High-Level Production of Phenylacetaldehyde using Fusion-Tagged Styrene Oxide Isomerase

Article abstract of DOI:10.1002/adsc.202001500

An order-of-magnitude improvement in the production of phenylacetaldehyde to 3.37 M (405 g L^?1) from the enzymatic isomerisation of styrene oxide was achieved. A small ubiquitin-related modifier (SUMO) tag increases the productivity of the whole-cell biocatalytic system by enhancing the expression of active membrane-bound styrene oxide isomerase (SOI) while retaining enzyme catalytic efficiency and broad natural substrate scope. The isomerisation was performed by using Escherichia coli expressing SUMO-tagged SOI in an organic-aqueous biphasic system to yield 96% of phenylacetaldehyde. (Figure presented.).

Full text of DOI:10.1002/adsc.202001500

FULL PAPER
DOI: 10.1002/adsc.202001500
Very Important Publication
High-Level Production of Phenylacetaldehyde using Fusion-
Tagged Styrene Oxide Isomerase
a
b
b,
a,
Joel P. S. Choo, Richard A. Kammerer, Xiaodan Li, * and Zhi Li *
a
Department of Chemical and Biomolecular Engineering
National University of Singapore
4
+
Engineering Drive 4, 117585 Singapore
65-65168416
E-mail: chelz@nus.edu.sg
b
Laboratory of Biomolecular Research, Division of Biology and Chemistry
Paul Scherrer Institut
CH-5232 Villigen PSI, Switzerland
+
41-563104469
E-mail: xiao.li@psi.ch
Manuscript received: December 2, 2020; Revised manuscript received: January 5, 2021;
Version of record online: Februar 5, 2021
Supporting information for this article is available on the WWW under https://doi.org/10.1002/adsc.202001500
À 1
Abstract: An order-of-magnitude improvement in the production of phenylacetaldehyde to 3.37 M (405 gL )
from the enzymatic isomerisation of styrene oxide was achieved. A small ubiquitin-related modifier (SUMO)
tag increases the productivity of the whole-cell biocatalytic system by enhancing the expression of active
membrane-bound styrene oxide isomerase (SOI) while retaining enzyme catalytic efficiency and broad natural
substrate scope. The isomerisation was performed by using Escherichia coli expressing SUMO-tagged SOI in
an organic-aqueous biphasic system to yield 96% of phenylacetaldehyde.
Keywords: Styrene oxide isomerase; phenylacetaldehyde; fusion enzyme; green chemistry; biotransformations;
whole-cell catalysis
[
4a,b]
Introduction
formation,
introducing difficulties in product iso-
lation. In contrast, the biocatalytic isomerisation
Phenylacetaldehyde is a fragrance compound used in (Scheme 1) by styrene oxide isomerase (SOI, EC
the perfume industry to give hyacinth or rose-like 5.3.99.7) enables an alternative route using signifi-
[1]
flavours, and can be applied in polymer manufactur- cantly milder conditions, while achieving high selec-
[
5]
ing as a rate-controlling additive during ring-opening tivity with no by-product formation.
polymerisation of polyesters. It may also be used as a
[2]
As first reported in 1989, SOI is a vital component
building block in the synthesis of pharmaceutical of the epoxidation-isomerisation-oxidation metabolic
compounds, pesticides and other fragrance compounds pathway in styrene-degrading bacteria, found between
[3]
such as 2-phenylethanol. As it does not occur styrene monooxygenase (SMO) and phenylacetalde-
[6]
naturally in high concentrations, industrial production hyde dehydrogenase (PAD). Isomerisation by SOI
of phenylacetaldehyde is typically carried out through
[1a]
synthetic means, among which an important atom-
efficient industrial route is the isomerisation of styrene
oxide 1a to phenylacetaldehyde 2a catalysed by metal
[1b]
oxides. Also referred to as the Meinwald rearrange-
[
4]
ment, this reaction typically involves Lewis acids,
requiring toxic organic solvents such as THF,
[4d–f]
as
Scheme 1. Isomerisation of styrene oxide and substituted
styrene oxides by SUMO-SOI.
well as elevated reaction temperatures up to
[4a–e]
2
00°C.
Such harsh conditions lead to by-product
Adv. Synth. Catal. 2021, 363, 1714–1721
1714
© 2021 Wiley-VCH GmbH

FULL PAPER
asc.wiley-vch.de
[6c]
from Pseudomonas sp. VLB120
was used effec- (SOI) were cultured in M9 medium under the same
tively in several publications as an intermediate step conditions. Cell density and whole-cell specific activity
[3d,e,7]
for biocatalytic cascade reactions.
However, the for the isomerisation of styrene oxide was monitored,
synthetic potential of the single-step SOI reaction to as shown in Figure 1A. Both cultures exhibited similar
produce phenylacetaldehyde has not yet been clearly growth kinetics following induction, reaching a cell
À 1
demonstrated, despite its exceptionally high activity density of 4 gcdw L . Notably, the specific isomer-
[6c,7]
for styrene oxide.
Thus far, enzymatic isomer-
isation of styrene oxide has been limited to the use of
[
5b]
[5c,d]
SOI in cell-free systems, in crude, enriched,
or
[5e]
immobilised forms. The highest phenylacetaldehyde
À 1
concentration (322 mM, or 39 gL ) was achieved by
using crude cell lysate of recombinant E. coli express-
[5b]
ing SOI from Sphingopyxis. Irreversible inhibition
of SOI by phenylacetaldehyde had been proposed as a
[5c,e]
possible limiting factor.
