Directed Evolution of Artificial Metalloenzymes: A Universal Means to

2 days ago - (9−13) We hoped that directed evolution would provide a general .... proteins that function in nature as molecular transport carriers f...
0 downloads 0 Views 4MB Size
Article Cite This: Acc. Chem. Res. XXXX, XXX, XXX−XXX

pubs.acs.org/accounts

Directed Evolution of Artificial Metalloenzymes: A Universal Means to Tune the Selectivity of Transition Metal Catalysts? Published as part of the Accounts of Chemical Research special issue “Artificial Metalloenzymes and Abiological Catalysis of Metalloenzymes”. Manfred T. Reetz*

Acc. Chem. Res. Downloaded from pubs.acs.org by IOWA STATE UNIV on 01/28/19. For personal use only.

Chemistry Department, Philipps-University, Hans-Meerwein-Strasse 4, 35032 Marburg, Germany Max-Planck-Institut für Kohlenforschung, Kaiser-Wilhelm-Platz 1, 45470 Mülheim Germany CONSPECTUS: Transition metal catalysts mediate a wide variety of chemo-, stereo-, and regioselective transformations, and therefore play a pivotal role in modern synthetic organic chemistry. Steric and electronic effects of ligands provide organic chemists with an exceedingly useful tool. More than four decades ago, chemists began to think about a different approach, namely, embedding achiral ligand/metal moieties covalently or noncovalently in protein hosts with formation of artificial metalloenzymes. While structurally fascinating, this approach led in each case only to a single (bio)catalyst, with its selectivity and activity being a matter of chance. In order to solve this fundamental problem, my group proposed in 2000− 2002 the idea of directed evolution of artificial metalloenzymes. In earlier studies, we had already demonstrated that directed evolution of enzymes constitutes a viable method for enhancing and inverting the stereoselectivity of enzymes as catalysts in organic chemistry. We speculated that it should also be possible to manipulate selectivity and activity of artificial metalloenzymes, which would provide organic chemists with a tool for optimizing essentially any transition metal catalyzed reaction type. In order to put this vision into practice, we first turned to the Whitesides system for artificial metalloenzyme formation, comprising a biotinylated diphosphine/Rh moiety, which is anchored noncovalently to avidin or streptavidin. Following intensive optimization, proof of principle was finally demonstrated in 2006, which opened the door to a new research area. This personal Account critically assesses these early studies as well as subsequent efforts from my group focusing on different protein scaffolds, and includes briefly some of the most important current contributions of other groups. Two primary messages emerge: First, since organic chemists continue to be extremely good at designing and implementing man-made transition metal catalysts, often on a large scale, those scientists that are active in the equally intriguing field of directed evolution of artificial metalloenzymes should be moderate when generalizing claims. All factors required for a truly viable catalytic system need to be considered, especially activity and ease of upscaling. Second, the most exciting and thus far very rare cases of directed evolution of artificial metalloenzymes are those that focus on selective transformations that are not readily possible using state of the art transition metal catalysts. • Sometimes product inhibition; • Most of the interesting transition metal catalyzed reaction types cannot be mediated by natural metalloenzymes.

1. INTRODUCTION Over the past 50 years organic chemists have invented an incredible number of synthetically useful transition metal catalyzed reaction types. In most cases, the transition metals are complexed to sterically and electronically tunable organic ligands, which govern activity and selectivity. As an alternative to man-made catalysts, the development of enzymes in synthetic organic chemistry has also progressed at a rapid pace.1,2 However, most organic chemists have been reluctant to include enzymes in their toolbox due to the following traditional limitations: • Narrow substrate acceptance (activity); • Often observed poor or wrong stereoselectivity; • Sometimes insufficient robustness under operating conditions; © XXXX American Chemical Society

During the past 25 years, these problems have been addressed by a protein engineering technique called directed evolution.3−5 It involves repeating cycles of gene mutagenesis, expression, and screening, which mimics natural evolution in the test tube. Since screening is generally the labor-intensive step in directed evolution (bottleneck) when evolving stereoand/or regioselectivity, methodology development for generating small but high-quality mutant libraries requiring Received: November 18, 2018

