Sebastian
Bormann†
a,
Alvaro
Gomez Baraibar†
b,
Yan
Ni†
b,
Dirk
Holtmann
*a and
Frank
Hollmann
*b
aDECHEMA Research Institute, Theodor-Heuss-Allee 25, 60486 Frankfurt am Main, Germany. E-mail: holtmann@dechema.de
bDelft University of Technology, Department of Biotechnology, Julianalaan 136, 2628 BL Delft, The Netherlands. E-mail: f.hollmann@tudelft.nl
First published on 9th January 2015
Peroxygenases enable H2O2-driven oxyfunctionalisation of organic compounds such as selective epoxidation, hydroxylation and heteroatom oxygenation. Therefore, peroxygenases are potentially very useful tools for the organic chemist. This contribution reviews the current state of the art in peroxygenases, their availability, engineering and reaction scope. Current challenges and possible solutions are critically discussed.
Sebastian Bormann† | Sebastian Bormann was born in Frankfurt, Germany in 1988. He received his Bachelor's degree in biotechnology from University of Applied Sciences Darmstadt in 2011 and his Master's degree in biotechnology from Braunschweig University of Technology in 2014. Currently, he is a PhD student in the biochemical engineering group of the DECHEMA Research Institute in Frankfurt, Germany. His research is focused on protein engineering of peroxygenases and biocatalysis using such enzymes. |
Alvaro Gomez Baraibar† | Alvaro Gomez Baraibar received his Master's in Biotechnology from the Technical University of Denmark in 2010, with a project dealing with transaminase catalysed reactions. He performed his PhD research at the Research Center Juelich (Germany) on enzymatic process development and characterization of carboligases. In his postdoctoral project in the Biocatalysis group at the Delft University of Technology (Netherlands) he focusses on enzymatic cascades, oxidative enzymes and photocatalysis. |
Yan Ni† | Yan Ni was born in 1986 in Shanghai, China. She obtained her PhD in biochemical engineering in 2013 from the East China University of Science and Technology. Since January 2014 she has been working as a postdoctoral fellow with Dr. Frank Hollmann in the Biocatalysis group at the Delft University of Technology (Netherlands). Her current research focuses on new and efficient (cofactor) regeneration approaches to promote enzymatic reductions, oxidations and specific oxyfunctionalisations. |
However, the rather complicated molecular architecture of P450 monooxygenases represents a significant challenge with respect to practical application. In P450 monooxygenases the catalytically active compound I is obtained via a sequence of electron transfer to the resting heme group, oxygen binding and further reduction followed by water elimination.7 The reducing equivalents needed for this reaction are delivered from reduced nicotinamide cofactors (NAD(P)H) to the enzymes through sometimes very complicated electron transport chains (Scheme 2, upper). Obviously, such complicated electron transport chains can be very difficult to optimise (both in vivo and in vitro).
An alternative class of enzymes that circumvents the issues described above and therefore is highly interesting for practical application are the so-called peroxygenases (E.C. 1.11.2).9 These enzymes are structurally related to the aforementioned P450 monooxygenases insofar as they also contain the heme prosthetic group coordinated by a cysteine ligand and therefore, in principle, give access to the same enormous range of reactions and products. In contrast to P450 monooxygenases peroxygenases do not catalyse the reductive activation of molecular oxygen but rather directly use hydrogen peroxide to form the catalytically active oxyferryl species (Scheme 2, lower).
It is important to point out the difference between peroxygenases and peroxidases as well as haloperoxidases originating from different reaction pathways of the oxyferryl species (compound I, Scheme 3) Depending on the enzyme, compound I can return into the resting state by various reactions. The so-called haloperoxidases mediate the oxygenation of halogenide ions (Cl−, Br−, I−) to the corresponding hypohalites (Scheme 3). Peroxidases react by two successive H-atom abstraction reactions (predominantly on phenolic substrates) yielding phenoxyradicals (Scheme 3). Both reactions produce reactive species (hypohalites and radicals) that mostly react further outside of the enzymes' active sites. As a consequence, the selectivity of the subsequent reactions generally is dictated by the chemical reactivity of the reactive species and the starting materials. In contrast, peroxygenases utilize compound I for H-atom abstraction and fast recombination forming new C–O-bonds within the enzyme active site (Scheme 3). As a consequence, the three-dimensional structure of the enzymes' active sites can exert control over the selectivity of the reaction.
