Integrated catalysis opens new arylation pathways via regiodivergent enzymatic C-H activation

Abstract

Despite major recent advances in C-H activation, discrimination between two similar, unactivated C-H positions is beyond the scope of current chemocatalytic methods. Here we demonstrate that integration of regioselective halogenase enzymes with Pd-catalysed cross-coupling chemistry, in one-pot reactions, successfully addresses this problem for the indole heterocycle. The resultant 'chemobio-transformation' delivers a range of functionally diverse arylated products that are impossible to access using separate enzymatic or chemocatalytic C-H activation, under mild, aqueous conditions. This use of different biocatalysts to select different C-H positions contrasts with the prevailing substrate-control approach to the area, and presents opportunities for new pathways in C-H activation chemistry. The issues of enzyme and transition metal compatibility are overcome through membrane compartmentalization, with the optimized process requiring no intermediate work-up or purification steps.

ARTICLE

Received 11 Jan 2016 | Accepted 9 May 2016 | Published 10 Jun 2016

Jonathan Latham1,2, Jean-Marc Henry1,2, Humera H. Sharif1,2, Binuraj R.K. Menon1,2, Sarah A. Shepherd1,2, Michael F. Greaney1 & Jason Mickleeld1,2

Despite major recent advances in CH activation, discrimination between two similar, unactivated CH positions is beyond the scope of current chemocatalytic methods. Here we demonstrate that integration of regioselective halogenase enzymes with Pd-catalysed cross-coupling chemistry, in one-pot reactions, successfully addresses this problem for the indole heterocycle. The resultant chemobio-transformation delivers a range of functionally diverse arylated products that are impossible to access using separate enzymatic or chemocatalytic CH activation, under mild, aqueous conditions. This use of different biocatalysts to select different CH positions contrasts with the prevailing substrate-control approach to the area, and presents opportunities for new pathways in CH activation chemistry. The issues of enzyme and transition metal compatibility are overcome through membrane compartmentalization, with the optimized process requiring no intermediate work-up or purication steps.

DOI: 10.1038/ncomms11873 OPEN

Integrated catalysis opens new arylation pathways via regiodivergent enzymatic CH activation

1 School of Chemistry, The University of Manchester, Oxford Road, Manchester M13 9PL, UK. 2 Manchester Institute of Biotechnology, The University of

Manchester, 131 Princess Street, Manchester M1 7DN, UK. Correspondence and requests for materials should be addressed to M.F.G.

(email: mailto:[email protected]

Web End [email protected] ) or to J.M. (email: mailto:[email protected]

Web End [email protected] ).

NATURE COMMUNICATIONS | 7:11873 | DOI: 10.1038/ncomms11873 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 1

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms11873

Selective activation of CH positions has transformed synthesis in recent years. Using transition metal catalysis, a plethora of direct, fundamental bond formations are now

achievable under mild, catalytic conditions, changing the way chemists make molecules1. The eld is currently dened by substrate control, with a relatively small number of CH motifs undergoing predictable and reliable metalation by a transition metal catalyst; for example, electron-rich heteroarenes, electron-poor uoroarenes, activation ortho (and more recently, meta), to a directing group via cyclometalation all being dominant examples (Fig. 1a)29. Effective catalyst control, where choice of catalyst can discriminate between neighbouring CH bonds in the absence of strong substrate-directing effects, is very rare and remains an outstanding challenge in the eld1013. Progress in this area will profoundly enhance the scope of CH activation, as the majority of CH bonds do not t the current substrate-control criteria for chemocatalytic activation chemistry.

We were interested in addressing this problem using a combination of bio- and chemo-catalysis. The complex coordination sphere of an enzyme active site enables exquisite selectivities that cannot be achieved by simple metal coordination complexes. Correspondingly, the manifold power of noble metal catalysis in CC bond formation presents synthesis pathways that are not possible through biocatalysis. Their integration offers a complementarity that could create new CH activation strategies, whereby CH positions beyond the scope of chemocatalysis alone can be targeted for CC, CN, CHal and other fundamental bond-forming steps. In addition, the merging of bio- and

chemo-catalysis into single process, one-pot operations creates a step-change in efciency through enhanced space-time yields, better energy consumption and elimination of auxiliary chemicals14,15.

Despite the potential advantages of combining biocatalysts with chemocatalysts, their successful integration has been largely restricted to organic solvents1517, which are incompatible with many enzymes and cofactorstherefore limiting the number of biocatalysts that can be utilized. In contrast, integration of chemocatalysts with enzymes in their preferred aqueous media is relatively rare14. Indeed, the inherent differences in operating conditions of enzymes and chemocatalysts present serious challenges, often requiring compartmentalization or immobilization of at least one of the catalysts1823.

Our plans for a catalyst-controlled, integrated CH activation system involve the merging of avin-dependent halogenase (Fl-Hal) enzymes with palladium-catalysed cross-coupling. The pivotal role played by aryl halide functionality in metal-catalysed transformations suggested to us that a regioselective, biohalogenation would have great versatility, with the incipient Ar-X group being adaptable to a multitude of Pd-catalysed chemistries. We chose to integrate Fl-Hals and SuzukiMiyaura coupling (SMC) with boronic acids in the rst instance, given the critical role of SMC in producing biaryl products for pharmaceutical, agrochemical and materials science applications.

