Enzyme Cascade Reactions for the Biosynthesis of Long Chain Aliphatic Amines from Renewable Fatty Acids

Article abstract of DOI:10.1002/adsc.201801501

Enzyme cascade reactions for the synthesis of long chain aliphatic amines such as (Z)-12-aminooctadec-9-enoic acid, 10- or 12-aminooctadecanoic acid, and 10-amino-12-hydroxyoctadecanoic acid from renewable fatty acids were investigated. (Z)-12-aminooctadec-9-enoic acid was produced from ricinoleic acid ((Z)-12-hydroxyoctadec-9-enoic acid) via (Z)-12-ketooctadec-9-enoic acid with a conversion of 71% by a two-step in vivo biotransformation involving a long chain secondary alcohol dehydrogenase (SADH) from Micrococcus luteus and a variant of the amine transaminase (ATA) from Vibrio fluvialis. 10-Aminooctadecanoic acid was prepared from oleic acid ((Z)-octadec-9-enoic acid) via 10-hydroxyoctadecanoic acid and 10-ketooctadecanoic acid by an in vivo three-step biocatalysis reaction involving not only the SADH and ATA variants, but also a fatty acid double bond hydratase (OhyA) from Stenotrophomonas maltophilia. 10-Aminooctadecanoic acid was produced at a total rate of 4.4 U/g dry cells with a conversion of 87% by recombinant Escherichia coli expressing the SADH and ATA variants, and OhyA simultaneously. In addition, bulky aliphatic amines could also be produced by the isolated enzymes (i. e., the SADH, the ATA variants, and a nicotinamide adenine dinucleotide (NADH) oxidase from Lactobacillus brevis) with methylbenzylamine or benzylamine as amino donor. This study thus contributes to the biosynthesis of long chain aliphatic amines having two large substituents next to the amine functionality. (Figure presented.).

Full text of DOI:10.1002/adsc.201801501

Title: Enzyme Cascade Reactions for the Biosynthesis of Long Chain
Aliphatic Amines from Renewable Fatty Acids
Authors: Da-Som Lee, Ji-Won Song, moritz Voß, Eva Schuiten, Ravi
Kumar A, Yong-Uk Kwon, Uwe Bornscheuer, and Jin-Byung
Park
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Adv. Synth. Catal. 10.1002/adsc.201801501
Link to VoR: http://dx.doi.org/10.1002/adsc.201801501

Advanced Synthesis & Catalysis
10.1002/adsc.201801501
FULL PAPER
DOI: 10.1002/adsc.201((will be filled in by the editorial staff))
Enzyme Cascade Reactions for the Biosynthesis of Long Chain
Aliphatic Amines from Renewable Fatty Acids
a
a
b
b
c
Da-Som Lee , Ji-Won Song , Moritz Voß , Eva Schuiten , Ravi Kumar Akula ,
c
b
a
Yong-Uk Kwon , Uwe Bornscheuer and Jin-Byung Park *
a
Department of Food Science & Engineering, Ewha Womans University, Seoul 03760, Republic of Korea
Fax: (+82)-2-3277-4213; phone: (+82)-2-3277-4509; e-mail: jbpark06@ewha.ac.kr
Institute of Biochemistry, Department of Biotechnology & Enzyme Catalysis, Greifswald University, 17487 Greifswald,
b
Germany
c
Department of Chemistry and Nanoscience, Ewha Womans University, Seoul 03760, Republic of Korea
Received: ((will be filled in by the editorial staff))
Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/adsc.201######.((Please
delete if not appropriate))
1
4
0-Aminooctadecanoic acid was produced at a total rate of
.4 U/g dry cells with a conversion of 87% by recombinant
Abstract. Enzyme cascade reactions for the synthesis of
long chain aliphatic amines such as (Z)-12-aminooctadec-9-
enoic acid, 10- or 12-aminooctadecanoic acid, and 10-
amino-12-hydroxyoctadecanoic acid from renewable fatty
acids were investigated. (Z)-12-aminooctadec-9-enoic acid
was produced from ricinoleic acid ((Z)-12-hydroxyoctadec-
Escherichia coli expressing the SADH and ATA variants,
and OhyA simultaneously. In addition, bulky aliphatic
amines could also be produced by the isolated enzymes
(
i.e., the SADH, the ATA variants, and a nicotinamide
adenine dinucleotide (NADH) oxidase from Lactobacillus
brevis) with methylbenzylamine or benzylamine as amino
donor. This study thus contributes to the biosynthesis of
long chain aliphatic amines having two large substituents
next to the amine functionality.
9
-enoic acid) via (Z)-12-ketooctadec-9-enoic acid with a
conversion of 71% by a two-step in vivo biotransformation
involving a long chain secondary alcohol dehydrogenase
(
SADH) from Micrococcus luteus and a variant of the amine
transaminase (ATA) from Vibrio fluvialis. 10-
Aminooctadecanoic acid was prepared from oleic acid ((Z)-
octadec-9-enoic acid) via 10-hydroxyoctadecanoic acid and
1
0-ketooctadecanoic acid by an in vivo three-step biocataly-
Keywords: enzyme cascade reactions, long chain aliphatic
amines, amine transaminase, whole cell biotransformation,
Escherichia coli
sis reaction involving not only the SADH and ATA variants,
but also a fatty acid double bond hydratase (OhyA) from
Stenotrophomonas maltophilia.
Introduction
The nicotinamide cofactors were provided by cellular
carbon metabolism, while the amino donor (i.e.,
alanine) was regenerated by the alanine
dehydrogenase inside cells.
Biocatalysis has been experiencing sustainable
growth due to fast and continuous developments in
enzyme engineering, metabolic engineering,
Amines are versatile building blocks for the
pharmaceutical and agrochemical industries.
However, bulky aliphatic amines are difficult to
[
1]
synthetic biology, and systems biocatalysis. In
particular, multi-step biotransformations with self-
sufficient redox recycling and utilization of whole-
cell biocatalysis are expanding the product range, as
[5]
produce by either enzymatic or chemical means.