Scheme 3 Reactions of compound I in the haloperoxidase pathway (upper), peroxidase pathway (middle) and the peroxygenase pathway (lower). |
Structurally, the main difference between peroxidases and peroxygenases appears to be the proximal Fe-ligand: a cysteine ligand (such as in P450 monooxygenases) is crucial for peroxygenase activity (not excluding peroxidase activity of these enzymes) whereas a histidine-ligand is found in peroxidases (only capable of H-atom reactions). It is worth mentioning here that haloperoxidases, in addition the heme-dependent ones, also comprise V-dependent enzymes, which have been discussed elsewhere.10–21
The catalytic cycle of peroxygenases has been studied extensively (Scheme 3).22 In essence it is similar to the so-called peroxide shunt of P450 monooxygenases. Specifically, the oxoferryl porphyrin cation radical intermediate (compound I) is formed from reaction of the resting heme iron with H2O2via heterolytic cleavage of the short lived ferric hydroperoxo intermediate compound 0 (Fe–OOH−).23 At the distal side of the heme, glutamic acid acts as acid–base catalyst promoting compound I formation.24–27 One electron transfer and protonation of the basic oxyferryl intermediate then affords the hydroxyferryl species (compound II) that hydroxylates the substrate radical.28–31 The basicity of the oxo ligand has been described for the chloroperoxidase from Caldariomyces fumago (CfuCPO) and has recently also been elucidated for P450 monooxygenases.28,30 Together with the redox-potential, the high pKa of compound II is the main driving force for oxidation of C–H bonds with high bond-dissociation energies. The heme cysteinate proximal ligand, which is also characteristic for P450 monooxygenases, is believed to increase the basicity of compound II due to electron ‘push’. It is also this high pKa of oxygenating enzymes that allows relatively moderate redox potentials which prevents oxidative self-destruction.30,32
As classified by EC, enzymes with a primary peroxygenase activity are summarized under EC 1.11.2.x which contains unspecific peroxygenase (UPO, EC 1.11.2.1, formerly aromatic peroxygenase), plant seed peroxygenase (EC 1.11.2.3), fatty-acid peroxygenase (EC 1.11.2.4), and myeloperoxidase (EC 1.11.2.2). The latter shows only little peroxygenation activity and is mainly grouped in this sub-class for historical reasons.33 Fatty-acid peroxygenase type enzymes (Cytochrome P450BSβ) are cytosolic bacterial peroxygenases that hydroxylate fatty acid substrates at the α- or β-position. Moreover, they carry out aromatic hydroxylations in the presence of short-chain fatty which supposedly act as acid–base catalysts.34 Plant seed peroxygenase is a membrane-bound heme-thiolate protein with a caleosin-like calcium binding motif that catalyses the oxygenation of unsaturated fatty acids using organic hydroperoxides as cosubstrate.35
Following Hager's seminal publications on the chloroperoxidase from Caldariomyces fumago (CfuCPO) in the 1960ies36 the first wave of exploration of the scope of this and related enzymes generated a lot of promise for organic synthesis. In 2000 van Rantwijk and Sheldon summarised the state of the art then.37 Unfortunately however, the promise of a practical peroxidase-based oxyfunctionalisation chemistry was not fulfilled ever since then. Possibly this is due to the relative poor reactivity of CfuCPO with (non)activated C–H-bonds (vide infra) but certainly also due to difficulties in (heterologous) expression of this enzyme and the resulting high price.
More recently, Hofrichter and coworkers have reported novel peroxygenases with potentially much higher impact in preparative organic synthesis.38–41 The same group have also recently reviewed the phylogenetic relations of peroxygenases33,42 as well as mechanistic aspects.22
In this respect it is interesting to note that N-alkylated flavins can react with H2O2 forming highly reactive flavin(hydro)peroxides capable of certain oxyfunctionalisation reactions (sulfoxidations, epoxidations, Baeyer–Villiger-Oxidations etc.). Recently, Fraaije and coworkers have demonstrated that riboflavin-binding proteins can also accommodate N-alkylated flavins resulting in artificial flavo-peroxygenases.43,44 Very promising first results were reported but further examples e.g. with engineered hybrid enzymes45 will be necessary to assess the full scope of this approach.
The aim of the current contribution therefore was to review the current state of the art in practical application of peroxygenases.
