Synthetic cascades are enabled by combining biocatalysts with artificial metalloenzymes

Article abstract of DOI:10.1038/nchem.1498

Enzymatic catalysis and homogeneous catalysis offer complementary means to address synthetic challenges, both in chemistry and in biology. Despite its attractiveness, the implementation of concurrent cascade reactions that combine an organometallic catalyst with an enzyme has proven challenging because of the mutual inactivation of both catalysts. To address this, we show that incorporation of a d 6 -piano stool complex within a host protein affords an artificial transfer hydrogenase (ATHase) that is fully compatible with and complementary to natural enzymes, thus enabling efficient concurrent tandem catalysis. To illustrate the generality of the approach, the ATHase was combined with various NADH-, FAD- and haem-dependent enzymes, resulting in orthogonal redox cascades. Up to three enzymes were integrated in the cascade and combined with the ATHase with a view to achieving (i) a double stereoselective amine deracemization, (ii) a horseradish peroxidase-coupled readout of the transfer hydrogenase activity towards its genetic optimization, (iii) the formation of L-pipecolic acid from L-lysine and (iv) regeneration of NADH to promote a monooxygenase-catalysed oxyfunctionalization reaction.

Full text of DOI:10.1038/nchem.1498

ARTICLES
PUBLISHED ONLINE: 25 NOVEMBER 2012 | DOI: 10.1038/NCHEM.1498
Synthetic cascades are enabled by combining
biocatalysts with artificial metalloenzymes
1
1
1
2
3
1
1
¨
¨
¨
V. Kohler , Y. M. Wilson , M. Durrenberger , D. Ghislieri , E. Churakova , T. Quinto , L. Knorr ,
1
3
2
D. Haussinger , F. Hollmann , N. J. Turner and T. R. Ward1
¨
*
*
*
Enzymatic catalysis and homogeneous catalysis offer complementary means to address synthetic challenges, both in
chemistry and in biology. Despite its attractiveness, the implementation of concurrent cascade reactions that combine an
organometallic catalyst with an enzyme has proven challenging because of the mutual inactivation of both catalysts. To
address this, we show that incorporation of a d6-piano stool complex within a host protein affords an artificial transfer
hydrogenase (ATHase) that is fully compatible with and complementary to natural enzymes, thus enabling efficient
concurrent tandem catalysis. To illustrate the generality of the approach, the ATHase was combined with various NADH-,
FAD- and haem-dependent enzymes, resulting in orthogonal redox cascades. Up to three enzymes were integrated in the
cascade and combined with the ATHase with a view to achieving (i) a double stereoselective amine deracemization, (ii) a
horseradish peroxidase-coupled readout of the transfer hydrogenase activity towards its genetic optimization, (iii) the
formation of L-pipecolic acid from L-lysine and (iv) regeneration of NADH to promote a monooxygenase-catalysed
oxyfunctionalization reaction.
ellular biochemistry requires the orchestration of metabolic
Results and discussion
pathways in which many enzyme-catalysed processes are Double stereoselective deracemization of amines. In recent years,
able to function simultaneously, resulting in the production evolved monoamine oxidases (MAO-N from Aspergillus niger) have
C
of a wide range of primary and secondary metabolites within the found widespread applications in the synthesis of enantiopure
cell. In an attempt to construct artificial cells using the principles amines33,34. For this purpose, a highly (S)-selective MAO is combined
34
.
of synthetic biology, compartmentalization of cellular processes with a stoichiometric reducing agent (for example, H3N BH3) or
will need to be mimicked to allow cascade reactions to take place with a heterogeneous reduction catalyst24. Efforts to combine MAO
in parallel in an efficient manner1–6. Whereas enzymes have with
a homogeneous imine reduction catalyst have proven
evolved in concert and in complex media, and are therefore well challenging thus far: indeed, on combining the transfer
suited for cascade reactions, compatibility problems and mutual hydrogenation catalyst [Cp*Ir(Biot-p-L)Cl] with MAO, we observed
inactivation are often encountered upon combining chemocatalysts mutual inactivation (Fig. 1). Only modest levels of amine could be
with biocatalysts7–9. Such incompatibility may be circumvented by detected starting from the imine, nor could imine be detected in
performing cascades in sequential steps or by site-isolation of the significant amounts starting from the racemic amine (Table 1,
individual catalysts through immobilization, heterogeneous or entries 1, 2). These observations suggest that the biocatalyst and
biphasic reaction conditions, encapsulation, and so on10–24
.
chemocatalyst are incompatible. In the chemo-enzymatic dynamic
Recently, we have described one approach to cellular compart- kinetic resolution of alcohols and amines, as pioneered by Ba¨ckvall,
mentalization in which an Escherichia coli cell was engineered to inactivation may be circumvented by performing the cascade reaction
simultaneously express an intracellular enzyme (monoamine in an organic solvent21,22. This strategy effectively provides a phase
oxidase) and also bind palladium nanoparticles in its outer mem- separation between the highly robust and enantioselective enzyme
brane, thereby allowing efficient chemo-enzymatic deracemization (for example, CALB) and the organometallic racemization catalyst.
