[gmx-users] Urea induced denaturation and urea force fields.

[Sagar Khavnekar]

Dear Gromacs Users,

I have some questions about MD simulations we performed following our
crystallographic investigations into urea induced denaturation.

We have performed some crystallographic experiments to capture initial
phases of urea induced denaturation of HEWL.

We have observed initial changes in protein structure such as breaking of
structurally important h bonds as a result of urea binding. We further
wanted to see if similar structural alterations occur when we simulate
native structure in the presence of these urea molecules, the positions of
which are taken from the crystal structure.

hence we prepared a starting system as follows

we superposed native structure onto HEWL structure from the urea complex.
Hence in starting system HEWL is in native conformation along with urea
molecules which are observed in the case of the complex.

we solvated this starting system in tip3p water and performed energy
minimization. we had restrained urea molecules during equilibration. for
production MD we removed these restrains onto urea molecules. it was
observed that urea molecules fly away from their positions soon after
starting production MD.

Hence we restrained urea molecules in order to see local changes that these
urea molecules will bring out in the native structure if they are present
at crystallographically observed positions. I would like to get opinions
from the gromacs community about this approach of restraining urea
molecules for observation of local dynamics or for observation of possible
conformational transition from native HEWL to the one in complex with urea.

Now the question that has come is that are these retrains introducing
crystallographic artifacts into MD.

Another question is if urea molecules are flying off then is it due to the
force field that might be biased by many solution studies which suggest
that urea denatures by weak interactions in contrast to strong interactions
such as we have observed in our crystallographic study. for the purpose of
information resolution of our structures is 1.6 Ang.

Another thing I would like to have experts opinion about is as follows.

In our opinion, the best thing to do MD without restraining urea molecules
will be to simulate nanocrystal spanning few unit cells in 9M urea/ But it
will be computationally expensive. We do not have enough computing power to
carry out such a massive simulation. Hence, will it be sensible to say we
didn't do it due to lack of computing power?

Further, most of MD simulations start from Protein molecule in mixed
solvent of water and urea, but in reality when you dissolve protein in 9
molar urea protein has its solvation shell around itself. Hence, isn't it
better to perform denaturation simulation starting from hydrated protein
with few hydration shells around it? and if we do so are present force
fields able to account for urea water diffusion, especially the penetration
of urea into hydration shell of protein?

PS we used amber 03 (Not amber 99 ildn. our apologies for mistake in the
earlier mail)


-- 
Sagar Khavnekar
Project student,
Structural Biology and Molecular Biophysics lab,
UM-DAE Centre for Excellence in Basic Sciences
University of Mumbai, Vidyanagari Campus, Kalina, Santacruz (East)
Mumbai 400098, India.