Suppression of water as a nucleophile in Candida antarctica lipase B catalysis

Article abstract of DOI:10.1002/cbic.200900743

A water tunnel in Candida antarctica lipase B that provides the active site with substrate water is hypothesized. A small, focused library created in order to prevent water from entering the active site through the tunnel was screened for increased transacylation over hydrolysis activity. A single mutant, S47L, in which the inner part of the tunnel was blocked, catalysed the transacylation of vinyl butyrate to 20mM butanol 14 times faster than hydrolysis. The single mutant Q46A, which has a more open outer end of the tunnel, showed an incresed hydrolysis rate and a decreased hydrolysis to transacylation ratio compared to the wild-type lipase. Mutants with a blocked tunnel could be very useful in applications in which hydrolysis is unwanted, such as the acylation of highly hydrophilic compounds in the presence of water.

Full text of DOI:10.1002/cbic.200900743

DOI: 10.1002/cbic.200900743
Suppression of Water as a Nucleophile in Candida
antarctica Lipase B Catalysis
[a, b]
[a]
[a]
[c]
Marianne Wittrup Larsen,
Dorota F. Zielinska, Mats Martinelle, Aurelio Hidalgo,
[d]
[b]
[a]
Lars Juhl Jensen, Uwe T. Bornscheuer, and Karl Hult*
A water tunnel in Candida antarctica lipase B that provides the
active site with substrate water is hypothesized. A small, fo-
cused library created in order to prevent water from entering
the active site through the tunnel was screened for increased
transacylation over hydrolysis activity. A single mutant, S47L, in
which the inner part of the tunnel was blocked, catalysed the
transacylation of vinyl butyrate to 20 mm butanol 14 times
faster than hydrolysis. The single mutant Q46A, which has a
more open outer end of the tunnel, showed an increased hy-
drolysis rate and a decreased hydrolysis to transacylation ratio
compared to the wild-type lipase. Mutants with a blocked
tunnel could be very useful in applications in which hydrolysis
is unwanted, such as the acylation of highly hydrophilic com-
pounds in the presence of water.
Introduction
[
1]
Candida antarctica (also known as Pseudozyma antarctica ) li-
pase B (CALB) is a widely used biocatalyst in both research as
well as in industrial processes. Its substrate specificity, activity
and stability in nonaqueous environments are some of the key
proceeds through an acyl enzyme intermediate. Any water
present acts as a competing nucleophile and affords hydrolysis
instead of the desired transacylation (Scheme 1). To circumvent
water as a favored nucleophile, reactions are performed under
dry conditions in organic solvents or directly in bulk. However,
absolutely dry conditions can be difficult to maintain and are
costly on large scale. It would therefore be of great interest to
minimize CALB’s preference for water as a nucleophile.
[
2]
features that make CALB very useful. In many applications hy-
drolysis is an undesired side reaction. The reaction mechanism
Specific tunnels are present in many enzymes. Several en-
zymes with two active sites have developed tunnels that allow
[3]
easy passage of reactive intermediates between the sites. In
other enzymes, tunnels allow for small substrates such as
gases and hydrogen peroxide to enter the active site from the
[
4–6]
external medium.
Based on structural studies, tunnels for
[7]
water passage have been suggested for photosystem II, hal-
[8]
oalkane dehalogenase as well as a lipase from Thermomyces
[9]
lanuginosa.
In its natural environment the lipase binds with its active
site facing the hydrophobic substrate phase; this obstructs
[
a] Dr. M. Wittrup Larsen, D. F. Zielinska, Dr. M. Martinelle, Prof. K. Hult
Department of Biochemistry, School of Biotechnology
Royal Institute of Technology, AlbaNova University Center
1
06 91 Stockholm (Sweden)
Fax: (+46)8-5537-8468
E-mail: kalle@biotech.kth.se
[
b] Dr. M. Wittrup Larsen, Prof. U. T. Bornscheuer
Department of Biotechnology and Enzyme Catalysis
Institute of Biochemistry, Greifswald University
Felix-Hausdorff-Strasse 4, 17487 Greifswald (Germany)
[
c] Dr. A. Hidalgo
Centro de Biologꢀa Molecular “Severo Ochoa” (UAM-CSIC)
Nicolas Cabrera 1, 28049 Madrid (Spain)
Scheme 1. The scheme shows the competition between water and alcohol
for the acyl enzyme. In addition it shows the principle of the analytical
method used to detect hydrolysis and acyl transfer activity simultaneously.