We therefore aimed at combining several ap-
proaches to improve the productivity of biocatalytic
isomerisation by SOI. Firstly, by applying the cheaper
and potentially more stable whole cells as the
biocatalyst, it could be possible to achieve a high
initial activity to outcompete product inhibition and
attain higher product concentrations. Additionally, we
[8]
sought to investigate the use of a fusion tag to
improve the expression as well as stability of the
enzyme in the whole-cell biocatalyst for further
improvement of productivity. Moreover, biphasic
aqueous-organic systems could also enhance produc-
tivity through the in situ removal of inhibitory
[5e]
aldehyde.
We selected the SUMO (small ubiquitin-related
modifier) tag as the fusion partner in the engineering
of a fusion SOI enzyme, as it has been utilised
effectively to improve the expression levels of the
[9]
[9c]
attached protein, notably membrane proteins.
In
the context of biocatalysis, there is currently only one
study documenting the use of the SUMO fusion tag,
which was for improved operational stability at 50°C,
observed in the soluble enzyme D-psicose 3-
[10]
epimerase.
Therefore, the synthetic utility of the
SUMO tag for whole-cell-based catalysis with a
membrane-bound enzyme is as yet unknown.
In this study, we report the engineering and
application of SUMO-SOI as a novel catalyst to
produce 2a–d (Scheme 1) in high concentrations,
À 1
Figure 1. (A) Time-course of cell density (solid line) and
specific activity (dashed line) for cell cultures of E. coli (SOI,
notably phenylacetaldehyde 2a at 3.37 M (405 gL ).
&, in orange) and E. coli (SUMO-SOI, ~, in blue). The data
Results and Discussion
shown represent the averages of independent cell-growth
experiments performed in triplicate, with error bars indicating
standard deviation. For more details, see SI. (B) SDS-PAGE for
the analysis of enzyme expression in the membrane fraction of
E. coli (SOI) and E. coli (SUMO-SOI). 1: Membrane proteins
The SUMO-SOI insert was prepared by splicing
together the DNA sequences encoding the individual
proteins by PCR, and it was subsequently cloned into
pRSFduet-1 (see SI). Another pRSFduet-1 plasmid of E. coli (SOI) after DDM treatment; 2: Purified SOI; M:
[7a]
containing SOI described in a previous publication
Protein marker. Numbers indicate molecular mass in kDa; 3:
was used as a control. Both plasmids were then Purified SUMO-SOI; 4: Membrane proteins of E. coli (SUMO-
transformed into E. coli for expression. To compare SOI) after DDM treatment. Arrows indicate bands correspond-
ing to monomeric SOI (calculated Mw: 19.7 kDa) and mono-
meric SUMO-SOI (calculated Mw: 30.9 kDa).
differences in active expression of the target enzyme
during cell growth, E. coli (SUMO-SOI) and E. coli
Adv. Synth. Catal. 2021, 363, 1714–1721
1715
© 2021 Wiley-VCH GmbH

FULL PAPER
asc.wiley-vch.de
isation activity of E. coli (SUMO-SOI) remained 1:1 stoichiometric conversion of phenylacetaldehyde
higher than E. coli (SOI) after 10 hours, continuously 2a catalysed in situ by EcALDH. Comparing the
À 1
increasing and reaching an impressive 93 kUgcdw at isomerisation activities of the purified enzymes (Ta-
1
6 hours (stationary phase). In contrast, a sharp decline ble 1), we found that molar activities of SOI
À 1
À 1
in the specific activity of E. coli (SOI) from 74 to (10.5 Unmol ) and SUMO-SOI (10.7 Unmol ) were
À 1
5
2 kUgcdw could be observed between 14–16 hours.
similar, with the fusion tag retaining the catalytic
To investigate whether the better whole-cell activity efficiency of isomerisation (k /K was found to be
cat
m
6
À 1
6
À 1
of E. coli (SUMO-SOI) was due to improved kinetics 4.09×10 (M·s) and 4.58×10 (M·s) for SUMO-
or enhanced active expression, both enzymes were SOI and SOI respectively). This provides further
purified for further characterisation. Since SOI has evidence for higher expression levels of functional
[5c]
been reported to be an integral membrane protein,
SUMO-SOI (as opposed to improved kinetics) as the
we first isolated the membrane fractions of the cell main reason to explain the higher whole-cell specific
lysates via ultracentrifugation and performed detergent activities (Figure 1A) observed in E. coli (SUMO-SOI)
solubilisation using n-dodecyl β-D-maltoside (DDM). compared to E. coli (SOI).
The enzymes (both engineered with an N-terminal
To explore the synthetic potential of isomerisation
6
xHis tag) were then purified by Ni-NTA affinity by SOI, we applied an aqueous-organic biphasic
chromatography. The SDS-PAGE gel shown in Fig- system: the aqueous phase contains E. coli resting cells
ure 1B shows evidence of the monomeric forms of SOI suspended in potassium phosphate buffer, while the
(
calculated Mw: 19.7 kDa) and SUMO-SOI (calculated organic layer holds most of the hydrophobic substrate
Mw: 30.9 kDa) present in the membrane fraction and product. Initially, we tested several organic
Lanes 1 and 4). After purification, a secondary band solvents for the biphasic system in small-scale reac-
(
was seen for both enzymes (Lanes 2 and 3), along with tions. Solvents such as hexadecane and cyclohexane
a weak third band observed for SOI—a possible were quickly eliminated due to the poor solubility for
indication of SDS-resistant oligomeric states of both the substrate and product. Solvents with better
SOI and SUMO-SOI. The additional bands at 30 kDa solubility such as toluene and hexyl acetate could give
and 55 kDa in Lane 2 most likely correspond to the about 4 M of the product, but they are highly toxic and
dimeric and tetrameric forms of SOI respectively, costly. On the other hand, organic phases such as
while the additional band at 60 kDa in Lane 3 most biodiesel, ethyl oleate, and phthalate esters have been
likely corresponds to the dimeric form of SUMO-SOI. utilised effectively in biocatalytic systems as a greener
Based on band intensities in Lanes 1 and 4, the solvent and to reduce contact of cells with the highly
[3d,e,13]
SUMO-SOI construct can be seen to be more highly toxic styrene oxide.