of amines24. Artificial metalloenzymes resulting from the encapsula-
In a biomimetic spirit, we speculated that molecular compart-
tion of an organometallic catalyst within a protein scaffold have mentalization of the organometallic imine reduction catalyst
been shown to combine attractive features of both chemocatalysts within a protein scaffold might shield it from the biocatalyst, thus
and biocatalysts for single-step transformations25–32. In the context preventing mutual inactivation. With this goal in mind, we set out
of concurrent cascade reactions, we reasoned that the artificial to investigate the potential of cascade reactions in the presence of
cofactor may be effectively shielded by its host protein, thus prevent- ATHase based on the supramolecular incorporation of a catalyti-
ing the mutual inactivation commonly encountered when combin- cally active [Cp*Ir(biot-p-L)Cl] complex within streptavidin (Figs
ing an organometallic catalyst with an enzyme (Fig. 1). To test the 1, 2). Initially, the individual reaction steps were performed with
validity of the concept, we examined the combination of an artificial a 1-red and 1-ox (1-methyl-1,2,3,4-tetrahydroisoquinoline and
transfer-hydrogenase (ATHase) and an oxidase, a catalase, an 1-methyl-3,4-dihydroisoquinoline) couple. This led to the identifi-
.
amino-acid oxidase and a monooxygenase (Fig. 2). For this cation of [Cp*Ir(biot-p-L)Cl] Sav S112T (59% enantiomeric excess
.
purpose, we rely on the incorporation of
[Cp*Ir(Biot-p-L)Cl] complex within streptavidin (Sav hereafter) as (S)-red-1) as well as MAO-N-9 for the oxidation of (S)-red-1
ATHase using sodium formate as hydride source (Figs 1 and 2).
(Table 1, entries 3 and 4; Supplementary Tables S2,S5)35,36
a biotinylated (e.e.) (R)-red-1) and [Cp*Ir(biot-p-L)Cl] Sav S112K (46% e.e.
.
1Department of Chemistry, University of Basel, Spitalstrasse 51, CH-4056 Basel, Switzerland, 2School of Chemistry, University of Manchester, Manchester
Institute of Biotechnology, 131 Princess Street, Manchester M1 7DN, UK, 3Department of Biotechnology, Delft University of Technology, Julianalaan 136,
2628BL Delft, The Netherlands. e-mail: F.Hollmann@tudelft.nl; Nicholas.Turner@manchester.ac.uk; thomas.ward@unibas.ch
*
NATURE CHEMISTRY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturechemistry
© 2012 Macmillan Publishers Limited. All rights reserved.
1

NATURE CHEMISTRY DOI: 10.1038/NCHEM.1498
ARTICLES
MAO-N
ATHase
catalase
HCO2Na
a
O
2 (air)
or
Biocatalyst
N
NH
NH
MOPS-buffer
pH 7.8, 37 °C
Substrates
1-ox
rac-1-red
(R)-1-red
MAO-N
ATHase
catalase
HCO2Na
O2 (air)
Intermediates
Mutual
inactivation
N
N
H
N
or
Products
H
MOPS-buffer
pH 7.8, 37 °C
2-ox
rac-2-red
(R)-2-red
MAO-N
Incorporation
b
c
ATHase
catalase
HCO2Na
O2 (air)
OH
*
NHMe
+
N
N
Me
Me
TM-catalyst
MOPS-buffer
pH 7.8, 37 °C
N
N
N
(S)-3-red
(R)-3-red
3-red-ol
Sav
TM-catalyst inside Sav
LAAO
ATHase
DAAO
catalase
HCO2Na
O
NH2
O
2 (air)
HN
NH
H2N
CO2H
N
H
CO2H
H
H
MOPS-buffer
pH 7.8, 37 °C
H
N
4
5
Ir
NH2
S
Cl
S
O
N
HbpA
d
ATHase
NAD
HCO2Na
O2 (air)
[Cp*Ir(biot-p-L)Cl]
OH
HO
OH
O
O
Figure 1 | Reaction cascades resulting from combining an ATHase with a
biocatalyst. The organometallic transfer-hydrogenation catalyst [Cp*Ir(biot-
p-L)Cl] and biocatalyst suffer from mutual inactivation, thus precluding the
implementation of reaction cascades. Relying on the strength of the biotin–
streptavidin interaction, incorporation of the biotin-bearing complex
[Cp*Ir(biot-p-L)Cl] within streptavidin (Sav) affords an ATHase that is fully
compatible with and complementary to a variety of natural enzymes, thus
leading to the development of concurrent orthogonal redox cascades.
KPi-buffer
(1-decanol)
pH 7.5, 30 °C
6
7
Figure 2 | Overview of reaction cascades scrutinized in this study. a,
Reduction of prochiral imines with subsequent deracemization of cyclic
amines. b, Stereoinversion of natural nicotine, which leads partially to the
formation of the chiral alcohol 3-red-ol. c, Formation of L-pipecolic acid from
L-lysine. d, Hydroxylation of 2-hydroxybiphenyl coupled to an ATHase-
catalysed NADH regeneration process. MAO-N is a monoamine oxidase
variant from A. niger. ATHase is an artificial transfer hydrogenase variant,
formed by incorporation of [Cp*Ir(biot-p-L)Cl] into a streptavidin variant.
LAAO is L-amino acid oxidase from C. atrox. DAAO is D-amino acid
oxidase from porcine kidney. HbpA is 2-hydroxybiphenyl monooxygenase
from P. azaleica.
On incorporating [Cp*Ir(biot-p-L)Cl] in Sav, only modest ATH
activity was observed in the presence of MAO-N-9 (Supplementary
Fig. S21). We speculated that the H2O2 produced by MAO as a side
product may poison the Ir catalyst (Supplementary Table S7)37.