There are only a few reports where aromatic β- and
γ-amino acids were prepared by kinetic resolution or
asymmetric synthesis using ω-transaminases from
Polaromonas sp., Burkholderia graminis, and
[2]
well as enhancing efficiency of the biocatalysis.
For instance, chiral amines and lactones were
produced in a high yield from alcohols through
enzyme cascade reactions involving alcohol
dehydrogenases (ADHs) and amine transaminases
[6]
Sphaerobacter thermophilus.
Aromatic amines
with bulky branched chains (e.g., 2,2-dimethyl-1-
phenylpropan-1-amine) were produced by variants of
(
ATAs) with redox-recycling ADHs (e.g., lactate
[7]
dehydrogenases and amine dehydrogenases) and
ATA from Vibrio fluvialis (VF-ATA) and from
Baeyer-Villiger
monooxygenases.[
3]
Another
Ruegeria sp. TM1040 (Rsp-ATA), of which the
[8]
example is the production of 12-aminododecanoic
acid methyl ester from dodecanoic acid methyl ester
by recombinant Escherichia coli expressing an
alkane hydroxylase, ω-transaminase, and alanine
dehydrogenase.[
active site has been engineered to accept bulky
ketones. Reshaping of the substrate-binding cavity of
amine dehydrogenases has also allowed asymmetric
synthesis of bulky aliphatic amines (e.g., (R)-2-
4]
[9]
heptanamine).
1
This article is protected by copyright. All rights reserved.

Advanced Synthesis & Catalysis
10.1002/adsc.201801501
In this study, we have investigated enzyme
A
1
OH
O
cascade reactions to provide bulky aliphatic amines
from renewable fatty acids (Schemes 1 and 2). One
example is the synthesis of (Z)-12-aminooctadec-9-
enoic acid (3), which is a precursor of antibacterial
lipoamino acid derivatives (e.g., (Z)-methyl-12-(2-
aminoacetamido)octadec-9-enoate) and phenolipids
OH
5
NAD+
NADH
oxidase
Alcohol dehydrogenase
NADH
(
e.g., (Z)-methyl-12-cinnamamidooctadec-9-enoate
2
O
O
and (Z)-methyl-12-(4-hydroxybenzamido)octadec-9-
OH
enoate).[ The long chain aliphatic amine 3 was
10]
5
produced
from
ricinoleic
acid
((Z)-12-
Methylbenzylamine
hydroxyoctadec-9-enoic acid, 1) via (Z)-12-
ketooctadec-9-enoic acid 2 (Scheme 1) with a high
conversion by recombinant E. coli expressing a long-
chain secondary alcohol dehydrogenase (SADH)
Amine Transaminase
3
NH2
5
O
from Micrococcus luteus[ and VF-ATA variants.
11]
[7]
OH
The amine product 3 was also produced using
isolated enzymes (i.e., the SADH, the VF-ATA
variants, and the NADH oxidase (NOX) from
Lactobacillus brevis[ with methylbenzylamine or
benzylamine as amino donor (Scheme 1).
B
12]
1
OH
O
OH
5
Lactate
NAD+
Results and Discussion
Alcohol dehydrogenase
NADH
O
2
Lactate
dehydrogenase
O
Enzyme Screening for the Synthesis of (Z)-12-
Aminooctadec-9-enoic Acid from Ricinoleic
Acid
OH
5
Alanine
Hydroxy fatty acids (e.g., ricinoleic acid) can be
converted into amino fatty acids via keto fatty acids
by enzyme cascade reactions involving a long chain
secondary alcohol dehydrogenase (SADH) and an
amine transaminase (ATA) (Scheme 1). A SADH
from M. luteus was reported to catalyze oxidation of
hydroxy fatty acids into the corresponding keto fatty
acids[ and was therefore included in the cascade.
ATAs are versatile enzymes that synthesize a
variety of amines, but most do not accept ketones
with two large substituents next to the carbonyl
Amine Transaminase
Pyruvate
3
NH2
5
O
OH
Scheme 1. Designed biotransformation pathway 1.
Ricinoleic acid (1) is converted into (Z)-12-aminooctadec-
-enoic acid (3) via (Z)-12-ketooctadec-9-enoic acid (2).
The NADH oxidase (NOX) was included for regeneration
of NAD , which is required for the SADH reaction (A).
11]
9
+
[13]
functionality. ATAs which have been reported to
have catalytic activity with sterically hindered
ketones, were thus considered as enzyme candidates.
These candidates included the variants of V. fluvialis
ATA (VF-ATA)[ and the variants of Ruegeria sp.
The lactate dehydrogenase was included for removal of
+
pyruvate and regeneration of NAD (B).
7]
Biotransformation of Ricinoleic Acid into (Z)-
2-Aminooctadec-9-enoic Acid
TM1040 ATA (Rsp-ATA).[8]
A
variant of
1
Arthrobacter sp. ATA, which had been developed to
[14]
accept the prositagliptin ketone, was also chosen
for initial studies (Table S1 and S2).
The biocatalytic activity of the VF-ATA variants was
quantitatively estimated by conducting the
biotransformation of ricinoleic acid 1 into (Z)-12-
aminooctadec-9-enoic acid 3 (Scheme 1). The
biotransformation was initiated by adding ricinoleic
The biocatalytic activity of the ATAs was evaluated
by measuring acetophenone formation with thin layer
chromatography (TLC).[ After reaction in 50 mM
HEPES buffer (pH 8.0) containing 1 mM 10-
15]
+
acid and NAD into the reaction medium containing
ketooctadecanoic
acid,
10
mM
racemic
SADH, ATAs, and NADH oxidase (NOX) from L.
methylbenzylamine (MBA), 0.1 mM pyridoxal
phosphate (PLP), and 10% dimethyl sulfoxide
[12]
brevis
(Fig. 1). The NOX was included for
+
regeneration of NAD , which is required for the
SADH reaction (Scheme 1A). The amino donor (i.e.,
rac-MBA) was added into the reaction medium at t =
(
DMSO), the resulting ketone product acetophenone
was analyzed by TLC. The acetophenone spot was
clearly visible from the VF-ATA and Rsp-ATA
mutants as well as the positive control (Fig. S1).