A

DOI: 10.1021/acs.accounts.8b00582 Acc. Chem. Res. XXXX, XXX, XXX−XXX

Article

Accounts of Chemical Research minimal screening was and still is a crucial endeavor.3 In addition to error-prone polymerase chain reaction (epPCR, a shotgun method) and DNA-shuffling (simulating natural sexual evolution),2−5 focused randomization by saturation mutagenesis at sites lining the enzyme’s binding pocket has emerged as a particularly effective approach, called Combinatorial Active-site Saturation Test (CAST).3 It is a systematization of saturation mutagenesis at the active site practiced earlier by our group. A site can be composed of one or more amino acid positions (residues). If the initial libraries do not ensure complete selectivity, then Iterative Saturation Mutagenesis (ISM) is ideally suited for further genetic optimization.3 The CAST/ISM strategy itself has undergone rapid methodology development, e.g., with the use of reduced amino acid alphabets as combinatorial building blocks,3 and novel ways to eliminate amino acid bias.6 These and other advancements have contributed to the emergence of directed evolution as a reliable tool in protein engineering. However, they do not resolve the disadvantage that enzymes cannot catalyze most of the transition metal mediated transformations that chemists have developed. By the year 2000, many bioconjugation methods had been developed, according to which man-made ligands, complexed to transition metals, are attached site-selectively to a variety of different protein scaffolds.7,8 In each case, this procedure generates a novel transition metal catalyst having a chiral local environment, but its promiscuous catalytic profile is unpredictable. Moreover, in such a process only a single artificial metalloenzyme is produced. Selectivity is then a matter of chance. In 2000−2002 ,my group at the Max-Planck-Institut für Kohlenforschung in Mülheim proposed the concept of directed evolution of artificial metalloenzymes, which we initially dubbed “hybrid catalysts”.9−13 We hoped that directed evolution would provide a general means to manipulate the stereoselectivity of any artificial metalloenzyme. Therefore, in principle, any transition metal catalyzed transformation that organic chemists have devised can be tuned for stereo- and/or regioselectivity as well as activity. It was also clear that at least three different routes to artificial metalloenzymes are possible:12,13 (1) covalent anchoring of a ligand/metal, (2) supramolecular noncovalent anchoring, and (3) the design of a protein scaffold that binds transition metal salts directly (Scheme 1). Yet another route to biocatalyzed promiscuous transformations of synthetic value was inspired by the seminal study of Dawson, Gellman, and Breslow, who showed in 1983 that a cytochrome P450 monooxygenase (CYP) is capable of inter- and intramolecular CH-activation when nitreneprecursors are used as substrates (Scheme 2).14 In these reactions, nitrenoid heme-FeNR intermediates undergo CHinsertion. This approach clearly expands the perspectives indicated in Scheme 1, but was not realized until later by Arnold and co-workers (see 2015 review).4 Rather than nitrene intermediates of the type heme-FeNR, the analogous carbene intermediates heme-FeCR2 enable promiscuous P450-catalyzed cyclopropanation using diazo compounds4,15 (Section 5). This Account is a personal commentary of directed evolution of artificial metalloenzymes with emphasis on scope, limitations and perspectives in an exciting research area.

Scheme 1. Different Ways To Generate Artificial Metalloenzymes, To Be Genetically Optimized for Stereoand/or Regioselectivity and Activity by Directed Evolution

Reproduced with permission from ref 13. Copyright 2009 Springer.

2. PROOF OF PRINCIPLE Many different artificial metalloenzymes were already available back in 2002,7,8 which could be used in directed evolution.9 We chose different scaffolds for anchoring designed metalcomplexes either covalently or noncovalently,9−13 with the latter being an example based on the Whitesides system7 comprising a biotinylated achiral diphosphine-complex biotinN[CH2CH2PPh2]2Rh which binds to avidin. In their seminal study, Wilson and Whitesides applied this novel artificial metalloenzyme as a catalyst in the asymmetric hydrogenation of N-acyl acrylic acid. It resulted in an ee-value of only 44% (S), but a uniquely structured metal-(bio)catalyst had been created!7 Since the expression system of avidin was not efficient enough to ensure sufficient amounts of protein in each well of 96-format microtiter plates, we turned to streptavidin, which also binds strongly to biotin, and for which an excellent expression system had been published.16 Unfortunately, despite months of work, we could not reproduce it. Before narrating our next steps, we cite the contributions of Ward et al., who were also working on the Whitesides system. In early work,17,18 they focused primarily on modifying the spacer-length between biotin and different Rh-diphosphine complexes, and varying the pH and pressure, while comparing avidin versus streptavidin with an improved expression system (streptavidin being superior to avidin). In the hydrogenation of N-acyl acrylic acid, an ee of 92% was achieved, which was increased to 96% ee by site-directed mutagenesis (Ser112Gly).17 This chemistry has been reviewed.19,20 In continuation of our efforts to provide proof of concept that directed evolution is principally suited for tuning artificial B

DOI: 10.1021/acs.accounts.8b00582 Acc. Chem. Res. XXXX, XXX, XXX−XXX

Article

Accounts of Chemical Research

Scheme 2. Promiscuous P450 Monooxygenase Catalyzed Nitrene Insertion into the CH2−Moiety of Cyclohexane without the Need To Introduce Mutations

Scheme 3. Model Reaction in the First Case of Directed Evolution of an Artificial Metalloenzyme

Figure 1. Excerpt of the biotinylated Rh-complex modeled in the X-ray structure of streptavidin/biotin, showing possible close (purple) and distal (blue) randomization residues for saturation mutagenesis. The two red dots mark the two slightly different positions of Rh. Reproduced with permission from ref 12. Copyright 2006 Royal Society of Chemistry.

led to a hit (Ser112Gly) showing an improved enantioselectivity of +35% ee. Interestingly, this is the same mutation that Ward et al. had introduced by rational design.17 In an ISM step, the gene of this variant was then used for randomizing position 49. The best double mutant Asn49Val/Ser112Gly showed a notable increase in enantioselectivity amounting to +54% ee (Scheme 4).12 In order to explore the role of point mutation Ser112Gly, the gene of the double mutant was subjected to NNK-based saturation mutagenesis at position 112. To our surprise, an even better hit was discovered, variant Asn49Val, showing an enantioselectivity of +65% ee. This means that in the double mutant, point mutation Ser112Gly is epistatically deleterious. Slight reversal of enantioselectivity was also achieved using ISM (Scheme 4), but due to the extensive labor required in screening, further mutagenesis was not undertaken. At this point we decided to publish this proof of principle study.12 Although the results are of no synthetic value (even if 100% ee had been achieved), our study showed for the first time that

metalloenzymes and thus of man-made metal-catalysts, we opted for 250 mL flasks rather than microtiter plates, which imposed rigid limits on the number of transformants (mutants) that could be screened for enantioselectivity. We used the methyl ester of N-acyl acrylic acid because the product is easier to extract (Scheme 3).12 As expected, the wildtype (WT) artificial metalloenzyme led to low enantioselectivity (+23% ee). We then modeled the biotinylated Rh-complex into the known X-ray structure of streptavidin/biotin (Figure 1).12 Two major poses with two slightly different positions of Rh in the “binding pocket”. CAST-type saturation mutagenesis3 was then considered at first-sphere residues Asn49, Leu110, Ser112, and Leu124, and at second-sphere residues Glu51, Tyr54, Trp79, Asn81, Arg84, and His89 (Figure 1).12 Only a few of these residues were chosen for limited iterative saturation mutagenesis (ISM).3 First, NNK-based saturation mutagenesis encoding all 20 canonical amino acids was performed at positions 110, 112 and 124, in each case 200 transformants being screened for 95% library coverage, which C

