Directed evolution of an amine transaminase for the synthesis of an Apremilast intermediate via kinetic resolution

Article abstract of DOI:10.1016/j.bmc.2021.116271

Apremilast is an important active pharmaceutical ingredient that relies on a resolution to produce the key chiral amine intermediate. To provide a new catalytic and enzymatic process for Apremilast, we performed the directed evolution of the amine transaminase from Vibrio fluvialis. Six rounds of evolution resulted in the VF-8M-E variant with > 400-fold increase specific activity over the wildtype enzyme. A homology model of VF-8M-E was built and a molecular docking study was performed to explain the increase in activity. The purified VF-8M-E was successfully applied to produce the key chiral amine intermediate in enantiopure form and 49% conversion via a kinetic resolution, representing a new enzymatic access towards Apremilast.

Full text of DOI:10.1016/j.bmc.2021.116271

Bioorg. Med. Chem. 43 (2021) 116271
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Directed evolution of an amine transaminase for the synthesis of an
Apremilast intermediate via kinetic resolution
,
,*
Chao Xianga, Shuke Wua b, Uwe T. Bornscheuera
a Department of Biotechnology and Enzyme Catalysis, Institute of Biochemistry, University of Greifswald, Felix Hausdorff-Str. 4, 17487 Greifswald, Germany
b State Key Laboratory of Agricultural Microbiology, College of Life Science and Technology, Huazhong Agricultural University, No. 1 Shizishan Street, Wuhan 430070,
PR China
A R T I C L E I N F O
A B S T R A C T
Keywords:
Apremilast is an important active pharmaceutical ingredient that relies on a resolution to produce the key chiral
amine intermediate. To provide a new catalytic and enzymatic process for Apremilast, we performed the directed
evolution of the amine transaminase from Vibrio fluvialis. Six rounds of evolution resulted in the VF-8M-E variant
with > 400-fold increase specific activity over the wildtype enzyme. A homology model of VF-8M-E was built and
a molecular docking study was performed to explain the increase in activity. The purified VF-8M-E was suc-
cessfully applied to produce the key chiral amine intermediate in enantiopure form and 49% conversion via a
kinetic resolution, representing a new enzymatic access towards Apremilast.
Biocatalysis
Chiral amine
Directed evolution
Kinetic resolution
Transaminase
1. Introduction
tumor necrosis factor-
α
.
11 Apremilast was developed by Celgene for the
treatment of autoimmune and inflammatory diseases (plaque psoriasis
and psoriatic arthritis) under the name of Otezla®. The (S)-enantiomer
of Apremilast is much more potent than the (R)-enantiomer,10 thus the
(S)-form was further developed into the API. The original synthetic route
to Apremilast involves the resolution of racemic 1-(3-ethoxy-4-
methoxyphenyl)-2-(methylsulfonyl)ethanamine (rac-1) to (S)-1 with
stoichiometric N-Ac-ʟ-leucine (Scheme 1a).10 The asymmetric hydro-
genation approach had been explored to produce (S)-1 from the corre-
sponding enamine or ketone with precious and toxic rhodium or
ruthenium complexes.12 To our knowledge, there is only one report of an
enzyme-mediated synthesis of (S)-1: lipase (Novozym 435) had been
employed for the resolution of rac-1 via acylation,13 yet the ee of acyl-
ated product was not perfect.