Organism | Enzyme | Fermentation scale | Max. enzyme concentration [mg L−1] | Ref. |
---|---|---|---|---|
SF: shake flask, BR: bioreactor.a Calculated from volumetric activities based on a specific activity of 1400 U mg−1. | ||||
Caldariomyces. fumago | CfuCPO | SF (0.3 L); BR (1 L) | 715–2310a | 52 and 53 |
Agrocybe aegerita | AaeUPO | BR (4 L) | 9 | 41 |
Marasmius rotula | MroUPO | SF (0.2 L) | 445 | 54 |
Coprinus radians | CraUPO | SF (0.2 L) | 7 | 39 |
It has not been conclusively elucidated why these heme-thiolate peroxygenases can only be heterologously expressed at relatively limited titers. Nonetheless, the increased interest in fungal peroxygenases due to the relatively novel class of unspecific peroxygenases will probably result in more efficient heterologous expression systems. This might also lead to broader availability of these enzymes since to date only CfuCPO is commercially available.
Issues to be resolved are (1) the sometimes limited enzyme activity and stability under process conditions and (2) the frequently very low reactant concentrations.
Limited enzyme activity and stability are frequently mentioned as factors limiting the practical usefulness of enzymes. The question however is, how to quantify these values and how to define efficient catalysts from less efficient ones. An enzyme's specific activity can be quantified best in terms of its specific activity (determined in units per milligram) or as turnover frequency (TF, essentially moles of product formed per mole of catalyst per time unit). Obviously, both numbers depend on the reaction conditions and the substrate used therefore.
The question about ‘efficiency’ is much harder to answer. Depending on an author's background and experience, efficiency can mean very different things. From an economic point of view efficiency may be seen as the costs related to the catalyst, i.e. the cost contribution of the catalyst to the overall production costs. In a recent seminal review Tufvesson et al. have provided guidelines to assess such numbers.56 Others have suggested similar numbers for economic evaluation.57 Even though the numbers shown in Table 3 should be seen as rule of thumb they give an impression about the economic feasibility of a given enzyme reaction. Apparently, other factors such as space time yields, downstream processing etc. are at least as important as the catalyst contribution. Also, a seemingly inefficient reaction (in terms of catalyst cost contribution) may lead to very significant simplifications in DSP or give access to cheaper feedstock resulting in an overall more efficient process.
Product price range [Euro kgProduct−1] | Allowable cost contribution of the catalyst [Euro kgProduct−1] | Minimal productivitya [kgProduct kgCatalyst−1] | Minimal productivityb [TTN] | |
---|---|---|---|---|
a Assuming an average catalyst cost of 200 Euro kg−1 (for crude enzyme preparations). b Assuming an enzyme molecular weight of 40 kDa and a product molecular weight of 200 gmol−1. | ||||
Pharma | >100 | 10 | 20 | 4000 |
Fine chemical | >15 | 1.5 | 133 | 26666 |
Speciality chemical | 5 | 0.25 | 800 | 160000 |
Bulk chemical | 1 | 0.05 | 4000 | 800000 |
The two main goals of peroxygenase engineering can be divided between reactivity and stability. Although fungal peroxygenases are dependent on H2O2, they are readily inactivated at elevated peroxide concentrations. While the replacement of easily oxidizable amino acids is a rational strategy for improved oxidative stability, random mutagenesis has been shown to generate further improvements elusive to rational design.62 Other obvious targets for protein engineering include enhanced stability at elevated temperatures and stability in the presence of organic solvents which are often necessary to solubilize highly polar substrates of peroxygenases in aqueous enzymatic reaction systems.
Desired changes concerning the reactivity of peroxygenases focus on the improvement of peroxygenation activity and the concomitant reduction of undesirable one-electron oxidation activity. Structural improvements to increase substrate spectrum and stereospecificity of peroxygenases are also highly desirable.