DOI: 10.1021/acs.accounts.8b00582 Acc. Chem. Res. XXXX, XXX, XXX−XXX

Article

Accounts of Chemical Research

that the inherent activity of a synthetic transition metal catalyst increases significantly?

Scheme 4. First Example of Directed Evolution of an Artificial Metalloenzyme, Based on the Whitesides System As the Catalyst in the Rh-Mediated Hydrogenation of Substrate 1 with Formation of (R)- and (S)-2 on an Optional Basis

3. GENERALIZATION OF THE STREPTAVIDIN SYSTEM Of the many far-reaching contributions of Ward and coworkers,17−20 only one study exploiting ISM is highlighted here.21 Several structurally different achiral Ru-complexes were biotinylated and anchored to streptavidin in order to test Noyori-type asymmetric transfer hydrogenation of prochiral ketones. Three different tuning tools were tested, spacer length, substitution pattern in the aromatic moiety of the piano stool Ru-ligand, and ISM-based directed evolution. Seven structurally different ketones were tested. Both (R)- and (S)selectivity were evolved in the reduction of 4-phenyl-2butanone and of other prochiral ketones.21 The crystal structure of one of the artificial streptavidin-based metalloenzymes revealed that the metal (Ru) itself is chiral (S)! 4. DEVELOPING OTHER PROTEIN SCAFFOLDS FOR USE IN DIRECTED EVOLUTION OF ARTIFICIAL METALLOENZYMES For potential use in directed evolution, we also considered many of the artificial metalloenzymes that had been reported previously.8 However, we decided to develop new ones which had the potential to be generalized for all kinds of transition metal mediated reaction types.9,22

Reproduced with permission from ref 11. Copyright 2004 Royal Society of Chemistry.

4.1. Cu(II)-Phthalocyanine Anchored Supramolecularly to Serum Albumins

a fundamentally different approach to metal-catalyst optimization is possible, namely directed evolution of artificial metalloenzymes. Nevertheless, we abandoned the Whitesides system and turned to other protein scaffolds (sections 4.1−4.3). As already noted, Ward and co-workers transformed the streptavidin system into a truly effective and general tool, as featured briefly in section 3. The proof of principle study aroused excitement at the time, because it suggested that a Darwinian approach to tuning manmade transition metal catalysts could be a general tool. Nevertheless, we repeatedly emphasized that evolving pronounced activity was necessary. Therefore, a fundamental question was posed:12,13 Can the local protein environment surrounding the transition metal be tuned by directed evolution so

One option was the use of serum albumins as scaffolds, proteins that function in nature as molecular transport carriers for such natural compounds as fatty acids, bile acids, bilirubin, and heme. Gross and co-workers had already anchored Mn(III)-corroles noncovalently to serum albumins and demonstrated that such supramolecular complexes catalyze the H2O2-mediated asymmetric sulfoxidation of prochiral thioethers (up to 74% ee).23 They also showed that the sodium salts of di-, tri-, and tetrasulfonic acid derivatives of porphyrins, phthalocyanines, and corroles likewise undergo strong supramolecular binding with serum albumins, although the exact point of attachment remained unknown. Fortunately, in a subsequent study by Curry and co-workers, an unambiguous

Figure 2. Water-soluble Cu(II)-phthalocyanine complex: (left) chemical structure; right) modeled into human serum albumin (HSA) based on the X-ray structure of apo-HSA. Adapted with permission from ref 25. Copyright 2006 Wiley-VCH. D

DOI: 10.1021/acs.accounts.8b00582 Acc. Chem. Res. XXXX, XXX, XXX−XXX

Article

Accounts of Chemical Research assignment was made on the basis of the X-ray structure of Feprotoporphyrin dimethyl ester.24 This structural data helped us to design a Cu(II)-based artificial metalloenzyme for use in asymmetric Diels−Alder cycloaddition and in other Cu-catalyzed reactions.25 We anticipated that the commercially available water-soluble Cu(II)-phthalocyanine (Figure 2, left) would bind analogously to HSA and other transport proteins such as bovine serum albumin (BSA). As a first step, we modeled the Cu(II)complex into HSA (Figure 2, right).25 This suggested that Diels−Alder cycloadditions (Scheme 5) would occur in the cavity directly opposite to the flat Cu(II)-complex. It is the same reaction that Roelfes, Feringa, and co-workers had used in DNA catalysis.26

Scheme 6. Chemical Modification of tHisF Regioselectively at Cys11 with Formation of Bioconjugates via Michael addition (A) or via SN2 Reaction (B)

Reproduced with permission from ref 27. Copyright 2008 WileyVCH.