The chiral amine group is a prestigious motif widely found in
bioactive natural products and man-made compounds, such as phar-
maceuticals and agrochemicals. Many optically pure amines are key
chiral building blocks for these bioactive chemicals and are also useful
chiral ligands, or chiral reagents for resolution. The common chemical
methods to access enantiopure chiral amines include traditional reso-
lution using chiral acids and enantioselective catalysis with metal cat-
alysts and chiral ligands.1 However, these methods suffer from the use of
stoichiometric chiral resolution reagents, or toxic transition metals and
costly complex ligands. On the other hand, nature’s catalysts, enzymes,
provide a catalytic, highly selective, and environmentally friendly
alternative method for the production of various chiral chemicals.2 For
the synthesis of chiral amines, the enzyme toolbox includes lipases,3
transaminases,4 amine dehydrogenases,5 monoamine oxidases,6 and
imine reductases.7 Despite the availability of these enzymes, due to the
relatively high substrate specificity, natural enzymes usually require
extensive directed evolution8 for the production of a specific target,
especially for the active pharmaceutical ingredients (APIs).9
As part of our research focus on amine transaminases (ATAs),14–16
we were interested in developing a novel ATA-based method to produce
(S)-1 in enantiopure form (Scheme 1b), and thus to provide a catalytic,
environmentally friendly, and economic process for Apremilast.
2. Results and discussion
A particularly interesting API containing a chiral amine group is
Apremilast, (S)-N-{2-[1-(3-Ethoxy-4-methoxyphenyl)-2-methanesulfo
nylethyl]-1,3-dioxo-2,3-dihydro-1H-isoindol-4-yl}acetamide.10 It is an
analog of thalidomide and a potent inhibitor of phosphodiesterase 4 and
2.1. Identification of a suitable ATA
We screened a range of wild-type (S)-selective ATAs and some of
* Corresponding author.
E-mail address: uwe.bornscheuer@uni-greifswald.de (U.T. Bornscheuer).
https://doi.org/10.1016/j.bmc.2021.116271
Received 29 March 2021; Received in revised form 30 May 2021; Accepted 4 June 2021
Available online 10 June 2021
0968-0896/© 2021 Elsevier Ltd. All rights reserved.

C. Xiang et al.
Bioorganic & Medicinal Chemistry 43 (2021) 116271
their mutants available in our laboratory for the conversion of rac-1 with
pyruvate, and analyzed the reaction products by HPLC. Although several
wild-type and engineered (S)-selective ATAs displayed some activities
towards bulky substrates,14–16 most of them had no detectable activity
towards rac-1. These results indicate the unique structural requirement
of 1 in comparison with the bulky amines previously investigated. The
only ATA with detectable activity was a double mutant of the ATA from
Vibrio fluvialis (F85L/V153A, abbreviated as VF-2M), which was previ-
ously engineered for (S)-phenylbutylamine with a specific activity of
4.99 U/mg.16a VF-2M was purified, and its specific activity towards rac-
1 was determined by the arylketone assay17 to be 0.62 mU/mg with
pyruvate (Table 1, entry 2). The wildtype VF-ATA had no detectable
activity (<0.1 mU/mg), demonstrating the importance of the F85L/
V153A mutations for the initial activity.
Table 1
VF-ATA variants and their corresponding activity towards rac-1/pyruvate, rac-
1/pentanala.
Variant
Mutation
Specific Activity [mU/mg]
Pyruvate
Pentanal
VF-wt
2M
–
< 0.1
< 0.1
F85L/V153A
0.62 ± 0.07
4.38 ± 0.12
5.43 ± 0.03
6.01 ± 0.07
8.06 ± 1.41
8.65 ± 0.23
12.45 ±
0.16
0.49 ± 0.34
0.96 ± 0.09
2.26 ± 0.11
4.42 ± 0.12
4.20 ± 0.33
6.03 ± 0.33
13.38 ±
0.23
3M
W57L/F85L/ V153A
4Mꢀ A
4Mꢀ E
5Mꢀ A
5Mꢀ E
6Mꢀ A
W57L/F85L/V153A/K163A
W57L/F85L/V153A/K163E
W57L/F85L/V153A/K163A/N286C
W57L/F85L/V153A/K163E/N286C
W57L/F85L/V153A/K163A/N286C/
R415K
6Mꢀ E
7Mꢀ E
8M-E
W57L/F85L/V153A/K163E/N286C/
R415K
15.40 ±
0.55
20.95 ±
0.36
F19H/W57L/F85L/V153A/K163E/
N286C/R415K
32.34 ±
1.82
14.86 ±
0.48
2.2. Directed evolution of the Vibrio fluvialis ATA
F19H/W57L/F85L/V153A/K163E/
Y249F/N286C/R415K
39.43 ±
3.68
n.m.b
To further improve the activity towards rac-1, we analyzed the
crystal structure of VF-ATA (PDB: 4E3Q)18 and focused on the residues
forming the substrate binding pocket (Figure 1), including 14 residues of
the small pocket (F19, L56, F85, F86, R88, R146, W147, V153, K163,
Y165, T268, E315, E316), and 11 residues of the large pocket (W57,
W150, A228, V258, I259, N286, R415, L417, C424, P426, F427). The
residue Y249 at the entry of the substrate binding site was also selected.
