Improved Enantioselectivity of Subtilisin Carlsberg towards Secondary Alcohols by Protein Engineering

Article abstract of DOI:10.1002/cbic.201700408

Generally, the catalytic activity of subtilisin Carlsberg (SC) for transacylation reactions with secondary alcohols in organic solvent is low. Enzyme immobilization and protein engineering was performed to improve the enantioselectivity of SC towards secondary alcohols. Possible amino-acid residues for mutagenesis were found by combining available literature data with molecular modeling. SC variants were created by site-directed mutagenesis and were evaluated for a model transacylation reaction containing 1-phenylethanol in THF. Variants showing high E values (>100) were found. However, the conversions were still low. A second mutation was made, and both the E values and conversions were increased. Relative to that shown by the wild type, the most successful variant, G165L/M221F, showed increased conversion (up to 36 %), enantioselectivity (E values up to 400), substrate scope, and stability in THF.

Full text of DOI:10.1002/cbic.201700408

Accepted Article
Title: Improved Enantioselectivity of Subtilisin Carlsberg Towards
Secondary Alcohols by Protein Engineering
Authors: Robin Dorau, Tamás Görbe, and Maria Svedendahl Humble
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: ChemBioChem 10.1002/cbic.201700408
Link to VoR: http://dx.doi.org/10.1002/cbic.201700408
A Journal of
www.chembiochem.org

10.1002/cbic.201700408
ChemBioChem
FULL PAPER
Improved Enantioselectivity of Subtilisin Carlsberg Towards
Secondary Alcohols by Protein Engineering
Robin Dorau[a],[b], Tamás Görbe[c] and Maria Svedendahl Humble*[a],[d]
Abstract: Generally, the catalytic activity of subtilisin Carlsberg
(SC) for transacylation reactions using secondary alcohols in
organic solvent is low. Enzyme immobilization and protein
engineering was performed to improve the enantioselectivity of
SC towards secondary alcohols. Possible amino acid residues
for mutagenesis were found by combining available literature
data with molecular modeling. SC variants were created by site-
directed mutagenesis and evaluated for a model transacylation
reaction containing 1-phenylethanol in THF. Variants showing
high E-values (>100) were found. However, the conversions
were still low. A second mutation was made and both E-values
and conversions were increased. The most successful variant,
G165L/M221F, showed increased conversion (up to 36%),
enantioselectivity (E-values up to 400), substrate scope and
stability in THF, compared to wild-type.
obtained by an enantioselective transacylation reaction using an
acyl donor in organic solvent. Then, one alcohol enantiomer is
acylated by an acyl donor and the other alcohol enantiomer is
enriched. The enantioselectivity of the transacylation reaction
can be controlled by using either (R)- or (S)-selective enzymes,
such as (R)-selective lipases or (S)-selective serine
proteases.[4,5] For (R)-selective KR reactions, Candida antarctica
lipase B (CalB, Novozyme 435) is commonly applied. For (S)-
selective KR reactions, subtilisin Carlsberg (SC) and CalB
W104A have been used. The later enzyme, CalB W104A, was
created by rational design to show reversed enantioselectivity
towards secondary alcohols.[6,7] CalB W104A has been applied
to resolve various secondary alcohols.[6-9]
SC is
a serine protease naturally secreted from Bacillus
licheniformis.[10] SC is industrially applied as an ingredient in
various detergent products.[11] For synthetic organic chemistry
applications, SC has been used to enrich the (S)-enantiomer of
a racemic alcohol mixture by dynamic kinetic resolution (DKR)
systems combining SC and a ruthenium complex in organic
solvent.[12-16]
Introduction
Nowadays, the industrial demand of stereospecific syntheses to
obtain enantiomerically enriched products using sustainable and
environmentally compatible methods is increasing. The ability of
enzymes to catalyze selective transformations under mild
conditions makes them highly attractive for green chemistry as
industrial biocatalysts.[1,2]
SC and CalB are serine hydrolases, which react by the similar
reaction mechanism and contain an oxyanion hole as well as a
catalytic triad (Asp-His-Ser). Despite that both enzymes have a
different overall fold, their active sites are mirror images.[17] Both
enzymes have limited space for one enantiomer of the
secondary alcohol, but the mirror-imaged orientation of their
catalytic sites favors the opposite enantiomer according to
Kazlauskas rule.[18-19] Both lipases and SC react faster as the
medium (M) sized group of a secondary alcohol is positioned
into a small (stereoselectivity or S1´) pocket and the large (L)
group is directed towards the solvent (Figure S1). In SC, the
methyl (M) group of 1-phenylethanol is positioned into the
S1´pocket and the phenyl (L) group is directed towards the
solvent (Figure 1). Generally, SC shows low enantioselectivity in
KR of secondary alcohols and low organic solvent tolerance,
compared to CalB.
By enzyme-catalyzed kinetic resolution (KR), enantiomerically
enriched alcohols of high purity can be prepared from alcohol
racemates on industrial scales. Enantiomerically enriched
alcohols are industrially important synthetic building blocks for
synthesis of various pharmaceuticals.[3,4] Normally, the KR is
[a]
[b]
R. Dorau, Dr. M. Svedendahl Humble
Division of Industrial Biotechnology
School of Biotechnology
KTH Royal Institute of Technology
AlbaNova University Center, 106 91 Stockholm, Sweden
R. Dorau
The activity and enantioselectivity of an enzyme in organic
solvent has been shown to depend on its structural flexibility and
thus on the preparation technique and the reaction conditions
(solvent and temperature).[20-22] Numerous preparatory methods
have been shown to affect the enantioselectivity and activity of
SC to various extents.[13, 23-27] However, the main bottleneck for
efficient application of such stabilized SC preparations is the low
stability in organic solvent.[28,29] Various methods (additives,
CLEA, surfactants and immobilization) have been explored to
improve its activity in organic solvent and at elevated
Division of Microbiology and Production
National Food Institute
Technical University of Denmark
Søltofts Plads, 2800 Kgs. Lyngby, Denmark
T. Görbe
Department of Organic Chemistry
Stockholm University
Arrhenius Laboratory, 106 91 Stockholm, Sweden
Dr. M. Svedendahl Humble
[c]
[d]
Pharem Biotech AB
Biovation Park, Forskargatan 20 J, 151 36 Stockholm, Sweden
Supporting information for this article is given via a link at the end of
the document.
This article is protected by copyright. All rights reserved.