It has been shown that directed evolution can generate enzyme variants with more than 100-fold improved oxidative and temperature stability in terms of a fungal peroxidase.63 Directed evolution of the peroxygenase AaeUPO resulted in increased enzyme titers, improved catalytic activity (18-fold increase in kcat/Km), and enhanced operational stability in the presence of organic solvents.48,55
Overall, we believe that the benefits and disadvantages of immobilisation have to be carefully evaluated. On the one hand immobilised enzymes tend to be more robust to demanding reaction conditions (mechanical, thermal and solvent-induced stress impairing the tertiary and quarternary structure of the biocatalyst). Furthermore, immobilisation is currently the method of choice for the application of peroxygenases in organic solvents as methods to solubilise peroxygenases in non-aqueous reaction media are scarce.75 On the other hand, immobilisation requires additional efforts in catalyst preparation going hand in hand with increased costs.56 Furthermore, the frequently observed kinetic limitations of immobilised catalysts further reduce their economic and practical attractiveness.
One of the major limitations of biocatalysis in general lies with the very poor water solubility of many (hydrophobic) reactants of interest. Unfortunately academic researchers tend to circumvent this limitation by adjusting the substrate loading of their reactions to the aqueous solubility of the reagents. The consequences of using such dilute reaction mixtures are:
(1) Poor practical relevance as dilute reaction mixtures are generally economically not feasible (inefficient use of reactor space) and because downstream processing efforts correlate to the volumes of reaction mixtures that have to be handled.
(2) Poor environmental impact as huge volumes of contaminated aqueous waste streams have to be dealt with. As a rule of thumb a 10 mM product solution produces approx. 500 kg of waste water per kg of product (of course depending on the molecular weight of the product).4
Therefore, one major line of research aiming at making peroxygenase-catalysed oxyfunctionalisation reactions more attractive should lie on increasing the overall reactant concentration! Apparently, using water-miscible organic cosolvents represents an interesting step in the right direction. Actually most of the contributions mentioned below make use of such cosolvents to some extent. For example ionic liquids have gained some attention as cosolvent.79–81 Also tertiary butanol has received significant attention as cosolvent and protectant against oxidative stress.82
From a substrate loading point of view (almost) non aqueous reaction media represent the most desirable system. Several groups have demonstrated that various peroxidases are active under non-aqueous reaction conditions. For example van Rantwijk, Sheldon and coworkers have applied immobilised CfuCPO in various organic solvents.83 The specific activity of the immobilised enzyme roughly correlated with the polarity of the solvent applied but no general conclusions could be drawn on the overall performance (in terms of TNs) underlining that our general understanding of enzyme stability in the presence of organic solvents still needs to be improved.84 Klibanov and coworkers have studied the effect of different solvents on HRP-catalysed sulfoxidation of thioanisole demonstrating dramatic effects of the solvent on activity and selectivity of the enzyme.85–88 Using carefully hydrated enzyme preparations Stamatis and coworkers could efficiently use CfuCPO for oxidation/oxyfunctionalisation reactions in organic solvents.89
Two liquid phase systems (2LPSs) represent an interesting compromise between pure aqueous and pure organic solvent-based reaction media (Scheme 4).
Such systems generally comprise a hydrophobic, water-inmixible organic phase, which due to its hydrophobicity serves as substrate reservoir and product-sink. Depending on the partitioning coefficient of the reagents the in situ concentration of the reagents in the aqueous, enzyme-containing layer can be controlled and thereby oxidative stress and inhibitory effects of substrate and product be controlled. For example supercritical CO2 (ref. 90) or 1-octanol91 or the substrate itself92 have been reported.
Also the use of organic hydroperoxides exhibiting a lower reactivity compared to H2O2 could not really establish itself as a practical method.71,83,95–103
Today's state-of-the-art in providing peroxygenases with suitable amounts of H2O2 to enable high reactivity while minimizing their oxidative inactivation is to generate H2O2in situ by partial reduction of molecular oxygen. The reducing equivalents needed for this reaction can be derived from different sources.
Oxidases have been considered very early as co-catalysts to provide peroxygenases with H2O2. Especially the compatibility of both catalysts and the homogeneous nature of the system (avoiding diffusion limitations) makes this approach attractive and popular. Recently Fraaije and coworkers reported an interesting approach of expressing a fusion protein containing oxidase and peroxidase.104 Certainly the most widespread approach is to use glucose oxidase-catalysed oxidation of glucose (Scheme 5).37,105–109
Scheme 5 Glucose oxidase (GOx)-catalysed in situ formation of H2O2 to promote peroxygenase-catalysed oxyfunctionalisations. |
Both, the starting material and the catalyst are readily available at very reasonable prices and all reagents are unproblematic. The formation of gluconic acid represents the major disadvantage of this system due to the concomitant pH-change and the additional pH-control needed at larger scale. A further issue, that will probably prevent applications of this system at scale is the high viscosity of concentrated glucose solutions. Furthermore, use of glucose as two-electron donor represents a rather wasteful process (generation of 198 g of by product per mol of H2O2) and thereby does not really comply with the principles of Green Chemistry.110,111 Also ethical issues arise considering that glucose is edible.