Using VF-2M as the starting template, single site-saturated mutagenesis
of these 26 residues was performed with the Q5 mutagenesis method
with NNK codons. For each site, 94 clones were picked and cultured in
deep 96-well plates. The activity of VF-ATA variants in the cell lysate
was screened for conversion of rac-1 with pyruvate. The characteristic
absorbance (λ = 310 nm) of the corresponding ketone 2, was measured
by a microtiter plate reader.
a
Arylketone assay conditions: The specific activities of the purified VF-ATA
variants for the conversion of rac-1 were determined by using the arylketone
assay in 96-well microtiter plates, and measured on the Infinite® 200 PRO
(TECAN) plate reader. The assay was performed with rac-1 (5 mM) as amine
donor and pyruvate (2.5 mM) or pentanal (2.5 mM) as amine acceptors in DMSO
(1.25–2.5%), HEPES buffer (50 mM, pH 6.5) at 30 ◦C. The formation of 2 was
quantified by following the increase of absorption at 310 nm over time. One unit
(U) was defined as the formation of 1 µmol 2 per minute. All measurements were
performed in triplicates.
b
n.m. = not measured.
F19H mutation), with an improved activity of 32.34 mU/mg (Table 1).
The sixth round of evolution of 7Mꢀ E led to the best variant, 8M-E
(additional Y249F mutation), with an activity of 39.43 mU/mg
(Table 1). The VF-8M-E variant thus contains eight mutations: F19H/
W57L/F85L/V153A/K163E/Y249F/N286C/R415K. In comparison with
the wildtype VF-wt (<0.1 mU/mg) and starting template VF-2M (0.62
mU/mg), the final VF-8M-E variant displays > 400-fold and > 60-fold
increased activity, respectively.
The iterative saturation mutagenesis (ISM)19 strategy was applied to
improve the activity of VF-ATA. The first round of directed evolution led
to the triple mutant VF-3M (bearing an additional W57L mutation) with
an activity of 4.38 mU/mg, representing
a 7-fold improvement
compared to the template VF-2M (Table 1). Using VF-3M as the next
template, the second round of directed evolution gave two variants,
4Mꢀ A (additional K163A mutation) and 4Mꢀ E (additional K163E
mutation), with slightly improved activities (Table 1). Further directed
evolution pursued two different routes with 4Mꢀ A and 4Mꢀ E as the
templates, respectively (Table 1). The third round of evolution led to
5Mꢀ A and 5Mꢀ E with the same additional N286C mutation (Table 1).
Continuing these two routes gave 6Mꢀ A (12.45 mU/mg) and 6Mꢀ E
(15.40 mU/mg) with the same additional R415K mutation (Table 1).
The fifth round of evolution with 6Mꢀ A and 6Mꢀ E was performed,
however, only the 6Mꢀ E direction gave a variant, 7Mꢀ E (additional
2.3. Characterization of the best three variants of VF-ATA
The best three variants, VF-6Mꢀ E, VF-7Mꢀ E, and VF-8M-E, were
selected for further characterization. Because the protein engineering
rounds were performed with rac-1, we further examined the enantio-
selectivity of the three variants for converting (S)-1. Fortunately, all the
VF-ATA variants did not show any detectable activity towards (S)-1
(<0.1 mU/mg). This proved the excellent enantioselectivity of VF-ATA
Scheme 1. a The original synthetic route to Apremilast involves resolution of rac-1 to (S)-1. b Kinetic resolution of rac-1 to (S)-1 by the evolved VF-ATA.
