Segmented Flow Processes to Overcome Hurdles of Whole-Cell Biocatalysis in the Presence of Organic Solvents

Article abstract of DOI:10.1002/anie.202015887

In modern process development, it is imperative to consider biocatalysis, and whole-cell catalysts often represent a favored form of such catalysts. However, the application of whole-cell catalysis in typical organic batch two-phase synthesis often struggles due to mass transfer limitations, emulsion formation, tedious work-up and, thus, low yields. Herein, we demonstrate that utilizing segmented flow tools enables the conduction of whole-cell biocatalysis efficiently in biphasic media. Exemplified for three different biotransformations, the power of such segmented flow processes is shown. For example, a 3-fold increase of conversion from 34 % to >99 % and a dramatic simplified work-up leading to a 1.5-fold higher yield from 44 % to 65 % compared to the analogous batch process was achieved in such a flow process.

Full text of DOI:10.1002/anie.202015887

A Journal of the Gesellschaft Deutscher Chemiker
Angewandte
Chemie
International Edition
www.angewandte.org
Accepted Article
Title: Segmented flow processes as an efficient tool to overcome
hurdles of whole-cell biocatalysis in organic solvents
Authors: Harald Gröger, Niklas Adenbar, Alina Nastke, and Jana Löwe
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: Angew. Chem. Int. Ed. 10.1002/anie.202015887
Link to VoR: https://doi.org/10.1002/anie.202015887

10.1002/anie.202015887
Angewandte Chemie International Edition
RESEARCH ARTICLE
Segmented Flow Processes to Overcome Hurdles of Whole-cell
Biocatalysis in Organic Solvents
Niklas Adebar[a], Alina Nastke[a], Jana Löwe[a] and Harald Gröger*[a]
Dedicated to Professor Dr. Shuji Akai on the occasion of his 60th birthday
In modern process development, it is imperative to consider
biocatalysis, and whole-cell catalysts often represent a favored form
of such catalysts. However, the application of whole-cell catalysis in
typical organic batch two-phase synthesis often struggles due to mass
transfer limitations, emulsion formation, tedious work-up and, thus,
low yields. Herein, we demonstrate that utilizing segmented flow tools
enables the conduction of whole-cell biocatalysis efficiently in biphasic
media. Exemplified for three different biotransformations, the power
of such segmented flow processes is shown. For example, a 3-fold
increase of conversion from 34% to >99% and a dramatic simplified
work-up leading to a 1.5-fold higher yield from 44% to 65% compared
to the analogous batch process was achieved in such a flow process.
In particular, constant product quality without human intervention
was pointed out as important for future process development. In
addition, these features are important not only in pharmaceutical
chemistry,6–8 as the bulk chemistry field can also benefit from flow
processing. Increased heat and mass transfer as well as high
safety for toxic or explosive compounds are general benefits.9,10
However, in spite of its tremendous application potential, so far
most continuous processes involving biocatalysis have been
limited to the application of lipases or alcohol dehydrogenases
(ADH).11–13 Due to their extreme resistance against temperature
and organic solvents as well as their commercial availability in
immobilised form, many processes involving lipases have been
developed.14,15 Recently, flow processes have also been
developed for many other biocatalytic systems.16–18 In contrast,
only few examples of continuous flow processes with whole-cell
catalysts have been reported.19 To the best of our knowledge, all
reported systems rely exclusively on the combination of
Introduction
The application of biocatalysis in organic synthesis has improved
substantially over the last decades leading to many successes.1,2
However, some issues remained unsolved for a long time. In
particular, the incompatibility of many biocatalytic systems with
various organic solvents still represents a major limitation in
synthesis and process design.3,4 This challenge to benefit from
organic solvents through an improved substrate availability while
at the same time avoiding its deactivation of the biocatalyst has
been addressed by us by means of flow chemistry techniques.
Herein we report on a system for improved whole-cell biocatalysis
based on liquid-liquid segmented flow, which remarkably
overcomes the known limitations when using analogous liquid-
liquid systems in the batch mode. With respect to such flow
processes it should be added that several years ago the U.S.
Food and Drug Agency (FDA) as well as the European Medicines
immobilised cells in packed bed reactors20–24
,
catalytic
biofilms25,26, wall coated reactors27 or hydrogel-immobilized cells
in segmented flow.28 These concepts are illustrated in Figure 2.