The total rate was determined by the release of acetaldehyde when the acyl
enzyme was formed. Hydrolysis was measured by the release of protons
from formed acid and the transacylation was calculated from the difference
of the two.
[d] Prof. L. J. Jensen
Novo Nordisk Foundation Center for Protein Research
Faculty of Health Sciences, University of Copenhagen
2200 Copenhagen N (Denmark)
7
96
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2010, 11, 796 – 801

Suppression of Water in Lipase Catalysis
substrate water from entering. We hypothesize that surround-
ing water molecules instead enter the active site of the
enzyme through a specific tunnel in the protein structure. To
test the tunnel hypothesis, a small CALB library was designed.
Hydrophobic amino acids were introduced at positions that
would block the tunnel and disfavour the entrance of water
into the active site, whereby transacylation would be favored
over hydrolysis. This library was expressed in Escherichia coli
and screened for the ratio of transacylation over hydrolysis
activity with a newly developed method. Some mutants were
selected for further characterization.
the desired amino acids and to avoid the unwanted amino
acids. CALB clones were randomly picked for sequencing to
confirm mutations and the quality of the library. The distribu-
tion of mutations in 64 clones showed that 41 were mutated
in position 42, 27 in position 47 and 44 in position 106. Among
the clones there were 31 single mutants, 13 double mutants
and 20 triple mutants. Overall the designed spiked primers re-
sulted in a distribution of mutations in CALB that correspond-
ed well to the expected distribution (Table 1).
Table 1. Distribution of introduced amino acids among 64 sequenced
clones picked by random. Among the clones there were 31 single mu-
tants, 13 double mutants and 20 triple mutants.
Results and Discussion
Number of mutations
Position
Distribution
Average Expected
Primer and library design
Amino acid
42
47
106
Total
[%]
[%]
Three positions in the CALB protein structure, 42, 47 and 106,
were chosen as targets to diminish water transport through
the hypothesized tunnel (Figure 1). A library was designed to
introduce hydrophobic amino acids at these positions to pre-
vent water from accessing the tunnel. In the library design the
following amino acids were included: alanine, valine, leucine,
isoleucine, methionine, proline and threonine. Cysteine was
excluded to avoid the formation of wrong disulfide bridges.
Tryptophan was also not included because of its large size,
which would disrupt the protein structure.
A
L
M
P
T
V
E
F
sum
2
5
8
5
0
5
16
1
2
5
3
4
0
4
11
0
1
6
6
7
1
4
20
0
5
16
17
16
1
13
47
1
4
13
14
14
1
11
42
1
8
16
16
16
4
8
32
0
41
27
44
117
The design of libraries with amino acid subpopulations is
constrained by the genetic code; phenylalanine and glycine
were thus excluded to obtain a more even distribution over
Two unexpected mutations, one glutamic acid and one phe-
nylalanine, were also found. The library was contaminated with
the wild-type gene, and this can be explained by incomplete
digestion with DpnI.
Characterization of mutants
In total almost one thousand mutants were screened with the
microtiter plate assay. Based on the ratio of transacylation
activity to hydrolysis several clones were selected for further
studies. In addition, mutants were selected to represent all the
sites that were mutated.
Vinyl butyrate was hydrolyzed with a similar rate as tributyr-
10]
[
in by the wild-type enzyme; this means that the deacylation
step in the catalysis was rate limiting. The consequences of
blocking the transport of water to the active site would
depend on whether the transport of water into the active site
is rate limiting or if the water reactivity is rate limiting. If water
transport is the rate-limiting step, mutations blocking the hy-
pothesized tunnel would show a decreased hydrolytic activity.