2

C. Xiang et al.
Bioorganic & Medicinal Chemistry 43 (2021) 116271
Figure 3. Kinetic resolution of rac-1 with the best three VF-ATA variants. Assay
conditions: 20 mM rac-1, 50 mM HEPES buffer pH 6.5, 0.1 mM PLP, 200 mM
pyruvate, 10% (vol/vol) DMSO, 0.3–0.8 mg/ml purified enzyme, 37 ◦C and
180 rpm.
Figure 1. Structure of VF-ATA (PDB: 4E3Q) with the key residues labeled. Key
residues are shown as sticks: pyridoxamine-5′-phosphate (PMP) in pink, the
catalytic lysine (K285) is highlighted in red, other sites for mutation are in blue.
2.4. Homology modeling and docking study of VF-8M-E
To elucidate the possible mechanism and the molecular rational for
the observed improved activity towards rac-1, a homology model of the
VF-8M-E was built with YASARA. As shown in Figure 4a and 4b, in
comparison to the wildtype VF-wt, the substrate binding pocket was
significantly expanded in the VF-8M-E variant, which hence can now
accommodate the bulky substrate (R)-1 much better. The molecular
docking of the reaction intermediate, PLP-(R)-1 complex, to the model
of VF-8M-E was simulated with YASARA. As depicted in Figure 4c, the
position of PLP in VF-8M-E was very similar to that of PMP in VF-wt,
which validated that PLP-(R)-1 complex was docked in a reasonable
conformation. It is clear that, the W57L and R415K mutations signifi-
cantly expanded the substrate binding pocket to accommodate the
challenging methylsulfonyl group. On the other side of the binding
pocket, F19H, F85L, and V153A contributed to host the 3-ethoxy-4-
methoxyphenyl group. The contribution of the other three mutations
is minor.
variants. The three variants were purified and tested in different reac-
tion buffer (Figure 2). The maximum specific activity was achieved at
the optimal pH of 6.0 or 6.5. Then, the best three variants were applied
for the kinetic resolution of rac-1 under the optimal conditions. As
shown in Figure 3, 20 mM of rac-1 was resolved by VF-8M-E in 24 h to
reach 51% conversion, while VF-7Mꢀ E and VF-6Mꢀ E took 48 h to
reach > 50% conversion. The optical purity of the remaining (S)-1 at 48
h was determined to be > 99% ee. During the reaction, a small per-
centage (~10%) of byproduct 1-(3-ethoxy-4-methoxyphenyl)ethenone
was observed, probably due to the α,β-elimination of (R)-1. Neverthe-
less, these results demonstrated that the highly enantioselective VF-ATA
variants could be applied to produce the enantiopure amine interme-
diate for Apremilast. The unpurified cell lysates were also proven as
catalysts for the kinetic resolution (see Figure S2 in SI). To demonstrate
the synthetic application, the kinetic resolution was performed on 20 mL
scale (100 mg) using the stable variant (VF-6Mꢀ E). The reaction took
48 h to reach a conversion of 50%. Simple workup (filtration, extraction,
and evaporation) afforded (S)-1 (>99% ee) in 41% isolated yield, and
ketone 2 in 43% isolated yield.
3. Conclusion
To provide a new enzymatic access to the chiral amine intermediate
(S)-1 for the synthesis of Apremilast, we identified a double mutant of
the ATA from Vibrio fluvialis, which had initial activity for the conver-
sion of rac-1. Subsequently, six rounds of directed evolution resulted in
the VF-8M-E variant with > 60-fold increase specific activity. A ho-
mology model of VF-8M-E was built and molecular docking study was
performed to explain the increase of activity. The purified VF-8M-E was
successfully applied in the kinetic resolution of rac-1 to produce enan-
tiopure (S)-1 in 49% conversion. The preparative kinetic resolution was
performed with the stable variant VF-6Mꢀ E to afford enantiopure (S)-1
in 41% isolated yield. The application of this ATA variant represents an
alternative method for the synthesis of the key chiral amine intermedi-
ate for the manufacturing of Apremilast.