Figure 2. Scheme of the developed organic/aqueous segmented flow system
utilizing whole-cell catalysts compared to classic approaches for continuous
biocatalysis utilising whole-cells.
Agency (EMA) added
a recommendation for continuous
manufacturing to their guidelines.5
Buehler et al. showed that mass-transfer limited reactions using
isolated and purified enzymes can benefit from segmented flow
systems.29,30 This work centers on an ADH from Lactobacillus
brevis for reducing heptanal to the corresponding alcohol.29
Thereafter, only few biphasic biocatalytic flow processes were
reported.31 Most recently, the Wirth group successfully applied
biocatalysis in
a
high performance counter current
chromatography (HPCCC) device, which enhanced the mass
transfer immensely.32
In our work presented here, we investigated the impact of such
segmented flow systems on a variety of whole-cell based systems,
which revealed different advantages. Besides the aldoxime
dehydratase (Oxd)-catalysed synthesis of octane nitrile (4), which
is used as bulk chemical,33-35 the preparation of 12-
oxophytodienoic acid (12-OPDA 5), a complex chiral plant
hormone intermediate,36 using a whole-cell catalyst containing an
allene oxide synthase and cyclase was investigated. Along with
these examples, a cofactor dependent imine reductase (IRED)
Figure 1. General comparison of (continuous) flow versus batch chemistry.
[a]
Niklas Adebar, Alina Nastke, Dr. Jana Löwe, Prof. Dr. Harald Gröger
Chair of Industrial Organic Chemistry and Biotechnology.
Faculty of Chemistry, Bielefeld University
Universitätsstr. 25, 33615 Bielefeld, Germany
E-mail: harald.groeger@uni-bielefeld.de
Supporting information for this article is given via a link at the end of the
document.
1
This article is protected by copyright. All rights reserved.

10.1002/anie.202015887
Angewandte Chemie International Edition
RESEARCH ARTICLE
was used for the synthesis of chiral cyclic imines (see Scheme 1).
For all experiments with these whole-cell catalysts, a segmented
flow system was applied and compared to a biphasic batch
approach. It is noteworthy that in each of these three whole-cell
processes a different current challenge in the field of biocatalysis,
e.g., phase separation of emulsions as well as mass transfer
issues, has been successfully addressed. In detail, different
stirring rates in the batch mode and different flow rates in the flow
mode were investigated and compared for the IRED-catalysed
reduction. The impact of solvent additives in the batch versus flow
mode was investigated for the oxime dehydration using OxdB as
a highly a solvent labile biocatalyst. In addition, the efficiency of
product isolation was studied in flow and compared to the batch
mode exemplified for the particularly challenging example of the
12-OPDA synthesis, since 12-OPDA is an emulsifier.
and cell deactivation have no correlation with each other,
predicting the solvent of choice represents a challenging task.
Thus, for our evaluation on the impact of solvent and stirring rate
on the whole-cell catalyzed imine reduction we chose several
water-immiscible solvents, ranging from highly non-polar solvents
such as cyclohexane to less non-polar solvents such as
methyl tert-butyl ether (MTBE) and 2-methyltetrahydrofuran
(MeTHF). We used preferably less hazardous solvent options,
e.g., cyclohexane rather than hexane or pentane. We also
avoided the use of chlorinated solvents due to their negative
environmental impact and known incompatibility with whole-cell
catalysts. The experiments were conducted at a typical whole-cell
catalyst loading (2 mgdcm mL-1 of dry cell mass per overall volume)
and a stirring rate of 850 rpm. At first, the substrate (40 mM) was
dissolved in the organic solvent, and then added to a KPi buffer
solution containing ᴅ-glucose, NADP+, methanol (4 vol%) and E.
coli BL21(DE3) whole-cells with IRED and GDH therein.
The less polar solvents proved to be the best performing ones
with sufficient conversions when using cyclohexane, isooctane
and methylcyclohexane (see Supporting Information). In contrast,
MTBE, MeTHF and ethyl acetate led to the complete deactivation
of the biocatalyst and, thus, no conversion was observed. For
further experiments we decided to use methylcyclohexane as an
organic solvent, as it provided the highest conversion in this initial
work. When conducting batch experiments with a magnetic stirrer
at a high stirring rate of 1100 rpm for an intensive mixing of the
two phases, a conversion of 36% after 6 h was observed (Figure
3). After optimizing the whole-cell catalyst loading to 10 mgdcm mL-
1, the reaction rate to amine (R)-2 strongly increased reaching a
quantitative conversion within only 2 h reaction time.