The most widely studied Fl-Hals are tryptophan halogenases, which are responsible for chlorinating different positions of the indole ring of tryptophan en route to various bacterial natural

R

a b

H

F

F

X

X = Br, I, OTf

R

DG

H

R

N H

H

F

F F

Biocatalysis Chemocatalysis

O

N H

Acidic CH Electron richCH

Ortho CH activation (cyclometalation)

Br, O2

FADH2

FAD NADH

NAD+ Glucose

Fl-Hal

Fre

GDH

[Pd] Ar-B(OH)2

DG

H2O

Gluconolactone

Ar

R

NH H

H

Meta CH sp3 CH

Br

c

R

R

Halogenase 1

Halogenase 2

Br

[Pd] Ar-B(OH)2

R

R

N H

N H

Biocat. Chemocat.

Membrane

N H

R

R

CH positions beyond scope of current arylation methods

[Pd] Ar-B(OH)2

R

N H

Br

N H

New arylation pathways via catalyst control

Figure 1 | Previous work on CH activation and proposed scheme of catalyst control using biohalogenation and Pd-catalysed cross-couplings.

(a) Previous work on CH activation utilizing substrate control of regioselectivity. (b) Catalytic cycle of halogenase biocatalysis using a Fl-Hal, along with

avin reductase (Fre) and glucose dehydrogenase (GDH) enzymes for cofactor recycling. (c) Proposed scheme of regioselective CH activation using

catalyst control by integration of a Fl-Hal and palladium-catalysed SMC. Biocat., biocatalysis; chemocat., chemocatalysis.

2 NATURE COMMUNICATIONS | 7:11873 | DOI: 10.1038/ncomms11873 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms11873 ARTICLE

products2429. Several fungal Fl-Hals have also been reported that chlorinate phenolic natural products3032. Together this suite of enzymes have the potential to regioselectively halogenate a range of aromatic scaffolds. With appropriate cofactor-recycling systems (Fig. 1b) Fl-Hals can afford halogenation reactions using benign inorganic halide (typically MgCl2 or NaBr), oxygen (from air), and glucose or isopropanol as the only stoichiometric reagents33,34. Studies on broadening the substrate scope of the tryptophan halogenases RebH, PrnA and PyrH3336, as well as improving their productivity and scaleability37, have meant that Fl-Hals are becoming increasingly viable as biocatalysts for aromatic halogenation. In addition, Fl-Hals have been used to halogenate natural products and synthetic substrates, which have been isolated or extracted and then derivatized in separate cross-coupling reactions under standard synthetic operating procedures3840. Biohalogenation is also attractive, given that traditional aromatic halogenation chemistry utilizes deleterious reagents, lacks regiocontrol and can afford mixtures, including halogenated by-products that can be toxic and persistent in the environment.

We anticipated that the unique regiocontrol inherent to Fl-Hal enzymatic halogenation could set-up an overall catalyst-controlled CH arylation system, if a working SMC could be successfully integrated under aqueous conditions. The indole heterocycle has been central to the development of contemporary CH activation chemistry, driving the discovery of new chemocatalytic methods for functionalizing the C2 and C3 positions (the innate positions for functionalization with an electrophile)4146. Direct CC bond formation at the arene C6H and C7H positions, however, is not currently possible using chemocatalysissomething we hoped to achieve with the integrated approach.

ResultsBiocatalytic halogenation. To effect one-pot Fl-Hal-SMC reactions (Fig. 1c), we required enzymes and substrates that afford high yields of structurally diverse brominated products. First, we exploited our previous ndings33 that the tryptophan 5-halogenase PyrH can accept anthranilamide (1) to develop conditions for effective enzymatic synthesis of 5-bromoanthranilamide (2) (Fig. 2). Given the exquisite natural regio-complementarity of the avin-dependent tryptophan halogenases we then set out to identify potential indole substrates for biocatalyst-controlled bromination and subsequent arylation. We found that tryptophol (3) could be efciently brominated by RebH (a tryptophan 7-halogenase) to afford 7-bromotryptophol (4), after optimizing the conditions (Supplementary Fig. 1) that had been previously reported for chlorination of this substrate34. Initial attempts to brominate the 5-position of tryptophol (3) using PyrH in place of RebH gave very low conversion. The larger substrate 3-indolepropionic acid (5) however, was transformed by PyrH to give 5-bromo-3-indolepropionate (6) as a single regioisomer in good yield. In addition, the 6-bromo-3-indolepropionate (7) could be obtained in good regioisomeric excess and yield by using the tryptophan 6-halogenase SttH27. Collectively, these three halogenase enzymes established a clean, regiodivergent synthesis of C5, C6 and C7 bromoindoles.

Finally, to extend the substrate diversity further we over-produced the radicicol halogenase RadH enzyme from the fungus Chaetomium chiversii30. RadH is similar in sequence to Rdc2a related fungal halogenase that had been shown previously to chlorinate 6-hydroxy-isoquinoline (8)31. Accordingly conditions were developed whereby RadH could be used to prepare 5-bromo-6-hydroxyisoquinoline (9) in good yields.

