The biotechnological advantages offered by the fusion of P450s to redox partner enzymes appear clear from studies on the P450BM-3 system. BM-3 has an efficient electron transport chain that enables fatty acid hydroxylation at rates ~100-fold faster than many eukaryotic P450s. In addition, it is catalytically self-sufficient, obviating the requirement for isolation of 2-3 different enzymes and determining ratios of redox partner-to-P450 that give optimal turnover [14, 72]. The same advantages are inherent in the A2/A3 systems and the other members of this class. While work is in its infancy for the CYP116B1/B2 systems, these are also catalytically self-sufficient and are likely to offer similar advantages in terms of efficient electron transfer apparatus. There are already several examples of the engineering of the P450BM-3 enzyme to enable its production of industrially relevant molecules. These include oxygenation of short-chain alkanoic acids, highly branched chain fatty acids (leading to chiral precursors for polyketide synthesis), and (+)-valencene (to produce the grapefruit flavor compound (+)-nootkatone) [109-111].
CYP102A3 has also been engineered to enable enhanced production of 1-octanol from octane . Virtually all efforts at engineering the BM-3 enzyme have been directed at altering the structure of the heme domain to effect changes in substrate selectivity and product formation. However, there are also some serious issues to be addressed with respect to the performance of the reductase domain, as discussed below in Section 5.6.
In the last few years (and as a direct result of the information generated from genome sequencing programs) new types of P450-redox partner fusions have been recognized, and a number have been expressed and characterized at the protein level. In Methylococcus capsulatus a P450 (N-terminal)-ferredoxin fusion protein occurs, and the P450 is clearly a member of the sterol demethylase (CYP51) family. The ferredoxin is likely to bind a 3Fe-4S cluster  (Fig. 5.3c). It appears likely that this system has evolved to fuse the P450's cognate ferredoxin, and now requires only a separate ferredoxin reductase to complete the novel form of a class I P450 redox system. An analogous system was recognized in the bacterium Rhodococcus rhodochrous (strain Y-11), in which a flavodoxin (N-terminal) is fused to a soluble P450 (Fig. 5.3d). The enzyme (XplA) was shown to degrade the explosive RDX (Royal Demolition eXplosive - the molecule hexahydro-1,3,5-trinitro-1,3,5-triazine) - although it is currently unclear whether the process involves oxygenase activity of the P450, or simply substrate reduction . As with the M. capsulatus enzyme, a reductase enzyme is clearly required to complete the Rh. rhodochrous P450-redox system, and work on this explosive degrading enzyme is clearly aimed at producing a plant-based system for degrading RDX in the soil (phytoremediation).
While not strictly a cytochrome P450, the similarity of the multidomain eu-karyotic nitric oxide synthase (NOS) enzymes to P450BM-3 (and its homologs) cannot be overlooked in the context of oxygenase fusion enzymes. The major mammalian NOS isoforms (neuronal, endothelial, and inducible) all have a thio-late-coordinated heme enzyme (N-terminal) fused to a CPR module . The enzymes are functional as dimers (as P450BM-3 is in fatty acid hydroxylation), and electron transfer occurs between the reductase of monomer 1 and the heme domain of monomer 2 in the dimer [88, 115]. The similarities between NOS enzymes and the BM-3 class of P450s extend to the fact that NOS enzymes perform hydroxylation/oxygenation chemistry. They hydroxylate the substrate L-arginine to N-hydroxy-L-arginine, and then (in a second reaction on this product) go on to form L-citrulline and nitric oxide (NO) at the heme center. However, there are important differences between NOS and BM-3 (and other P450s). The protein effector calmodulin plays important regulatory roles in the control of electron transfer in NOS, and the cofactor tetrahydrobiopterin (H4B) is bound close to the heme and also participates in catalytic redox reactions . In addition, while the reductase module is structurally closely related to rat CPR, the oxygenase domain has a considerably different structure to the P450s, with a much more solvent exposed active site (Fig. 5.5).
Notwithstanding the important differences between the NOS and P450-CPR structures, attempts have been made to create heterologous fusion proteins involving the different domains. In studies of fusions between BM-3 and rat neuronal NOS, catalytically functional chimeras were generated - with the most successful fusion being that between the BM-3 reductase domain and the nNOS heme domain, which catalyzed NO production quite efficiently . Numerous attempts have also been made to mimic the natural BM-3-type P450 fusion systems by fusing eukaryotic P450s with CPR. Frequently, some enhancement of activity is obtained - although no fusions created to date have produced chimeras with activity levels that approach those of P450BM-3 or its A2/A3 relatives. Examples of stable, active P450-CPR fusions reported include ones involving the mammalian P450s CYP3A4 and CYP17A [117, 118]. Other interesting fusion proteins generated included three-protein chimeras involving P450cam and its PdR and Pd redox partners. The most efficient chimera created (a PdR-Pd-P450 chimera from N- to C-terminus) had a kcat for camphor hydroxylation of ~30 min-1, around 100-fold lower than can be obtained by reconstituting the isolated redox partner proteins .
Thus, artificial P450-redox partner fusion protein constructs have provided useful tools to enable production of single-component, catalytically self-sufficient
Fig. 5.5. Atomic structures of the oxygenase and reductase domains of eukaryotic nitric oxide synthase. (a) Atomic structure of murine iNOS bound to L-arginine (PDB code 1DWV) . The distinctly different protein fold to those of the P450s is obvious from comparison with the structures shown in Figs 5.1 and 5.3. The relatively exposed heme is shown at the center of the structure in red spacefill. L-Arginine is shown in colored ball and stick representation above the plane of the heme. (b) Atomic structure of rat neuronal NOS reductase. The overall structure shows strong similarities to that of rat CPR . Domains of the NOS reductase protein are color coded as follows: NADPH-binding domain in red; FAD-binding domain
in dark blue; alpha helical connecting domain (CD, which orientates the two flavin cofactors) in cyan; and FMN-binding domain in orange. Other prominent features are the autoinhibitory helix (purple) at the bottom left of the structure that is located in the FMN-binding domain and which interferes with CaM binding and likely modulates both intra and inter-domain electron transfer; and the P-finger/CD2A section (green beta sheet section and associated yellow loop region) extending from the CD and which also influences regulation by Ca2+/CaM. The NADP+, FAD and FMN cofactors are shown in green, orange, and yellow stick representation .
protein entities, and have potential applications in biotechnology. However, their activity levels rarely exceed those achieved by reconstituting their separate component enzymes, and thus do little to explain the efficiency of the electron transfer systems seen in the "natural" (e.g. P450BM-3) P450-redox partner fusion enzymes.
Was this article helpful?