Because the reactivation of VX-inhibited AChE by 2-PAM when Glu202 is unprotonated seem not feasible in the light of the QM/MM results presented in the previous sub-section, a new protonation state should be envisaged for this residue. In the simulations presented in this sub-section, Glu202 is protonated (i.e. in its acid state)
Since a two-dimensional scan had to be resorted to for the unprotonated Glu202 case, a bidimensional scan was also used for protonated Glu202. Again, both P-Ox and P-OSer203 distances are controlled and scanned by 0.05 Å steps. This time 282 points were gathered, all optimized in B3LYP-D3 with def2-SV(P). All points of this two-dimensional scan yield an energy value letting three-dimensional potential energy surface to be plotted (Figure V-3). On this potential energy surface, the high energy area combining a P-OSer203 superior to 2.0 Å and a
P-Ox exceeding 2.25 Å was not fully scanned. This area, similarly to the Glu202- scan, was
probed by several mono-dimensional scan and shown to be of very high energy and thus not worth the computational cost of a full two-dimensional scan.
Figure V-3. Three-dimensional potential energy surface for the reactivation of VX-inhibited Glu202H0 AChE by 2-
PAM.
Both the reactant and products were optimized by obtained by releasing constraints on scan points with reactant-like and product-like geometries respectively. In the reactant,
called structure 1P (Figure V-4), there is a 1.67 Å covalent bond between the phosphorus and the oxygen of Ser203 while the P-Ox is 2.75 Å long. In the product, structure 2P (Figure V-4), the covalent bond is between the oxygen of the oxime and the phosphorus and is 1.72 Å long. The P-OSer203 has a length of 2.50 Å. To confirm the absence of energy variations between the edge of the potential energy surface an additional monodimensional scan with both P- OSer203 and P-Ox under constraints was realised. It confirmed the flatness of the potential energy surface in this section.
Figure V-4. Optimized structures of (1P) the reactant, (1*P) the stable intermediate, (2P) the product, and (TS1P-
2P) the transition state for the reactivation of VX-inhibited Glu202H0 AChE by 2-PAM.
The reactant and product can be connected on the potential energy surface by the minimum energy path. Following this minimum energy path, the two reaction distances P-Os and P-Oser203 vary sequentially. The shortening of the P-Ox to the range of 1.80-1.90 Å precedes the lengthening of the P-OSer203 that marks the departure of the leaving group, in
that case the whole enzyme, from the phosphorus. Along this energy path we see that the reduction of the P-Ox distance is associated with close to no energy change. After this reduction in the P-Ox distance, a stable reaction intermediate (1*P) was found and optimised. This reaction intermediate has a long P-Ox bond of 2.07 Å and a shorter P-OSer203 of 1.76 Å. No saddle point was located between this intermediate and the reactant due to a very flat potential energy surface in that section. When the P-Ox has been shortened to the range of 1.80-1.90 Å the minimum energy path shifts and the P-OSer203 increases. Along this increase of P-OSer203 along the minimum energy path the P-Ox remain almost constant and shifts form the range of 1.80-1.90 Å to the range of 1.70-1.80 Å. Following the minimum energy path as P-OSer203 is stretched, a saddle point is encountered. The transition state (structure TS1P-
2P in Figure V-4) corresponding to this saddle point was localized by a finer scan on the saddle
point area by 0.01 Å steps incremental variation of the P-Ox and P-OSer203. In this transition state, the P-Ox is 1.79 Å and the P-OSer203 is 2.11 Å. Interestingly it almost mirrors the structure of the pentavalent intermediate.
The potential energy surface is reminiscent of an addition-elimination mechanism where the two reactive distance vary sequentially and not simultaneously. The optimisation of a stable bipyramidal trigonal reaction intermediate further confirms the reactivation to likely have a two-step mechanism with a stable intermediate in between. We were unable to find the energy barrier for the first step between the reactant (1P) and the stable intermediate (1*P) due to the flatness of the surface. The stable intermediate is the pentavalent structure expected for an addition elimination mechanism on a phosphorus centre although distorted with a much longer phosphorus-oxime distance (2.07 Å) than the phosphorus-Ser203 distance (1.76 Å) where a more symmetrical intermediate could be expected. The TS1P-2P structure with its long phosphorus-Ser203 distance (2.11 Å) and the much shorter phosphorus-oxime distance (1.79 Å) is the transition state of the elimination step. Once again, there is no exchange of the proton between Glu334 and His447. This proton remains covalently bonded to His447 and maintains a hydrogen bonds with Glu334.
At the B3LYP-D3/def2-SV(P) level, used for geometry optimization, the reactivation of
AChE by 2-PAM has an energy barrier of 6.76 kcal.mol-1 and is slightly endothermic with an
energy difference of 2.97 kcal.mol-1. The stable intermediate is only 1.35 kcal.mol-1 higher in
energy than the product. Single points with the def2-TZVP basis set give an energy barrier of
8.84 kcal.mol-1 and an almost isoenergetic reaction with an energy difference of -0.19
kcal.mol-1. The stable intermediate has an energy 4.42 kcal.mol-1 higher than that of the
reactant. The DLPNO-CCSD(T) single points show the same trend with an energy gain of -6.87 kcal.mol-1 and a small reaction barrier of 1.89 kcal.mol-1. The stable intermediate is -0.14
kcal.mol-1 more stable than the reactant. The reliability of the DFT calculations is once again
confirmed. This case, contrary to the unprotonated Glu202 case, thus leads to reactivation. The energy barrier is small enough and the enthalpy is close to zero.