An alternative nucleophile, such as butan-1-ol, which must
enter the active site through the normal substrate route, might
then be able to restore (or even increase) the total consump-
tion rate of the ester substrate by transacylation. If this does
not occur, then the mutation has only decreased the total effi-
ciency of the enzyme. The second scenario is that blocking the
water tunnel only affects the affinity of water to the active site.
In the normal situation without any competing nucleophile an
unchanged hydrolytic rate could still be observed as the
enzyme anyhow might be saturated by water. On the other
Figure 1. Lipase B from C. antarctica with bound Tween 80 (1 Lbt). The ester
function of Tween 80 (in line representation) can be seen through the pro-
posed water tunnel. The two ends of the Tween molecule can be seen on
top of the protein sticking out of the substrate binding cleft. The amino
acids Q46 and S47 are shown in yellow. Hydrophobic and hydrophilic resi-
dues are shown in grey and blue, respectively. The glycosylation is shown in
CPK colours.
ChemBioChem 2010, 11, 796 – 801
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chembiochem.org
797

K. Hult et al.
hand, a decrease of water affinity to the active site would
open up the competition of alternative nucleophiles and a
higher increase in transacylation compared to the wild-type
enzyme should be observed.
shows that the mutation had some effect on water as a nucle-
ophile.
The triple mutant T42L S47A Q106V was expressed at a too
low level to be characterized in detail. From these results it
was hard to draw any conclusions about position 42. On the
other hand, position 47 seemed to be much more important.
All single mutants at this position S47L, S47M and S47V had
hydrolysis rates similar to that of the wild-type enzyme in the
absence of butanol. In addition, a strong decrease in hydrolysis
with increasing concentration of butanol was seen. For all mu-
tants the transacylation activity increased with increasing buta-
nol concentration and it was dramatically high for S47L. The
total consumption of vinyl butyrate could in all cases be in-
creased by the addition of 20 mm butanol; this shows that the
deacylation is rate limiting for hydrolysis. In conclusion, S47
gave results consistent with an inlet to the active site that kept
the concentration of water high for catalysis.
After inspection of the hydrolytic and transacylation rates of
mutants picked from the original library with mutations at po-
sitions 42, 47 and 106, one could draw the following conclu-
sions (Figure 2): Mutations at position 106 resulted in mutants
After defining the end of the tunnel close to the active site
we performed further structure analyses. Q46 was a good can-
didate for an amino acid at the outer end of the tunnel as it is
surrounded by a cluster of water molecules in the crystal struc-
ture. Point mutation of Q46 to large hydrophobic amino acids,
which would block the entrance, failed to afford any protein.
The less drastic mutation Q46A, which opened up the tunnel,
gave a mutant with high hydrolytic activity that was sensitive
to butanol inhibition. In addition, butanol could not compete
as well with water as in the wild-type enzyme; this shows that
water activity in the active site of Q46A should be higher than
in that of the wild-type. In conclusion, Q46 is a good candidate
for the entrance of a water transport tunnel. The double
mutant Q46A S47L, which represents mutations both at the
inner and outer ends of the tunnel, was prepared. It showed a
strong inhibition of hydrolysis as well as a strong increase in
transacylation rate and kept the properties of the single muta-
tion S47L.
Conclusions
In a separate study we have detected water tunnels in hun-
dreds of hydrolases by using molecular modeling (unpublished
results). There are several good reasons for lipases to have a
separate entrance to their active site for water. Lipases bind to
the interface between water and their insoluble substrates,
and they face the substrate to let it pass into the active site
without passing through the water phase. At the same time
they hide the entrance from water, which thus has to pass
either through the hydrophobic phase or find an alternative
way. Here we propose the existence of a tunnel in CALB. From
our results it can be concluded that the transport of water is
not the rate-limiting step in hydrolysis. Instead, the rate-limit-
ing step seems to be in reaching the transition state after
water has bound. In a situation in which water still can satu-
rate the enzyme, no effect on hydrolysis would be seen from a
reduced transport rate. The situation is different as soon as
water has to compete with other nucleophiles, an effect we
could show with butanol. In its natural environment a lipase
will face the substrate phase with a subsequently increasing
Figure 2. Rates of A) hydrolysis and B) transacylation of vinyl butyrate in the
presence of butan-1-ol in MOPS-NaOH pH 7.2. Rates of transacylation were
calculated as the difference between aldehyde formation and hydrolysis as
depicted in Scheme 1. The results represent single measurements. Triplicate
analyses done different days for wild-type enzyme showed a standard devia-
tion of around 20% from the mean in all data points.