0
(
, just after ricinoleic acid (1) was transformed into
Z)-12-ketooctadec-9-enoic acid (2).
2
This article is protected by copyright. All rights reserved.

Advanced Synthesis & Catalysis
10.1002/adsc.201801501
Notably, the H3_RA and H1_R variants of VF-ATA,
which had been designed to accept the bulky
aliphatic ketone,[ converted (Z)-12-ketooctadec-9-
enoic acid (2) into the amine product (3) at a rate of
over 200 U/L with over 100 U/g proteins at t < 0.25
h (Fig. 1AB and Table 1). The conversion, which
was determined based on the product concentration
measured by GC/MS at t = 1 h, reached
approximately 90% (Fig. S2A). In contrast, another
VF-ATA variant, H3_RAV, showed only little
bioconversion (Fig. 1C).
Figure 1. Time course of the biotransformation of
ricinoleic acid (1) into (Z)-12-aminooctadec-9-enoic acid
(3). The biotransformation was initiated by adding
ricinoleic acid and NAD+ into the reaction medium
containing the cascade enzymes (i.e., secondary alcohol
dehydrogenase (SADH), amine transaminase (ATA), and
NADH oxidase (NOX)). The amino donor (i.e.,
methylbenzylamine (MBA)) was added into the reaction
medium at t = 0, just after ricinoleic acid (1) was
transformed into (Z)-12-ketooctadec-9-enoic acid (2). The
V. fluvialis ATA (VF-ATA) variants (i.e., H1_R (triple
mutant of VF-ATA (L56V, W57F, F85V)) (A), H3_RA
(triple mutant of VF-ATA (L56V, W57C, V153A)) (B),
and H3_RAV (quadruple mutant of VF-ATA (L56V,
W57C, F85V, V153A)) (C) were used for amination of
(Z)-12-ketooctadec-9-enoic acid (2). Symbols indicate the
concentrations of ricinoleic acid (1) (●), (Z)-12-
ketooctadec-9-enoic acid (2) (▽), (Z)-12-aminooctadec-9-
enoic acid (3) (■).
7]
The isolated yield of the amine product (3) from the
reaction medium was over 70%; (Z)-12-
aminooctadec-9-enoic acid (3) (54 mg , purity >
9
0%) was obtained as a liquid form from 50 mL
reaction medium containing 70 mg amine (Fig. S2B).
The structure of 3 was confirmed by GC/MS and
NMR analysis (Fig. S2B and S3). This indicates that
the VF-ATA variants H3_RA and H1_R are able to
catalyze transamination of the bulky aliphatic ketone
(
2) into the bulky aliphatic amine (3). Chirality of the
amine product remained to be investigated.
Biosynthesis of (Z)-12-Aminooctadec-9-enoic
Acid with Benzylamine and Alanine as Amino
Donor
A
Benzylamine is one of the cheapest amino donors for
[16]
ATAs. Thus, the ricinoleic acid biotransformation
into (Z)-12-aminooctadec-9-enoic acid (3) was
conducted with benzylamine (Fig. 2A). All reaction
conditions were equal to those for the
biotransformations with MBA (Fig. 1B), except that
octane was added into the reaction medium up to
2
0% (v/v) to in situ remove possible ATA inhibitors
[1c, 16]
(
e.g., benzaldehyde).
(Z)-12-Ketooctadec-9-
enoic acid (2) was transformed into the amine
product (3) by H3_RA at a rate of 83 U/L and 42 U/g
proteins at t < 0.5 h (Fig. 2A and Table 1). The
conversion reached approximately 80%. Thereby, it
was confirmed that benzylamine can be used as
amino donor for the transamination of bulky
aliphatic ketones.
B
Alanine has been widely used as an amino donor
and therefore, the biosynthesis of (Z)-12-
aminooctadec-9-enoic acid (3) was investigated with
alanine as amino donor (Fig. 2B). The
biotransformation of ricinoleic acid (1) into (Z)-12-
aminooctadec-9-enoic acid (3) was initiated by
adding 6 mM ricinoleic acid and 50 mM alanine into
the reaction medium containing the enzymes, SADH,
H3_RA, and lactate dehydrogenase (LDH) from
C
+
porcine heart, and 1 mM NAD (Scheme 1B). The
+
LDH was included for regeneration of NAD , which
is required for the SADH reaction and for removal of
[1c]
pyruvate, which is one of the ATA inhibitors.
Ricinoleic acid was converted into the amine product
3) via (Z)-12-ketooctadec-9-enoic acid (2) at a rate
(
of 79 U/L at t < 0.5 h (Table 1). The bioconversion
yield reached approximately 65%. This result
indicates that alanine can also be used as amino
donor for transamination of the long chain aliphatic
ketones.
3
This article is protected by copyright. All rights reserved.

Advanced Synthesis & Catalysis
10.1002/adsc.201801501
A
Whole-cell Biotransformation of Ricinoleic
Acid into (Z)-12-Aminooctadec-9-enoic Acid
Next, the cascade described above using isolated
enzymes (Scheme 1A) was introduced into E. coli by
expressing the SADH and one of the different VF-
ATA variants simultaneously. NOX was not
+
overexpressed because the NAD cofactor could be
provided through cellular carbon metabolism. The
fatty acid transporter FadL was overexpressed in the
outer membrane of the E. coli to facilitate substrate
[17]
transport into the cell.
The two-step
biotransformation was initiated by adding ricinoleic
acid as substrate and (S)-MBA as amino donor into
the reaction medium containing 7.4 g dry cells/L of
the recombinant E. coli.