Upon the addition of a catalase (from bovine liver), quantitative con-
version and 99% e.e. in favour of (R)-1-red was obtained using
.
0.5 mol% [Cp*Ir(biot-p-L)Cl] Sav S112T with 0.19 mg MAO/ml.
data highlight the versatility of the double stereoselective cascade,
The orthogonal redox cascade yielded identical results on starting with the ATHase and MAO working in concert to expediently
from the racemic amine (rac-1-red) or from the prochiral imine produce the enantiopure amine.
(1-ox; Figs 2, 3; Table 1, compare entries 5 and 6). In stark contrast,
only negligible MAO activity could be observed with [Cp*Ir(biot-p- the overall enantioselectivity of the reaction; (ii) a synergistic effect
L)Cl], inthe presenceorabsence of catalase, highlightingthe beneficial
These results demonstrate the following: (i) the MAO determines
for the matched (R)-selective ATHase–(S)-selective MAO-N–catalase
effect of the Sav host protein in effectively site-isolating the reductase cascade is operative; (iii) the stereoselectivity of ATHase is unaffected
and the oxygenase (Table 1, entries 1 and 2; Supplementary Table S6). by the presence of both MAO and catalase, as highlighted by the enan-
We next investigated the influence of the Sav isoform on the rate tioselectivities at the start of the reaction (samples taken after 5 min);
of the cascade reaction. For this purpose, a moderately (S)-selective (iv) the presence of a His-tag on the MAO-N-9 does not affect the per-
.
[Cp*Ir(biot-p-L)Cl] Sav S112K ATHase was combined with MAO-
formance of the ATHase, as judged by the initial rates.
N-9. Although the reaction with Sav S112K had a faster initial rate, it
took significantly longer to reach high enantiomeric excess in favour Other amine–imine couples. To investigate the scope of the
of the (R)-enantiomer: 93% e.e. after 23 h with S112K versus 99% concurrent cascade reaction, a secondary amine with purely
e.e. after 8 h with S112T (Fig. 3b; Table 1, entries 6 and 7). The aliphatic substituents (rac-2-red) and a tertiary amine (S)-3-red,
kinetic profile reflects a combined effect of kcat, KM, Kinhib and selec- as well as their respective oxidation products (2-ox, 3-ox), were
tivity (kcat(app) ¼ 5.4 min21 [+0.2 min21];KM(app) ¼ 55 mM [+5 mM]) subjected to the above enzyme cascade including ATHase. Before
compared to the (R)-selective ATHase S112T (kcat(app) ¼ 1.09 min21 this, suitable enzymes for the individual reaction steps (reduction
[+0.05 min21], KM(app) ¼ 4.5 mM [+0.4 mM] with a Kinhib of and oxidation) were identified (see Table 1, entries 8 and 9;
104 mM [+12 mM]) (Fig. 3c, Supplementary Fig. S19). These Supplementary Tables S3 and S4).
2
NATURE CHEMISTRY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturechemistry
© 2012 Macmillan Publishers Limited. All rights reserved.

NATURE CHEMISTRY DOI: 10.1038/NCHEM.1498
ARTICLES
Table1 | Orthogonal redox cascades combining ATHase with oxidases or oxygenase.
Entry Substrate
Product
Sav-mutant
Oxidase/oxygenase Catalase Conv.* (%) e.e. (%) (abs.config)
1
–
MAO-N-9
Yes
36†
rac.
N
NH
1-ox
1-red
2
3
4
5
6
7
rac-1-red
1-ox
1-ox
rac-1-red
1-ox
1-ox
1-red
1-red
1-red
1-red
1-red
1-red
–
MAO-N-9
–
–
MAO-N-9
MAO-N-9
MAO-N-9
No
No
No
Yes
Yes
Yes
,1†
rac.
S112T
S112K
S112T
S112T
S112K
.99
.99
.99
.99
98
59 (R)
46 (S)
.99 (R)
.99 (R)†
99 (R)‡
8
S112A
–
No
84§
86 (R)
N
N
H
2-ox
2-red
9
10
11
2-ox
2-ox
rac-2-red
2-red
2-red
2-red
S112A-K121N
S112A
S112A-K121N MAO-N-9
–
No
Yes
Yes
.99§
98§
63 (R)
99 (R)
.99 (R)
MAO-N-9
99§
76ꢁ
79 (R)
O
12
S112A-K121T
–
No
MeHN
N
Me
N
3-ox
N
3-red
79}
30 (ND)
OH
13
3-ox
S112G
–
No
MeHN
3-red-ol
N
14
(S)-3-red
3-red
S112A-K121T
MAO-N-5
Yes
65#
.99 (R)
1 (R)
NH2
CO2H
15
–
LAAO
Yes
20
H2N
N
H
CO2H
5
4
16
17
4
(S)-4
5
5
S112A
S112A
LAAO
LAAO
1
Yes
Yes
89
88**
29 (S)
86 (S)
DAAO
HO
HO
OH
18
–
HbpA
No
3
NA
Ph
Ph
6
7
19
6
7
S112A
HbpA
No
.99
NA
The most relevant entries are highlighted in bold. *Conversion determined by HPLC or gas chromatography (GC). †After 8 h reaction time with 1 mol% ATHase or after 24 h with 0.5 mol%. ‡After 65 h reaction
time with 1 mol% ATHase. §Conversion determined by response of 2-ox versus 2-red. ꢁRatio of 3-red to 3-red-ol ¼ 6.6:1, conversion refers to 3-red only. }Ratio of 3-red to 3-red-ol ¼ 1:22.8, conversion refers to
3-red-ol only. #Additionally ꢀ10% of 3-red-ol were formed. **Determined by 2D NMR. See Methods and Supplementary Information for full experimental details. ND, not determined; NA, not applicable.