with low hydrolytic activity. No or little reduction of the hydro-
lytic activity was seen with increasing concentrations of buta-
nol; the transacylation to butanol did not increase substantial-
ly. It was concluded that position 106 could not be mutated to
increase transacylation over hydrolysis. No single mutants at
position 42 were found in the screening. The double mutant
T42A S47V had a specific hydrolytic activity four times lower
than that of the wild-type enzyme. On the other hand, the
transacylation activity decreased only by a factor of two; this
7
98
www.chembiochem.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2010, 11, 796 – 801

Suppression of Water in Lipase Catalysis
product concentration in the form of the leaving alcohol. In
this situation product inhibition would be severe if water
cannot reach the active site efficiently. We hypothesize that
this is the reason for the tunnel in CALB.
amplification, the reaction product (15 mL) was added along with
FastDigest DpnI (1 mL; Fermentas, MBI Fermentas, Munich, Germa-
ny) and FastDigest buffer (2 mL; Fermentas). This mixture was incu-
bated at 378C for 5 h to digest the parental DNA.
Directed mutagenesis: Position 46 in the CALB amino acid sequence
was subjected to directed mutagenesis by using pET-22b(+) CALB
as template DNA (10 ng), forward primer F-CALB Q46L and reverse
primer R-CALB Q46L (15 pmol of each), Phusion Hot start High-Fi-
delity DNA polymerase (1 U; Finnzymes Oy, Espoo, Finland), Phu-
sion HF buffer (1ꢁ; Finnzymes), dNTP (0.2 mm), DMSO (4 mL) and
milliQ water (to a total volume of 50 mL). Initial denaturation for
Experimental Section
Strains and media: Cloning of the CALB gene into pET22-b(+) and
its expression, which was controlled by the T7lac promoter by
using the pelB signal sequence for periplasmic secretion in the
[11]
E. coli strain Rosetta (DE3), are described elsewhere.
3
6
0 s at 988C was followed by 18 cycles of 10 s at 988C, 10 s at
58C, 6 min at 728C. Hereafter template DNA was digested by ad-
For cultivation of E. coli, Luria–Bertani (LB) medium (10 g tryptone,
dition of FastDigest DpnI (1 mL; Fermentas) followed by incubation
at 378C for 1.5 h to digest the parental DNA.
5
g yeast extract, 5 g NaCl, 1 mL of 1m NaOH in 1 L water) was
used whereas the rich Superbroth medium (32 g tryptone, 20 g
yeast extract, 5 g NaCl, 5 mL of 1m NaOH in 1 L water) was used
for the protein expression. Positive transformants were selected on
In addition a double mutant was created by using wild-type CALB
as template DNA. The reaction mixture conditions were in this case
as following: pET-22b(+) CALB (15 ng), forward primer F-Q46A
S47L and reverse primer R-Q46A S47L (10 pmol of each), Pfu poly-
merase (2.5 U; Fermentas), Pfu buffer+MgSO4 (5 mL; Fermentas),
dNTP (0.2 mm) and milliQ water (to a total volume of 50 mL). Initial
denaturation for 1 min at 958C was followed by 18 cycles of 1 min
at 958C, 1 min at 608C, 12 min and 30 s at 728C, followed by a
final extension for 15 min at 728C. Parental template DNA was di-
gested by the addition of FastDigest DpnI (1 mL) and FastDigest
buffer (5 mL; Fermentas) followed by an incubation at 378C for 4 h.
The digest was purified by using a PCR Purification Kit (Stratagene,
La Jolla, CA, USA) and the final product (1 mL) was used for trans-
formation.