Primary alcohols (such as ethanol) represents a somewhat more efficient process (44 g by product per mol H2O2) but is hampered by the stoichiometric accumulation of aldehydes, which is problematic from an enzyme stability- and environmental point-of-view.112,113
Also amino acid oxidases are worth mentioning as in situ H2O2 generation catalysts.114 Very similar to the above-mentioned GOx-scenario, the reagents (amino acids as staring material and α-ketoacid products) are cheap, readily applicable and unproblematic. Interestingly, enough, this approach has not been pursued widely, yet.
Electrochemical reduction of molecular oxygen, from an environmental point-of-view, certainly represents one of the cleanest processes to provide peroxygenases with H2O2 (Scheme 6). Especially if the electrical power is obtained from renewable resources, electrochemical H2O2-generation can be considered to be essentially waste-free. Maybe due to the extra equipment needed or due to a bias amongst some researchers, this promising method is developed and applied by a limited number of research groups.79,115–120
Scheme 6 Electrochemical in situ formation of H2O2 to promote peroxygenase-catalysed oxyfunctionalisations. |
Leitner and coworkers90 proposed partial hydrogenation of O2 using Pd-nanoparticle catalysts in an autoclave setting. Even though this method is also appealing from an atom economy point-of-view, issues of equipment (especially to prevent Knallgas reactions) remain to be solved.
We have demonstrated that photocatalytic reduction of O2 (using visible light and simple flavin photocatalysts) may be a viable method for in situ generation of H2O2 (Scheme 7).92,121–123 ETDA however, despite its simplicity to use, yields undesirable formaldehyde as by-product necessitating further improvements. Other sacrificial electron donors such as formate or simple amino acids may represent interesting alternatives to EDTA.
Scheme 7 Photochemical in situ formation of H2O2 to promote peroxygenase-catalysed oxyfunctionalisations. |
Finally, also simple and mild reducing agents such as Hantzsch esters or nicotinamide derivatives (Scheme 8)124,125 or organometallic hydrides126 may be interesting methods to provide peroxygenases with suitable amounts of H2O2.
Scheme 8 Chemical in situ formation of H2O2 to promote peroxygenase-catalysed oxyfunctionalisations. |
A finite judgment on which in situ H2O2 generation system is ‘the best’ is not yet (and probably will never be) possible. Every system exhibits specific advantages and disadvantages; but the variety offers the practitioner sufficient choices for the specific task given. One criterion may be the turnover number (TN) achievable with the single systems. Table 4 compares the reported TNs of some systems.
Mode of H2O2 addition | TN | STY (g L−1 d−1) | ee (%) | Ref. | |
---|---|---|---|---|---|
CfuCPO | Co-catalyst | ||||
Step-wise addition | 41000 | — | 0.7 | 99 | 127 |
Continuous addition | 109000 | — | 84 | 99 | 82 |
GOx/glucose/O2 | 250000 | n.d. (GOD) | 13 | 99 | 108 |
Pd/H2/O2 | 6120 | 12.8 (Pd) | 12–67 | 80–94 | 90 |
Cathode/O2 | 95000 | — | 30 | 98.5 | 128 |
3D electrode/O2 | 145000 | — | 104 | 99.4 | 129 |
Flavin/EDTA/hv/O2 | 22400 | 1250 (FMN) | 42 | >98 | 121 |
One major challenge of all methods mentioned above lies with the very poor solubility of molecular oxygen in aqueous media (as a rule of thumb being ca. 0.25 mM at room temperature and atmospheric pressure). Hence, the dissolved oxygen is generally depleted very quickly and diffusion of O2 into the reaction medium becomes overall rate-limiting.
