Pushing the limits: Cyclodextrin-based intensification of bioreductions

Article abstract of DOI:10.1016/j.jbiotec.2020.11.017

The asymmetric reduction of ketones is a frequently used synthesis route towards chiral alcohols. Amongst available chemo- and biocatalysts the latter stand out in terms of product enantiopurity. Their application is, however, restricted by low reaction output, often rooted in limited enzyme stability under operational conditions. Here, addition of 2-hydroxypropyl-β-cyclodextrin to bioreductions of o-chloroacetophenone enabled product concentrations of up to 29 % w/v at full conversion and 99.97 % e.e. The catalyst was an E. coli strain co-expressing NADH-dependent Candida tenuis xylose reductase and a yeast formate dehydrogenase for coenzyme recycling. Analysis of the lyophilized biocatalyst showed that E. coli cells were leaky with catalytic activity found as free-floating enzymes and associated with the biomass. The biocatalyst was stabilized and activated in the reaction mixture by 2-hydroxypropyl-β-cyclodextrin. Substitution of the wild-type xylose reductase by a D51A mutant further improved bioreductions. In previous optimization strategies, hexane was added as second phase to protect the labile catalyst from adverse effects of hydrophobic substrate and product. The addition of 2-hydroxypropyl-β-cyclodextrin (11 % w/v) instead of hexane (20 % v/v) increased the yield on biocatalyst 6.3-fold. A literature survey suggests that bioreduction enhancement by addition of cyclodextrins is not restricted to specific enzyme classes, catalyst forms or substrates.

Full text of DOI:10.1016/j.jbiotec.2020.11.017

Journal of Biotechnology xxx (xxxx) xxx
Contents lists available at ScienceDirect
Journal of Biotechnology
journal homepage: www.elsevier.com/locate/jbiotec
Pushing the limits: Cyclodextrin-based intensification of bioreductions
,
,
Christian Rappa, Bernd Nidetzkya b, Regina Kratzera *
a Institute of Biotechnology and Biochemical Engineering, Graz University of Technology, NAWI Graz, Petersgasse 12, 8010 Graz, Austria
b Austrian Centre of Industrial Biotechnology (acib), Krenngasse 37, 8010 Graz, Austria
A R T I C L E I N F O
A B S T R A C T
Keywords:
The asymmetric reduction of ketones is a frequently used synthesis route towards chiral alcohols. Amongst
available chemo- and biocatalysts the latter stand out in terms of product enantiopurity. Their application is,
however, restricted by low reaction output, often rooted in limited enzyme stability under operational condi-
tions. Here, addition of 2-hydroxypropyl-β-cyclodextrin to bioreductions of o-chloroacetophenone enabled
product concentrations of up to 29 % w/v at full conversion and 99.97 % e.e. The catalyst was an E. coli strain co-
expressing NADH-dependent Candida tenuis xylose reductase and a yeast formate dehydrogenase for coenzyme
recycling. Analysis of the lyophilized biocatalyst showed that E. coli cells were leaky with catalytic activity found
as free-floating enzymes and associated with the biomass. The biocatalyst was stabilized and activated in the
reaction mixture by 2-hydroxypropyl-β-cyclodextrin. Substitution of the wild-type xylose reductase by a D51A
mutant further improved bioreductions. In previous optimization strategies, hexane was added as second phase
to protect the labile catalyst from adverse effects of hydrophobic substrate and product. The addition of 2-
hydroxypropyl-β-cyclodextrin (11 % w/v) instead of hexane (20 % v/v) increased the yield on biocatalyst 6.3-
fold. A literature survey suggests that bioreduction enhancement by addition of cyclodextrins is not restricted
to specific enzyme classes, catalyst forms or substrates.
Biotransformation additive
2-Hydroxypropyl-β-cyclodextrin
Solvent-free bioreduction
Biocatalyst stabilization
(S)-1-(2-Chlorophenyl)ethanol
1. Introduction
limited to a relatively low number of examples. The majority of them
relate to reaction enhancement by solubilization of sparingly
¨
Cyclodextrins (CDs) are macrocyclic oligosaccharides composed of
six to more than 100 -(1,4) linked D-glucopyranose units. These
water-soluble substrates (Bungaruang et al., 2016; Schmolzer et al.,
2018; Yuan et al., 2019) or by inhibitor blocking in product- or
substrate-inhibited reactions (Bonnet et al., 2010; Gubica et al., 2011).
Improved reaction rates and product concentrations upon addition of
CDs have been also assigned to stabilization and activation of catalytic
enzyme(s) under process conditions (Cao et al., 2017; Wang et al., 2019;
Zhang et al., 2016). Enhanced enantioselectivity by addition of CDs has
been reported for lipases (Bonnet et al., 2010; Mine et al., 2003). Cell
permeabilization was identified as an additional effect of CDs in
whole-cell biocatalysis (He et al., 2016; Giorgi et al., 2019).
α
naturally occurring products of starch degradation most commonly
consist of 6, 7 or 8 glucose units that form a truncated cone with a hy-
drophilic exterior surface and a more hydrophobic interior cavity. CDs
are able to accommodate appropriately sized, hydrophobic molecules in
host∙guest complexes (S∙CD). Their ability to form inclusion com-
pounds is widely exploited to enhance aqueous solubility and stability of
hydrophobic molecules (Del Valle, 2004; Jambhekar and Breen, 2016).
Less known are specific interactions between CDs and proteins: CDs host
aromatic surface residues in host∙guest complexes and ‘cap’ solvent
exposed amino acids. The most often reported effect is protein stabili-
zation by hydrophobic surface reduction and masking of protease-attack
sites (Aachmann et al., 2011, 2004; Serno et al., 2011). Application of
CDs in biocatalysis has hence potential to overcome two frequently
encountered problems: low solubilities (and stabilities) of substrates and
products and deterioration of biocatalysts under process conditions.
However, the application of CDs to promote biotransformations is still
Here, we studied the addition of CDs to bioreductions of o-chlor-
oacetophenone to (S)-1-(2-chlorophenyl)ethanol. The product consti-
tutes a chiral building block for L-clorprenaline, polo-like kinase 1
inhibitors, lysophosphatidic acid receptor agonists and cannabinoid
receptor 2 agonists (Nettekoven et al., 2016; Sato et al., 2012). The used
enzyme was xylose reductase from Candida tenuis (CtXR), an enzyme
well known for its ability to convert a broad spectrum of ketones with
high enantioselectivities (Kratzer et al., 2006; Krump et al., 2014).
* Corresponding author.
E-mail addresses: rapp@tugraz.at (C. Rapp), bernd.nidetzky@tugraz.at (B. Nidetzky), regina.kratzer@tugraz.at (R. Kratzer).
https://doi.org/10.1016/j.jbiotec.2020.11.017
Received 4 August 2020; Received in revised form 16 November 2020; Accepted 16 November 2020
Available online 18 November 2020
0168-1656/© 2020 The Authors.
Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Please cite this article as: Christian Rapp, Journal of Biotechnology, https://doi.org/10.1016/j.jbiotec.2020.11.017

C. Rapp et al.
Journal of Biotechnology xxx (xxxx) xxx
However, CtXR showed moderate activities and low stabilities under
reaction conditions. Therefore, a more stable E. coli whole-cell catalyst
based on CtXR and Candida boidinii formate dehydrogenase (CbFDH) for
NADH recycling was designed. in situ extraction of hydrophobic sub-
strates and products into hydrophobic solvents was used to further
separate the biocatalyst from toxic substrates and products. A catalyst
loading of ~40 gCDW/L E. coli whole-cell catalyst led to full conversion of
maximally 300 mM o-chloroacetophenone (total turnover number 1.1
gproduct/gCDW). The reaction was still limited by catalyst lifetime in the
presence of hydrophobic o-chloroacetophenone and (S)-1-(2-chlor-
complex. Binding constants (K) can be calculated by rearranging the
chemical equilibrium for cyclodextrin and guest molecules where solely
the slope (k) and the intrinsic solubility are needed for determination.