LB plates prepared as LB medium supplemented with agar
À1
(
20 gL ). All media were supplemented with appropriate antibiot-
À1
ics, including chloramphenicol (Cam, 34 mgL ) and ampicillin
À1
(
Amp, 100 mgL ).
Primer design
Library: Three amino acid positions in the CALB sequence were tar-
geted, and spiked oligonucleotides were designed to reduce the
complexity of the library and hence facilitate the following screen-
ing procedure. A computational algorithm was used to optimize
the nucleotide mixtures to produce specified amino acid subpopu-
lations.
[12]
Two spiked forward primers were used, one for positions 42 and
Transformation to E. coli
4
7: F-CALB 42–47 (5’-CCG GAA CCG GCV YKA CAG GTC CAC AAV
YKT TCG ACT CGA A-3’) and another one for position 106: F-CALB
06 (5’-CGT GCT TAC CTG GTC TVY KGG TGG TCT GGT TGC AC-3’).
Electrocompetent E. coli cells (40 mL; Rosetta (DE3), Novagen) were
used for transformation directly with the digested PCR product
and plated on LBAmp,Cam plates after phenotypic expression at 378C
at 220 rpm for 1 h in LB media. Positive transformants appeared
after incubation overnight at 378C. The DNA sequences of the mu-
tated genes in the positive transformants were verified by DNA
sequencing.
1
In all three cases the targeted positions are underlined and the
mixtures of nucleotides were defined as: V(G=40.0%, A=40.0%,
C=20.0%); Y(T=80.2%, C=19.8%); K(G=49.8%, T=50.2%). Two
silent mutations were introduced (shown in bold), one in each
primer to minimize a critical DG value and thereby to disfavour
primer dimerisation and hairpin structure formation. Furthermore,
the primers were phosphorylated at the 5’-end to facilitate the
direct ligation to the linear amplification product.
Protein expression
96-deep-well plates: Single cell colonies of CALB variants that grew
on selection plates were used to inoculate 96-deep-well plates
Directed mutagenesis: Primers were designed to mutate posi-
tion 46: F-Q46A (5’-GGA ACC GGC ACC ACA GGT CCA GCG TCG
TTC GAC TCG-3’) and the reverse primer: R-Q46A (5’-CGA GTC GAA
CGA CGC TGG ACC TGT GGT GCC GGT TCC-3’). Moreover primers
for a double mutant Q46A S47L were designed: F-Q46A S47L (5’-
GGC ACC ACA GGT CCA GCG TTA TTC GAC TCG AAC TGG-3’) and
reverse primer: R-Q46A S47L (5’-CCA GTT CGA GTC GAA TAA CGC
TGG ACC TGT GGT GCC-3’). The point mutations are underlined in
all cases. Design was also used to minimise the binding energy of
hairpin and homodimer structures.
(
2.2 mL storage plate, ABgene House, Surrey, UK) containing Super-
brothAmp,Cam (1 mL). The plates were incubated at 220 rpm and
78C for maximum 16 h. ABgene gas-permeable adhesive seals
3
(
(
ABgene House) were used to seal the plates. This preculture
50 mL) was used to inoculate the main culture, which consisted of
SuperbrothAmp,Cam (1 mL) in 96-deep-well-plates (Bel-Art, Science-
ware, Pequannock, NJ, USA) that were incubated at 220 rpm and
3
78C until an OD600 0.5–0.8 (~1 to 2 h) was reached. These plates
have a lid that was used to seal the cultures and prevent cross
contamination. After incubation on ice for 15 min, isopropyl b-d-1-
thiogalactopyranoside (IPTG, final concentration of 0.1 mm) was
added and the plates were incubated at 220 rpm and 168C for
Mutagenesis
[13]
Multiple mutagenesis: Multiple mutagenesis reactions were per-
formed by using pET-22b(+) CALB template DNA (150 ng), spiked
primer F-CALB 42–47 and F-CALB 106 (10 pmoles), AccuPrime Pfx
DNA polymerase (1 U; Invitrogen, Carlsbad, CA, USA), AccuPrime
Pfx reaction mix (1ꢁ; Invitrogen), Ampligase (15 U; Epicentre, Mad-
2
4 h.