expressed by E. coli following induction. The biocatalytic isomerisation in 1:1 v/v biphasic
Next, we compared the specific activities for systems was screened using the non-toxic organic
isomerisation of styrene oxide 1a. We note that the solvents biodiesel, ethyl oleate and dioctyl terephtha-
well-established aldehyde-specific assays described in late. Surprisingly, we observed that resting E. coli
[11]
the literature are incompatible with the SOI reaction, expressing SOI or SUMO-SOI were able to perform
when performed simultaneously at pH 8. Thus, a novel the isomerisation at appreciable rates even with a high
assay by using a second enzyme was developed, styrene oxide concentration of 4 M in the organic
utilising the enzyme E. coli phenylacetaldehyde dehy- phase (Figure 2). E. coli (SUMO-SOI) performed
drogenase (EcALDH), which readily oxidises phenyl- significantly better than the control E. coli (SOI),
acetaldehyde with concomitant production of enhancing yields by an average of 790 mM (Fig-
[12]
NADH. Product generation by SOI could therefore ure 2A). Evidently, this was the result of the higher
be coupled to NADH absorbance increase due to the whole-cell activity of E. coli (SUMO-SOI) achieving
Table 1. Activity and kinetic data of SOI and SUMO-SOI for the isomerisation of styrene oxide.
[c]
À 1 [c]
Enzyme
Specific activity by mass
of enzyme [Umg ]
Specific activity by mole
of enzyme [Unmol ]
K [mM]
m
kcat [s ]
k /K
cat m
[(M·s) ]
À 1 [a]
À 1 [b]
À 1 [c]
2
2
6
SOI
SUMO-SOI
534�37
345�46
10.5�0.7
10.7�1.4
0.045
0.049
2.06×10
2.01×10
4.58×10
6
4.09×10
[a]
Activity determined by using a coupled enzyme assay, by following NADH generation from EcALDH-catalysed oxidation of
phenylacetaldehyde 2a (Scheme S1) through detection of Abs340nm at regular time intervals. Reaction was conducted at 25°C in a
+
cuvette with 1 mL potassium phosphate buffer (0.05 M, pH 8) with 0.25 mM styrene oxide, 2 mM NAD , 6 U of purified
EcALDH and 0.15 μg of SOI or 0.27 μg of SUMO-SOI. One U is defined as 1 μmol of product formation per minute.
Determined using the calculated molecular weights of SOI (19.7 kDa) and SUMO-SOI (30.9 kDa).
Determined by non-linear regression of activity data for 0.015–0.2 mM styrene oxide (Figure S1).
[
b]
c]
[
Adv. Synth. Catal. 2021, 363, 1714–1721
1716
© 2021 Wiley-VCH GmbH

FULL PAPER
asc.wiley-vch.de
lowered reaction temperature from 30°C to 25°C for
better enzyme stability (Figure S3). As the screened
organic solvents were all equally viable for the
reaction, biodiesel was chosen since it can be inex-
pensively obtained from sustainable sources, which
would reduce the cost and environmental burden of the
process.
After optimisation, the biphasic system with E. coli
(
SUMO-SOI) could attain near-quantitative yields of
phenylacetaldehyde 2a from 3.5 M styrene oxide 1a
Table 2). No by-products were detected in the reac-
tion, with only unreacted styrene oxide remaining
Figures S6 and S14). Assuming full extraction of the
(
(
phenylacetaldehyde to the organic phase, product
concentration was determined to be 3.37 M
À 1
(
405 gL ), an order-of-magnitude improvement over
[5b]
a previous report of isomerisation by SOI.
Such
elevated product concentrations are not typically seen
in biocatalysis and is a demonstration of the strong
synthetic potential of SOI.
To explore the substrate scope and the synthetic
application of the SUMO-SOI construct, we performed
the isomerisation with halogen-substituted styrene
oxides 1b–d to produce the corresponding aldehydes,
which are used as intermediates for pharmaceutical
[14]
compounds.
The products 2b–d were sufficiently
hydrophobic to be completely extracted into the
organic phase, with no detectable product in the
aqueous phase (Figures S14–S16). Increase in bulki-
ness of the substituent resulted in a decrease in product
À 1
Figure 2. (A) Effect of organic phase (DOTP: dioctyl tereph-
thalate, EO: ethyl oleate and biodiesel) on the isomerisation of
concentration, from 94 gL (F-substituted 2b) to
À 1
1
1 gL (Br-substituted 2d), likely as a result of steric
4
M styrene oxide 1a to phenylacetaldehyde 2a in a biphasic
interference in the binding pocket (Table 2). Never-
theless, product concentrations remained significantly
higher for cells expressing SOI with the SUMO tag
than without (Table 2); we even observed the complete
consumption of substrate in the reaction catalysed by
E. coli (SUMO-SOI) (Figures S8, S10 and S12).
Evidently, the SUMO fusion construct retained the
system of total volume 6 mL (DOTP and EO) or 30 mL
(
biodiesel) conducted at 30°C for 2 hours using a 1:1 v/v
mixture of organic solvent and potassium phosphate buffer
(
À 1
0.4 M, pH 8) containing 5 gcdw
L
resting E. coli (SOI) or E.