Equilibrium constants were calculated using Eq. 1 for 1:1 complexes and
Eq. 2 for 2:1 complexes (Del Valle, 2004; Loftsson et al., 2004).
k
K1:1
=
(1)
S0∙(1 ꢀ k)
k
K2:1
=
(2)
(S20∙(2 ꢀ k))
¨
ophenyl)ethanol (Schmolzer et al., 2012). In the present study, the
replacement of organic solvents by a CD resulted in substantial
improvement of product concentrations: First, we selected the most
suitable CD for reaction intensification. Then, effects of the CD on sub-
strate solubilization and catalyst performance (activity and stability)
were dissected and bioreductions were optimized aiming at full con-
version at high substrate loading. Finally, a survey amongst literature
reports on bioreduction enhancement by cyclodextrins was carried out.
Quantification of o-chloroacetophenone and (S)-1-(2-chlorophenyl)
¨
ethanol) was accomplished by HPLC as described previously (Schmolzer
et al., 2012).
2.2.1. HBC∙o-chloroacetophenone complex preparation
As the complex stoichiometry for (S)-1-(2-chlorophenyl)ethanol)
with HBC resulted in two product molecules being enclosed at once, a
two-fold substrate excess with regard to HBC was applied for complex-
ation. HBC∙o-chloroacetophenone complex formation was prepared by
liquid complexation: HBC was dissolved in potassium phosphate buffer
followed by stoichiometric substrate addition. The mixture was vortexed
for 10 s. Alternatively, in case of a substrate loading of 2 M, cyclodex-
trins were directly mixed with the stoichiometric amounts of substrate
2. Material and methods
2.1. Chemicals and strains
2-Hydroxypropyl-β-cyclodextrin (HBC, degree of substitution 2–9,
product code OH05393, batch number OH053931501) and γ-cyclo-
dextrin were purchased at Carbosynth (Berkshire, UK), β-cyclodextrin
and o-chloroacetophenone were from Sigma-Aldrich (Vienna, Austria),
1-(2-chlorophenyl)ethanol was from Alfa Aesar (Kandel, Germany),
NADH-disodium salt, NAD+ and Pierce™ BCA Protein Assay Kit were
from Roth (Karlsruhe, Germany), B-PER© cell lysis buffer was from
Thermo Scientific (Vienna, Austria). 2-mL reaction tubes were from
Eppendorf (Vienna, Austria). Other chemicals were from Sigma-
Aldrich/Fluka or Roth, and were of the highest purity available. Two
E. coli Rosetta2 strains co-expressing CtXR variants (GenBank ID wild-
type CtXR AF074484) and CbFDH (GenBank ID AJ011046) were used:
along with 50 μL potassium phosphate buffer (100 mM, pH 6.2).
2.3. Catalyst preparations, bioreductions and product analysis
2.3.1. Catalyst preparations for bioreductions
All catalyst preparations started from lyophilized whole-cell catalyst.
Whole biomass (as suspended and rehydrated). Lyophilized biomass
(40 mg or 80 mg) was rehydrated in 1 mL potassium phosphate buffer
(100 mM, pH 6.2) containing NAD+ (0.5ꢀ 6 mM) and sodium formate
(50 mM excess on [substrate]). Used 2-mL Eppendorf tubes were placed
on a thermomixer for 30 min at 25 ◦C and 1400 rpm.
Pelleted biomass. The pellet was obtained by centrifugation (3220 g,
25 ◦C, 10 min) of the respective reaction mixture (1 mL) subsequently to
the bioreduction.
¨
XR_wild-type strain based on CtXR wild-type (Schmolzer et al., 2012)
and XR_D51A strain based on the CtXR mutant D51A. CtXR D51A was
previously described (Kratzer et al., 2006). Co-transformation of E. coli
Rosetta2 with pET11a carrying the CtXR D51A gene and pRSF-1b car-
rying the CbFDH gene was accomplished by electroporation. Whole-cell
biocatalyst production was previously described by Leis et al. (2017)
and is summarized in the Supplementary data.
Supernatant (cell-free catalyst). Lyophilized and rehydrated whole-
cell catalyst (40 or 80 mg/mL in 1 mL) was centrifuged at 3220 g and
25 ◦C for 10 min.
2.3.2. Bioreductions
2.2. Intrinsic solubilities (S0), complex stoichiometries (S∙CD) and
With whole biomass (as suspended and rehydrated). Rehydrated cells
were combined with HBC and o-chloroacetophenone and filled up with
potassium phosphate buffer (100 mM, pH 6.2) to a total working volume
of 1 mL (biomass concentrations 1.5–40 mgCDW/mL). Reaction mixtures
with 300 mM substrate were incubated for 24 h, mixtures with higher
substrate concentrations for 48 h. Incubation took place on an end-over-
end rotator (30 rpm) at room temperature.
binding constants (KS∙CD
)
Phase solubility studies were carried out according to the method
described by Higuchi and Connors (1965). HBC stock solutions were
prepared in potassium phosphate buffer (100 mM, pH 6.2) and diluted
to final volumes of 1 mL. An average molecular weight of 1453 g/mol
was assumed for HBC (degree of 2-hydroxypropyl substitution 2–9 per
β-CD molecule, Carbosynth information). Guest molecules (o-chlor-
oacetophenone or (S)-1-(2-chlorophenyl)ethanol) were added in excess
amounts to HBC solutions of defined concentrations. Mixtures of HBC
and guest molecules were equilibrated for 24 h on an end-over-end ro-
tator at 30 rpm and room temperature. Subsequent to equilibration,
samples were centrifuged for 10 min at 13.2k rpm. Defined aliquots
were taken from the HBC and guest molecule containing aqueous phase,
diluted into isopropanol/anhydrous ethanol (1:9) and mixed on a vortex
mixer. The decreased reaction medium polarity favoured decom-
plexation. Thus, HPLC measurements allowed determination of the
apparently dissolved guest molecule concentration (Stot), i.e. concen-
tration of the guest molecule’s inherent solubility (S0) plus concentra-
tion of the formed complex (S∙CD). A plot of Stot versus HBC
concentration reveals S0 as intersect. The obtained slope (k) details
complex stoichiometry: k ≤ 1 suggests a 1:1 complex, k > 1 a 2:1
With pelleted biomass. Rebatch bioreductions were prepared with the
pelleted biomass instead of the whole biomass under otherwise identical
conditions.
With the supernatant (cell-free catalyst). Bioreductions with the cell-
free catalyst were prepared with the supernatant instead of the whole
biomass under otherwise identical conditions.
2.3.3. Bioreductions of 1.9 M o-chloroacetophenone and product isolation
Rehydrated whole-cell catalyst (40 mg), o-chloroacetophenone
(1.9 M), sodium formate (1.95 M), HBC (75 mM) and NAD+ (6 mM)
were adjusted to 1 mL with potassium phosphate buffer (100 mM, pH
6.2) and placed on an end-over-end rotator (30 rpm) for 48 h. Subse-
quent to the reaction, the product was extracted thrice by adding 1.5 (v/
v) ethyl acetate followed by centrifugal phase separation at 3220 g,
25 ◦C for 10 min. The product (S)-1-(2-chlorophenyl)ethanol was
concentrated under reduced pressure.
2

C. Rapp et al.
Journal of Biotechnology xxx (xxxx) xxx
2.4. Determination of catalyst activity and effect of HBC on the catalyst
or adsorbed to plastics (tubes, tips). Generally, higher recoveries were
obtained after harsh reaction conditions, i.e. higher substrate concen-
trations and longer reaction times (Leis et al., 2017). CD concentrations
played a minor role in analyte recovery.