4. Experimental section
4.1. Site-directed mutagenesis
All variants were prepared using Q5® site-directed mutagenesis kit
from New England BioLabs. The degenerated primers were designed
non-overlapping by using the standard setting of NEBaseChanger. For
Figure 2. Specific activity of the best three VF-ATA variants for conversion of
rac-1 in different buffers. Assay conditions: 5 mM rac-1, 50 mM potassium
phosphate buffer (pH 5.5–6.5) or HEPES buffer (pH 6.5–8.5), 0.1 mM PLP, 2.5
mM pyruvate, 1% (vol/vol) DMSO, 30 ◦C.
the PCR, template plasmid (0.25 ng/
μ
L, carrying the VF-2M gene),
forward and reverse primers (0.5
μM each), Q5® hot start high-fidelity
3

C. Xiang et al.
Bioorganic & Medicinal Chemistry 43 (2021) 116271
Figure 4. Structural study of VF-wt and VF-8M-E. a The substrate binding pocket of VF-wt; b the substrate binding pocket of the homology model of VF-8M-E; c
docking of PLP-(R)-1 complex to the model of VF-8M-E. The PMP and PLP-(R)-1 are in pink, the catalytic lysine (K285) is highlighted in red, and the mutated residues
are highlighted in blue.
2X master mix were used. The PCR was performed as follows: (i) 98 ◦C,
30 s; (ii) 30 cycles: 98 ◦C, 10 s; 50–72 ◦C, 30 s; 72 ◦C, 0.5 min/kbp; (iii)
72 ◦C, 2 min. The resulting PCR product was directly treated with the
kinase, ligase & DpnI (KLD enzyme mix; NEB) at room temperature for
30 min and then transformed in chemically competent E. coli TOP10
cells. After confirmation of the codon distribution in the case of site-
saturation mutagenesis, the PCR products were transformed in chemi-
cal competent E. coli BL21(DE3) cells for expression.
mg/mL). The reaction was incubated in a shaking incubator at 37 ◦C and
180 rpm. At 1, 2, 4, 6, 8, 24, and 48 h of the reaction, samples (100 L)
were taken and mixed with acetonitrile (500 L) and trifluoracetic acid
solution (0.1%, 400 L) for HPLC analysis of the conversion of 1. The
samples (125 L) were basified with NaOH (100 mM, 125 L) and
μ
μ
μ
μ
μ
extracted with ethyl acetate (0.5 mL). The organic phase was dried over
Na2SO4, filtered and the supernatant was subjected to evaporation. The
dried residues were diluted in a mixture of hexane/ethanol/isopropanol
(50:40:10) for chiral HPLC analysis of ee.
4.2. Protein expression and purification of the enzyme variants
4.5. Preparation of (S)-1 with VF-ATA variants
The plasmid constructs containing the genes of VF-ATA variants
were transformed into E. coli BL21(DE3) cells, and the resulted cells
were incubated in an LB-medium (Lysogeny Broth, 5 mL) preculture
with kanamycin (50 µg/mL) at 37 ◦C overnight. The preculture was
transferred into TB-medium (Terrific Broth, 100 mL) with kanamycin
(50 µg/mL) and incubated at 37 ◦C, 180 rpm. The expression of the Vf-
ATA variants was induced with IPTG (0.5 mM, isopropyl β-ᴅ-thio-
galactopyranoside) at an optical density of approx. 0.6 at 600 nm, and
incubated at 26 ◦C overnight. The cells were harvested by centrifugation
(20 min, 4000 g).