B
The target product (3) was produced at a rate of 24
U/L at t < 1 h by recombinant E. coli expressing the
SADH and H3_RA (Fig. 3A). The specific
biotransformation rate was determined to be 3.5 U/g
dry cells. This corresponds to approximately 42 U/g
ATAs, as estimated from the SDS-PAGE where the
H3_RA band in the soluble fraction accounted for
approximately 15% of the total amount of proteins
(
Fig. S4). As compared to the biotransformation by
isolated enzymes (Fig. 1B and Table 1), the specific
enzyme activity in the recombinant E. coli was ca.
4
0% of the isolated ATA. Thereby, it could be
concluded that the enzyme cascade (Scheme 1) was
efficiently working even in a whole-cell system.
The biotransformation activity of the H1_R
expressed in the recombinant E. coli was comparable
to that of the H3_RA (Fig. 3B). On the other hand,
H3_RAV expressed in the recombinant E. coli was
markedly lower than that of H3_RA and H1_R (Fig.
3C), as shown in the biotransformation with isolated
enzymes (Table 1).
Figure 2. Time course of the biotransformation of
ricinoleic acid (1) into (Z)-12-aminooctadec-9-enoic acid
(
3). The biotransformation conditions were identical to the
experiments shown in Fig. 1B, except for the amino donor.
Benzylamine (A) and alanine (B) was used instead of
MBA. In case of Fig. 2A, octane was added into the
reaction medium to 20% (v/v) to in situ remove one of the
ATA inhibitors (e.g., benzaldehyde). Symbols indicate the
concentrations of ricinoleic acid (1) (●), (Z)-12-keto-
octadec-9-enoic acid (2) (▽), (Z)-12-aminooctadec-9-enoic
acid (3) (■).
Table 1. Biotransformation activity of the V. fluvialis amine transferase variants
Initial
Initial
Conversion
Enzymes involved
Amino donor
Bioconversion rate
bioconversion rate
(U/g proteins)[a]
(
%)[b]
(
U/L)[a]
Biotrans 1
Biotrans 2
Biotrans 3
Biotrans 4
SADH/H1_R/ NOX
SADH/H3_RA/NOX
SADH/H3_RAV/NOX
SADH/H3_RA/NOX
SADH/H3_RA/LDH
MBA
MBA
210
108
115
10
88
90
17
79
65
230
20
MBA
Benzylamine
Alanine
83
42
Biotrans 5
79
40
[
[
a]
The initial biotransformation rates were determined based on the product concentrations, which were measured at t =
.25 or 0.5 h by gas chromatography/mass spectrometry (GC/MS)
The conversion was calculated based on the substrate and product concentrations determined by GC/MS.
0
b]
4
This article is protected by copyright. All rights reserved.

Advanced Synthesis & Catalysis
10.1002/adsc.201801501
A
our previous studies,[11] into corresponding amino
fatty acids was investigated as well. The initial
biotransformation rate of 12-hydroxyoctadecanoic
acid into 12-aminooctadecanoic acid by the
recombinant E. coli expressing H3_RA was 4.0 U/g
dry cells. This value was similar to that of ricinoleic
acid ((Z)-12-hydroxyoctadec-9-enoic acid), while the
reaction rate of 10-hydroxyoctadecanoic acid
increased up to 9.1 U/g dry cells (Table 2). On the
other hand, the biotransformation rate of 10,12-
dihydroxyoctadecanoic acid into 10-amino-12-
hydroxyoctadecanoic acid was significantly lower
than that of ricinoleic acid. These results indicated
that structure of the substrates and especially the
position of the carbonyl group in the carbon skeleton
of fatty acids could be one of the key factors to
determine bioconversion rates. The structure of 12-
B
aminooctadecanoic
hydroxyoctadecanoic acid was confirmed by GC/MS
Fig. S5 and S6).
acid
and
10-amino-12-
(
Biosynthesis of 10-Aminooctadecanoic Acid
from Oleic Acid
Finally, the recombinant biotransformation of oleic
acid (4) into 10-aminooctadecanoic acid (7) was
examined (Scheme 2). The recombinant E. coli was
engineered to express not only the SADH, H3_RA
and FadL, but also a fatty acid double bond hydratase
C
[18]
(
i.e., OhyA) from S. maltophilia.
Oleic acid will be transported into the cytoplasm via
FadL in the outer membrane and subjected to
hydration by OhyA into 10-hydroxyoctadecanoic
acid (5).
OH
7
O
H2O
Hydratase
4
OH
Figure 3. Time course of the biotransformation of
ricinoleic acid (1) into (Z)-12-aminooctadec-9-enoic acid
OH
7
O
(
3) by the recombinant E. coli BL21(DE3) expressing a
NAD+
5
secondary alcohol dehydrogenase (SADH) of M. luteus,
the VF-ATA variants (i.e., H3_RA (A), H1_R (B), and
H3_RAV (C)), and a long chain fatty acid transporter
Alcohol dehydrogenase
NADH
O
(
6
FadL). The biotransformation was initiated by adding 5 to
mM ricinoleic acid and 10 mM MBA into the
OH
7
O
recombinant E. coli culture broth (cell density: 7.4 g dry
cells/L). Symbols indicate the concentrations of ricinoleic
acid (1) (●), (Z)-12-ketooctadec-9-enoic acid (2) (▽), (Z)-
6
Methylbenzylamine
Amine transaminase
1
2-aminooctadec-9-enoic acid (3) (■).
NH2
OH
7
Biotransformation of Other Hydroxy Fatty
Acids
O
7
Whole-cell biotransformation of 10-hydroxy-
octadecanoic acid, 12-hydroxyoctadecanoic acid, and
Scheme 2. Designed biotransformation pathway 2. Oleic
acid (4) is converted into 10-aminooctadecanoic acid (7)
via 10-hydroxyoctadecanoic acid (5) and 10-keto-
octadecanoic acid (6).
