Fluorescence-Based Kinetic Assay for High-Throughput Discovery and Engineering of Stereoselective ω-Transaminases

Article abstract of DOI:10.1002/adsc.201500215

ω-Transaminases are a valuable class of enzymes for the production of chiral amines with either (R)- or (S)-configuration in high optical purity and 100% yield by the biocatalytic reductive amination of prochiral ketones. A versatile new assay was developed to quantify ω-transaminase activity for the kinetic characterization and enantioselectivity typing of novel or engineered enzymes based on the conversion of 1-(6-methoxynaphth-2-yl)alkylamines. The associated release of the acetonaphthone product can be monitored by the development of its bright fluorescence at 450 nm with very high sensitivity and selectivity. The assay principle can be used to quantify ω-transaminase catalysis over a very broad range of enzyme activity. Because of its simplicity and low substrate consumption in microtiter plate format the assay seems suitable for liquid screening campaigns with large library sizes in the directed evolution of optimized transaminases. For assay substrates that incorporate structural variations, an efficient modular synthetic route was developed. This includes racemate resolution by lipase-catalyzed transacylation to furnish enantiomerically pure (R)- and (S)-configured amines. The latter are instrumental for the rapid enantioselectivity typing of ω-transaminases. This method was used to characterize two novel (S)-selective taurine-pyruvate transaminases of the subtype 6a from thermophilic Geobacillus thermodenitrificans and G. thermoleovorans.

Full text of DOI:10.1002/adsc.201500215

FULL PAPERS
DOI: 10.1002/adsc.201500215
Very Important Publication
Fluorescence-Based Kinetic Assay for High-Throughput
Discovery and Engineering of Stereoselective w-Transaminases
Thomas Scheidt,a Henrik Land,b Mattias Anderson,b Yujie Chen,c Per Berglund,b
Dong Yi,c and Wolf-Dieter Fessnera,*
a
Institut für Organische Chemie und Biochemie, Technische Universität Darmstadt, Alarich-Weiss-Str. 4, 64287 Darmstadt,
Germany
Fax: (+49)-6151-166636; e-mail: fessner@tu-darmstadt.de
KTH Royal Institute of Technology, School of Biotechnology, Division of Industrial Biotechnology, AlbaNova University
b
Center, SE-106 91 Stockholm, Sweden
State Key Laboratory of Bioreactor Engineering, Newworld Institute of Biotechnology, East China University of Science
c
and Technology, Meilong Road 130, Shanghai 200237, Peopleꢀs Republic of China
Received: March 2, 2015; Revised: April 10, 2015; Published online: May 7, 2015
Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/adsc.201500215.
Abstract: w-Transaminases are a valuable class of campaigns with large library sizes in the directed
enzymes for the production of chiral amines with evolution of optimized transaminases. For assay sub-
either (R)- or (S)-configuration in high optical purity strates that incorporate structural variations, an effi-
and 100% yield by the biocatalytic reductive amina- cient modular synthetic route was developed. This
tion of prochiral ketones. A versatile new assay was includes racemate resolution by lipase-catalyzed
developed to quantify w-transaminase activity for transacylation to furnish enantiomerically pure (R)-
the kinetic characterization and enantioselectivity and (S)-configured amines. The latter are instrumen-
typing of novel or engineered enzymes based on the tal for the rapid enantioselectivity typing of w-trans-
conversion of 1-(6-methoxynaphth-2-yl)alkylamines. aminases. This method was used to characterize two
The associated release of the acetonaphthone prod- novel (S)-selective taurine-pyruvate transaminases of
uct can be monitored by the development of its the subtype 6a from thermophilic Geobacillus ther-
bright fluorescence at 450 nm with very high sensitiv- modenitrificans and G. thermoleovorans.
ity and selectivity. The assay principle can be used to
quantify w-transaminase catalysis over a very broad
range of enzyme activity. Because of its simplicity Keywords: biocatalysis; chiral amines; high-through-
and low substrate consumption in microtiter plate put screening; protein engineering; reductive amina-
format the assay seems suitable for liquid screening tion
Introduction
tion process that suffered from insufficient levels of
enantioselectivity and chemical purity. Starting out
Methods for the asymmetric synthesis of chiral from a transaminase scaffold inactive on the target
amines are of high topical interest, as these functional substrate and from a strongly truncated substrate ana-
moieties represent prevalent motifs in many active logue, several rounds of rational and evolutive engi-
pharmaceutical ingredients.[1,2] Recently, w-transami- neering were needed to incrementally adapt the large
nases have enjoyed a particular focus for develop- and small binding pockets of the enzyme to finally
ment because the enzymes in this family offer a green yield an active catalyst that was able to substantially
route for the reductive amination of prochiral ketones improve the manufacturing process (Scheme 1).
to produce (R)- and (S)-amines in high optical purity
“Systems biocatalysis”,[5] the concept of organizing
and 100% yield.[3] A recent informative example is independent enzyme modules from various origins in
the engineering of an (R)-selective w-transaminase vitro to generate novel artificial metabolisms[6] for the
for the industrial production of sitagliptin,[4] an oral synthetic transformation of non-natural precursors,
anti-diabetic drug, as a replacement for the rhodium- likewise requires transaminase modules of either (R)-
based, high-pressure chemical asymmetric hydrogena- or (S)-enantiospecificity that are applicable to
Adv. Synth. Catal. 2015, 357, 1721 – 1731
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1721

FULL PAPERS
Thomas Scheidt et al.
Scheme 1. The pro-sitagliptin structure (left) illustrating the size requirements for w-transaminase substrate binding pockets.
a broad variety of substrate scaffolds.[7] Particularly, substrate morphing[16]) approach for protein engineer-
thermostable enzymes are expected to offer several ing.
advantages over normal w-transaminase in biotech-
nology, such as maintaining high activity under con-
tinuous process operation, improving tolerance for or- Results and Discussion
ganic co-solvent and reducing overall reaction times
due to increased catalytic activity at higher reaction Assay Principle
temperatures. Both the discovery of novel w-transa-
minases and determination of their enantioselectivity, Transaminases are PLP-dependent enzymes that con-
as well as the protein engineering of known w-transa- vert ketones into chiral amines and vice versa. In the
minases to improve their substrate scope rely on ef- reverse mode, the amine functionality is transferred
fective methods to rapidly evaluate biocatalyst activi- to PLP with formation of PMP and release of
ty and stereoselectivity with high accuracy. Although a ketone. PLP is regenerated by transfer of the amino
X-ray protein crystal structures have been determined group from PMP to another carbonyl acceptor. We
for both (R)- and (S)-selective transaminases[8–11] and reasoned that the non-fluorescent 1-(6-methoxy-
although it could be shown that the enantioselectivity naphth-2-yl)ethylamine 1a (R=CH3) in the appropri-
and substrate scope of novel transaminases can be ate enantiomeric form should be able to act as a sub-
predicted from specific signatures in their genetic se- strate of transaminase and furnish the corresponding
quence,[12] an ultimate typing of their enantioprefer- acetonaphthone derivative 2a which, in an aqueous
ence and substrate scope requires experimental proof environment, shows a bright blue fluorescence emis-
by an appropriate assay technology.
sion around 450 nm with a quantum yield of about 0.2
Most conventional methods to assay for transami- that can be readily detected using a common plate
nase activity are time consuming and inappropriate reader (Scheme 2).[17] Indeed, when incubating the
for a determination of substrate scope and enantiose- amine 1a, commercially available as the hydrochlo-
lectivity in high-throughput screening mode. Many ride salt in racemic form, in the presence of PLP and
assays also suffer from interference with amine con- pyruvate as an appropriate amine acceptor with the
taining metabolites in crude cell extracts, which re- (R)-selective transaminase from Neosartorya fischeri
quires prior enzyme purification. More recently, at an optimum pH of 8.2, a steady development of
a number of colorimetric assays were developed for the bright blue fluorescence could be monitored, veri-
kinetic determinations that are based on pH change, fying the general assay concept. In comparison, a mi-
metal complexation or redox initiated dye forma-
tion.[10,13,14] A common method consists in monitoring
the UV absorption at 245 nm of acetophenone gener-
ated from the widely used model substrate 1-phenyl-
ethylamine (“a-methylbenzylamine”, MBA).[15] Much
higher sensitivity and specificity are to be expected
from measurement of fluorescence emission rather
than UV absorption. However, to the best of our
knowledge no fluorescence assay for transaminases
exists.[13] Here we report on the development of
a novel fluorogenic assay for measurement of transa-
minase activity that is based on the strong blue emis-
sion of the well-known 6-methoxynaphth-2-yl-carbon-
yl fluorophor formed upon transamination of a corre-
Scheme 2. Fluorogenic assay principle for transaminases by
release of fluorescent ketone 2 upon reaction of amines 1;
the large and small substrate binding pockets occupied by
sponding amine precursor. The assay concept allows
a systematic modular variation of the substrate struc-
ture to facilitate a substrate walking[4] (also termed the fluorogenic substrate are indicated by dashed lines.
