Can the Structure of the Hydrophobic Core Determine the Complexation Site? - Identification of Ligand Binding Site and Protein-Protein Interaction Area - page 43

points in the protein body. Following normalization both distributions may be used
to calculate differences between the theoretical and observed hydrophobicity (or its
likelihood) at any coordinates.
Since both distributions are normalized (via a coefficient which appears in both
equations) the irregularity of the hydrophobicity distribution in actual proteins may
be measured by comparing idealized and observed values, applying the following
expression:

t e

Δ= −

HHH

i i i

Δ H )
are thought to correspond to cavities capable of binding ligands. On the other hand,
it is assumed that excess hydrophobicity (expressed by low negative values of

Local hydrophobicity de fi ciencies (expressed by large positive values of

Δ H ),
particularly when observed on the protein surface, may trigger protein-protein
complexation.
Thus, the positions of local minima and maxima in the Δ H pro fi le may indicate

residues involved in protein-protein interactions, ligand complexation or other types
of interaction.
The validity of the presented model may be verified by analyzing actual proteins,
both accordant with and divergent from theoretical assumptions. Identifying a
protein as structurally accordant can be treated as an argument in support of the
model, reflecting the influence of the aqueous environment on the protein body.
On the other hand, when serious discrepancies between actual and predicted struc-
tures are observed, a thorough analysis of their underlying causes may lead to useful
conclusions. Determining the reasons behind irregularities in the structure of the
hydrophobic core may yield fresh insight into the mechanisms of protein folding.
A sample differential profile (highlighting the discrepancies between the expected
and observed hydrophobicity distributions) is shown in Figs. 3.1 and 3.2 . Figure 3.1
depicts a protein whose structure is highly accordant with theoretical predictions
(1BDD) (Gouda et al. 1992 ) , while Fig. 3.2 represents a case of poor agreement
between the model and observed properties (1 G58) (Ramoni et al. 2001 ) .
1BDD (60 aa) is a recombinant B domain (FB) of the staphylococcal protein A,
which specifically binds to the Fc portion of immunoglobulin G. Its

Δ H pro fi le is
shown in Fig. 3.1a 1 G58 (159 aa) is an odorant-binding protein in form of homodi-
mer which complexes its ligand (1-octen-3-ol). Figure 3.2a presents the Δ H pro fi le

for this protein. Residues involved in ligand complexation and monomeric unit
binding have been highlighted.
Figure 3.1b depicts the theoretical (T) and observed (O) distributions of hydro-
phobicity for the 1BDD protein while Fig. 3.2b presents the corresponding distributions
for 1 G85. While the theoretical and observed distributions are in good agreement
for 1BDD (as can be seen in Fig. 3.1b ) they remain substantially divergent in the
case of 1 G85 (Fig. 3.2b )