br Cytochrome P monooxygenases Cytochrome P monooxygenases P

Cytochrome P450 monooxygenases Cytochrome P450 monooxygenases (P450s) belong to one of Nature’s most versatile group of enzymes for C–H functionalization and are able to perform challenging regioselective and stereoselective activations of remote, unactivated C(sp3)–H bonds, for which there are only a limited number of industrially viable chemical methods [10, 11, 12]. The exceptional catalytic versatility of P450s has led to numerous studies to alter and adapt the enzyme’s properties by (semi-)rational protein engineering and directed evolution [13, 14, 15]. However, despite decades of research on P450 engineering, inefficient Trichostatin A transfer, uncoupling, low stability and activity have thus far prevented a broad use of P450s in industrial processes other than the small-scale production of drug metabolites [16, 17, 18]. Many engineering efforts have focused on the enzyme P450-BM3, as it is a redox self-sufficient, natural fusion of heme-domain and reductase-domain that exhibits high activity and good expression in different heterologous hosts. Moreover, the enzyme also works as a ‘peroxygenase’, accepting H2O2 as an oxidant without the need for NAD(P)H or redox partners [11,19]. Directed evolution has proven to be a powerful tool to expand the enzyme’s substrate scope from long-chain fatty acids towards small alkanes, terpenes and drug compounds, including alteration or improvement of regioselectivity and stereoselectivity [20, 21, 22] (Figure 1a). A recent example of the biocatalytic potential of such engineered P450s was reported by Loskot et al., who screened a panel of P450-BM3 variants and identified a mutant for the site-selective oxidation of an allylic C–H bond at the late-stage of a natural product total synthesis [23]. Engineering efforts also targeted to address the limitations associated with P450s, in particular the uncoupling of the reaction cycle, which leads to low product formation and enzyme instability. Morlock et al. recently developed a P450 screening assay, which enables the distinction of desired product formation and undesired uncoupling events by the simultaneous detection of NAD(P)H consumption and peroxide formation using horseradish peroxidase (HRP) and AmplifluTM Red [24]. Other groups focused on engineering P450 systems that use alternative electron sources [25]. In the last years, P450 engineering has gained new momentum by the discovery that P450-BM3 can also catalyze a variety of abiological reactions, including carbene and nitrene transfer reactions for cyclopropanations [26], C–H aminations [27], aziridinations [28] and sulfimidations [29] (Figure 1a). Again, P450-BM3 was found to be highly evolvable, and activities as well as stereoselectivities were improved by accumulating beneficial mutations through repeated diversification and screening [7,30]. Further expansion of P450 chemistry has been achieved by optimization of a promiscuous side reactivity of P450-LA1, the anti-markovnikov oxidation of styrene. Through several rounds of random mutagenesis of the heme-binding domain followed by site-saturated mutagenesis of active site residues, a variant was obtained which catalyzes the styrene oxidation with greatly improved activity and selectivity over the kinetically favored epoxidation and even out-competes current small molecule catalysts [].
Peroxygenases The complexities and limitations associated with P450 systems have spurred research on alternative enzyme classes. Besides P450 monooxygenases, many other enzymes are known which catalyze selective oxyfunctionalizations of C–H bonds, including both heme-dependent enzymes and non-heme iron-dependent enzymes, as well as bicoatalysts dependent on copper, pterin and flavin [32]. Particularly flavin-dependent monooxygenases possess diverse oxygenation activities. A well-studied enzyme is p-hydroxybenzoate hydroxylase (PHBH), which catalyzes the regioselective hydroxylation of 4-hydroxy benzoate substrates [33]. While hydroxylation of sp2-centers is most common for this enzyme class, an FMN-dependent monooxygenase exists (LadA) that is able to hydroxylate unactivated C–H bonds of long-chain alkane substrates [34]. Additionally, a subclass of flavin-dependent monooxygenases catalyze the synthetically important Baeyer–Villiger oxidations [32].