1722
asc.wiley-vch.de
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2015, 357, 1721 – 1731

FULL PAPERS
Fluorescence-Based Kinetic Assay for High-Throughput Discovery
Scheme 3. Synthesis of amine substrates and racemate resolution of 1a by lipase-catalyzed transacylation.
crobial (S)-selective transaminase of unspecified
Enantiomers of 1a were kinetically resolved by
origin (Sigma No. 77856) under identical conditions lipase-catalyzed transacylation in organic medium
did not produce any fluorescence above background, using ethyl methoxyacetate and immobilized lipase B
indicating that the extended naphthyl fluorophore from Candida antarctica (CAL-B) according to a liter-
structure is instrumental in effectively discriminating ature precedent.[20] Both antipodes were obtained
enzymes that are substrate tolerant (e.g., w-TA from with >99% ee (HPLC analysis) in excellent yield, in-
N. Fischeri) from those having a rather narrow specif- cluding recrystallization and hydrolysis of the easily
icity. Such discrimination of the “large” substrate separable methoxyacetamide (R)-5a. Although not
binding pocket of transaminases is not possible with yet performed in this study, kinetic resolution of
the common acetophenone assay, which shows activi- higher homologous amines using this method should
ty for both enzymes.
be similarly straightforward.[20b] Configurational as-
signments follow the general predictions of the Ka-
zlauskas rule,[21] but were proven by independent syn-
thesis of the (S)-configured enantiomer of amine 1a
from the inexpensive NSAID naproxen (Scheme 4),
Substrate Synthesis
Generally, the synthesis of racemic amines 1 can be which is marketed as the enantiomerically pure drug
achieved by reductive amination of the corresponding in the (S)-configuration. Conversion of the acid by
6-methoxy-2-acylnaphthalenes 2. In view of the high a published Curtius degradation gave erratic results,[22]
costs of higher homologous ketones but with a desire but Hofmann rearrangement of the corresponding
to investigate a broader variety of structural varia- carboxamide provided the expected (S)-amine, if only
tions in the R group, and the need for separation of in modest yield. Because the Hofmann rearrangement
regioisomers obtained by Friedel–Crafts acylation,[18] proceeds stereospecifically with retention of configu-
we decided for an alternative modular approach con- ration, the amine product derived from (S)-naproxen
sisting of a Grignard addition to the related aldehyde must have the absolute (S)-configuration.
3
followed by Jones oxidation of the carbinol
(Scheme 3). Efficient reductive amination of the re-
sultant ketones 2a–c was performed by titanium- Assay Development and Performance
mediated imination with ammonia followed by boro-
hydride reduction[19] to furnish the racemic amines According to the fluorescence spectrum of ketone 2a
1a–c in high overall yields.
(see Figure 3 in the Supporting Information) kinetic
measurements were performed with excitation at
Adv. Synth. Catal. 2015, 357, 1721 – 1731
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1723

FULL PAPERS
Thomas Scheidt et al.
of fluorescence due to formation of ketone 2a showed
excellent linearity with time, and varying the amount
of protein confirmed a linear relationship between
signal intensity and enzyme activity (Figure 4 in the
Supporting Information). No fluorescence signal was
recorded in the absence of the biocatalyst or of pyru-
vate, as long as no additional PLP cofactor was added
(which could act as a co-substrate) beyond that bound
to the enzyme active site. Measurements were tested
Scheme 4. Enantiospecific synthesis of (S)-amine from opti-
cally pure (S)-naproxen.
330 nm (lmax =310 nm) and emission observed at in both cuvette and microtiter plate formats with
460 nm (lmax =450 nm). For low concentrations of the good reproducibility, while the latter has the added
ketone (at least up to 6 mM) a linear relationship be- advantage that simultaneous recording of standard
tween fluorescence intensity and the ketone concen- curve dilutions can be achieved under identical condi-
tration was confirmed, which allows continuous quan- tions. As a variation, a concentrated stock solution of
tification of enzyme activity by correlation with the substrate 1a in DMSO was also prepared for dilution
standard curve. The limit of detection (LOD) and to a concentration of less than 1% of DMSO in the
limit of quantification (LOQ) values derived by linear final assay with no influence on data quality. Activity
regression were 0.35 nmolminÀ1 and 1.2 nmolminÀ1 could be determined also from crude cell extracts
(vmax for 2a), respectively.[23] Exemplary studies for ve- with good signal intensity against E. coli host as a neg-
rification of the assay principle and optimization of ative control. Despite different instrumentation set-
conditions were conducted first by using the commer- ups, data reproducibility was quite well between the
cial w-transaminase from N. fischeri as catalyst. Reac- laboratories involved in this study.
tions were run under slightly alkaline conditions
Reaction rates were easily determined even for
(pH 8.2), which is a typical optimum for enzymes in concentrations of amine 1a as low as mM with good
this class, in the presence of 4.5 mM pyruvate as a co- linearity between substrate concentration and signal
substrate that is suitable for most w-transaminases. intensity (Figure 1). However, strongly reduced signal
Kinetic data determined in phosphate or HEPES buf- intensities were observed at amine concentrations
fers were found to be very similar (Table 1, entries 1 higher than 1 mM, indicating a strong inhibition effect
and 2), so the influence of buffers was not further that is thought to arise from substrate inhibition of
studied. The resultant time curves for the appearance the transaminase. The reason for this substrate inhibi-
Table 1. Determination of kinetic data for various w-transaminases by using substrates (R)- and (S)-1a.
Enzyme
Enantiomer of 1
KM [mM]
Ki [mM]
vmax [mmol/minmg]
vmax/KM
E
Neosartorya fischeri[a]
(S)
(R)
(S)
(R)
(S)
(R)
(S)
(R)
(S)
(R)
(S)
(R)
(S)
(R)
(S)
(R)
(S)
(R)
(S)
(R)
1.2
4.9
3.2
3.3
0.72
1.4
5.1
2.0
2.0
1.3
0.73
2.0
1.1
0.31
0.52
1.1
0.68
6.1
5.6
6.1
–
0.00015
0.0023
0.00044
0.018
0.76
0.099
0.00013
0.0039
0.00056
0.020
1.7
0.090
29
0.054
0.0026
0.11
30 (R)
0.59
0.79
0.88
0.45
1.1
0.77
0.65
0.77
0.27
1.1
Neosartorya fischeri
Chromobacterium violaceum
Vibrio fluvialis
36 (R)
19 (S)
>200 (S)
42 (R)
80 (R)
50 (R)
110 (S)
38 (S)
– (S)
22
0.035
Aspergillus fumigatus
Aspergillus terreus
0.0020
0.031
0.00011
0.00040
0.044
0.25
0.012
0.023
0.00039
0.000020
0.000031
n.d.[c]
0.00010
0.0080
0.022
1.1
0.050
2.0
ATA 117[b]
0.23
0.0025
0.55
0.80
1.6
ATA 113[b]
4.8
0.042
0.00049
0.000013
0.000053
–
Geobacillus thermodenitrificans[a]
Geobacillus thermoleovorans[a]
0.58
–
[a]
Measured in phosphate buffer, all other data measured in HEPES buffer.