1
0,12-dihydroxyoctadecanoic acid, which had been
produced from renewable fatty acids according to
5
This article is protected by copyright. All rights reserved.

Advanced Synthesis & Catalysis
10.1002/adsc.201801501
When the oleic acid and MBA were added at the
same time into culture broth of the recombinant E.
coli, oleic acid was not converted into 10-
hydroxyoctadecanoic acid. One of the reasons might
include inhibition of OhyA by the amine donor.
Therefore, MBA was added after production of 10-
ketooctadecanoic acid (6) from oleic acid (Fig. 4).
Addition of MBA into culture broth of the recombi-
nant E. coli at t = 0 allowed to produce 4.5 mM 10-
aminooctadecanoic acid (7) in the medium. The con-
version of oleic acid into 10-aminooctadecanoic acid
reached 87%, which is comparable to that of 10-
hydroxyoctadecanoic acid (Table 2). The structure of
Comparison with Chemical Procedure
(
Z)-Methyl-12-aminooctadec-9-enoate, a precursor
of antibacterial lipoamino acid derivatives (e.g., (Z)-
methyl-12-(2-aminoacetamido)octadec-9-enoate) and
phenolipids,
methyl ricinoleate by
[10]
was chemically synthesized from
three-step chemical
a
[19]
procedure. Methyl ricinoleate which was prepared
by the esterification of ricinoleic acid was treated
with methanesulfonyl chloride to give the
corresponding mesylate, which was then treated with
sodium azide to afford methyl-12-azidooctadec-9-
enoate. The third step involved the selective
reduction of azido group using Zn-AlCl
provide the corresponding methyl-12-aminooctadec-
-enoate in 77% yield over three steps from methyl
7
was confirmed by GC/MS and NMR analysis (Fig.
3
⋅6H
2
O to
S7 and S8).
9
ricinoleate. As compared to the chemical procedures
for the conversion of hydroxyl group to amino group,
the whole-cell biocatalysis is much simpler and
environmentally benign. It can be operated in a one-
pot process without toxic chemicals.
Conclusion
This study demonstrates that long chain aliphatic
amines such as (Z)-12-aminooctadec-9-enoic acid,
1
0- or 12-aminooctadecanoic acid, and 10-amino-12-
hydroxyoctadecanoic acid can be prepared from
renewable fatty acids by multi-step
Figure 4. Time course of the biotransformation of oleic
acid (4) into 10-aminooctadecanoic acid (7) by the
recombinant E. coli BL21(DE3) expressing a fatty acid
double bond hydratase (OhyA) of S. maltophilia, the
SADH of M. luteus, the VF-ATA variant (i.e., H3_RA),
and FadL. The biotransformation was initiated by adding
biotransformations involving a fatty acid double
bond hydratase (OhyA) from S. maltophilia, a long
chain SADH from M. luteus, and a variant of V.
fluvialis amine transaminase. Not only screening of
an active enzyme but also tight control of putative
enzyme inhibitors (e.g., amino donor, byproducts)
and cofactor regeneration allows for the production
of bulky aliphatic amines at a rather high rate and
yield. This study will be continued by determining
the chirality of the amine products and by
engineering of the enzymes to improve the
5
.3 mM oleic acid into the recombinant E. coli culture
broth (cell density: 7.4 g dry cells/L) at t = -3 h. MBA was
added at t = 0. Symbols indicate the concentrations of
oleic acid (4) (●), 10-hydroxyoctadecanoic acid (5) (△),
1
0-ketooctadecanoic acid (6) (▽), and 10-amino-
octadecanoic acid (7) (■).
transamination
activity.
Table 2. Substrate specificity of the V. fluvialis amine transferase variants
Initial bioconversion
Initial bioconversion
rate
Conversion
rate
Substrates
(
%)[b]
(
U/L)[a]
(U/g dry cells)[a]
1
1
1
1
2-Ketooctadec-9-enoic acid
2-Ketooctadecanoic acid
24
3.5 (42)
71
87
94
60
28
63
4.0 (48)
9.1 (110)
0.9 (10)
0-Ketooctadecanoic acid
0-Keto-12-hydroxyoctadecanoic acid
6.0
[
a]
The initial biotransformation rates were determined based on the product concentrations, which were measured at t =
0
.25 or 0.5 h by GC/MS. The values in the parenthesis indicate initial bioconversion rates per g ATAs (U/g ATAs),
which were determined from the SDS-PAGE analysis (Fig. S4).
The conversion was calculated based on the substrate and product concentrations determined by GC/MS.
[
b]
6
This article is protected by copyright. All rights reserved.

Advanced Synthesis & Catalysis
10.1002/adsc.201801501
Experimental Section
samples were analyzed by the TLC with different staining
methods and mobile phases. The mobile phase was hexane
and ethyl acetate (70:30). Staining of ketones was
conducted using 2,4-dinitrophenylhydrazine.
Microbial Strains and Cultivation Media
Recombinant E. coli BL21(DE3) was cultivated in the
Riesenberg medium, which was supplemented with 10 g/L
glucose and the appropriate antibiotics for plasmid
maintenance. The Riesenberg medium consisted of 4 g/L
Enzymatic Transformation of Ricinoleic Acid
into (Z)-12-Aminooctadec-9-enoic Acid (3)
Enzymatic Transformation of ricinoleic acid into (Z)-12-
aminooctadec-9-enoic acid was performed in 50 mM
HEPES buffer (pH 8.0) containing 5 mM ricinoleic acid, 2
(
NH
MgSO
.25 g/L ZnSO
g/LNa , 2.0 g/L CaCl
The recombinant gene expression was induced by adding
4
)
2
HPO
, and 10 mL/L trace metal solution [10 g/L FeSO
, 1.0 g/L CuSO , 0.5 g/L MnSO , 0.23
7
, and 0.1 g/L (NH Mo O24].