coli (SUMO-SOI) whole cells. (B) Time course of phenyl-
acetaldehyde 2a produced in the organic phase (solid line) and
aqueous phase (dashed line) for the 30 mL biphasic biotransfor-
mation of 4 M styrene oxide 1a conducted at 30°C using a 1:1 broad substrate scope of SOI, an indication that the N-
v/v mixture of biodiesel and potassium phosphate buffer terminal SUMO tag does not interfere with substrate
À 1
(
0.4 M, pH 8) containing 5 g
L
resting E. coli (SOI, &, in binding. It might therefore serve as a useful technique
cdw
orange) or E. coli (SUMO-SOI, ~, in blue) whole cells. All
data points shown represent the averages of triplicate measure-
ments, with error bars indicating standard deviation.
for maximisation of expression levels when applied to
similar membrane enzymes. While the exact mecha-
nism of action of the SUMO tag has not yet been
determined, it has been suggested that the rapid folding
of the SUMO tag at the N-terminal can function as a
2
0–30% higher product yields within the first half- nucleation site to encourage better expression and
[9a]
hour of the reaction (Figure 2B). Additionally, in situ folding of the fused partner enzyme.
removal of inhibitory aldehyde was observed in all To demonstrate the industrial feasibility of the
experiments, as approximately 90–95% of the detect- process, a scale-up of the reaction was performed. The
able product could be extracted to the organic phase in biodiesel applied in this study was found to be less
our biphasic system (Figure 2).
volatile than phenylacetaldehyde, enabling product
Encouraged by the high productivity of the system, purification through distillation of the organic phase.
we further performed optimisation of yields. Increasing To reduce the likelihood of product decomposition, the
À 1
cell density to 15 gcdw
L
could push the isomerisation distillation was carried out at a lower temperature
reaction closer to completion (Figure S2). We also under reduced pressure. Using whole E. coli cells
Adv. Synth. Catal. 2021, 363, 1714–1721
1717
© 2021 Wiley-VCH GmbH

FULL PAPER
asc.wiley-vch.de
[a]
Table 2. Isomerisation of styrene oxide and para-halogen-substituted styrene oxides with E. coli expressing SOI or SUMO-SOI.
[b,c]
Substrate
Substrate concentration
Time
[h]
Catalyst
Product concentration
Conversion [%]
[b]
À 1
[mM]
mM
gL
[
d]
d]
[d]
[d]
1
a
a
b
b
c
c
d
d
3500
3500
750
750
250
250
100
100
2
2
2
2
4
4
4
4
E. coli (SOI)
E. coli (SUMO-SOI)
E. coli (SOI)
E. coli (SUMO-SOI)
E. coli (SOI)
E. coli (SUMO-SOI)
E. coli (SOI)
3030�10
3370�30
623�12
681�4
144�3
183�4
39�1
364�1
405�4
86�2
94�1
22�1
28�1
86.3�0.2
96.4�0.9
83.1�1.6
95.0�0.8
61.3�1.4
73.1�1.7
38.6�1.0
56.0�1.1
[
1
1
1
1
1
1
1
7.8�0.2
E. coli (SUMO-SOI)
56�1
11.1�0.2
[a]
Biotransformation was carried out at 25°C in 6 mL biphasic systems consisting of 1:1 v/v mixtures of biodiesel (organic phase)
À 1
and potassium phosphate buffer (0.4 M, pH 8) containing 15 g
L
resting E. coli cells (aqueous phase) expressing SOI or
cdw
SUMO-SOI. The data shown represent the averages of triplicate experiments, with standard errors presented.
Concentration in organic phase (assuming full extraction).
[
b]
c]
[
Products in the organic phase and aqueous phases were analysed by GC and HPLC respectively. For substrates 1b–d, no
significant amounts of product were found in the aqueous phase (Figures S15–S17).
Product concentrations of 2a in the aqueous phase were determined to be 0.137 M (16 gL ) and 0.315 M (38 gL ) for the
reactions catalysed by E. coli (SOI) and E. coli (SUMO-SOI) respectively.
[d]
À 1
À 1
expressing SUMO-SOI, a biphasic isomerisation reac- reaction to a point where it could rival traditional
tion was carried out with 3.5 M of styrene oxide, in a chemical Meinwald rearrangement systems. SUMO-
total reaction volume of 100 ml (21 g substrate). By SOI also retained the broad natural substrate scope of
subjecting the biodiesel layer to a simple short-path SOI, enabling the production of halogen-substituted
À 1
distillation, an isolated yield of 65% (13.7 g product) phenylacetaldehyde compounds at up to 94 gL .
could be obtained. Since phenylacetaldehyde self- Hence, the SUMO tag might even find greater
polymerises by aldol condensation and acetal forma- synthetic utility as a tool to improve expression levels
[15]
tion even at room temperature, product separation of other membrane-bound enzymes. This reported
by distillation likely resulted in major product losses. system clearly demonstrates a viable high-yielding and
Better isolated yields might be obtained by carrying greener biocatalytic alternative for the isomerisation of
out the distillation at lower temperatures with a styrene oxide and related compounds to their respec-
stronger vacuum or by purification using column tive aldehydes.
chromatography. This highly productive and simple
whole-cell system has potential for scale-up and could
lay the foundations for a practical biocatalytic method
Experimental Section
to produce phenylacetaldehyde.