Scheme 1. Investigated model processes with whole-cells: A) IRED-catalysed
reduction of 1-methyl-3,4-dihydroisoquinoline (1) to the corresponding amine
(R)-2, B) OxdB-catalysed dehydration of octanal oxime (3) to octanenitrile (4)
and C) Oxidative cyclisation towards 12-OPDA (8) starting from 13-(S)-hydro-
peroxylinolenic acid (6).
Results and Discussion
Imine reduction in a continuous segmented flow mode: The
impact of organic solvents and studies on mass transfer
As a first example we chose the enantioselective reduction of the
C=N-double bond of 1-methyl-3,4-dihydroisoquinoline (1) to the
corresponding amine (R)-2 as a model reaction for the biocatalytic
synthesis of cyclic amines. The reaction was carried out using an
IRED with NADPH as cofactor. The resulting oxidized species of
the cofactor was regenerated in situ using ᴅ-glucose and glucose
dehydrogenase (GDH) under formation of one equivalent of
gluconolactone, which irreversibly opens to gluconic acid and
therefore moves the equilibrium of the reaction towards the
product side. For this process, a recombinant E. coli whole-cell
catalyst containing an IRED from Streptomyces viridochromo-
genes and a GDH from Bacillus subtillis in overexpressed form,
recently developed in our research group,37 was used. Initially, we
investigated the reaction system in a batch mode in order to
obtain a benchmark and to gain insight into some important
reaction parameters, such as choice of organic solvent and
stirring rate. The solvent can have a tremendous influence on the
reaction system, as it affects compound solubility and distribution
as well as cell and/or enzyme deactivation. As substrate solubility
Figure 3. Time course measurements of the conversion to amine (R)-2 at
whole-cell catalyst loadings of 2 mgdcm mL-1 and 10 mgdcm mL-1 (left) and
investigation of different stirring rates (right). Reaction conditions: Biphasic
system with 40 mM of substrate 1 in methylcyclohexane and whole-cell catalyst
in different concentrations, 0.2 mM NADP+, 240 mM glucose, 2% MeOH in KPi
buffer (50 mM, pH 7); stirred at 30 °C and 500 rpm or 1100 rpm.
As mass transfer is often a limiting factor in catalytic reactions, we
became interested in how sensitive this biocatalytic reaction is
with respect to the stirring rate. Thus, we investigated the reaction
using 10 mgdcm mL-1 of whole-cells in batch mode with a lower
stirring rate of 500 rpm (at which the phases remained separated).
These experiments revealed a strong mass transfer limitation of
the reaction as a 6-fold higher conversion was achieved within 1 h
reaction time when increasing the stirring rate (Figure 3).
With these results from batch biotransformations in hand, we
investigated the reaction in a segmented flow setup (Figure 4).
2
This article is protected by copyright. All rights reserved.

10.1002/anie.202015887
Angewandte Chemie International Edition
RESEARCH ARTICLE
We chose identical conditions to the batch experiments to ensure
comparability between the batch and flow processes. Thus, an
organic solution with substrate 1 (40 mM) in methylcyclohexane
as well as an aqueous solution containing ᴅ-glucose, NADP+, E.
coli BL21(DE3) whole-cells with IRED and GDH (10 mgdcm mL-1)
and methanol (2 vol%), were transferred into syringes and
combined with each other through a Y-mixer. This then resulted
in a segmented flow system which was fed through a coil reactor
(PFE, 0.8 mm inner diameter). It should be added that by means
of the utilized (two-channel) syringe pump, disruption of the
whole-cells caused by forces in the pump (high pressure or shear
forces) is highly unlikely. The residence time was set to 0.5 h and
an equilibration time of two residence times was considered to
ensure a steady state. Fractions of the reaction mixture were then
collected in glass vials containing a quenching solution (0.4 mL,
2 M aq. NaOH solution). The quenching method was validated in
advance. The segment sizes strongly depended on the flow rate,
inner tube diameter and on the utilized mixer (Y- or T-piece with
0.5 to 1 mm bore). In our experiments presented here, the
segment sizes were found to be in a range of 0.2 to 0.8 mm.
When conducting the enzymatic reduction of 1 under these (initial)
flow conditions, we were pleased to find already a significant
increase in conversion compared to the batch process (Figure 4).