Commercial enzyme of unspecified origin obtained from Codexis Inc.
n.d.=no detectable activity.
[b]
[c]
1724
asc.wiley-vch.de
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2015, 357, 1721 – 1731

FULL PAPERS
Fluorescence-Based Kinetic Assay for High-Throughput Discovery
a range of common w-transaminases that previously
had been utilized for systematic studies in method de-
velopment and for preparative applications, and was
also extended to commercial samples of known spe-
cificity (Table 1). The fluorogenic assay shows gener-
ally good capacity for the determination of the kinetic
parameters KM and vmax of w-transaminases. Substrate
concentration was varied between 0.00043–7.0 mM by
serial dilutions. Kinetic constants could be determined
easily over a large range of data values for KM (three
orders of magnitude), KI (one order of magnitude)
and vmax (six orders of magnitude) (Table 1). The rela-
tively low KM values as compared to values published
for the benchmark substrate 1-phenylethylamine
(35.03 mM for the w-transaminase from Bacillus thur-
ingiensis[25] and 16 mM for the w-transaminase from
Figure 1. Kinetic profile for activity of w-transaminase from
C. violaceum upon incubation with increasing substrate con-
centration of (S)-1a. Data fit for Michaelis–Menten kinetics
(solid line) indicates strong substrate inhibition beyond
1 mM.
tion has not been further investigated yet but the hy- Vibrio fluvialis[26] implies that all enzymes investigated
drophobic nature of the large naphthyl fluorophore is in this study have a rather good affinity for the hydro-
seen as a plausible reason. Therefore a modified ver- phobic substrate 1a.
sion of the Michaelis–Menten equation [Eq. (1)] has
The enantioselectivity of an enzyme can be ex-
been used for the fit of all kinetic data to take inhibi- pressed by the kinetic E-factor,[27] which is the ratio
tion effects into account. Quenching of ketone fluo- of the specificity constants kcat/KM for the two enan-
rescence in the presence of higher concentrations of tiomers. As a rough estimate, we calculated the enan-
the amine can be excluded as an alternative cause for tioselectivity of the transaminases screened by the flu-
ketone concentrations as high as 0.5 mM.
orogenic assay from the tabulated vmax/KM data. We
note that operationally this method of E-factor deter-
mination is as fast as, but probably more accurate
than, the quickE method introduced for hydrolases,[28]
which is based on a direct rate comparison for enan-
tiomeric substrates only and is limited by the fact that
competitive binding to the enzyme active site, as typi-
cal for a racemate resolution, has to be re-introduced
In order to evaluate and validate the assay for its by addition of a reference compound.
potential in high-throughput screening applications,
Although we have not yet explored the option for
we further determined the Z-factor,[24] a dimensionless measuring the influence of further reaction parame-
statistical parameter that reflects both the assay signal ters that might have an influence on enzyme activity,
dynamic range and the data variation associated with such as medium pH, temperature or addition of co-
the signal measurements. For a substrate concentra- solvent, it appears to be a rather simple task to study
tion of 0.7 mM we determined an excellent Z-factor enzyme properties by serial variation of any such ef-
of 0.95, whereas even at a lower substrate concentra- fectors or physical parameters under the standard
tion of 0.35 mM a very high Z-factor of 0.80 resulted. assay conditions. Especially, the replacement of pyru-
Thus, the fluorescent assay seems to be very appropri- vate by other (non-fluorescent) ketones or aldehydes
ate for screening efforts of variant libraries in the di- as alternative co-substrates seems a promising option
rected evolution of w-transaminases with improved or to rapidly determine the amino acceptor specificity of
altered protein properties, while consuming only very a given transaminase.
small substrate quantities (e.g., 4.5 mg of 1a hydro-
chloride per 96-well titer plate at 1.0 mM substrate
concentration).
Typing of Novel Transaminases
Among the aminotransferases, the Class III or w-
transaminases constitute a large family of enzymes in
which its members are sub-classified into six clades by
their sequence homology and the diversity of their
Determination of Enantioselectivity and Kinetic
Constants
The enantioselectivity of known and unknown en- natural substrates.[9] Although knowledge about their
zymes can be determined by their behavior towards natural functions and substrates is still sparse, many
the pure individual antipodes of 1a. The suitability of have proven useful for biotechnological applications,
this approach for enzyme typing was confirmed for such as those included in Table 1 (entries 1–8).[1–3] The
Adv. Synth. Catal. 2015, 357, 1721 – 1731
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1725

FULL PAPERS
Thomas Scheidt et al.
assumed wide substrate diversity is matched by struc-
tural variations in the substrate-binding tunnels of w-
transaminases, which are interesting for the selective
binding and conversion of various unnatural ketone
acceptors towards the development of future applica-
tions in the synthesis of a broad diversity of chiral
amines. Therefore, searching for yet unidentified w-
transaminases in databases for novel specificities is
still a key topic in the w-transaminase research arena.
Taurine-pyruvate transaminase is classified as
a member of subclade 6a among w-transaminases,
which catalyzes transamination between taurine and
pyruvate in several microorganisms.[9,29] However, its
substrate tolerance so far was only tested towards
a very limited set of keto acids,[30] and no synthetic
studies with unnatural acceptors have been published
Figure 2. Relative activity of w-transaminase from C. viola-
ceum with amine substrates rac-1a (left bar), rac-1b (middle
bar), and rac-1c (right bar).
as yet. Therefore, we included in this study two novel with substrate 1a carrying the smallest methyl sub-
thermostable taurine-pyruvate transaminases that stituent, the amine 1b having a larger ethyl moiety
were cloned from Geobacillus thermodenitrificans was indeed converted with a relative velocity of 14%,
(DSM465)
(DSM5366) and that we found to be highly expressed propyl substituent was below 1% (Figure 2).
in soluble form in recombinant E. coli host (details of The reason for this discrimination between the dif-
and
Geobacillus
thermoleovorans while the relative activity for 1c with the largest n-
the cloning and characterization to be published else- ferent alkyl substituents is the limited space in the
where). Here, we conducted an enantioselectivity small substrate pocket of the C. violaceum w-transa-
typing of these enzymes and measured their kinetic minase where a phenylalanine (Phe88) residue effec-
constants with 1a by the fluorogenic screening assay tively closes the pocket.[8] Enzyme engineering in the
(Table 1, entries 9 and 10). Accordingly, while both small substrate pocket, focusing on Phe88 and its
Geobacillus enzymes are (S)-selective, the transami- neighboring residues, should improve the activity for
nase from G. thermoleovorans showed a very high substrates bearing larger substituents like the one
enantioselectivity, although the absolute activity at present in pro-sitagliptin (Scheme 1). The assay prin-
378C with substrate 1a was only rather low. A poten- ciple reported here, and the modular synthesis of ap-
tial advantage of such thermostable w-transaminases propriate screening substrates, offer a fast and prom-
might be that higher reaction temperatures could fa- ising method to support such an experimental strat-
cilitate the evaporative removal of ketone, which egy.