10.1002/cbic.201700408
ChemBioChem
FULL PAPER
temperatures.[13,16,30-35] Previously, pre-treatments of SC with KCl
salt or surfactants (Brij 56 and octyl-β-D-glycopyranoside) prior
lyophilization have improved the enzyme performance in terms
of activity, operational stability and enantioselectivity in organic
solvent.[13,36] It has previously been proposed that the enzyme
treatment method applied prior to immobilization is more
important for the enzyme activity and enantioselectivity, than the
type of immobilization technique.[30]
(40) compared to (a). Importantly, the use of free enzyme in
water solution, immobilized enzyme on EziG 2 or Accurel
MP1000 without surfactant or KCl treatment, did not result in any
transacylation activity for the model reaction. Despite this, all
enzyme preparations showed activity in a proteolytic activity
assay using a synthetic peptide, N-Succinyl-Ala-Ala-Pro-Phe p-
nitroanilide.
Since only moderate E-values (40) were obtained using the pre-
treatment methods (surfactants and KCl) and immobilization
techniques (EziG 2 and Accurel MP1000), protein engineering
guided by molecular modeling based on literature data was
applied.
As an alternative method to the previously mentioned chemical-
or reaction condition engineering, protein engineering can be
applied to genetically modify the enzyme. SC belongs to the
protein family of subtilisins, which has numerous members
sharing high sequence identities, structure similarity and
catalytic properties.[37,38] The subtilisins served as model proteins
for the development of the early protein engineering strategies.
As a result, over 50% of the amino acid residues in SC have
been subjected to mutagenesis.[37] To date, there are few reports
on protein engineering of SC for improved transacylation activity
in organic solvent. However, a variant of Subtilisin Bacillus
lentus (SBL) (M222C) showing both increased stereoselectivity
and conversion for the transacylation of 1-phenylethanol using
N-acetyl-L-phenylalanine vinyl ester as acyl donor in acetonitrile,
was created by site-directed mutagenesis.[39]
By molecular modeling, the (S)- and (R)-configuration of the
tetrahedral intermediate of the model transacylation reaction
(Scheme 1) were built into a crystal structure of SC (PDB ID:
1YU6[41]. In SC, the methyl group of (S)-1-phenylethanol should
be directed into the S1´ pocket and the phenyl group should be
directed towards the solvent (Figure
1
and S1B).[18-19,42]
Theoretically, the large phenyl group should not fit well into the
S1´ pocket and therefore the (R)-1-phenylethanol should react
slower than the (S)-1-phenylethanol.
The size of the S1´ pocket is restricted by a conserved
methionine (M221). The numbering of the amino acid residues in
1YU6[41] is shifted one residue, starting from amino acid residue
56, compared to the mature protein sequence of SC
(AGN35600.1)[40]. Therefore, the amino acid residue M222 in
Due to the current lack of efficient (S)-selective enzymes for
enantioselective transacylation of racemic secondary alcohols in
organic solvent, this work aims at improving the
enantioselectivity of SC by protein engineering.
1YU6[41] corresponds to residue M221 in SC AGN35600.1[40]
.
The same type of restriction of the alcohol binding pocket is
shown in lipases. For example, CalB has a tryptophan (W104) at
the similar position. The size of the side chain of the amino acid
residue at this position has been shown to affect both the
Results and Discussion
Initially, the enantioselectivity of SC (AGN35600.1)[40] towards 1-
phenylethanol ((rac)-1a) in THF was explored using a model
transacylation reaction (Scheme 1). For this, the enzyme was
applied in (a) free form, (b) free and treated with surfactants (Brij
56 and octyl-β-D-glycopyranoside), (c) immobilized on EziG 2
and treated with surfactants and (d) immobilized on Accurel
MP1000 and treated with KCl. All enzyme preparations were
lyophilized 24 h prior use and all reactions were performed
under argon atmosphere.
enzymes
enantiomeric
preference
and
the
enantioselectivity.[6,7,43] Previous mutagenesis studies on
subtilisin Bacillus amyloliquefaciens (SBA) have shown that the
methionine sulfur atom next to the catalytic serine (M222 in
SBA) is prone to oxidize. The formation of a sulfoxide at this
position results in enzyme inactivation, probably due to
destabilization of the transition state intermediate.[44-46] This may
be the reason why an inert atmosphere (addition of argon) is
required to obtain conversion for the model reaction using SC.
Estell et al. created 19 variants on position M222 in SBA and
evaluated the oxygen resistance. Two variants, M222A and
M222S, showed increased resistance towards oxidation.[44] Also,
the mutation M222F in subtilisin BPN´ resulted increased
peptide ligation activity in organic solvent.[47] Based on these
literature data, mutagenesis of SC on the corresponding position
(M221) might result in a more stable enzyme. The initial
molecular modeling indicated that an exchange of M221 by
other amino acid residues might result in a reshaped S1´ pocket.
O
SC wt or variants
Na2CO3
OH
OH
O
O
+
+
THF
O
(rac)-1a
(S)-3a
(R)-1a
2
Scheme 1. The model transacylation reaction used to evaluate the
enantioselectivity of SC wild-type (wt) and variants.
The use of (a) free SC in solution for the model reaction resulted
in an E-value close to 4. Previously, E-values of 2-3 have been
published.[13] The use of treated enzyme preparations, in free or
in immobilized form (b-d), resulted in a tenfold increased E-value
Initially,
five
enzyme
variants
on
position
M221
(M221A/C/S/F/W) were created by site-directed mutagenesis. All
created variants showed proteolytic activity both on skim milk
This article is protected by copyright. All rights reserved.

10.1002/cbic.201700408
ChemBioChem
FULL PAPER
agar plates and in a proteolytic activity assay. However, the
proteolytic activity of M221F and M221W was low. It is important
that the enzyme variants retain the natural proteolytic activity.
Otherwise, the pro-peptide cannot be cleaved off after secretion
and the enzyme will remain inactive.[48,49]
G165L/M221C
G165L/M221F
G165L/M221S
18
25
6
98
98
99
16
20
6
>100
>100
>100
[a] The enantiomeric ratio (E) was calculated using Equation 2.
Two variants, M221C and M221F, showed higher E-values for
the model transacylation reaction (Scheme 1) compared to wild-
type (Table 1). Variant M221C showed a synthetically useful E-
value (>100), followed by M221F (E = 67). However, the
conversions were still unsatisfying (5%) after 48 h. Variant
M221S showed similar E-value as the wild-type. No E-values
could be determined for variants M221A and M221W due to the
low conversion (<1%) after 48 h. Possibly, more accurate E-
values could have been determined if higher conversions were
obtained.
[b] The enzymes were immobilized on EziG 2 material and co-lyophilized
with surfactants.
[c] Reaction conditions (1 mL): Immobilized SC on EziG 2 (10 mg), Na2CO3
(53 mg), vinyl butyrate 2 (75 mM), 1-phenylethanol (rac)-1a (50 M),
dodecane (100 mM) in THF at 25 °C. The reactions were flushed with
argon and analyzed by chiral GC.
[d] Not determined due to low conversion (<1%).
The substrate specificity of the M221C and G165L/M221C/F/S
variants for (S)- and (R)-1-phenylethanol was determined and
compared to the wild-type enzyme (Table 2). In order to
increase the E-value towards (S)-1-phenylethanol ((S)-1a), the
substrate specificity (kcat/KM) towards the (S)-1a should be
increased, while the corresponding value for the (R)-1-
phenylethanol (R)-1a should be decreased. Both SC M221C and
G165L/M221F showed increased substrate specificity towards
(S)-1a and reduced substrate specificity for (R)-1a. Both SC
G165L/M221C and SC G165L/M221S showed higher substrate
specificities for both (S)- and (R)-1a. The G165L/M221F variant
showed the highest E-value of 160.
With the prospect to improve the conversion of the model
reaction, while retaining the high enantioselectivity shown by the
M221 variants, further mutation points were searched for. At the
bottom of the S1 pocket, a glycine (G166 in 1YU6[41] and G165 in
SC AGN35600.1[40]) is found. Previous mutagenesis studies on
the corresponding position in SBA affected both the size of the
S1 pocket as well as the natural peptide amino acid side chain
preference.[50] In addition, variants of SC on this position
(G165L/I/Y) were previously found in a gene library aiming on
improved perhydrolysis activity by increasing the substrate
specificity towards methyl esters.[51-52]
Table 2. Kinetic parameters[a] (kcat/KM) determined for immobilized[b] SC wild-
type and variants (M221C, G165L/M221C, G165L/M221S and G165L/M221F)
for transacylation[c] of (S)- and (R)-1a, separately, in THF at 25 °C.
Again, the literature data and molecular modeling was used as
the base for the design of one variant on position G165. The SC
G165L variant was created by site-directed mutagenesis and
was after enzyme preparations tested for the model
transacylation reaction (Table 1). SC G165L showed a slightly
higher E-value (46) and conversion (8%) compared to wild-type
enzyme.
(kcat/KM)R
(kcat/KM)S
(KM)R
[M]
(KM)S
[M]
SC variant
E (S)
[s-1]
[s-1]
Wild-type
0.064
0.042
0.11
2.5
5.9
15
39
140
140
160
57
0.41
0.25
0.38
0.51
0.64
1.4
M221C
2.1
3.4
G165L/M221C
G165L/M221F
G165L/M221S
Then, the G165L mutation was combined with the M221C/F/S
variants and further evaluated for the model reaction (Table 1).
The double variants showed good E-values (>100). Variant
G165L/M221F showed the highest conversion (20%) after 48 h.
0.047
0.22
7.7
13
>6.8
6.8
Table 1. Enantioselectivities[a] (E-values) of immobilized[b] SC wild-type and
variants for transacylation of (rac)-1a and vinyl butyrate in THF measured
after 48 h.[c]
[a] Enzyme kinetics was performed under pseudo-one-substrate conditions.
The concentration of vinyl butyrate 2 was kept constant at 200 mM. Separate
reactions were performed for (S) and (R)-1a.
[b] The enzymes were immobilized on EziG 2 material and co-lyophilized with
surfactants.
[c] Reaction conditions (1 mL): Immobilized SC on EziG 2 (10 mg), Na2CO3
(53 mg), vinyl butyrate 2 (200 mM), 1-phenylethanol (rac)-1a (0.1-1.0 M),
dodecane (100 mM) in THF at 25 °C. The reactions were flushed with argon
and monitored by chiral GC.
SC variant
eeS [%]
eeP [%]
Conv. [%]
E (S)
Wild-type
M221A
M221C
M221F
M221S
M221W
G165L
3
0
6
5
3
0
8
95
87
99
97
98
86
95
3
39
0.2
5
n.d.[d]
>100
67
The enantioselectivity of SC G165L/M221F was explored for
different secondary alcohols in comparison to SC wild-type
(Table 3). Within the functionalization of the para-position (R1),
the electronic effects of the substituents were explored. The 1-
(4-methylphenyl)ethan-1-ol (rac)-1b worked well with SC,
G165L/M221F. However, the enantioselectivity was a bit lower
43 (Table 3, entry b). The electron donating 1-(4-
5
3
41
0.1
8
n.d.[c]
46
This article is protected by copyright. All rights reserved.