Cross-coupling optimization. We next sought conditions whereby the SMC could be integrated with the biocatalysts, using the coupling of 5-bromoanthranilamide (2) with phenyl boronic acid as a test reaction (Fig. 3). On the basis of previous studies of SMCs in aqueous media with an open atmosphere47,48, a number of pyrimidine and guanidine ligands were selected for screening (L1L4). Of these, L2 afforded highest conversion to 5-phenylanthranilamide (10) at 2.0 mM substrate (2) concentration, 2.5 mol% Pd loading, in potassium phosphate buffer in the absence of any cofactors or enzymes. In addition, a number of water-soluble phosphine ligands were also screened (L5L8). Of these, tppts (L6) and sulfonated SPhos (L5) gave full conversion to the desired product (10). These two ligands were therefore carried forward in a base and palladium source screen (Supplementary Fig. 2). The reaction was found to be tolerant of many bases, including DIPEA, whilst the lack of any base afforded no product. In addition, conducting the reaction using L5 and L6 under air as opposed to inert atmosphere afforded no conversion, whilst the same reaction using L2 under open air afforded good conversion.

The tolerance of the selected Pd catalysts towards the biocatalytic halogenation conditions was then probed. As NAD /NADH and FAD/FADH2 are rich in heteroatoms and redox active they could possibly interfere with Pd-catalysis either through redox chemistry or coordination to metal species. In addition, it has been previously reported that soft donor atoms (for example, sulfur) on the surface of proteins can coordinate to transition metals potentially inhibiting their catalytic activity. PdII is also known to react with glucose under certain conditions (utilized to recycle NADH). The test reaction (Fig. 3) was therefore run in the presence of each of these components to

Br

PyrH

1 2 (91%)

3 4 (92%)

NH2

CONH2

NH2

CONH2

OH

OH

RebH

N H

N H

Br

Br

Br

5

7

CO2H

CO2H

PyrH N H

6 (87%)

5

N H

CO2H

SttH

6

N H

7 (72%)

Br

HO RadH

N

HO

N

8 9 (77%)

Figure 2 | The regioselective and regio-divergent bromination of a range

of aromatic scaffolds. Regioselective halogenation of anthranilamide (1),

tryptophol (3), 3-indole propionate (5) and 6-hydroxy isoquinoline (8)

using a range of Fl-Hals to access different regiochemistries.

NATURE COMMUNICATIONS | 7:11873 | DOI: 10.1038/ncomms11873 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 3

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms11873

Br

Ph

[Pd]

2

10

PhB(OH)2

NH2

CONH2

NH2

CONH2

SO2Na

NH2

HO OH

N N

MeO OMe

N N

NMe2

HO OH

H2N

NH

N N

NH N

L1 L2 L3 L4

PCy2

SO3H

P t-Bu3P

NMe3Cl Cy2P

SO3H

L5

NaO3S

L6 L7 L8

3

Ligand Temperature (C) Conversion

L1 L2 L3 L4 L5 L6 L7 L8

80 80 80 80 50 50 50 50

32

32

86a

56

100 100b n.c. n.c.

Figure 3 | Ligand screening for the SMC of 2 with phenyl boronic acid. Pd(OAc)2 was used as the palladium source in each case in a 2:1 ligand:Pd ratio.

Conversions determined by analytical high-performance liquid chromatography (HPLC) or liquid chromatography mass spectrometry (LC-MS). n.c., no

conversion. a81% isolated yield; b80% isolated yield. Further optimization results are available in the Supplementary Material (Supplementary Fig. 2).

determine their effect on SMC conversion. This revealed a somewhat deleterious effect from the presence of the cofactors and a major reduction in cross-coupling yields in the presence of proteins (Supplementary Fig. 3). Increasing the loading of palladium catalyst was generally advantageous in terms of conversion in the presence of NADH, FAD and glucose, but did not enhance cross-coupling when either the reductase or dehydrogenase proteins were present (Supplementary Fig. 4).

Halogenase SMC integration. Given that the presence of proteins caused the greatest reduction in cross-coupling yields, we reasoned that the enzymes would need to be compartmentalized or removed before cross-coupling, at least in the rst instance. Pleasingly, we found that passing the PyrH biotransformation of anthranilamide (1) through a 10-kDa molecular-weight cut-off (MWCO) cellulose membrane to remove protein before cross-coupling was effective in this regard, affording very good conversion to arylated product (10). Following further optimization around this process (Supplementary Fig. 5), we could achieve an overall isolated yield of 79% of arylated product (10) using tppts ligand L6, Pd(OAc)2 (50 mol%), and an excess of phenyl boronic acid (Fig. 4). Moreover, we showed that the enzymes that had been removed using the membrane lter could be recycled affording intermediate bromide (2) in 62% and 42% in the second and third cycles, respectively. Subsequent arylation of the combined biotransformation ltrates afforded arylated product (10)

in 54% overall yield; a ca. twofold increase in the amount of material that would be obtained from the same amount of enzyme in a single arylation reaction. After determining the scope of this method with respect to boronic acids (Supplementary Fig. 6) and nding the SMC compatible with the conditions required for RadH-catalysed bromination of 8 (Supplementary Fig. 7), we then extended this membrane ltration method (method A) to the selective 5-arylation of 6-hydroxyisoquinoline (8), using RadH to produce 1517 in good yields.

