************第３１回尾張コンプレックスセミナー*************

題　目：　Molecular Dynamics Study on Structure and Properties of Membrane  
　     　 From Lipid bilayer to DDS(Drug Delivery System)


発表者：　Veeramol Vailikhit　（Faculty of Liberal Arts and Science, Kasetsart University）

日　時：　１２月１５日（月）午後１時３０分～

場　所：　情報科学研究科６階６１６号室

内　 容： 
Modelling of compounds in solution can be computationally cheap using 
a static gas-phase model or using IEFPCM [1]. However, these methods 
give especially poor predictions for acidic protons of solutes in a polar solvent. 
The hydrogen bonding between the solute and solvent molecules adds a 
complexity to the electronic environment that requires more sophisticated modelling, 
which comes at a higher computational cost. 
Using an MD-ONIOM2 model limits the increase in cost while still giving good predictions. 
This uses standard molecular dynamics (MD) to first model the dynamics of 
solute-solvent interactions and find the solvation shell. 
The solute molecule and its solvation shell is then optimised quantum 
mechanically using the ONIOM2 method. 
ONIOM2 treats the solute molecule at a higher level of theory, 
and the whole solute-solvation shell system at a lower level of theory.

MD-ONIOM2 has been used previously to model the anti-HIV drug molecule, 
nevirapine in DMSO solution [2], and to predict its 1H-NMR chemical shifts 
with good accuracy (<±0.4ppm). 
This included one acidic amine-proton whose NMR shift was poorly predicted 
with a gas-phase model and with IEFPCM. 
MD-ONIOM2 is now being used to model other amines in DMSO solution to 
demonstrate its application in the general study of dynamic solvent-solute 
interactions, including solutes with multiple acidic protons.

With this model, the system consists of an optimised solute molecule placed 
in a box of solvent with periodic boundary conditions. 
Standard MD is run to simulate the dynamic interactions, including the formation of H-bonds. 
Radial distribution functions (RDFs) of the solvent atoms around the acidic 
amine protons show the number of solvent molecules in the solvation shell. 
>From the final 1ns of the production period, a snapshot of the system is 
taken every 100ps for ONIOM2 calculations which optimise the structure of 
the system within the solvation shell using [3]:
EHigh, Solvated   EONIOM2 = EHigh, Solute + ELow, Solvated - ELow, Solute
and generate the NMR shift of each proton from the corresponding:
 (1Hi)ONIOM2 =  (1Hi)High, Solute +  (1Hi)Low, Solvated -  (1Hi)Low, Solute
where   High = high level of quantum theory, 
             Low = low level of quantum theory, 
             Solute = solute molecule only,
             Solvated = solute with solvent molecules in the solvation shell.
              (1Hi) = shielding tensor of the ith proton in the solute molecule
The 10 NMR shifts for each proton are averaged for final NMR shift values.

To date, aniline, N-methylaniline, acetamide and valerolactam have now been 
successfully modelled using MD-ONIOM2. 
The dynamics show the H-bonded solvent molecules, as expected in number and 
position around the acidic anime protons, and their exchange with the bulk. 
Also, the different NMR shifts of the cis- and trans- protons in acetamide were found. 
In addition, this model is computationally affordable giving final results 
in 3-4 days running on a modern standalone PC.

REFERENCES 
1.　Shi, Q.; Lu, R.; Jin, K.; Zhang, Z.; Zhao, D. Green Chem. 2006, 8, 868-870.
2.　Vailikhit, V.; Treesuwan, W.; Hannongbua, S. J. Mol. Struct. (Theochem). 2007, 806, 99-104.
3.　Dapprich, S.; Komaromi, I.; Byun, K. S.; Morokuma, K.; Frisch, M. J. J. Mol. Struct. (Theochem). 1999, 461, 1-21.


***********************************************************