Scheme 5. Diels−Alder Reactions Catalyzed by the Cu(II)Phthalocyanine (Figure 2) Which Is Anchored Supramolecularly to Various Serum Albumins

implemented. Scheme 7 displays some of the precursors that were anchored via covalent bioconjugation to tHisF, thereby Scheme 7. Molecular Precursors That Were Synthesized and Anchored Covalently to the Host Protein tHisF

Most of the serum albumins led to high enantioselectivity and pronounced endoselectivity, e.g., in the case of BSA, eevalues up to 93% and endo/exoratios reaching 96:4 were observed. In the case of substrate 3d, essentially complete enantioselectivity resulted (98% ee).25 We postulated that Cu(II) is chelated by the O atom of the carbonyl group and the N atom of the pyridine moiety, thereby activating the substrate. Since enantio- and diastereoselectivity proved to be so high, further tuning of this artificial metalloenzyme by directed evolution was not necessary, except for the low activity (3 days for 75−90% conversion). Other Cu(II)- and Cu(I)-catalyzed transformations have yet to be studied, e.g., click reactions. The use of Fe would be a step toward an artificial P450 enzyme.

setting the stage for complexation of transition metal salts (which is also possible prior to bioconjugation). It can also be seen that the cationic species are precursors of nucleophilic carbenes which, upon base-mediated deprotonation in the environment of tHisF, could mediate Stetter-type organocatalytic reactions.27,28

4.2. tHis-F Scaffold for Anchoring Metal Chelating Agents, Cofactors, and Organocatalysts

4.3. tHisF as a Scaffold for Direct Transition Metal Complexation

Another protein that we considered as a scaffold for artificial metalloenzyme formation is the so-called tHisF enzyme from Thermotoga maritima, which plays a crucial role in the biosynthesis of histidine. It is unusually thermostable and can be expressed efficiently in E. coli. In order to ensure regioselective bioconjugation on the way to a family of novel artificial metalloenzymes, we first generated the mutant Cys9Ala/Asp11Cys having Cys11 at a prominent position for ready bioconjugation via Michael additions or SN2 reactions (Scheme 6), in which the R-group represents an achiral moiety that chelates transition metals.27 We completed the miniaturization of the experimental platform so that potential application in directed evolution became possible, but systematic exploration of the many possibilities has still to be

As an example of an artificial metalloenzyme generated by direct transition metal complexation (Scheme 1, bottom), we again chose tHisF.29 Nature has evolved several different ways to effectively bind transition metals in proteins, e.g., as heme− metal entities, but also upon direct complexation by perfectly positioned residues having donor capacity in the side chains, with prominent examples being many Cu-containing proteins with two histidines at the metal (in addition to water molecules, and sometimes cysteine). The X-ray structure of tHisF shows an aspartate residue at position 11 at the top of the typical TIM barrel, which could loosely bind metals. We therefore engineered an Asp/His/His triad for coordinating Cu(II) tightly.29 Such a triad occurs in certain Fe(II) proteins. The putative metal binding site comprising Asp11/His50/ E

DOI: 10.1021/acs.accounts.8b00582 Acc. Chem. Res. XXXX, XXX, XXX−XXX

Article

Accounts of Chemical Research

5. IRON AND IRIDIUM HEME-BASED ARTIFICIAL METALLOENZYMES As already indicated in the Introduction, another way to enable promiscuous enzyme-catalyzed reactions is the use of CYPs. The structure, and mechanism and catalytic profile of many different cytochrome P450 monooxygenases had already been studied, hydroxylation and epoxidation being the most common reactions.30−33 In the attempt to discover new promiscuous P450-catalyzed transformations, Arnold and co-workers used the standard cyclopropanation reaction of styrene with diazo acetic acid ethyl ester leading to four stereoisomeric cyclopropane products (Scheme 8).15 P450-BM3 was chosen as the enzyme,

His52 near the top rim of tHisF was designed on the basis of crude modeling (Figure 3). Since Cys9 could interfere, it was

Scheme 8. Model Reaction Used in the Directed Evolution of P450-BM3

a well-known self-sufficient CYP30−33 that the group had used in a number of previous studies to control the regioselectivity of oxidative hydroxylation.4 Small libraries of mutants, produced earlier for other purposes, were tested in the model cyclopropanation, in addition to new libraries generated by saturation mutagenesis in the vicinity of the enzyme’s bind pocket (CAST sites). The best mutant was found to deliver mainly the cis-product (S,R) in yields of 30−60% and with high diastereoselectivity (cis:trans = 92:8) and enantioselectivity (97% ee).15 The beneficial influence upon mutating the axial Fe-binding residue (cysteine) to serine was also discovered.4,34 Some of the parallel studies by Arnold et al.4 and Fasan et 35 al. focused on amine syntheses by nitrenoid insertion of heme-Fe = NR intermediates into CH-bonds, work that has been reviewed elsewhere.4 Fasan and co-workers opened a new door by turning to myoglobin as the biocatalyst in carbene and nitrene reactions, a heme-dependent protein which appears to have a number of advantages over P450-BM3.35,36 In view of the fact that Fe in hemoglobins and myoglobins had been substituted by other metals such as cobalt37 and manganese,38 Hartwig, Clark, and co-workers went a crucial step further by substituting the natural metal (Fe) in P450BM3 by iridium, which provided an access to many types of catalytic transformations such as carbene insertions into CHbonds, in these reactions showing, kinetics of natural enzymes and up to 98% enantioselectivity.39 Other reaction types such as stereo- and regioselective cyclopropanation were also reported.40 In the meantime, Fasan et al. reported the use of such metals as Co, Ir, Rh, and Ru in myoglobin.41 My group was interested in testing a different concept using evolved mutants of P450-BM3. We speculated that in the Fe(II)-state, WT or an appropriate mutant could catalyze the Kemp elimination via a unique redox mechanism (Scheme 9).42 The Kemp elimination as such is of little synthetic use, but it has been used for decades as a platform for designing metal-free protein scaffolds which catalyze the reaction by an acid/base mechanism, letting researchers learn many mechanistic intricacies. The most active Kemp eliminase reaching the

