The Quest for Accurate Theoretical Models of Metalloenzymes …
449
4.1 Explicit Solvent Effects Captured with QM/MM Models
Another extremely diverse superfamily of oxygenases are the heme-dependent metalloenzymes, the cytochrome P450s, which are represented by over 20,000 unique
sequences and are extremely well conserved throughout the natural world [6, 56–60].
Humans also possess a very large number of P450 homologs localized mainly in
liver hepatocytes, where they are responsible for the metabolism of harmful xenobiotics, such as drug molecules [61] but also phthalates from cosmetics products [62];
however, with this family of biocatalysts being so functionally diverse they also catalyze other pathways, including hormone biosynthesis [63–65]. Such a massively
diverse functionality is hard to correlate to the large amount of structural conservation seen throughout the P450 superfamily of metalloenzymes. However, much of
the small amount of structural differences are localized close to the enzyme active
site, which amplifies the effects of such changes. Several members of this superfamily such as P450 cam possess relatively tight-binding pockets, whereas those that
catalyze the oxidation of large lipid structures, such as P450 BM3 may have very open
substrate-binding regions [66–69]. There has been a large amount of biotechnological interest in many P450 isozymes owing primarily to their much superior product
regio/stereo—selectivity, when compared to more traditional catalysts used by the
pharmaceutical industry [70, 71].
This superdiverse family of biocatalysts conserve many core structural as well as
spectroscopic features, chiefly an absorbance feature at 450 nm, which is observed
once CO binds the iron(III) co-factor of their heme co-enzyme [72]. Figure 5 shows
the active site of P450 CAM and illuminates the other main feature conserved throughout all P450s, namely the thiolate linkage to iron that connects the heme co-factor
to the protein via a cysteinate residue [73–75]. In the case of P450 CAM, Cys 357 occupies this coordination position. Figure 5 represents the substrate-bound 5-coordinate
iron(III) complex that after O 2 binding and two protonation steps is converted into
an iron(IV)-oxo heme cation radical species called Compound I (Cpd I) [76–78].
Theoretical models have been able to predict the likelihood of competitive doublet
and quartet spin states that bifurcate the reaction mechanisms catalyzed by these
enzymes and leads to the possible spin-state selective product composition [78–82]
(see Scheme 2). Cpd I is responsible for an initial hydrogen atom transfer (HAT)
step, with the potential energy surfaces of each spin state then deviating; one spin
state gives low rebound barriers (low-spin), while the other spin state (high-spin) has
significant barriers that implicate a finite lifetime of the radical intermediate leading
to rearrangement pathways and by-products of side reactions [80, 83].
Many computational studies on P450 systems have been performed and generally they reproduce experimental findings well [84–90]. Although, these small DFT
cluster models had proved useful in justifying a possible mechanism to explain the
presence of the two distinct and competitive product pathways seen in experiment,
they were insufficient when it came to describe whether these differences were caused
by some adaptation in the protein environments or by some innate properties of the
substrate(s) themselves. Because of this, our first example in this section concerns
449
4.1 Explicit Solvent Effects Captured with QM/MM Models
Another extremely diverse superfamily of oxygenases are the heme-dependent metalloenzymes, the cytochrome P450s, which are represented by over 20,000 unique
sequences and are extremely well conserved throughout the natural world [6, 56–60].
Humans also possess a very large number of P450 homologs localized mainly in
liver hepatocytes, where they are responsible for the metabolism of harmful xenobiotics, such as drug molecules [61] but also phthalates from cosmetics products [62];
however, with this family of biocatalysts being so functionally diverse they also catalyze other pathways, including hormone biosynthesis [63–65]. Such a massively
diverse functionality is hard to correlate to the large amount of structural conservation seen throughout the P450 superfamily of metalloenzymes. However, much of
the small amount of structural differences are localized close to the enzyme active
site, which amplifies the effects of such changes. Several members of this superfamily such as P450 cam possess relatively tight-binding pockets, whereas those that
catalyze the oxidation of large lipid structures, such as P450 BM3 may have very open
substrate-binding regions [66–69]. There has been a large amount of biotechnological interest in many P450 isozymes owing primarily to their much superior product
regio/stereo—selectivity, when compared to more traditional catalysts used by the
pharmaceutical industry [70, 71].
This superdiverse family of biocatalysts conserve many core structural as well as
spectroscopic features, chiefly an absorbance feature at 450 nm, which is observed
once CO binds the iron(III) co-factor of their heme co-enzyme [72]. Figure 5 shows
the active site of P450 CAM and illuminates the other main feature conserved throughout all P450s, namely the thiolate linkage to iron that connects the heme co-factor
to the protein via a cysteinate residue [73–75]. In the case of P450 CAM, Cys 357 occupies this coordination position. Figure 5 represents the substrate-bound 5-coordinate
iron(III) complex that after O 2 binding and two protonation steps is converted into
an iron(IV)-oxo heme cation radical species called Compound I (Cpd I) [76–78].
Theoretical models have been able to predict the likelihood of competitive doublet
and quartet spin states that bifurcate the reaction mechanisms catalyzed by these
enzymes and leads to the possible spin-state selective product composition [78–82]
(see Scheme 2). Cpd I is responsible for an initial hydrogen atom transfer (HAT)
step, with the potential energy surfaces of each spin state then deviating; one spin
state gives low rebound barriers (low-spin), while the other spin state (high-spin) has
significant barriers that implicate a finite lifetime of the radical intermediate leading
to rearrangement pathways and by-products of side reactions [80, 83].
Many computational studies on P450 systems have been performed and generally they reproduce experimental findings well [84–90]. Although, these small DFT
cluster models had proved useful in justifying a possible mechanism to explain the
presence of the two distinct and competitive product pathways seen in experiment,
they were insufficient when it came to describe whether these differences were caused
by some adaptation in the protein environments or by some innate properties of the
substrate(s) themselves. Because of this, our first example in this section concerns