4 2 4
, 13.5 g/L KH PO , 1.7 g/L citric acid, 1.4 g/L
4
4
,
2
4
4
4
+
mM PLP, 2.5 mM NAD , and 10 mM rac-MBA or 50
2
B
4
O
7
2
4
)
6
mM alanine. After the cascade enzymes (i.e., SADH (4
mg/mL), VF-ATA variants (2 mg/mL), and NOX (1
mg/mL) or LDH (45 U/mL) were added into the reaction
media, they were incubated at 30ꢀ°C for 3 h with agitation
at 200 rpm (working volume: 2 mL). Biotransformation of
ricinoleic acid into (Z)-12-aminooctadec-9-enoic acid was
also carried out with benzylamine as an amine donor. The
reaction conditions were similar to the experiment with
rac-MBA or alanine as amino donor. The difference was
that 50mM Tris-HCl buffer (pH 7.0) and octane (ratio =
0
.1 mM isopropyl β-D-thiogalactopyranoside (IPTG) an
OD600 of 0.6-0.8 into the cultivation medium.
Chemicals and materials
Ricinoleic acid, N-methyl-N-(trimethylsilyl)trifluoro-
acetamide (TMS), (S)-methy-lbenzylamine (MBA), and
rac-MBA were obtained from TCI (Tokyo, Japan). Oleic
acid was purchased from Daejung (Siheung, Korea). 10-
Hydroxyoctadecanoic acid, 12-hydroxyoctadecanoic acid,
and 10,12-dihidroxyoctadecanoic acid were purified in our
5
:1) was used as the reaction medium.
Whole-Cell Biotransformations
[
11]
lab as described earlier. Antibiotics, trace elements for
culture medium, IPTG, Tween80, L-alanine, benzylamine,
The recombinant E. coli cells, expressing the SADH, VF-
ATA variants, and FadL, were used for the
biotransformation of hydroxy fatty acids, whereas the
recombinant E. coli cells, expressing not only the SADH,
VF-ATA variants, and FadL but also a fatty acid double
bond hydratase (i.e., OhyA) were used for the
biotransformation of oleic acid, based on our previous
+
nicotinamide adenine dinucleotide (NAD ), pyridoxal 5′-
phosphate (PLP), and 4-(2-hydroxyethyl) piperazine-1-
ethanesulfonic acid (HEPES), N-(2-hydroxyethyl)
piperazine-N′-(2-ethanesulfonic acid) were purchased
from Sigma (St. Louis, MO, USA). Lactate dehydrogenase
from porcine heart was obtained from MP bio (Santa Ana,
California, USA).
[
17]
studies. The recombinant E. coli cells were cultivated in
Riesenberg medium at 30 °C and expressions of target
genes were induced with 0.1 mM IPTG at 20 °C based on
Protein Purification
[
20]
our earlier works. When added glucose was completely
consumed, the recombinant cells were harvested by
centrifugation at 3,500 rpm (2,602 ×g) for 15 min at
After the recombinant E. coli BL21(DE3) expressing the
cascade enzymes were cultivated in the Riesenberg
medium, the cells were harvested through centrifugation at
4
ꢀ°C. The biotransformation was initiated by adding 5 mM
of a fatty acid, 1 mM PLP, and 10 mM (S)-MBA into 50
mM HEPES buffer (pH 8.0) containing 0.5 g/L Tween80
and 7.4 g/dry cells. The reaction mixture were incubated at
3
,500 rpm (2,602 ×g) for 15ꢀmin and washed with 50 mM
HEPES buffer. The washed cells were resuspended into
lysis buffer containing 10 mM imidazole and 0.1 mM
3
5ꢀ°C and 200 rpm.
+
cofactor PLP or NAD and subjected to the cell lysis by
sonication. The cell debris was removed by centrifugation
at 13,000 rpm (15,814 ×g) for 7 min at 4ꢀ°C. The
supernatant was mixed with 1.0 mL per g dry cells of
Ni -NTA agarose beads solution (Quiagen, Hilden,
Germany) and incubated at 4 °C, 250 rpm for 1 h. The
mixture was loaded onto the column and flowed out. The
column was washed twice with a wash buffer containing
Product Analysis by Gas Chromato
graphy/Mass Spectrometry (GC/MS)
2+
Concentrations of ricinoleic acid, (Z)-12-aminooctadec-9-
enoic acid and other intermediates were measured
[
20]
according to the method based on our previous works.
The reaction medium was mixed with 2.8 times volume of
ethyl acetate containing 0.5 g/L palmitic acid as an
internal standard. The organic phase was harvested after
vigorous vortexing and subjected to derivatization with N-
methyl-N-(trimethylsilyl)trifluoroacetamide (TMS). The
trimethylsilyl derivatives were analyzed using a Thermo
Ultra Trace GC system connected to San ion trap mass
detector (Thermo ITQ1100GC-ion Trap MS, Thermo
Scientific, Indianapolis, IN, USA) and were separated on a
non-polar capillary column (30 m length, 0.25 mm film
thickness, HP-5MS, Agilent technologies, Palo Alto, CA,
USA). A linear temperature gradient was programmed as
160 °C, 25 °C /min to 235 °C, 3 °C /min to 253 °C.
Injection port temperature was 230 °C. Mass spectra were
obtained by electron impact ionization at 70 eV. Scan
spectra were obtained within the range of 100-600 m/z.
Selected ion monitoring was used for detection and the
fragmentation analysis of the reaction products.
2
0 mM imidazole. The target proteins were then eluted by
increasing the imidazole concentration to 250 mM.
Fractions containing the recombinant proteins were pooled
and dialyzed to remove imidazole.
Acetophenone Assay
The biocatalytic activity of the ATAs was evaluated by
measuring acetophenone formation with thin layer
[
15]
chromatography (TLC).