Figure 3. Designed direct metal ligating site in tHisF. (a) Close-up view of the metal binding triad Asp11/His50/His52. (b) View of genetically modified tHisF showing the position of the binding triad at the top rim. Reproduced with permission from ref 29. Copyright 2010 Wiley-VCH.

changed to Ala9, which means that mutant Cys9Ala/ Leu50His/IleHis52 served as the starting point for complexation (in addition to the already present Asp11 and possibly H2O). Upon treating the modified tHisF protein with Cu(II) salts, the resulting complexes were characterized by electron paramagnetic resonance spectroscopy (EPR). Unambiguous evidence for the putative binding mode was obtained by monitoring the magnetic interactions of the 14N nuclei in the Cu(II)-coordinating ligands using hyperfine sublevel correlation (HYSCORE) spectroscopy.29 Subsequently, the new artificial metalloenzyme was tested as a catalyst in the Diels−Alder reaction 3a + 4 → 5 (Scheme 5). Without any directed evolution, this biocatalyst led to an enantioselectivity of 46% ee and an endoselectivity ratio of 13:1.29 The system is set up for directed evolution, especially with the aim of enhancing and reversing enantioselectivity and in increasing activity. The binding triad should also be useful for creating Fe(II)- and Mn(II)-based artificial metalloenzymes. F

DOI: 10.1021/acs.accounts.8b00582 Acc. Chem. Res. XXXX, XXX, XXX−XXX

Article

Accounts of Chemical Research Scheme 9. Redox Mechanism of P450-BM3 catalyzed Kemp Elimination As Supported by QM Calculations

Reproduced with permission from ref 42. Copyright 2017 Nature.

al.50 as well as Green et al.51 and Young and Schultz52 have recently summarized the use of noncanonical amino acids for engineering artificial metalloenzymes and for other purposes. This approach allows site-specific coordination of transition metals with formation of novel artificial metalloenzymes, as also reported by Yu et al.53 Further advances reflect other types of perspectives.54 In all of the studies on artificial metalloenzyme-catalysis and in similar contributions from other groups, CAST- or CAST/ ISM-methods3 and/or “rational” site-specific mutagenesis at CAST sites constitute the currently best strategies for controlling stereoselectivity and improving the activity of artificial metalloenzymes. Not every new study stands for true efficiency, nor can this be expected at this stage. After all, it is a challenge to beat the efficiency of privileged man-made catalysts. It is also interesting to note that receptor-based artificial metalloenzymes have been installed on living cells where, inter alia, the Diels−Alder cycloaddition (Scheme 5) was shown to occur.55 In a very different approach, Bäckvall and co-workers recently described a biohybrid catalyst by embedding Pd-nanoparticles in cross-linked lipase CALB, which was used in cascade reactions.56 Some degree of modesty is called for in the general field of artificial metalloenzymes because it has not reached the developmental state of traditional transition metal catalysis. It appears that it is easier to evolve stereoselectivity than activity, although progress has been made.39 In a stellar contribution, Hilvert and co-workers recently described an artificial metalloenzyme that displays saturation kinetics in ester hydrolysis comparable to natural hydrolases (70,000-fold improvement by directed evolution).57 While the transformation itself is of moderate synthetic interest, this study demonstrates that truly pronounced enzyme rates can indeed be achieved, and that directed evolution of artificial metalloenzymes provides a stage for learning fundamental mechanistic lessons. Hopefully, researchers will address this issue more so than in the past, in addition to providing processes that allow more than just a few milligrams of product to be isolated. Most exciting are those selective transformations that are not readily possible using modern man-made transition metal catalysts or organocatalysts.

efficiency of natural enzymes was recently reported by Hilvert and co-workers using a designed metal-free protein.43 We were not interested in beating this record, but in establishing a fundamentally new mechanism for a completely different Kemp eliminase. In our study, a P450-BM3 mutant was identified which proved to be unusually active as a Kemp eliminase (kcat = 11.5 s−1; kcat/Km = 8800 s−1 M−1; kcat/kuncat = 107).42 What looks like a complicated mechanism, is actually rather simple, as supported by extensive QM computations: The substrate receives an electron from heme-Fe(II), which leads to the spontaneous rupture of the weakest bond (O−N), followed by a conformational change which enables H-transfer and reconstitutes the Fe(II)-valence state (Scheme 9).42 We speculated that P450-BM3 in the Fe(II)-state can also catalyze a range of other redox-based organic transformations, yet to be found. The general theme for discovering new promiscuous reactions using this system is the succession of (1) single electron transfer, (2) chemical change, and (3) return of the single electron to iron with formation of heme-Fe(II). Perhaps other reaction types or polymerizations can be mediated by this kind of catalytic electron transfer.