Small-scale cultures: Selected candidates from the screening were
cultivated in LB medium (100 mL) in baffled shake flasks (1 L). Glyc-
erol stocks (5 mL) were used to inoculate LBAmp,Cam, (10 mL). This
was then incubated for a maximum of 16 h at 378C and 220 rpm.
This preculture (1 mL) was inoculated to SuperbrothAmp,Cam
(100 mL) and incubated until the OD600 reached 0.5–0.8. Before in-
duction with IPTG (0.1 mm) the culture was cooled on ice and the
+
ison, WI, USA), NAD (0.5 mm) and in addition DMSO (8% (v/v)) in
a total volume of 25 mL. Initial denaturation for 5 min at 958C was
followed by amplification reaction performed during 35 cycles of
3
0 s at 958C, 30 s at 558C and 6 min 15 s at 658C. After this linear
ChemBioChem 2010, 11, 796 – 801
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chembiochem.org
799

K. Hult et al.
following protein expression was carried out for 24 h at 220 rpm
and 168C.
reaction time, hydrazone (165 mL, 2.5 mm 3-methyl-2-benzothiazoli-
nonehydrazone hydrochloride; TCI Europe) and an oxidizing solu-
tion (60 mL of 60 mm iron(III) chloride (BDH, Poole, UK) and 0.16m
sulfamic acid (Sigma–Aldrich)) dissolved in water were added. The
strongly acidic conditions stopped the enzyme-catalyzed reaction.
After 20 min (to allow for complete colour development) the ab-
sorbance at 720 nm gave the acetaldehyde concentration and the
total consumption of vinyl acetate. To verify the formation of ester,
incubation samples were extracted with diethyl ether. The ether
phase was quickly washed with sodium bicarbonate (0.5%) and
dried with anhydrous magnesium sulfate before being subjected
to gas chromatography.
Protein harvest
96-deep-well plates: CALB variants expressed in the periplasmic
space were recovered by osmotic shock. After 24 h expression the
cells were isolated by centrifugation at 3300 g, and 48C for 10 min.
The supernatant was carefully discarded. Thereafter, of Tris-HCl
buffer (765 mL; 30 mm Tris-HCl, 1 mm EDTA, 20% (w/v) sucrose,
pH 8.0) was added to the wells containing cell pellets. The plates
were incubated at room temperature on a shaker for 10 min. After
centrifugation (3300g, 48C for 10 min) the supernatant was dis-
carded and the cells were re-suspended in ice cold MgSO (5 mm;
4
Kinetics of enzyme variants: The kinetics of purified enzymes from
selected clones cultivated in shake flasks were analyzed in more
detail by using a spectrophotometer. The assay solution consisted
of 4-nitrophenol (1 mm), vinyl butyrate (8 mm diluted in acetoni-
trile; >97%, TCI Europe), butan-1-ol (final concentration 0–40 mm),
protein solution (0.05–0.1 mL) and MOPS-NaOH (10 mm, pH 7.2) to
a final volume of 1 mL. The reaction was followed continuously at
405 nm (4-nitrophenolate) and samples for analysis of acetalde-
hyde (50 mL) were taken after the absorbance in absence of buta-
nol had decreased 0.05 and 0.1 absorbance units, respectively.
Samples were taken at the same timepoints for incubations with
and without butanol. Oxidizing solution (200 mL; 4 mm iron(III)
chloride (BDH) and 16 mm sulfamic acid (Sigma–Aldrich), dissolved
in water) was immediately added to the samples for acetaldehyde
analysis to stop the reaction. This was followed by the addition of
hydrazone (800 mL, 5 mm; 3-methyl-2-benzothiazolinonehydrazone
hydrochloride) dissolved in water. The samples were incubated at
room temperature for 30 min for complete colour development.
The absorbance was hereafter measured at 720 nm. This wave-
length was chosen to decrease the sensitivity of the method.