With a workable integrated arylation process in hand, we looked at ways to improve our method, principally with a view to increasing scale and reducing the high Pd content in the SMC reaction. The requirement for high catalyst loadings appeared to arise from a combination of dilute reaction volumes and inhibitory effect of cofactors present in the reaction medium. Indeed, other attempts to conduct palladium-catalysed chemistry in biological media and in the presence of proteins have required very high loading or suffered from poor yields47,48. Enhancing enzyme efciency would mitigate these two issues, by affording higher concentrations of aryl bromide for cross-coupling. The group of Sewald have recently reported improvements in RebH productivity through the use of crosslinked enzyme aggregates (CLEAs)37. The individual subunits of RebH are crosslinked with each other, as well as the partner reductase and dehydrogenase proteins, via surface lysines using glutaraldehyde forming a RebH-reductase-dehydrogenase CLEA that is more efcient and tractable to handle37. Accordingly, we successfully prepared these

4 NATURE COMMUNICATIONS | 7:11873 | DOI: 10.1038/ncomms11873 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms11873 ARTICLE

Membrane

R

Br

[Pd] Ar-B(OH)2

R

Halogenase, Fre, (GDH2),

FAD, NADH, O2, NaBr

R

R

Biocat. Chemocat.

NO2

N H

F

S

PyrH 10 84%b (79%)a

PyrH 11 81%b

PyrH 12 73%b

PyrH 13 71%b

NH2

CONH2

NH2

CONH2

NH2

CONH2

NH2

CONH2

NO2

F

PyrH 14 65%b

RadH 15 66%a

RadH 16 64%a

RadH 17 78%a

HO

HO

HO

N

N

N

NH2

CONH2

OH

OH

OH

N H

OH

RebH 18 75%b

N H

N H

RebH 19 75%b

RebH 20 71%b

RebH 21 68%b

N

NO2

F

CO2H CO2H

N H

OH

RebH 22 54%b

N H

N H

PyrH 23 68%b

SttH 24 75%c

Figure 4 | Isolated yield of the regioselective and regiodivergent arylation. aObtained using puried enzymes and a MWCO membrane (method A).

bObtained using a CLEA of the appropriate Fl-Hal and 10 mol% Pd-catalyst loading (method B). cObtained from three successive cycles of halogenation

using the same Fl-Hal. Biocat., biocatalysis; chemocat., chemocatalysis.

materials as active heterogeneous biocatalysts containing either PyrH, SttH or RebH. Although active CLEAs of the phenolic halogenase RadH were also obtained, they did not prove to be signicantly more efcient than puried enzyme.

Initial screening of the PyrH-CLEA with anthranilamide (1) demonstrated improved efciency, affording up to 3.0 mM aryl bromide. This enabled one-pot halogenase-SMC reactions to proceed to reasonable conversion (ca. 50%) without the need to remove or compartmentalize the CLEA from the Pd catalyst and other reagents, presumably as many heteroatoms of the protein had been effectively protected during the crosslinking process (Supplementary Fig. 8). We also found that the arylation of tryptophol (3) could be carried out in the same one-pot manner, with the RebH-CLEA and the Pd catalyst both present throughout (Supplementary Fig. 8). However, highest conversions

were obtained when the CLEA was removed, by centrifugation or ltration, before cross-coupling of the supernatant from biocatalytic halogenation (Supplementary Figs 8 and 9). The scope of this method (method B) was found to be broad with respect to boronic acids (Supplementary Fig. 10), allowing the efcient arylation of anthranilamide (1) with both electron-rich, electon-poor and heterocyclic boronic acids using PyrH-CLEA with reduced palladium loading (10 mol%), affording the 5-arylated compounds in very good yield (1014, Fig. 4). Tryptophol (3) was also efciently coupled using electron-rich, electron-poor, heterocyclic and alkenyl boronic acids to afford 7-substituted products 18 to 22 in good yield with RebH-CLEA at lower Pd loading. 3-Indole propionate (5) was likewise effectively arylated to biaryl products 23 and 24, with PyrH- and SttH-CLEAs, respectively. Along with the tryptophol examples

NATURE COMMUNICATIONS | 7:11873 | DOI: 10.1038/ncomms11873 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 5

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms11873

Glass PDMS

[Pd]and ArB(OH)2

Br

OH

OH

2

10

18

Br

N H

N H

NH2

CONH2

74%a (59%)b

57%c

PyrH

NH2

CONH2

NH2

CONH2

1

[Pd] ArB(OH)2

O

NH2

CONH2

Ar

25

26

64%c

61%c

Enzymes NADH, FAD

NH2

CONH2

NH2

CONH2

Figure 5 | PDMS compartmentalization of the regioselective halogenation cross-coupling cascade. aObtained using puried enzyme and 10 mol% of

L2.Pd(OAc)2. bObtained using halogenase CLEA and 2 mol% of L2.Pd(OAc)2. cObtained using halogenase CLEA and 10 mol% of L2.Pd(OAc)2.