2.4.1. Catalyst activity measured prior to bioreductions
Total enzyme activities of the whole biomass (extracellular and intra-
cellular enzymes). The suspension was directly subjected to cell lysis and
protein extraction with B-PER© cell lysis reagent. Enzyme activities
were assayed subsequently to the protein extraction.
3. Results and discussion
Enzyme activities of the supernatant (extracellular enzymes). The
rehydrated biomass was centrifuged and the activities of xylose reduc-
tase and formate dehydrogenase that had leaked out of the cells were
determined in the supernatant.
3.1. Cyclodextrin selection
The initial intention of the present study was to protect the catalyst in
the reaction mixture by complexing hydrophobic substrate and product.
Therefore, we searched for CDs capable of hosting o-chlor-
oacetophenone and (S)-1-(2-chlorophenyl)ethanol. The main determi-
nant for the formation of a tight S∙CD complex is the cavity size of the
Spectrophotometric determination of enzyme activities. Reductase and
dehydrogenase activities were assayed by monitoring the reduction or
oxidation of NAD(H) at 340 nm (Beckman Coulter DU 800® spectro-
photometer). Rates of 0.05ꢀ 0.10 ΔA/min were measured over a time
cyclodextrin. In most applications, CDs with 6, 7 or 8
α
-(1,4) linked D-
period of 5 min. 1
μ
mol of NADH formed or consumed per minute equals
glucopyranose units, referred to as -, β- or γ-CDs, are used. α
α
-CD is able
one unit of enzyme activity. The assay for CtXR wild-type contained
to complex low molecular weight molecules or aliphatic compounds.
β-CD accommodates aromatics and heterocycles and is mostly used to
complex drugs, flavors, cosmetic ingredients and pesticides. Main
drawback of β-CD is its low water-solubility (18 g/L) caused by H-bond
aided self-aggregation in aqueous solution. Derivatization of the 2-, 3-,
and 6-OH groups disturbs crystallization and leads to amorphous
structures favoring gel-formation over precipitation. Etherification of
the glucose C2-OH groups with 2-hydroxypropyl groups increased the
solubility of β-CD to ~45 % (w/v) (Caligur, 2008; Szejtli, 2004). γ-CD
can accommodate even larger molecules. However, γ-CD has weaker
complex-forming ability than β-CDs (Del Valle, 2004). We tested
2-hydroxypropyl β-CD (abbreviated HBC), unmodified β-CD and γ-CD as
additives for bioreductions of 300 mM o-chloroacetophenone. The
addition of substrate to an aqueous solution of 75 mM HBC resulted in a
visible increase in apparent solubility (Stot) of o-chloroacetophenone.
Loading of 4 gCDW/L lyophilized and rehydrated XR_wild-type cells to
mixtures of HBC and 300 mM substrate yielded 89 % product (Fig. 1A).
Addition of substrate to solutions of 75 mM β-CD or γ-CD led to visible
precipitation. Conversions obtained with 4 gCDW/L cells and 300 mM
substrate in the presence of β-CD or γ-CD were 10 and 9 %, respectively,
similar to reaction mixtures without CDs (data with β-CD or γ-CD not
shown). Therefore, we focused on complex-formation of o-chlor-
oacetophenone and (S)-1-(2-chlorophenyl)ethanol with HBC.
700 mM D-xylose and 300 μM NADH; that for CbFDH contained 200 mM
sodium formate and 2 mM NAD+. Assays were performed in 100 mM
potassium phosphate buffer (pH 6.2) and started by adding NADH or
NAD + . Non-specific background oxidation/reduction was considered
by measuring blank activities.
2.4.2. Effect of HBC on the catalyst (whole biomass as suspended and
rehydrated)
Rehydrated cells (40 mg in 500 μL) were added to 2-mL Eppendorf
tubes containing pre-dissolved HBC (0, 38, 75, 150, 310 mM) and were
filled up to 1 mL (potassium phosphate buffer, 100 mM, pH 6.2). The
mixtures were placed on an end-over-end rotator (30 rpm). Samples
(10 μL) were taken over time and directly subjected to spectrophoto-
metric determination of enzyme activities.
2.5. Effect of HBC on enzyme stabilities in the presence of heptane
Experiments were performed with CtXR (wild-type) and CbFDH
double-purified from the whole-cell catalyst by affinity chromatog-
raphy. The protocol followed the standard protocol for CtXR purification
using a Red 31 dye ligand column (6 × 2.6 cm, Mayr et al., 2000)
(Supplementary data). The specific activity for CtXR was calculated to
12.4 U/mg, while partially purified FDH was determined to 0.64 U/mg.
Enzymes were stored at ꢀ 20 ◦C. 0.5 mL of heptane was added to 2-mL
Eppendorf tubes containing CtXR (0.1 mg/mL) or CbFDH (0.1 mg/mL)
and HBC (0, 38, 75, 150, 310 mM) in potassium phosphate buffer
(100 mM, pH 6.2). Samples were placed on an end-over-end rotator at
30 rpm and room temperature for 24 h. Enzyme activities were
measured over time. All samples were prepared in triplicates.
3.2. Complex characterization
Formation of host∙guest complexes leads to an apparent solubility
(Stot) increase of the hydrophobic guest molecule. We observed linear
increases in Stot for o-chloroacetophenone and (S)-1-(2-chlorophenyl)
ethanol at stepwise elevated HBC concentrations (Fig. 2). It has been
previously shown that slopes with k ≤ 1 indicate 1:1 complexes, slopes
with k > 1 imply 2:1 complexes, respectively. In the latter case, two
guest molecules bind to one cyclodextrin molecule (Del Valle, 2004;
Loftsson et al., 2004). The obtained slope of k 0.92 for o-chlor-
oacetophenone suggested hence a 1:1 complex, while a slope of k 1.29
for (S)-1-(2-chlorophenyl)ethanol a 2:1 complex. Intersects reflect the
intrinsic solubilities of guest molecules (S0), which were 5.8 mM for
o-chloroacetophenone and 19.0 mM for (S)-1-(2-chlorophenyl)ethanol.
Binding constants (K[S∙CD]) can be calculated by rearranging the
chemical equilibrium for S∙CD formation where solely the slope (k) and
the intrinsic solubility are needed for determination (equs. 1, 2; Del
Valle, 2004). For o-chloroacetophenone K[S∙CD] was calculated to
2.6. HPLC analysis
Subsequent to the reaction, mixtures were vortexed vigorously. Ali-
quots (100–500 μL) were taken, diluted stepwise in 7:3 ethanol/dH2O
and vortexed. Alternatively, the work up was performed by transferring
the whole reaction mixture (1 mL) into 15 mL Sarstedt tubes followed
by addition of 4 mL isopropanol. The mixture was vortexed and
centrifuged for 10 min, 25 ◦C and 3220 g. The resulting supernatant was
isolated, syringe-filtered and properly diluted using isopropanol.
Quantification of o-chloroacetophenone, (S)-1-(2-chlorophenyl)
ethanol) and (R)-1-(2-chlorophenyl)ethanol) was accomplished by chi-
2.0 mMꢀ 1
, while K[S∙CD] for (S)-1-(2-chlorophenyl)ethanol equals
¨
ral HPLC as described previously (Schmolzer et al., 2012). Conversions
5.0 mMꢀ 1. The higher the constants are, the more effective complexa-
tion is.
were calculated as Eq. (3):
[Product]
Conversion =
(3)
[Product] + [Substrate]
3.3. Bioreductions of 300 mM o-chloroacetophenone
Total recoveries of substrate and product in small-scale bio-
reductions were between 66 and 99 %. Analyte loss suggested that parts
of hydrophobic substrates and products remained in the cell sludge and/
3.3.1. Bioreductions using the whole biomass as catalyst
Previously, 300 mM o-chloroacetophenone represented the highest
3

C. Rapp et al.
Journal of Biotechnology xxx (xxxx) xxx
Fig. 1. (A) Bioreductions of 300 mM o-chloroacetophenone with lyophilized and rehydrated biomass (XR_wild-type strain). Product concentrations obtained at
different cell and HBC concentrations are shown. (NAD+ 0.5 mM, reaction time 24 h. Recoveries were 65 – 88 %, maximal std. dev. ± 8.6, each experiment was done
in triplicates. Data with std. dev. summarized in the Supplementary data.) (B) Time course of the reaction with a catalyst loading of 4 gCDW/L and 75 mM HBC.
3.3.2. Rebatch using the pelleted biomass as catalyst
High catalyst activity over several hours indicated reusability of the
whole-cell catalyst. For this purpose, the catalyst (whole cells, cell
debris) was separated from the first reaction mixtures by centrifugation
and used as catalyst in further, identical batch reactions. However,
conversions in second batches were between 22 and 53 % of the first
batches (batch1, batch2, Fig. 1A, Table 1). Therefore, the supernatant of
rehydrated cells was investigated to quantify catalyst loss by centrifugal
catalyst recovery.
3.3.3. Bioreductions with the supernatant
Conversions with the supernatants of 1.5 and 4 gCDW/L resulted in 48
and 62 % of the conversions obtained with the whole biomass (both in
the presence of 75 mM HBC). The supernatant of 10 and 40 gCDW/L led
to full conversions (≥98 %) (Table 1).
3.3.4. Biocatalyst activity
Activities of the whole biomass (extracellular and intracellular en-
zymes), measured after cell lysis and protein extraction for xylose
reductase and formate dehydrogenase were 1590 U/gCDW and 154 U/
g
CDW, respectively. The rehydrated biomass was centrifuged and the
Fig. 2. Phase solubility relationships. Apparent guest molecule solubility
activities of xylose reductase and formate dehydrogenase that had
○
(Stot=S0+S∙CD) related to HBC concentrations (● o-chloroacetophenone, (S)-
leaked out of the cells were determined to 512 U/gCDW and 72 U/gCDW
,
1-(2-chlorophenyl)ethanol,
linear
fit
for
o-chloroacetophenone
respectively, in the supernatant. Hence, approximately 40 % of the
catalytic activities were free-floating enzymes and not associated with
the biomass.
y = 0.92x+5.84, for (S)-1-(2-chlorophenyl)ethanol y = 1.29x+19.04).
substrate concentration that was fully reduced by 40 gCDW/L XR_wild-
type cells (conversion 97 %). A sufficiently long catalyst lifetime was
ensured by in situ substrate supply and product removal with hexane as
Catalyst activities as measured prior to bioreductions in initial rate
measurements and total turnover numbers obtained in bioreductions of
300 mM o-chloroacetophenone were summarized in Table 1. The su-
pernatants of 1.5 gCDW/L achieved approximately half of the total
turnover number of the whole biomass in the presence of 75 mM HBC
(expressed as g product per gCDW, referring to the used biomass,
Table 1). The supernatants of 10 and 40 gCDW/L fully reduced 300 mM of
substrate with total turnover numbers similar to the whole biomass.
¨
second solvent (Schmolzer et al., 2012). Here, we replaced the second
solvent by HBC and reduced the amount of catalyst. Concentrations of
XR_wild-type cells were varied from 1.5–10 gCDW/L and HBC from
0–310 mM (Fig. 1A). 310 mM HBC corresponded to the amount of HBC
necessary to fully dissolve 300 mM o-chloroacetophenone (Fig. 2). In
each case, addition of HBC increased conversions compared to unmod-
ified mixtures. 10 gCDW/L catalyst and addition of 75 mM HBC resulted
in full conversion (97 %). At higher cyclodextrin concentrations, con-
versions dropped again. Time course analysis of the reaction mixture
with 4 gCDW/L and 75 mM HBC showed high catalyst activity over 24 h.
The conversion increased upon prolongation of the bioreduction from
24 to 48 h further 7 % (Fig. 1B). Conversions with 0, 75- and 150-mM
HBC were investigated for prolonged reaction times. The increase in
reaction time from 24 to 48 h resulted in up to 25 % higher conversions
without HBC and up to 69 % with HBC (summarized in the Supple-
mentary data).
Table
1
shows furthermore conversions of 300 mM o-chlor-
oacetophenone in the presence of the highest amount of HBC used
(310 mM). The first batch (batch 1) was performed with the whole
biomass (resuspended cells comprising extracellular and intracellular
enzymes). The second batch (batch 2) was performed with the pelletized
catalyst from batch 1 i.e. with the enzymes attached to the biomass
(intracellular enzymes). In the present case, catalyst activity is divided
into free-floating activity (extracellular enzymes) and activity attached
to the biomass after centrifugation (intracellular enzymes). Addition of
HBC stabilizes extracellular enzymes seen as high activity of the
4

C. Rapp et al.
Journal of Biotechnology xxx (xxxx) xxx
2011). The CtXR activity (as measured under substrate-saturated con-
ditions) corrected for low coenzyme and o-chloroacetophenone con-
centrations led to a rest activity of ~70 % (using v = vmax[NADH]
[o-chloroacetophenone]/(Ki,
Table 1
Comparison of different catalyst preparations in the whole cell reduction of
300 mM o-chloroacetophenone based on conversions and yields on bio-
catalystsa. (Product concentrations are shown. Experiments with whole biomass
and pelleted biomass were done in triplicates; mean values with std. dev. are
shown. Experiments with supernatant and 40gCDW/L were done in duplicates;
mean values are shown).
NADHKm,o-chloroacetophenone+Km,o-chloroacetophenone[NADH]+Km,
NADH[o-chloroacetophenone]+[NADH][o-chloroacetophenone]);
(Ki,NADH is an apparent dissociation constant for NADH, a value of
Biomass amount
K
m,NADH was estimated). Furthermore, formate exerted non-competitive
Catalyst preparationb
1.5gCDW
/
4gCDW/L
10gCDW/L
40gCDW
/
inhibition on the wild-type CtXR (Ki,formate 182 mM; Neuhauser et al.,
1998). Inclusion of inhibition by formate further decreased rest activ-
ities to ~25 % equal to 80 U/gCDW. This value was in the order of the
initial rate and well below the FDH activity with 154 U/gCDW. The
wild-type reductase displayed a kcat/Km value of ~430 Mꢀ 1sꢀ 1 for
o-chloroacetophenone (no full saturation achieved within limits of sol-
ubility, (Kratzer et al., 2011). An Asp51 to Ala replacement (D51A
mutant) resulted in a ~13-fold higher kcat/Km value (5750 Mꢀ 1sꢀ 1). In
order to improve the bioreduction, the catalytic reductase was replaced
by the D51A mutant (XR_ D51A strain). Additionally, a high catalyst
loading of 40 gCDW/L was applied. The XR_ D51A strain accomplished
97.6 % conversion in a reaction mixture containing 1.5 M substrate,
75 mM HBC and 6 mM NAD+. In comparison, 62 % were obtained with
the original strain under otherwise identical conditions. Comparison of
XR_wild-type and XR_D51A strains is summarized in Table 2.
L
L
Whole biomass as rehydrated
(total activity XR 1590 U/
107 ± 18
268 ± 15
292 ± 3.6
297
1.2
11.2
10.5
4.6
g
CDW, FDH 154 U/gCDW i.e.
extracellular and
intracellular enzymes)
Conversion (mM) (with
75 mM HBC)
Yield on biocatalyst
(gproduct/gCDW
)
Supernatant (centrifugation
after rehydration, XR 510 U/
51
166
6.5
293
4.6
300
1.2
5.3
g
CDW, FDH 72 U/gCDW i.e.
extracellular enzymes)
Conversion (mM) (with
75 mM HBC)
Yield on biocatalyst
(gproduct/gCDW
)
3.4.2. NAD+ and substrate concentrations
Whole biomass as rehydrated
(total activity XR 1590 U/
gCDW, FDH 154 U/gCDW i.e.
extracellular and
66 ± 10
158 ± 10
248 ± 8
n.d.
n.d.
Effects of increased NAD+ and substrate concentrations are sum-
marized in Table 2. NAD+ added at concentrations of 0.5 and 3 mM led
to conversions of 20 and 95 % for 1.5 M substrate with the XR_D51A
strain. The 4.6-fold higher conversion of 1.5 M o-chloroacetophenone by
an increase of NAD+ concentration from 0.5 to 3 mM suggested a ~3ꢀ 5-
fold higher Km,NADH of the D51A CtXR compared to the wild-type CtXR
(0.17 mM, Krump et al., 2014). Higher substrate concentrations of 2 M,
2.5 M and 3.2 M led to conversions of 92 %, 33 % and 3.1 %, respec-
tively, in the presence of 75 mM HBC. We experienced an increase in
reaction mixture viscosity and a decrease in miscibility at higher sub-
strate concentrations. Reaction mixtures started to lose fluidity at sub-
strate concentrations ≥ 2 M.
6.9
6.2
3.9
intracellular enzymes)
Conversion (mM) (with
310 mM HBC, batch 1)
Yield on biocatalyst
(gproduct/gCDW
)
Pelleted biomass from batch 1
(intracellular enzymes)
Conversion (mM) (with
310 mM HBC, batch 2)
Yield on biocatalyst
14 ± 5
69 ± 2
133 ± 6
1.5
2.7
2.1
(gproduct/gCDW
)
n.d. not determined. aNAD+ 0.5 mM, reaction time 24 h. bXR activities
measured in initial rate measurements using xylose.
3.4.3. Concentration of HBC
HBC was varied between 38 and 310 mM in reaction mixtures with
2 M substrate and high loading of the XR_ D51A strain (Fig. 3). Con-
versions in reactions with HBC followed the trend observed for re-
ductions of 300 mM substrate: Positive effects on conversions were
observed up to 75 mM HBC while conversions dropped at higher HBC
concentrations. The expected correlation of HBC and substrate concen-
tration on conversion was not experienced. Only low amounts of sub-
strate and product were complexed at HBC-to-substrate ratios of 0.25 to
supernatant and intracellular enzymes seen as activity in pelletized
biomass from reaction mixtures.
Note that the reductase activities were measured with xylose and not
o-chloroacetophenone for the following reasons: o-Chloroacetophenone
and 1-(2-chlorophenyl)ethanol have deactivating effects on free xylose
¨
reductase (Schmolzer et al., 2012). The low solubility of o-chlor-
oacetophenone impedes full saturation of the biocatalyst. o-Chlor-
oacetophenone forms microemulsions (Xie et al., 2006), the true
concentration of a ≥5.8 mM o-chloroacetophenone solution is hence
flawed (5.8 mM represents the solubility limit S0, Fig. 2). The measured
xylose reductase activity of 1590 U/gCDW on xylose corresponds to
approximately 350 U/gCDW on o-chloroacetophenone.
Table 2
Comparison of XR_wild-type and XR_D51A strains in bioreductions of o-chlor-
oacetophenone. Effects of increasing NAD+ (0.5
– 6 mM) and o-chlor-
oacetophenone (1500 – 3200 mM) concentrations. (Product concentrations with
std. dev. are shown. Each experiment was done in triplicates.).
3.4. Optimization of the bioreduction
Strain
NAD+ (mM)
Substrate (mM)
Product (mM)
3.4.1. Catalyst activity limitation
XR_D51A
XR_D51A
XR_wild-type
XR_D51A
XR_D51A
XR_wild-type
XR_D51A
XR_D51A
XR_D51A
0.5
1.5
1.5
3
1500
1000
1500
1500
1500
1500
2000
2500
3200
308 ± 14
840 ± 95
740 ± 42
1430 ± n.d.
1464 ± 20
930 ± 36
1840 ± 42
833 ± 70
99 ± 45
The time course of a typical bioreduction using the whole biomass
(suspended and rehydrated) as catalyst in the presence of 75 mM HBC
and 0.5 M NAD+ (Fig. 1B) showed an initial rate of 102 U/gCDW at a
substrate concentration of 300 mM. The value was 66 % of the totally
applied CbFDH activity with 154 U/gCDW. The corresponding CtXR ac-
tivity on o-chloroacetophenone was 350 U/gCDW. However, the activity
of the wild-type CtXR was diminished by low o-chloroacetophenone and
NADH concentrations: The S0 of o-chloroacetophenone was 5.8 mM at a
Km,o-chloroacetophenone of 5.1 mM, the one of NADH was maximally
0.5 mM at a Km,NADH of 0.17 mM (Krump et al., 2014; Kratzer et al.,
6
6
6
6
6
n.d. not determined. Catalyst loading was 40 gCDW/L whole biomass as sus-
pended and rehydrated, HBC concentration was 75 mM, reaction time 48 h.
Recoveries 65–88%.
5

C. Rapp et al.
Journal of Biotechnology xxx (xxxx) xxx
complexes (Loftsson and Brewster, 2012). Vice versa, isopropanol was
used to decrease reaction media polarity and thereby increase solubil-
ities of substrate and product. Supplementation of reaction mixtures
containing 150 mM HBC with 50 and 100 mM potassium chloride or
0.5–5 % v/v isopropanol to reactions had no effect on conversions.
Addition of 10 % isopropanol led to 10 % lower conversions. No positive
effects by altered reaction medium polarity were observed compared to
unmodified reaction mixtures (summarized in the Supplementary data.)
3.4.6. Full conversion of 1.9 M o-chloroacetophenone and product isolation
We aimed at the maximally achievable product concentration at full
conversion. Main variables tested were concentrations of substrate
(300 mM to 3.2 M), HBC (0–310 mM), catalyst (1.5–40 gCDW/L) and
NAD+(0.5–6 mM). Highest conversions were always obtained with
75 mM HBC, irrespective of the substrate, catalyst and NAD+ concen-
trations (Figs. 1A, 3). A substrate concentration of 1.9 M was converted
in 98.1 ± 2.3 % yield and 99.97 ± 0.05 % e.e. with 40 gCDW/L XR_D51A
cells (whole biomass as suspended and rehydrated), 75 mM HBC and
6 mM NAD+. Extractive work-up of the product led to an isolated yield
of 84 %. (Note that the high amount of product loss during work-up is
partly due to a small reaction volume of 1 mL).
3.5. HBC effect on the catalyst
Reusability of the pelleted biomass from bioreduction mixtures
containing HBC (Fig. 1A, Table 1) prompted studying catalyst activity
and stability in the presence of HBC. We therefore determined CtXR and
CbFDH activities in the presence of 0–310 mM HBC (Table 3). HBC
addition after cell rehydration and sampling directly from suspensions
of lyophilized cells showed up to 300 % activity for CtXR and 180 %
activity for CbFDH compared to initial values without HBC (no cell lysis
and protein extraction was performed.) The highest activities were ob-
tained after 6 h with 75 mM HBC (activities in samples without HBC also
showed highest values after 6 h). After 24 h, CtXR rest activities in the
presence of HBC were 40–190 % of the starting activity without HBC.
The CtXR rest activity was only 5 % without HBC. CbFDH rest activity
after 24 h was 90 % in the absence of HBC, while in the presence
120–170 % of the starting CbFDH activity without HBC was found.
Fig. 3. Bioreductions of 2 M o-chloroacetophenone with 40 gCDW/L lyophilized
and rehydrated XR_D51A strain. Product concentrations obtained at different
HBC and hexane concentrations are shown. (Catalyst loading 40 gCDW/L, NAD+
6 mM, reaction time 48 h. Recoveries 93-99 %, maximal std. dev. ± 15 %, each
experiment was done in triplicates.) (Data with std. dev. summarized in the
Supplementary data.).
3.5.1. Catalyst stability
0.02 (Figs. 1A, 3). Complexation of catalyst-deactivating substrate and
product with HBC was therefore ruled out as major reason for
bioprocess-intensification. Obviously, HBC exerted another effect on the
reaction.
Isolated and free-floatig CtXR and CbFDH were efficiently stabilized
by HBC against adverse effects of the water-immiscible heptane or
hexane and water-miscible co-solvent isopropanol (section 3.4.4; sum-
marized in the Supplementary data). In previous studies we experienced
deactivation of the free enzymes within minutes and of whole-cell cat-
alysts within maximally 5 h in the presence of substrate and product
3.4.4. Addition of water-immiscible organic solvents
CDs were used in previous studies as ‘shuttles’ to increase substrate/
product transfer from the organic to the aqueous phase and vice versa
(Bricout et al., 2009). Here, we added low amounts of hexane and
heptane to reaction mixtures with 2 M substrate and 0–75 mM HBC
(Fig. 3). No improvement was experienced upon addition of 5–20 %
hexane or heptane to the reaction mixture. Addition of 5 % hexane had
no effect on conversions, higher concentrations of hexane had a negative
effect. Addition of heptane had a similar effect as hexane (summarized
in the Supplementary data.) A study of the effect of heptane on the
isolated, free-floating enzymes showed residual activities of 86 and 50 %
of CtXR wild-type and CbFDH in plain buffer after 24 h (0 mM HBC). In
the presence of 50 % v/v heptane residual activities decreased to 1 and 2
% within 24 h (0 mM HBC) The additional presence of 75 mM HBC
stabilized free reductase and dehydrogenase tremendously. Residual
activities of 98 and 80 % were determined for CtXR wild-type and
CbFDH, respectively (summarized in the Supplementary data).
¨
above their solubility limits (Schmolzer et al., 2012).
3.5.2. Catalyst form and activity
In the present case, cell integrity was severely impaired by biomass
lyophilization: Approximately 40 % of total CtXR and CbFDH activities
were found in the supernatant of lyophilized, rehydrated and centri-
fuged cells. The addition of HBC to suspended catalyst (whole biomass)
led to an increase in enzyme activities over time (measured as initial
rates from cell suspension samples and hence mainly covering free-
floating enzymes, Table 3). The measured activities were up to 170 %
and 108 % of the totally applied CtXR and CbFDH units, respectively
(measured as initial rates after cell lysis and protein extraction, Table 3).
Increasing activities upon addition of HBC might indicate catalyst acti-
vation and further protein extraction from the biomass. Therefore, high
conversions and product concentrations were possible in a scenario of
(1) abolished mass transfer by free-floating enzymes, (2) substantial
stabilization of the free enzymes by HBC and (3) an up to 1.7-fold
activation of CtXR by HBC.
3.4.5. Additives to alter reaction medium polarity
The polarity of the reaction medium was increased with potassium
chloride in order to strengthen hydrophobic interactions in host∙guest
6

C. Rapp et al.
Journal of Biotechnology xxx (xxxx) xxx
Table 3
Effect of HBC on specific enzyme activitiesa. The totally applied activities were 1590 and 154 U/gCDW for CtXRb and CbFDH, respectivelyc (extracellular and intra-
cellular enzymes). Data with std. dev., each experiment was done in triplicates.
b
CtXR activity (WT) (U/gCDW
)
CbFDH activity (U/gCDW)
HBC (mM)
0 h
6 h
24 h
0 h
6 h
24 h
0
874 ± 173
1097 ± 144
1679 ± 98
1778 ± 221
1660 ± 119
959 ± 264
913 ± 121
2668 ± 84
1006 ± 68
1636 ± 324
40 ± 16
92 ± 6
151 ± 12
123 ± 15
166 ± 8
154 ± 9
142 ± 18
84 ± 15
124 ± 26
107 ± 6
157 ± 18
108 ± 3
38
611 ± 72
1620 ± 117
349 ± 22
576 ± 127
116 ± 12
102 ± 5
136 ± 13
64 ± 20
75
150
310
aSamples were taken directly from suspensions of lyophilized cells. bXR activities measured in initial rate measurements using xylose. cThe totally applied activities
were measured after cell lysis and protein extraction.
3.6. Literature survey
Table 4
Effect of additives on chemistry metrics of o-chloroacetophenone bioreductions.
Cyclodextrin-promoted bioreductions have been previously re-
Aqueous, no
20 %
75 mM
75 mM
HBC,
ported. In the present study, an E. coli whole-cell catalyst co-expressing
an aldo-keto reductase superfamily member (CtXR) and a member of the
D-specific dehydrogenases of 2-hydroxy acids (CbFDH) was used.
Enhancement of bioreductions by addition of cyclodextrins is not
restricted to specific enzyme classes. Stabilizing and activating effects on
coupled oxidoreductase systems (free enzymes or whole-cell catalysts)
were observed for all main oxidoreductase superfamilies. Zelinski et al.
(1999) reported on bioreductions of hydrophobic ketones by a coupled
additivea
Hexaneb
HBCc
high
NAD+,
red.
mutantd
Product
10.6
(68)
51.4
(82)
45.6
(97)
292
concentration (g/
(98.1)
L)
(Conversion (%))
Yield on biocatalyst
oxidoreductase system. The used carbonyl reductase was
a
0.27
1.28
4.56
7.30
Zn-dependent alcohol dehydrogenase from Candida parapsilosis and the
coenzyme regenerating enzyme CbFDH. The continuous reduction of
2-acetylnaphthalene was performed in a membrane reactor in the
presence of 50 mM heptakis-(2,6-di-O-methyl)-β-cyclodextrin. Only 1.2
% enzyme deactivation was experienced in 24 h in comparison to 85 %
in control experiments without cyclodextrin. Improvement of bio-
reduction rates upon addition of β-CD have been reported for an E. coli
whole-cell catalyst based on Candida magnoliae carbonyl reductase and
Bacillus megaterium glucose dehydrogenase, both short-chain dehydro-
genase/reductases. Addition of 0.2 and 0.4 mol/mol β-CD to reductions
of 5-hydroxymethylfurfural and ethyl 4-chloro-3-oxobutanoate led to up
to 6-fold reaction rate improvements (He et al., 2018, 2015). Similarly,
addition of 0.06 mol/mol β-CD to bioreductions of ethyl 4-chloro-3-ox-
obutanoate increased reduction rates of an E. coli whole-cell catalyst
co-expressing Sporidiobolus salmonicolor β-carbonylcarboxylic ester
reductase (an aldo-keto reductase) and Bacillus megaterium glucose de-
hydrogenase 1.8-fold (He et al., 2016). It was previously demonstrated
that the interaction of CDs takes place at specific sites on the protein
surface. A hydrophilic ‘cap’ (the CD) is placed on solvent exposed aro-
matic residues. Thereby, aggregation is suppressed if hosted residues are
responsible for aggregation and degradation is prevented in cases where
the point of protease-attack is sterically ‘masked’ by the CD (Aachmann
et al., 2011, 2004). Further effects such as altered stereoselectivities,
reduction of mass transport limitations and changes in enzyme dynamics
were also reported (Fasoli et al., 2006).
(gproduct/gCDW
)
aNAD+ 0.5 mM, substrate 100 mM, 40 gCDW/L XR_wild-type strain (data from
Kratzer et al., 2011).
bNAD+ 0.5 mM, substrate 400 mM, 40 gCDW/L XR_wild-type strain (data from
¨
Schmolzer et al., 2012).
cNAD+ 0.5 mM, substrate 300 mM, 10 gCDW/L XR_wild-type strain.
dNAD+ 6 mM, substrate 1.9 M, 40 gCDW/L XR_D51A strain.
became comparable to the most positive bioreduction examples using
highly active and stable oxidoreductases (Chamouleau et al., 2007; Chen
et al., 2016; Ema et al., 2008; Jakoblinnert et al., 2011; Kataoka et al.,
1999). The previously experienced low catalyst stability under process
conditions was overcome by HBC addition. Thereby the main limiting
factor was shifted from catalyst stability towards reaction mixture
miscibility and viscosity (section 3.4.2). Literature reports suggest that
bioreduction enhancement by addition of cyclodextrins is a universal
method that is neither restricted to specific enzyme classes nor to
catalyst form (whole-cell, free-floating enzymes) (section 3.6. Literature
survey).
Funding
This research did not receive any specific grant from funding
agencies in the public, commercial, or not-for-profit sectors.
Appendix A. Supplementary data
4. Conclusions
Supplementary material related to this article can be found, in the
online version, at doi:https://doi.org/10.1016/j.jbiotec.2020.11.017.
The bioreduction of o-chloroacetophenone is limited by catalyst
instability. Application of hexane (20 %) as in situ substrate reservoir
and product sink with 20 % hexane as co-solvent increased product
References
concentrations to maximally 328 mM (total turnover 1.3 gproduct/gCDW
)
¨
(Schmolzer et al., 2012). Unexpectedly high product concentrations of
up to 1.5 M were obtained with the same catalyst upon replacement of
hexane by HBC. Main effects of HBC were enzyme stabilization and cell
permeabilization (He et al., 2016, 2015). Minor effects were enzyme
activation, substrate and product complexation. Use of a more active
catalytic reductase (D51A CtXR) led to 1.9 M product and total turnover
numbers of 7.3 gproduct/gCDW (Table 4). Through the use of HBC, product
concentrations obtained with the moderately active and labile CtXR
Aachmann, F.L., Otzen, D.E., Larsen, K.L., Wimmer, R., 2004. Structural background of
cyclodextrin-protein interactions. Protein Eng. Des. Sel. 16, 905–912. https://ac
ademic.oup.com/peds/article/16/12/905/1513678.
Aachmann, F.L., Larsen, K.L., Wimmer, R., 2011. Interactions of cyclodextrins with
aromatic amino acids: A basis for protein interactions. J. Incl. Phenom. Macro. 73,
349–357. https://doi.org/10.1007/s10847-011-0071-y.
Bonnet, V., Gervaise, C., Favrelle, A., Sarazin, C., Djedaini-Pilard, F., 2010. Enzymatic
catalysis in presence of cyclodextrins. Curr. Org. Chem. 14, 1323–1336. https://doi.
org/10.2174/138527210791616849.
Bricout, H., Hapiot, F., Ponchel, A., Tilloy, S., Monflier, E., 2009. Chemically modified
cyclodextrins: An attractive class of supramolecular hosts for the development of
7

C. Rapp et al.
Journal of Biotechnology xxx (xxxx) xxx
aqueous biphasic catalytic processes. Sustainability 1, 924–945. https://doi.org/
Kratzer, R., Pukl, M., Egger, S., Vogl, M., Brecker, L.M., Nidetzky, B., 2011. Enzyme
identification and development of a whole-cell biotransformation for asymmetric
reduction of o-chloroacetophenone. Biotechnol. Bioeng. 2011 (108), 797–803.
https://doi.org/10.1002/bit.23002.
10.3390/su1040924.
Bungaruang, L., Gutmann, A., Nidetzky, B., 2016. β-Cyclodextrin improves solubility and
enzymatic C-glucosylation of the flavonoid phloretin. Adv. Synth. Catal. 358,
486–493. https://doi.org/10.1002/adsc.201500838.
Krump, C., Vogl, M., Brecker, L., Nidetzky, B., Kratzer, R., 2014. Acceleration of an aldo-
keto reductase by minimal loop engineering. PEDS 27, 245–248. https://doi.org/
10.1093/protein/gzu021.
Caligur, V., 2008. Cyclodextrins. BioFiles 3, 32.
Cao, H., Jiang, Y., Zhang, H., Nie, K., Lei, M., Deng, L., Wang, F., Tan, T., 2017.
Enhancement of methanol resistance of Yarrowia lipolyticalipase 2 using
β-cyclodextrin as an additive: Insights from experiments and molecular dynamics
simulation. Enzyme Microb. Technol. 96, 157–162. https://doi.org/10.1016/j.
enzmictec.2016.10.007.
Leis, D., Lauß, B., Macher-Ambrosch, R., Pfennig, A., Nidetzky, B., Kratzer, R., 2017.
Integration of bioreduction and product isolation: Highly hydrophobic co-solvents
promotein situ substrate supply and extractive product isolation. J. Biotechnol. 257,
110–117. https://doi.org/10.1016/j.jbiotec.2016.11.021.
¨
Chamouleau, F., Hagedorn, C., May, O., Groger, H., 2007. Biocatalytic aldehyde
Loftsson, T., Brewster, M.E., 2012. Cyclodextrins as functional excipients: Methods to
enhance complexation efficiency. J. Pharm. Sci. 2012 (101), 3019–3032. https://
doi.org/10.1002/jps.23077.
reduction using tailor-made whole-cell catalysts: A novel synthesis of the aroma
chemical cinnamyl alcohol. Flavour Fragr. J. 22, 169–172. https://doi.org/10.1002/
ffj.1768.
´
Loftsson, T., Masson, M., Brewster, M., 2004. Self-association of cyclodextrins and
Chen, K., Li, K., Deng, J., Zhang, B., Lin, J.-P., Wei, D., 2016. Carbonyl reductase
identification and development of whole-cell biotransformation for highly efficient
synthesis of (R)-[3,5-bis(trifluoromethyl)phenyl] ethanol. Microb. Cell Fact. 15, 191.
https://doi.org/10.1186/s12934-016-0585-5.
cyclodextrin complexes. J. Pharm. Sci. 1091–1099. https://doi.org/10.1002/
jps.20047.
Mayr, P., Brüggler, K., Kulbe, K.D., Nidetzky, B., 2000. D-Xylose metabolism by Candida
intermedia: isolation and characterisation of two forms of aldose reductase with
different coenzyme specificities. J. Chromatogr. B 737, 195–202. https://doi.org/
10.1016/S0378-4347(99)00380-1.
Del Valle, E.M.M., 2004. Cyclodextrins and their uses: A review. Process Biochem. 39,
1033–1046. https://doi.org/10.1016/S0032-9592(03)00258-9.
Ema, T., Ide, S., Okita, N., Sakai, T., 2008. Highly efficient chemoenzymatic synthesis of
methyl (R)-o-chloromandelate, a key intermediate for Clopidogrel, via asymmetric
reduction with recombinant Escherichia coli. Adv. Synth. Catal. 350, 2039–2044.
https://doi.org/10.1002/adsc.200800292.
Mine, Y., Fukunaga, K., Itoh, K., 2003. Enhanced enzyme activity and enantioselectivity
of lipases in organic solvents by crown ethers and cyclodextrins. J. Biosci. Bioeng.
95, 441–447. https://doi.org/10.1016/S1389-1723(03)80042-7.
Nettekoven, M., Adam, J.-M., Bendels, S., Bissantz, C., Fingerle, J., Grether, U.,
Grüner, S., Guba, W., Kimbara, A., Ottaviani, G., Püllmann, B., Rogers-Evans, M.,
Fasoli, E., Castillo, B., Santos, A., Silva, E., Ferrer, A., Rosario, E., Griebenow, K.,
Secundo, F., Barletta, G., 2006. Activation of subtilisin Carlsberg in organic solvents
by methyl-β-cyclodextrin: Lyoprotection versus substrate and product-complex
effect. J. Mol. Catal. 42, 20–26. https://www.sciencedirect.com/science/article/pii/
S138111770600169X.
¨
¨
Rover, S., Rothenhausler, B., Schmitt, S., Schuler, F., Schulz-Gasch, T., Ullmer, C.,
2016. Novel triazolopyrimidine-derived cannabinoid receptor 2 agonists as potential
treatment for inflammatory kidney diseases. ChemMedChem 11, 179–189. https://
doi.org/10.1002/cmdc.201500218.
´
Giorgi, V., Menendez, P., García-Carnelli, C., 2019. Microbial transformation of
Neuhauser, W., Steininger, M., Haltrich, D., Kulbe, K.D., Nidetzky, B., 1998. A pH-
controlled fed-batch process can overcome inhibition by formate in NADH-
dependent enzymatic reductions using formate dehydrogenase-catalyzed coenzyme
regeneration. Biotechnol. Bioeng. 60, 277–282. https://doi.org/10.1002/(SICI)
1097-0290(19981105)60:3<277.
cholesterol: reactions and practical aspects-an update. World J. Microbiol.
Biotechnol. 35, 131. https://doi.org/10.1007/s11274-019-2708-8.
´
Gubica, T., Pełka, A., Pałka, K., Temeriusz, A., Kanska, M., 2011. The influence of
cyclomaltooligosaccharides (cyclodextrins) on the enzymatic decomposition of L-
phenylalanine catalyzed by phenylalanine ammonia-lyase. Carbohydr. Res. 346,
1855–1859. https://doi.org/10.1016/j.carres.2011.06.008.
Sato, T., Sugimoto, K., Inoue, A., Okudaira, S., Aoki, J.T.H., 2012. Synthesis and
biological evaluation of optically active Ki16425. Bioorg. Med. Chem. Lett. 22,
4323–4326. https://doi.org/10.1016/j.bmcl.2012.05.012.
He, Y.-C., Zhang, D.-P., Tao, Z.-C., Lu, Y., Ding, Y., Liu, F., Zhu, Z.-Z., Rui, H., Zheng, G.-
W., Zhang, X., 2015. Improved biosynthesis of ethyl (S)-4-chloro-3-
hydroxybutanoate by adding L-glutamine plus glycine instead of NAD+ in
β-cyclodextrin–water system. Bioresour. Technol. 182, 98–102. https://doi.org/
10.1016/j.biortech.2015.01.111.
¨
¨
Schmolzer, K., Madje, K., Nidetzky, B., Kratzer, R., 2012. Bioprocess design guided by in
situ substrate supply and product removal: process intensification for synthesis of (S)-
1-(2chlorophenyl)ethanol. Bioresour. Technol. 108, 216–223. https://doi.org/
10.1016/j.biortech.2012.01.009.
¨
He, Y.-C., Tao, Z.-C., Di, J.-H., Chen, L., Zhang, L.-B., Zhang, D.-P., Chong, G.-G., Liu, F.,
Ding, Y., Jiang, C.-X., Ma, C.-L., 2016. Effective asymmetric bioreduction of ethyl 4-
chloro-3-oxobutanoate to ethyl (R)-4-chloro-3-hydroxybutanoate by recombinant
E. coli CCZU-A13 in [Bmim]PF6-hydrolyzate media. Bioresour. Technol. 214,
411–418. https://doi.org/10.1016/j.biortech.2016.04.134.
Schmolzer, K., Lemmerer, M., Nidetzky, B., 2018. Glycosyltransferase cascades made fit
for chemical production: Integrated biocatalytic process for the natural polyphenol
C-glucoside nothofagin. Biotechnol. Bioeng. 115, 545–556. https://doi.org/
10.1002/bit.26491.
Serno, T., Geidobler, R., Winter, G., 2011. Protein stabilization by cyclodextrins in the
liquid and dried state. Adv. Drug Deliv. Rev. 63, 1086–1106. https://doi.org/
10.1016/j.addr.2011.08.003.
He, Y.-C., Jiang, C.-X., Chong, G.-G., Di, J.-H., Ma, C.-L., 2018. Biological synthesis of
2,5-bis(hydroxymethyl)furan from biomass-derived 5-hydroxymethylfurfural
byE. coli CCZU-K14 whole cells. Bioresour. Technol. 247, 1215–1220. https://doi.
org/10.1016/j.biortech.2017.09.071.
Szejtli, J., 2004. Past, present and future of cyclodextrin research. Pure Appl. Chem. 76,
1825–1845. https://doi.org/10.1351/pac200476101825.
Higuchi, T., Connors, K.A., 1965. Phase solubility techniques, advances in analytical
chemistry and instrumentation. Interscience 4, 117–212.
Wang, P., Ke, Z., Yi, J., Liu, X., Hao, L., Kang, Q., Lu, J., 2019. Effects of β-cyclodextrin on
the enzymatic hydrolysis of hemp seed oil by lipase Candida sp. 99–125. Ind. Crops
Prod. 129, 688–693. https://doi.org/10.1016/j.indcrop.2018.11.046.
Xie, Q., Wu, J., Xu, G., Yang, L., 2006. Asymmetric reduction of o-chloroacetophenone
with Candida pseudotropicalis 104. Biotechnol. Prog. 2006 (22), 1301–1304. https://
doi.org/10.1021/bp0600583.
Jakoblinnert, A., Mladenov, R., Paul, A., Sibilla, F., Schwaneberg, U., Ansorge-
Schumacher, M.B., Domínguez de María, P., 2011. Asymmetric reduction of ketones
with recombinant E. Coli whole cells in neat substrates. Chem. Commun. 47,
12230–12232. https://doi.org/10.1039/C1CC14097C.
Jambhekar, S.S., Breen, P., 2016. Cyclodextrins in pharmaceutical formulations I:
structure and physicochemical properties, formation of complexes, and types of
complex. Drug Discov. Today 21, 356–362. https://doi.org/10.1016/j.
drudis.2015.11.017.
Yuan, X., Liu, G., Zhang, P., Xu, W., Tang, K., 2019. Lipase-catalyzed production of (S)-
Carprofen enhanced by hydroxyethyl-β-cyclodextrins: experiment and optimization.
Org. Process Res. Dev. 23, 891–899. https://doi.org/10.1021/acs.oprd.9b00009.
Zelinski, T., Liese, A., Wandrey, C., Kula, R.-M., 1999. Asymmetric reductions in aqueous
media: enzymatic synthesis in cyclodextrin containing buffers. Tetrahedron
Asymmetry 10, 1681–1687. https://doi.org/10.1016/S0957-4166(99)00137-8.
Zhang, X., Nie, K., Zheng, Y., Wang, F., Deng, L., Tan, T., 2016. Lipase Candida sp. 99-125
coupled with β-cyclodextrin as additive synthesized the human milk fat substitutes.
J. Mol. Catal. 125, 1–5. https://doi.org/10.1016/j.molcatb.2015.12.009.
Kataoka, M., Yamamoto, K., Kawabata, H., Wada, M., Kita, K., Yanase, H., Shimizu, S.,
1999. Stereoselective reduction of ethyl 4-chloro-3-oxobutanoate by Escherichia coli
transformant cells coexpressing the aldehyde reductase and glucose dehydrogenase
genes. Appl. Microbiol. Biotechnol. 51, 486–490. https://doi.org/10.1007/
s002530051421.
Kratzer, R., Leitgeb, S., Wilson, D.K., Nidetzky, B., 2006. Probing the substrate binding
site of Candida tenuis xylose reductase (AKR2B5) with site-directed mutagenesis.
Biochem. J. 393, 51–58. https://doi.org/10.1042/BJ20050831.
8