Tuning a P450 Enzyme for Methane Oxidation

Cytochrome P450 (CYP) enzymes are heme-dependent monooxygenases that catalyze the oxidation of C H bonds of endogenous and exogenous organic compounds with formation of the respective alcohols. The mechanism involves the intermediacy of a high-spin oxyferryl porphyrin radical cation which abstracts a hydrogen atom from the substrate, and the short-lived alkyl radical then undergoes C Obond formation. The binding pockets of CYPs are relatively large, therefore small compounds do not have a statistically high enough probability of being properly oriented near the oxyferryl moiety for rapid oxidation to occur; additionally there are other effects that slow down or prevent catalysis. A notorious challenge is the oxidation of methane to methanol by chemical catalysis or using enzymes of the type methane monooxygenases (MMOs). It is not only the smallest alkane, but also has the strongest C H bond (104 kcalmol ). Although CYPs represent a superfamily of monooxygenases, none have been shown to accept methane, whereas MMOs are complex enzymes (many membrane bound) that have not been expressed in heterologous hosts in any significant quantities, among other problems. Herein we show that chemical tuning of a CYP, which is based on guest/host activation using perfluoro carboxylic acids as chemically inert guests, activates the enzyme for oxidation of not only medium-sized alkanes such as n-hexane, but also of small gaseous molecules such as propane and even methane as the ultimate challenge. In the present study we chose, for practical reasons, the enzyme P450 BM3 (CYP102A1) from Bacillus megaterium, which is a self-sufficient fusion protein composed of a P450 monooxygenase and an NADPH diflavin reductase. Several crystal structures of this CYP harboring a fatty acid or fatty acid derived inhibitors, as well in the absence of such compounds have been published. To engineer mutants of P450 BM3 and of other CYPs for enhanced activity and selectivity toward a variety of different compounds, including such difficult substrates as small alkanes, rational design as well as directed evolution have proven to be successful to some extent. For example, P450 BM3 variants characterized by numerous point mutations were obtained in extensive laboratory evolution, and showed for the first time notable activity toward propane by formation of the respective alcohols (2-propanol/1-propanol= 9:1); however, the ethane to ethanol conversion remains problematic and methane oxidation has not been achieved to date. Higher activity in ethane oxidation was accomplished using mutants of P450cam, but here again methane oxidation was not reported. Our chemical approach involves a chemically inert compound that serves as a guest in the binding pocket of P450 BM3, thereby filling the space and reducing the translational freedom of small alkanes or of any other substrate. On the basis of previous reports involving CYPs harboring various substrates, such guest/host interactions can be expected to induce other modes of activation effects as well, specifically water displacement at the Fe/heme site accompanied by a change in the electronic state from the inactive low-spin state to the catalytically active high-spin states. Moreover, many studies have shown that P450 enzymes and mutants thereof can harbor two different substrates simultaneously, thus leading to cooperative effects; one example is lauric acid and palmitic acid in which cooperativity has been demonstrated by isotope labeling experiments. In yet another study regarding the metabolism of bilirubin, the addition of lauric acid or the perfluorinated analogue was reported to facilitate NADPH oxidation and substrate degradation, a finding that has implications for the treatment of jaundice, uroporphyria, and possibly cancer. It has also been shown for the case of a distantly related H2O2dependent P450 enzyme that its peroxidase activity can be influenced by the addition of fatty acids, wherein increased or decreased activity is observed depending upon their chain length. In our endeavor we were guided by the binding mode of the natural substrates, fatty acids, of P450 BM3. The binding includes H-bonds originating from their carboxy function and residues Arg 47 and Tyr 51, as well as hydrophobic interactions. The use of perfluoro carboxylic acids such as 1a–h as chemically inert, yet activating guests was therefore envisioned, because perfluoro alkyl groups are known to be resistant to oxidation while having a hydrophobic character. Moreover, it is known that a CF3 residue is sterically comparable to a CH(CH3)2 group, [11a] which means that a perfluoro fatty acid fills much more space in a P450 binding pocket than a traditional fatty acid, and can additionally induce the crucial low-spin to high-spin conversion of Fe/heme. In exploratory studies, the oxidation of n-octane and n-hexane as well as isomers thereof was studied using P450 [*] Dr. F. E. Zilly, Dr. J. P. Acevedo, Dr. W. Augustyniak, A. Deege, U. W. H usig, Prof. M. T. Reetz Max-Planck-Institut f r Kohlenforschung Kaiser-Wilhelm-Platz 1, 45470 M lheim an der Ruhr (Germany) reetz@mpiso-muelheim.mpg.de