(1822), this demonstrates the efcient and highly selective manipulation of each of the 5-, 6- and 7- CH positions of the indole nucleus, under catalyst control, which is unattainable using existing synthetic methodology. In addition, the heterogeneous nature of the CLEA biocatalyst allowed it to be recycled over repeated cycles of bromination, with little loss in efciency (Supplementary Fig. 11). In the case of the 6-arylation of 3-indolepropionic acid (5), this allowed an overall yield of 75% of 24 over three cycles with the SttH-CLEA. The increased stability and ease of preparation of the heterogeneous biocatalyst allowed this methodology to be scaled up to a 1.0-mmol scale to afford reasonable yields of 10 (4100 mg, 52%).

Polydimethylsiloxane compartmentalization. To further enhance the scope and efciency of the halogenase-SMC process, we looked for alternative ways of separating the proteins from the palladium catalyst. While membrane ltration worked well, it did introduce an additional operation in the middle of the process and necessitated a deoxygenation step before SMC. To better streamline the process and develop a more efcient one-pot procedure, we examined polydimethylsiloxane (PDMS) thimbles as a compartmentalization strategy. This approach had been reported for a number of reactions that require the separation of incompatible reagents19,49, including the combination of a Pd-catalyzed oxidation with a biocatalytic reduction19.

Typically, one of the reactions is assembled inside a small thimble made of PDMS, which is then placed inside the larger bulk reaction vessel. As PDMS is a hydrophobic polymer, non-polar compounds (that is, substrates and intermediates in the cascade) should dissolve sufciently well to pass through the PDMS walls, whilst charged reagents (Pd catalysts, enzymes and cofactors) would be unable to diffuse through PDMS and might therefore be compartmentalized. Initial tests showed that 5-bromoanthranilamide (2) would ux through PDMS well while glucose, NADH, FAD and the L2:Pd(OAc)2 complex were all unable to effectively penetrate PDMS (Supplementary Fig. 12). SMC conditions using L2:Pd(OAc)2 were therefore chosen to test the PDMS compartmentalization approach, as this catalyst functions well under open air and because alternative phosphine ligands have been shown to promote ux through PDMS.

Accordingly conditions were optimized, which allowed the PyrH-halogenation of anthranilamide (2) using pure protein to be carried out with the cross-coupling reagents within the PDMS thimble (Fig. 5), resulting in a one-pot reaction to give 5-phenylanthranilamide (10) in 74% yield. The immediate advantages of this approach (method C) are that protein removal and deoxygenation of the reaction buffer are not required. Moreover a lower catalyst loading (10 mol%) is achievable under this regime, compared with the removal of pure protein using ltration (50 mol%), due to the compartmentalization of cofactors and enzymes from the Pd catalyst. By using a PyrH-CLEA to afford higher concentration of aryl bromide in conjunction with PDMS compartmentalization (method D), the Pd loading was further reduced to 2 mol% to afford 10 in good yield (59%). This combination also allowed the coupling of deactivated, electron-rich, boronic acids such as 3-furyl boronic acid to give 25 in good yield (64%), a reaction that was unsuccessful using a cellulose membrane (MWCO lter). Finally, this method was also extended to include coupling of the 7-position of tryptophol (3), using RebH-CLEA, to give 7-pent-1-enyl tryptophol (18) and 40-tertbutyl-7-phenyl tryptophol (26) in good yields.

DiscussionIn summary, we have combined four regioselective avin-dependent halogenase enzymes with a palladium-catalysed SuzukiMiyaura cross-coupling to afford the regio-controlled arylation of a number of aromatic scaffolds (benzamides, isoquinolines and indoles). In each case, the regioselective CH arylation transformations were previously inaccessible using stand-alone chemocatalysis or biocatalysis methods. Specically, we report the selective CH arylation of the 5-, 6- and 7- positions of the indole nucleus under catalyst control.

To overcome issues of catalyst incompatibility, we initially developed a ltration method using a 10-kDa MWCO lter (method A). This allowed efcient biocatalyst recycling for a number of cycles, affording approximately double the overall productivity that the same batch would in a single cycle. Biocatalyst efciency was further improved by using CLEAs, which afforded higher aryl bromide concentrationenabling lower palladium loading to be employed (method B). We also

6 NATURE COMMUNICATIONS | 7:11873 | DOI: 10.1038/ncomms11873 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms11873 ARTICLE

developed a PDMS thimble system to compartmentalize the chemo- and bio-catalysts. This method afforded similar yields of regioselectively arylated product as the MWCO lter system, but without the need for protein removal or solution deoxygenation before the cross-coupling chemistry (method C). Finally, the implementation of CLEAs with PDMS compartmentalization led to a further reduction in the required Pd loading down to 2 mol% (method D). Combined with directed evolution and other research aimed at improving the productivity and scalability of halogenase enzymes37,50, the integrated method described here sets out a pathway to new CH activation transformations that cannot be achieved through stand-alone biocatalysis or chemocatalysis. Research in this area is ongoing in our laboratories.