In the cascade reaction, the conversion of rac-2-red or 2-ox
into (R)-red-2 was achieved by combining MAO-N-9 with reaction, MAO-N-5 was combined with [Cp*Ir(biot-p-L)Cl] Sav
For the stereoinversion of (S)-nicotine red-3 using the cascade
.
.
.
[Cp*Ir(biot-p-L)Cl] Sav S112A-K121N or [Cp*Ir(biot-p-L)Cl] Sav S112A-K121T and catalase. After 14 h, (R)-nicotine was formed
S112A and a catalase to afford (R)-red-2 (e.e. of 99% (R), from (S)-nicotine in .99% e.e. (R) at 65% yield, with ꢀ10% yield
Table 1, entries 10 and 11). In a preparative deracemization of alcohol 3-red-ol (35% e.e.; Table 1, entry 14).
reaction, nearly quantitative yields of red-2 were isolated in
95% e.e. (R).
Activity screen of ATHases by a coupled colorimetric assay.
In the case of nicotine red-3, we found that mutations on Sav had Horseradish peroxidase (HRP) has found broad use in coupled
a significant influence on product distribution. As reported by assays, both in medicine and in biotechnology39. In this context, a
Branda¨nge et al., the tertiary iminium generated upon oxidation (pro)-dye is rapidly oxidized by HRP, consuming H2O2 and
of the tertiary amine red-3 exists in equilibrium with the keto- leading to an optical signal. We speculated that substitution of the
amine ox-3 in aqueous solution38. The relative rate of keto-amine catalase by HRP may allow the performance of the ATHase in the
versus iminium reduction by the ATHase determines the product cascade reaction to be estimated (Fig. 4).
distribution 3-red versus 3-red-ol (Table 1, entries 12 and 13).
For the reduction step alone, [Cp*Ir(Biot-p-L)Cl] Sav S112A- 1 mM for MAO-N-9) indicates that the ATHase step is rate-limiting
K121T led to the highest e.e. for nicotine red-3 (79% e.e. (R)) in the enzyme cascade under the considered reaction conditions
Comparison of the kinetic constants (kcat . 250 min21, KM
,
.
with a nicotine red-3 to alcohol red-3-ol ratio of 6.6:1.
(Supplementary Section S2.9). Thus, the HRP-generated optical
NATURE CHEMISTRY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturechemistry
3
© 2012 Macmillan Publishers Limited. All rights reserved.

NATURE CHEMISTRY DOI: 10.1038/NCHEM.1498
ARTICLES
(R)-amine (accumulates)
a
b
+
CO2
O2 (air)
(S)-amine
H2O
MAO-N-9/ATHase (S112T)
or
MAO-N-9/ATHase (S112K)
Highly '(S)-selective'
'(R)-selective'
ATHase variant
N
Catalase
NH
MAO-N variant
1-ox
(R)-1-red
Imine
(iminium)
HCO2H
H2O2
(R)-selective
S
ATHase (S112T)
c
100
80
60
40
20
1.0
0.8
0.6
0.4
0.2
4
3
2
1
0
(S)-selective
ATHase (S112K)
0
–20
–40
–60
–80
–100
5
10
15
20
Time (h)
(S)-selective
ATHase (S112K)
ATHase S112T
ATHase S112K
0.0
0
R
(R)-selective
ATHase (S112T)
50
100
150
0
50
100
150
1-ox concentration (mM)
1-ox concentration (mM)
Figure 3 | Enzyme cascade for the double stereoselective deracemization of amines. a, Schematic representation of the double stereoselective
deracemization with MAO-N/ATHase (see Methods and Supplementary Section S2 for experimental details). b, Reaction scheme for the production of
(R)-1-red via a reaction cascade, and time plot of conversion (dotted lines, empty circles) and enantiomeric excess (solid lines, filled circles) for
.
.
[Cp*Ir(biot-p-L)Cl] Sav S112T ATHase (red) and [Cp*Ir(biot-p-L)Cl] Sav S112K ATHase (blue). ATHase loading was 1 mol%. c, Saturation kinetic curves
.
.
for[Cp*Ir(biot-p-L)Cl] Sav S112T ATHase (red trace) and [Cp*Ir(biot-p-L)Cl] Sav S112K ATHase (blue trace). Error bars indicate +1 standard deviation,
determined from the triplicate measurement of each rate. Although the ATHase rate of the (R)-selective ATHase (S112T variant) is significantly lower than
that of the (S)-selective variant (S112K) at the beginning of the reaction, the matched enzyme cascade (that is, S112T MAO-N-9) reaches a high
enantiomeric excess in favour of the (R)-amine 1 more rapidly than the mismatched case (that is, S112K MAO-N-9).
