442
M. G. Quesne and S. P. de Visser
substrate position to compartmentalize their reactivity [28]. However, in the view
of these authors such objections should be system-specific and when done correctly
(and on appropriate systems), cluster model techniques remain powerful tools for
studying enzymatic systems in general and metalloenzymes in particular.
2.2 Quantum Mechanics/Molecular Mechanics (QM/MM)
QM/MM studies begin with selecting heavy atom coordinates obtained from X-ray
crystallography [29] and deposited at the protein databank. It is often impossible to
crystallize a protein-substrate complex; therefore, competitive inhibitors are often
used to bind to an active site in a manner that closely mimics the natural substrate
or alternatively the metal is replaced by an unreactive metal such as Zn(II). As a
consequence, QM/MM calculations involve a detailed setup procedure starting from
a protein databank file (pdb) [29] that will need to be carefully analyzed and often
modified and/or corrected to create an appropriate starting structure for the QM/MM
calculations. After such a structure is obtained, some heavy atom coordinates need
to be modified into those of the starting structure, for instance, the user may want
to change the resting state crystal structure into a QM/MM model of the active
oxidant of the enzyme. This could for instance, mean replacing an iron(III)-water
complex with an iron(IV)-oxo structure [28]. Next, hydrogen atoms must be readded, since the X-ray scattering ability of their single electron is not great enough
to be resolved. Several protonation procedures exist but a popular one utilizes a web
server a common Web service for proteins uses the PROPKA technique [30]. This is
typically done at an estimated pH value of approximately 7 [30], which is the active
pH of most proteins. Crystallization techniques necessitate the removal of water from
the protein suspension, and this combined with difficulties in resolving their positions
leads to a reduced amount of crystal waters, present in the electron density maps.
Therefore, there is a need to re-solvate the metalloenzyme using a water sphere or box
centered around the enzyme/substrate system. After the solvation step, a molecular
dynamics (MD) simulation is run, with pressure being exerted from the periphery
of the system to force water molecules into the protein. Biomolecular force fields
that have been extensively parameterized for use on biopolymers, such as AMBER
[31], CHARMM [32], and GROMOS [33], are highly recommended. For a detailed
overview of how to set up a QM/MM calculation and the difficulties and challenges
encountered, a tutorial review on the subject is available in the literature [21]. As
described below, one of the real advantages in this type of setup relates to the explicit
solvation approach; whereby, a systematic solvation protocol allows explicit water
molecules into the active site and so well-prepared QM/MM models often replicate
the specific polarity of the metalloenzyme more accurately.
Once the system is solvated, several different QM/MM protocols can be followed,
whereby the complete system is split into two regions as explained in Fig. 1 above:
a QM region for the inner core of the protein and the MM region for the rest. The
interaction between these two regions is important for the electronic description of the
M. G. Quesne and S. P. de Visser
substrate position to compartmentalize their reactivity [28]. However, in the view
of these authors such objections should be system-specific and when done correctly
(and on appropriate systems), cluster model techniques remain powerful tools for
studying enzymatic systems in general and metalloenzymes in particular.
2.2 Quantum Mechanics/Molecular Mechanics (QM/MM)
QM/MM studies begin with selecting heavy atom coordinates obtained from X-ray
crystallography [29] and deposited at the protein databank. It is often impossible to
crystallize a protein-substrate complex; therefore, competitive inhibitors are often
used to bind to an active site in a manner that closely mimics the natural substrate
or alternatively the metal is replaced by an unreactive metal such as Zn(II). As a
consequence, QM/MM calculations involve a detailed setup procedure starting from
a protein databank file (pdb) [29] that will need to be carefully analyzed and often
modified and/or corrected to create an appropriate starting structure for the QM/MM
calculations. After such a structure is obtained, some heavy atom coordinates need
to be modified into those of the starting structure, for instance, the user may want
to change the resting state crystal structure into a QM/MM model of the active
oxidant of the enzyme. This could for instance, mean replacing an iron(III)-water
complex with an iron(IV)-oxo structure [28]. Next, hydrogen atoms must be readded, since the X-ray scattering ability of their single electron is not great enough
to be resolved. Several protonation procedures exist but a popular one utilizes a web
server a common Web service for proteins uses the PROPKA technique [30]. This is
typically done at an estimated pH value of approximately 7 [30], which is the active
pH of most proteins. Crystallization techniques necessitate the removal of water from
the protein suspension, and this combined with difficulties in resolving their positions
leads to a reduced amount of crystal waters, present in the electron density maps.
Therefore, there is a need to re-solvate the metalloenzyme using a water sphere or box
centered around the enzyme/substrate system. After the solvation step, a molecular
dynamics (MD) simulation is run, with pressure being exerted from the periphery
of the system to force water molecules into the protein. Biomolecular force fields
that have been extensively parameterized for use on biopolymers, such as AMBER
[31], CHARMM [32], and GROMOS [33], are highly recommended. For a detailed
overview of how to set up a QM/MM calculation and the difficulties and challenges
encountered, a tutorial review on the subject is available in the literature [21]. As
described below, one of the real advantages in this type of setup relates to the explicit
solvation approach; whereby, a systematic solvation protocol allows explicit water
molecules into the active site and so well-prepared QM/MM models often replicate
the specific polarity of the metalloenzyme more accurately.
Once the system is solvated, several different QM/MM protocols can be followed,
whereby the complete system is split into two regions as explained in Fig. 1 above:
a QM region for the inner core of the protein and the MM region for the rest. The
interaction between these two regions is important for the electronic description of the