arises from the sacrificial aminoalkane donor support,
to push the reaction equilibrium towards amine for-
mation.[31]
Conclusion and Outlook
The generation of brightly fluorescent acetonaph-
thone derivatives 2 from non-fluorescent chiral 1-(6-
methoxynaphth-2-yl)alkylamines 1 has been demon-
Determination of Substrate Promiscuity
An exemplary test was conducted to evaluate the po- strated to be a useful principle to quantify w-transa-
tential of the fluorogenic assay method to determine minase catalysis over a very broad range of enzyme
the substrate tolerance of w-transaminases, as a pre- activity. We could show that this new assay format is
requisite for protein engineering projects. Particularly, applicable for the reliable determination of kinetic
the option for systematic structural variation of the constants and, when employing enantiomerically pure
substrate by altering the R group in the amine amine substrates, for the enantioselectivity typing of
1 should be a valuable asset to probe and expand the a large variety of known enzymes, as well as for the
small substrate pocket of transaminases, as had to be discovery of novel w-transaminases of as yet unknown
performed in the search for an enzyme variant capa- specificity, as demonstrated for the two thermostable
ble of the processing of pro-sitagliptin.[4] Such a strat- taurine-pyruvate transaminases from G. thermodeni-
egy was also recently employed for enlarging the trificans and G. thermoleovorans. The assay is charac-
small substrate pocket of the Vibrio fluvialis w-transa- terized by high sensitivity in microtiter plate format
minase.[32] Thus, the transaminase from C. violaceum which, because of its simplicity and low substrate con-
as one of the most proficient and versatile candidates sumption, suggests its suitability even for large-scale
was tested against the three homologous substrates liquid screening campaigns in the directed evolution
1a–c in racemic form. When compared to the activity of optimized transaminases. The modular scheme de-
1726
asc.wiley-vch.de
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2015, 357, 1721 – 1731

FULL PAPERS
Fluorescence-Based Kinetic Assay for High-Throughput Discovery
the same supplemented with imidazole (500 mM; elution
buffer), respectively, were adjusted to pH 7.4.
veloped for the synthesis of substrates (including the
option of their kinetic racemate resolution) allows
a simple and rapid generation of broad structural var-
iations that offers a promising approach for the di-
rected evolution of transaminases, in particular for
the targeted modification of the small binding pocket.
Also, the high product stability under the assay condi-
tions particularly supports the monitoring of very
slow conversion rates from low-activity enzyme var-
iants, further suggesting the assayꢀs suitability for the
directed evolution of transaminases towards demand-
ing substrate structure changes when starting out
from incipient activity.
In principle, using the same substrates of type 1 the
same fluorogenic assay technology seems applicable
to the enzyme class of amine oxidases that likewise
are of high interest for complementary synthetic ap-
plications.[1] Research in this direction is currently on-
going in our laboratories and results will be communi-
cated in due course.
Enzyme Expression and Purification
The transaminases from Chromobacterium violaceum (Cv),
Vibrio fluvialis (Vf), Aspergillus fumigatus (AspFum), As-
pergillus terreus (AspTer) and Neosartorya fischeri (NeoFis)
were cloned and transformed as previously described.[12a,33]
Overnight cultures of 20 mL LB medium were prepared.
Antibiotics (50 mgmLÀ1 kanamycin for Cv and Vf,
100 mgmLÀ1 ampicillin for AspFum, AspTer and NeoFis)
were added and the cultures were inoculated. Cultures were
incubated overnight (378C, 220 rpm), then used to inoculate
180 mL of fresh LB medium (antibiotics added). Cv, Vf,
AspFum and NeoFis were induced by the addition of
0.4 mM IPTG and AspTer was induced by the addition of
0.2% (w/v) l-rhamnose. The induced cultures were incubat-
ed for 24 h (258C, 250 rpm). After expression, cells were re-
covered by centrifugation (8000 rpm, 15 min, 48C). Cell pel-
lets were dissolved in 5 mL buffer solution [IMAC binding
buffer for Cv and Vf and HEPES (50 mM, pH 8.2) for
AspFum, AspTer and NeoFis]. Cells were disrupted by soni-
cation, and cell debris was separated from the crude extract
by centrifugation (20000 rpm, 15 min, 48C). Cv was purified
as previously described[34] and the same protocol was used
for Vf. AspFum, AspTer and NeoFis were subjected to am-
monium sulfate precipitation and lyophilization as previous-
ly described.[3c] The Geobacillus enzymes were produced in
a similar fashion by expression from plasmids pET21a(+) in
E. coli BL21(DE3) cells in the presence of ampicillin with
IPTG induction. Proteins were purified by IMAC and con-
centrated by dialysis (MC>10 kD).
Experimental Section
Materials and Equipment
Column chromatography was performed on Merck 60 silica
gel (0.063–0.200 mesh); analytical thin layer chromatogra-
phy was performed on Merck silica gel plates 60 GF254
using UV absorption or anisaldehyde stain for detection.
Ketone 2a was obtained from TCI Europe and aldehyde 3
from Alfa Aesar. Commercial reagents/solvents were used
as received without further purification. NMR spectra were
recorded on Bruker AX300 or DRX500 spectrometers;
chemical shifts are referenced to CHCl3 (d=7.24 ppm).
Chiral HPLC analysis was performed on a Shimadzu 20A
system using a DaiCell Chiralpak IB column (4”250 mm)
and n-hexane containing 10–11% isopropyl alcohol and
0.6% TFA as eluent at a flow rate of 1.1 mLminÀ1 with pho-
tometric detection at 330 nm. Kinetic measurements in mi-
croplate format were performed on BMG Optima and
BMG Clariostar microplate readers. Transaminases were
purchased from Sigma (No. 77856, and 93006) and Codexis
Inc (ATA113 and ATA117). Immobilized lipase from C. ant-
arctica B was obtained from Sprin Technologies.
Enzyme Concentration and Activity Determination
Commercially available enzymes from Sigma (No. 77856,
ꢀ0.9 U/mg and 93006 NeoFis, ꢀ0.4 U/mg) were dissolved to
1.5 mgmLÀ1. Specific activity of non-commercial enzymes
was determined with the acetophenone assay[15] using
2.5 mM 1-phenylethylamine [(S)-enantiomer for Cv and Vf
and (R)-enantiomer for AspFum, AspTer and NeoFis] and
2.5 mM sodium pyruvate. Reaction rate was measured spec-
trophotometrically at 245 nm and the specific activity calcu-
lated. Enzyme concentrations of Cv and Vf were determined
according to Pace et al.[35]
Assay Set-up and Determination of Kinetic Constants
The assay was developed in Sarstedt 96-well plates using
a BMG Optima plate reader equipped with 330 nm filters
for excitation and 460 nm filters for emission. The standard
curve of the assay was determined by using a serial dilution
of ketone 2a in a concentration range of 0–5.0 mM in the
corresponding assay buffer in a 96-well plate. The relation-
ship between rate for formation of 2a and fluorescence was
measured to provide the standard curve as y=ax+b. LOD
was defined by LOD=3.3 Sb/¼, and LOQ was defined as
LOD=10 Sb/¼, where Sb is the standard deviation of the
control and ¼ is the slope of the standard curve.[23] Each
enzyme–substrate pair was measured in duplicate using
a serial 2” dilution from 10 to 0.16 mM of amine hydrochlo-
ride (2.4 mgmLÀ1) in PBS 2” or phosphate buffer. Each
Buffers and Solutions
PBS 2x buffer: An aqueous solution of NaCl (274 mM),
KCl (5.4 mM), Na2HPO4 (16.2 mM) and KH2PO4 (2.95 mM)
was adjusted to pH 7.2, and DMF was added (5%, v/v) as
co-solvent and the mixture diluted to 1 L.
Phosphate buffer:
A
0.1M solution of Na2HPO4
(14.2 gLÀ1) was adjusted to pH 8.20 by addition of 0.1M
HCl.
HEPES buffer: A solution of 50 mM (11.9 gLÀ1) HEPES
free acid was prepared and heated to 378C before the pH
was adjusted to 8.2 by addition of 1M NaOH.
IMAC buffers: Solutions containing Na2HPO4 (20 mM)
and NaCl (500 mM) in distilled water (binding buffer), and
Adv. Synth. Catal. 2015, 357, 1721 – 1731
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1727

FULL PAPERS
Thomas Scheidt et al.
well contained 140 mL of substrate solution, 10 mL of pyru-
vate stock (90 mM sodium pyruvate) and 50 mL of enzyme
solution. For measurements at higher temperature, the in-
strument was preheated to 388C. Raw data were processed
with Origin 8.1 for linear or hyperbolic fitting.
The assay was evaluated using the w-transaminases from
Cv (0.66 U/mg), Vf (15.8 U/mg), AspFum (0.019 U/mg),
AspTer (0.00078 U/mg), NeoFis (0.013 U/mg) and Codexis
enzymes ATA113 (0.45 U/mg) and ATA117 (1.9 U/mg).