In detail, the conversion raised up from 25% (after 0.5 h reaction
time in batch, see Figure 3) to 41% under these non-optimized
flow conditions (with a residence time of 0.5 h, see Figure 4).
parameters. In accordance with our expectations, at an elevated
flow rate of 4 mL h-1 the conversion could be further increased
significantly, leading to the formation of amine (R)-2 with 58%
conversion (Figure 4). These findings confirm our hypothesis that
mass transfer-limited biotransformations in biphasic media can
tremendously benefit from conducting them in a segmented flow
mode: even compared to the analogous reaction in batch mode
running at a high stirring rate of 1100 rpm, in the analogous flow
process the conversion has been more than doubled (Figure 4).
Completing the investigations on the imine reduction, finally
preparative experiments with product isolation were carried out in
batch as well as segmented flow mode (for details, see
Supporting Information). In both cases the formed product (R)-2
was isolated from reaction mixtures, which showed full conversion.
In case of the batch reaction, isolation proved to be tedious due
to poor phase separation. Thus, centrifugation and an additional
two-fold extraction was needed. However, even with this time-
consuming procedure, only 75% (>99% purity) of the product (R)-
2 could be isolated. In contrast, in case of the segmented-flow
approach, phase separation was not a critical issue and, thus,
95% yield of the isolated product (R)-2 (95% purity) was obtained
after a simple phase separation.
Aldoxime dehydration in a continuous segmented flow
mode: The impact of a surfactant on process efficiency
Encouraged by this positive impact of the segmented flow
technology on whole-cell processes running in biphasic media,
we became interested in demonstrating the generality of this flow
method by expanding it to further biocatalytic applications. In
addition, we wanted to explore if whole-cell catalysis in flow is also
superior to the batch mode (and enables improved mass trans-
fer) when surfactants are used as additives. As the model reaction
for this study we chose the enzymatic key step of a recently
developed cyanide-free route for nitriles, which consists of an
Oxd-catalyzed dehydration of oximes to nitriles.38–40 When
focusing on aliphatic nitriles, this biocatalytic method goes beyond
the common application range of enzymes in the fields of fine
chemicals and pharmaceuticals, as the resulting n-octanitrile 4
serves as a bulk chemical.41 The aldoxime dehydratase from
Bacillus sp. OxB-1 (OxdB), overexpressed in E. coli BL21(DE3),42
proved to be a suitable whole-cell catalyst for this purpose.41
As a surfactant we used the polysorbate-type non-ionic surfactant
Tween 20 for our studies, as this compound was successfully
applied Buehler et al. in another previous biotransformation.30
Again, we started with initial batch experiments in order to set a
benchmark for the subsequent flow experiments. When
conducting the Oxd-catalyzed dehydration of n-octanal oxime (3)
in a batch mode with whole-cells at a reaction time of 30 min, a
conversion of 13 % to nitrile 4 was observed without the Tween
20 additive. In the presence of Tween 20, the dehydration in the
presence of the Oxd-containing whole-cell proceeds with an
improved conversion of 32 %, thus indicating the beneficial
impact of a surfactant to make the water-immiscible substrate
accessible to the enzyme. Furthermore, this experiment indicates
the mass transfer limitation also for this type of reaction.
Figure 4. Segmented flow reaction of imine
1 to amine (R)-2 in
methylcyclohexane and KPi buffer solution, with an overall flow rate of 2 (left),
respectively 4 mL min-1 (right) corresponding to 0.5 h residence time. Reaction
conditions: Biphasic system with 40 mM of substrate 1 in methylcyclohexane
and 10 mg mL-1 whole-cell catalyst, 0.2 mM NADP+, 240 mM glucose,
2% MeOH in KPi buffer (50 mM, pH 7) at 30 °C and different flow rates.
When conducting the same reaction in a segmented flow mode,
we found again a dramatic improvement of the conversion (Figure
5). Without the additive, an already increased average conversion
to nitrile 4 of 68% was observed, and with Tween 20 as an additive
a nearly complete conversion (average of 96%) was achieved
within a short residence time of 30 min.
Next, we focused on optimizing the flow process. To improve
mass transfer, which turned out as a crucial issue for process
efficiency already in the batch mode experiments (Figure 3), we
increased the flow rate and elongated the reactor length in order
to intensify phase mixing without changing other process
3
This article is protected by copyright. All rights reserved.

10.1002/anie.202015887
Angewandte Chemie International Edition
RESEARCH ARTICLE
4 after simple phase separation and solvent evaporation. As for
the batch reaction, however, the product 4 was only obtained in a
decreased yield of 64% (>95% purity) in spite of a more tedious
work-up consisting of centrifugation, phase separation, two-times
extraction with cyclohexane (and subsequent centrifugation).
Enzymatic cascades in a continuous segmented flow mode:
The impact on conversion and product isolation of the
emulsifying plant hormone intermediate 12-OPDA
Finally, we studied the segmented flow system for the synthesis
12-OPDA (8), which represents a challenging product in terms of
synthetic complexity as well as product isolation. 12-OPDA (8) is
a prostaglandin-like metabolite in plants43 and a precursor of
jasmonic acid.44 At the same time it has emulsifying properties,
which made product isolation in batch syntheses difficult and led
to a tedious work-up and non-satisfactory yields. In addition, the
biosynthetic steps towards this plant hormone intermediate are
challenging since this cascade is based on a highly labile
intermediate.45 In contrast, the biosynthesis of 12-OPDA (8) is
much shorter than any reported chemical total syntheses,43 which
makes this route attractive for synthetic purpose. We recently
reported a bioprocess in batch-mode based on this biosynthesis
(Scheme 2),45 which starts with the formation of hydroperoxide 6
(HPOT) from α-linolenic acid (5) by means of a 13-lipoxygenase.
Subsequently, an allene oxide synthase then catalyzes the
formation of a highly labile epoxide intermediate (7), which then is
transformed into the desired 12-OPDA (8) in the presence of an
allene oxide cyclase.
Figure 5: Upper part: Schematic reaction setup. Lower part: Conversion to
nitrile 4 against the reactors run time (starting from switching on of the system)
of a cyclohexane/buffer segmented flow approach for the OxdB-catalysed
dehydration of octanal oxime (3) with and without addition of Tween 20.
For the synthetic evaluation of this bioprocess in a segmented
flow mode, we focused on the critical final two-step cascade
consisting of the formation of the highly labile allene oxide 7 and
subsequent cyclization to 12-OPDA (8). The first intermediate 13-
HPOT (6) was synthesized beforehand, in accordance with our
previously reported protocol.45 Starting from 13-HPOT (6), we
conducted various batch-type experiments, which served as a
benchmark for the subsequent segmented flow reactions. In the
initial batch experiments, we evaluated the influence of buffer-
solvent systems as well as the amount of biocatalyst and additive.
Details of the batch optimization are given in the Supporting
Information. The reaction time was set to 30 min as for the
subsequent flow process we also chose 30 min as a residence
time. At this reaction time, the highest conversion to 12-OPDA (8)
in batch reactions was found to be 34%. For the optimized batch-
experiment, 20 mgwcm mLaq-1 of whole-cell catalyst, 1 vol% Tween
20 and isooctane as organic solvent were used.
When switching from batch to flow mode, all other parameters
were kept constant to preserve comparability between the
systems (except in case of the flow process a quenching step was
conducted using a 2 M aqueous HCl solution).
It is noteworthy that compared to the corresponding batch
experiments, the flow system performed much better (as shown
in Figure 6). Thus, by means of this flow approach a space-time
yield (STY) of 12.5 g L-1 h-1 was achieved. In addition, compared
to a previous result in the literature28 the biocatalyst’s efficiency
could be more than doubled (0.16 vs. 0.38 mgproduct mgwcm-1).
Figure 6. Batch versus flow approach for the OxdB-catalysed dehydration of
octanal oxime (3) with and without addition of Tween 20 after 30 min reaction/
residence time.
After investigating the conversion of the reaction, we then studied
the product isolation in preparative experiments. Toward this end,
we compared the product isolation and resulting yields from fully
converted reaction mixtures of a batch as well as a segmented
flow reaction (for details, see Supporting Information). In the flow
reaction, a yield of 82% (>94% purity) was obtained for the nitrile
Scheme 2. Synthesis of 12-OPDA (8) starting from 13-HPOT (6).
4
This article is protected by copyright. All rights reserved.

10.1002/anie.202015887
Angewandte Chemie International Edition
RESEARCH ARTICLE
For the synthesis of 12-OPDA (8) in the segmented flow mode the
same reaction conditions as in the batch-reactions were used.
Accordingly, the hydroperoxide substrate 6 was dissolved in
isooctane and the tailor-made whole-cell catalyst consisting of the
allene oxide synthase and allene oxide cyclase was suspended
in a buffer with 1 vol% addition of Tween 20 as a surfactant. When
conducting this synthesis in the segmented flow mode, we were
pleased to find again a dramatic increase of the conversion, which
reached in average >99% along with a high bioprocess stability
over at least a run time of 5 h (Figure 7). Thus, compared to the
batch experiment (34%) the conversion could be almost tripled
when carrying out this process in a segmented flow mode while
keeping all other parameters unchanged, which underlines the
high efficiency of flow processes also for whole-cell
transformations.
mixtures resulting from the biotransformations in the batch mode
(Figure 8A) and segmented flow mode (Figure 8B).
Figure 8. Reaction mixtures of the synthesis of 12-OPDA (8) in batch (A) before
quenching (left, A1) and after quenching (right, A2), as well as segmented flow
(B), directly after the reaction (left, B1) and after sedimentation of the cells (right,
B2). Arrow symbolises the required unit operations leading to an isolated yield
of 8.
In the batch process, a clear phase separation does not occur and
an interphase is formed, which can be attributed to an emulsion
of cells and fatty acid (Figure 8, A1). Furthermore, after quenching
(A2) the organic phase is nearly fully mixed with the interphase,
which makes further work-up rather difficult (Figure 8, A2). Thus,
for this batch-type bioprocess a tedious further downstream-
processing consisting of, e.g., a phase separation centrifugation
turned out to be mandatory.
In contrast, we found that in the flow experiment a clear phase
boundary was observed directly at the end of the reaction without
further treatment of the reaction mixture and even without
centrifugation. Furthermore, after
a short time the cells
sedimented and the organic phase could be easily separated
without further process operations. Afterwards, the organic phase
was decanted, and the solvent was removed. Whereas for the
batch process a yield of only 44% could be achieved, the flow
process with its more simplified work-up led to an increased, still
non-optimized yield of 65%. This improvement of the yield
underlines the benefit of the application of whole-cell catalysis in
a segmented flow system not only for increasing the catalytic
efficiency itself but also to simplify and improve product isolation.
Figure 7. Synthesis of 12-OPDA (8) in a segmented flow system with
20 mgwcm mLaq whole-cells in sodium phosphate buffer (pH 8, 100 mM),
1% Tween 20 and isooctane with 13-HPOT (6) at room temperature and 30 min
residence time.
-1
Note that utilizing whole-cells being stored only at 4 °C and not at
-20 °C is crucial for this bioprocess, as with frozen cells no
conversion was observed, which can be rationalized by
permeabilization of the cell membranes leading to less protection
of the enzymes against the surrounding organic solvents.
Conclusion
Even after significantly improving the catalytic efficiency of this
cascade with whole-cells by means of a segmented flow process,
product isolation as a further challenge remained. The target
molecule 12-OPDA as well as many related fatty acid-derived
compounds are known from the literature47,48 to be difficult to
isolate from aqueous solutions due to their emulsifying properties.
Thus, we became interested in determining if the segmented flow
technology is not only able to dramatically increase the catalytic
efficiency, but also to simplify the downstream processing by
enabling an improved phase separation. For a better comparison
of the isolation efficiency of batch versus flow processes, we
compared the yields of isolated 12-OPDA (8) from the batch-
mode with those from the segmented flow process when starting
in both cases with fully converted reaction mixtures (Figure 8).
Notably, there is a clear optical difference between both reaction
In conclusion, we demonstrated that utilizing whole-cell catalysts
in a segmented flow mode provides an effective and at the same
time easy-to-use methodology for enabling highly efficient
biotransformations, thus overcoming existing limitations
previously known from the widely used batch-processes. In
addition to up to tripled conversions to the desired products,
downstream processing was found to be dramatically improved.
For
a
selected biotransformation with known extremely
challenging work-up due to emulsification and problematic phase
separation, a 1.5-fold higher yield was obtained for the isolated
product with this flow bioprocess compared to the standard batch
process. Furthermore, we showed that whole-cell catalysis can be
simply implemented into a flow process even without specific prior
experience or needed equipment. In addition, large-scale
production of such whole-cell processes can potentially be easier
in flow (via “numbering up” of the optimized process) since these
5
This article is protected by copyright. All rights reserved.

10.1002/anie.202015887
Angewandte Chemie International Edition
RESEARCH ARTICLE
[23] P. Zambelli, L. Tamborini, S. Cazzamalli, A. Pinto, S. Arioli, S. Balzaretti,
F. J. Plou, L. Fernandez-Arrojo, F. Molinari, P. Conti, D. Romano, Food
Chem. 2016, 190, 607-613.
biotransformations have been shown to be very sensitive to mass
transfer and stirring rate in the batch mode, which makes scale-
up challenging. A further advantage of flow processes is that
shear forces can be minimized, which is beneficial for enabling a
high biocatalyst stability. Moreover, constant product quality from
batch to batch can also not be necessarily ensured due to the
sensitivity of the system in contrast to a segmented flow system.
We believe that this whole-cell flow process technology has a
broad reaction scope and can be applied to numerous biocatalytic
systems, thus contributing to a further increase of the number of
whole-cell catalytic applications in organic synthesis.
[24] D. Valikhani, P. L. Srivastava, R. K. Allemann, T. Wirth, ChemCatChem
2020, 12, 2194-2197.
[25] B. Halan, R. Karande, K. Buehler, A. Schmid, J. Flow Chem. 2016, 6, 39-
42.
[26] R. Gross, K. Buehler, A. Schmid, Biotechnol. Bioeng. 2013, 110, 424-
436.
[27] N. Miložič, G. Stojkovič, A. Vogel, D. Bouwes, P. Žnidaršič-Plazl, N.
Biotechnol. 2018, 47, 18-24.
[28] N. Adebar, H. Gröger, Eur. J. Org. Chem. 2020, 6062-6067.
[29] R. Karande, A. Schmid, K. Buehler, Langmuir 2010, 26, 9152-9159.
[30] R. Karande, A. Schmid, K. Buehler, Adv. Synth. Catal. 2011, 353, 2511-
2521.
Acknowledgements
[31] R. Karande, A. Schmid, K. Buehler, Org. Process Res. Dev. 2016, 20,
361-370.
[32] F. Huynh, M. Tailby, A. Finniear, K. Stephens, R. K. Allemann, T. Wirth,
Angew. Chem. 2020, 132, 16632-16637; Angew. Chem. Int. Ed. 2020,
59, 16490-16495.
The authors gratefully acknowledge generous support from the
European Union (EU) within the EU-Research Project ONE-
FLOW (“Catalyst Cascade Reactions in ‘One-Flow’ within a
Compartmentalized, Green-Solvent ‘Digital Synthesis Machinery’
– End-to-End Green Process Design for Pharmaceuticals”), Work
programme: EU proposal 737266. Furthermore, we thank Manuel
Warkentin for technical assistance and Ms. Georgina Willms from
the Ellipsis English Writing Support Centre at Bielefeld University
for editorial support.
[33] N. Laurain, S. Zard, PCT Patent Applic. WO 01/02320 A1, 2001.
[34] K. Abe, T. Hattori, Y. Matsumori, US Patent US 9742033 B2, 2017.
[35] J. Smets, H. R. G. Denuette, A. Pintens, K. Van Aken, F. A. C. Vrielynck,
US Patent US 8609603 B2, 2013.
[36] K. Floková, K. Feussner, C. Herrfurth, O. Miersch, V. Mik, D. Tarkowská,
M. Strnad, I. Feussner, C. Wasternack, O. Novák, Phytochemistry 2016,
122, 230-237.
[37] N. Zumbrägel, D. Wetzl, H. Iding, H. Gröger, Heterocycles 2017, 95,
1261-1271.
Keywords: Biocatalysis • biphasic reaction • flow chemistry •
segmented flow processes • whole-cell catalysis
[38] T. Betke, P. Rommelmann, K. Oike, Y. Asano, H. Gröger, Angew. Chem.
2017, 129, 12533-12538; Angew. Chem. Int. Ed. 2017, 56, 12361-12366.
[39] P. Rommelmann, T. Betke, H. Gröger, Org. Process Res. Dev. 2017, 21,
1521-1527.
[1]
[2]
K. Drauz, H. Gröger, O. May, Enzyme Catalysis in Organic Synthesis,
Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim, Germany, 2012.
U. T. Bornscheuer, G. W. Huisman, R. J. Kazlauskas, S. Lutz, J. C.
Moore, K. Robins, Nature 2012, 485, 185-194.
[40] C. Plass, A. Hinzmann, M. Terhorst, W. Brauer, K. Oike, H. Yavuzer, Y.
Asano, A. J. Vorholt, T. Betke, H. Gröger, ACS Catal. 2019, 9, 5198-
5203.
[41] A. Hinzmann, S. Glinski, M. Worm, H. Gröger, J. Org. Chem. 2019, 84,
4867-4872.
[3]
[4]
[5]
P. Domínguez de María, F. Hollmann, Front. Microbiol. 2015, 6, 1257.
A. M. Klibanov, Nature 2001, 409, 241-246.
[42] Y. Kato, K. Nakamura, H. Sakiyama, S. G. Mayhew, Y. Asano,
Biochemistry 2000, 39, 800-809.
S. L. Lee, T. F. O’Connor, X. Yang, C. N. Cruz, S. Chatterjee, R. D.
Madurawe, C. M. V. Moore, L. X. Yu, J. Woodcock, J. Pharm. Innov.
2015, 10, 191-199.
[43] D. Maynard, H. Gröger, T. Dierks, K. J. Dietz, J. Exp. Bot. 2018, 69, 5341-
5354.
[6]
[7]
M. Baumann, I. R. Baxendale, Beilstein J. Org. Chem. 2015, 11, 1194-
1219.
[44] P. Zerbe, E. W. Weiler, F. Schaller, Phytochemistry 2007, 68, 229-236.
[45] J. Löwe, K.-J. Dietz, H. Gröger, Adv. Science 2020, 7, 1902973.
B. Gutmann, D. Cantillo, C. O. Kappe, Angew. Chem. 2015, 127, 6788-
6832; Angew. Chem. Int. Ed. 2015, 54, 6688-6728.
R. Porta, M. Benaglia, A. Puglisi, Org. Process Res. Dev. 2016, 20, 2-25.
T. Wirth, Microreactors in Organic Chemistry and Catalysis, Wiley-VCH,
2013.
[8]
[9]
[10] M. Movsisyan, E. I. P. Delbeke, J. K. E. T. Berton, C. Battilocchio, S. V.
Ley, C. V. Stevens, Chem. Soc. Rev. 2016, 45, 4892-4928.
[11] L. Hajba, A. Guttman, J. Flow Chem. 2016, 6, 8-12.
[12] E. Laurenti, A. dos Santos Vianna Jr., Biocatalysis 2016, 1, 148-165.
[13] J. Döbber, M. Pohl, S. V. Ley, B. Musio, React. Chem. Eng. 2018, 3, 8-
12.
[14] I. Denčić, S. De Vaan, T. Noël, J. Meuldijk, M. De Croon, V. Hessel, Ind.
Eng. Chem. Res. 2013, 52, 10951-10960.
[15] A. S. de Miranda, M. V. de M. Silva, F. C. Dias, S. P. de Souza, R. A. C.
Leão, R. O. M. A. de Souza, React. Chem. Eng. 2017, 2, 375-381.
[16] X. Tang, R. K. Allemann, T. Wirth, Eur. J. Org. Chem. 2017, 414-418.
[17] A. Brahma, B. Musio, U. Ismayilova, N. Nikbin, S. B. Kamptmann, P.
Siegert, G. E. Jeromin, S. V. Ley, M. Pohl, Synlett 2016, 27, 262-266.
[18] P. De Santis, L.-E. Meyer, S. Kara, React. Chem. Eng. 2020, published
as Advanced Article; DOI 10.1039/d0re00335b.
[19] A. Pinto, M. L. Contente, L. Tamborini, Curr. Opin. Green Sustain. Chem.
2020, 25, 100343.
[20] L. H. Andrade, W. Kroutil, T. F. Jamison, Org. Lett. 2014, 16, 6092-6095.
[21] G. Stojkovič, P. Žnidaršič-Plazl, Process Biochem. 2012, 47, 1102-1107.
[22] M. Polakovič, J. Švitel, M. Bučko, J. Filip, V. Neděla, M. B. Ansorge-
Schumacher, P. Gemeiner, Biotechnol. Lett. 2017, 39, 667-683.
6
This article is protected by copyright. All rights reserved.

10.1002/anie.202015887
Angewandte Chemie International Edition
RESEARCH ARTICLE
Entry for the Table of Contents
Utilizing tools of segmented flow chemistry enables the conduction of whole-cell biocatalysis efficiently in biphasic media. Exemplified
for different biotransformations with an imine reductase, aldoxime dehydratase, allene oxide synthase and cyclase, the power of flow
processes is demonstrated, leading to a 3-fold increase of conversion as well as a dramatic simplified work-up and a 1.5-fold higher
yield compared to the analogous batch processes.
7
This article is protected by copyright. All rights reserved.