3
00 mL). The plates were again incubated on a shaker at room tem-
perature for 10 min. Finally the periplasmic fraction containing the
protein was isolated by centrifugation (3300g, 48C for 10 min).
Small-scale cultures: Cells expressing lipase variants were cultivated
in shake flasks (100 mL) and were harvested after 24 h incubation
by centrifugation at 3300g and 48C for 15 min. The supernatant
was discarded and the cells slowly suspended in Tris-HCl buffer
(
10 mL; 30 mm Tris-HCl, 1 mm EDTA, 20% (w/v) sucrose, pH 8.0).
After incubation on an end-over-end rotor for 10 min and centrifu-
gation (3300 g, 48C for 15 min) the supernatant was discarded and
the cells were resuspended in ice cold MgSO (5 mm, 10 mL). After
4
a final incubation on an end-over-end rotor for 10 min and centri-
fugation (3300g, 48C for 15 min) the supernatant contained the
periplasmic fraction.
Lipase purification
Lipase expressed in small-scale cultures was purified before further
characterisation. His-tagged protein variants were captured by
using Immobilized Metal Affinity Chromatography (IMAC) as de-
[11]
scribed previously,
followed by a desalting step in order to
Standard curves for acid and aldehyde production were construct-
ed as follows: vinyl butyrate (0–8 mm) was added to the acid assay
solution. CALB was added and the hydrolysis was allowed to go to
completion. The final absorbance at 405 nm was used for the acid
standard curve. Samples (50 mL) were withdrawn and treated as
described above for the analysis of acetaldehyde. All incubations
were carried out so that only the linear parts of the standard
curves were used.
remove the imidazole used to elute the protein variants from the
IMAC columns. Protein fractions were loaded onto PD-10 desalting
columns (GE Healthcare Bio-Sciences AB, Uppsala, Sweden), and
eluted with MOPS (3.5 mL, 10 mm, pH 7.2).
Characterisation of enzyme variants
The enzyme variants were screened and analysed for hydrolytic
and acyl transfer activity in a combined new assay (Scheme 1). Hy-
drolysis of vinyl esters was detected by following the production
of acid with the pH indicator p-nitrophenol as is done in the
Vinyl butyrate was kept over activated molecular sieves (3 ꢃ) and
stored at +48C. Before use aliquots were purified on a silica
column to remove any free acid.
[14]
Quick-E method. The total reaction rate was detected by follow-
ing the production of acetaldehyde with 3-methyl-2-benzothiazoli-
[15]
Active-site titration: The active site inhibitor methyl 4-methylumbel-
nonehydrazone (TCI Europe N.V., Belgium). The transacylation re-
action rate was calculated as the difference between acetaldehyde
formation and acid formation.
[16]
liferyl hexylphosphonate was prepared as previously described.
Purified inhibitor (6 mmol) was lyophilized and stored at À208C in
sealed glass ampoules. The lyophilized inhibitor was diluted in
acetonitrile (luminescence quality, Fluka). The inhibition assay con-
tained Tris-HCl (ultra pure, Research organics, Cleveland, Ohio,
Library: Cells from 96-deep-wells plates were subjected to osmotic
shock. The isolated protein solution (36 mL) was transferred to a
microtiter plate (F 96, NUNC, Thermo Fisher, Roskilde, Denmark)
and the enzyme library was analysed as follows: solution contain-
ing butan-1-ol (36 mL, 5 mm; Riedel-de Haꢂn, Seelze, Germany) and
USA) buffer (700–800 mL, 100 mm, pH 8.0, 1 mm CaCl (Merck), puri-
2
fied protein (100–200 mL) and inhibitor (100 mL 5 mm in acetoni-
trile), total reaction volume was 1 mL.