Methods

Method A: regioselective arylation 6-hydroxyisoquinoline. 6-hydroxyisoquinoline (8) (0.5 mM), NaBr (10 mM), FAD (1 mM), Fre (4 mM), RadH (25 mM) and NADH(2.5 mM) were added to 10 mM potassium phosphate buffer (pH 7.2) containing 1% v/v EtOH (40 ml) and incubated with shaking overnight (30 C) before ltration through a 10-kDa MWCO lter (Vivaspin 20). CsF (4.8 eq) and boronic acid (30 eq) were then added to the ltrate. Following freeze-thaw degassing and backll with N2,

tppts (0.5 mM) and Na2PdCl4 (0.25 mM) were added as deoxygenated stock solutions in water. The resultant solution was heated to 80 C with stirring for 24 h. After cooling, pH was adjusted to 6 using HCl, and the reaction partitioned into CH2Cl2 before the removal of solvent in vacuo and purication.

Method B: regioselective arylation using CLEAs. Substrate (3.0 mM), FAD (10 mM), NaBr (30 mM), NADH (100 mM) were dissolved in 15 mM sodium phosphate buffer with 5% v/v isopropanol (30 ml, pH 7.4). Halogenase CLEA from1.5 l culture (prepared as described on page S46 of the Supplementary Methods) was then resuspended into the reaction buffer and the resultant suspension incubated at room temperature with orbital shaking overnight. After removal of CLEA by centrifugation (10,000 r.p.m., 4 C, 30 min), K3PO4/K2CO3 (10 eq) and boronic acid (5 eq) were added to the supernatant before degassing of the solution with a stream of nitrogen. Tppts (20 mol%) and Na2PdCl4 (10 mol%) were then added and the solution was stirred at 80 C overnight. On cooling, the reaction was partitioned into EtOAc/CH2Cl2 and concentrated before purication.

Method C: regioselective arylation using PDMS thimbles. To a solution containing anthranilamide (2.0 mM), NaBr (100 mM), FAD (1 mM), glucose(20 mM), PyrH (20 mM), Fre (2 mM) and GDH2 (12 mM) in 10 mM potassium phosphate buffer (30 ml, pH 7.2) was added NADH (100 mM). A PDMS thimble containing L2.Pd(OAc)2 (10 mol%), CsF (10 eq) and PhB(OH)2 (5 eq) in water was then placed inside the Erlenmeyer ask containing the biotransformation. After incubation at room temperature overnight, the reaction was heated to 80 C for 8 h. After cooling, the inner and outer chambers were combined and basied with 4 N NaOH. The thimble was then soaked in EtOAc to ux out any residual product, and the reaction mixture extracted into EtOAc before concentration and purication.

Method D: regioselective arylation using CLEAs and PDMS. Halogenase CLEA from 3 l of culture (prepared as described on page S46 of the Supplementary Methods) was suspended into 30 ml of reaction buffer containing anthranilamide (1) or tryptophol (3) (5.0 mM), NADH (100 mM), FAD (10 mM) and NaBr(30 mM) in 15 mM sodium phosphate buffer with 5% v/v isopropanol (30 ml, pH 7.4). To this suspension was added a PDMS thimble containing L2.Pd(OAc)2 (10 mol%), CsF (10 eq) and ArB(OH)2 (5 eq). After incubation with shaking at room temperature overnight, the reaction was heated to 80 C for a further 24 h. Reactions were then worked up as previously described. To facilitate the recycling of the biocatalysts, the CLEA can be removed using centrifugation (10,000 r.p.m., 30 min., 4 C) after overnight incubation at room temperature before addition of the PDMS thimble containing the SMC components.

Data availability. The authors declare that the data supporting the ndings of this study are available within the article and its Supplementary Information Files.

References

1. Yu, J.-Q. & Shi, Z. C-H Activation Vol. 292 (Springer, 2010).2. Clot, E. et al. CF and CH bond activation of uorobenzenes and uoropyridines at transition metal centers: how uorine tips the scales. Acc. Chem. Res. 44, 333348 (2011).

3. Rossi, R., Bellina, F., Lessi, M. & Manzini, C. Cross-coupling of heteroarenes by CH Functionalization: recent progress towards direct arylation and

heteroarylation reactions involving heteroarenes containing one heteroatom. Adv. Synth. Catal. 356, 17117 (2014).4. Engle, K. M., Mei, T. S., Wasa, M. & Yu, J. Q. Weak coordination as a powerful means for developing broadly useful CH functionalization reactions. Acc. Chem. Res. 45, 788802 (2012).

5. Kuhl, N., Hopkinson, M. N., Wencel-Delord, J. & Glorius, F. Beyond directing groups: transition-metal-catalyzed CH activation of simple arenes. Angew. Chem. Int. Ed. 51, 1023610254 (2012).

6. Leow, D., Li, G., Mei, T. S. & Yu, J. Q. Activation of remote meta-CH bonds assisted by an end-on template. Nature 486, 518522 (2012).

7. Wang, X. C. et al. Ligand-enabled meta-CH activation using a transient mediator. Nature 519, 334338 (2015).

8. Saidi, O. et al. Ruthenium-catalyzed meta sulfonation of 2-phenylpyridines.J. Am. Chem. Soc. 133, 1929819301 (2011).9. McNally, A., Haffemayer, B., Collins, B. S. L. & Gaunt, M. J. Palladium-catalysed CH activation of aliphatic amines to give strained nitrogen heterocycles. Nature 512, 338338 (2014).

10. Ueda, K., Yanagisawa, S., Yamaguchi, J. & Itami, K. A general catalyst for the beta-selective CH Bond arylation of thiophenes with iodoarenes. Angew. Chem. Int. Ed. 49, 89468949 (2010).