Costar 96 well plates were used in a BMG Clariostar plate
reader with 330 nm excitation and 460 nm emission set for
detection. All stock solutions were prepared with HEPES
buffer (50 mM, pH 8.2). A stock solution of 10 mM amine
hydrochloride 1a was made and diluted in a series of 2” di-
lutions to get a good coverage of the Michaelis–Menten ki-
d=157.6 (C-6’), 140.0 (C-2), 134.1 (C-4a’), 129.4 (C-8’),
128.7 (C-1’), 127.1 (C-4’), 124.7 (C-3’), 124.5 (C-8a’), 118.9
(C-7’), 105.7 (C-5’), 76.2 (C-1), 55.3 (OCH3), 34.9 (C-2), 10.2
(C-3).
1-(6-Methoxynaphth-2-yl)butan-1-ol (4c): Yield: 1.74 g
(75%); colorless crystals; mp 758C; 1H NMR (CDCl3,
300 MHz): d=7.70–7.43 (m, 3H, 4’-, 8’-, 1’-H), 7.41 (m, 1H,
5’-H), 7.15 (m, 2H, 3’-, 7’-H), 4.67 (t, J=6.5 Hz, 1H, 1-H),
3.88 (s, 3H, OCH3), 1.95–1.76 (m, 2H, 2-H), 1,72 (br, 1H,
OH), 1.53–1,40 (m, 2H, 3-H), 1.38 (t, J=7.2 Hz, 3H, 4-H);
13C NMR (CDCl3, 75 MHz): d=157.6 (C-6’), 140.0 (C-2),
134.1 (C-4a’), 129.4 (C-8’), 128.7 (C-1’), 127.1 (C-4’), 124.7
(C-3’), 124.5 (C-8a’), 118.9 (C-7’), 105.7 (C-5’), 74.5 (C-1),
55.5 (C-9’), 41.1 (C-2), 19.1 (C-3), 14.0 (C-4).
netics. Furthermore,
a sodium pyruvate stock solution
(90 mM) was prepared and enzymes were applied in differ-
ent amounts to give readable rates for each enzyme and
substrate enantiomer (see Table 2 in the Supporting Infor-
mation). Wells were charged as above. All measurements
were done in triplicate at 378C. Raw data were exported to
MS Excel where they were treated and fitted appropriately.
b) Oxidation to Ketone
The alcohol 4 (22 mmol) was dissolved in 40 mL dry, freshly
distilled acetone and cooled to 08C in an ice bath., A 4M
solution of CrO3 (8 mL, 32 mmol) in 15% H2SO4 was added
with under argon at such a rate that the temperature re-
mained below 58C. After stirring for 1 h aqueous NaHSO3
was added to remove excess oxidant. The product was pre-
cipitated by addition of water (30 mL) and isolated by filtra-
tion. The residue was re-dissolved in diethyl ether and the
solution washed consecutively with 10% aqueous Na2CO3,
water and brine. Evaporation provided ketones 2 as slightly
brownish solids, which can be purified by recrystallization
from aq ethanol.
Determination of Substrate Promiscuity
Comparison of the specific activities of Cv w-transaminase
towards the substrates rac-1a–c was performed with the
same assay set-up as before. Amine concentration was set to
1 mM and sodium pyruvate concentration to 4.5 mM. Quan-
tities of 0.0005 U of enzyme were added to 1a, 0.01 U to 1b
and 0.05 U to 1c.
1-(6-Methoxynaphth-2-yl)propan-1-one (2b): Yield: 4.46 g
(90%); colorless crystals; mp 1128C (Lit.:[18b] 111.58C);
1H NMR (CDCl3, 300 MHz): d=8.42 (d, J=1.7 Hz 1H, 1’-
H), 8.04 (dd, J=8.7, 1.8 Hz, 1H, 8’-H), 7.86 (d, J=8.7 Hz,
1H, 3’-H), 7.78 (d, J=8.7 Hz, 1H, 4’-H), 7.18 (m, 2H, 5’-, 7’-
H), 3.96 (s, 3H, OCH3), 3.12 (q, J=7.3 Hz, 2H, 2-H), 1.30
(t, J=7.3 Hz, 3H, 3-H); 13C NMR: (CDCl3, 75 MHz): d=
200.5 (C-1), 159.7 (C-6’), 137.2 (C-4a’), 132.3 (C-2’), 131.1
(C-8’), 129.4 (C-8a’), 127.9 (C-1’), 127.1 (C-4’), 124.7 (C-3’),
119.6 (C-7’), 105.7 (C-5’), 55.4 (OCH3), 31.7 (C-2), 8.5 (C-3).
1-(6-Methoxynaphth-2-yl)butan-1-one (2c): Yield: 4.26 g
(86%); colorless crystals; mp 888C (Lit:[36] 89–908C);
1H NMR (CDCl3, 300 MHz): d=8.42 (d, J=1.7 Hz 1H, 1’-
H), 8.04 (dd, J=8.7, 1.8 Hz, 1H, 8’-H), 7.86 (d, J=8.7 Hz,
1H, 3’-H), 7.78 (d, J=8.7 Hz, 1H, 4’-H), 7.18 (m, 2H, 5’-, 7’-
H), 3.96 (s, 3H, OCH3), 3.05 (t, J=7.3 Hz, 2H, H-2), 1.82
(m, 2H, H-3), 1.10 (t, J=7.2 Hz, 3H, H-4); 13C NMR
(CDCl3, 75 MHz): d=200.5 (C-1), 159.7 (C-6’), 137.2 (C-
4a’), 132.3 (C-2’), 131.1 (C-8’), 129.4 (C-8a’), 127.9 (C-1’),
127.1 (C-4’), 124.7 (C-3’), 119.6 (C-7’), 105.7 (C-5’), 55.4
(OCH3), 40.4 (C-2), 18.0 (C-3), 14.0 (C-4).
General Procedure for Substrate Synthesis
a) Grignard Coupling
Magnesium turnings (0.5 g, 22 mmol) were ground in
a mortar under an octane layer to activate the metal surface
then transferred into a two-neck flask equipped with reflux
condenser and addition funnel and dry diethyl ether
(20 mL) was added under an argon atmosphere. A solution
of the appropriate bromoalkane (22 mol) in 20 mL dry di-
ethyl ether was slowly added through the addition funnel to
maintain mild reflux. When all magnesium was consumed,
a
solution of 6-methoxynaphth-2-aldehyde
3
(2.0 g,
11 mmol) in 5 mL dry THF was slowly added. After com-
plete consumption of 3, the mixture was quenched by addi-
tion of saturated aqueous ammonium chloride. The organic
layer was separated and the remaining aqueous layer was
extracted with ethyl acetate (3”25 mL). The combined or-
ganic layers were dried over Na2SO4, decolorized with char-
coal and concentrated to dryness under vacuum. The residue
was re-dissolved in 20 mL of THF and 10 mL n-heptane
were added. The solution was concentrated under vacuum
at 458C until turbidity, warmed to clearness again and the
product was allowed to crystallize overnight at 28C.
c) Reductive Amination
Ketone 2 (20 mmol) was suspended in titanium(IV) isoprop-
oxide (12 mL, 40 mmol) under argon and the mixture stirred
for 15 min. Dry ethanol (55 mL) was cooled to À208C and
saturated with gaseous ammonia to a final concentration of
approx. 2M (100 mmol) and the solution was added quickly
to the ketone suspension. The resulting reaction mixture
1-(6-Methoxynaphth-2-yl)propan-1-ol (4b): Yield: 1.87 g
(76%); colorless crystals; mp 718C; 1H NMR (CDCl3,
300 MHz): d=7.70–7.43 (m, 3H 4’-, 8’-, 1’-H), 7.41 (m, 1H,
5’-H), 7.15 (m, 2H, 3’-, 7’-H), 4.67 (t, J=6.5 Hz, 1H, 1-H),
3.88 (s, 3H, OCH3), 2.19 (br, 1H, OH) 1.85 (m, 2H, 2-H),
0.91 (t, J=7.5 Hz, 3H, 3-H); 13C NMR (CDCl3, 75 MHz):
1728
asc.wiley-vch.de
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2015, 357, 1721 – 1731

FULL PAPERS
Fluorescence-Based Kinetic Assay for High-Throughput Discovery
was stirred overnight. Sodium borohydride (1.1 g, 26 mmol)
was added in one portion and the mixture stirred for 3 h at
room temperature. The reaction was then quenched by
pouring into ammonium hydroxide (2M, 50 mL), the result-
ing inorganic precipitate was removed by filtration and
washed with ethyl acetate (3”25 mL). The aqueous layer
was extracted with ethyl acetate (3”75 mL). The combined
organic layers were dried over Na2SO4 and evaporated. The
solid residue was dissolved in MTBE (25 mLgÀ1) and
a stream of dry HCl gas was passed through the solution for
ten seconds using a single-use Pasteur pipette as nozzle. The
precipitate was isolated by filtration through a G3 frit,
washed with cold ether and dried under vacuum. Analytical
samples were obtained by crystallization from dry acetone.