NH
(R)
+
H2O
CO2
O2
+
NH
(S)
Oxidation products
MAO-N
ATHase
HRP
MeO
HCO2H
H2O2
N
HO
O
O
Figure 4 | Colorimetric assay for the determination of ATHase activity in an enzyme cascade. The activity of the ATHase in the enzyme cascade is
revealed by HRP: the H2O2 generated by MAO-N reacts with a dye (scopoletin, red) in the presence of HRP and leads to bleaching of scopoletin.
signature allows comparison of the relative ATHase activities, production of (S)-red-1. As the ATHases display modest selectiv-
depending on the Sav isoform.
ities (E values ≤ 4), sufficient (S)-red-1 is produced to allow the
After screening several dyes/prodyes, we identified scopoletin as implementation of the coupled assay. By following the decrease in
a suitable substrate as it loses its fluorescence upon oxidation: the fluorescence of scopoletin in the ATHase/MAO/HRP coupled
because the MAO-N variant is highly (S)-selective, the optical assay, imine-reductases with strongly increased activity (compared
.
readout generated by the oxidation of scopoletin correlates to the to S112T) could be identified. The ATHase [Cp*Ir(biot-p-L)Cl]
4
NATURE CHEMISTRY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturechemistry
© 2012 Macmillan Publishers Limited. All rights reserved.

NATURE CHEMISTRY DOI: 10.1038/NCHEM.1498
ARTICLES
a
b
Accumulates
N
H
CO2H
O2 (air)
NADH
CO2
H2O
O2 (air)
HO
N
H
CO2H
CO2
(S)-selective
Highly (R)-selective
Catalase
ATHase S112A
DAAO
ATHase S112A
HbpA
HO
HO
HCO2H
H2O2
HCO2H
NAD
N
CO2H
LAAO
NH2
H2O
O2 (air)
H2N
CO2H
Figure 5 | Expanding the concept of orthogonal redox cascades to include other enzymes. a, Production of L-pipecolic acid from L-lysine requires the
combination of an ATHase with an LAAO and a DAAO. Catalase is added to decompose the H2O2 generated by the AAOs. b, Concurrent regeneration of
NADH by the ATHase in the presence of a monooxygenase.
K121A displayed kinetic constants of kcat(app) ¼ 20 min21 commonly encountered44,45. To test the validity of the molecular
(+3 min21), KM(app) ¼ 12 mM (+3 mM) and Ki ¼ 20 mM compartmentalization concept outlined in Fig. 1, we investigated
(+4 mM) (Supplementary Fig. S19).The relative order of activities the regeneration of NADH in the presence of an NADH-
in the assay was in good agreement with the rates determined by dependent enzyme using an ATHase. Although significantly more
high-performance liquid chromatography (HPLC), although for active than [Cp*Rh(bipy)Cl] for NADH regeneration in the
high activities the relative activity was underestimated by the assay absence of an NADH-dependent enzyme, [Cp*Ir(biot-p-L)Cl] was
(Supplementary Fig. S20).
rapidly deactivated in the presence of 2-hydroxybiphenyl
monooxygenase (HbpA from Pseudomonas azaleica, an NADH-
Combination of ATHase with amino acid oxidases for the and FAD-dependent enzyme). In the presence of Sav, however,
formation of L-pipecolic acid. Encouraged by the results with the mutual inactivation of [Cp*Ir(biot-p-L)Cl] and HbpA was
MAO, we sought to combine ATHases in more complex cascades efficiently prevented, and robust hydroxylation activity was
with other oxidases. For this purpose, we selected L-lysine (4), achieved (Table 1, compare entries 18 and 19; Fig. 5b). We
which, upon oxidation with an L-amino acid oxidase (LAAO) and conclude that Sav shields the ATHase from the downstream
reduction with the ATHase, yields pipecolic acid (5). The enzyme, allowing NADH regeneration with formate as the
enantiomeric excess of 5 can be upgraded by combining the hydride source (KM(app) ¼ 165 mM [+6 mM], kcat(app) ¼ 1.37 min21
LAAO/ATHase/catalase with a D-selective amino acid oxidase [+0.01 min21]; Supplementary Fig. S27). Full conversion of
(DAAO) (Fig. 5a)40,41
.
2-hydroxybiphenyl (6) to 2,3-dihydroxybiphenyl (7) was
In a first round of screening, the performances of a range of accomplished in 2 h with a crude enzyme extract (Fig. 5b,
ATHase mutants were screened for the reduction of D1-piper- Table 1, entries 18 and 19; Supplementary Fig. S25).
idine-2-carboxylic acid generated in situ by oxidation of L-lysine
The system could be run either in the pure aqueous phase or as a
with LAAO (from snake venom, Crotalus atrox). The Ir-cofactor biphasic system with 1-decanol functioning as a substrate reservoir
[Cp*Ir(biot-p-L)Cl] in the absence of Sav showed only a low conver- and product sink, thereby highlighting the applicability of the
sion (TON ¼ 11, turnover number (TON) = moles product/moles ATHase under a variety of reaction conditions. The biphasic
catalyst) and led to racemic product, highlighting again the inacti- system again displayed strong inactivation in the absence of Sav,
vation of the naked catalyst (Table 1, entry 15). In contrast, the whereas a TON of .100 (versus [Cp*Ir(biot-p-L)Cl]) was achieved
ATHase demonstrated promising conversions (up to 90%; TON ¼ when Sav was present (Supplementary Fig. S26).