Enzymea | TNb | TFc [min−1] | STYd [g L−1 d−1] | ee [%] | Ref. |
---|---|---|---|---|---|
a CfuCPO: chloroperoxidase from C. fumago; V-phytase: vanadate-loaded phytase (from A. ficuum); VBPO: vanadium bromoperoxidase from A. nodosum or C. pilulifera; HRP: horseradish peroxidase; MyG: myoglobin mutant; Lip: lignin peroxidase from P. chrysosporium; TfuDyP: DyP-type peroxidase from T. fusca. b TN: turnover number (moles (product) × moles (catalyst)−1). c TF: turnover frequency (TN × (reaction time)−1). d STY: space time yield. | |||||
CfuCPO | 109000 | 905 | 84 | 99 (R) | 82 |
V-phytase | 25000 | 5.5 | 5.3 | 68 (S) | 130 |
VBPO | 750 | 0.4 | 0.8 | 85 (R) | 19 |
VBPO | 180 | 0.2 | 0.03 | 55 (S) | 19 |
HRP | 100 | 1.7 | 3.3 | 60 (S) | 131 |
MyG | n.d. | 5.5 | n.d. | 97 (R) | 132 |
Lip | 270 | 4.5 | 0.5 | 37 (S) | 133 |
TfuDyP | 312 | 0.1 | 0.2 | 61 (R) | 134 |
Product | n, Ra | TN | ee (%) | Ref. |
---|---|---|---|---|
a n, R refer to the product structures given in the table body. | ||||
n = 0 | 109000 | 99 (R) | 82 | |
n = 1 | 90000 | 99 (R) | 82 | |
n = 2 | 3300 | 27 (R) | 82 | |
CH3 | 90000 | 99 (R) | 82 | |
OCH3 | 58000 | 99 (R) | 82 | |
Cl | 85000 | 99 (R) | 82 | |
Br | 16000 | 99 (R) | 82 | |
NO2 | 21000 | 99 (R) | 82 | |
— | 3300 | 99 (R) | 82 | |
— | 40000 | 99 (R) | 82 | |
— | 109000 | 99 (R) | 82 | |
R = cyclopentyl; R′ = CH3 | 63000 | >98 (R) | 135 | |
R = cyclohexyl; R′ = CH3 | 53000 | 85 (R) | 135 | |
R = ipropyl; R′ = CH3 | 63000 | >98 (R) | 135 | |
R = tbutyl; R′ = CH3 | 50000 | 85 (R) | 135 | |
R = tbutyl; R′ = C2H5 | 19000 | 35 (R) | 135 | |
R = C5H11; R′ = CH3 | 47000 | >98 (R) | 135 | |
R = C8H17; R′ = CH3 | 25000 | 54 (R) | 135 | |
n = 1 | 44000 | 99 (R) | 136 | |
n = 2 | 4400 | 96 (R) | 136 | |
n = 0 | 36000 | 99 (R) | 137 | |
n = 1 | 36000 | 99 (R) | 137 | |
n = 2 | 9300 | 99 (R) | 137 | |
— | 16000 | 95 (R) | 137 | |
— | 36000 | 95 (R) | 137 | |
— | 7200 | 95 (R) | 137 | |
— | 50 | 82 (R) | 138 |
One feature of CPO-catalysed sulfoxidation is that overoxidation to the corresponding sulfone, which is frequently an issue with chemical sulfoxidation reactions,139 is virtually never observed.37 The substrate scope is fairly broad with sometimes excellent enantioselectivity. Also the catalytic efficiency, expressed as turnover number varies depending on the substrate used. It can, however, be asserted that TNs of CPO in sulfoxidation generally are significantly higher than in other oxygenation reactions (vide infra).