6. CONCLUSIONS AND PERSPECTIVES It took some time to provide proof of principle illustrating experimentally the idea of directed evolution of artificial metalloenzymes (Scheme 2).12 Fortunately, since then this exciting new research area has blossomed with the emergence of many novel protein scaffolds and applications using a variety of different reaction types. Some of them are synthetically more interesting4,25,44 than the original study involving asymmetric Rh-catalyzed olefin hydrogenation.12 Arnold and co-workers have expanded their interest in designing and engineering artificial metalloenzymes, inter alia, as P450catalysts in chiral borane synthesis,44 construction of strained carbocycles,45,46 and synthesis of chiral silicon compounds.47 Okuda, Schwaneberg, and co-workers have described an artificial metalloenzyme that catalyzes Grubbs−Hoveyda-type olefin metathesis,48 while Budisa, Süssmuth, and co-workers designed a ribosomal lasso peptide using noncanonical amino acids for anchoring the same type of Ru-catalyst.49 Budisa et G

DOI: 10.1021/acs.accounts.8b00582 Acc. Chem. Res. XXXX, XXX, XXX−XXX

Article

Accounts of Chemical Research



(11) Reetz, M. T. Controlling the enantioselectivity of enzymes by directed evolution: Practical and theoretical ramifications. Proc. Natl. Acad. Sci. U. S. A. 2004, 101, 5716−5722. (12) Reetz, M. T.; Peyralans, J. J.-P.; Maichele, A.; Fu, Y.; Maywald, M. Directed Evolution of Hybrid Enzymes: Evolving Enantioselectivity of an achiral Rh-complex Anchored to a Protein. Chem. Commun. 2006, 41, 4318−4320. (13) M. T. Reetz, M. T. Directed Evolution of Stereoselective Hybrid Catalysts. In Topics in Organometallic Chemistry, Vol. 25; Ward, T. R., Ed.; Springer: Berlin, Heidelberg, 2009; pp 63−92. (14) Svastits, E. W.; Dawson, J. H.; Breslow, R.; Gellman, S. H. Functionalized nitrogen atom transfer catalyzed by cytochrome P450. J. Am. Chem. Soc. 1985, 107, 6427−6428. (15) Coelho, P. S.; Brustad, E. M.; Kannan, A.; Arnold, F. H. Olefin cyclopropanation via carbene transfer catalyzed by engineered cytochrome P450 enzymes. Science 2013, 339, 307−310. (16) Wu, S.-C.; Wong, S.-L. Engineering of a Bacillus subtilis strain with adjustable levels of intramolecular biotin for secretory production of functional streptavidin. Appl. Environ. Microbiol. 2002, 68, 1102−1108. (17) Collot, J.; Gradinaru, J.; Humbert, N.; Skander, M.; Zocchi, A.; Ward, T. R. Artificial Metalloenzymes for Enantioselective Catalysis Based on Biotin-Avidin. J. Am. Chem. Soc. 2003, 125, 9030−9031. (18) Skander, M.; Humbert, N.; Collot, J.; Gradinaru, J.; Klein, G.; Loosli, A.; Sauser, J.; Zocchi, A.; Gilardoni, F.; Ward, T. R. Artificial Metalloenzymes: (Strept)avidin as Host for Enantioselective Hydrogenation by Achiral Biotinylated Rhodium-Diphosphine Complexes. J. Am. Chem. Soc. 2004, 126, 14411−14418. (19) Review of artificial metalloenzymes with emphasis on the (strept)avidin system: Hyster, T. K.; Ward, T. R. Genetic optimization of metalloenzymes: Enhancing Enzymes for Non-natural Reactions. Angew. Chem., Int. Ed. 2016, 55, 7344−7357. (20) Newest review of artificial metalloenzymes with emphasis on the (strept)avidin system: Schwizer, F.; Okamoto, Y.; Heinisch, T.; Gu, Y.; Pellizzoni, M. M.; Lebrun, V.; Reuter, R.; Köhler, V.; Lewis, J. C.; Ward, T. R. Artificial Metalloenzymes: Reaction Scope and Optimization Strategies. Chem. Rev. 2018, 118, 142−231. (21) Creus, M.; Pordea, A.; Rossel, T.; Sardo, A.; Letondor, C.; Ivanova, A.; Letrong, I.; Stenkamp, R. E.; Ward, T. R. X-ray structure and designed evolution of an artificial transfer hydrogenase. Angew. Chem., Int. Ed. 2008, 47, 1400−1404. (22) Ilie, A.; Reetz, M. T. Directed Evolution of Artificial Metalloenzymes. Isr. J. Chem. 2015, 55, 51−60. (23) Mahammed, A.; Gross, Z. Albumin-conjugated corrole metal complexes: extremely simple yet very efficient biomimetic oxidation systems. J. Am. Chem. Soc. 2005, 127, 2883−2887. (24) Zunszain, P. A.; Ghuman, J.; Komatsu, T.; Tsuchida, E.; Curry, S. Crystal structural analysis of human serum albumin complexed with hemin and fatty acid. BMC Struct. Biol. 2003, 3, 1−9. (25) Reetz, M. T.; Jiao, N. Copper-phthalocyanine conjugates of serum albumins as enantioselective catalysts in Diels-Alder reactions. Angew. Chem., Int. Ed. 2006, 45, 2416−2419. (26) Review of DNA-based transition metal catalysis: Boersma, A. J.; Megens, R. P.; Feringa, B. L.; Roelfes, G. DNA-based Asymmetric Catalysis. Chem. Soc. Rev. 2010, 39, 2083−2092. (27) Reetz, M. T.; Rentzsch, M.; Pletsch, A.; Taglieber, A.; Hollmann, F.; Mondiere, R. J. G.; Dickmann, N.; Höcker, B.; Cerrone, S.; Haeger, M. C.; Sterner, R. A Robust Protein Host for Anchoring Chelating Ligands and Organocatalysts. ChemBioChem 2008, 9, 552−564. (28) Reetz, M. T. Directed Evolution of Promiscuity: Artificial Enzymes as Catalysts in Organic Chemistry. In Reetz, M. T. Directed Evolution of Selective Enzymes: Catalysts for Organic chemistry and Biotechnology; Wiley-VCH: Weinheim, 2016; Chapter 7, pp 237−266. (29) Podtetenieff, J.; Taglieber, A.; Bill, E.; Reijerse, E. J.; Reetz, M. T. An Artificial Metalloenzyme: Creation of a Design Copper Binding Site in a Thermostable Protein. Angew. Chem., Int. Ed. 2010, 49, 5151−5155.