48), albeit with low enantioselectivity. The most (S)-selective
mutant (Sav S112A; TON ¼ 47; 29% e.e.) was selected for further Outlook
studies (Table 1, entry 16; Supplementary Table S8). Combining We have demonstrated herein that an ATHase consisting of
the system with a DAAO (from porcine kidney) allowed the [Cp*Ir(biot-p-L)Cl] anchored within a streptavidin isoform is
enantiomeric excess of the L-pipecolic acid to be upgraded by re- complementary and compatible with a variety of redox enzymes
oxidizing the D-enantiomer; under the chosen conditions, an enan- relying on NADH, FADH2 and haem cofactors. Such artificial
tiomeric excess of 86% in favour of L-4 was obtained (Fig. 5a; metalloenzymes display attractive features that are reminiscent of
Table 1, entry 17). The formation of pipecolic acid was unambigu- both biocatalysts and chemocatalysts: precious metal reactivity,
ously established by 2D NMR analysis of the reaction mixtures using genetic optimization potential and well-defined second coordi-
L-lysine-2-13C as substrate (Supplementary Fig. S22).
nation sphere provided by a protein scaffold. This last feature
could be further exploited with a view to achieving the immobiliz-
NADH regeneration for monooxygenases. The chemical and ation of the entire enzyme cascade.
electrochemical recycling of NAD(P)H and analogues has been
To optimize such cascades, directed evolution protocols are
intensively investigated as an alternative to enzymatic highly desirable. With this goal in mind, we have shown that the
regeneration12,42–49. In this context, [Cp*Rh(bipy)Cl] has emerged ATHase can be integrated with a colorimetric coupled assay,
as the redox mediator or catalyst of choice. However, in the leading to the identification of a genetically improved ATHase.
presence of the downstream enzyme, mutual inactivation is These proof-of-principle examples open fascinating perspectives
NATURE CHEMISTRY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturechemistry
© 2012 Macmillan Publishers Limited. All rights reserved.
5

NATURE CHEMISTRY DOI: 10.1038/NCHEM.1498
ARTICLES
20. Lopez-Gallego, F. & Schmidt-Dannert, C. Multi-enzymatic synthesis. Curr.
Opin. Chem. Biol. 14, 174–183 (2010).
towards complementing biocatalytic cascades with molecularly
compartmentalized organometallic catalysts.
`
21. Pamies, O. & Ba¨ckvall, J-E. Combination of enzymes and metal catalysts.
A powerful approach in asymmetric catalysis. Chem. Rev. 103,
3247–3261 (2003).
22. Kim, Y., Park, J. & Kim, M-J. Dynamic kinetic resolution of amines and amino
acids by enzyme–metal cocatalysis. ChemCatChem. 3, 271–277 (2011).
Methods
Typical procedure for a double stereoselective deracemization with MAO-
N/ATHase (analytical scale). The following stock solutions were prepared.
Catalase from bovine liver (50 kU ml21) was dissolved in MOPS/sodium
formate buffer (0.6 M in MOPS, 3.0 M in sodium formate, pH adjusted with aq.
KOH to pH 7.8). Sav S112T (16.4 mg ml21, 0.75 mM free binding sites, assuming
three free binding sites per tetramer) was dissolved in the catalase containing
buffer. [IrCp*(biot-p-L)Cl] was dissolved in DMF (37.5 mM). Affinity purified
MAO-N-9 (buffer exchange with MOPS, 0.6 M, pH adjusted with KOH to 7.8)
was diluted in MOPS buffer (0.6 M, pH adjusted with KOH to 7.8) to a
concentration of 0.38 mg ml21. Substrate stock was prepared by dissolving the
hydrochloride of 1-ox in H2O (1 M). The ATHase was prepared by adding
[IrCp*(biot-p-L)Cl]-stock (10 ml ml21) to the Sav-stock solution. MAO-N-stock
was placed in 1.5 ml PP-tubes (100 ml) and ATHase was added (100 ml). The
reactions were initiated by addition of substrate stock solution (7.5 ml) and
incubated at 37 8C and 250 r.p.m. in a lying position. For work-up aq. NaOH
solution was added (50 ml of a 10 M solution) and the mixture extracted with
CH2Cl2 (1 × 1 ml). The organic phase was dried over Na2SO4 and analysed by
HPLC (Supplementary Sections S2.3.1 and S2.4.1).
´
23. Yusop, R. M., Unciti-Broceta, A., Johansson, E. M. V., Sanchez-Martin, R. M. &
Bradley, M. Palladium-mediated intracellular chemistry. Nature Chem. 3,
239–243 (2011).
24. Foulkes, J. M., Malone, K. J., Coker, V. S., Turner, N. J. & Lloyd, J. R. Engineering
a biometallic whole cell catalyst for enantioselective deracemization reactions.
ACS Catal. 1, 1589–1594 (2011).
25. Lu, Y., Yeung, N., Sieracki, N. & Marshall, N. M. Design of functional
metalloproteins. Nature 460, 855–862 (2009).
26. Ward, T. R. Artificial metalloenzymes based on the biotin–avidin
technology: enantioselective catalysis and beyond. Acc. Chem. Res. 44,
47–57 (2011).
27. Bos, J., Fusetti, F., Driessen, A. J. M. & Roelfes, G. Enantioselective artificial
metalloenzymes by creation of a novel active site at the protein dimer interface.
Angew. Chem. Int. Ed. 51, 7472–7475 (2012).
28. Jing, Q. & Kazlauskas, R. J. Regioselective hydroformylation of styrene using
rhodium-substituted carbonic anhydrase. ChemCatChem 2, 953–957 (2010).