4
-nitrophenol (1.3 mm; Sigma–Aldrich) in MOPS-NaOH (20 mm,
pH 7.2; Ducheta Biochemie, Haarlem, The Netherlands) were added
to the protein solution. The reaction was started by the addition of
vinyl acetate (3 mL, 200 mm, >99%, Aldrich) dissolved in acetoni-
trile (Carlo Erba Reactifs-Sds, Chaussee Du Vexin, France) to a final
concentration 8 mm. Acetic acid formation was followed at 405 nm
for 9 min by using a microplate reader (FLUOstar OPTIMA, BMG,
Offenburg, Germany). From the absorbance change at this wave-
length the amount of produced acid was calculated. After 9 min
The inhibition reaction was followed by monitoring the leaving
group 4-methylumbelliferone (97%, Sigma), which increased the
fluorescence intensity, as detected on a Perkin–Elmer LS50B (lex
=
360 nm, lem =445 nm). The reaction was allowed to continue until
no further increase in fluorescence was observed. Released 4-meth-
ylumbelliferone had a linear correlation to the concentration of
enzyme added. Three different concentrations of enzyme were
used in each case. No significant background hydrolysis of the in-
8
00
www.chembiochem.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2010, 11, 796 – 801

Suppression of Water in Lipase Catalysis
hibitor was detected. Standard curves were prepared by using free
[6] J. W. Murray, J. Barber, J. Struct. Biol. 2007, 159, 228–237.
[
7] M. Silberstein, J. Damborsky, S. Vajda, Biochemistry 2007, 46, 9239–
249.
4
-methylumbelliferone.
9
[
8] S. Santini, J. M. Crowet, A. Thomas, M. Paquot, M. Vandenbol, P. Thonart,
J. P. Wathelet, C. Blecker, G. Lognay, R. Brasseur, L. Lins, B. Charloteaux,
Biophys. J. 2009, 96, 4814–4825.
Acknowledgements
[
9] A. Weeks, L. Lund, F. M. Raushel, Curr. Opin. Chem. Biol. 2006, 10, 465–
4
Financial support from the Swedish Research Council, The Swed-
ish Foundation for International Cooperation in Research and
Higher Educations and The Swedish Institute.
72.
[10] M. W. Larsen, U. T. Bornscheuer, K. Hult, Protein Expression Purif. 2008,
2, 90–97.
11] L. J. Jensen, K. V. Andersen, A. Svendsen, T. Kretzschmar, Nucl. Acids Res.
998, 26, 697–702.
12] C. Hames, S. Halbedel, O. Schilling, J. Stulke, Appl. Environ. Microbiol.
005, 71, 4097–4100.
6
[
[
1
Keywords: Candida antarctica lipase B · hydrolysis · library
screening · mutagenesis · rational design · transacylation
2
[
[
13] L. E. Janes, R. J. Kazlauskas, J. Org. Chem. 1997, 62, 4560–4561.
14] E. Sawicki, T. R. Hauser, T. W. Stanley, W. Elbert, Anal. Chem. 1961, 33,
[1] T. Boekhout, J. Gen. Appl. Microbiol. 1995, 41, 359–366.
93–96.
[2] E. M. Anderson, K. M. Larsson, O. Kirk, Biocatal. Biotransform. 1998, 16,
[
[
15] A. O. Magnusson, J. C. Rotticci-Mulder, A. Santagostino, K. Hult, Chem-
BioChem 2005, 6, 1051–1056.
16] D. Rotticci, J. C. Rotticci-Mulder, S. Denman, T. Norin, K. Hult, ChemBio-
Chem 2001, 2, 766–770.
181–204.
[
[
3] L. Chen, A. Y. Lyubimov, L. Brammer, A. Vrielink, N. S. Sampson, Biochem-
istry 2008, 47, 5368–5377.
4] F. Leroux, S. Dementin, B. Burlat, L. Cournac, A. Volbeda, S. Champ, L.
Martin, B. Guigliarelli, P. Bertrand, J. Fontecilla-Camps, M. Rousset, C.
Lꢄger, Proc. Natl. Acad. Sci. USA 2008, 105, 11188–11193.
5] J. Saam, I. Ivanov, M. Walther, H.-G. Holzhꢅtter, H. Kuhn, Proc. Natl. Acad.
Sci. USA 2007, 104, 13319–13324.
[
Received: December 4, 2009
Published online on March 16, 2010
ChemBioChem 2010, 11, 796 – 801
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chembiochem.org
801