orientations are occasionally encountered in experimental struc-
tures (Fig.3b). Sometimes, thecispeptide is present in the native
form of the molecule and may be biologically important, particu-
larly if the peptide contains proline (Fig.3c). In other cases,cis
peptides are introduced during automated refinement of experi-
mental structures, or during computational modeling. In the case
of the example project for this chapter, the crystallographic struc-
ture contains twocispeptides: one between Gly162 and Ser163 and
another between Arg212 and Pro213. The former involves a mem-
ber of the aryl acylamidase’s catalytic triad, Lys84-cisSer163-
Ser187, and plays a key role in its hydrolysis mechanism [13]. The
latter contains a proline, which exhibits similar structural energetics
fortransandcisorientations, and is, thus, considered realistic for
the model. In the event that an artifactualcispeptide is found, it can
be adjusted to thetransorientation using a program like cispeptide,
available in VMD (seeNote 4).3.2 Immersing the
Protein–Drug Complex
in a Realistic
Environment
Once the initial structural model has been generated and verified,
its surrounding environment must be appropriately accounted for.
Careful mimicking of realistic solvent conditions during simulation
is key to producing results that are comparable to experiment or
have predictive power toward biological questions (seeNote 5).
The first environmental factor to address is pH. For classical
simulation methodology, performing a simulation at a given pH
means constructing the initial model to represent the likely proton-
ation state at that pH (seeNote 6). This can be achieved by
calculating local pKa values and assigning hydrogen atom positions
accordingly. Programs like PDB2PQR [23], which performs pKa
calculations using PROPKA [24], can be employed to place hydro-
gen atoms, accounting for both protein protonation states and the
likely neutral states of histidine residues, which can carry a hydro-
gen on either their delta or epsilon nitrogen. The accurate assign-
ment of hydrogen atoms to the system may be critical for
reproducing hydrogen bond networks within the higher-order
protein structure, as well as between the protein and drug mole-
cule. Further, accounting for differences in these networks as a
function of pH is critical to simulation accuracy.
For the example system in this chapter, the protein–drug com-
plex is studied at pH 7. As the project aims to investigate the
interaction of an aryl acylamidase with the drug acetaminophen
within a biological sample, such as a patient’s blood, physiological
pH of 7 is appropriate. However, as aryl acylamidases are known
experimentally to exhibit optimum enzymatic activity at pH 10 or
above [25] and some biosensing approaches recommend increasing
system pH to enhance hydrolysis [9], an interesting follow-up
study could investigate the system at pH 10 to shed light on the
structural origin of increased activity at high pH.Molecular Dynamics Simulations of Protein-Drug Complexes 249