29. Podtetenieff, J., Taglieber, A., Bill, E., Reijerse, E. J. & Reetz, M. T. An artificial
metalloenzyme: creation of a designed copper binding site in a thermostable
protein. Angew. Chem. Int. Ed. 49, 5151–5155 (2010).
Received 10 May 2012; accepted 10 October 2012;
published online 25 November 2012
30. Deuss, P. J., den Heeten, R., Laan, W. & Kamer, P. C. J. Bioinspired catalyst
design and artificial metalloenzymes. Chem. Eur. J. 17, 4680–4698 (2011).
31. Ueno, T., Abe, S., Yokoi, N. & Watanabe, Y. Coordination design of artificial
metalloproteins utilizing protein vacant space. Coord. Chem. Rev. 251,
2717–2731 (2007).
32. Matsuo, T. et al. Meso-unsubstituted iron corrole in hemoproteins: remarkable
differences in effects on peroxidase activities between myoglobin and
horseradish peroxidase. J. Am. Chem. Soc. 131, 15124–15125 (2009).
33. Turner, N. J. Directed evolution drives the next generation of biocatalysts. Nature
Chem. Biol. 5, 567–573 (2009).
References
1. Wo¨rsdo¨rfer, B., Woycechowsky, K. J. & Hilvert, D. Directed evolution of a
protein container. Science 331, 589–592 (2011).
2. Choudhary, S., Quin, M. B., Sanders, M. A., Johnson, E. T. & Schmidt-
Dannert, C. Engineered protein nano-compartments for targeted enzyme
localization. PLoS ONE 7, e33342 (2012).
3. Dueber, J. E. et al. Synthetic protein scaffolds provide modular control over
metabolic flux. Nature Biotechnol. 27, 753–759 (2009).
4. Keasling, J. D. Synthetic biology for synthetic chemistry. ACS Chem. Biol. 3,
64–76 (2008).
34. Turner, N. J. Enantioselective oxidation of C–O and C–N bonds using oxidases.
Chem. Rev. 111, 4073–4087 (2011).
5. Bromley, E. H. C., Channon, K., Moutevelis, E. & Woolfson, D. N. Peptide and
protein building blocks for synthetic biology: from programming biomolecules
to self-organized biomolecular systems. ACS Chem. Biol. 3, 38–50 (2008).
6. Weissman, K. J. & Leadlay, P. F. Combinatorial biosynthesis of reduced
polyketides. Nature Rev. Microbiol. 3, 925–936 (2005).
7. Mutti, F. G. et al. Simultaneous iridium catalysed oxidation and enzymatic
reduction employing orthogonal reagents. Chem. Commun. 46,
8046–8048 (2010).
35. Du¨rrenberger, M. et al. Artificial transfer hydrogenases for the enantioselective
reduction of cyclic imines. Angew. Chem. Int. Ed. 50, 3026–3029 (2011).
36. Rowles, I., Malone, K. J., Etchells, L. L., Willies, S. C. & Turner, N. J. Directed
evolution of the enzyme monoamine oxidase (MAO-N): highly efficient chemo-
enzymatic deracemisation of the alkaloid (+)-crispine A. ChemCatChem 4,
1259–1261 (2012).
37. Heiden, Z. M. & Rauchfuss, T. B. Homogeneous catalytic reduction of
dioxygen using transfer hydrogenation catalysts. J. Am. Chem. Soc. 129,
14303–14310 (2007).
8. Haak, R. M. et al. Dynamic kinetic resolution of racemic b-haloalcohols:
direct access to enantioenriched epoxides. J. Am. Chem. Soc. 130,
13508–13509 (2008).
9. Maid, H. et al. Iron catalysis for in situ regeneration of oxidized cofactors
by activation and reduction of molecular oxygen: a synthetic metalloporphyrin
as a biomimetic NAD(P)H oxidase. Angew. Chem. Int. Ed. 50,
2397–2400 (2011).
10. Wasilke, J-C., Obrey, S. J., Baker, R. T. & Bazan, G. C. Concurrent tandem
catalysis. Chem. Rev. 105, 1001–1020 (2005).
11. Zhou, J. Recent advances in multicatalyst promoted asymmetric tandem
reactions. Chem. Asian J. 5, 422–434 (2010).
12. Betanzos-Lara, S. et al. Organometallic ruthenium and iridium transfer-
hydrogenation catalysts using coenzyme NADH as a cofactor. Angew. Chem. Int.
Ed. 51, 3897–3900 (2012).
13. Wingstrand, E., Laurell, A., Fransson, L., Hult, K. & Moberg, C. Minor
enantiomer recycling: metal catalyst, organocatalyst and biocatalyst working in
concert. Chem. Eur. J. 15, 12107–12113 (2009).
14. Simons, C., Hanefeld, U., Arends, I. W. C. E., Maschmeyer, T. & Sheldon, R. A.
Towards catalytic cascade reactions: asymmetric synthesis using combined
chemo-enzymatic catalysts. Top. Catal. 40, 35–44 (2006).
15. Wieczorek, B. Covalent anchoring of a racemization catalyst to CALB-beads:
towards dual immobilization of DKR catalysts. Tetrahedron Lett. 52,
1601–1604 (2011).
38. Branda¨nge, S., Lindblom, L., Pilotti, Å. & Rodriguez, B. Ring-chain tautomerism
of pseudooxynicotine and some other iminium compounds. Acta Chem. Scand.
B 37, 617–622 (1983).