1-(6-Methoxynaphth-2-yl)ethylamine·HCl (rac-1a): Yield:
3.42 g (85%); colorless crystals; mp 2208C; 1H NMR
(DMSO-d6, 300 MHz): d=8.63 (br, 3H, NH), 7.95 (m, J=
12.0 Hz, 1H, 8’-H), 7.84 (dd, J=8.7, 16,0 Hz, 2H, 1’-, 4’-H),
7.66 (dd, J=8.8, 1.9 Hz, 1H, 5’-H), 7.35 (d, J=2.5 Hz, 1H,
7’-H), 7.20 (dd, J=2,5, 8.8 Hz, 1H, 3’-H), 4.50 (q, J=6.7 Hz,
1H, 1-H), 3.88 (s, 3H, OCH3), 1.60 (d, J=6.7 Hz, 3H, 2-H);
13C NMR: (CDCl3, 75 MHz): d=157.6 (C-6’), 134.3 (C-2’),
134.0 (C-4a’), 129.3 (C-8’), 128.0 (C-1’), 127.2 (C-4’), 125.7
(C-3’), 125.1 (C-8a’), 119.1 (C-7’), 105.8 (C-5’), 55.2 (OCH3),
50.0 (C-1), 20.6 (C-2).
methoxyacetate (8.0 mL, 62 mmol) and 7.0 g immobilized
CAL-B lipase the mixture was gently stirred at 42–458C for
about 40 h with monitoring by chiral HPLC. As the reaction
proceeded a white precipitate was formed, mainly consisting
of the (S)-amine while the (R)-methoxyacetamide remains
in solution. After complete consumption of the (R)-enantio-
mer the mixture was filtered. The clear filtrate was cooled
to 08C and treated with HCl gas to precipitate remaining
dissolved (S)-amine as hydrochloride, which was filtered off.
Combined solids obtained by filtration containing (S)-1a
and the remaining solution containing the (R)-amide 5a
were further treated separately.
(S)-1-(6-Methoxynaphth-2-yl)ethylamine·HCl
[(S)-1a]:
The collective solids obtained from the filtration steps
above were treated with water to re-dissolve the (S)-amine
salts for separation from the enzyme beads. Solid NaOH
was added to the solution until pH>10 and the amine was
extracted with diethyl ether. The organic layer was separat-
ed, dried over Na2SO4 and processed by treatment with HCl
gas as described above to yield the (S)-amine as hydrochlo-
ride in >99% ee; yield: 4.79 g (48%); spectroscopic data
were identical to those for rac-1a.
(R)-2-Methoxy-N-[1-(6-methoxynaphth-2-yl)ethyl]acet-
amide [(R)-5a]: The remaining filtrate was evaporated to
yield the (R)-amide 5a in >95% ee. A single recrystalliza-
tion of the amide from aqueous ethanol raised the ee to
>99%; yield: 4.7 g (41%); colorless crystalline solid; mp
1-(6-Methoxynaphth-2-yl)propylamine·HCl
(rac-1b):
Yield: 2.6 g (65%); colorless crystals; mp 255–2578C;
1H NMR (DMSO-d6, 300 MHz): d=8.63 (br, 3H, NH), 7.95
(m, J=12.0 Hz, 1H, 8’-H), 7.84 (dd, J=8.7, 16,0 Hz, 2H, 1’-,
4’-H), 7.66 (dd, J=8.8, 1.9 Hz, 1H, 5’-H), 7.35 (d, J=2.5 Hz,
1H, 7’-H), 7.20 (dd, J=2,5, 8.8 Hz, 1H, 3’-H), 4.50 (q, J=
6.7 Hz, 1H, 1-H), 3.88 (s, 3H, OCH3), 2.08 (m, 1H, H-2),
1.94 (m, 1H, H-2), 0.76 (t, J=7,5 Hz, 3H, 3-H); 13C NMR
(CDCl3, 75 MHz): d=157.6 (C-6’), 134.3 (C-2’), 134.0 (C-
4a’), 129.3 (C-8’), 128.0 (C-1’), 127.2 (C-4’), 125.7 (C-3’),
125.1 (C-8a’), 119.1 (C-7’), 105.8 (C-5’), 55.2 (OCH3), 52.2
(C-1), 27.2 (C-2), 10.0 (C-3); EI-MS: m/z=215.1304, calcd.
for C14H17NO [M]+: 215.1310.
1-(6-Methoxynaphth-2-yl)butylamine·HCl (rac-1c): Yield:
2.7 g (65%); colorless crystals; mp 272–2738C; 1H NMR
(DMSO-d6, 300 MHz): d=8.63 (br, 3H, NH), 7.95 (m, J=
12.0 Hz, 1H, 8’-H), 7.84 (dd, J=8.7, 16,0 Hz, 2H, 1’-, 4’-H),
7.66 (dd, J=8.8, 1.9 Hz, 1H, 5’-H), 7.35 (d, J=2.5 Hz, 1H,
7’-H), 7.20 (dd, J=2,5, 8.8 Hz, 1H, 3’-H), 4.50 (q, J=6.7 Hz,
1H, 1-H), 3.88 (s, 3H, OCH3), 1.24 (m, 1H, 2-H), 1.13 (m,
1H, 3-H), 0.86 (t, J=7,5 Hz, 3H, 4-H); 13C NMR (CDCl3,
75 MHz): d=157.6 (C-6’), 134.3 (C-2’), 134.0 (C-4a’), 129.3
(C-8’), 128.0 (C-1’), 127.2 (C-4’), 125.7 (C-3’), 125.1 (C-8a’),
119.1 (C-7’), 105.8 (C-5’), 56.2 (OCH3), 54.4 (C-1), 36.7 (C-
2), 18.5 (C-3), 13.4 (C-4); EI-MS: m/z=229.1479, calcd. for
C15H19NO [M]+: 229.1467.
1
948C; H NMR (DMSO-d6, 300 MHz): d=7.71 (m, 3H, 8’-,
1’-, 4’-H), 7.69 (dd, J=8.8, 1.9 Hz, 1H, 5’-H), 7.15 (m, 2H,
3’-, 7’-H), 6.83 (bd, J=8.5 Hz, 1H, NH), 5.33 (dq, J=8.5,
6.9 Hz, 1H, 1-H), 3.98 (s, 1H, 2“-H), 3.92 and 3.41 (2 s, 2”
3H, 2 OCH3), 1.60 (d, J=6.9 Hz, 3H, 2-H); 13C NMR
(CDCl3, 75 MHz): d=167.6 (C1”), 156.7 (C-6’), 136.3 (C-2’),
132.8 (C-4a’), 128.3 (C-8’), 127.7 (C-1’), 126.3 (C-4’), 124.2
(C-3’), 123.4 (C-8a’), 118.0 (C-7’), 104.6 (C-5’), 70.9 (C-2’’),
58.1 and 56.2 (2”OCH3), 47.0 (C-1), 20.7 (C-2).