11. Hickman, A. J. & Sanford, M. S. Catalyst control of site selectivity in the Pd-II/IV-catalyzed direct arylation of naphthalene. ACS Catal. 1, 170174 (2011).

12. Neufeldt, S. R. & Sanford, M. S. Controlling site selectivity in palladium-catalyzed CH bond functionalization. Acc. Chem. Res. 45, 936946 (2012).

13. Gormisky, P. E. & White, M. C. Catalyst-controlled aliphatic CH oxidations with a predictive model for site-selectivity. J. Am. Chem. Soc. 135, 1405214055 (2013).

14. Grger, H. & Hummel, W. Combining the two worlds of chemocatalysis and biocatalysis towards multi-step one-pot processes in aqueous media. Curr. Opin. Chem. Bio. 19, 171179 (2014).

15. Lohr, T. L. & Marks, T. J. Orthogonal tandem catalysis. Nat. Chem. 7, 477482 (2015).

16. Allen, J. V. & Williams, J. M. J. Dynamic kinetic resolution with enzyme and palladium combinations. Tetrahedron Lett. 37, 18591862 (1996).

17. Larsson, A. L. E., Persson, B. A. & Backvall, J.-E. Enzymatic resolution of alcohols coupled with ruthenium-catalyzed racemization of the substrate alcohol. Angew. Chem. Int. Ed. 36, 12111212 (1997).

18. Denard, C. A. et al. Cooperative tandem catalysis by an organometallic complex and a metalloenzyme. Angew. Chem. Int. Ed. 53, 465469 (2014).

19. Sato, H., Hummel, W. & Grger, H. Cooperative catalysis of noncompatible catalysts through compartmentalization: Wacker oxidation and enzymatic reduction in a one-pot process in aqueous media. Angew. Chem. Int. Ed. 54, 44884492 (2015).

20. Makkee, M., Kieboom, A. P. G., Van Bekkum, H. & Roels, J. A. Combined action of enzyme and metal catalyst, applied to the preparation of D-mannitol.J. Chem. Soc. Chem. Commun. 19, 930931 (1980).21. Khler, V. et al. Synthetic cascades are enabled by combining biocatalysts with articial metalloenzymes. Nat. Chem. 5, 9399 (2013).

22. Wang, Z. J., Clary, K. N., Bergman, R. G., Raymond, K. N. & Toste, F. D. A supramolecular approach to combining enzymatic and transition metal catalysis. Nat. Chem. 5, 100103 (2013).

23. Tenbrink, K., Seler, M., Schatz, J. & Grger, H. Combination of olen metathesis and enzymatic ester hydrolysis in aqueous media in a one-pot synthesis. Adv. Synth. Catal. 353, 23632367 (2011).

24. Keller, S. et al. Purication and partial characterization of tryptophan 7-halogenase (PrnA) from Pseudomonas uorescens. Angew. Chem. Int. Ed. 39, 23002302 (2000).

25. Heemstra, J. R. & Walsh, C. T. Tandem action of the O2- and FADH2-dependent halogenases KtzQ and KtzR produce 6,7-dichlorotryptophan for kutzneride assembly. J. Am. Chem. Soc. 130, 1402414025 (2008).

26. Yeh, E., Garneau, S. & Walsh, C. T. Robust in vitro activity of RebF and RebH, a two-component reductase/halogenase, generating 7-chlorotryptophan during rebeccamycin biosynthesis. Proc. Natl Acad. Sci. USA 102, 39603965 (2005).

27. Zeng, J. & Zhan, J. Characterization of a tryptophan 6-halogenase from Streptomyces toxytricini. Biotechnol. Lett. 33, 16071613 (2011).

28. Zehner, S. et al. A regioselective tryptophan 5-halogenase is involved in pyrroindomycin biosynthesis in Streptomyces rugosporus LL-42D005. Chem. Biol. 12, 445452 (2005).

29. Wagner, C., El Omari, M. & Knig, G. M. Biohalogenation: natures way to synthesize halogenated metabolites. J. Nat. Prod. 72, 540553 (2009).

30. Wang, S. et al. Functional characterization of the biosynthesis of radicicol, an Hsp90 inhibitor resorcylic acid lactone from Chaetomium chiversii. Chem. Biol. 15, 13281338 (2008).

31. Zeng, J., Lytle, A. K., Gage, D., Johnson, S. J. & Zhan, J. Specic chlorination of isoquinolines by a fungal avin-dependent halogenase. Bioorg. Med. Chem. Lett. 23, 10011003 (2013).

NATURE COMMUNICATIONS | 7:11873 | DOI: 10.1038/ncomms11873 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 7

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms11873

32. Zhou, H., Qiao, K., Gao, Z., Vederas, J. C. & Tang, Y. Insights into radicicol biosynthesis via heterologous synthesis of intermediates and analogs. J. Biol. Chem. 285, 4141241421 (2010).

33. Shepherd, S. A. et al. Extending the biocatalytic scope of regiocomplementary avin-dependent halogenase enzymes. Chem. Sci. 6, 34543460 (2015).

