The purpose of this investigation is the development of techniques for the modeling of the outcome of an interaction between a substrate molecule and the active site of a P450 enzyme.
Due to the high cost of calculations on such a system it would be prohibitively expensive to determine the orientation of the substrate molecule within the active site from first principles. We will therefore restrict ourselves to cases in which data is available on this orientation.
The investigation will consist of three principle stages as follows:
In this phase a basic approach will be developed using the
substrate free and substrate bound structures of 
. A
table of substrates for which the bound structure is known is given in
Table 5.1. This data defines the orientation of the
substrate within the active site.
Table 5.1: Determined structures of substrate bound 
.
The first stage of the reaction cycle (See Section 5.2)
will be studied in an attempt to model the displacement of the water
ligand and the low spin to high spin transition of the 
ion on binding of a reactive substrate. The energy difference between
the low spin, water-bound and high spin, water-free states determines
the fraction of low spin character retained by the bound substrate
system, and hence the redox potential of the iron ion. This in turn
allows the estimation of the relative rates of reaction.
Initially the interaction will be modelled using only the substrate
molecule, the haem moiety and other 
ligated groups (See
Figure 5.4 for an example). If this minimalistic approach
proves to be insufficient other contact residues will be introduced in
order of their accepted importance until an adequate system to model
the reaction has been found, or the system grows too large for the
computational resources available.
Figure 5.4: The
orientation of the haem moiety, CYS357 residue and 2-phenyl imidazole
molecule in bound state. (C=Grey, O=Red, Fe=Orange, S=Yellow, N=Blue, H
not shown)
The behaviour of the substrates in Table 5.1 are well known. This will allow the validation of the computational techniques used by comparison of results with experimental data.
In order to fully validate the methods developed in Phase 1, these will be used to predict experimental results for a set of substrates to be determined by Professor Tucker and Dr Ellis from published data. The experimental data will not be disclosed, facilitating a blind test of the techniques used.
The substrates will be chosen from those substrates of
for which a bound structure has been determined.
This enzyme was chosen as it is thought to have a structure closer to
Human P450 than that of 
. The use of substrates for
which the 3-dimensional bound structure is known will remove any
uncertainty caused by incomplete knowledge of the orientation of the
substrate molecule.
On completion of the validation of the model the techniques developed will be applied to problems of a more speculative nature. In particular the reactions of various substrates of Human P450 CYP2D6 will be investigated.
This enzyme is of special interest in pharmacology as it is responsible for the metabolism of many pharmacologically active molecules. The 3-dimensional structure of this CYP2D6 is as yet unknown, however many features of the active site have been determined. The computational techniques will be used to test theories regarding the processes giving rise to unexplained experimental results and in order to predict new data. At this stage the introduction of some of the determined contact residues of the active site into the model will allow theories of the orientation of substrate molecules to be tested.