(R)-1-(6-methoxynaphth-2-yl)ethylamine·HCl
[(R)-1a]:
The (R)-amide obtained from two batches (10.0 g,
36.3 mmol) was dispersed in 500 mL of half concentrated
aqueous HCl and the mixture was stirred at 708C under
argon until the solution became clear. The acid was carefully
removed under vacuum and the solid residue dispersed in
ether. The suspension was saturated with HCl gas and fil-
tered through a G3 frit. The residue was washed with cold
ether and dried under vacuum to afford the (R)-amine as
hydrochloride in >99% ee; yield: 5.54 g (64%); spectro-
scopic data identical to those for rac-1a.
Chiral HPLC Analysis
HPLC: (ChiralPac IB column, mobile phase: n-hexane/iso-
propyl
1.3 mLminÀ1, absorption at 330 nm): Rf 10.2 min=(R)-
amide, 12.6 min=(S)-amide, 20.9 min=(S)-amine·HCl,
24.8 min=(R)-amine·HCl.
alcohol/TFA=100/11/0.5
(v/v/v),
flow
rate
d) Racemate Resolution
Enantioselective Substrate Synthesis from Naproxene
The hydrochloride of amine rac-1a (10.0 g, 42 mmol) was
suspended in 250 mL of MTBE in a 500-mL Erlenmeyer
flask and 4N NaOH (15 mL, 60 mmol) was added with vigo-
rous stirring. After 15 min the layers were separated, and
the organic phase was washed with water (100 mL) then
dried over anhydrous Na2CO3. After addition of isopropyl
(2S)-(6-Methoxynaphth-2-yl)propanoyl chloride: Naproxene
acid was extracted from commercial drug tablets by Soxhlet
extraction with CH2Cl2 (quant.). To a stirred solution of
5.0 g (21.7 mmol) naproxen acid in 40 mL of CH2Cl2 con-
taining 4 drops of dry DMF was added thionyl chloride
Adv. Synth. Catal. 2015, 357, 1721 – 1731
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1729

FULL PAPERS
Thomas Scheidt et al.
(2.0 mL, 27.5 mmol) dropwise at room temperature over References
10 min. The resulting mixture was stirred at room tempera-
ture under a nitrogen atmosphere for 3 h, then concentrated
under vacuum to afford crude naproxen acid chloride as
a yellowish solid, which was used directly for the next step
without purification.
[1] D. Ghislieri, N. J. Turner, Top. Catal. 2014, 57, 284–300.
[2] M. Hoehne, U. T. Bornscheuer, ChemCatChem 2009, 1,
42–51.
[3] a) D. Koszelewski, K. Tauber, K. Faber, W. Kroutil,
Trends Biotechnol. 2010, 28, 324–332; b) M. Hoehne,
U. T. Bornscheuer, in: Enzyme Catalysis in Organic
Synthesis, 3rd edn., Vol. 2, (Eds.: K. Drauz, H. Grçger,
O. May), Wiley-VCH, Weinheim, 2011, pp 779–820;
c) S. Schaetzle, F. Steffen-Munsberg, A. Thontowi, M.
Hoehne, K. Robins, U. T. Bornscheuer, Adv. Synth.
Catal. 2011, 353, 2439–2445; d) S. Mathew, H. Yun,
ACS Catal. 2012, 2, 993–1001; e) W. Kroutil, E.-M. Fi-
schereder, C. S. Fuchs, H. Lechner, F. G. Mutti, D.
Pressnitz, A. Rajagopalan, J. H. Sattler, R. C. Simon, E.
Siirola, Org. Process Res. Dev. 2013, 17, 751–759;
f) R. C. Simon, N. Richter, E. Busto, W. Kroutil, ACS
Catal. 2014, 4, 129–143; g) R. C. Simon, N. Richter, W.
Kroutil, in: Cascade Biocatalysis, (Eds.: S. Riva, W.-D.
Fessner), Wiley-VCH, Weinheim, 2014, pp 65–86.
[4] C. K. Savile, J. M. Janey, E. C. Mundorff, J. C. Moore,
S. Tam, W. R. Jarvis, J. C. Colbeck, A. Krebber, F. J.
Fleitz, J. Brands, P. N. Devine, G. W. Huisman, G. J.
Hughes, Science 2010, 329, 305–309.
(2S)-(6-Methoxynaphth-2-yl)propanoylamide: A 500-mL
flask equipped with a mechanical stirrer, a 100-mL dropping
funnel and external cooling by an ice-salt mixture was
charged with 50 mL of cold concentrated aqueous ammonia
(25%), and a solution of naproxen acid chloride (5.0 g,
20 mmol) in 50 mL MTBE was carefully added dropwise
with rapid stirring, maintaining the temperature below
158C. After the addition was complete, the mixture was
stirred for one hour at room temperature. After evaporation
to dryness under vacuum the solid residue was boiled with
100 mL of dry ethyl acetate for 10 min, and the hot solution
was filtered rapidly through a pre-warmed glass frit. The
filter cake was extracted twice with warm ethyl acetate (2”
15 mL). The combined organic extracts were cooled to 08C,
and the precipitated crystalline amide was isolated by filtra-
tion. The mother liquor was concentrated to furnish
a second crop of amide. The combined solids were dried,
first in an oven at 708C for 3 h, then in a vacuum desiccator;
1
yield: 3.92 g (85%); colorless crystals. H NMR (DMSO-d6,
300 MHz): d=7.80–7.65 (m, 3H, 4’-, 8’-, 1’-H), 7.41 (dd, J=
8.6, 1.9 Hz, 1H, 5’-H), 7.24–7.09 (m, 2H, 3’-, 7’-H), 5,59 and
5,42 (2 bd, 2H, NH), 3.93 (s, 3H, OCH3), 3.74 (q, J=7.2 Hz,
1H, 2-H), 1.61 (d, J=7.2 Hz, 3H; 3-H); 13C NMR (CDCl3,
75 MHz): d=176.8 (C-1), 127.8 (C-6’), 136.3 (C-4a’), 133.8
(C-2’), 129.2 (C-8’), 129.0 (C-1’), 127.6 (C-4’), 126.1 (C-3’),
126.1(C-8a’), 119.2 (C-7’), 105.7 (C-5’), 55.3 (OCH3) 45.5 (C-
2), 18.2 (C-3).
(S)-1-(6-Methoxynaphth-2-yl)ethylamine·HCl [(S)-1a]: In
a 500-mL flask equipped with reflux condenser and septum
were placed 100 mL of 1N NaOH, and elemental bromine
(1.0 mL, 20 mmol) was added through a cannula with inten-
sive stirring at 08C. After 15 min naproxenamide (4.0 g,
17.5 mmol) was added in one portion and the mixture was
stirred with heating to 608C until TLC indicated complete
conversion. Excess of bromine was removed by addition of
saturated sodium sulfite solution (ca. 5 mL) and the aqueous
phase was extracted with MTBE (6”50 mL). The combined
organic phases were dried over Na2SO4 and the product (S)-
1a was precipitated as the hydrochloride as described
above; yield: 2.10 g (50%); colorless powder.
[5] a) W.-D. Fessner, New Biotechnol. 2014, 31, S74; b) W.-
D. Fessner, New Biotechnol. 2015, 32, DOI:10.1016/
j.nbt.2014.11.007; c) D. Tessaro, L. Pollegioni, L. Piubel-
li, P. D’Arrigo, S. Servi, ACS Catal. 2015, 5, 1604–1608.
[6] W. D. Fessner, C. Walter, Angew. Chem. 1992, 104,
643–645; Angew. Chem. Int. Ed. Engl. 1992, 31, 614–
616.
[7] J. H. Sattler, M. Fuchs, F. G. Mutti, B. Grischek, P.
Engel, J. Pfeffer, J. M. Woodley, W. Kroutil, Angew.
Chem. Int. Ed. 2014, 53, 14153–14157.