Engineering of E. coli (SUMO-SOI) and E. coli
SOI)
(
Conclusion
SUMO-SOI was constructed by overlap extension PCR using
Phusion polymerase. Oligonucleotide primer sets (SUMO 5’
BamHI, 5’-ctcgtcggatccgATGAGCGATAGCGAAGTGAACC-
In conclusion, we have developed a system for
efficient enzymatic isomerisation of styrene oxide to
produce phenylacetaldehyde in high concentrations
3
’
and SUMO-SOI 3’, 5’-CGCGTGTAACATGCCGCC-
GATCTGTTCACGATG-3’) and (SOI 3’ XhoI, 5’-TACCA-
and yields using SUMO-SOI expressed in E. coli. The GACTCGAGTTATTACTCG-3’ and SUMO-SOI 5’, 5’-CA-
SUMO tag was shown to be effective at yielding a GATCGGCGGCATGTTACACGCGTTTG-3’) were designed
high expression level of functional SOI while retaining to amplify cDNA fragments using a synthetic SUMO gene
optimised for expression in E. coli and SOI (Pseudomonas sp.
catalytic efficiency. By using resting cells expressing
[7a]
VLB120),
as templates, respectively. The following PCR
SUMO-SOI, a higher productivity of isomerisation
was achieved compared to cells expressing SOI with-
out the SUMO tag. Phenylacetaldehyde was produced
conditions were used: 98°C for 1 min (1 cycle); 98°C for
1
5 sec, 50°C for 30 sec, 72°C for 30 sec (25 cycles), 100 μL
reactions. After gel extraction from a 1% agarose gel (QIAGEN
QIAquick Gel Extraction Kit) 25 μL each of the amplified
DNA fragments were combined and the PCR reaction repeated
using primers SUMO 5’ BamHI and SOI 3’ XhoI. The final
PCR product was purified (QIAGEN QIAquick PCR Purifica-
tion Kit) and digested with 30 units each of BamHI and XhoI
with no detectable by-products at a concentration of
À 1
4
05 gL by far the highest reported concentration for
biocatalytic production of phenylacetaldehyde. By
combining the productivity of whole cells with an
expression-enhancing tag, we were able to significantly
enhance the synthetic ability of the SOI-catalysed
Adv. Synth. Catal. 2021, 363, 1714–1721
1718
© 2021 Wiley-VCH GmbH

FULL PAPER
asc.wiley-vch.de
restriction enzymes (NEB) for 1 h at 37°C (100 μL reaction
volume). After purification (QIAGEN QIAquick PCR Purifica-
tion Kit), the amplified product was ligated into the BamHI/
XhoI site of pRSFduet-1 using Quick Ligase (NEB) (5 μl Quick
Ligase reaction Buffer (2X), 2 μL vector, 3 μL PCR product,
Desalting column, with desalting buffer (50 mM potassium
phosphate buffer, pH 8, 0.05% DDM). SDS-PAGE was
performed for the membrane fraction and purified enzymes
using a 12% gel with 4–5 μg of total protein per lane.
1
1
μL Quick Ligase; control: H O instead of PCR product). After
5 min, 5 μl of the ligation reactions were transformed in 100 ul
2
Measurement of Kinetic Parameters of SOI and
SUMO-SOI
of chemically competent XL-1 Blue bacterial cells. After
incubation on ice for 30 min, cells were heat-shocked at 42°C
for 60 sec. After cooling on ice for 2 min, cells were plated on
LB plates containing 100 μg/mL kanamycin(Kan) and incubated
For isomerisation activity measurement, the following coupled
enzyme assay was carried out (Scheme S1). To a 1.5 mL
cuvette, 1 mL reaction for 0.25 mM (~5 K ) styrene oxide was
m
overnight at 37°C. Individual colonies were grown overnight in
carried out at 25°C in reaction buffer (0.05 M potassium
Kan
4
mL LB medium at 37°C and recombinant plasmids were
+
phosphate buffer, pH 8) containing 2 mM NAD , 6 U
isolated (QIAGEN QIAquick Miniprep Kit). Insert sequences
were verified by DNA sequencing (GATC/Eurofins).
[7b]
EcALDH, and 0.15 μg SOI or 0.27 μg SUMO-SOI. Increase
in absorbance at 340 nm was recorded at regular time intervals
using a Hitachi U2900 spectrophotometer, and absorbance units
were converted to concentration using the ɛ340nm of NADH
Culture and Expression of E. coli for Biotransfor-
mation
(6.22 Abs/mM/cm). The slope of the concentration-time plot for
the first 15 seconds was subsequently used to determine activity,
with one U defined as corresponding to one micromole of
product generated in one minute. Calculated molecular weights
(19659.9 Da for SOI; 30903.56 Da for SUMO-SOI) were used
to determine the molar activity of the respective isomerases.
E. coli (SOI) and E. coli (SUMO-SOI) plasmids were trans-
formed into E. coli C41(DE3) Rosetta cells. Cells were grown
Kan
on LB at 37°C overnight. Single colonies were inoculated
Kan
into 2 ml LB and grown for 6–8 hours (250 rpm, 37°C). Each
inoculum was subsequently transferred into a culture flask of
To determine kinetic parameters Km and k , isomerisation
cat
5
0
0–1000 mL M9 (0.1 M pH 7.5 potassium phosphate buffer,
activity was determined with varying styrene oxide concen-
trations of 0.015–0.2 mM. The slope of the concentration-time
plot for the first 15 seconds was used to determine the initial
rate of each reaction in mM/min, to obtain a relationship
between initial activity and concentration. Non-linear regression
was then used to determine K and V based on the Michaelis-
.5 g NaCl, 1 g NH Cl, 20 g glucose, 6 g yeast extract, 0.1 mM
4
CaCl , 1 mM MgSO , and 1 ml of a 1000x trace metals
2
4
solution) containing 50 μg/mL kanamycin. OD600 of the cultures
were measured at various time points and used to calculate cell
density in cell dry weight (gcdw). Once OD600 reached a value of
m
max
0.6–0.7, IPTG was added (0.5 mM final concentration), then
Menten model, with the MATLAB function nlinfit (Figure S1).
The Vmax calculated was further converted to kcat by the
relationship kcat =Vmax/[E].
temperature was adjusted to 22°C for overnight growth.
