The best-studied BVMO is cyclohexanone monooxygenase (CHMO) from Acineto-bacter sp. NCIMB 9871. This type I BVMO was first isolated in 1976  which set off many biocatalytic studies involving this monooxygenase . Over the years a large number of different substrates (>100) have been reported for this bacterial monooxygenase. The enzyme has been used as prototype BVMO to show that these enzymes are not only capable of performing Baeyer-Villiger oxidations but can also oxygenate sulfides, selenides, amines, phosphines, olefins, and iodide-and boron-containing compounds. In a seminal paper in 1982 the first pre-steady-state kinetic analysis of CHMO was described . Using the stopped-flow technique it was shown that CHMO is able to form and stabilize an oxygenated form of the bound FAD cofactor: a C4a-peroxy-FAD (Fig. 3.2). This intermediate is formed after reduction of the flavin cofactor by NADPH, which in turn triggers a fast reaction with molecular oxygen yielding C4a-peroxy-FAD.
It was shown that external addition of hydrogen peroxide to CHMO does not result in formation of this peroxyflavin intermediate . Apparently the enzyme is unable to use hydrogen peroxide and can only form the reactive intermediate by using the NADPH coenzyme and molecular oxygen. Substrate binding has no effect on the formation of the oxygenated flavin cofactor. Such a seemingly uncontrolled consumption of valuable NADPH and formation of a reactive enzyme species is not observed other monooxygenase systems.
Fig. 3.2 Catalytic mechanism of type I BVMOs.
In cytochrome P450 monooxygenases and FAD-containing hydroxylases, formation of the oxygenating enzyme intermediate is tightly coupled to substrate binding, which prevents unproductive coenzyme consumption. In CHMO, such an uncontrolled consumption of NADPH (uncoupling) is largely prevented by the effective stabilization of the peroxyflavin intermediate. In the presence of NADPH and the absence of a suitable substrate only a very slow rate of hydrogen peroxide formation is observed for CHMO. For phenylacetone monooxygenase (PAMO) and 4-hydroxyacetophenone monooxygenase (HAPMO) very slow uncoupling reactions have also been observed, with rates lower than 0.1 s-1 (M.W. Fraaije unpublished results).
The reactive peroxyflavin enzyme intermediate mimics the function of a per-oxyacid as it can perform a nucleophilic attack on the carbonyl carbon of a ketone or aldehyde resulting in a Baeyer-Villiger oxidation. In analogy with peroxyacid-catalyzed Baeyer-Villiger reactions, the reaction would involve formation of a flavin-substrate Criegee intermediate in the active site of CHMO. Decay of this intermediate results in formation of the ester product.
While the oxygenated C4a-peroxy-FAD intermediate has been observed in CHMO [30, 31], no Criegee-intermediate has ever been observed in pre-steady-
state kinetic studies, indicating that this is not a long-lived enzyme species. In case of CHMO it has been found that release of the ester product precedes decay of the hydroxyflavin that is formed upon collapse of the Criegee intermediate  (Fig. 3.2). The catalytic cycle is completed by release of NADP+. A recent kinetic study of CHMO has revealed that release of the oxidized coenzyme, NADP+, is limiting the rate of catalysis . This indicates that the conversion rate is not always determined by the nature of the substrate or product and is in line with the observation that other substrates display similar steady-state kinetic parameters when compared with cyclohexanone. A similar kinetic behavior has been observed for eukaryotic flavin-containing monooxygenases which are distantly related in sequence .
In our laboratory we have observed that HAPMO and PAMO share several kinetic properties with CHMO (unpublished results). With these BVMOs also the peroxyflavin is formed and stabilized in the absence of a substrate, no Criegee intermediate is observed and NADP+ has been found to represent a competitive inhibitor with respect to NADPH. Recently we have been able to demonstrate by using electrospray ionization mass spectrometry (ESI-MS) that the coenzyme NADPH/NADP+ is bound to the enzyme throughout the catalytic cycle .
In 2004 the first crystal structure of a BVMO, PAMO, was published . This BVMO represents a thermostable enzyme as it originates from the mesothermo-philic bacterium Thermobifida fusca . Substrate profiling studies have shown that the enzyme prefers aromatic substrates while it also accepts some aliphatic substrates. The crystal structure of PAMO was solved at 1.7 A resolution and revealed the presence of two dinucleotide binding domains and a helical subdomain (Fig. 3.3). The dinucleotide binding domain formed by residues from the N- and C-termini binds the FAD cofactor in an extended conformation. Such a mode of FAD binding has been observed for many other flavoproteins. The combination of an FAD-binding domain with an NADPH-binding domain is also reminiscent of other FAD-containing oxidoreductases. Contrarily, the helical domain that flanks both cofactor-binding domains and which forms part of the active site appears to be unique to type I BVMOs.
Unfortunately, we have been unable to obtain crystals that contain a bound NADPH/NADP+ or a substrate/product which would reveal the binding mode of these ligands. Nevertheless, binding of NADPH could be easily modeled into the structure using the observed binding modes in other protein structures as a guide. This modeling exercise has suggested that two conserved basic residues, R217 and K336, are crucial for recognition of the adenine and 2'-phosphate moieties of NADPH, respectively.
Elucidation of the structure of PAMO enabled identification of an active site residue that appears to be key to catalysis in type I BVMOs: R337 (Fig. 3.3). R337 is located at the re side of the flavin and points towards the reactive part of the FAD cofactor. This residue is strictly conserved in all type I BVMOs. Replacing R337 in PAMO (unpublished results) and HAPMO by an alanine yielded inactive fad cofactor
binding domair binding domair fad cofactor
binding domair binding domair
enzymes . Based on the structure of PAMO we have proposed that the positive charge of this arginine may be crucial (a) to stabilize the negatively charged per-oxy-FAD intermediate and/or (b) to promote and stabilize the negatively charged Criegee intermediate (Fig. 3.2). The delicate function of this active site residue is also apparent from the absolute strict conservation of this residue within all identified type I BVMO sequences. It is also striking to note that this active site arginine is the only active site residue that is conserved among type I BVMOs. The extensive variation of active site residues might reflect the plasticity in substrate specificities observed for these monooxygenases.
As no structure is available of PAMO with a substrate or product bound in the active site, it is difficult to assign residues that form the substrate-binding pocket. In a recent enzyme redesign study it was demonstrated that by deleting two residues in a loop region that is close to R337 the substrate specificity of PAMO can be significantly altered . While PAMO does not accept 2-phenylcyclohexanone, the deletion mutant AA441/L442 was found to convert this relatively bulky substrate. It is envisaged that by shortening the loop interacting with R337, the substrate-binding pocket is enlarged and can accommodate bulky substrates (Fig. 3.3). A role of the targeted loop region in substrate binding was substantiated by results obtained from a directed evolution study with CHMO. Screening a library of CHMO mutants created by error-prone PCR resulted in the discovery of three single mutants that displayed significantly altered enantioselective behavior [38, 39]. Locating the analogous residues in the PAMO structure reveals that one of them (L443) is part of the abovementioned loop region while the others, Q152 and T393, are in close proximity (Fig. 3.3).
These findings confirm that substrates bind at the re side of the flavin in PAMO. A similar situation will hold true for other type I BVMOs as the si side of the flavin in PAMO is occupied by the protein and the corresponding residues are highly conserved among type I BVMOs.
Was this article helpful?