[8] a) M. S. Humble, K. E. Cassimjee, M. Haakansson,
Y. R. Kimbung, B. Walse, V. Abedi, H.-J. Federsel, P.
Berglund, D. T. Logan, FEBS J. 2012, 279, 779–792;
b) M. S. Humble, K. E. Cassimjee, V. Abedi, H.-J. Fed-
ersel, P. Berglund, ChemCatChem 2012, 4, 1167–1172;
c) C. Sayer, M. N. Isupov, A. Westlake, J. A. Littlechild,
Acta Crystallogr. Sect. D 2013, 69, 564–576.
[9] C. Rausch, A. Lerchner, A. Schiefner, A. Skerra, Pro-
teins Struct. Funct. Bioinf. 2013, 81, 774–787.
[10] K. S. Midelfort, R. Kumar, S. Han, M. J. Karmilowicz,
K. McConnell, D. K. Gehlhaar, A. Mistry, J. S. Chang,
M. Anderson, A. Villalobos, J. Minshull, S. Govindara-
jan, J. W. Wong, Protein Eng. Des. Sel. 2013, 26, 25–33.
[11] a) F. Steffen-Munsberg, C. Vickers, A. Thontowi, S.
Schaetzle, T. Tumlirsch, M. Svedendahl Humble, H.
Land, P. Berglund, U. T. Bornscheuer, M. Hoehne,
ChemCatChem 2013, 5, 150–153; b) A. Lyskowski, C.
Gruber, G. Steinkellner, K. Steiner, M. Schurmann, H.
Schwab, K. Gruber, PloS one 2014, 9, e87350; c) M.
Thomsen, L. Skalden, G. J. Palm, M. Hohne, U. T.
Bornscheuer, W. Hinrichs, Acta Crystallogr. Sect. D
2014, 70, 1086–1093; d) C. Sayer, R. J. Martinez-Torres,
N. Richter, M. N. Isupov, H. C. Hailes, J. A. Littlechild,
J. M. Ward, FEBS J. 2014, 281, 2240–2253.
Acknowledgements
This work was funded by TU Darmstadt and KTH Royal In-
stitute of Technology, as well as by the National Natural Sci-
ence Funds of China (Grant No. 21406069 to DY) and the
Fundamental Research Funds for the Central Universities
and China Postdoctoral Science Foundation (Project No.
2013M541490 to DY). We thank M. Briesenick for his assis-
tance in developing the enantioselective synthesis of amine
1a. We gratefully acknowledge ESF project COST CM1303
for supporting this collaboration.
[12] a) M. Hoehne, S. Schaetzle, H. Jochens, K. Robins,
U. T. Bornscheuer, Nat. Chem. Biol. 2010, 6, 807–813;
b) E.-S. Park, M. Kim, J.-S. Shin, Appl. Microbiol. Bio-
1730
asc.wiley-vch.de
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2015, 357, 1721 – 1731

FULL PAPERS
Fluorescence-Based Kinetic Assay for High-Throughput Discovery
technol. 2012, 93, 2425–2435; c) F. Steffen-Munsberg, C.
Vickers, A. Thontowi, S. Schaetzle, T. Meinhardt, M.
Svedendahl Humble, H. Land, P. Berglund, U. T. Born-
scheuer, M. Hoehne, ChemCatChem 2013, 5, 154–157.
[13] S. Mathew, G. Shin, M. Shon, H. Yun, Biotechnol. Bio-
process Eng. 2013, 18, 1–7.
[14] a) A. R. Martin, R. DiSanto, I. Plotnikov, S. Kamat, D.
Shonnard, S. Pannuri, Biochem. Eng. J. 2007, 37, 246–
255; b) T. Sehl, R. C. Simon, H. C. Hailes, J. M. Ward,
U. Schell, M. Pohl, D. Rother, J. Biotechnol. 2012, 159,
188–194; c) K. Weinhandl, M. Winkler, A. Glieder, A.
Camattari, Tetrahedron 2012, 68, 7586–7590; d) A. P.
Green, N. J. Turner, E. OꢀReilly, Angew. Chem. Int. Ed.
2014, 53, 10714–10717.
[15] S. Schaetzle, M. Hoehne, E. Redestad, K. Robins, U. T.
Bornscheuer, Anal. Chem. 2009, 81, 8244–8248.
[16] A. K. Samland, M. Rale, G. A. Sprenger, W.-D. Fess-
ner, ChemBioChem 2011, 12, 1454–1474.
[17] a) J. Wierzchowski, W. P. Dafeldecker, B. Holmquist,
B. L. Vallee, Anal. Biochem. 1989, 178, 57–62; b) B.
List, C. F. Barbas, R. A. Lerner, Proc. Natl. Acad. Sci.
USA 1998, 95, 15351–15355.
[22] a) H. Spahn, P. Langguth, Pharm. Res. 1990, 7, 1262–
1268; b) A. G. M. Al-Sehemi, A. M. S. El-Sharief, Y. A.
Ammar, Ind. J. Chem. 2006, 45B, 450–455.
[23] a) G. L. Long, J. D. Winefordner, Anal. Chem. 1983, 55,
712A–724A; b) Analytical Methods Committee, Ana-
lyst 1987, 112, 199–204.
[24] J.-H. Zhang, T. D. Y. Chung, K. R. Oldenburg, J.
Biomol. Screening 1999, 4, 67–73.
[25] J. S. Shin, B. G. Kim, Biotechnol. Bioeng. 1998, 60, 534–
540.
[26] J. S. Shin, B. G. Kim, Biotechnol. Bioeng. 2002, 77, 832–
837.
[27] C. S. Chen, Y. Fujimoto, G. Girdaukas, C. J. Sih, J. Am.
Chem. Soc. 1982, 104, 7294–7299.
[28] L. E. Janes, R. J. Kazlauskas, J. Org. Chem. 1997, 62,
4560–4561.
[29] K. Yonaha, S. Toyama, M. Yasuda, K. Soda, FEBS
Lett. 1976, 72, 21–24.
[30] H. Laue, A. M. Cook, Eur. J. Biochem. 2000, 267,
6841–6848.
[31] H. Yun, B.-K. Cho, B.-G. Kim, Biotechnol. Bioengin.
2004, 87, 772–778.
[32] A. Nobili, F. Steffen-Munsberg, H. Kohls, I. Trentin, C.
Schulzke, M. Hçhne, U. T. Bornscheuer, ChemCat-
Chem 2015, 7, 757–760.
[33] a) K. E. Cassimjee, M. S. Humble, V. Miceli, C. G. Co-
lomina, P. Berglund, ACS Catal. 2011, 1, 1051–1055;
b) J. S. Shin, H. Yun, J. W. Jang, I. Park, B. G. Kim,
Appl. Microbiol. Biotechnol. 2003, 61, 463–471.
[34] M. Anderson, S. Afewerki, P. Berglund, A. Córdova,
Adv. Synth. Catal. 2014, 356, 2113–2118.
[18] a) R. D. Haworth, G. Sheldrick, J. Chem. Soc. 1934,
864–867; b) W. Johnson, J. Am. Chem. Soc. 1950, 72,
925–935; c) C. Giordano, M. Villa, R. Annuziata,
Synth. Commun. 1990, 20, 383–392.
[19] B. Miriyala, S. Bhattacharyya, J. S. Williamson, Tetrahe-
dron 2004, 60, 1463–1471.
[20] a) F. Balkenhohl, K. Ditrich, B. Hauer, W. Ladner, J.
prakt. Chem. 1997, 339, 381–384; b) U. Karl, A. Simon,
Chim. Oggi 2009, 27, 66–69.
[35] C. N. Pace, F. Vajdos, L. Fee, G. Grimsley, T. Gray, Pro-
tein Sci. 1995, 4, 2411–2423.
[36] R. Gay, A. Horeau, Bull. Soc. Chim. Fr. 1955, 955–962.
[21] R. J. Kazlauskas, A. N. E. Weissfloch, A. T. Rappaport,
L. A. Cuccia, J. Org. Chem. 1991, 56, 2656–2665.
Adv. Synth. Catal. 2015, 357, 1721 – 1731
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1731