Purification of SOI and SUMO-SOI
The cell pellet of 1 L of overnight M9 culture of E. coli (SOI) Isomerisation of 1a–d to 2a–d using E. coli (SOI)
or E. coli (SUMO-SOI) was resuspended in lysis buffer and E. coli (SUMO-SOI)
(
50 mM sodium phosphate buffer, pH 8, 0.3 M sodium chloride)
M9 cultures of the respective E. coli clones were centrifuged at
000×g, 6 min and the pellet was washed twice, once with
deionised water and once in reaction buffer (0.4 M potassium
phosphate buffer, pH 8). Cell density of the resulting suspension
was adjusted by repeated dilution and measuring Abs at
and subjected to high-pressure lysis at 550 bar by an SPX Flow
APV 1000 Lab Homogeniser. The cell lysate was then
centrifuged at 6000×g for 15 min to remove unbroken cells.
The remaining supernatant was subjected to high-speed
centrifugation at 32300×g for 1 h 45 min, and the red-brown
membrane pellet was then suspended in 35 mL of binding
buffer (50 mM sodium phosphate buffer, pH 8, 10 mM
imidazole, 0.3 M sodium chloride). 1% DDM was added, and
the solution was incubated on ice for 1 hour before being
centrifuged at 32300×g for 2 h. The solubilized red membrane
fraction was then subject to affinity chromatography with the
HisTrap HP column (GE Healthcare). After loading of sample,
the column was washed with wash buffer 1 (50 mM sodium
phosphate buffer, pH 8, 60 mM imidazole, 0.3 M sodium
chloride) and wash buffer 2 (50 mM sodium phosphate buffer,
pH 8, 100 mM imidazole, 0.3 M sodium chloride). Following
this step, the column had a brownish-red apprearance. Elution
was then carried out with elution buffer (50 mM sodium
phosphate buffer, pH 8, 500 mM imidazole, 0.3 M sodium
6
6
00 nm. In all biphasic systems described, the organic phase
containing dissolved substrate 1a–d in a known concentration
was added into a conical flask containing the cell suspension.
Each reaction was carried out in an orbital shaker with
temperature control and shaking speed set to 250 rpm. At
specific time points, 150 μL samples of the emulsion was taken,
and centrifuged at 20,000×g, 5 min. 10 μL of the organic phase
was subjected to GC analysis following dilution in 90 μL ethyl
acetate containing 17.32 g/L of ethylbenzene as internal
standard, before being subject to further dilution in ethyl acetate
(10–100×) for GC analysis. 12 μL of the aqueous phase was
diluted in 480 μL acetonitrile containing 0.5 mM of benzyl
alcohol as internal standard, along with 708 μL of a quenching
solution containing 0.76 M H PO4 to hydrolyse unreacted
3
[5c,16]
styrene oxide to 1-phenyl-1,2-ethanediol (detectable by reverse
phase HPLC, see Figure S13). The sample was then filtered
through a 0.2 μL Minisart RC4 filter into a vial for reverse
phase HPLC analysis.
chloride) until the red fraction of SOI was eluted(.
eluate containing purified SOI was then subject to concentration
using a 30 kDa concentrator and desalted using a HiTrap
The
Adv. Synth. Catal. 2021, 363, 1714–1721
1719
© 2021 Wiley-VCH GmbH

FULL PAPER
asc.wiley-vch.de
À
Preparative Biotransformation of Styrene Oxide 1a
to Produce Phenylacetaldehyde 2a
49.81. HRMS (APCI-): m/z 153.0114 (Calcd. for [MÀ H] :
1
53.0113)
1
(
4-Bromophenyl)acetaldehyde 2d: H NMR (400 MHz, CDCl )
To 50 mL of resting E. coli (SUMO-SOI) cell suspension
3
δ 9.74 (t, J=2.1 Hz, 1H, CHO), 7.54–7.44 (m, 2H, ArH), 7.13–
(0.4 M potassium phosphate buffer, pH 8) of cell density
1
3
7
.05 (m, 2H, ArH), 3.67 (d, J=2.1 Hz, 2H, CH2). C NMR
1
5 gcdw/L in a 500 mL conical flask, 50 mL of biodiesel
(
4
1
400 MHz, CDCl ) δ 198.51, 132.11, 131.30, 130.78, 121.57,
containing 3.5 M of styrene oxide 1a (21.03 g) was added. The
reaction was carried out for 2 h at 25°C in an orbital shaker set
at 250 rpm. 5 mL of H PO (85 wt%) was then added to quench
3
À
9.87. HRMS (APCI-): m/z 196.9603 (Calcd. for [MÀ H] :
96.9608)
3
4
the reaction, followed by 10 minutes of incubation to com-
pletely hydrolyse any remaining styrene oxide. The reaction
mixture was then saturated with NaCl and transferred to two
centrifuge tubes for centrifugation at 15000×g for 10 min. The
organic phase was then carefully transferred to a 50 mL round-
bottomed flask and distilled under reduced pressure in two
batches, using the short-path distillation setup shown in Fig-
ure S4. Distillation temperature: 100–140°C. Isolated yield:
Acknowledgements
We acknowledge financial support for this work from the
Ministry of Education, Singapore through an MOE Tier 2 Grant
(MOE 2016-T2-2-140; NUS WBS R-279-000-501-112).
1
3.66 g. Appearance: clear liquid.
References
1
Phenylacetaldehyde 2a: H NMR (400 MHz, CDCl ) δ 9.78–
3
[
1] a) C. Kohlpaintner, M. Schulte, J. Falbe, P. Lappe, J.
