants was explored through binding energy.
The recently evolved 8C8 variant was compared to related variants 2H4 and 4E9 to investigate calcium movement and coordination environment in stereospecific mutants for toxic (-) isomer coumarin-OP hydrolysis. A distinct calcium environment was seen, with calcium residue coordination reflective of that seen in related crystal structures by the Weizmann Institute. Coumarin-OPs in 8C8 were shown to not mimic their associated G-agents in binding pose, which is corroborated by USAMRICD experimental results. The 8C8 model exhibited a predominance of higher energy poses for (-) isomer coumarin-OPs that were set up for hydrolysis by our proposed Asp269/His285/water mechanism. The recurring conformations seen for (S) isomer coumarin-OPs in the 8C8 model may indicate key residues for (-) isomer stereospecificity.