10.1002/cbic.201700408
ChemBioChem
FULL PAPER
methoxyphenyl)ethan-1-ol (rac)-1d gave an E-value of 139,
which is a huge increase compared to the wild type enzyme,
which provides very low activity towards this specific substrate
(Table 3, entry d). SC G165L/M221F showed a high E-value
f
-H
-H
-H
-C3H7
-C4H9
-C7H11
0, 0%
220, 8%
210, 6%
210, 7%
g
h
0, 0%
0, 0%
(110)
for
1-(4-chlorophenyl)ethan-1-ol
(rac)-1c.
The
[a] The enzymes were immobilized on Accurel MP1000 in phosphate buffer
(0.1 M, pH 7.8) containing 1 M KCl.
[b] Reaction conditions (1 mL): Immobilized SC on Accurel MP1000 (10 mg),
Na2CO3 (53 mg), vinyl butyrate 2 (750 mM), secondary alcohol (rac)-1a-h (500
M), dodecane (100 mM) in THF at 23 °C.[13] The reactions were flushed with
argon and monitored by chiral GC.
corresponding value for the SC wild-type for the same substrate
was 41 (Table 3, entry c). The aliphatic part of the model
substrate (rac)-1a was also substituted with longer alkyl chains
(R2-groups) (Table 3): 1-phenylpropan-1-ol (rac)-1e (entry e), 1-
phenylbutan-1-ol (rac)-1f (entry f), 1-phenylpentan-1-ol (rac)-1g
(entry g) and 1-phenylheptan-1-ol (rac)-1h (entry h). SC
G165L/M221F showed high E-values (>200) with great
enantiomeric excess (99%) for all substrates with longer alkyl
chains. However, the conversions were still low.
Finally, the enantioselectivity of SC wild-type and G165L/M221F
for the model reaction was explored at different reaction
temperatures (Table 4). In contrast to the wild-type enzyme, the
enantioselectivity of G165L/M221F variant was shown to be
significantly affected by the temperature. The highest E-value
(400) was determined for the G165L/M221F variant at 7 °C. The
corresponding value for SC wild-type at 7 °C was 40. The E-
value shown by the G165L/M221F variant at 23 °C was lower
than previously measured in the kinetic study using the two
alcohol enantiomers in separate reaction vials.
The alcohols, (S)-1a and (S)-1e-h, were modelled into a model
of the G165L/M221F variant based on 1YU6[41]. According to the
modeling, the new variant has opened up a new cavity in the S1´
pocket that harbors longer alkyl chains of secondary alcohols
(Figure 2). In the wild-type enzyme (1YU6[41]), the side chain of
M221 is directed towards the catalytic S220. In the model of
G165L/M221F, the amino acid side chain of F221 is positioned
in the opposite direction compared to the position of the native
methionine side chain. Consequently, a new cavity is created in
the S1´ pocket of G165L/M221F due to the M221F mutation. All
modeled secondary alcohols (Table 3 a, e-h and Figure 2) were
able to fit the alkyl group (methyl to heptyl) into the S1´ pocket of
the G165L/M221F variant while retaining the required hydrogen
bonding network. This was not possible in the wild-type model,
which can only fit a methyl group in its S1´ pocket (Figure 1). The
amino acid residue G165 (or L165) is located in the bottom of
the S1 pocket (Figure S4 and S5). In the wild-type model, the
alkyl chain of vinyl butyrate does not fill up the space in the S1
pocket. In order to reduce this pocket and to improve the binding
of the butyl chain, amino acid residue G165 was substituted by a
leucine. According to the model, the G165L mutation restricts
the S1 cavity (Figure S4 and S5).
Table 4. Enantioselectivities (E-values) and conversion [%] of immobilized[a]
SC wild-type and G165L/M221F for transacylation[b] of (rac)-1-phenylethanol
and vinyl butyrate in THF measured after 120 h.
Temperature [°C]
Wild-type
40, 3%
G165L/M221F
400, 11%[c]
69, 20%
7
50
21, 3%
[a] The enzymes were immobilized on Accurel MP1000 in phosphate buffer
(0.1 M, pH 7.8) containing 1 M KCl.
[b] Reaction conditions (1 mL): Immobilized SC on Accurel MP1000 (10 mg),
Na2CO3 (53 mg), vinyl butyrate (750 mM), 1-phenylethanol (500 M), dodecane
(100 mM) in THF.[13] The reactions were flushed with argon and monitored by
chiral GC.
[c] Measured after 43 h.
Finally, the stability of the most successful variant SC
G165L/M221F in THF was compared to wild-type (Figure 3).
Purified enzyme solutions (1 mg/mL) were prepared in sodium
phosphate buffer and diluted in THF. The remaining relative
activity of SC wild-type and G165L/M221F during storage in THF
was measured using a proteolytic activity assay. The wild-type
enzyme lost 30% activity after 15 minutes of incubation and after
1 h <1% activity remained. Variant G165L/M221F remained 50%
of the initial activity after 1 h of storage in THF. In order to
evaluate the stability of each mutation in G165L/M221F, the
single variants G165L and M221F were also explored. The three
SC variants were found to be more stable than the wild-type
enzyme in THF. The stability of the double variant was shown to
depend on the M221F mutation.
Table 3. Enantioselectivity (E) and conversion [%] of immobilized[a] SC wild-
type and G165L/M221F for transacylation of (rac)-secondary alcohols (rac)-
1a-h and vinyl butyrate in THF measured after 48 h.[b]
O
SC wt or double mutant
Na2CO3
OH
R2
OH
R2
O
O
+
+
THF
R2
O
R1
R1
R1
(rac)-1a-h
(R)-1a-h
2
(S)-3a-h
Entry[b]
R1
-H
R2
Wild-type[a]
G165L/M221F[a]
120, 36%
43, 31%
a
b
c
d
e
-CH3
-CH3
-CH3
-CH3
-C2H5
44, 3 %
7, <1%
41, 6%
2, <1%
0, 0%
-CH3
-Cl
110, 36%
140, 26%
210, 6%
-OMe
-H
This article is protected by copyright. All rights reserved.

10.1002/cbic.201700408
ChemBioChem
FULL PAPER
Figure 2. The tetrahedral intermediates of (S)-α-MBB (cyan), (S)-α-
ethylbenzyl butyrate (green), (S)-α-propylbenzyl butyrate (yellow), (S)-α-
butylbenzyl butyrate (magenta) and (S)-α-hexylbenzyl butyrate (grey) are
coordinated into the active site of SC G165L/M221F. The alkyl chains of the
secondary alcohol are coordinated into the S1´ binding pocket and the phenyl
group of the alcohol is directed towards the solvent. The alkyl chain of the acyl
donor, vinyl butyrate, is coordinated into the S1 binding pocket. L165 is located
in the bottom of the S1 pocket and is displayed in Figure S5. The enzyme
structure is shown in surface mode in element colors. Molecular graphics
created with YASARA[53,54] and PovRay[55]
.
Figure 1. The tetrahedral intermediates of the model reaction (Scheme 1) built
into SC wild-type (PDB ID: 1YU6[41]). The enzyme structure is shown in
surface mode in element colors. The (S)-α-methylbenzyl butyrate ((S)-α-
MBB) is shown in cyan and the (R)-α-methylbenzyl butyrate ((R)-α-MBB) is
shown grey. The catalytic important amino acid residues (H63 and S220) and
the amino acid residues that build up the oxyanion hole (N154 and S220) are
labeled. The amino acid residue M221 is labeled as well as the S1 (acyl) and
the S1´ (alcohol) binding pockets. The alkyl chain of the acyl donor, vinyl
butyrate, is coordinated into the S1 binding pocket. The alkyl chain of the
secondary alcohol is directed into the S1´ binding pocket and the phenyl group
is directed towards the solvent. G165 is located in the bottom of the S1 pocket
and is displayed in Figure S4. Molecular graphics created with YASARA[53,54]
Figure 3. Remaining relative activity of SC wild-type, G165L, M221F and
G165L/M221F during storage in THF by a proteolytic activity assay containing
N-succinyl-Ala-Ala-Pro-Phe-p-nitroanilide in phosphate buffer (0.1 M, pH 7.8).
The residual activity is expressed as a percent of an enzyme control.
Conclusions
As previously been concluded by others[30], the enantioselectivity
of SC was more dependent on the treatment of the enzyme prior
to immobilization, than the type of immobilization technique.
However, immobilization of SC to Accurel MP1000 in phosphate
buffer (0.1 M, pH 7.8) supplemented with 1 M KCl resulted in
efficient immobilization (5% enzyme was remaining in solution
after 90 min) (Figure S6). In addition, the lyophilized enzyme
preparations were stable for several months when stored at
room temperature in a desiccator saturated with LiCl to provide
a low water activity. A new SC variant, G165L/M221F, was
created by site-directed mutagenesis. This variant showed E-
values >100 towards (S)-1a in the model reaction. By enzyme
kinetics, the G165L/M221F mutations were shown to increase
the enzyme substrate specificity towards the desired (S)-1a,
while the undesired (R)-enantiomer was reduced. In contrast to
the wild-type enzyme, the enantioselectivity of G165L/M221F
was shown to be significantly temperature dependent. The
highest E-value was obtained at 7 °C for the model reaction.
The M221F mutation opened up new space in the S1´pocket as
well as improved the enzyme stability in THF. The
G165L/M221F variant showed E-values >200 for secondary
alcohols ((S)-1e-h) harboring longer alkyl chains (C2H5 to C7H11)
and PovRay[55]
.
F221
S1´
(S)-α-(CH3-C7H11)-
benzyl butyrate
S220
H63
N154
S1
This article is protected by copyright. All rights reserved.

10.1002/cbic.201700408
ChemBioChem
FULL PAPER
Mutations were made to the two wild-type models containing either (R)-
or (S)-α-MBB (Figure S2 and S3). The selected amino acid residues for
amino acid exchange were swapped into other amino acid residues. After
energy minimization (using the protocol above), the following was
measured: (i) Hydrogen bond distances between; S221 OG and H64
HE2 or S221 O6 (from alcohol) and H64 HE2; (ii) Angles between; S221
OG, H64 HE2 and H64 NE2 or S221 O6 (from alcohol), H64 HE2 and
H64 NE2. The atom numbering (according to the mature protein) is
shown in Figure S2 and the obtained data is shown in Table S1.
than a methyl group. These substrates, with propyl to heptyl
groups, were not possible to convert using the wild-type enzyme.
Experimental Section
Materials
All commercially available chemicals were purchased from Sigma-Aldrich
and used without further purification. Other chemicals (alcohols and
esters) for KR were synthesized as described in section “Synthesis of
substrates and products”.
In order to explore the acceptance of SC G165L/M221F towards 1-
phenyl alcohols with longer alkyl chains than methyl, the previous model
containing (S)-α-MBB was modified from methyl to propyl, butyl, pentyl
and heptyl (C2H5 to C7H11) and saved as separate files. All files were
energy minimized (using the protocol above). The obtained models were
The B. subtilis WB600 cells and the gene AGN35600.1[40] encoding for
subtilisin Carlsberg AprE [Bacillus paralicheniformis ATCC 9945a]
inserted in a commercial E. coli-Bacillus shuttle vector pHY300PLK
(Takara Bio, Inc., Shiga, Japan) together with its promoter and pre-pro-
sequence[51] were kindly provided by the research group of Prof. Dr.
Ulrich Schwaneberg (RWTH Aachen University in Germany).
structurally aligned and compared (Figure 2) using Mustang[58]
.
H6-tag insertion
The gene of SC was provided in the pHY300PLK shuttle vector.[51]
A
Histidine-tag (H6-tag) was inserted at the C-terminus of the SC
AGN35600.1[40] gene. The primers for the H6-tag insertion were designed
to partially overlap in the area of the insert. The PCR product was
purified, the pHY300SC template was removed by DpnI (FD1704,
Thermo Fisher Scientific) treatment (37 °C, over night) and the digested
PCR-product was repurified. The obtained PCR construct (pHY300SCH6)
was sequenced after transformation and stored at -20 °C.
All primers (for his-tag insertion and site-directed mutagenesis) are found
in the list of primers in the Supplementary Information. The primers were
purchased from Eurofins Genomics. Also, Eurofins Genomics performed
DNA sequencing.
For PCR product purification and DNA plasmid extraction, a GeneJET
PCR Purification Kit (K0701) and a GeneJET Plasmid MiniPrep Kit
(K0502) from Thermo Fisher Scientific were used.
Site-directed mutagenesis
Molecular modeling
Point mutations were introduced in pHY300SCH6 by site-directed
mutagenesis using a two-stage PCR method.[59,60] Starting with two PCR
reactions (50 µl), each containing either the forward or the reverse primer
(0.5 µM), single stranded mutated fragments of the gene were created in
separate tubes.
Molecular modeling was performed using the software YASARA (Version
16.4.16)[53,54] and the crystal structure with PDB ID: 1YU6[41] obtained
from the RCSB Protein Data Bank. All molecular numbers mentioned in
this section is according to the pdb file.
Initially, the molecules A, C and D in 1YU6[41] were deleted. The
remaining object (Mol B) was cleaned. The hydrogen atom (HG) on S221
OG was deleted and added on H64 NE2 as HE2. A simulation cell of 5 Å
around all atoms was set.
PCR reaction, first step: 98 °C for 30 s, one cycle; 98 °C for 10 s, 55 °C
for 30 s, 72 °C for 3 min, 3 cycles. The reactions were then combined in
one tube (100 µL). Second step: (98 °C for 30 sec, one cycle; 98 °C for
10 sec, 55 °C for 30 sec, 72 °C for 3 min, 15 cycles; 72 °C for 5 min, one
cycle).
To remove bumps and correct the covalent geometry, the structure was
energy-minimized with the Amber99 force field[56], using an 8.0 Å force
cutoff and the Particle Mesh Ewald algorithm[57 to treat long range
electrostatic interactions. After removal of conformational stress by a
short steepest descent minimization, the procedure continued by
simulated annealing (timestep 2 fs, atom velocities scaled down by 0.9
every 10th step) until convergence was reached, i.e. the energy
improved by less than 0.05 kJ/mol per atom during 200 steps.
The obtained PCR products were purified, digested with DpnI and re-
purified. The obtained PCR constructs were stored at -20 °C.
Transformation
The mutated DNA plasmids were transformed into E. coli DH5α by
electroporation (2.5 kV, 200 Ω, 25 µF). Aliquots of E. coli DH5α (40 µl)
were thawed on ice (30 min) prior to addition of purified DNA plasmid (1
µl, 50 ng/µl). After the electroporation, SOC-medium (0.5 mL) was
immediately added to the cells. The transformation solution was
incubated for 2 h at 200 rpm and 37 °C. Then, the transformation solution
was spread out on LB-Amp agar plates and incubated at 37 °C overnight.
The tetrahedral intermediates of (R)- and (S)-α-MBB were stepwise built
into the structure from S221 OG. Energy minimizations (using the
protocol above) were performed on the tetrahedral intermediate
structures of (R)- and (S)-α-MBB. During the energy minimizations, the
protein backbone remained fixed and the S221 residue remained free. As
a final step, energy minimizations were performed on free structures.
This article is protected by copyright. All rights reserved.

10.1002/cbic.201700408
ChemBioChem
FULL PAPER
Then, one colony of E. coli DH5α containing pHY300SCH6 (wild-type or
mutant) was cultivated in LB-Amp medium overnight at 37 °C and at 200
rpm. The plasmid was isolated and purified. The wild-type and the
mutations were confirmed by sequencing.
1000 Spectrophotometer at 280 nm. The purified enzyme solutions
were stored at 4 °C prior further use.
Immobilization on EziG and co-lyophilization with surfactants
After sequencing, the pHY300SCH6 (wild-type and mutants) were
transformed to B. subtilis WB600 for protein expression. The
electroporation of B. subtilis WB600 was performed according to Lu et
al.[61] Aliquots of electrocompetent B. subtilis WB600 were thawed on ice
for 30 min before addition of plasmid DNA (1 µl, 50 ng/µl). Subsequently,
the cells were transferred to the electroporation cuvette, incubated on ice
for additional 5-10 min and subjected to electroporation (2.4 kV, 200 Ω,
25 µF). LB medium (1 mL) supplemented with D-sorbitol (0.5 M), and D-
mannitol (0.38 M) was immediately added to each cuvette. The
transformation solution transferred to 15 mL Falcon tubes before
incubation for 3-6 h at 37 °C and 220 rpm. Then, the transformation
solution was spread on LB-Tet-Milk agar plates containing 15 µg/mL
tetracycline and 1.5 % skim-milk in LB agar.
Immobilization of SC on hydrophobic EziG 2 material was made using
sterile filtered (0.45 and 0.2 µm) cell culture supernatant in order to
obtain simultaneous protein purification and immobilization.[64] The EziG
2 material was added in 4-fold excess, compared to the amount of
protein in solution. For SC wild-type, about 300 mg EziG 2 per L cell
culture supernatant was applied. The amount of protein was measured
by Bradford[63]. The immobilization solution was incubated at 30 °C and
200 rpm for 1 h in 0.5 mL media bottle. Then, the carrier material was
collected by centrifugation at 750 x g for 2 min, washed twice with
sodium phosphate buffer (0.1 M, pH 7.8), washed twice with surfactant
solution (Octyl β-D-glucopyranoside (2.5 mg/ml) and BrijTM 58 (2.5
mg/ml) dissolved (37 °C, 220 rpm) in sodium phosphate buffer (0.1 M,
pH 7.8) according to Borén et al.[13], pelleted by centrifugation at 1000 x g
for 1 min, and finally, the supernatant was removed. The immobilized
enzyme preparation on EziG 2 was rapidly frozen in liquid nitrogen and
lyophilized overnight.
Protein expression
The E. coli-Bacillus shuttle vector pHY300PLK allows protein expression
in B. subtilis.[62] The pHY300SCH6 (wild-type or variants) were produced
in B. subtilis WB600, which continuously secrets the expressed enzyme
into the cell culture supernatant. One colony of B. subtilis WB600
containing pHY300SCH6 (wild-type or variants) was grown in LB medium
(4 mL) containing 15 ug/mL tetracycline overnight at 37 °C and 180 rpm
in 50 mL Falcon tubes. A fraction of the overnight culture (0.5-1 mL) was
used to inoculate shake flask cultures (200-250 mL). The protein
expression was performed in unbaffled glass shake flasks (2 L) using
1.5x LB medium (tryptone 15 g/L, NaCl 7.5 g/L, yeast extract 7.5 g/L)
supplemented with tetracycline (15 µg/mL) during 60-70 h at 37 °C at 180
rpm. Then, the supernatant was separated from the cells by
Immobilization on Accurel MP1000
Accurel MP1000 (particle size <1500 m) was pre-wetted in ethanol (95%).
Thereafter, the beads were washed 3-fold with sodium phosphate buffer
(100 mM, pH 7.8). Purified enzyme solution (~10 mg) was diluted to in
sodium phosphate buffer (100 mM, pH 7.8) containing KCl (1 M). The
diluted enzyme solution (10 mL, 1 mg/mL) was added to wet Accurel
MP1000 (1 g) in a glass vial (20 mL). The immobilization solution was
incubated at room temperature on an end-over-end rotation wheel for 1-2
h. The enzyme activity was measured before, during and after the
immobilization using a proteolytic activity assay (SI Figure S6). As no
proteolytic activity could be detected (1-2 h) in solution, the liquid solution
was discarded by pipetting. The immobilized enzyme preparation was
lyophilized overnight, without previous freezing.
centrifugation (JA-10, 8500 rpm, 30 min,
4 °C). The supernatant,
containing the secreted SC, was collected and sterile filtered (0.45 and
0.2 µm). The total protein amount was determined by Bradford[63] using
Bio-Rad Protein assay solution (Bio-Rad, #5000001). The pH of the
protein solution was set to 7.8 prior to further application or storing at
4 °C.
Storage of immobilized enzyme preparations
The immobilized and lyophilized enzyme preparations were stored at
23 °C (room temperature) in a desiccator containing a saturated water
solution of LiCl.
Protein purification
The supernatant solutions containing his-tagged SC wild-type or variant
enzymes were manually purified on a HisTrap™ HP column (17-5247-01,
GE Healthcare). The column was pre-washed with a sodium phosphate
buffer (20 mM, pH 7.8) containing sodium chloride (500 mM) and the
protein was eluted with a sodium phosphate buffer (20 mM, pH 7.8)
containing sodium chloride (500 mM) and imidazole (500 mM). The
collected protein solutions were concentrated and the buffer was
exchanged to sodium phosphate buffer (100 mM, pH 7.8) using Amicon
Ultra (15 mL) centrifugal filters with a molecular weight cut-off of 10,000
(Merck). The amount of purified protein was measured on NanoDropTM
Transacylation reactions using SC immobilized on EziG
The enantioselectivity of the obtained immobilized SC and variants on
EziG 2 were explored for the model reaction (Scheme 1) using: 50 mM
(rac)-1a, 75 mM vinyl butyrate, 25 mM dodecane in THF. The THF was
dried over activated molecular sieves (3 Å) for at least 12 h prior use. To
the reaction solution (1 mL), both enzyme prepared on EziG 2 (10
mg/mL), dry Na2CO3 (53 mg/mL) was added. In order to avoid air
exposure, argon gas was added to the reaction tubes. Spontaneous
This article is protected by copyright. All rights reserved.

10.1002/cbic.201700408
ChemBioChem
FULL PAPER
Equation 3:
transacylation reactions without enzyme were not observed. The reaction
mixture was incubated at 30 °C and 220 rpm in an incubator shaker.
Reaction samples (50 µL) were taken to GC vials containing cyclohexane,
after 0 h, 2 h, 4 h, 24 h and 48 h.
ꢀꢀ!
ꢀꢀ! + ꢀꢀ!
Conversion =
Synthesis of substrates and products
The kinetic parameters for SC wild-type and SC variants were measured
using the (R)- and (S)-1a in separate reactions tubes in a pseudo-one-
substrate approach. Therefore, vinyl butyrate was used at a constant
concentration of 200 mM[20,28,33,42] and dodecane (100 mM) was added as
an internal standard. The concentration of 1-phenylethanol was varied
from 10 mM to 1000 mM (10 mM, 20 mM, 30 mM, 40 mM, 50 mM, 75
mM, 100 mM, 200 mM, 500 mM, 1000 mM). The reaction mixtures were
incubated at 30 °C and shaking at 220 rpm in an incubator shaker.
Reaction samples (50 µl) were taken at 0 h, 2 h, 4 h, 6 h, 24 h and 48 h
to cyclohexane prior to GC analysis.
Substrates and products, that were not commercially available, were
synthesized according to literature procedures.
Procedure A: The secondary alcohols, (rac)-1f-h, were synthesized
from the corresponding ketones. The ketone (27 mmol) was solved in
MeOH (50 mL) and NaOH was added to get the solution basic. The
mixture was cooled down to 0 °C and 1.6 equivalents of NaBH4 was
added in small portions. The solution was stirred overnight in room
temperature. The reaction was diluted with 50 mL water and acidified
with concentrated HCl solution. The MeOH was evaporated under
vacuum and the water solution was extracted with EtOAc. The organic
phase was washed with brine solution, separated, dried with anhydrous
Na2SO4. The solvent was evaporated. The product was analyzed by
NMR spectroscopy. The yield was quantitative in each case and further
purification was not necessary.
Transacylation reactions using SC immobilized on Accurel MP1000
Due to a low stability of SC on EziG, further reactions were made with
enzyme immobilized on Accurel MP1000. In addition, the substrate
concentrations were increased (1 mL): 500 mM racemic alcohol, 750 mM
vinyl butyrate, 100 mM dodecane in dry THF[13]
.
1-phenylbutan-1-ol ((rac)-1f): The synthesis followed Procedure A. The
product appeared as a colorless oil with the mass of 3.77 g (93% yield).
1H NMR (400 MHz, CDCl3): δ [ppm] 7.37-7.26 (m, 5H) 4.70-4.66 (m, 1H),
1.85-1.64 (m, 3H), 1.51-1.25 (m, 2H), 0.94 (t, 3H, J = 7.3 Hz), 13C NMR
(100 MHz, CDCl3): δ [ppm] 144.9, 128.4, 127.5, 125.9, 74.4, 41.3, 19.0,
14.0.
GC analysis
All collected reaction samples were analyzed by gas chromatography
(GC) using a chiral column (CP-ChiraSil-DEX CB 25 m x 0.32 mm x 0.25
µm) without prior storage. For the model reaction (Scheme 1) the
following GC-method was used: 90 °C hold 4 min, 5 °C/min to 150 °C,
70°C/min until 200 °C and 200 °C hold 2 min.
1-phenylpentan-1-ol ((rac)-1g): The synthesis followed Procedure A.
The product appeared as a colorless oil with the mass of 4.26 g (96%
yield). 1H NMR (400 MHz, CDCl3): δ [ppm] 7.37-7.27 (m, 5H) 4.69-4.65
(m, 1H), 1.85-1.68 (m, 3H), 1.45-1.23 (m, 4H), 0.90 (t, 3H, J = 7.1 Hz),
13C NMR (100 MHz, CDCl3): δ [ppm] 145.1, 128.6, 127.6, 126.0, 74.9,
39.0, 28.1, 22.8, 14.2.
A calibration curve was prepared in THF using rac-α-MBB in different
concentrations: 1 mM, 5 mM, 10 mM, 25 mM, 50 mM, 75 mM and 100
mM, while keeping the internal standard (dodecane) constant at 100 mM.
The ratio between the concentration of one enantiomer and dodecane
was used to calculate the concentration (mM) of the obtained products.
1-phenylheptan-1-ol ((rac)-1h): The synthesis followed Procedure A.
The product appeared as a colorless oil with the mass of 4.78 g (92%
yield). 1H NMR (400 MHz, CDCl3): δ [ppm] 7.36-7.27 (m, 5H) 4.68-4.65
(m, 1H), 1.85-1.66 (m, 3H), 1.45-1.25 (m, 8H), 0.87 (t, 3H, J = 6.8 Hz),
13C NMR (100 MHz, CDCl3): δ [ppm] 145.1, 128.6, 127.6, 126.0, 74.9,
39.3, 31.9, 29.3, 26.0, 22.8, 14.2.
Calculation of enantiomeric ratio and reaction yield
The enantiomeric ratio (E-value) was calculated using two different
methods. As racemic secondary alcohols were applied in the reaction
solution, equation (1) was used for calculation of the E-values.[65] As the
two enantiomers were applied in separate reaction solutions, equation (2)
was applied for the calculation of the E-values.[66] The conversion of the
reaction was calculated according to equation (3).[66]
Procedure B: To be able to identify the enantioselectivity (ee) of the
products, the (rac)-1-phenylethyl butyrate derivatives (rac)-3a-h were
synthesized. The alcohol substrate (rac)-1a-h (1 mmol) was solved in
CH2Cl2 (3 mL). 3.4 equivalents of butyric anhydride and 0.25 equivalents
of DMAP were added to the solution and stirred at room temperature
overnight. The reaction mixture was diluted with CH2Cl2 and washed with
water four times. The organic phase was separated and dried with
anhydrous Na2SO4. The solvent was evaporated. The product was
purified by column chromatography with Al2O3 to be able to remove the
acid residues and it was analyzed by NMR spectroscopy.
Equation 1:
1 − ꢀꢀ!
ln
1 + ꢀꢀ! / ꢀꢀ!
1 + ꢀꢀ!
1 + ꢀꢀ! / ꢀꢀ!
ꢀ =
ln
Equation 2:
1-phenylbutyl butyrate ((rac)-3f): The synthesis followed Procedure B.
The product appeared as a colorless oil with the mass of 212 mg (96%
yield). 1H NMR (400 MHz, CDCl3): δ [ppm] 7.34-7.26 (m, 5H), 5.77-5.74
(m, 1H), 2.35-2.26 (m, 2H), 1.94-1.85 (m, 1H), 1.78-1.61 (m, 3H), 1.41-
ꢀꢁꢂꢃ ꢀꢁ ꢀ
ꢀꢁꢂꢃ ꢀꢀ ꢀ
ꢀ ꢀ
ꢀ ꢀ
ꢀ =
=
This article is protected by copyright. All rights reserved.

10.1002/cbic.201700408
ChemBioChem
FULL PAPER
1.22 (m, 2H), 0.94-0.90 (m, 6H), 13C NMR (100 MHz, CDCl3): δ [ppm]
173.1, 141.2, 128.5, 127.9, 126.6, 75.7, 38.7, 36.7, 18.9, 18.6, 14.0, 13.8.
1-phenylpentyl butyrate ((rac)-3g): The synthesis followed Procedure B.
The product appeared as a colorless oil with the mass of 220 mg (94%
yield). 1H NMR (400 MHz, CDCl3): δ [ppm] 7.36-7.27 (m, 5H) 5.76-5.72
(m, 1H), 2.33-2.29 (m, 2H), 1.95-1.86 (m, 1H), 1.81-1.73 (m, 1H), 1.70-
1.61 (m, 2H), 1.34-1.20 (m, 4H), 0.92 (t, 3H, J = 7.5 Hz), 0.87 (t, 3H, J =
7.1 Hz), 13C NMR (100 MHz, CDCl3): δ [ppm] 173.1, 141.2, 128.5, 127.9,
126.6, 76.0, 36.7, 36.3, 27.8, 22.6, 18.6, 14.1, 13.8.
[1]
[2]
[3]
R. N. Patel, Coord. Chem. Rev. 2008, 252, 659-701.
T. Itoh, U. Hanefeld, Green Chem. 2017, 19, 331-332.
M. Breuer, K. Ditrich, T. Habicher, B. Hauer, M. Kesseler, R. Stürmer, T.
Zelinski, Angew. Chem., Int. Ed. 2004, 43, 788-824.
O. Verho, J.-E. Bäckvall, J. Am. Chem. Soc. 2015, 137, 3996-4009.
D. Muñoz Solano, P. Hoyos, M.J. Hernáiz, A.R. Alcántara, J.M.
Sánchez-Montero, Bioresour. Technol. 2012, 115, 196-207.
A. O. Magnusson, M. Takwa, A. Hamberg, K. Hult, Angew. Chem. Int.
Ed. 2005, 44, 4582-4585.
[4]
[5]
[6]
[7]
[8]
[9]
A. O. Magnusson, J. C. Rotticci-Mulder, A. Santagostino, K. Hult,
ChemBioChem 2005, 6, 1051-1056.
K. Engström, M. Vallin, P.O. Syren, K. Hult, J.-E. Bäckvall, Org. Biomol.
Chem. 2011, 9, 81-82.
1-phenylheptyl butyrate ((rac)-3h): The synthesis followed Procedure B.
The product appeared as a colorless oil with the mass of 234 mg (89%
yield). 1H NMR (400 MHz, CDCl3): δ [ppm] 7.35-7.27 (m, 5H) 5.75-5.72
(m, 1H), 2.36-2.26 (m, 2H), 1.94-1.85 (m, 1H), 1.80-1.61 (m, 3H), 1.32-
1.22 (m, 8H), 0.92 (t, 3H, J = 7.4 Hz), 0.86 (t, 3H, J = 6.9 Hz), 13C NMR
(100 MHz, CDCl3): δ [ppm] 173.1, 141.2, 128.5, 127.8, 126.6, 76.0, 36.7,
36.6, 31.8, 29.1, 25.6, 22.7, 18.6, 14.2, 13.8.
K. Engström, M. Vallin, K. Hult, J.-E. Bäckvall, Tetrahedron 2012, 68,
7613-7618.
[10] K. Linderstrøm-Lang, M. Ottesen, Nature 1947, 159, 807-808.
[11] D. Kumar, Savitri, N. Thakur, R. Verma and T. C. Bhalla, Res. Microbiol.
2008, 3, 661-672.
[12] M.-J. Kim, Y. I. Chung, Y. K. Choi, H. K. Lee, D. Kim, J. Park, J. Am.
Chem. Soc. 2003, 125, 11494-11495.
[13] L. Borén, B. Martín-Matute, Y. Xu, A. Córdova, J-E. Bäckvall, Chem.
Eur. J. 2006, 12, 225-232.
Stability measurement
[14] J. Norinder, K. Bogár, L. Kanupp, J.-E. Bäckvall, Org. Lett. 2007, 9,
5095-5098.
SC wild-type and variants were purified and diluted (1 mg/mL) in sodium
phosphate buffer (0.1 M, pH 7.8). A tenfold dilution of the enzyme
solution was made in THF in glass vials. Samples were taken during 60
min and immediately analyzed for remaining activity using a proteolytic
activity assay.
[15] M.-J. Kim, H. K. Lee, J. Park, Bull. Korean Chem. Soc. 2007, 28, 2096-
2098.
[16] L. K. Thalén, A. Sumic, K. Bogár, J. Norinder, A. K. Å. Persson, J.-E.
Bäckvall, J. Org. Chem. 2010, 75, 6842-6847.
[17] D. L. Ollis, E. Cheah, M. Cygler, B. Dijkstra, F. Frolow, S. M. Franken,
M. Harel, S. J. Remington, I. Silman, J. Schrag, J. L. Sussman, K. H. G.
Verschueren, A. Goldman, Protein Eng. 1992, 30, 321-331.
[18] R. J. Kazlauskas, A. E. Weissfloch, A. T. Rappaport, L. A. Cuccia, J.
Org. Chem. 1991, 56, 2656-2665.
Proteolytic activity assay
[19] R. J. Kazlauskas, A. E. Weissfloch, J. Mol. Catal. B: Enzym. 1997, 3,
65-72.
A proteolytic activity assay based on hydrolysis of a synthetic peptide N-
Succinyl-Ala-Ala-Pro-Phe p-nitroanilide (S7388 Sigma) was applied in
order measure the proteolytic activity of SC wild-type, different SC
variants, batches and immobilization. The proteolytic activity was
monitored after all working steps and during the immobilization (on
Accurel MP1000). Aliquots of the synthetic peptide was diluted to 20 mM
in DMSO and stored at -20 °C. Reactions (1 mL) containing; sodium
phosphate buffer (990 µl, 0.1 M, pH 7.8), synthetic peptide (5 µl, 20 mM)
and enzyme solution (5 µl), were prepared. The initial hydrolysis reaction
[20] H. Noritomi, Ö. Almarsson, G. L. Barletta, A. M. Klibanov, Biotechnol.
Bioeng. 1996, 51, 95-99.
[21] J. Broos, Biocatal. Biotransform. 2002, 20, 291-295.
[22] J. Mukherjee, M. N. Gupta, Biotechnol. Rep. 2015, 6, 119-123.
[23] W.-T. Hsu, D. S. Clark, Biotechnol. Bioeng. 2001, 73, 231-237.
[24] I. Montañez-Clemente, E. Alvira, M. Macias, A. Ferrer, M. Fonceca, J.
Rodriguez, A. Gonzalez, G. Barletta, Biotechnol. Bioeng. 2002, 78, 53-
59.
[25] V. Stepankova, S. Bidmanova, T. Koudelakova, Z. Prokop, R.
Chaloupkova, J. Damborsky, ACS Catal. 2013, 3, 2823-2836.
[26] J. Mukherjee, M. N. Gupta, Tetrahedron Lett. 2015, 56, 4397-4401.
[27] J. Mukherjee, P. Mishra, and M. N. Gupta. Tetrahedron Lett. 2015, 56,
1976-1981.
rates were measured at 410 nm on
a
Varian Cary® 50 UV-Vis
spectrophotometer. A standard curve of p-nitroanilide (31569 Sigma-
Aldrich) was prepared (0.02-0.14 mM).
[28] S. G. Martínez, E. Alvira, L. V. Cordero, A. Ferrer, I. Montañés-
Clemente, G. Barletta, Biotechnol. Prog. 2002, 18, 1462-1466.
[29] S. Prasad, V. S. Negi, J. K. Laha, I. Roy, J. Mol. Catal. B: Enzym. 2016,
134, 32-36.
Acknowledgements
[30] B. Orsat, G. J. Drtina, M. G. Williams, A. M. Klibanov, Biotechnol.
Bioeng. 1994, 44, 1265-1269.
This work was funded by KTH Royal Institute of Technology and the
Erasmus+ Traineeship program. The research group of Prof. Dr. Ulrich
Schwaneberg at RWTH Aachen University in Germany is gratefully
acknowledged for providing both the gene of Subtilisin Carlsberg and the
Bacillus subtilis WB600 cells. EnginZyme AB is acknowledged for
providing the EziG 2 material.
[31] Y. Guo, D. S. Clark, BBA Protein Struct. M. 2001, 1546: 406-411.
[32] K. Sangeetha, T. E. Abraham, Int. J. Biol. Macromol. 2008, 43, 314-319.
[33] B. Castillo, Y. Delgado, G. Barletta, K. Griebenow, Tetrahedron 2010,
66, 2175-2180.
[34] J.-K. Bang, S.-O. Jung, Y.-W. Kim, M.-J. Kim, Bull. Korean Chem. Soc.
2011, 32, 2871-2872.
[35] M. N. Gupta, J. Mukherjee, D. Malhotra, Univers. Org. Chem. 2013, 1,
1.
Keywords: Biocatalysis • Enzyme immobilization • Kinetic
[36] Y. L. Khmelnitsky, R. Hilhorst, A. J. Visser, C. Veeger, Eur. J. Biochem.
1993, 211, 73-77.
resolution • Molecular modeling • Transacylation
[37] N. B. Philip, BBA Protein Struct. Mol. Enzymol. 2000, 1543, 203-221.
This article is protected by copyright. All rights reserved.

10.1002/cbic.201700408
ChemBioChem
FULL PAPER
[38] R. Gupta, Q.K. Beg, P. Lorenz, Appl. Microbiol. Biotechnol. 2002, 59,
15-32.
[52] D. Despotovic, L. Vojcic, M. Blanusa, K.-H. Maurer, M. Zacharias, M.
Bocola, R. Martinez, U. Schwaneberg, J. Biotechnol. 2013, 167, 279-
286.
[39] M. Dickman, R. C. Lloyd, J. B. Jones, Tetrahedron: Asymmetry 1998, 9,
4099-4102.
[53] E. Krieger, G. Vriend, Bioinformatics 2014, 30, 2981-2982.
[54] The YASARA software, http://www.yasara.org. 2017 (assessed
17.06.17).
[40] M. Rachinger, S. Volland, F. Meinhardt, R. Daniel, H. Liesegang,
Genome Announc. 2013, 1, e00525-13.
[41] J. T. Maynes, M. Cherney, A. Qasim, J. Laskowski, N. G. James, Acta
Crystallogr., Sect. D. 2005, 61, 580-588.
[55] The PovRay software, http://www.povray.org. 2017 (assessed
17.06.17).
[42] P. A. Fitzpatrick, A. M. Klibanov, J. Am. Chem. Soc. 1991, 113, 3165-
3171.
[56] J. Wang, P. Cieplak, P. A. Kollman, J.Comput.Chem. 2000, 21, 1049-
1074.
[43] J. Uppenberg, M.T. Hansen, S. Patkar, T.A. Jones, Structure 1994, 2,
293-308.
[57] U. Essman, L. Perera, M. L. Berkowitz, T. Darden, H. Lee, L. G.
Pedersen, J. Chem. Phys. B. 1995, 103, 8577-8593.
[58] A. S. Konagurthu, J. C. Whisstock, P. J. Stuckey, A. M. Lesk, Proteins,
2006, 64, 559-574.
[44] D. A. Estell, T. P. Graycar, J. A. Wells, J. Biol. Chem. 1985, 260, 6518-
6521.
[45] R. Bott, M. Ultsch, A. Kossiakoff, T. Graycar, B. Katz, S. Power, J. Biol.
Chem. 1988, 263, 7895-906.
[59] W. Wang, B. A. Malcolm, Methods Mol. Biol. 2002, 182, 37-43.
[60] O. Edelheit, A. Hanukoglu, I. Hanukoglu, BMC Biotechnol. 2009, 9, 1-8.
[61] Y.-P. Lu, C. Zhang, F.X. Lv, X.M. Bie, Z.-X. Lu, Lett. Appl. Microbiol.
2012, 55, 9-14.
[46] H. Gron, L. M. Bench, S. Branner, K. Breddam, Eur. J. Biochem. 1990,
194, 897-901.
[47] S. Atwell, J. A. Wells, Proc. Natl. Acad. Sci. 1999, 96, 9497-9502.
[48] S.D. Power, R. M. Adams, J. A. Wells, Proc. Natl. Acad. Sci. 1986, 83,
3096-3100.
[62] H. Ishiwa, H. Shibahara, Jpn. J. Genet. 1985, 60, 235-243.
[63] M. M. Bradford, Anal. Biochem. 1976, 72, 248-254.
[64] K. Engelmark Cassimjee, M. Kadow, Y. Wikmark, M. Svedendahl
Humble, M. L. Rothstein, D.M. Rothstein, J.-E Bäckvall, Chem.
Commun. 2014, 50, 9134-9137.
[49] J. Eder, M. Rheinnecker, A. L. Fersht, J. Mol. Biol. 1993, 233, 293-304.
[50] J. J. Perona, C. S. Craik, Protein Sci. 1995, 4, 337-360.
[51] D. Despotovic, L. Vojcic, R. Prodanovic, R. Martinez, K.-H. Maurer, U.
Schwaneberg, J. Biomol. Screen. 2012, 17, 796-805.
[65] C. S. Chen, Y. Fujimoto, G. Girdaukas, C. J. Sih, J. Am. Chem. Soc.
1982, 104, 7294–7299.
[66] J. L. Rakels, A. J. Straathof, J. J. Heijnen, Enzyme Microb. Technol.
1993, 15, 1051-1056.
This article is protected by copyright. All rights reserved.

10.1002/cbic.201700408
ChemBioChem
FULL PAPER
FULL PAPER
Robin Dorau, Tamás Görbe, and Maria
Svedendahl Humble
Page No. – Page No.
Improved Enantioselectivity of
Subtilisin Carlsberg Towards
Secondary Alcohols by Protein
Engineering
The performance of the serine protease subtilisin Carlsberg (SC) was improved
towards transacylation reactions using secondary alcohols in THF. By enzyme
immobilization, the enantioselectivity of SC was increased tenfold. By further
protein engineering, an enzyme variant (G165L/M221F) showing increased;
conversion, enantioselectivity (E >100), substrate scope and stability in THF, was
created.
This article is protected by copyright. All rights reserved.