Weber, in Ullmann’s Encyclopedia of Industrial
Chemistry, Wiley-VCH Verlag GmbH & Co. KGaA,
Weinheim, Germany, 2000, pp. 1–39; b) C. Kohlpaintner,
M. Schulte, J. Falbe, P. Lappe, J. Weber, G. D. Frey, in
Ullmann’s Encyclopedia of Industrial Chemistry, Wiley-
VCH Verlag GmbH & Co. KGaA, Weinheim, Germany,
9
7
2
1
.73 (t, 1H, J=2.4 Hz, CHO), 7.41–7.35 (m, 2H, ArH), 7.34–
.28 (m, 1H, ArH), 7.25–7.20 (m, 2H, ArH), 3.69 (d, J=
1
3
.4 Hz, 2H, CH2). C NMR (400 MHz, CDCl ) δ 199.59,
3
31.97, 129.75, 129.15, 127.56, 50.72. HRMS (APCI-): m/z
À
119.0501 (Calcd. for [MÀ H] : 119.0502)
Preparative Biotransformation of Substituted
Styrene Oxide 1b–d to Produce the Corresponding
Phenylacetaldehydes 2b–d
2
013, pp. 1–7; c) J. Panten, H. Surburg, in Ullmann’s
Encyclopedia of Industrial Chemistry, Wiley-VCH Ver-
lag GmbH & Co. KGaA, Weinheim, Germany, 2016, pp.
1
–45.
To 30 mL of resting E. coli (SUMO-SOI) cell suspension
[
2] E. Hosseini Nejad, A. Paoniasari, C. E. Koning, R.
Duchateau, Polym. Chem. 2012, 3, 1308–1313.
3] a) X. Zhang, S. Sarkar, R. C. Larock, J. Org. Chem.
(0.4 M potassium phosphate buffer, pH 8) of cell density
15 gcdw/L in a 125 mL conical flask, 15 mL of DOTP containing
0.75 M or 0.25 M of 1b or 1c was added respectively; to
50 mL of resting E. coli (SUMO-SOI) cell suspension (0.4 M
[
2
006, 71, 236–243; b) P. B. Rasmussen, S. Bøwadt,
Synthesis 1989, 1989, 114–117; c) M. Sakai, H. Hirata,
H. Sayama, K. Sekiguchi, H. Itano, T. Asai, H. Dohra,
M. Hara, N. Watanabe, Biosci. Biotechnol. Biochem.
2007, 71, 2408–2419; d) B. R. Lukito, S. Wu, H. J. J.
Saw, Z. Li, ChemCatChem 2019, 11, 831–840; e) B. S.
Sekar, B. R. Lukito, Z. Li, ACS Sustainable Chem. Eng.
potassium phosphate buffer, pH 8) of cell density 15 gcdw/L in a
500 mL conical flask, 30 mL of biodiesel containing 0.25 M of
1
d was added. The reactions were carried out for 2 h in an
orbital shaker set at 250 rpm. The reaction mixtures were then
saturated with NaCl and placed into a centrifuge tube for
centrifugation at 15000×g for 10 min. The organic phases were
then carefully transferred to a 50 mL round-bottomed flask and
distilled under reduced pressure to obtain a crude distillate.
Distillation temperatures: 100–135°C (2b); 130–160°C (2c);
2
019, 7, 12231–12239.
[
4] a) R. Umeda, M. Muraki, Y. Nakamura, T. Tanaka, K.
Kamiguchi, Y. Nishiyama, Tetrahedron Lett. 2017, 58,
1
50–180°C (2d). As the products were found to be insuffi-
2
2
393–2395; b) X.-F. Zhang, J. Yao, X. Yang, Catal. Lett.
017, 147, 1523–1532; c) R. Srirambalaji, S. Hong, R.
ciently pure, a further purification was performed using a short
silica column (1:20 ethyl acetate:hexane) and the fractions with
highest purity were dried with sodium sulfate and used as
product standards.
Natarajan, M. Yoon, R. Hota, Y. Kim, Y. Ho Ko, K. Kim,
Chem. Commun. (Camb.) 2012, 48, 11650–11652; d) N.
Humbert, D. J. Vyas, C. Besnard, C. Mazet, Chem.
Commun. (Camb.) 2014, 50, 10592–10595; e) D. J.
Vyas, E. Larionov, C. Besnard, L. Guenee, C. Mazet, J.
Am. Chem. Soc. 2013, 135, 6177–6183; f) I. Karamé,
M. L. Tommasino, M. Lemaire, Tetrahedron Lett. 2003,
1
(
4-Fluorophenyl)acetaldehyde 2b: H NMR (400 MHz, CDCl )
3
δ 9.74 (t, J=2.2 Hz, 1H, CHO), 7.21–7.14 (m, 2H, ArH), 7.10–
13
7
.01 (m, 2H, ArH), 3.68 (d, J=2.2 Hz, 2H, CH2). C NMR
(
400 MHz, CDCl ) δ 198.98, 162.23 (d, J=246.1 Hz), 131.18
3
(
4
1
d, J=8.1 Hz), 127.54 (d, J=3.3 Hz), 115.91 (d, J=21.4 Hz),
4
4, 7687–7689; g) M. W. C. Robinson, K. S. Pillinger, I.
Mabbett, D. A. Timms, A. E. Graham, Tetrahedron 2010,
6, 8377–8382.
À
9.66. HRMS (APCI-): m/z 137.0412 (Calcd. for [MÀ H] :
37.0408)
6
1
[
5] a) R. Kourist, P. Domínguez de María, K. Miyamoto,
Green Chem. 2011, 13, 2607–2618; b) M. Oelschlagel,
L. Richter, A. Stuhr, S. Hofmann, M. Schlomann, J.
Biotechnol. 2017, 252, 43–49; c) M. Oelschlagel, J. A.
