PROTEIN-WATER
INTERACTIONS
Sushmita Pal
1
PROBLEM
1) The simulations for unfolded state of protein is highly collapsed
Molten globule state or non-native trap states
2) The protein-water interactions are not captured properly by the
current force-fields
1) The proteins are not solvated
2) solvation free energies of side chain analogs are unfavorable
3) Protein association are slightly more favorable
2
SOLUTIONS
Parametrization strategy
The Lennard-Jones parameters were parametrized to capture the correct unfolded
state of the protein without hampering the folded state of the protein
The standard Lorentz-Berthelot mixing rule was scaled
The parameter is the adjustable scaling parameter.
Best, R.B., Zheng, W. and Mittal, J., 2014. Balanced protein–water interactions improve properties of disordered proteins and non-specific
protein association. Journal of chemical theory and computation, 10(11), pp.5113-5124.
3
Temperature (K)
Black symbols – Radius of gyration with Amber ff03w force field
Red symbols – Radius of gyration with Amber ff03ws force field
Red dot-dashed and dashed lines indicate the results obtained by reweighting at γ = 1.05 and 1.15,respectively.
Amber ff03w- force field with TIP4P/2005 water model
Amber ff03ws- force field with TIP4P/2005 water model with rescaled parameter
4
The rescaled parameter provided evidence that it correctly characterized the unfolded
proteins.
The solvation free energies were calculated for various force fields and error in comparison
to experiments were calculated
Global Measures of Deviation from Experimental Solvation Free Energies (kJ/mol)
5
To validate the protein-protein interaction association of villin HP36 was studied
Villin HP36 is soluble upto concentration of 1.5 mM so Kd > 1.5 mM
Kinetic and Equilibrium Parameters of Villin HP36 Association
A binding event as when the minimum distance between the
molecules falls below 0.2 nm and unbinding when it exceeds 1.0
nm.
Protein concentration = 8.5 mM
6
WATER MODELS
A water model is defined by its geometry along with its partial charges and interaction parameters such
as Lennard-Jones parameters.
Water has a tetrahedral geometry with two lone pairs and V shape.
Water model is characterized by 1) the sites 2) whether the model is rigid or flexible and 3)
polarization effects
http://www1.lsbu.ac.uk/water/water_models.html
7
A new water model (TIP4P-D)
A new, four-point water model, called TIP4P-D, in which the water dispersion coefficient C6 is constrained to be
∼50% larger than in current water models, and the remaining nonbonded parameters are optimized with respect
to experimental liquid water properties.
Dispersion energy
It works accurate only in the
asymptotic regime in which r
exceeds the molecular sizes.
Short range interactions are not
accounted
Various Quantum mechanical approaches and calculated
suggests:
At the typical nearest-neighbor distances
(∼3 Å between the oxygen atoms) observed in liquid
water, the cumulative effect of the higher-order C8/r8
and C10/r10 terms is no longer a small correction and
can even exceed the C6/r6 contribution to EDS.
Piana, S., Donchev, A.G., Robustelli, P. and Shaw, D.E., 2015. Water dispersion
interactions strongly influence simulated structural properties of disordered protein
states. The journal of physical chemistry B, 119(16), pp.5113-5123.
8
Increasing the dispersion energy with 50% results in more expanded unfolded protein state
9
the apo N-terminal zinc-binding domain of HIV-1 integrase (IN)
the immunoglobulin-binding domain of protein L
CONCLUSION
• = 1.1 parametrization works optimally in
capturing the true unfolded state of the protein in
context to solvation free energy and protein
association
• Changing/increasing the dispersion coefficient
could enhance the protein water interaction.
10