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Manfred T. Reetz: 0000-0001-6819-6116 Notes

The author declares no competing financial interest. Biography Manfred T. Reetz was born in Germany 1943, immigrated to the United States in 1952, and obtained Bachelor and Master degrees at Washington University and University of Michigan, respectively, before obtaining the doctoral degree at Göttingen University/ Germany in 1969 in synthetic organic chemistry under the guidance of Ulrich Schöllkopf. Following a postdoctoral stay with Reinhard Hoffmann at Marburg University and several academic positions in Germany, he served as Full Professor in the Chemistry Department of Marburg University 1980−1991. From 1991 to 2011, he was Director at the Max-Planck-Institut für Kohlenforschung in Mülheim. In 2011, he retired and became Emeritus Professor at Marburg University and external emeritus group leader of the Mülheim MPI, and more recently Adjunct Professor at Tianjin Institute of Industrial Biocatalysis, Chinese Academy of Sciences. Since the 1997 proof of principle study, he has published ∼200 papers on directed evolution of stereo- and regioselective enzymes for use in organic chemistry and biotechnology.



ACKNOWLEDGMENTS Generous support by the Max-Planck-Society is gratefully acknowledged. I also thank Guangyue Li and Yijie Dong for preparing some figures and formatting the manuscript.



REFERENCES

(1) Drauz, K.; Gröger, H.; May, O., Eds. Enzyme Catalysis in Organic Chemistry; Wiley-VCH: Weinheim, 2012. (2) Reetz, M. T. Biocatalysis in Organic Chemistry and Biotechnology: Past, Present, and Future. J. Am. Chem. Soc. 2013, 135, 12480−12496. (3) Review of directed evolution of stereoselective enzymes based on the CAST and CAST/ISM methods: Reetz, M. T. Laboratory Evolution of Stereoselective Enzymes: A Prolific Source of Catalysts for Asymmetric Reactions. Angew. Chem., Int. Ed. 2011, 50, 138−174. (4) Review of engineered artificial metalloenzymes based on CYPs: Renata, H.; Wang, Z. J.; Arnold, F. H. Expanding the Enzyme Universe: Mechanism-Guided Accessing Non-Natural Reactions by Directed Evolution. Angew. Chem., Int. Ed. 2015, 54, 3351−3367. (5) Recent review of directed evolution: Zeymer, C.; Hilvert, D. Directed Evolution of Protein Catalysis. Annu. Rev. Biochem. 2018, 87, 131−157. (6) Li, A.; Sun, Z.; Reetz, M. T. Solid-Phase Gene Synthesis for Mutant Library Construction: The Future of Directed Evolution? ChemBioChem 2018, 19, 2023−2032. (7) Wilson, M. E.; Whitesides, G. M. Conversion of a Protein to a Homogeneous Asymmetric Hydrogenation Catalyst by Site-Specific Modification with a Diphosphine-rhodium(I) Moiety. J. Am. Chem. Soc. 1978, 100, 306−307. (8) Early review of artificial metalloenzymes: Qi, D.; Tann, C.-M.; Haring, D.; Distefano, M. D. Generation of new enzymes via covalent modification of existing proteins. Chem. Rev. 2001, 101, 3081−3111. (9) Reetz, M. T. Optimierung von synthetischen Katalysatoren durch gerichtete Evolution (Optimization of synthetic catalysts by directed evolution). Patent DE-A 10129187.6; WO 92, 18645, 2001. (10) Reetz, M. T. Directed evolution of selective enzymes and hybrid catalysts. Tetrahedron 2002, 58, 6595−6602. H

DOI: 10.1021/acs.accounts.8b00582 Acc. Chem. Res. XXXX, XXX, XXX−XXX

Article

Accounts of Chemical Research

Enzymology Meets Xenobiology. Angew. Chem., Int. Ed. 2017, 56, 9680−9703. (51) Hayashi, T.; Hilvert, D.; Green, A. P. Engineered Metalloenzymes with Non-Canonical Coordination Environments. Chem. Eur. J. 2018, 24, 11821−11830. (52) Young, D. D.; Schultz, P. G. Playing with the Molecules of Life. ACS Chem. Biol. 2018, 13, 854−870. (53) Yu, Y.; Hu, C.; Xia, L.; Wang, J. Artificial Metalloenzyme Design with Unnatural Amino Acids and Non-Native Cofactors. ACS Catal. 2018, 8, 1851−1863. (54) Petrick, I. D.; Hosseinzadeh, P.; Nilges, M. J.; Lu, Y. A designed heme-[4Fe-4S] metalloenzyme catalyzes sulfite reduction like native enzyme. Science 2018, 361, 1098−1101. (55) Ghattas, W.; Dubosclard, V.; Wick, A.; Bendelac, A.; Guillot, R.; Ricoux, R.; Mahy, J.-P. Receptor-based artificial metalloenzymes in human cells. J. Am. Chem. Soc. 2018, 140, 8756−8762. (56) Görbe, T.; Gustafson, K. P. J.; Verho, O.; Kervefors, G.; Zheng, H.; Zou, X.; Johnston, E. V.; Bäckvall, J.-E. Design of a Pd(0)-CalB CLEA Biohybrid Catalyst and its Application in a One-Pot Cascade Reaction. ACS Catal. 2017, 7, 1601−1605. (57) Studer, S.; Hansen, D. A.; Pianowski, Z. L.; Mittl, P. R. E.; Debon, A.; Guffy, S. L.; Der, B. S.; Kuhlman, B.; Hilvert, D. Evolution of a highly active and enantiospecific metalloenzyme from short peptides. Science 2018, 362, 1285−1288.