To a 100-mL flask, HEPES buffer (50 mM, pH 6.5, including 0.1 mM
PLP and 100 mM pyruvate), rac-1 (100 mg), and purified VF-6 Mꢀ E (20
mg) were added to form a 20 mL reaction volume system. The reaction
was maintained at pH 6.5 and 30 ◦C until reaching 50% conversion
analyzed by HPLC. Afterwards, the reaction was quenched with HCl
(100 mM) to pH 2.0 and the mixture was filtered. The filtrate was
washed with CH2Cl2 (2 × 5 mL) and then NaOH solution (1 M) was
added to adjust the pH to 12. The basified solution was extracted with
CH2Cl2 (2 × 25 mL), and the CH2Cl2 was evaporated to afford (S)-1 (41
mg, 41% yield) as yellow solid. 1H NMR (DMSO‑d6) δ: 7.02 (m, 1H),
6.89 (m, 2H), 4.28–4.23 (m, 1H), 4.03–3.98 (q, J = 6.9 Hz, 2H), 3.73 (s,
3H), 3.45–3.20 (m, 2H), 2.96 (s, 3H), 2.07 (brs, 2H), 1.35–1.30 (t, J =
6.9 Hz, 3H). To isolate 2, the solid of the initial filtration was washed
with HCl (100 mM), dissolved in acetonitrile, and filtered again, the
filtrate was evaporated to give ketone 2 (43 mg, 43% yield) as white
solid. 1H NMR (DMSO‑d6) δ: 7.73–7.72 (m, 1H), 7.50–7.49 (m, 1H),
7.13–7.09 (m, 1H), 5.04 (s, 2H), 4.13–4.05 (q, J = 7.2 Hz, 2H), 3.87 (s,
3H), 3.13 (s, 3H), 1.38–1.32 (t, J = 7.2 Hz, 3H).
For purification of the VF-ATA variants, the cell pellets were resus-
pended in HEPES buffer (50 mM, pH 7.5) containing NaCl (300 mM),
PLP (0.1 mM), imidazole (10 mM), and then lysed by ultrasonication
(50% power, 50% pulse, 2 × 5 min). The lysate was clarified by
centrifugation (1 h, 10,000 g, 4 ◦C) and purified by affinity chroma-
tography (Ni-NTA agarose) with the following buffers: washing buffer
containing HEPES (50 mM, pH 7.5) PLP (0.1 mM), NaCl (300 mM), and
imidazole (20 mM), and elution buffer containing HEPES (50 mM, pH
7.5) PLP (0.1 mM), NaCl (300 mM), and imidazole (300 mM). The VF-
ATAs were desalted in HEPES buffer (50 mM pH 7.5) with PLP (0.1
mM) using the PD-10 desalting column (GE Healthcare). The purified
VF-ATAs were stored in 30% glycerol at ꢀ 20 ◦C.
Declaration of Competing Interest
4.3. Determination of activity of VF-ATA variants
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
the work reported in this paper.
The specific activities of the purified VF-ATA variants for the con-
version of rac-1 were determined by using the arylketone assay in 96-
well microtiter plates, and measured on the Infinite® 200 PRO
(TECAN) plate reader. The assay was performed with rac-1 (5 mM) as
amine donor and pyruvate (2.5 mM) or pentanal (2.5 mM) as amine
acceptors in DMSO (1.25–2.5%), HEPES buffer (50 mM, pH 6.5) at
30 ◦C. The formation of 2 was quantified by following the increase of
absorption at 310 nm over time.
Acknowledgments
X. C. thanks the China Scholarship Council for financial support of a
PhD thesis project (File No.: 201808330394). S. W. thanks the Alex-
ander von Humboldt-Stiftung for a Humboldt Research Fellowship. The
authors thank Ina Menyes for support with HPLC analysis.
4.4. Kinetic resolution of rac-1 with VF-ATA variants
Appendix A. Supplementary material
To a 100-mL flask, the following components were added to form a
10-mL system: HEPES buffer (50 mM, pH 6.5) with PLP (0.1 mM), rac-1
(20 mM), pyruvate (200 mM), DMSO (10%), purified VF-ATA (0.3–0.8
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.bmc.2021.116271.