Product | n | TN | Selectivityb (%) | ee (%) | Ref. |
---|---|---|---|---|---|
a n refers to the product structures given in the table body. b Selectivity: percentage of epoxide of the total product; generally, allylic hydroxylation represents the major side reaction of epoxidation reactions. | |||||
— | 3400 | 100 | 70 (R) | 140 | |
n = 1 | 5000 | 61 | 62 (R) | 141 | |
n = 2 | 5000 | 93 | 88 (R) | 141 | |
n = 3 | 3250 | 89 | 95 (R) | 141 | |
— | 840 | 71 | 46 (R) | 141 | |
— | 4200 | 100 | 94 (R) | 140 | |
— | 4250 | 100 | 94 (R) | 142 | |
— | 2100 | 100 | 95 (R) | 143 | |
— | 34 | 100 | 10 (S) | 143 |
In contrast, the newly discovered AaeUPO (Table 8) already now shows significantly higher TTNs than CfuCPO. The enantioselectivity of both enzymes appears to be very similar. In those cases where allylic C–H bonds are present in the starting material, epoxidation competes with hydroxylation resulting in sometimes very complex product mixtures.122,144,145
Product | n, Ra | TN | Selectivityb (%) | ee (%) | References |
---|---|---|---|---|---|
a n, R refer to the product structures given in the table body. b Selectivity: percentage of epoxide of the total product; generally, allylic hydroxylation represents the major side reaction of epoxidation reactions. | |||||
R = Me | 308 | 100 | n.d. | 144 | |
R = n-C5H11 | 26 | 85 | 72 (S) | 144 | |
R = n-C6H13 | 511 | 55 | 66 (S) | 144 | |
5000 | 100 | — | 144 | ||
5000 | 70 | n.d | 144 | ||
— | 2630 | 85 | n.d. | 144 | |
— | 6174 | 100 | 7 (R) | 144 | |
— | 10600 | 100 | >99 (1R, 2S) | 145 | |
— | 10700 | 100 | 29 (n.d.) | 145 | |
n = 1 | 10700 | 100 | 2.3 (1R, 2S) | 145 | |
n = 2 | 10600 | 100 | 32 (n.d.) | 145 |
One of the earliest reactions examined for AaeUPO was the hydroxylation of aromatic compounds. Detailed mechanistic studies revealed that the apparent hydroxylation in fact is an epoxidation followed by spontaneous protonation/rearrangement (Scheme 9).
Scheme 9 Aromatic hydroxylation mediated by AaeUPO as a sequence of epoxidation and spontaneous isomerization.146 |
So far, a rather limited amount of aromatic compounds (amongst them many pharmaceutical ingredients) have been evaluated (Table 9). It is interesting to note that in case of alkyl substituted arenes different selectivity patterns have been observed for different peroxygenases. For example, the UPO from Coprinellus radians exhibits preference for benzylic C–H-bonds whereas AaeUPO appears to be selective for aromatic hydroxylation (via epoxidation as described above). Further systematic investigations will be necessary to fully understand (and predict) the UPOs' selectivities; but we believe that a very powerful toolkit for organic synthesis may result here!
The selective hydroxylation of sp3-C–H bonds has been in focus already during the ‘CfuCPO era’ (Table 10). Many research groups have evaluated CfuCPO as catalyst for the hydroxylation of activated benzylic, allylic and propargylic C–H bonds. However, as shown in Table 10, the catalytic efficiency of CfuCPO (in terms of turnover numbers) falls back by some orders of magnitude behind preparative usefulness. Functionalisation of non-activated C–H-bonds such as in linear and cyclic alkanes appears to be out of scope for CfuCPO. Therefore, just as in case of epoxidation, CfuCPO does not seem to be a very useful catalyst for C–H-hydroxylation. However, a different picture evolves with the newly discovered UPOs such as AaeUPO. Already at the present, very early stage of development, this enzyme shows TNs that appear to reach synthetically useful levels in optimised reaction systems.
Product | n, R | TN | ee (%) | Ref. |
---|---|---|---|---|
AaeAPO | ||||
R = CH3 | 10600 | >99 | 145 | |
R = CH2CH3 | 7100 | >99 | 145 | |
R = (CH2)2CH3 | 5800 | 40 | 145 | |
n = 1 | 8600 | 87 | 145 | |
n = 2 | 9400 | >99 | 145 | |
— | 22000 | n.d | 152 | |
— | 1882 | 62.5 (R) | 153 | |
17000 | — | 122 | ||
CfuCPO | ||||
— | 2100 | — | 154 | |
— | 12000 | — | 91 | |
22000 | — | 91 | ||
R = CH3 | 420 | 97 (R) | 154 | |
R = (CH2)CH3 | 420 | 88 (S) | 154 | |
— | 590 | 94 (n.d) | 155 | |
— | 32000 | — | 156 | |
— | 394 | — | 157 | |
— | 960 | — | 71 |
In addition to CC-double bonds also activated aromatic systems such as phenols undergo electrophilic attack by hypohalites, which we recently demonstrated at the halogenation of thymol (Scheme 11).119
Footnote |
† The authors, listed in alphabetical order, contributed equally to this work. |
This journal is © The Royal Society of Chemistry 2015 |