In brief, the enzyme reaction
was carried out in 1 mL scale with the following
conditions in the kinetic resolution mode: 1 mM 10-
ketooctadecanoic acid, 10 mM rac-MBA, 0.1 mM PLP
and 10% DMSO in 50 mM HEPES buffer pH 8.0. The
reaction was incubated over night at 30 °C, 600 rpm. For
product analysis, the reactions were terminated by
basification with NaOH and extracting the reaction
products with ethyl acetate (3x with 200 μL each). The
7
This article is protected by copyright. All rights reserved.

Advanced Synthesis & Catalysis
10.1002/adsc.201801501
Isolation of (Z)-12-Aminooctadec-9-enoic
[
[
[
5] a) A. Gomm, E. O'Reilly, Curr. Opin. Chem. Biol.
Acid (3) after Biotransformation
2
018, 43, 106-112; b) F. Guo, P. Berglund, Green
Chem. 2017, 19, 333-360; b) I. Slabu, J. L. Galman, R.
C. Lloyd, N. J. Turner, ACS Catal. 2017, 7, 8263-8284.
(
Z)-12-Aminooctadec-9-enoic acid was isolated from the
biotransformation medium by simple extraction with ethyl
acetate. The combined organic layer was washed with
6] a) S. Mathew, S. P. Nadarajan, T. Chung, H. H. Park,
H. Yun, Enz. Microb. Technol. 2016, 87–88, 52–60; b)
M. Shon, R. Shanmugavel, G. Shin, S. Mathew, S. H.
Lee, H. Yun, Chem. Commun. 2014, 50, 12680-12683.
2 4
saline, dried over anhydrous Na SO , filtered, and
evaporated in vacuo. The crude compounds were
separated and purified by silica gel column
chromatography. The varying concentrations of ethyl
acetate/hexane were used to remove the fatty acid
impurities. The amine products were finally recovered
with 10-20% methanol in dichloromethane as the eluent.
The isolation/purification yield of the amine product (3)
was over 70%, based on the concentration of the amine
product after whole-cell biotransformation. The purified
7] M. Genz, O. Melse, S. Schmidt, C. Vickers, M. Dörr,
T. van den Bergh, H. J. Joosten, U. T. Bornscheuer,
ChemCatChem 2016, 8, 3199-3202.
1
amine product (54 mg) was a liquid form; H NMR (300
[8] a) M. S. Weiß, I. V. Pavlidis, P. Spurr, S. P. Hanlon, B.
MHz, CD
3
OD)  5.69–5.60 (m, 1H), 5.41–5.32 (1H, m),
Wirz, H. Iding, U. T. Bornscheuer, Org. Biomol. Chem.
3
.17 (quin, J = 6.7 Hz, 1H), 2.42–2.31 (m, 2H), 2.19 (t, J
2
016, 14, 10249-10254; b) M. S. Weiß, I. V. Pavlidis,
=
7.5 Hz, 2H), 2.12-2.05 (m, 2H), 1.68–1.52 (m, 4H),
13
P. Spurr, S. P. Hanlon, B. Wirz, H. Iding, U. T. Born-
scheuer, ChemBioChem 2017, 18, 1022-1026; c) I. V.
Pavlidis, M. S. Weiß, M. Genz, P. Spurr, S. P. Hanlon,
B. Wirz, H. Iding, U. T. Bornscheuer, Nat. Chem. 2016,
1
.41–1.29 (m, 16H), 0.91 (t, J = 6.0 Hz, 3H); C NMR
OD)  181.2 (COOH), 135.8 (CH=CH),
23.8 (CH=CH), 71.5 (CNH ), 53.1, 37.6, 33.4, 32.7, 31.4,
0.26, 30.22, 30.21, 30.0, 28.2, 27.0, 26.2, 23.6, 14.4
).
(
75 MHz, CD
3
1
3
2
(
CH
3
8
, 1076-1082.
[
[
9] F. F. Chen, G. W. Zheng, L. Liu, H. Li, Q. Chen, F. L.
Acknowledgements
Li, C. X. Li, J. H. Xu, ACS Catal. 2018, 8, 2622−2628.
10] a) Y. Mohini, R. B. N. Prasad, M. S. L. Karuna, Y.
Poornachandra, C. Ganesh Kumar, Bioorg. Med.
Chem. Lett. 2016, 26, 5198–5202; b) M. Yarra, S. S.
Kaki, R. B. N. Prasad, K. S. L. Mallampall, P. Yedla,
G. K. Chityala, Eur. J. Lipid Sci. Technol. 2016, 118,
This study was supported by the National Research Foundation
of Korea (NRF) grant funded by the Korea government (MEST)
(No. 2017008670) and the Marine Biomaterials Research Center
grant from the Marine Biotechnology Program funded by the
Ministry of Oceans and Fisheries, Republic of Korea (No.
D11013214H480000100).
6
22–630.
[
11] a) H. J. Cha, E. J. Seo, J. W. Song, H. J. Jo, A. R.
References
Kumar, J. B. Park, Adv. Synth. Catal. 2018, 360, 696–
7
03; b) H. J. Oh, S. U. Kim, J. W. Song, J. H. Lee, W.
[
1] a) Y. J. Yeon, J. B. Park, in : Lipid Modification by
Enzymes and Engineered Microbes (Eds.: U. T.
Bornscheuer), Elsevier, Amsterdam, 2018, pp 139-
R. Kang, Y. S. Jo, K. R. Kim, U. T. Bornscheuer, D.
K. Oh, J. B. Park, Adv. Synth. Catal. 2015, 357, 408–
416; c) J. W. Song, E. Y. Jeon, D. H. Song, H. Y.
Jang, U. T. Bornscheuer, D. K. Oh, J. B. Park, Ang.
Chem. Int. Ed. 2013, 52, 2534-2537.
1
55; b) X. Chen, C. Gao, L. Guo, G. Hu, Q. Luo, J.