4

C. Xiang et al.
Bioorganic & Medicinal Chemistry 43 (2021) 116271
5 Grogan G. Curr. Opin. Chem. Biol. 2018;43:15–22.
References
6 Batista VF, Galman JL, Pinto DCGA, Silva AMS, Turner NJ. ACS Catal. 2018;8:
11889–11907.
1 (a) Breuer M, Ditrich K, Habicher T, et al. Angew Chem Int Ed. 2004;43:788–824.
https://doi.org/10.1002/anie.200300599;
7 Mangas-Sanchez J, Montgomery SL, Aleku GA, et al. Curr. Opin. Chem. Biol. 2017;37:
19–25.
(b) Nugent TC, El-Shazly M. Adv Synth Catal. 2010;352:753–819. https://doi.org/
10.1002/adsc.200900719.
8 (a) Zeymer C, Hilvert D. Ann Rev Biochem. 2018;87:131–157. https://doi.org/
10.1146/annurev-biochem-062917-012034;
2 (a) Sheldon RA, Brady D, Bode ML. Chem Sci. 2020;11:2587–2605. https://doi.org/
10.1039/C9SC05746C;
(b) Angew AFH. Chem Int Ed. 2019;58:14420–14426. https://doi.org/10.1002/
anie.201907729;
(b) Hauer B. ACS Catal. 2020;10:8418–8427. https://doi.org/10.1021/
acscatal.0c01708;
(c) Bornscheuer UT, Hauer B, Jaeger KE, Schwaneberg U. Angew Chem Int Ed. 2019;
58:36–40. https://doi.org/10.1002/anie.201812717;
(c) Hollmann F, Opperman DJ, Paul CE. Angew Chem Int Ed. 2021;60:5644–5665.
https://doi.org/10.1002/anie.202001876;
(d) Qu G, Li A, Acevedo-Rocha CG, Sun Z, Reetz MT. Angew Chem Int Ed. 2020;59:
13204–13231. https://doi.org/10.1002/anie.201901491.
9 (a) Patel RN. Bioorg Med Chem. 2018;26:1252–1274. https://doi.org/10.1016/j.
bmc.2017.05.023;
(d) Winkler CK, Schrittwieser JH, Kroutil W. ACS Cent Sci. 2021;7:55–71. https://
doi.org/10.1021/acscentsci.0c01496;
(e) Wu S, Snajdrova R, Moore JC, Baldenius K, Bornscheuer UT. Angew Chem Int Ed.
2021;60:88–119. https://doi.org/10.1002/anie.202006648;
(f) Marshall JR, Mangas-Sanchez J, Turner NJ. Tetrahedron. 2021;82. https://doi.
org/10.1016/j.tet.2021.131926.
(b) Sun H, Zhang H, Ang EL, Zhao H. Bioorg. Med. Chem. 2018;26:1275–1284.
https://doi.org/10.1016/j.bmc.2017.06.043;
(c) Li G, Wang J, Reetz MT. Bioorg. Med. Chem. 2018;26:1241–1251. https://doi.
org/10.1016/j.bmc.2017.05.021;
3 (a) Ghislieri D, Turner NJ. Top Catal. 2014;57:284–300. https://doi.org/10.1007/
s11244-013-0184-1;
(d) Adams JP, Brown MJB, Diaz-Rodriguez A, Lloyd RC, Roiban G. Adv. Synth. Catal.
2019;361:2421–2432. https://doi.org/10.1002/adsc.201900424;
¨
(b) Kohls H, Steffen-Munsberg F, Hohne M. Curr Opin Chem Biol. 2014;19:180–192.
´
´
´
(e) Albarran-Velo J, Gonzalez-Martínez D, Gotor-Fernandez V. Biocatal.
Biotransformat.. 2018;36:102–130. https://doi.org/10.1080/
10242422.2017.1340457;
https://doi.org/10.1016/j.cbpa.2014.02.021.
