Can the Structure of the Hydrophobic Core Determine the Complexation Site? - Identification of Ligand Binding Site and Protein-Protein Interaction Area - page 40

Simulating dynamic processes in an environment where the presence of water
is modeled as a large quantity of individual water molecules (each of which
comprises either a single effective atom or three separate atoms arranged into a
geometric shape and associated with a specific charge) is computationally dif fi cult
due to the large number of atoms involved. More specifically, the number of
interacting atoms (in the protein's neighborhood as well as in the protein itself)
may reach several dozen or even several hundred (Zobnina and Roterman 2009 ) .
Compounding this problem is the fact that interactions between individual water
molecules and specific atoms belonging to the protein body are highly local - thus,
modeling them individually does not reflect the holistic influence of water on
the protein (which, as discussed in (Zobnina and Roterman 2009 ) , drives the
structural arrangement and optimization of the entire molecule rather than its
constituent parts).
Stabilization mediated by hydrophobic interactions (expressing the influence of
water on the protein as a whole) can be explained on the basis of the “oil drop”
model (Kauzmann 1959 ). The model introduces the notion of a “hydrophobic core”.
It claims that the eventual distribution of hydrophobic and hydrophilic residues in
the protein body is determined by the aqueous environment. Hydrophilic residues
tend to migrate to the surface of the protein while hydrophobic residues are internal-
ized (Kauzmann 1959 ) .
Structural stabilization of the protein is also associated with optimization of
nonbinding interactions (electrostatic and van der Waals potentials), although the
optimization processes involved differ from those covered by the “fuzzy oil drop”
model (discussed in this chapter). The distribution of hydrophobic interactions may
indicate active sites, responsible for binding ligands and protein complexation.

3.2 Description of the Model

3.2.1 Theoretical (Idealized) Hydrophobicity Distribution

Our model assumes the existence of an idealized hydrophobicity distribution which
is treated as the “target”. This distribution involves a hydrophobic core (where the
concentration of hydrophobicity reaches its highest value) located at the geometric
center of the molecule. As we move away from the center, hydrophobicity decreases
stochastically, reaching a value close to 0 on the molecule surface.
In order to accurately model such a structure, a 3D Gauss function can be applied
(Konieczny et al. 2006 ). Traditionally, Gauss functions are used to model stochastic
distribution, whereas in our model they reflect the distribution of hydrophobicity.
For the sake of interpretational consistency, we can state that the values of this func-
tion correspond to the probability that hydrophobic conditions will be encountered
at specific locations within the protein body.