Liu, J. Nielsen, J. Chen, L. Liu, Chem. Rev. 2018, 118,
4
3
–72; c) F. Guo, P. Berglund, Green Chem. 2017, 19,
33-360; d) Y. Liang, J. Wei, X. Qiu, N. Jiao, Chem.
[
12] G. Rehn, A. T. Pedersen, J. M. Woodley, J. Mol.
Catal. B: Enz. 2016, 134, 331–339.
Rev. 2018, 118, 4912−4945; e) I. Slabu, J. L. Galman,
R. C. Lloyd, N. J. Turner, ACS Catal. 2017, 7, 8263-
8
[13] a) S. Calvelage, M. Dçrr, M. Hçhne, U. T. Born-
scheuer, Adv. Synth. Catal. 2017, 359, 4235-4243; b)
F. Steffen-Munsberg, C. Vickers, H. Kohls, H. Land,
H. Mallin, A. Nobili, L. Skalden, T. van den Bergh, H.
J. Joosten, P. Berglund, M. Hchne, U. T. Bornscheuer,
Biotechnol. Adv. 2015, 33, 566-604.
284; f) R. A. Sheldon, J. M. Woodley, Chem. Rev.
018, 118, 801−838.
2
[
2] a) J. Muschiol, C. Peters, N. Oberleitner, M. D. Miho-
vilovic, U. T. Bornscheuer, F. Rudroff, Chem. Com-
mun. 2015, 51, 5798; b) J. H. Schrittwieser, S. Veli-
kogne, M. Hall, W. Kroutil, Chem. Rev. 2018, 118,
[14] C. K. Savile, J. M. Janey, E. C. Mundorff, J. C.
Moore, S. Tam, W. R. Jarvis, J. C. Colbeck, A. Kreb-
ber, F. J. Fleitz, J. Brands, P. N. Devine, G. W.
Huisman, G. J. Hughes, Science 2010, 329, 305-309.
2
2
2
70-348; c) J. M. Sperl, V. Sieber, ACS Catal. 2018, 8,
385−2396; d) S. Wu, Z. Li, ChemCatChem 2018, 10,
164–2178.
[
[
3] a) J. H. Sattler, M. Fuchs, K. Tauber, F. G. Mutti, K.
Faber, J. Pfeffer, T. Haas, W. Kroutil, Angew. Chem.
Int. Ed. 2012, 51, 9156–9159; b) S. Schmidt, C. Scher-
kus, J. Muschiol, U. Menyes, T. Winkler, W. Hummel,
H. Groger, A. Liese, H. G. Herz, U. T. Bornscheuer,
Angew. Chem. Int. Ed. 2015, 54, 2784–2787.
[15] S. Schätzle, M. Höhne, E. Redestad, K. Robins, U. T.
Bornscheuer, Anal. Chem. 2009, 81, 8244-8248.
[
16] J. S. Shin, B. G. Kim, Biotechnol. Lett. 2009, 31,
595−1599; b) S. Sung, H. Jeon, S. Sarak, M.
1
Murshidul Ahsan, M. D. Patil, W. Kroutil, B. G. Kim,
H. Yun, Green Chem. 2018, 20, 4591–4595.
4] N. Ladkau, M. Assmann, M. Schrewe, M. K. Julsing,
A. Schmid, B. Bühler, Metab. Eng. 2016, 36, 1–9.
8
This article is protected by copyright. All rights reserved.

Advanced Synthesis & Catalysis
10.1002/adsc.201801501
[
[
17] a) E. Y. Jeon, J. H. Seo, W. R. Kang, M. J. Kim, J. H.
Lee, D. K. Oh, J. B. Park, ACS Catal. 2016, 6, 7547-
7
553; b) E. Y. Jeon, J. W. Song, H. J. Cha, S. M. Lee,
J. Lee, J. B. Park, J. Biotech. 2018, 281, 161-167.
18] a) E. Y. Jeon, J. H. Lee, K. M. Yang, Y. C. Joo, D. K.
Oh, J. B. Park, Process Biochem. 2012, 47, 941-947;
b) Y. C. Joo, E. S. Seo, Y. S. Kim, K. R. Kim, J. B.
Park, D. K. Oh, J. Biotechnol. 2012, 158, 17-23; c) E.
J. Seo, Y. J. Yeon, J. H. Seo, J. H. Lee, J. P. Boñgol,
Y. Oh, J. M. Park, S. M. Lim, C. G. Lee, J. B. Park,
Biores. Technol. 2018, 251, 288–294.
[
[
19] Y. Mohini, R. B. N. Prasad, M. S. L. Karuna, Y.
Poornachandra, C. Ganesh Kumar, Bioorg. Med.
Chem. Lett. 2014, 24, 5224–5227.
20] a) S. Koppireddi, J. H. Seo, E. Y. Jeon, P. S. Chow-
dhury, H. Y. Jang, J. B. Park, Y. U. Kwon, Adv. Synth.
Catal. 2016, 358, 3084-3092; b) E. J. Seo, Y. J. Yeon,
J. H. Seo, J. H. Lee, J. P. Boñgol, Y. Oh, J. M. Park,
S. M. Lim, C. G. Lee, J. B. Park, Biores. Technol.
2
018, 251, 288-294; c) J. M. Woo, E. Y. Jeon, E. J.
Seo, J. H. Seo, D. Y. Lee, Y. J. Yeon, J. B. Park, Sci.
Rep. 2018, 8, 10280.
9
This article is protected by copyright. All rights reserved.

Advanced Synthesis & Catalysis
10.1002/adsc.201801501
FULL PAPER
Enzyme cascade reactions for the biosynthesis of
long chain aliphatic amines from renewable fatty
acids
Adv. Synth. Catal. Year, Volume, Page – Page
Da-Som Lee, Ji-Won Song, Moritz Voß, Eva
Schuiten, Ravi Kumar Akula, Yong-Uk Kwon,
Uwe Bornscheuer and Jin-Byung Park*
1
0
This article is protected by copyright. All rights reserved.