A Bi-Enzymatic Cascade Pathway towards Optically Pure FAHFAs**

Article abstract of DOI:10.1002/cbic.202100070

Recently discovered endogenous mammalian lipids, fatty acid esters of hydroxy fatty acids (FAHFAs), have been proved to have anti-inflammatory and anti-diabetic effects. Due to their extremely low abundancies in vivo, forging a feasible scenario for FAHFA synthesis is critical for their use in uncovering biological mechanisms or in clinical trials. Here, we showcase a fully enzymatic approach, a novel in vitro bi-enzymatic cascade system, enabling an effective conversion of nature-abundant fatty acids into FAHFAs. Two hydratases from Lactobacillus acidophilus were used for converting unsaturated fatty acids to various enantiomeric hydroxy fatty acids, followed by esterification with another fatty acid catalyzed by Candida antarctica lipase A (CALA). Various FAHFAs were synthesized in a semi-preparative scale using this bi-enzymatic approach in a one-pot two-step operation mode. In all, we demonstrate that the hydratase-CALA system offers a promising route for the synthesis of optically pure structure-diverse FAHFAs.

Full text of DOI:10.1002/cbic.202100070

Full Papers
doi.org/10.1002/cbic.202100070
ChemBioChem
A Bi-Enzymatic Cascade Pathway towards Optically Pure
FAHFAs**
Yan Zhang,[a] Bekir Engin Eser,*[a] and Zheng Guo*[a]
Recently discovered endogenous mammalian lipids, fatty acid
esters of hydroxy fatty acids (FAHFAs), have been proved to
have anti-inflammatory and anti-diabetic effects. Due to their
extremely low abundancies in vivo, forging a feasible scenario
for FAHFA synthesis is critical for their use in uncovering
biological mechanisms or in clinical trials. Here, we showcase a
fully enzymatic approach, a novel in vitro bi-enzymatic cascade
system, enabling an effective conversion of nature-abundant
fatty acids into FAHFAs. Two hydratases from Lactobacillus
acidophilus were used for converting unsaturated fatty acids to
various enantiomeric hydroxy fatty acids, followed by esterifica-
tion with another fatty acid catalyzed by Candida antarctica
lipase A (CALA). Various FAHFAs were synthesized in a semi-
preparative scale using this bi-enzymatic approach in a one-pot
two-step operation mode. In all, we demonstrate that the
hydratase-CALA system offers a promising route for the syn-
thesis of optically pure structure-diverse FAHFAs.
Introduction
Diabetes mellitus is a metabolic disorder caused by insulin
deficiency or insulin receptor resistance and characterized by
high blood glucose levels. It was estimated that more than
400 million people are suffering from diabetes worldwide.[1]
Diabetes gives rise to several complications such as cardiovas-
cular diseases, eye disease, renal and neurological problems,
resulting in higher risk of morbidity and mortality.[2] Thus,
diabetes has become a global disease with high economic
burden on health systems.[3] The traditional drugs used for
diabetes treatment often have various side effects, like
xerostomia or burning sensation in the mouth.[4] Therefore,
effective prevention and treatments using molecules of natural
origin, with low production cost and minimal side effects, are
desired. Recently, a new class of endogenous mammalian lipids
i.e. fatty acid esters of hydroxy fatty acids (FAHFAs), discovered
by Kahn and coworkers in 2014, have shown promising effects
for potential treatment of diabetes and inflammatory diseases.[5]
Based on the structure of acyl chains and the ester bond
position, FAHFAs are grouped in over 20 FAHFA families, which
have been found in organisms ranging from plants to humans,
especially in mammalian adipose tissue.[5a,6] As shown in
Figure 1, we illustrate the structures of several representative
FAHFAs and FAHFA-containing triacylglycerols (FAHFA-TGs) that
have been identified in mammalian tissues. Among them,
palmitic acid esters of hydroxy stearic acids (PAHSAs) are
Figure 1. Structures of representative FAHFA families and lipid classes that
have been identified in mammalian cells and tissues. 9-PAHSA displays
potent anti-diabetic and anti-inflammatory effects.[5a,7] Docosahexaenoic
acid-derived FAHFAs (e.g. 13-DHAHLA) have also been shown to exert anti-
inflammatory effects in both mice and humans.[6c] FAHFA-TGs play key roles
in the regulation of inflammation and metabolic health.[20]
representing one of the most abundant FAHFA families. To
date, eight different regioisomers of PAHSAs (5-, 7-, 8-, 9-, 10-,
11-, 12-, and 13-) have been identified, potentially with
significant variation in biological function and regulation. For
instance, through oral administration to mice, 9-PAHSA has
been demonstrated to display effective anti-diabetic effects via
promoting insulin secretion and glucose transport, as well as
prominent anti-inflammatory effects in dendritic cells. However,
5-PAHSA is only correlative with insulin sensitivity, i.e. anti-
diabetic effects.[5a,7]
Due to the therapeutic potential toward diabetes and
inflammation, the stereospecific and environmentally friendly
synthesis of FAHFAs has received substantial attention and
remains of high interest (Scheme 1). However, studies on
enzymes that synthesize FAHFAs in vivo are limited and the
abundancies of FAHFAs in biological samples are extremely low
and always naturally occurring as isomer mixtures, which make it
highly challenging to enrich and isolate the FAHFA isomers from
nature.[8] In order to identify and characterize the biological
properties of FAHFAs, several chemical synthesis strategies have
been developed recently. The first synthetic case of FAHFAs was
[a] Dr. Y. Zhang, Dr. B. E. Eser, Prof. Dr. Z. Guo
Department of Biological and Chemical Engineering,
Aarhus University
Gustav Wieds Vej 10, 8000 Aarhus (Denmark)
E-mail: bekireser@bce.au.dk
guo@bce.au.dk
[**] A previous version of this manuscript has been deposited on a preprint
server https://doi.org/10.26434/chemrxiv.13650629.v1.
Supporting information for this article is available on the WWW under
https://doi.org/10.1002/cbic.202100070
ChemBioChem 2021, 22, 1–9
1
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

Full Papers
doi.org/10.1002/cbic.202100070
ChemBioChem
hydration leading to generation of various HFAs with different
À OH positions (Table 1).[14] Moreover, we have shown semi-
preparative scale enzymatic synthesis of a number of HFA
products, with excellent regio- and stereoselectivity, using these
wild-type and engineered FAHs (Table S1). Inspired by our
diversiform FAH toolbox enabling the synthesis of various HFAs,
we envisioned that we could use lipases in tandem with FAHs for
the efficient synthesis of various FAHFAs, through a novel bi-
enzymatic cascade pathway as shown in Scheme 1. Thus, on the
basis of FAH driven conversion of FAs to HFAs, we executed the
following work strategy: enabling the biotransformation of HFA
to FAHFA, and then assembling the two conversion steps in
tandem in one reaction system, affording a fully enzymatic
synthesis of FAHFAs.