(
4-Chlorophenyl)acetaldehyde 2c: H NMR (400 MHz, CDCl )
3
δ 9.75 (t, J=2.1 Hz, 1H, CHO), 7.39–7.27 (m, 2H, ArH), 7.19–
.11 (m, 2H, ArH), 3.68 (d, J=2.2 Hz, 2H, CH2). C NMR
400 MHz, CDCl ) δ 198.64, 133.51, 130.95, 130.27, 129.15,
13
7
(
3
Adv. Synth. Catal. 2021, 363, 1714–1721
1720
© 2021 Wiley-VCH GmbH

FULL PAPER
asc.wiley-vch.de
Groning, D. Tischler, S. R. Kaschabek, M. Schlomann,
Appl. Environ. Microbiol. 2012, 78, 4330–4337; d) M.
T. S. Larch, F. Rudroff, M. Winkler, Adv. Synth. Catal.
2020, 362, 4673–4679.
Oelschlagel, J. Zimmerling, M. Schlomann, D. Tischler, [12] J. S. Rodriguez-Zavala, A. Allali-Hassani, H. Weiner,
Microbiology (Reading) 2014, 160, 2481–2491; e) M. Protein Sci. 2006, 15, 1387–1396.
Oelschlagel, A. Riedel, A. Zniszczol, K. Szymanska, [13] a) H. Toda, R. Imae, N. Itoh, Tetrahedron: Asymmetry
A. B. Jarzebski, M. Schlomann, D. Tischler, J. Biotech-
nol. 2014, 174, 7–13.
6] a) S. Hartmans, J. P. Smits, M. J. Van der Werf, F.
Volkering, J. A. M. De Bont, Appl. Environ. Microbiol.
2012, 23, 1542–1549; b) D. Kuhn, M. K. Julsing, E.
Heinzle, B. Bühler, Green Chem. 2012, 14, 645–653;
c) J. H. Han, M. S. Park, J. W. Bae, E. Y. Lee, Y. J.
Yoon, S.-G. Lee, S. Park, Enzyme Microb. Technol.
2006, 39, 1264–1269.
[
1
989, 55, 2850–2855; b) M. Oelschlagel, J. Zimmerling,
D. Tischler, Front. Microbiol. 2018, 9, 1–17; c) S. Panke, [14] a) J. U. Peters, H. Kuhne, H. Dehmlow, U. Grether, A.
B. Witholt, A. Schmid, M. G. Wubbolts, Appl. Environ.
Microbiol. 1998, 64, 2032–2043.
7] a) S. Wu, J. Liu, Z. Li, ACS Catal. 2017, 7, 5225–5233;
b) S. Wu, Y. Zhou, D. Seet, Z. Li, Adv. Synth. Catal.
Conte, D. Hainzl, C. Hertel, N. A. Kratochwil, M.
Otteneder, R. Narquizian, C. G. Panousis, F. Ricklin, S.
Rover, Bioorg. Med. Chem. Lett. 2010, 20, 5426–5430;
b) Y. Haga, S. Mizutani, A. Naya, H. Kishino, H. Iwaasa,
M. Ito, J. Ito, M. Moriya, N. Sato, N. Takenaga, A.
Ishihara, S. Tokita, A. Kanatani, N. Ohtake, Bioorg.
Med. Chem. 2011, 19, 883–893; c) M. Brindisi, S.
Maramai, S. Gemma, S. Brogi, A. Grillo, L. Di Cesar-
e Mannelli, E. Gabellieri, S. Lamponi, S. Saponara, B.
Gorelli, D. Tedesco, T. Bonfiglio, C. Landry, K. M. Jung,
A. Armirotti, L. Luongo, A. Ligresti, F. Piscitelli, C.
Bertucci, M. P. Dehouck, G. Campiani, S. Maione, C.
Ghelardini, A. Pittaluga, D. Piomelli, V. Di Marzo, S.
Butini, J. Med. Chem. 2016, 59, 2612–2632; d) X. D.
Wang, W. Wei, P. F. Wang, L. C. Yi, W. K. Shi, Y. X.
Xie, L. Z. Wu, N. Tang, L. S. Zhu, J. Peng, C. Liu, X. H.
Li, S. Tang, Z. P. Xiao, H. L. Zhu, Bioorg. Med. Chem.
2015, 23, 4860–4865.
[
2
017, 359, 2132–2141.
8] H. Yang, L. Liu, F. Xu, Appl. Microbiol. Biotechnol.
016, 100, 8273–8281.
[
[
2
9] a) J. G. Marblestone, S. C. Edavettal, Y. Lim, P. Lim, X.
Zuo, T. R. Butt, Protein Sci. 2006, 15, 182–189; b) M. P.
Malakhov, M. R. Mattern, O. A. Malakhova, M. Drinker,
S. D. Weeks, T. R. Butt, J. Struct. Funct. Genomics 2004,
5
, 75–86; c) X. Zuo, S. Li, J. Hall, M. R. Mattern, H.
Tran, J. Shoo, R. Tan, S. R. Weiss, T. R. Butt, J. Struct.
Funct. Genomics 2005, 6, 103–111.
[
10] S. N. Patel, M. Sharma, K. Lata, U. Singh, V. Kumar,
R. S. Sangwan, S. P. Singh, Bioresour. Technol. 2016,
2
16, 121–127.
11] a) R. Lauchli, K. S. Rabe, K. Z. Kalbarczyk, A. Tata, T.
Heel, R. Z. Kitto, F. H. Arnold, Angew. Chem. Int. Ed. [15] J. L. E. Brickson, G. N. Grammer, J. Am. Chem. Soc.
Engl. 2013, 52, 5571–5574; b) C. S. Mesquita, R. 1958, 80, 5466–5469.
Oliveira, F. Bento, D. Geraldo, J. V. Rodrigues, J. C. [16] D. Tischler, Microbial Styrene Degradation, 1st ed.,
Marcos, Anal. Biochem. 2014, 458, 69–71; c) M. Horvat, Springer International Publishing, 2015.
[
Adv. Synth. Catal. 2021, 363, 1714–1721
1721
© 2021 Wiley-VCH GmbH