39. Truppo, M. D., Escalettes, F. & Turner, N. J. Rapid determination of both the
activity and enantioselectivity of ketoreductases. Angew. Chem. Int. Ed. 47,
2639–2641 (2008).
40. Yasuda, M., Ueda, M., Muramatsu, H., Mihara, H. & Esaki, N. Enzymatic
synthesis of cyclic amino acids by N-methyl-L-amino acid dehydrogenase from
Pseudomonas putida. Tetrahedron Asymm. 17, 1775–1779 (2006).
41. Gatto, G. J., Boyne, M. T., Kelleher, N. L. & Walsh, C. T. Biosynthesis of pipecolic
acid by RapL, a lysine cyclodeaminase encoded in the rapamycin gene cluster.
J. Am. Chem. Soc. 128, 3838–3847 (2006).
42. Hollmann, F., Hofstetter, K. & Schmid, A. Non-enzymatic regeneration of
nicotinamide and flavin cofactors for monooxygenase catalysis. Trends
Biotechnol. 24, 163–171 (2006).
43. Hollmann, F., Arends, I. W. C. E. & Buehler, K. Biocatalytic redox reactions for
organic synthesis: nonconventional regeneration methods. ChemCatChem 2,
762–782 (2010).
44. Poizat, M., Arends, I. W. C. E. & Hollmann, F. On the nature of mutual
inactivation between [Cp*Rh(bpy)(H2O)]2þ and enzymes — analysis and
potential remedies. J. Mol. Catal. B 63, 149–156 (2010).
45. Hildebrand, F. & Lu¨tz, S. Stable electroenzymatic processes by catalyst
separation. Chem. Eur. J. 15, 4998–5001 (2009).
46. Haquette, P. et al. Chemically engineered papain as artificial formate
dehydrogenase for NAD(P)H regeneration. Org. Biomol. Chem. 9,
5720–5727 (2011).
´
´
´
16. Rocha-Martın, J., de las Rivas, B., Mun˜oz, R., Guisan, J. M. & Lopez-Gallego, F.
Rational co-immobilization of bi-enzyme cascades on porous supports and their
applications in bio-redox reactions with in situ recycling of soluble cofactors.
ChemCatChem 4, 1279–1288 (2012).
17. Hanefeld, U., Gardossi, L. & Magner, E. Understanding enzyme immobilisation.
Chem. Soc. Rev. 38, 453–468 (2009).
18. Brady, D. & Jordaan, J. Advances in enzyme immobilisation. Biotechnol. Lett.
31, 1639–1650.
19. Mateo, C., Palomo, J. M., Fernandez-Lorente, G., Guisan, J. M. & Fernandez-
Lafuente, R. Improvement of enzyme activity, stability and selectivity via
immobilization techniques. Enzyme Microb. Tech. 40, 1451–1463 (2007).
47. Maenaka, Y., Suenobu, T. & Fukuzumi, S. Efficient catalytic interconversion
between NADH and NADþ accompanied by generation and consumption of
hydrogen with a water-soluble iridium complex at ambient pressure and
temperature. J. Am. Chem. Soc. 134, 367–374 (2012).
6
NATURE CHEMISTRY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturechemistry
© 2012 Macmillan Publishers Limited. All rights reserved.

NATURE CHEMISTRY DOI: 10.1038/NCHEM.1498
ARTICLES
ˇ
ˇ
ˇ
48. Canivet, J., Su¨ss-Fink, G. & Stepnicka, P. Water-soluble phenanthroline
complexes of rhodium, iridium and ruthenium for the regeneration of NADH in
the enzymatic reduction of ketones. Eur. J. Inorg. Chem. 4736–4742 (2007).
49. Ryan, J. D., Fish, R. H. & Clark, D. S. Engineering cytochrome P450 enzymes for
improved activity towards biomimetic 1,4-NADH cofactors. ChemBioChem 9,
2579–2582 (2008).
Author contributions
V.K., F.H., N.T. and T.W. conceived the catalytic cascades. V.K., F.H., N.T. and T.W.
supervised the project. V.K., Y.W., M.D., D.G., E.C. and T.Q. performed the experiments.
V.K., Y.W., M.D., D.G., E.C., T.Q., F.H., N.T. and T.W. analysed the data. V.K., F.H., N.T.
and T.W. co-wrote the paper. V.K., Y.W., M.D., D.G. and L.K. contributed materials. D.H.
analysed the conversion of 13C-labelled lysine by 2D NMR.
Acknowledgements
Additional information
This work was supported by the Marie Curie Initial Training Network (Biotrains FP7-ITN-
238531). T.R.W. acknowledges financial support from the SNF (Schweizerische
Nationalfonds, grant no. 200020_126366) and the National Centre of Competence in
Research Nanosciences. N.J.T. acknowledges the Royal Society for a Wolfson Research
Merit Award. F.H. thanks A. Schmid (Dortmund University of Technology) for the kind
provision of HbpA. The authors also thank M. Corbett, S. Willies and K. Malone for helpful
Supplementary information and chemical compound information are available in the
online version of the paper. Reprints and permission information is available online at http://
www.nature.com/reprints. Correspondence and requests for materials should be addressed to
F.H., N.J.T. and T.R.W.
advice and materials, R. Pfalzberger for help with the graphic material, and Umicore for a Competing financial interests
precious metal loan.
The authors declare no competing financial interests.
NATURE CHEMISTRY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturechemistry
7
© 2012 Macmillan Publishers Limited. All rights reserved.