(30) Ortiz de Montellano, P. R. Hydrocarbon Hydroxylation by Cytochrome P450 Enzymes. Chem. Rev. 2010, 110, 932−948. (31) Isin, E. M.; Guengerich, F. P. Complex reactions catalyzed by cytochrome 450 enzymes. Biochim. Biophys. Acta, Gen. Subj. 2007, 1770, 314−329. (32) Shaik, S.; Cohen, S.; Wang, Y.; Chen, H.; Kumar, D.; Thiel, W. P450 Enzymes: Their Structure, Reactivity, and Selectivity - Modeled by QM/M Calculations. Chem. Rev. 2010, 110, 949−1017. (33) Poulos, T. L. Heme Enzyme Structure and Function. Chem. Rev. 2014, 114, 3919−3962. (34) Coelho, P. S.; Wang, Z. J.; Ener, M. E.; Baril, S. A.; Kannan, A.; Arnold, F. H.; Brustad, E. M. A serine-substituted P450 catalyzes highly efficient carbene transfer to olefins in vivo. Nat. Chem. Biol. 2013, 9, 485−487. (35) Singh, R.; Bordeaux, M.; Fasan, R. P450-catalyzed Intramolecular sp3 CH-Amination with Arylsulfonyl Azide Substrates. ACS Catal. 2014, 4, 546−552. (36) Bordeaux, M.; Tyagi, V.; Fasan, R. Highly Diastereoselective and Enantioselective Olefin Cyclopropanation Using Engineered Myoglobin-Based Catalysts. Angew. Chem., Int. Ed. 2015, 54, 1744− 1748. (37) Yonetani, T.; Yamamoto, H.; Woodrow, G. V., III. Studies of Cobalt Myoglobins and Hemoglobins. J. Biol. Chem. 1974, 249, 682− 690. (38) Bordeaux, M.; singh, R.; Fasan, R. Intramolecular C(sp3)-H amination of arylsulfonyl azides with engineered and artificial myoglobin-based catalysts. Bioorg. Med. Chem. 2014, 22, 5697−5704. (39) Dydio, P.; Key, H. M.; Nazarenko, A.; Rha, J. Y.-E.; Seyedkazemi, V.; Clark, D. S.; Hartwig, J. F. Science 2016, 354, 102−106. (40) Dydio, P.; Key, H. M.; Hayashi, H.; Clark, D. S.; Hartwig, J. F. J. Am. Chem. Soc. 2017, 139, 1750−1753. (41) Sreenilayam, G.; Moore, E. J.; Steck, V.; Fasan, R. Metal substitution modulates the reactivity and extends the reaction scope of myoglobin carbene transfer catalysis. Adv. Synth. Catal. 2017, 359, 2076−2089. (42) Li, A.; Wang, B.; Ilie, A.; Dubey, K. D.; Bange, G.; Korendovych, I. V.; Shaik, S.; Reetz, M. T. A redox-mediated Kemp eliminase. Nat. Commun. 2017, 8, 14876. (43) Blomberg, R.; Kries, H.; Pinkas, D. M.; Mittl, P. R.; Grütter, M. G.; Privett, H. K.; Mayo, S. L.; Hilvert, D. Precision is essential for efficient catalysis in an evolved Kemp eliminase. Nature 2013, 503, 418−421. (44) Kan, S. B. J.; Huang, X.; Gumulya, Y.; Chen, K.; Arnold, F. H. Genetically Programmed Chiral Organoborane Synthesis. Nature 2017, 552, 132−138. (45) Chen, K.; Huang, X.; Kan, S. B. J.; Zhang, R. K.; Arnold, F. H. Enzymatic construction of highly strained carbocycles. Science 2018, 360, 71−75. (46) Lewis, R. D.; Garcia-Borras, M.; Chalkley, M. J.; Buller, A. R.; Houk, N. K.; Kan, S. B. J.; Arnold, F. H. Catalytic iron-carbene intermediates revealed in a cytochrome c carbene transferase. Proc. Natl. Acad. Sci. U. S. A. 2018, 115, 7308−7313. (47) Kan, S. B. J.; Lewis, R. D.; Chen, K.; Arnold, F. H. Directed evolution of cytochrome c for carbon-silicon bond formation: Bringing silicon to life. Science 2016, 354, 1048−1051. (48) Sauer, D. F.; Himiyama, T.; Tachikawa, K.; Fukumoto, K.; Onoda, A.; Mizohata, E.; Inoue, T.; Bocola, M.; Schwaneberg, U.; Hayashi, T.; Okuda, J. A Highly Active Biohybrid Catalyst for Olefin Metathesis in Water: Impact of a Hydrophobic Cavity in a β-Barrel Protein. ACS Catal. 2015, 5, 7519−7522. (49) Al Toma, R. S.; Kuthning, A.; Exner, M. P.; Denisiuk, A.; Ziegler, J.; Budisa, N.; Süssmuth, R. D. Site-directed and global incorporation of orthogonal and isostructural noncanonical amino acids into the ribosomal lasso peptide capistruin. ChemBioChem 2015, 16, 503−509. (50) Agostini, F.; Völler, J.-S.; Koksch, B.; Acevedo-Rocha, C. G.; Kubyshkin, V.; Budisa, N. Biocatalysts with Unnatural Amino Acids: I

DOI: 10.1021/acs.accounts.8b00582 Acc. Chem. Res. XXXX, XXX, XXX−XXX