322
X. Yang et al.
16.4.2 Refining the Complex
The biomacromolecule is conformationally altered during the binding process. For
example, the position of helix 12 in the human ER underwent a large rearrangement
when the ligand changed from an agonist to an antagonist [72]. Not accounting for
such structural changes during docking may prevent identification of the true binding
pose. Docking which fully considers target flexibility is very computationally expensive. Instead, semi-flexible docking where the flexibility ligands are taken in account
while the protein is kept rigid is more commonly used to prepare the toxicant–target
complex.
To overcome this problem, molecular dynamics (MD) simulation or hybrid quantum mechanics/molecular mechanics (QM/MM) simulation can further refine the
complex generated from docking [73–75]. MD simulation leverages classical molecular mechanics force fields to predict particle motions as a function of time [76]. In
contrast, QM/MM simulation queries the site of interest (QM region) with quantum
mechanics. The rest of the system (MM region) is studied with classical molecular
mechanics force fields [77]. Those two provide detailed information on the conformational changes and fluctuations of the molecules in the complex, and both are
now routinely employed to refine molecular structures, investigate the dynamics of
a given molecular system, and elucidate atomic-level interactions [78]. Various software including AMBER [79], GROMACS [80], NAMD [81], CHARMM [82] could
be employed to refine the complex.
As described so far, the desired toxicant–target complex is prepared on the basis of
molecular docking and/or MD simulation or QM/MM simulation. Next, the types of
intermolecular interactions, bond distances, and binding affinities can be inferred and
used to reveal the underlying interaction mechanism between EDCs and endocrine
system targets.
16.5 Probing the Underlying Binding Mechanism of Action
16.5.1 Analyzing Binding Patterns
As stated above, the binding conformation informs the binding site location and
binding orientation of ligands within the LBD of a given target. Thus, by analyzing the
binding pattern, we can determine the dominant orientation and preferential binding
region for the functional groups in EDCs. Then, we would be able to further analyze
the binding conformation similarities and differences between endocrine hormone
and EDCs, and/or among the different EDCs. For example, by analyzing simulated
ligand conformations and hTTR crystal structures, we found that the ionized function
group (e.g. O
− , COO
− ) in the ligand had a dominant orientation and pointed toward
the entry port of the binding site. However, no dominant orientation was observed
for the neutral form of the ionizable group [83, 84].
X. Yang et al.
16.4.2 Refining the Complex
The biomacromolecule is conformationally altered during the binding process. For
example, the position of helix 12 in the human ER underwent a large rearrangement
when the ligand changed from an agonist to an antagonist [72]. Not accounting for
such structural changes during docking may prevent identification of the true binding
pose. Docking which fully considers target flexibility is very computationally expensive. Instead, semi-flexible docking where the flexibility ligands are taken in account
while the protein is kept rigid is more commonly used to prepare the toxicant–target
complex.
To overcome this problem, molecular dynamics (MD) simulation or hybrid quantum mechanics/molecular mechanics (QM/MM) simulation can further refine the
complex generated from docking [73–75]. MD simulation leverages classical molecular mechanics force fields to predict particle motions as a function of time [76]. In
contrast, QM/MM simulation queries the site of interest (QM region) with quantum
mechanics. The rest of the system (MM region) is studied with classical molecular
mechanics force fields [77]. Those two provide detailed information on the conformational changes and fluctuations of the molecules in the complex, and both are
now routinely employed to refine molecular structures, investigate the dynamics of
a given molecular system, and elucidate atomic-level interactions [78]. Various software including AMBER [79], GROMACS [80], NAMD [81], CHARMM [82] could
be employed to refine the complex.
As described so far, the desired toxicant–target complex is prepared on the basis of
molecular docking and/or MD simulation or QM/MM simulation. Next, the types of
intermolecular interactions, bond distances, and binding affinities can be inferred and
used to reveal the underlying interaction mechanism between EDCs and endocrine
system targets.
16.5 Probing the Underlying Binding Mechanism of Action
16.5.1 Analyzing Binding Patterns
As stated above, the binding conformation informs the binding site location and
binding orientation of ligands within the LBD of a given target. Thus, by analyzing the
binding pattern, we can determine the dominant orientation and preferential binding
region for the functional groups in EDCs. Then, we would be able to further analyze
the binding conformation similarities and differences between endocrine hormone
and EDCs, and/or among the different EDCs. For example, by analyzing simulated
ligand conformations and hTTR crystal structures, we found that the ionized function
group (e.g. O
− , COO
− ) in the ligand had a dominant orientation and pointed toward
the entry port of the binding site. However, no dominant orientation was observed
for the neutral form of the ionizable group [83, 84].