Scheme 1. Overview of the synthesis strategies for FAHFA preparation. In
vivo: biosynthesis in de novo lipogenesis organs, like adipose tissue, liver and
kidney; Chemical synthesis: starting from 1,9-nonanediol, S/R-(+)-epichlorohy-
drin, methyl 9-oxononanoate or monoprotected α,ω-diols; This work: bi-
enzymatic cascade reaction by combination of hydratase and lipase for
producing FAHFAs from renewable biomass.
Results and Discussion
reported in 2014, where 9-PAHSA was produced starting from
1,9-nonanediol in eleven steps, however, synthesis of the other
analogs of 9-PAHSA were not achieved with this strategy.[5a] In
another method, several branched FAHFAs have been totally
synthesized from terminal alkenes and alkynes in seven steps,
including epoxide ring opening with acetylide carbanions
mediated by boron trifluoride and hydrogenation of the alkyne
function.[9] Subsequently, several similar strategies using Grignard
reagents have also been developed starting from non-racemic
epichlorohydrin and monoprotected α,ω-diols to accomplish the
introduction of the hydroxy group at different positions of the
long chain, followed by esterification with another fatty acid
molecule for enantioselective synthesis of FAHFAs.[10]
The key step of FAHFA synthesis is the introduction of the
hydroxy group at different positions of the fatty acids (FAs) to
produce hydroxy fatty acids (HFAs). Thus, the complexity of
chemical synthetic strategies are caused by the challenge of
hydroxylation, which involve harsh reaction conditions like high
temperatures, hazardous organic solvents and elaborate group
protecting steps.[11] Many chemical methods have low selectivity
and do not render the biologically active enantiomer.[10a]
However, long-term biological studies as well as potential
medical treatments and dietary supplements require a sufficient
supply of the desired FAHFAs. In this respect, enzymatic methods
have great potential for high-scale production of desired FAHFAs
with excellent selectivity in a green manner. In contrast to
chemical catalysts, Nature has developed an extensive list of
enzymes to generate HFAs from FAs, including lipoxygenases,
cytochrome P450s and fatty acid hydratases.[12] Among them,
fatty acid hydratases (FAHs) are able to perform asymmetric
synthesis of enantiomerically pure HFAs from abundant renew-
able unsaturated fatty acids. FAHs use water as a substrate and
catalyze the hydration without the need for continuous supply of
reducing cofactors (e.g. NAD(P)H), which present an attractive
green pathway for the production of HFAs.[13] Although, until
recently, low substrate promiscuity and limited regioselectivity of
FAHs had been the main hurdle for their wider applicability, we
demonstrated earlier that wild-type and rationally engineered
variants of two FAHs from Lactobacillus acidophilus greatly
expanded substrate scope and regioselectivity of fatty acid
Biotransformation of HFA to FAHFA
In a first round of experiments, we screened 21 commercially
available lipases to explore their esterification activity toward
palmitic acid (PA) and rac-12-hydroxystearic acid (rac-12-HSA),
selected as a model reaction for HFA to FAHFA conversion.
°
Small-scale (1 mL) activity assays were performed at 60 C using
anhydrous toluene (<0.001% water) as solvent. To avoid the
self-esterification of rac-12-HSA, the concentration of PA was
excessive (2-fold) over rac-12-HSA. After 24 h, the reactions
were analyzed by GC-FID (see Supporting Information for
method), and half of the tested lipases exhibited activity as
shown in Table 2 (column 2). Among them, five candidates
have shown high to excellent activities, with >55% conversion.
The active lipases can be categorized into two sources of
strains: Candida antarctica A and Candida rugosa (C. cylindracea
is a former name of C. rugosa); even though they may be from
different producers of different activities with/without genetic
modification. However, the structures and substrate binding
pockets of these lipases are highly similar with tunnel-like
binding sites.[15] Considering the compatibility of FAH and a
robust lipase, we decreased the screening reaction temperature
°
to 30 C in the second round of experiments. Interestingly, only
CALA retained high conversion rate under lower temperature
°
°
(85.3% at 30 C vs. 91.9% at 60 C, Table 2), which indicated
that CALA displays a prominent activity and a broad temper-
ature range for the synthesis of FAHFAs. Actually, the substrate
specificity of CALA had been explored in many studies, which
demonstrated that CALA exhibited a relatively high activity
towards straight-chained primary alcohols and carboxylic acids,
as well as sterically hindered secondary and tertiary alcohols.[16]
However, low or no activity was observed towards very short-
chain acids.[16] In addition, CALA was also used for enzymatic
synthesis of fatty acid polyesters (estolide) from 12-hydroxy-9-
cis-octadecenoic acid (ricinoleic acid).[17] These results indicate
that CALA has a unique applicability for the esterification of
secondary alcohols (like HFAs) and long-chain FAs, which is an
excellent candidate for production of FAHFAs. Thus, CALA was
chosen for further investigation in our case.
ChemBioChem 2021, 22, 1–9
www.chembiochem.org
2
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

Full Papers
doi.org/10.1002/cbic.202100070
ChemBioChem
Table 1. Hydroxy fatty acids that can be synthesized from fatty acid precursors by two FAHs from L. acidophilus and their variants as reported in previous
studies.[14]
Enzyme
Fatty acid
Hydroxy fatty acid[a]
Conv. %[a]
FA-HY2
FA-HY2
FA-HY2
FA-HY2
FA-HY2
FA-HY1
FA-HY1
FA-HY1
FA-HY1
FA-HY1
FA-HY1
FA-HY1
FA-HY1
FA-HY1
FA-HY1
Palmitoleic acid (16:1Δ9
)
10-Hydroxy-16:0
10-Hydroxy-18:0
89.3
96.9
40.9
60.8
1.7
58.8
48.1
57.0
56.6
53.8
12.1
9.2
18.7
51.7
14.1
4.7
0.4
0.2
45.7
10.3
6.0
0.6
35.6
2.8
92.1
69.4
[b]
Oleic acid (18:1Δ9
)
Linoleic acid (18:2Δ9,Δ12
)
(S)-10-Hydroxy-cis-12-18:1[a]
10-Hydroxy-cis-12,cis-15-18:2
12-Hydroxy-cis-5,cis-8,cis-14-20:3
12-Hydroxy-18:0
[b]
α-Linolenic acid (18:3Δ9,Δ12,Δ15
)
[b]
Arachidonic acid (20:4Δ5,Δ8,Δ11,Δ14
)
cis-Vaccenic acid (18:1Δ11
)
[b]
Linoleic acid (18:2Δ9,Δ12
Pinolenic acid (18:3Δ5,Δ9,Δ12
γ-Linolenic acid (18:3Δ6,Δ9,Δ12
)
(R)-13-Hydroxy-cis-9-18:1
13-Hydroxy-cis-5,cis-9-18:2
13-Hydroxy-cis-6,cis-9-18:2
(S)-13-Hydroxy-cis-9,cis-15-18:2
13-Hydroxy-cis-6,cis-9,cis-15-18:3
15-Hydroxy-cis-11-20:1
[b]
)
)
α-Linolenic acid (18:3Δ9,Δ12,Δ15
Stearidonic acid (18:4Δ6,Δ9,Δ12,Δ15
Eicosadienoic acid (20:2Δ11,Δ14
Mead acid (20:3Δ5,Δ8,Δ11
Sciadonic acid (20:3Δ5,Δ11,Δ14
Dihomo-γ-linolenic acid (20:3Δ8,Δ11,Δ14
)
[b]
)
)
)
12-Hydroxy-cis-5,cis-8-20:2
)
15-Hydroxy-cis-5,cis-11-20:2
15-Hydroxy-cis-8,cis-11-20:2
12-hydroxy-cis-8,cis-14-20:2
12-Hydroxy-cis-14,cis-17-20:2
15-Hydroxy-cis-11,cis-17-20:2
15-Hydroxy-cis-5,cis-8,cis-11-20:3
12-Hydroxy-cis-5,cis-8,cis-14,cis-17-20:4
15-Hydroxy-cis-5,cis-8,cis-11,cis-17-20:4
14-Hydroxy-cis-4,cis-7,cis-10,cis-16,cis-19-22:5
10-Hydroxy-cis-6,cis-12-18:2
14-Hydroxy-cis-4,cis-7,cis-10,cis-16,cis-19-22:5
(S)-12-Hydroxy-cis-8,cis-14-20:2
(S)-12-Hydroxy-cis-5, cis-8, cis-14,cis-17-20:4
)
FA-HY1
Eicosatrienoic acid (20:3Δ11,Δ14,Δ17
Arachidonic acid (20:4Δ5,Δ8,Δ11,Δ14
)
FA-HY1
FA-HY1
)
Eicosapentaenoic acid (20:5Δ5,Δ8,Δ11,Δ14,Δ17
)
FA-HY1
Docosahexaenoic acid (22:6Δ4,Δ7,Δ10,Δ13,Δ16,Δ19
)
FA-HY2-T391S[c]
γ-Linolenic acid (18:3Δ6,Δ9,Δ12
Docosahexaenoic acid (22:6Δ4,Δ7,Δ10,Δ13,Δ16,Δ19
Dihomo-γ-linolenic acid (20:3Δ8,Δ11,Δ14
Eicosapentaenoic acid (20:5Δ5,Δ8,Δ11,Δ14,Δ17
)
FA-HY2-H393S
FA-HY2-T391S/H393S/I378P
FA-HY2-T391S/H393S/I378P
)
)
)
[a] The conversions and absolute configurations of the hydroxy fatty acids were reported in previous studies.[14] [b] Indicates that the hydration of these
substrates were used to verify the cascade reaction in a semi-preparative scale in this study. [c] Only the variants that exhibited the highest conversion were
displayed.
Table 2. Lipase screening for fatty acid and hydroxy fatty acid esterification.[a]
°
°
Commercial lipase
Conv. % (60 C)
Conv. % (30 C)
CALA (Immobilized Candida antarctica lipase A from Codexis)
CALB (Liquid/Unimmobilized Candida antarctica lipase B)
Novozyme 435 FG (Candida antarctica lipase B immobilized on a macroporous acrylic resin)
Lipozyme 435 (Candida antarctica lipase B immobilized on silica)
Lipozyme TL IM (Lipase from Thermomyces lanuginosus immobilized on silica)
Lipozyme RM IM (Lipase from Rhezomucor miehei immobilized on a macroporous ion-exchange resin)
NS-40042 (Genetically modified lipase from Thermomyces lanuginosus)
NS-40079 (Genetically modified lipase from Thermomyces lanuginosus)
Lipase A “Amano” 12 (From Aspergillus niger)
Lipase D “Amano” 350 (From Rhizopus delemar)
Lipase G “Amano” 50 (From Penicillium camemberti)
Lipase M “Amano” 10 (From Mucor javanicus)
Lipase OF (From Candida cylindracea (C. rugosa))
91.9
4.0
6.6
27.5
1.7
ND
4.1
10.2
ND
ND
ND
ND
56.5
81.5
ND
ND
63.4
69.7
1.2
85.3
ND[b]
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
5.1
19.5
ND
ND
21.1
9.5
ND
ND
ND
Lipase MY (From Candida cylindracea (C. rugosa))
Lipase F-AP15 (From Rhizopus oryzae)
Lipase PSÀ D “Amano” I (From Burkholderia cepacia)
Lipase from Candida rugosa (Sigma, powder)
Lipase from Candida cylindracea (Sigma, powder)
Lipase from porcine pancreas (Sigma, powder)
Lipase from Candida rugosa immobilized on Immobead 150 (Sigma, powder)
Lipase AK from Pseudomonas fluorescens immobilized on Immobead 150 (Sigma, powder)
3.6
ND
°
[a] Reaction system: Lipase-10 mg, 10 mg/ml PA, 5 mg/ml rac-12-HSA, Toluene, 30 or 60 C, 250 rpm, 24 h. [b] ND: not detected.
Subsequently, the reaction was scaled up to 20 mL for product
preparation and identification (palmitic acid esters of 12-hydrox-
ystearic acids, 12-PAHSA) under the same conditions. From time-
course analysis (Figure S6), it was obvious that there was no
°
significant difference when reactions were performed at 30 C or
°
°
60 C (final conversion was slightly higher at 60 C), which was
ChemBioChem 2021, 22, 1–9
www.chembiochem.org
3
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

Full Papers
doi.org/10.1002/cbic.202100070
ChemBioChem
consistent with the small scale screening result. The reaction rate
started decreasing after 1.5–2.0 h, and conversion levels reached
Assembling and optimization of the bi-enzymatic cascade
°
°
86.8% or 90.5% at 8 h for 30 C or 60 C, respectively. Less than
complete conversion is probably due to the generation of water
as the by-product, driving the equilibrium towards reactant side.[18]
Afterwards, purification of the product was carried out on
preparative TLC plate, giving 78.1% isolated molar yield and
90.1% purity. The structure of isolated product was confirmed by
1H and 13C NMR (Figure S18–19).
As described above, no esterification reaction takes place
between PA and rac-12-HSA when CALA was subjected to
single-phase aqueous buffer. Thus, in order to identify the
optimal solvent system for the cascade, we evaluated the
conversion rate of FA-HY1 and CALA in different biphasic
systems (Figure 2a). Several organic solvents with the log P
value varying from 2.5 to 5.8 were selected, as well as three
versatile eco-friendly solvents (2-methyltetrahydrofuran, 2-
MeTHF; cyclopentyl methyl ether, CPME; and γ-valerolactone,
GVL), which have been shown as promising bio-solvents in
lipase-catalyzed reactions.[20] Obviously, due to the poor organic
solvent tolerance and low stability, the conversion rate of LA by
FA-HY1 in nine of the biphasic systems were lower than 10%
compared with 74.5% in single Kpi buffer. CALA displayed
similar performance, with less than 30% esterification in most
of the biphasic systems. However, 90.4% conversion was
observed in Kpi/Toluene system, which seems to be optimum
for the bi-enzymatic cascade. The three green solvents are not
suitable for either hydratase or CALA in our case.
We further compared two different operation modes in Kpi/
Toluene biphasic system; one-pot one-step mode where all
enzymes and substrates were added at the start of the reaction;
and one-pot two-step mode in which the hydration reaction of
unsaturated fatty acid was performed first (in single Kpi buffer),
followed by the addition of CALA, toluene and another fatty
acid. As shown in Figure 2c, different Kpi/Toluene ratios were
evaluated for both reaction modes. The highest overall
conversion was achieved at 65% with one-pot two-step mode
Preliminary optimization of the two conversion steps
We also simply compared the catalytic capacity of CALA under
different conditions, in terms of reaction solvents and the ratio
of PA/rac-12-HSA (Table S2). As anticipated, the esterification
activity of CALA was completely lost in single aqueous phase
(Kpi, 100 mM potassium phosphate buffer, pH 6.0). When
toluene was included, the conversion rates achieved were
43.3% and 69.9% in Kpi/Toluene (1:1 v/v) and pure toluene,
respectively (Table S2, entry 1–3). We next investigated the
effect of the PA/rac-12-HSA ratio on the production of 12-
PAHSA. The yield of 12-PAHSA improved when a higher PA/rac-
12-HSA ratio was reached. By increasing the PA concentration
to 10 mg/mL and lowering the rac-12-HSA concentration to
1 mg/mL, we were able to achieve the highest conversion at
95.1% (Table S2, entry 3–5). Moreover, we performed a control
reaction without PA substrate (only involved rac-12-HSA). As
expected there was no peak observed for 12-PAHSA, however,
a new peak was observed with less than 10% conversion which
might be caused by the self-esterification of rac-12-HSA (Fig-
ure S10). Interestingly, this product was not observed when PA
was present in the reaction system, indicating a lower reactivity
of the rac-12-HSA as an acyl donor.
After demonstrating efficient enzymatic conversion of a
HFA to FAHFA, we next explored a bi-enzymatic cascade
reaction where FAH and CALA are combined for direct
conversion of renewable FAs into FAHFAs. As described in the
literature, one of the hurdles in the biotransformation of plant
oils is the low fatty acid transport rate across the membrane of
the cells from or into the media.[19] Since we initially aimed at
using whole cells expressing hydratase as the biocatalyst in the
first step of our cascade, we firstly investigated FA and HFA
transport through the membrane of E. coli cells. After perform-
ing the hydration of linoleic acid (LA) using the E. coli whole-
cells (containing FA-HY1) as catalyst, both LA and 13-hydroxy-
cis-octadec-9-enoic acid (13-HODA) were observed at higher
concentration in the cells (over 4-fold than the supernatant,
Figure S7), which indicated that the HFA product was trapped
in majority inside the cells. To ensure a sufficient HFA
interaction with the second enzyme CALA, we chose to use cell
lysates of E. coli expressing FAHs in the cascade reaction,
facilitating FAHFAs preparation.
Figure 2. Cascades for the transformation of LA into (R)-13-PAHODA.
a) Organic solvent tolerance of FA-HY1 and CALA. Kpi means only Kpi buffer
without organic solvent, the volume ratio of Kpi/organic solvent for other
reactions is 1:1. b) Reaction scheme for the synthesis of (R)-13-PAHODA
from LA with the bi-enzymatic cascade. c) The influence of reaction mode
and solvent ratio to the cascade reaction. d) Time dependent one-pot two-
step bi-enzymatic cascade for biotransformation of LA and PA into (R)-13-
PAHODA.
ChemBioChem 2021, 22, 1–9
www.chembiochem.org
4
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

Full Papers
doi.org/10.1002/cbic.202100070
ChemBioChem
using 1:1 Kpi/Toluene ratio. With one-pot one-step mode, all of
the tested solvent ratios resulted in less than 20% overall
conversion, which can be attributed to the instability of FAHs in
the presence of organic solvent. Therefore, for all further
experiments, a one-pot two-step operation was followed. Under
the optimized one-pot two-step conditions, the enzymatic
synthesis of (R)-palmitic acid ester of 13-hydroxy-cis-octadec-9-
enoic acid ((R)-13-PAHODA) from LA and PA was scaled up to
100 mL in a semi-preparative scale. FA-HY1 almost completely
converted LA into its 13-OH product ((R)-13-HODA) in 8 h
(Figure 2d). The second step took significantly longer after
adding CALA, PA and toluene to the sample pot. The final
composition of (R)-13-PAHODA amounted up to 56% at plateau
stage. In other words, the cascade reaction reaches an
equilibrium, which might be caused by a reversible balance
between hydrolysis and esterification.
FAHFA preparation. By using either FA-HY1 or FA-HY2, oleic
acid (OA), LA, cis-vaccenic acid (CVA) and α-linolenic acid (ALA)
were converted to corresponding HFAs with different À OH
position, followed by the esterification with PA catalyzed by
CALA. From the semi-preparative reactions in 100 mL media,
several branched FAHFA products were synthesized, and the
overall conversion of FAHFA products can be achieved at 28–
67%. Since the polarity of the remaining excess fatty acid
substrates and FAHFA products are quite similar, trailing
phenomena of these compounds on silica gel rendered the
separation of fatty acids and FAHFAs very difficult. To eliminate
the trailing phenomena caused by the strong interaction
between the carboxyl groups of FAHFA and silica gel for a
better FAHFA isolation, methylation of reaction mixture before
purification was performed to convert the carboxyl groups to
corresponding methyl esters. On the other hand, methyl esters
(or other esters) of FAHFAs can afford similar reactivity as free
FAHFAs as acyl donors for interesterification or biological
incorporation into triglycerides, which are the major form of
FAHFAs in the tissues.[21] As shown in Figure S8, a better
resolution of fatty acids and FAHFA products was observed on
TLC plate after methylation. Then, the methylated FAHFA
products were purified by preparative TLC plate. Among them,
palmitic acid ester of 12-hydroxy steric acid (12-PAHSA),
prepared by converting CVA to 12-HSA (catalyzed by FA-HY1)
and followed by esterification with PA (catalyzed by CALA), was
isolated in a highest molar yield of 42%. The majority of the
purified FAHFA products had a purity of >90%, as determined
by GC-FID analysis. The structures of isolated products were
confirmed by 1H and 13C NMR (Figures S20–31). The absolute
configurations of three products i.e. palmitic acid ester of 10-
hydroxy-cis-octadec-12-enoic acid (10-PAHODA), palmitic acid
ester of 13-hydroxy-cis-octadec-9-enoic acid (13-PAHODA) and
palmitic acid ester of 13-hydroxy-cis-9,cis-15-octadecadienoic
acid (13-PAHODDA) can be given as (S), (R) and (S), respectively,
based on our previous analysis of the corresponding chiral
HFAs. The enantiomeric excess (ee) of relevant HFA products
were >95%.[14]
Substrate scope exploration
As described above, FAHFA regioisomers display differential
regulation and biological activity,[5a,7] which have also been
observed for FAHFA enantiomers. For example, only (R)-9-
PAHSA enantiomer was identified as the predominant enan-
tiomer that accumulates in adipose tissues from transgenic
mice.[10a] Encouraged by the broad substrate promiscuity of our
FAHs library, the substrate scope of this proof-of-concept bi-
enzymatic cascade was further expanded by employing the
above one-pot two-step approach for the production of differ-
ent enantiomeric FAHFAs (Figure 3). The typical hydratase, FA-
HY2, can only add water on Δ9 double bonds to produce 10-
OH products, whereas the unique FA-HY1 preferably hydrates
double bonds other than Δ9 to generate 12-OH, 13-OH, 14-OH
and 15-OH products. As we reported previously, FA-HY1 and
FA-HY2 are highly stereospecific enzymes that generate a single
stereoisomer product with an enantiomeric excess of over
95%.[14] Therefore, several pairs of FAH-substrates that exhibited
high conversion rates in our previous study,[14a] were chosen for
Conclusion
In summary, we have successfully demonstrated that the
medically important endogenous lipids of FAHFAs can be
synthesized enzymatically from renewable fatty acids using the
FAHs-CALA catalytic system, either by a two-step synthesis or
by a bi-enzymatic cascade of a one-pot two-step process. To
our knowledge, the only other study that reported synthesis of
fatty acid estolides (including FAHFAs) by an enzymatic cascade
approach utilizing hydratase and lipase is a patent,[22] where
oleate hydratase (OhyA) was used for the hydration of
unsaturated FAs at the Δ9-position to produce only 10-OH
products. By utilizing two different FAHs from L. acidophilus and
various fatty acids, our system greatly expands the product
scope by enabling preparation of various enantiomeric struc-
tures of FAHFAs, possibly associated with differential regulatory
and biological functions. Our results suggested that the two
Figure 3. Substrate scope of hydratase-lipase cascade reaction. The reactions
were performed in a one-pot two-step procedure as described above.
Conversion=[FAHFA]final/[Unsaturated fatty acid]initial ×100%; determined via
GC-FID as described in the Supporting Information. Yield: isolated molar
yield of methylated FAHFA products.
ChemBioChem 2021, 22, 1–9
www.chembiochem.org
5
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

Full Papers
doi.org/10.1002/cbic.202100070
ChemBioChem
from Pseudomonas fluorescens immobilized on immobead 150
(90678-10G) are purchased from Sigma-Aldrich (now Merck). fa-hy1
(GenBank ID: LC030242.1) and fa-hy2 (Genebank ID LC030243.1)
genes from the strain Lactobacillus acidophilus NTV001 were
received (from Prof. Jun Ogawa’s group at Kyoto University) as
cloned into pET-21b vector using XhoI and NdeI sites including a
stop codon at the end of the genes.
types of lipases from C. antarctica A and C. rugosa are highly
selective and active catalysts for esterification between FAs and
HFAs. Lipases from both strains exhibit similar tunnel-like
binding sites with ample space, which is capable to accept
bulky secondary alcohol donors; thus allowing effective inter-
action of two reactants with large steric hindrance.[15,23] Overall,
we demonstrate a promising method for enzymatic synthesis of
a variety of FAHFAs from sustainable fatty acids. We admit that
our approach is still in the conceptual stage, with low substrate
loading and incomplete transformation, probably caused by the
thermodynamic equilibrium of the second step in a biphasic
reaction system. It has been suggested that esterification is a
thermodynamically controlled process; the molar ratio of
substrates and water content strongly affect the initial reaction
rate and the final conversion, which can be enhanced by
maintaining the water activity sufficiently low to push the
thermodynamic equilibrium towards the formation of the
product.[24] Therefore, further optimization and reaction engi-
neering studies should enable feasible large-scale production of
FAHFAs for use in pre-clinical and clinical studies as well as for
future applications as medications and as dietary supplements.
Moreover, due to the limited regioselectivity of FAHs as well as
the limited double bond positions of natural fatty acids, the
proposed method cannot achieve the synthesis of some
FAHFAs, like 5-FAHFAs and 9-FAHFAs. Protein engineering
efforts to further increase the activity as well as to expand the
substrate scope and regiodiversity of FAHs is ongoing in our
lab. Use of lipoxygenases and cytochrome P450s, which also
add À OH groups on FAs with complementary selectivity and
substrate scope to FAHs, can further diversify the product
spectra of enzymatically synthesized FAHFAs.
Protein expression
Following transformation of E.coli Rosetta (DE3) cells (Merck
Millipore) with pET-21b-fa-hy1 or pET-21b-fa-hy2 plasmids, starter
°
cultures of 5 mL were grown overnight at 37 C in LB media
supplemented with a final ampicillin concentration of 100 μg/mL.
For protein expression, 0.5 L of LB media was inoculated with 5 mL
of overnight culture and the expression was induced by addition of
0.2 mM isopropyl-β-D-thiogalactopyranoside (IPTG) when the OD600
reached 0.6-1.0. After IPTG addition, the expression was continued
°
at 20 C and 180 rpm for 16–20 h. Cells were harvested by
°
centrifugation at 5,000 rpm and 4 C for 30 min. After washing the
cells twice with 0.85% NaCl solution, the cell pellet was stored at
°
À 80 C. For cell lysis, the cell pellet was resuspended in an ice-cold
lysis buffer of 100 mM potassium phosphate buffer (Kpi) at pH 6.0.
Sonicator (Bandelin Sonopuls) was used for cell disruption on ice.
After sonication, cell lysate was centrifuged at 12,000 rpm for
°
30 min at 4 C and the supernatant was used for following
reactions.
Small-scale reactions
Small-scale reactions were set up aerobically (we have determined
that there is no significant oxidation of the unsaturated fatty acids
up to 12 h, as shown in Figure S5) on 1 mL scale and were placed
in 5-mL glass screw top vials. After all materials were added, the
vials were then sealed with a cap and moved to a shaker. After
shaking at the indicated temperature, speed and time, ethyl acetate
was added for extraction. The mixture was vortexed for a few
seconds, and the phases were separated by centrifugation at room
temperature and 12,000 rpm for 10 min. For totally organic phase
reactions, the extraction step can be escaped. The organic phase
was moved to a clean Eppendorf tube, and the solvent was
completely evaporated under a gentle nitrogen flow. Subsequently,
100 μL BSTFA+TMCS (99:1) was added to convert the extracted
Experimental Section
Materials
Fatty acid substrates were purchased from Sigma-Aldrich (now
Merck), except cis-vaccenic acid (CVA) which was from MP
Biomedicals. All FAs had a purity of at least 98.5%. N,O-Bis
(trimethylsilyl)trifluoroacetamide with trimethylchlorosilane (BSTFA
+TMCS; 99:1) was purchased from Fisher Scientific. All other
chemicals were commercially available and analytical grade.
Immobilized Candida antarctica lipase A (CALA) is kindly provided
by Codexis. Liquid/unimmobilized Candida antarctica lipase B
(CALB), CALB immobilized on a macroporous acrylic resin (Novozym
435 FG), CALB immobilized on silica (Lipozyme 435), lipase from
Thermomyces lanuginosus immobilized on silica (Lipozyme TL IM),
lipase from Rhezomucor miehei immobilized on a macroporous ion-
exchange resin (Lipozyme RM IM), genetically modified lipase from
Thermomyces lanuginosus (NS-40042, NS-40079) are kindly provided
by Novozymes (Denmark). Lipase A “Amano” 12 from Aspergillus
niger, lipase D “Amano” 350 from Rhizopus delemar, lipase G
“Amano” 50 from Penicillium camemberti, lipase M “Amano” 10 from
Mucor javanicus, lipase OF from Candida cylindracea (C. rugosa),
lipase MY from Candida cylindracea (C. rugosa), lipase F-AP15 from
Rhizopus oryzae and lipase PSÀ D “Amano” I from Burkholderia
cepacia are kindly provided by Amano Enzyme. Lipase from
Candida rugosa (L1754-5G), lipase from Candida cylindracea (62316-
50G), lipase from porcine pancreas (L3126-25G), lipase from
Candida rugosa immobilized on immobead 150 (89444-10G), lipase
°
reaction mixture into their TMS derivatives at 60 C for 30 min. After
silylation, samples were moved to a 2 mL glass GC screw top vial
with a glass insert and analyzed via GC-FID.
Optimization of the cascade reaction
The effect of various organic solvents, including decane, isooctane,
n-heptane, n-hexane, cyclohexane, toluene, 2-MeTHF, CPME, and
GVL on enzyme activity of FA-HY1 and CALA were tested to
determine an optimal biphasic reaction system of the bi-enzymatic
cascade reaction. For FA-HY1 reaction system, 1 mL reaction
included 500 μL organic solvent, 500 μL supernatant of E. coli cell
lysate expressing FA-HY1 (sonication of 50 mg/mL cells in 100 mM
Kpi, pH 6.0) and 2 mg linoleic acid (LA). For CALA reaction system,
1 mL reaction included 500 μL organic solvent, 500 μL Kpi (100 mM,
pH 6.0), 10 mg palmitic acid (PA) and 5 mg rac-12-hydroxystearic
°
acid (rac-12-HSA). After shaking at 30 C and 220 rpm for 12 h,
reactions were extracted by adding 1 mL ethyl acetate. After
evaporation and silylation, samples were analyzed by GC-FID. The
optimum reaction operation mode was also determined by
comparing the overall conversions in one-pot one-step (all of the
starting materials and enzymes were added at the beginning) or
ChemBioChem 2021, 22, 1–9
www.chembiochem.org
6
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

Full Papers
doi.org/10.1002/cbic.202100070
ChemBioChem
one-pot two-step approach (the hydration reaction of one unsatu-
rated fatty acid was first performed in single Kpi buffer, followed by
the addition of CALA, toluene and another fatty acid). For one-pot
one-step reactions, all of the starting materials were mixed at the
beginning, including 500–1000 μL supernatants of FA-HY1, 100–
500 μL toluene, 10 mg CALA, 2 mg LA and 4 mg PA. Reactions were
Table 3. List of abbreviations
Abbreviation
Explanation
FAHFAs
CALA
FAHFA-TGs
PAHSA
FAs
HFAs
FAHs
Fatty acid esters of hydroxy fatty acids
C. antarctica lipase A
FAHFA-containing triacylglycerols
Palmitic acid esters of hydroxy stearic acid
Fatty acids
Hydroxy fatty acids
Fatty acid hydratases
Palmitic acid
rac-12-Hydroxystearic acid
Linoleic acid
Hydroxy-octadec-enoic acid
Oleic acid
cis-Vaccenic acid
°
performed at 30 C and 220 rpm for 24 h, then ethyl acetate was
added for extraction. For one-pot two-step reactions, the hydration
reaction of the LA was first performed in single Kpi buffer including
1 mL supernatants of FA-HY1 and 2 mg LA. After 12 h, then the
esterification was activated by the addition of 10 mg CALA, 1–2 mL
PA
rac-12-HSA
LA
HODA
OA
CVA
°
toluene and 4 mg PA. After shaking at 30 C and 220 rpm for
another 12 h, reactions were centrifuged and the toluene phases
were collected. After evaporation and silylation, samples were
analyzed by GC-FID.
ALA
α-Linolenic acid
HODDA
2-MeTHF
CPME
GVL
Hydroxy-octadecadienoic acid
2-Methyltetrahydrofuran
Cyclopentyl methyl ether
γ-Valerolactone
Preparative-scale reactions and product purification
All preparative-scale reactions were set at a scale of 20 mL
(esterification of PA and rac-12-HSA by CALA) or 100 mL (bi-
enzymatic cascade) total reaction volume in Erlenmeyer flasks
Kpi
Potassium phosphate buffer
°
under aerobic condition. Reaction mixtures were incubated at 30 C
with shaking at 220 rpm. For esterification of PA and rac-12-HSA,
100 mg rac-12-HSA, 200 mg PA, 200 mg CALA and 20 mL toluene
°
Injector and detector temperature was set at 320 C and the column
flow was maintained at a constant pressure of 22 Psi throughout
the analysis with helium as the carrier gas. Conversion was
calculated by the standard curves of palmitic acid (PA), 12-
hydroxystearic acid (rac-12-HSA) and palmitic acid ester of 9-
hydroxystearic acid (9-PAHSA).
°
were mixed and reactions were performed at 30 and 60 C,
200 rpm, respectively. The bi-enzymatic cascade reactions started
with an unsaturated FA substrate (100 mg) and 50 mL supernatants
of E. coli cells expressing the appropriate hydratase (50 mg/mL) in
100 mM Kpi buffer, pH 6.0 in first hydration step. The second step
was started by the addition of CALA (500 mg), PA (200 mg) and
50 mL toluene. Extraction of the lipid components from the
reaction mixture was performed by using ethyl acetate, similar to
that for the 1 mL scale assays, except the volumes of solvents used
for extraction were increased 100-fold and the extraction was
performed in several 50 mL centrifuge tubes. The upper organic
phase was recovered, the solvent was evaporated by a Buchi rotary
evaporator. The reaction mixture after evaporation was methylated
by following steps: add 2 mL of BF3-methanol and 0.6 mL of
Acknowledgements
This work was supported by Novo Nordisk Foundation (grant
NNF16OC0021740) and by AUFF-NOVA from Aarhus Universitets
Forskningsfond (AUFF-E-2015-FLS-9-12). We thank Prof. Jun
Ogawa from Kyoto University for providing the plasmids of
FA-HY1 and FA-HY2.
°
hydroquinone; heat at 80 C on the heating block for 2 min, shake
once in a while, cool to room temperature; add 0.4 mL of salt
solution, shake for 10 sec; add 2 mL of heptane, shake for 10 sec;
centrifuge at 4,000 rpm for 5 min; take the supernatant for
evaporation. Then, the methylated mixture was re-dissolved in ethyl
acetate (1–2 mL) for purification of FAHFA products on silica gel by
preparative TLC plates (L×W, 20 cm×20 cm, Merck). The solvent
path was 18 cm and the development system was hexane/acetone
(20:1, v/v). A Hanessian’s stain was used to stain the TLC plates.
The bands with the Rf value at 0.41, which belonged to the FAHFA
products, were scraped out of the TLC plates. The products in silica
gel powder were extracted with ethyl acetate and evaporated by a
Buchi rotary evaporator. The amount of dried products was
measured for yield calculation. Chemical structure identification of
FAHFA products were recorded on a Bruker Avance III spectrometer
(400 MHz).
Conflict of Interest
The authors declare no conflict of interest.
Keywords: bioactive FAHFAs · enzymatic synthesis · fatty acid
hydratase · lipase · renewable biomass
[1] K. Ogurtsova, U. Linnenkamp, R. F. J. D. da, Y. Huang, L. Guariguata, D.
Cavan, N. H. Cho, J. E. Shaw, L. E. Makaroff, Diabetes Res. Clin. Pract.
2017, 128, 40–50.
[2] a) O. Pinhas-Hamiel, P. Zeitler, Lancet 2007, 369, 1823–1831; b) Diabetes
Care 2013, 36, S67–S74; c) M. Stolar, Am. J. Med. 2010, 123, S3–S11.
[3] Y. Huang, S. Karuranga, B. Malanda, D. R. R. Williams, Diabetes Res. Clin.
Pract. 2018, 140, 351–352.
[4] I. D. Federation, Int. Diabetes Fed. 2017, Brussels, Belgium.
[5] a) M. M. Yore, I. Syed, P. M. Moraes-Vieira, T. Zhang, M. A. Herman, E. A.
Homan, R. T. Patel, J. Lee, S. Chen, O. D. Peroni, A. S. Dhaneshwar, A.
Hammarstedt, U. Smith, T. E. McGraw, A. Saghatelian, B. B. Kahn, Cell
2014, 159, 318–332; b) J. Lee, P. M. Moraes-Vieira, A. Castoldi, P. Aryal,
E. U. Yee, C. Vickers, O. Parnas, C. J. Donaldson, A. Saghatelian, B. B.
Kahn, J. Biol. Chem. 2016, 291, 22207–22217; c) L. Balas, C. Feillet-
Coudray, T. Durand, Chem. Eur. J. 2018, 24, 9463–9476.
GC-FID analysis
All samples analyzed via GC-FID were treated by silylation reagent
to convert the reaction mixture into their TMS derivatives. GC-FID
analysis was performed using Scion 436-GC (Bruker, Billerica, MA)
system equipped with a flame ionization detector (FID) and a non-
polar fused silica capillary Zebron ZB-5HT column (20 m length×
0.18 mm I.D.×0.18 μm film thickness; Phenomenex). The oven
program was a starting temperature of 150 C, a ramp of 4 Cmin
was settled to 310 C, followed by a hold for 20 min at 310 C.
À 1
°
°
°
°
ChemBioChem 2021, 22, 1–9
www.chembiochem.org
7
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

Full Papers
doi.org/10.1002/cbic.202100070
ChemBioChem
[6] a) M. Brezinova, O. Kuda, J. Hansikova, M. Rombaldova, L. Balas, K.
Bardova, T. Durand, M. Rossmeisl, M. Cerna, Z. Stranak, J. Kopecky,
Biochim. Biophys. Acta Mol. Cell Biol. Lipids 2018, 1863, 126–131; b) T. Hu,
M. Lin, D. Zhang, M. Li, J. Zhang, Anal. Bioanal. Chem. 2018, 410, 7415–
7428; c) O. Kuda, M. Brezinova, M. Rombaldova, B. Slavikova, M. Posta, P.
Beier, P. Janovska, J. Veleba, J. Kopecky, E. Kudova, T. Pelikanova,
Diabetes 2016, 65, 2580–2590; d) O. Kuda, M. Brezinova, J. Silhavy, V.
Landa, V. Zidek, C. Dodia, F. Kreuchwig, M. Vrbacky, L. Balas, T. Durand,
N. Hubner, A. B. Fisher, J. Kopecky, M. Pravenec, Diabetes 2018, 67,
1190–1199; e) Q. F. Zhu, J. W. Yan, T. Y. Zhang, H. M. Xiao, Y. Q. Feng,
Anal. Chem. 2018, 90, 10056–10063; f) Y. Ma, T. Kind, A. Vaniya, I.
Gennity, J. F. Fahrmann, O. Fiehn, J. Cheminf. 2015, 7, 53/51–53/55.
[7] a) P. M. Moraes-Vieira, A. Saghatelian, B. B. Kahn, Diabetes 2016, 65,
1808–1815; b) E. Pflimlin, M. Bielohuby, M. Korn, K. Breitschopf, M.
Loehn, P. Wohlfart, A. Konkar, M. Podeschwa, F. Baerenz, A. Pfenninger,
U. Schwahn, T. Opatz, M. Reimann, S. Petry, N. Tennagels, Cell Metab.
2018, 28, 217–227.e213.
[8] a) M. E. Ertunc, B. P. Kok, W. H. Parsons, J. G. Wang, D. Tan, C. J.
Donaldson, A. F. M. Pinto, J. M. Vaughan, N. Ngo, K. M. Lum, C. L. Henry,
A. R. Coppola, M. J. Niphakis, B. F. Cravatt, E. Saez, A. Saghatelian, J. Biol.
Chem. 2020, 295, 5891–5905; b) M. G. Kokotou, Biomolecules 2020, 10,
1092.
[9] L. Balas, J. Bertrand-Michel, F. Viars, J. Faugere, C. Lefort, S. Caspar-
Bauguil, D. Langin, T. Durand, Org. Biomol. Chem. 2016, 14, 9012–9020.
[10] a) A. T. Nelson, M. J. Kolar, Q. Chu, I. Syed, B. B. Kahn, A. Saghatelian, D.
Siegel, J. Am. Chem. Soc. 2017, 139, 4943–4947; b) H. Wang, T. Chang, S.
Konduri, J. Huang, A. Saghatelian, D. Siegel, J. Antibiot. 2019, 72, 498–
506; c) O. G. Mountanea, D. Limnios, M. G. Kokotou, A. Bourboula, G.
Kokotos, Eur. J. Org. Chem. 2019, 2019, 2010–2019.
[14] a) B. E. Eser, M. Poborsky, R. Dai, S. Kishino, A. Ljubic, M. Takeuchi, C.
Jacobsen, J. Ogawa, P. Kristensen, Z. Guo, ChemBioChem 2020, 21, 550–
563; b) A. Hirata, S. Kishino, M. Takeuchi, N. Kitamura, J. Ogawa, S.-B.
Park, J. Lipid Res. 2015, 56, 1340–1350.
[15] a) M. Widmann, P. B. Juhl, J. Pleiss, BMC Genomics 2010, 11, 123; b) S. H.
Albayati, M. Masomian, S. N. H. Ishak, M. S. B. M. Ali, A. L. Thean, F. B. M.
Shariff, N. D. B. M. Noor, R. N. Z. R. A. Rahman, Catalysts 2020, 10, 747.
[16] O. Kirk, M. W. Christensen, Org. Process Res. Dev. 2002, 6, 446–451.
[17] a) A. Arslan, A. Rancke-Madsen, J. Brask, Catalysts 2020, 10, 835; b) A.
Bodalo-Santoyo, J. Bastida-Rodriguez, M. F. Maximo-Martin, M. C. Mon-
tiel-Morte, M. D. Murcia-Almagro, Biochem. Eng. J. 2005, 26, 155–158.
[18] T. Kobayashi, S. Adachi, Biotechnol. Lett. 2004, 26, 1461–1468.
[19] 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, 7547–7553; b) E. Y. Jeon, J. W. Song, H. J. Cha, S. M.
Lee, J. Lee, J. B. Park, J. Biotechnol. 2018, 281, 161–167.
[20] a) G. de Gonzalo, A. R. Alcántara, P. Domínguez de María, ChemSusChem
2019, 12, 2083–2097; b) K. Watanabe, N. Yamagiwa, Y. Torisawa, Org.
Process Res. Dev. 2007, 11, 251–258; c) C. Aranda, G. de Gonzalo,
Molecules 2020, 25, 3016; d) F. G. Calvo-Flores, M. J. Monteagudo-
Arrebola, J. A. Dobado, J. Isac-Garcia, Top. Curr. Chem. 2018, 376, 1–40.
[21] D. Tan, M. E. Ertunc, S. Konduri, J. Zhang, A. M. Pinto, Q. Chu, B. B. Kahn,
D. Siegel, A. Saghatelian, J. Am. Chem. Soc. 2019, 141, 8798–8806.
[22] C. G. Boeriu, A. Todea, I. W. C. E. Arends, L. G. Otten, US20180073046
2018.
[23] a) D. J. Ericsson, A. Kasrayan, P. Johansson, T. Bergfors, A. G. Sandstroem,
J.-E. Baeckvall, S. L. Mowbray, J. Mol. Biol. 2008, 376, 109–119; b) P.
Grochulski, Y. Li, J. D. Schrag, F. Bouthillier, P. Smith, D. Harrison, B.
Rubin, M. Cygler, J. Biol. Chem. 1993, 268, 12843–12847.
[24] a) G. Sandoval, J. S. Condoret, P. Monsan, A. Marty, Biotechnol. Bioeng.
2002, 78, 313–320; b) D. T. Parikh, K. J. Lanjekar, V. K. Rathod, Biotechnol.
Lett. 2019, 41, 1163–1175.
[11] a) J. Jin, U. Hanefeld, Chem. Commun. (Camb.) 2011, 47, 2502–2510;
b) V. Resch, U. Hanefeld, Catal. Sci. Technol. 2015, 5, 1385–1399.
[12] a) J. C. Lewis, P. S. Coelho, F. H. Arnold, Chem. Soc. Rev. 2011, 40, 2003–
2021; b) F. Feyza Ozgen, M. E. Runda, B. O. Burek, P. Wied, J. Z. Bloh, R.
Kourist, S. Schmidt, Angew. Chem. Int. Ed. 2020, 59, 3982–3987; c) K.
Zorn, I. Oroz-Guinea, H. Brundiek, U. T. Bornscheuer, Prog. Lipid Res.
2016, 63, 153–164.
[13] a) H. Groeger, Angew. Chem. Int. Ed. 2014, 53, 3067–3069; Angew. Chem.
2014, 126, 3128–3130; b) Y. Zhang, B. E. Eser, P. Kristensen, Z. Guo, Chin.
J. Chem. Eng. 2020, 28, 2051–2063.
Manuscript received: February 16, 2021
Revised manuscript received: March 31, 2021
Accepted manuscript online: April 1, 2021
Version of record online: ■■■, ■■■■
ChemBioChem 2021, 22, 1–9
www.chembiochem.org
8
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

FULL PAPERS
Dr. Y. Zhang, Dr. B. E. Eser*,
Prof. Dr. Z. Guo*
1 – 9
A Bi-Enzymatic Cascade Pathway
towards Optically Pure FAHFAs
Full enzymatic synthesis of FAHFAs:
Various hydroxy fatty acids with
different À OH positions can be syn-
thesized from renewable fatty acids
by wild-type and engineered fatty
acid hydratases, which can be
cascaded with lipase to prepare value-
added FAHFAs with potential medical
value.