34. Payne, J. T., Andorfer, M. C. & Lewis, J. C. Regioselective arene halogenation using the FAD-dependent halogenase RebH. Angew. Chem. Int. Ed. 52, 52715274 (2013).

35. Hlzer, M., Burd, W., Reiig, H.-U. & Pe, K.-H. V. Substrate specicity and regioselectivity of tryptophan 7-halogenase from Pseudomonas uorescens BL915. Adv. Synth. Catal. 343, 591595 (2001).

36. Frese, M., Guzowska, P. H., Vo, H. & Sewald, N. Regioselective enzymatic halogenation of substituted tryptophan derivatives using the FAD-dependent halogenase RebH. ChemCatChem. 6, 12701276 (2014).

37. Frese, M. & Sewald, N. Enzymatic halogenation of tryptophan on a gram scale. Angew. Chem. Int. Ed. 54, 298301 (2015).

38. Glenn, W. S., Nims, E. & OConnor, S. E. Reengineering a tryptophan halogenase to preferentially chlorinate a direct alkaloid precursor. J. Am. Chem. Soc. 133, 1934619349 (2011).

39. Roy, A. D., Grschow, S., Cairns, N. & Goss, R. J. M. Gene expression enabling synthetic diversication of natural products: chemogenetic generation of pacidamycin analogs. J. Am. Chem. Soc. 132, 1224312245 (2010).

40. Durak, L. J., Payne, J. T. & Lewis, J. C. Late-stage diversication of biologically active molecules via chemoenzymatic CH activation. ACS Catal. 6, 14511454 (2016).

41. Sandtorv, A. H. Transition metal-catalyzed CH activation of indoles. Adv. Synth. Catal. 357, 24032435 (2015).

42. Itahara, T., Ikeda, M. & Sakakibara, T. Alkenylation of 1-acylindoles with olens bearing electron-withdrawing substituents and palladium acetate.J. Chem. Soc. Perkin. Trans. 1, 13611363 (1983).43. Grimster, N. P., Gauntlett, C., Godfrey, C. R. A. & Gaunt, M. J. Palladium-catalyzed intermolecular alkenylation of indoles by solvent-controlled regioselective CH functionalization. Angew. Chem. Int. Ed. 44, 31253129 (2005).

44. Stuart, D. R. & Fagnou, K. The catalytic cross-coupling of unactivated arenes. Science 316, 11721175 (2007).

45. Kalyani, D., Dick, A. R., Anani, W. Q. & Sanford, M. S. A simple catalytic method for the regioselective halogenation of arenes. Org. Lett. 8, 25232526 (2006).

46. Lebrasseur, N. & Larrosa, I. Room temperature and phosphine free palladium catalyzed direct C-2 arylation of indoles. J. Am. Chem. Soc. 130, 29262927 (2008).

47. Chalker, J. M., Wood, C. S. C. & Davis, B. G. A convenient catalyst for aqueous and protein Suzuki-Miyaura cross-coupling. J. Am. Chem. Soc. 131, 1634616347 (2009).

48. Gao, Z., Gouverneur, V. & Davis, B. G. Enhanced aqueous Suzuki-Miyaura coupling allows site-specic polypeptide 18F-labeling. J. Am. Chem. Soc. 135, 1361213615 (2013).

49. Runge, M. B., Mwangi, M. T., Miller, A. L., Perring, M. & Bowden, N. B. Cascade reactions using LiAlH4 and grignard reagents in the presence of water.

Angew. Chem. Int. Ed. 47, 935939 (2008).50. Poor, C. B., Andorfer, M. C. & Lewis, J. C. Improving the stability and catalyst lifetime of the halogenase RebH by directed evolution. ChemBioChem. 15, 12861289 (2014).

Acknowledgements

We acknowledge BBSRC (grant BB/K00199X/1), EPSRC, CoEBio3 and GlaxoSmithKline for support. We also thank Dr Anna-Winona Struck for helpful conversations.

Author contributions

J.M. and M.F.G. conceived and supervised the project; J.L., J.-M.H. and H.H.S. performed integrated arylation reactions; J.L. optimized the integrated process whilst J.L. and H.H.S. screened conditions for the SMC; B.R.K.M. and S.A.S. developed expression and initial reaction conditions for RadH and SttH, respectively; J.L. and S.A.S. prepared functional CLEAS; J.L. prepared and carried out integrated reactions with PDMS thimbles;

J.L., M.F.G. and J.M. wrote the manuscript.

Additional information

Supplementary Information accompanies this paper at http://www.nature.com/naturecommunications

Web End =http://www.nature.com/ http://www.nature.com/naturecommunications

Web End =naturecommunications

Competing nancial interests: The authors declare no competing nancial interests.

Reprints and permission information is available online at http://npg.nature.com/reprintsandpermissions/

Web End =http://npg.nature.com/ http://npg.nature.com/reprintsandpermissions/

Web End =reprintsandpermissions/

How to cite this article: Latham, J. et al. Integrated catalysis opens new arylation pathways via regiodivergent enzymatic CH activation. Nat. Commun. 7:11873 doi: 10.1038/ncomms11873 (2016).

This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the articles Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/

Web End =http://creativecommons.org/licenses/by/4.0/

8 NATURE COMMUNICATIONS | 7:11873 | DOI: 10.1038/ncomms11873 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications