Laboratory evolution of enantiocomplementary Candida antarctica lipase B mutants with broad substrate scope

Article abstract of DOI:10.1021/ja310455t

Candida antarctica lipase B (CALB) is a robust and easily expressed enzyme used widely in academic and industrial laboratories with many different kinds of applications. In fine chemicals production, examples include acylating kinetic resolution of racemic secondary alcohols and amines as well as desymmetrization of prochiral diols (or the reverse hydrolytic reactions). However, in the case of hydrolytic kinetic resolution of esters or esterifying kinetic resolution of acids in which chirality resides in the carboxylic acid part of the substrate, rate and stereoselectivity are generally poor. In the present study, directed evolution based on iterative saturation mutagenesis was applied to solve the latter problem. Mutants with highly improved activity and enantioselectivity relative to wild-type CALB were evolved for the hydrolytic kinetic resolution of p-nitrophenyl 2-phenylpropanoate, with the selectivity factor increasing from E = 1.2 (S) to E = 72 (S) or reverting to E = 42 (R) on an optional basis. Surprisingly, point mutations both in the acyl and alcohol pockets of CALB proved to be necessary. Some of the evolved CALB mutants are also efficient biocatalysts in the kinetic resolution of other chiral esters without performing new mutagenesis experiments. Another noteworthy result concerns the finding that enantiocomplementary CALB mutants for α-substituted carboxylic acid esters also show stereocomplementarity in the hydrolytic kinetic resolution of esters derived from chiral secondary alcohols. Insight into the source of stereoselectivity was gained by molecular dynamics simulations and docking experiments.

Full text of DOI:10.1021/ja310455t

Article
pubs.acs.org/JACS
Laboratory Evolution of Enantiocomplementary Candida antarctica
Lipase B Mutants with Broad Substrate Scope
Qi Wu,†,‡ Pankaj Soni,‡,§ and Manfred T. Reetz*,‡,∥
†Department of Chemistry, Zhejiang University, Hangzhou, 310027, People’s Republic of China
‡Max-Planck-Institut fur Kohlenforschung, Kaiser-Wilhelm-Platz 1, 45470 Mulheim an der Ruhr, Germany
̈
̈
§CSIR-Institute of Microbial Technology, Chandigarh, 160036, India
∥Fachbereich Chemie der Philipps-Universitat, Hans-Meerwein-Strasse, 35032 Marburg, Germany
̈
S
* Supporting Information
ABSTRACT: Candida antarctica lipase B (CALB) is a robust
and easily expressed enzyme used widely in academic and
industrial laboratories with many different kinds of applica-
tions. In fine chemicals production, examples include acylating
kinetic resolution of racemic secondary alcohols and amines as
well as desymmetrization of prochiral diols (or the reverse
hydrolytic reactions). However, in the case of hydrolytic
kinetic resolution of esters or esterifying kinetic resolution of
acids in which chirality resides in the carboxylic acid part of the
substrate, rate and stereoselectivity are generally poor. In the
present study, directed evolution based on iterative saturation mutagenesis was applied to solve the latter problem. Mutants with
highly improved activity and enantioselectivity relative to wild-type CALB were evolved for the hydrolytic kinetic resolution of p-
nitrophenyl 2-phenylpropanoate, with the selectivity factor increasing from E = 1.2 (S) to E = 72 (S) or reverting to E = 42 (R)
on an optional basis. Surprisingly, point mutations both in the acyl and alcohol pockets of CALB proved to be necessary. Some of
the evolved CALB mutants are also efficient biocatalysts in the kinetic resolution of other chiral esters without performing new
mutagenesis experiments. Another noteworthy result concerns the finding that enantiocomplementary CALB mutants for α-
substituted carboxylic acid esters also show stereocomplementarity in the hydrolytic kinetic resolution of esters derived from
chiral secondary alcohols. Insight into the source of stereoselectivity was gained by molecular dynamics simulations and docking
experiments.
Structurally, CALB belongs to the α/β-hydrolase family and
follows the same reaction mechanism as other serine
hydrolases.1 The active site is characterized by the catalytic
triad Ser-His-Asp, an oxyanion hole that stabilizes the transition
state of the reaction, and a binding pocket composed of two
domains, one for the acyl moiety of the ester and another one
for the alcohol part, respectively.8 Consequently, when chirality
is in the alcohol part of an ester, researchers have focused on
the alcohol pocket of CALB (or of any other lipase) for
interpreting enantioselectivity or manipulating it by protein
engineering. Kazlauskas’s rule regarding the kinetic resolution
of racemic secondary alcohols has served as a useful guide in
the prediction of stereoselectivity.9 Indeed, this transformation
is one of the most important applications of CALB. When the
stereogenic center belongs to the acid moiety, attention is
traditionally paid to the acyl pocket of lipases.1,2
INTRODUCTION
■
Lipases (EC 3.1.1.3) catalyze the hydrolysis of carboxylic acid
esters in aqueous medium but also esterification or trans-
esterification in organic solvents.1 Among the commercially
available lipases that can be expressed industrially in large
quantities, Candida antarctica lipase B (CALB; Novozyme 435)
continues to be one of the most extensively used biocatalysts in
both academia and industry.1,2 It is robust, especially in
immobilized form, and has a broad range of applications that
include kinetic resolution of racemic alcohols and amines and
desymmetrization of diols and diacetates as well as utility in
polymer chemistry3 and protection/deprotection technology.4
Numerous examples pertain to the stereoselective synthesis of
chiral intermediates for the production of pharmaceuticals and
plant-protecting agents.1,2 Dynamic kinetic resolution of
alcohols and amines utilizing CALB and a racemizing
transition-metal catalyst has proven to be particularly effective
In contrast to the broad spectrum of chiral alcohols (Scheme
1a), diols, and amines accepted by CALB often with high
enantioselectivity, only a limited number of substrates
as pioneered by Backvall et al.5 and Kim et al.6 Moreover, a
̈
number of fascinating cases of enzyme promiscuity employing
this lipase have been reported by Berglund, Hult, and others,
including aldol reactions, Michael additions, vinyl polymer-
ization, and even oxidations.7
Received: October 23, 2012
Published: January 9, 2013
© 2013 American Chemical Society
1872
dx.doi.org/10.1021/ja310455t | J. Am. Chem. Soc. 2013, 135, 1872−1881

Journal of the American Chemical Society
Article
ed.17g Svendsen has written an informative review covering
protein engineering of CALB and other lipases for the period
up to the year 2000.17h
Scheme 1. CALB-Catalyzed Kinetic Resolution of (a) Chiral
Alcohols and of (b) Chiral Esters
Given the robustness and efficient expression of CALB (tons
are produced each year), it is surprising that few attempts to
apply protein engineering for solving these problems have been
reported to date. Brask and co-workers described the
replacement of helix a5 of CALB by the corresponding lid
region from the CALB homologues from Neurospora crassa and
Gibberella zeae, which resulted in two new CALB mutants with
increased enantioselectivity in the hydrolysis of racemic ethyl 2-
phenylpropanoate as measured by the selectivity factor (E =
50), although the absolute configuration was not determined.18
Lutz and co-workers reported a 2-fold increase in the E value in
the esterification of 2-phenylpropanoic acid as a result of
applying circular permutation.15a Relevant is also a recent study
containing structurally different acids or esters react at
reasonable rates and acceptable stereoselectivity (Scheme
1b).1,2 CALB can convert straight chain fatty acid esters of
different length rapidly only when the chain length varies
between C5 and C12.10 The substrate scope of CALB toward
sterically demanding chiral α-substituted carboxylic acid esters
is narrow, with low activity and poor stereoselectivity generally
being observed.1,2,11 For example, 2-arylpropionic acid esters of
the iboprufen-type, which are of interest as pain relievers in the
pharmaceutical industry,12 prove to be poor substrates, reacting
slowly (or not at all) with low degrees of stereoselectivi-
ty.1,8c,10,11 Some of these problems can be solved by using the
analogous N-acylazoles in combination with CALB, but this
requires additional synthetic steps.11e It is interesting to note
that enzyme-free systems using chiral organocatalysts have been
developed only recently for performing esterifying kinetic
resolution of chiral acids as reported by Shiina et al.13a and
Birman et al.13b
In order to further improve thermostability, activity,
stereoselectivity, and expression rate of CALB and to tailor it
for different applications, a range of mutants have been
developed using rational design or directed evolution with
notable degrees of success, generally focusing on kinetic
resolution of chiral alcohols.1,14−17 For example, error-prone
polymerase chain reaction (PCR) applied to CALB provided
mutants with >20-fold improvement in half-life at 70 °C
relative to the wild type (WT).14b In the quest to increase the
hydrolysis activity, circular permutation was elegantly used to
produce a CALB variant with an up to 175-fold enhanced kcat/
KM value over WT CALB.15a−c Various mutants such as
V190A,15d L278A/W104F,15e and S47L15f were designed to
improve the rate of transesterification in nonaqueous medium.
Random mutagenesis at single residues or multiple positions
can effectively improve the expression yield of CALB in
Escherichia coli, Pichia pastoris, and Hansenula polymorpha.16 In
order to increase the stereoselectivity of CALB toward certain
secondary alcohols, Hult and co-workers resorted to rational
design utilizing site-specific mutagenesis in the generation of
mutations S47A,17a T42V, and T40A17 situated next to the
alcohol binding pocket, which dramatically improved the
enantioselectivity toward pentane-2-ol or halohydrins. Similarly,
they discovered that the W104A mutant is able to
accommodate much larger groups than WT CALB. Interest-
ingly, this change transformed the highly R-selective WT into
an S-selective mutant.17c,d Two recent studies of variant W104A
describe its use in S-selective kinetic resolution and dynamic
kinetic resolution of 1-phenylalkanols with an alkyl chain of up
to eight carbon atoms, with the S-enantiomer being
preferred.17e,f Moreover, when introducing single point
mutations I189A, L278V, and A282L, which can modify the
width and shape of the hydrophobic tunnel leading to the active
site, significant modification of enantioselectivity was observ-
by Backvall and co-workers, who reported the CAST-based
̈
directed evolution of the related C. antarctica lipase A (CALA)
resulting in the acceptance of 2-phenylpropionic acid esters and
similar substrates with high R-selectivity; the structurally related
and medicinally important ibuprofen-ester was not accepted by
the mutants nor was reversal of enantioselectivity in favor of S-
products reported.19a Fortunately, in subsequent attempts
utilizing a more effective strategy based on saturation
mutagenesis at multiple-residue sites, R-selective mutants
were evolved that accept ibuprofen-type esters.19b
We reasoned that if CALB could be tuned to accept 2-
arylpropionic acid esters and similar bulky esters with
acceptable rates and high R- and S-enantioselectivity on an
optional basis, then the range of applications of this industrially
viable lipase would be broadened considerably.
Directed evolution20 of stereoselective enzymes21 as catalysts
for asymmetric transformations in synthetic organic chemistry
and/or biotechnology is now well established. Nevertheless,
some uncertainty still exists as to the optimal method and
strategy for probing protein sequence space optimally. Any one
of the known gene mutagenesis methods such as error-prone
polymerase chain reaction (epPCR), saturation mutagenesis, or
DNA shuffling can be expected to provide mutants with
enhanced stereoselectivity, with the degree of improvement
depending upon the amount of laboratory work that the
experimenter is willing to invest.20,21 Since the bottleneck of
directed evolution is the screening step, higher-quality and
therefore smaller libraries are needed. Our contribution in this
endeavor is iterative saturation mutagenesis (ISM) in its two
structure-based embodiments, namely, the combinatorial
active-site saturation test (CAST),21 which can be applied to
control stereo- and or regioselectivity, to enhance rate and to
expand substrate scope of enzymes, and B-FIT for increasing
the stability of proteins.22 When applying iterative CASTing,
appropriate sites next to or near the binding pocket of an
enzyme, each comprising one or more amino acid positions, are
first subjected to saturation mutagenesis with formation of
focused libraries. The genes of the hits are then utilized as
templates for performing saturation mutagenesis at the
respective other sites, with the process being repeated until
the desired degree of catalyst improvement has been achieved.
This strategy relies on data from structural biology and is quite
different from the more conventional approach in which point
mutations originating from two or more libraries are simply
combined.20,21 In contrast to the latter, ISM maximizes the
probability of cooperative effects acting between sets of
mutations at each branching point in an upward climb and
1873
dx.doi.org/10.1021/ja310455t | J. Am. Chem. Soc. 2013, 135, 1872−1881

Journal of the American Chemical Society
Article
Scheme 2. Stereoselective Hydrolysis of rac-1 Catalyzed by R- and S-Selective CALB Mutants
between point mutations within a set.21 In order to beat the
numbers problem in directed evolution further, the use of
reduced amino acid alphabets as defined by appropriate codon
degeneracies has proven to be a valuable tool because it cuts
oversampling drastically.21
In the present study, we wanted to test the ISM approach in
an attempt to evolve active and stereoselective CALB mutants
suitable for the hydrolytic kinetic resolution of chiral α-
substituted carboxylic acid esters, while keeping the screening
effort as low as possible. Obtaining both R- and S-selective
CALB mutants was also part of the goal, which has not been
achieved prior to this investigation.
RESULTS AND DISCUSSION
■
Experimental Platform. The hydrolytic kinetic resolution
of racemic 2-phenyl propionic acid p-nitrophenyl ester (rac-1)
was chosen as the model reaction (Scheme 2), with a UV/
visible plate reader monitoring the appearance of p-nitro-
phenolate (3) as a function of time.23 In order to detect hits
with enhanced or inverted enantioselectivity, separately
synthesized (R)- and (S)-1 were used pairwise on 96-well
microtiter plates, with the relative slopes of initially observed
rates indicating the degree of stereoselectivity. In order to
obtain reliable selectivity factors E, the hits were subsequently
isolated and tested as catalysts in the reaction of rac-1. WT
CALB is a poor catalyst in this reaction, as the substrate reacts
sluggishly with very slight preference for (S)-2 (E = 1.2).
In order to choose appropriate sites for saturation muta-
genesis, we were guided by the X-ray structure of CALB (PDB
code 1TCA).8 As pointed out in the Introduction, all lipases are
traditionally considered to have two binding domains for
harboring esters: one for the acid (acyl) part and the other for
the alcohol part of the ester. Catalytically active serine as part of
the triad Asp187-His224-Ser105 lies between them (Figure 1a).
The structure of the acyl binding site can be subdivided into a
hydrophilic region and a hydrophobic region, the so-called
hydrophobic crevice. The hydrophobic crevice is defined by
residues A141, L144, V149, and I285 that are located at the top
of the binding site. The hydrophilic region of the binding site is
characterized by residues D134, T138, and Q157. Residues
A281, A282, and I285 point toward the alcohol moiety of the
substrate and limit the size of the channel. The alcohol binding
pocket is defined by the mainly hydrophobic residues W104,
L278, A281, A282, and I285. Initially, we focused saturation
mutagenesis only on sites in the acyl binding domain (sites B,
C, and E as shown in Figure 1b), but as the investigation
Figure 1. X-ray structure of WT CALB (PDB: 1TCA)8 used as a guide
in directed evolution. (a) The catalytic triad S105-H224-D187 is
shown as purple sticks, the alcohol binding pocket as a yellow surface,
and the acid binding pocket as a combination of a blue surface
(hydrophilic region) and a red surface (hydrophobic region). (b)
Proposed CAST sites A (W104/S105), B (L144/V149), C (I189/
V190), D (A281/A282), and E (V154/Q157).
proceeded, residues in the alcohol binding pocket were also
considered (sites A and D).
The 10 residues at sites A−E can in principle be treated
separately as single residue sites that would call for 10 initial
saturation mutagenesis experiments. Although such a strategy
can be successful as shown in other studies,21 we decided to
group them into five two-residue sites, because this maximizes
the possibility of cooperative epistatic effects within a site and
between sets of mutations as the evolutionary ISM process
proceeds.20n,21 When choosing NNK codon degeneracy
encoding all 20 canonical amino acids, saturation mutagenesis
at a two-residue site requires the screening of about 3000
1874
dx.doi.org/10.1021/ja310455t | J. Am. Chem. Soc. 2013, 135, 1872−1881

Journal of the American Chemical Society
Article
transformants for 95% coverage (assuming the absence of
amino acid bias).21 The total initial screening effort in the
present case would amount to about 15 000 measurements. In
order to reduce this work, we chose a reduced amino acid
alphabet by turning to NDT codon degeneracy that encodes 12
amino acids (Phe, Leu, Ile, Val, Tyr, His, Asn, Asp, Cys, Arg,
Ser, and Gly) and requires the screening of only 430
transformants for 95% library coverage.21
chosen routes: one directed toward R-selectivity and the other
toward S-selectivity. Attention was first paid to enhancing the
reversal of enantioselectivity in favor of (R)-2 by utilizing the
gene of variant RG101 as a template for saturation mutagenesis
at site D, which is likewise at the alcohol binding pocket. Three
mutants with moderate R-selectivity were identified, namely,
RG201, RG202, and RG204, with E values in the range 11−15
(Figure 2 and Table 2). The two best mutants were then used
Initial Libraries Generated by Saturation Mutagene-
sis. In exploratory mutagenesis experiments at positions lining
the acyl binding pocket, namely, sites B (Leu144/Val149), C
(Ile189/Val190), and E (Val154/Gln157) (Figure 1b), we were
surprised to discover that these libraries failed to contain
notably improved variants. Although in ISM it is possible to
continue upward climbs even if a library does not harbor
improved variants simply by choosing nonimproved or even
inferior mutants,24 we decided to extend the search in protein
sequence space by including sites lining the alcohol binding
pocket, specifically A (Trp104/Ser105) and D (Ala281/
Ala282) (Figure 1b). Note that Ser105 is part of the catalytic
triad, which means that normally it would not be included in a
randomization site. Nevertheless, we were curious to see if
amino acid substitutions such as threonine would improve
catalyst performance.
Following the generation and screening of the libraries
resulting from randomization at sites A and D at the alcohol
binding pocket, mutants were identified showing somewhat
improved activity and enantioselectivity in the model reaction
(Table 1). For example, variant SG101 (A281G/A282V) with
Table 1. Improved First Generation CALB Mutants
Obtained by Saturation Mutagenesis at Sites A and D
Aligning the Alcohol Binding Pocket, with NDT Codon
Degeneracy Being Employed
Figure 2. Iterative CASTing in the evolution of R- and S-
enantioselective CALB mutants as catalysts in the hydrolytic kinetic
resolution of rac-1.
reaction time conversion
eep
E
E
mutant
(h)
(%)
(%) (R) (S)
sequence
W104C
W104I
1.2 WT
RG101
RG102
WT
6.5
6.5
20
1
29
28
41
45
62
39
11
57
5
3
to perform saturation mutagenesis at site B next to the acyl
binding pocket; this ISM-branch provided notably improved
variants RG311 (E = 29) and RG301 (E = 40.) (Figure 2 and
Table 2). Only one of the pathways was continued, namely, A
→ D → B → E, which provided variant RG401 (E = 42). In the
case of S-selectivity, a likewise limited number of mutagenesis
experiments were performed, and the short pathway D → C →
B resulted in the best variant SG301 with E = 72 (Figure 2 and
Table 3). An attempt to continue on this pathway by
considering site E failed to induce further improvement. The
total screening effort involved the evaluation of less than 10 000
transformants, which is a good score considering the fact that
two different catalysts of opposite enantioselectivity were
evolved. Upscaling the model reaction of rac-1 (40 mg)
catalyzed by mutant SG301 also proved to be successful
(Supporting Information).
Substrate Scope of S- and R-Selective CALB Mutants.
Although no (bio)catalyst can ever be “universal”, organic
chemists seek to develop catalysts that function well for a range
of different substrates. Therefore, without performing any more
mutagenesis experiments, we tested some of the best R- and S-
selective CALB mutants evolved for the model substrate rac-1
as catalysts in the hydrolytic kinetic resolution of other racemic
esters 4−13 (Table 4). Upon employing the best mutants
SG301 and RG401, remarkably enhanced activity and
selectivity relative to WT CALB were observed in many cases
SG101
6
A281G/
A282 V
SG102
1
42
19
1.6 A281G
enhanced S-selectivity (E = 6) was found in library D, and two
mutants displaying reversed enantioselectivity in favor of (R)-2
occurred in library A. Interestingly, the genetic code of Ser105
in these two mutants was changed from TCC to AGT, but the
amino acid is still serine, which is the essential amino acid in the
catalytic triad. Rather than repeating saturation mutagenesis at
sites B, C, and E using traditional NNK codon degeneracy that
would ensure a complete amino acid alphabet of 20 canonical
members and therefore entail higher structural diversity, we
decided to use the best R- and S-selective mutants (RG101 and
SG101, respectively) as templates for saturation mutagenesis at
the other sites.
Iterative Rounds of Saturation Mutagenesis. When
considering an ISM system consisting of five sites, 120 different
pathways are theoretically possible.21 Rather than embarking on
a systematic exploration of all pathways,24 the goal of the
present study was to see how far one can get by restricting the
number of options severely but still achieving good R- and S-
selectivity in separate pathways. Therefore, the decision was
made to continue the evolutionary process along two arbitrarily
1875
dx.doi.org/10.1021/ja310455t | J. Am. Chem. Soc. 2013, 135, 1872−1881

Journal of the American Chemical Society
Article
experiments would be necessary. Some evolved mutants
showing lower enantioselectivity than the two best hits as
catalysts in the reaction of rac-1 proved to be superior catalysts
in the hydrolysis of the other substrates as well (Tables S1−S6
in the Supporting Information).
Table 2. Second, Third, and Fourth Generation R-Selective
CALB Mutants as Catalysts in the Hydrolytic Kinetic
Resolution of rac-1
reaction conversion eep
E
mutant
time (h)
(%)
(%) (R)
sequence
Enantiocomplementarity in the Hydrolytic Kinetic
Resolution of the Acetate of a Chiral Secondary Alcohol.
As already mentioned, the primary application domain of WT
CALB involves the stereoselective acylation of chiral secondary
alcohols or the hydrolytic reverse reaction, which in numerous
cases, hardly needs improvement. Nevertheless, we were
interested to see how some of the best mutants evolved for
rac-1 perform when used as catalysts in the hydrolytic kinetic
resolution of acetate rac-14 with formation of the R- or S-
alcohol 15, respectively (Scheme 3).
There was no stringent reason to expect analogous
stereocomplementarity for this substrate or in fact any degree
of stereoselectivity. We were therefore pleased to discover that
some of the CALB mutants which are S-selective for rac-1 show
in the reaction of the acetate rac-14 high enantioselectivity,
comparable to WT CALB (Table 5). Moreover, reversal of
enantioselectivity in favor of the alcohol (S)-15 is possible using
mutants that were again evolved for rac-1, although only up to
an E value of 23 (Table 5).
Kinetic Studies. Although the UV/visible-based pretest
indicated qualitatively that the CALB mutants are more active
than the WT, it was important to obtain precise kinetic data.
For this purpose, a select group of purified mutants were used
as catalysts in the hydrolysis of separately synthesized R- and S-
substrates 1, 4, 6, and 11, thereby leading to E values based on
kinetic data that may differ slightly from the values obtained by
the Sih equation.25 The results summarized in Table 6 reveal
some remarkable trends. For example, upon going from WT
CALB to mutant SG303, kcat/Km increases drastically from 165
to 45 800 M−1 s−1. In the case of S-selectivity in reactions of
substrates 4, 6, and 11, a similar phenomenon occurs. When
aiming for reversal of stereoselectivity in favor of the R-
configurated products, rate enhancement relative to the WT is
not as pronounced. Nevertheless, in the case of the model
compound (rac-1), the R-selective mutant (RG401) has a 15-
fold higher activity for the R-enantiomer and a 2.6-fold lower
activity for the S-enantiomer relative to WT CALB. Further
kinetic studies were performed for substrates 5, 7, 8, 10, and 12
(see the Supporting Information, Table S8)
RG201
RG202
RG204
RG301
7
8
7
9
29
30
40
45
79
83
74
90
12 W104C/A281C/A282F
15 W104C/A281V/A282L
11 W104C/A281V/A282V
40 W104C/L144Y/V149I/
A281C/A282F
RG302
RG303
RG305
RG308
RG309
RG310
RG311
RG313
RG401
4
8
4
4
6
6
6
4
3
32
39
38
34
40
34
52
33
49
87
88
87
88
86
88
82
86
89
21 W104C/L144Y/A281C/
A282F
27 W104C/V149D/A281C/
A282F
24 W104C/L144Y/V149D/
A281C/A282F
23 W104C/L144Y/V149H/
A281C/A282F
23 W104C/L144V/V149G/
A281V/A282L
24 W104C/L144F/V149C/
A281V/A282L
29 W104C/144F/149G/
A281V/A282L
20 W104C/L144F/V149L/
A281C/A282F
42 W104C/L144Y/V149I/
V154I/A281C/A282F
Table 3. Second and Third Generation S-Selective CALB
Mutants as Catalysts in the Kinetic Resolution of rac-1
reaction
time (h)
conversion eep
E
mutant
SG201
(%)
(%) (S)
sequence
3,5
49
92
84
78
71 I189V/V190C/A281G/
A282V
SG202
SG203
1
48
43
27 I189V/V190L/A281G/
A282V
12
15 I189C/V190C/A281G/
A282V
SG204
SG205
1
49
18
90
31
56 V190C/A281G/A282V
20
2
I189M/V190W/
A281G/A282V
SG206
SG301
8
1
47
50
50
91
5
I189V/A281G/A282V
72 L144Q/I189V/V190C/
A281G/A282V
SG302
SG303
1
1
39
45
91
93
40 L144R/I189V/V190C/
A281G/A282V
62 V149D/I189V/V190C/
A281G/A282V
Thermostability of Evolved CALB Mutants. The
introduction of point mutations in an enzyme may lead to a
decrease in thermostability.20,26 While performing the above
experiments, we noticed qualitatively that the mutants appear
to retain the thermostability characteristic of WT CALB. In
order to check quantitatively whether the thermostability of the
CALB variants has suffered to any notable degree as a
(Table 4). For example, good to excellent results were observed
in the cases of substrates rac-4, rac-5, rac-11, and rac-12, with
both R- and S-selectivity being possible in a stereocomple-
mentary manner. Noteworthy are also the reactions of
substrates lacking aryl groups in the respective carboxylic
acids, for example, rac-12 in which the R- and S-selective
mutants are capable of differentiating effectively between ethyl
and n-butyl substituents at the stereogenic center. When
comparing the ethyl ester rac-13 with the respective model p-
nitrophenyl ester rac-1 used in the screening process, only R-
selectivity is accessible. Other limitations were also observed.
For example, methyl or isobutyl groups in the para-position of
the phenyl moiety (rac-5, rac-6) are tolerated with regard to
substrate acceptance, but only moderate enantioselectivity
resulted (Table 4, entries 6−14). In the case of more bulky
substitutents in the para-position, unacceptably low activity was
observed. In these and other cases, additional mutagenesis
45
consequence of mutagenesis, the T50 values were measured
for the best mutants, which is the temperature at which 50%
residual activity is maintained following a heat treatment of 45
min.26 The measured T5045 values of WT CALB (56.7 °C) and
those of the best mutants SG201 (55 °C) and RG401 (55.2
°C) show that the evolved mutations responsible for enhanced
activity and stereoselectivity do not impair the thermostability
of the enzyme to any significant degree.
Unveiling the Source of Inverted Enantioselectivity.
In order to obtain some insight regarding the source of inverted
enantioselectivity induced by the ISM in the stereoselective
hydrolysis of rac-1, molecular dynamics (MD) simulations and
1876
dx.doi.org/10.1021/ja310455t | J. Am. Chem. Soc. 2013, 135, 1872−1881

Journal of the American Chemical Society
Article
Table 4. Hydrolytic Kinetic Resolution of Different Substrates Using R- and S-Selective CALB Mutants Originally Evolved for
rac-1
a
b
Mean value of 2−3 reactions. Best mutant for the specified substrate.
docking using GROMACS27 and Autodock 4.028 were
performed, respectively, for the evolved mutants SG303 and
RG401 that are two of the best variants observed in the present
study. The X-ray crystal structure of CALB lipase (PDB code
1TCA)8 was utilized as a starting point, and amino acid
mutations were introduced using the PyMOL.99rc6 program.29
1877
dx.doi.org/10.1021/ja310455t | J. Am. Chem. Soc. 2013, 135, 1872−1881

Journal of the American Chemical Society
Article
and RG401 were observed. In mutant SG303, the bottom
region of the binding site surrounded by C190, V189, T138,
L140, and Q157 (referred to as cavity A) is clearly enlarged
relative to the respective space in RG401 and the WT. In
Figure 3, the respective distances between the above-mentioned
Scheme 3. Stereoselective Hydrolysis of rac-14 Catalyzed by
CALB Mutants
Table 5. Hydrolytic Kinetic Resolution of rac-14 with
Different R- and S-Selective CALB Mutants Originally
Evolved for rac-1
reaction conversion
eep
E
(S)
a
a
mutant
mutation
time (h)
(%)
(%) E (R)
WT
−
12
12
50
47
>99
>99
>300
SG303
V149D/V190C/
I189V/A281G/
A282V
>300
SG204
V190C/A281G/
A282V
12
30
98
126
RG105 W104G
RG104 W104L
RG103 W104M
RG101 W104C
11
11
11
11
11
49
50
44
44
50
98
87
56
21
81
>300
40
7
2
RG309 W104C/L144V/
V149G/A281V/
A282L
23
Figure 3. (a) Equilibrated conformer of S-selective mutant SG303. (b)
Equilibrated conformer of R-selective mutant RG401. Some important
residues surrounding the acyl binding sites and catalytic triad (S105-
H224-D187) are shown as sticks. Distances defining the dimension of
cavity A and crevice B are labeled as red or black dotted lines. Three
groups of H bonds indicated by green dotted lines occur between
C190 and A132 (SG303), D149 and K290 (SG303), and Q191 and
V139 (RG401).
RG310 W104C/L144F/
V149C/A281V/
A282L
11
47
82
22
a
Mean values are available in the Supporting Information.
First, 4 ns MD simulations were carried out for WT CALB,
SG303, and RG401, and their equilibrated conformers were
extracted for structure analysis. Remarkable differences in the
space dimension of the acyl binding pocket of variants SG303
Table 6. Kinetic Data of WT CALB and Different Mutants as Catalyst in the Hydrolysis of Both Enantiomers of Substrates 1, 4,
6, and 11
enzyme
Km (mM)
kcat (s−1
)
kcat/Km (M−1 s−1
)
Km (mM)
kcat (s−1
)
kcat/Km (M−1 s−1
)
E (R)
E (S)
(R)-1
1.52
(S)-1
0.36
RG401
WT
0.20
0.62
2.06
7600
161
1.78
5.71
0.17
200
38
0.10
0.94
165
1
SG303
1.48
718
7.65
45 800
64
(R)-4
0.35
(S)-4
0.03
RG401
WT
0.12
1.81
2.27
2920
10
0.69
6.19
0.37
43
68
5
0.02
0.33
53
5
SG303
0.04
17
0.26
702
41
(R)-6
0.04
(S)-6
0.11
RG401
WT
0.05
0.84
0.80
800
59
0.66
0.57
0.41
167
59
0.05
0.03
1
SG303
0.14
175
0.81
1980
11
(R)-11
0.07
(S)-11
0.06
RG401
WT
0.08
0.43
1.27
875
279
940
0.86
0.46
0.93
70
13
0.12
0.23
500
2
SG303
1.20
43.81
47 110
50
1878
dx.doi.org/10.1021/ja310455t | J. Am. Chem. Soc. 2013, 135, 1872−1881

Journal of the American Chemical Society
Article
residues in SG303 are about 2−3 Å larger than those
corresponding distances in RG401. This conformational change
is induced by the presence of a new H bond involving C190−
SH···OC−A132 in SG303 (Figure 3a), which is different
from the existent H bond of Q191−CO···HN−V139 in
RG401 (Figure 3b). Furthermore, mutation I189V in variant
SG303 also contributes to the enlargement of the cavity. In
contrast, in variant RG401, the H bond of Q191−C
O···HN−V139 induces two effects that lead to the size
reduction of cavity A: (1) movement of loop 187−193 closer
to loop 136−140; (2) movement of residue I189 closer to helix
280−286. Moreover, mutation A282F also leads to a decrease
in the size of cavity A. In variant SG303, a different crevice
defined by several residues located at the top of the whole
binding site, such as L144, V154, D149, I285, and V286
(referred to as crevice B), is reduced in size. A completely
opposite change occurs in the respective crevice of variant
RG401. Thus the acyl binding sites in SG303 and RG401 have
been reshaped considerably, leading to different binding modes
as shown in Figures 4a and S3 (see the Supporting
Information), respectively. In SG303, the large group (phenyl)
at the stereogenic center in the acid moiety of substrate 1 binds
preferentially in cavity A eventually resulting in S-selectivity. In
contrast, crevice B in mutant RG401 is comparatively large
enabling the accommodation of the phenyl group; this is the
ultimate reason for reversed R-selectivity. Point mutation
W104C also contributes to the reversal of stereoselectivity
because the alcohol moiety (p-nitrophenyl ester) of substrate
(R)-1 would undergo severe steric clash with W104.
Next, the equilibrated conformers of the mutants were used
in docking experiments involving substrates (R)-1 and (S)-1.
The energetically favorable poses of substrates (R)-1 and (S)-1
binding to the reshaped active sites of mutants SG303 and
RG401 were extracted from the results of applying Autodock
4.0. In the case of the SG303 mutant, the acyl moiety of the
favored substrate (S)-1 is readily accommodated in the
enlarged cavity surrounded by V189, T138, L140, and Q157;
both the large substituent (phenyl group) and the medium-
sized substituent (methyl group) fit in a specified manner. All
the hydrogen bonds required for catalysis including the normal
stabilization of the oxyanion can occur readily (Figure 4b). In
sharp contrast, the slow-reacting enantiomer (R)-1 substrate
cannot undergo any favorable stabilization of the oxyanion due
to steric reasons (Figure 4c). In the R-selective mutant RG401,
the reduced space of cavity A cannot accept the whole acyl
moiety of substrate (R)-1, which forces the large substituent
(phenyl group) to be accommodated in the enlarged crevice B.
Compared to the disfavored substrate (S)-1, enantiomeric (R)-
1 in variant RG401 experiences the necessary H bonds for
smooth catalysis, and at the same time, the above-mentioned
distances for nucleophilic attack are within an appropriate range
(Figure S3b,c in the Supporting Information). These findings
are in full accord with the results of the kinetic study (Table 6).
CONCLUSIONS
■
The lipase from C. antarctica B (CALB) is one of the most
important enzymes in synthetic organic chemistry and
biotechnology. It is readily expressed in large quantities and
widely used in both academic and industrial applications.1,2
Although CALB has remarkably high stereoselectivity toward
many chiral secondary alcohols and various amines, its ability to
catalyze stereoselective reactions of α-substituted carboxylic
acid esters is limited.
Using structure-based saturation mutagenesis in an iterative
manner (ISM)20n,21 coupled with a reduced amino acid
alphabet resulting from NDT codon degeneracy,21,22 we were
able to evolve both R- and S-selective mutants as catalysts for a
variety of structurally different α-chiral esters. Both stereo-
selectivity and activity were dramatically enhanced. We
discovered that targeting randomization sites not only at the
acid binding pocket but also at the alcohol binding pocket leads
to success. Thus, we conclude, inter alia, that the traditional
strategy of protein engineers to focus on the alcohol binding
site when considering the reaction of chiral alcohols and to
focus on the acid binding pocket when working with the esters
of chiral carboxylic acids may be successful in some cases, but it
diminishes the probability of success. Indeed, long-range effects
of point mutations have been reported in other cases.20
Therefore, we recommend not to adhere to the long-standing
practice of dissecting the total binding pocket of a lipase into
two separate binding domains.
Figure 4. (a) Optimized poses of substrates (S)-1 and (R)-1 in S-
selective mutant SG303. (S)-1 and (R)-1 are shown as yellow and
purple sticks, respectively. S105, H224, and D187 define the catalytic
triad. (b) Necessary H bonds for catalysis and the stabilization of the
oxyanion indicated by green dotted lines occur in the complex of
SG303 and (S)-1; the distances for effective nucleophilic attack of
serine in the catalytic triad at the carbonyl function of (S)-1 are labeled
as black dotted lines. (c) Oxyanion-stabilizing H bonds in the
nonproductive complex of SG303 and (R)-1 are lost, and the distance
between catalytically active serine and the carbonyl function is too
long.
In the present case, insight regarding the source of
stereoselectivity was developed using MD simulations and
1879
dx.doi.org/10.1021/ja310455t | J. Am. Chem. Soc. 2013, 135, 1872−1881

Journal of the American Chemical Society
Article
Bruijn, R.; Vekemans, J. A. J. M.; Palmans, A. R. A; Meijer, E. W.
docking experiments, especially in the case of inverted
enantioselectivity. A reasonable model resulted from these
efforts. On the practical side, the enantioselective mutants
essentially retained the thermostability of WT CALB. Thus, the
range of CALB applications has been expanded considerably.
Macromolecules 2006, 39, 5021−5027.
́
(4) (a) Alatorre-Santamaría, S.; Gotor-Fernandez, V.; Gotor, V. Eur.
J. Org. Chem. 2011, 6, 1057−1063. (b) Patterson, L. D.; Miller, M. J. J.
Org. Chem. 2010, 75, 1289−1292. (c) Santaniello, E.; Casati, S.;
Ciuffreda, P. Curr. Org. Chem. 2006, 10, 1095−1123.
(5) (a) Pam
(b) Pam
(c) Edin, M.; Steinreiber, J.; Backvall, J.-E. Proc. Natl. Acad. Sci. U.S.A.
̀
ies, O.; Backvall, J.-E. Chem. Rev. 2003, 103, 3247−3261.
̈
ASSOCIATED CONTENT
* Supporting Information
■
̀
ies, O.; Backvall, J.-E. Trends Biotechnol. 2004, 22, 130−135.
̈
S
̈
2004, 101, 5761−5766. (d) Deska, J.; Ochoa, C. D.; Backvall, J.-E.
̈
Experimental details including library generation, expression
and screening of libraries, enzyme purification and kinetic
measurements, kinetic resolution of substrates rac-1−14,
protocol for upscaling the model reaction using substrate rac-
1, determination of thermostability, and molecular modeling
and MD simulation; all details and data of the hydrolytic kinetic
resolution of rac-4−14 catalyzed by WT CALB and a series of
mutants; kinetic data of WT CALB and mutant-catalyzed
hydrolysis of racemic substrates 5, 7, 8, 10, 12; primers used in
this work; chiral GC separation conditions; SDS-PAGE of pure
WT CALB, RG 401, and SG 303; equilibrated conformer of
WT CALB after MD simulation; optimized poses of substrate
(R)-1 and of (S)-1 in R-selective mutant RG401. This material
is available free of charge via the Internet at http://pubs.acs.org.
Chem.Eur. J. 2010, 16, 4447−4451.
(6) (a) Kim, M.-J.; Kim, H. M.; Kim, D.; Ahn, Y.; Park, J. Green
Chem. 2004, 6, 471−474. (b) Kim, M.-J.; Kim, W.-H.; Han, K.; Choi,
Y. K.; Park, J. Org. Lett. 2007, 9, 1157−1159. (c) Ko, S.-B.; Baburaj, B.;
Kim, M.-J.; Park, J. J. Org. Chem. 2007, 72, 6860−6864.
(7) Promiscuous reactions catalyzed by CALB: (a) Branneby, C.;
Carlqvist, P.; Hult, K.; Brinck, T.; Berglund, P. J. Mol. Catal. B: Enzym.
2004, 31, 123−128. (b) Branneby, C.; Carlqvist, P.; Magnusson, A.;
Hult, K.; Brinck, T.; Berglund, P. J. Am. Chem. Soc. 2003, 125, 874−
875. (c) Carlqvist, P.; Svedendahl, M.; Branneby, C.; Hult, K.; Brinck,
T.; Berglund, P. ChemBioChem 2005, 6, 331−336. (d) Svedendahl, M.;
Hult, K.; Berglund, P. J. Am. Chem. Soc. 2005, 127, 17988−17989.
(e) Carlqvist, P.; Eklund, R.; Hult, K.; Brinck, T. J. Mol. Model. 2003,
9, 164−171. (f) Carboni-Oerlemans, C.; Domínguez de María, P.;
Tuin, B.; Bargeman, G.; van der Meer, A.; van Gemert, R. J. Biotechnol.
2006, 126, 140−151. (g) Rustoy, E. M.; Sato, Y.; Nonami, H.; Erra-
Balsells, R.; Baldessari, A. Polymer 2007, 48, 1517−1525. (h) Sharma,
U. K.; Sharma, N.; Kumar, R.; Kumar, R.; Sinha, A. K. Org. Lett. 2009,
11, 4846−4848. (i) Wu, Q.; Liu, B.-K.; Lin, X.-F. Curr. Org. Chem.
2010, 14, 1966−1988.
AUTHOR INFORMATION
Corresponding Author
reetz@mpi-muelheim.mpg.de
■
Author Contributions
(8) (a) Uppenberg, J.; Hansen, M. T.; Patkar, S.; Jones, T. A.
Structure 1994, 2, 293−308. (b) Uppenberg, J.; Oehrner, N.; Norin,
M.; Hult, K.; Kleywegt, G. J.; Patkar, S.; Waagen, V.; Anthonsen, T.;
Jones, T. A. Biochemistry 1995, 34, 16838−16851. (c) Otto, R. T.;
Scheib, H.; Bornscheuer, U. T.; Pleiss, J.; Syldatk, C.; Schmid, R. D. J.
Mol. Catal. B: Enzym. 2000, 8, 201−211.
This manuscript was written through contributions of all
authors.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
(9) Kazlauskas, R. J.; Weissfloch, A. N. E.; Rappaport, A. T.; Cuccia,
L. A. J. Org. Chem. 1991, 56, 2656−2665.
■
We thank M. Hermes for the synthesis of all chiral p-
nitrophenyl esters (1−14) and H. Hinrichs for high-perform-
ance liquid chromatography analyses. Q.W. acknowledges the
visiting scholarship from “Future Academic Stars” project of
Zhejiang University and China Scholarship Council. This work
was financed by the Max-Planck-Society.
(10) Kirk, O.; Bjorkling, F.; Godtfredsen, S.; Ostenfeld Larsen, T.
̈
Biocatal. Biotransform. 1992, 6, 127−134.
(11) (a) Arroyo, M.; Sinisterra, J. V. J. Org. Chem. 1994, 59, 4410−
4417. (b) Hollmann, F.; Grzebyk, P.; Heinrichs, V.; Doderer, K.;
Thum, O. J. Mol. Catal. B: Enzym. 2009, 57, 257−261. (c) Klomp, D.;
Peters, J. A.; Hanefeld, U. Tetrahedron: Asymmetry 2005, 16, 3892−
3896. (d) Ong, A. L.; Kamaruddin, A. H.; Bhatia, S.; Aboul-Enein, H.
Y. J. Sep. Sci. 2008, 31, 2476−2485. (e) Wang, P.-Y.; Chen, Y.-J.; Wu,
A.-C.; Lin, Y.-S.; Kao, M.-F.; Chen, J.-R.; Ciou, J.-F.; Tsai, S.-W. Adv.
Synth. Catal. 2009, 351, 2333−2341. (f) Morrone, R.; D’Antona, N.;
Lambusta, D.; Nicolosi, G. J. Mol. Catal. B: Enzym. 2010, 65, 49−51.
(g) Juhl, P. B.; Doderer, K.; Hollmann, F.; Thum, O.; Pleiss, J. J.
Biotechnol. 2010, 150, 474−480.
REFERENCES
■
(1) Reviews of lipases in the production of fine chemicals:
(a) Schmid, R. D.; Verger, R. Angew. Chem., Int. Ed. 1998, 37,
1608−1633. (b) Jaeger, K.-E.; Dijkstra, B. W.; Reetz, M. T. Annu. Rev.
Microbiol. 1999, 53, 315−351. (c) Bornscheuer, U. T.; Kazlauskas, R. J.
Hydrolases in Organic Synthesis: Regio- and Stereoselective Biotransfor-
mations; Wiley-VCH: Weinheim, Germany, 1999. (d) Naik, S.; Basu,
A.; Saikia, R.; Madan, B.; Paul, P.; Chaterjee, R.; Brask, J.; Svendsen, A.
J. Mol. Catal. B: Enzym. 2010, 65, 18−23. (e) Sharma, D.; Sharma, B.;
Shukla, A. K. Biotechnology 2011, 10, 23−40. (f) Baldessari, A. Methods
Mol. Biol. 2012, 861, 445−456. (g) Andualema, B.; Gessesse, A.
Biotechnology 2012, 11, 100−118. (h) Houde, A.; Kademi, A.; Leblanc,
D. Appl. Biochem. Biotechnol. 2004, 118, 155−170. (i) Song, X.; Qi, X.;
Hao, B.; Qu, Y. Eur. J. Lipid Sci. Technol. 2008, 110, 1095−1101.
(12) (a) Evans, A. M. Eur. J. Clin. Pharmacol. 1992, 42, 237−256.
(b) Davies, N. M. Clin. Pharmacokinet. 1998, 34, 101−154.
(13) (a) Shiina, I.; Nakata, K.; Ono, K.; Onda, Y.; Itagaki, M. J. Am.
Chem. Soc. 2010, 132, 11629−11641. (b) Yang, X.; Birman, V. B.
Chem.Eur. J. 2011, 17, 11296−11304.
(14) (a) Patkar, S.; Vind, J.; Kelstrup, E.; Christensen, M. W.;
Svendsen, A.; Borch, K.; Kirk, O. Chem. Phys. Lipids 1998, 93, 95−101.
(b) Zhang, N. Y.; Suen, W.-C.; Windsor, W.; Xiao, L.; Madison, V.;
Zaks, A. Protein Eng., Des. Sel. 2003, 16, 599−605. (c) Patkar, S.;
Svendsen, A.; Kirk, O.; Clausen, I. G.; Borch, K. J. Mol. Catal. B:
Enzym. 1997, 3, 51−54. (d) Chodorge, M.; Fourage, L.; Ullmann, C.;
(j) Paravidino, M.; Groger, H.; Hanefeld, U. In Enzyme Catalysis in
̈
Organic Synthesis; Drauz, K., Groger, H., May, O., Eds.; Wiley-VCH:
̈
Weinheim, Germany, 2012; pp 251−362.
(2) Kirk, O.; Christensen, M. W. Org. Prog. Res. Dev. 2002, 6, 446−
451.
Duvivier, V.; Masson, J.-M.; Lefev
1022−1026. (e) Costa, L.; Brissos, V.; Lemos, F.; Ramo
̀
re, F. Adv. Synth. Catal. 2005, 347,
a Ribeiro, F.;
̂
Cabral, J. M. S. Bioprocess Biosyst. Eng. 2009, 32, 53−61. (f) Suen, W.-
C.; Zhang, N.; Xiao, L.; Madison, V.; Zaks, A. Protein Eng., Des. Sel.
2004, 17, 133−140.
(3) Examples of CALB-catalyzed polymerization: (a) Gross, R. A.;
Kumar, A.; Kalra, B. Chem. Rev. 2001, 101, 2097−2124. (b) Kobayashi,
S.; Uyama, H.; Kimura, S. Chem. Rev. 2001, 101, 3793−3818.
(c) Takwa, M.; Wittrup Larsen, M.; Hult, K.; Martinelle, M. Chem.
Commun. 2011, 47, 7392−7394. (d) van der Mee, L.; Helmich, F.; de
(15) (a) Qian, Z.; Fields, C. J.; Lutz, S. ChemBioChem 2007, 8,
1989−1996. (b) Qian, Z.; Horton, J. R.; Cheng, X. D.; Lutz, S. J. Mol.
1880
dx.doi.org/10.1021/ja310455t | J. Am. Chem. Soc. 2013, 135, 1872−1881

Journal of the American Chemical Society
Article
Biol. 2009, 393, 191−201. (c) Qian, Z.; Lutz, S. J. Am. Chem. Soc.
(28) Morris, G. M.; Goodsell, D. S.; Halliday, R. S.; Huey, R.; Hart,
W. E.; Belew, R. K.; Olson, A. J. J. Comput. Chem. 1998, 19, 1639−
1662.
(29) DeLano, W. L. The PyMOL Molecular Graphics System; DeLano
Scientific: San Carlos, CA, 2002; http://www.pymol.org.
́
2005, 127, 13466−13467. (d) Syren, P.-O.; Lindgren, E.; Hoeffken, H.
W.; Branneby, C.; Maurer, S.; Hauer, B.; Hult, K. J. Mol. Catal. B:
Enzym. 2010, 65, 3−10. (e) Liu, D.; Trodler, P.; Eiben, S.;
Koschorreck, K.; Muller, M.; Pleiss, J.; Maurer, S. C.; Branneby, C.;
̈
Schmid, R. D.; Hauer, B. ChemBioChem 2010, 11, 789−795.
(f) Wittrup Larsen, M.; Zielinska, D. F.; Martinelle, M.; Hidalgo, A.;
Juhl Jensen, L.; Bornscheuer, U. T.; Hult, K. ChemBioChem 2010, 11,
796−801.
(16) (a) Jung, S.; Park, S. Biotechnol. Lett. 2008, 30, 717−722.
(b) Budisa, N.; Wenger, W.; Wiltschi, B. Mol. BioSyst. 2010, 6, 1630−
1639. (c) Wittrup Larsen, M.; Bornscheuer, U. T.; Hult, K. Protein
Expression Purif. 2008, 62, 90−97. (d) Kim, S. Y.; Sohn, J. H.; Pyun, Y.
R.; Yang, I. S.; Kim, K. H.; Choi, E. S. J. Microbiol. Biotechnol. 2007, 17,
1308−1315.
(17) (a) Rotticci, D.; Rotticci-Mulder, J. C.; Denman, S.; Norin, T.;
Hult, K. ChemBioChem 2001, 2, 766−770. (b) Magnusson, A. O.;
Hult, K.; Holmquist, M. J. Am. Chem. Soc. 2001, 123, 4354−4355.
(c) Magnusson, A. O.; Rotticci-Mulder, J. C.; Santagostino, A.; Hult,
K. ChemBioChem 2005, 6, 1051−1056. (d) Magnusson, A. O.; Takwa,
M.; Hamberg, A.; Hult, K. Angew. Chem., Int. Ed. 2005, 44, 4582−
4585. (e) Vallin, M.; Syren
́
, P.-O.; Hult, K. ChemBioChem 2010, 11,
, P. O.; Hult, K.; Backvall,
411−416. (f) Engstrom, K.; Vallin, M.; Syren
́
̈
̈
J. E. Org. Biomol. Chem. 2011, 9, 81−82. (g) Marton, Z.; Leo
́
nard-
Nevers, V.; Syren, P.-O.; Bauer, C.; Lamare, S.; Hult, K.; Tranc, V.;
́
Graber, M. J. Mol. Catal. B: Enzym. 2010, 65, 11−17. (h) Svendsen, A.
Biochim. Biophys. Acta 2000, 1543, 223−238.
(18) Skjøt, M.; De Maria, L.; Chatterjee, R.; Svendsen, A.; Patkar, S.
A.; Østergaard, P. R.; Brask, J. ChemBioChem 2009, 10, 520−527.
́
(19) (a) Engstrom, K.; Nyhlen, J.; Sandstrom, A. G.; Backvall, J.-E. J.
̈ ̈
̈
Am. Chem. Soc. 2010, 132, 7038−7042. (b) Sandstrom, A. G.;
̈
Wikmark, Y.; Engstrom, K.; Nyhlen
́
, J.; Backvall, J.-E. Proc. Natl. Acad.
̈
̈
Sci. U.S.A. 2012, 109, 78−83.
(20) Recent reviews of directed evolution: (a) Lutz, S.; Bornscheuer,
U. T. Protein Engineering Handbook; Wiley-VCH: Weinheim,
Germany, 2009; Vol. 1−2. (b) Quin, M. B.; Schmidt-Dannert, C.
ACS Catal. 2011, 1, 1017−1021. (c) Turner, N. J. Nat. Chem. Biol.
2009, 5, 567−573. (d) Bommarius, A. S.; Blum, J. K.; Abrahamson, M.
J. Curr. Opin. Chem. Biol. 2011, 15, 194−200. (e) Dalby, P. A. Curr.
Opin. Struct. Biol. 2011, 21, 473−480. (f) Jackel, C.; Hilvert, D. Curr.
̈
Opin. Biotechnol. 2010, 21, 753−759. (g) Brustad, E. M.; Arnold, F. H.
Curr. Opin. Chem. Biol. 2011, 15, 201−210. (h) Bershtein, S.; Tawfik,
S. Curr. Opin. Chem. Biol. 2008, 12, 151−158. (i) Shivange, A. V.;
Marienhagen, J.; Mundhada, H.; Schenk, A.; Schwaneberg, U. Curr.
Opin. Chem. Biol. 2009, 13, 19−25. (j) Kumar, A.; Singh, S. Crit. Rev.
Biotechnol. 2012, DOI: 10.3109/07388551.2012.716810. (k) Siloto, R.
M. P.; Weselake, R. J. Biocatal. Agric. Biotechnol. 2012, 1, 181−189.
(l) Bolt, A.; Berry, A.; Nelson, A. Arch. Biochem. Biophys. 2008, 474,
318−330. (m) Wang, M.; Si, T.; Zhao, H. Bioresour. Technol. 2012,
115, 117−125. (n) Reetz, M. T. Tetrahedron 2012, 68, 7530−7548.
(21) Review of directed evolution of stereoselective enzymes20n with
emphasis on iterative saturation mutagenesis: Reetz, M. T. Angew.
Chem., Int. Ed. 2011, 50, 138−174.
(22) Reetz, M. T.; Carballeira, J. D. Nat. Protoc. 2007, 2, 891−903.
(23) Reetz, M. T.; Zonta, A.; Schimossek, K.; Liebeton, K.; Jaeger, K.-
E. Angew. Chem., Int. Ed. Engl. 1997, 36, 2830−2832.
(24) Gumulya, Y.; Sanchis, J.; Reetz, M. T. ChemBioChem 2012, 13,
1060−1066.
(25) Chen, C.-S.; Fujimoto, Y.; Girdaukus, G.; Sih, C. J. J. Am. Chem.
Soc. 1982, 104, 7294−7299.
(26) Reviews of protein engineering of thermostabiliity:22
(a) O’Fagain, C. Enzyme Microb. Technol. 2003, 33, 137−149.
(b) Eijsink, V. G. H.; Gaseidness, S.; Borchert, T. V.; van den Burg,
̊
B. Biomol. Eng. 2005, 22, 21−30. (c) Wintrode, P. L.; Arnold, F. H.
Adv. Protein Chem. 2001, 55, 161−225.
(27) (a) Berendsen, H. J. C.; van der Spoel, D.; van Drunen, R.
Comput. Phys. Commun. 1995, 91, 43−56. (b) van der Spoel, D.;
Lindahl, E.; Hess, B.; Groenhof, G.; Mark, A. E.; Berendsen, H. J. C. J.
Comput. Chem. 2005, 26, 1701−1718.
1881
dx.doi.org/10.1021/ja310455t | J. Am. Chem. Soc. 2013, 135, 1872−1881