4 (a) Hohne M, Bornscheuer UT. ChemCatChem. 2009;1:42–51. https://doi.org/
10.1002/cctc.200900110;
(f) Fryszkowska A, Devine PN. Curr. Opin. Chem. Biol. 2020;55:151–160. https://doi.
org/10.1016/j.cbpa.2020.01.012.
(b) Koszelewski D, Tauber K, Faber K, Kroutil W. Trends Biotechnol. 2010;28:
324–332. https://doi.org/10.1016/j.tibtech.2010.03.003;
(c) Tufvesson P, Lima-Ramos L, Jensen JS, Al-Haque N, Neto W, Woodley JM.
Biotechnol Bioeng. 2011;108:1479–1493. https://doi.org/10.1002/bit.23154;
(d) Mathew S, Yun H. ACS Catal. 2012;2:993–1001. https://doi.org/10.1021/
cs300116n;
10 Man H-W, Schafer P, Wong LM, et al. J. Med. Chem. 2009;52:1522–1524.
11 (a) Schett G, Sloan VS, Stevens RM, Schafer P. Ther. Adv. Musculoskel. Dis. 2010;2:
271–278. https://doi.org/10.1177/1759720X10381432;
(b) Papp K, Reich K, Leonardi CL, et al. J. Am. Acad. Dermatol. 2015;73:37–49.
https://doi.org/10.1016/j.jaad.2015.03.049.
(e) Fuchs M, Farnberger JE, Eur KW. J Org Chem. 2015;6965–6982. https://doi.org/
10.1002/ejoc.201500852;
12 Ruchelman AL, Connolly TJ. Tetrahedron Asymmetry. 2015;26:553–559.
13 Doubsky, J.; Klvana, R.; Richter, J.; Lehnert, P. WO2016192694, 2016.
14 Pavlidis IV, Weiß MS, Genz M, et al. Nat. Chem. 2016;8:1076–1082.
15 Voss M, Das D, Genz M, et al. ACS Catal. 2018;8:11524–11533.
16 (a) Nobili A, Steffen-Munsberg F, Kohls H, et al. ChemCatChem. 2015;7:757–760.
https://doi.org/10.1002/cctc.201403010;
(f) Steffen-Munsberg F, Vickers C, Kohls H, et al. Biotechnol Adv. 2015;33:566–604.
https://doi.org/10.1016/j.biotechadv.2014.12.012;
(g) Guo F, Berglund P. Green Chem. 2017;19:333–360. https://doi.org/10.1039/
C6GC02328B;
(h) Slabu I, Galman JL, Lloyd RC, Turner NJ. ACS Catal. 2017;7:8263–8284. https://
doi.org/10.1021/acscatal.7b02686;
(b) Genz M, Vickers C, van denBergh T, et al. Int. J. Mol. Sci. 2015;16:26953–26963.
https://doi.org/10.3390/ijms161126007;
(i) Kelly SA, Pohle S, Wharry S, et al. Chem Rev. 2018;118:349–367. https://doi.org/
10.1021/acs.chemrev.7b00437;
(c) Genz M, Melse O, Schmidt S, et al. ChemCatChem. 2016;8:3199–3202. https://
doi.org/10.1002/cctc.201601007.
(j) Gomm A, O’Reilly E. Curr Opin Chem Biol. 2018;43:106–112. https://doi.org/
10.1016/j.cbpa.2017.12.007;
¨
¨
17 Schatzle S, Hohne M, Redestad E, Robins K, Bornscheuer UT. Anal. Chem. 2009;81:
8244–8248.
(k) Patil MD, Grogan G, Bommarius A, Yun H. Catalysts. 2018;8:254. https://doi.
org/10.3390/catal8070254;
18 Midelfort KS, Kumar R, Han S, et al. Protein Eng. Des. Sel. 2013;26:25–33.
19 Reetz MT, Carballeira JD. Nat. Prot. 2007;2:891–903.
(l) Rocha JF, Pina AF, Sousa SF, Cerqueira NMFSA. Catal Sci Technol. 2019;9:
4864–4876. https://doi.org/10.1039/C9CY01210A.
5