Mapping the substrate selectivity of new hydrolases using colorimetric screening: Lipases from Bacillus thermocatenulatus and Ophiostoma piliferum, esterases from Pseudomonas fluorescens and Streptomyces diastatochromogenes

Article abstract of DOI:10.1016/S0957-4166(01)00072-6

Recent advances in biochemistry and molecular biology have simplified the discovery and preparation of new hydrolases. Although these hydrolases might solve problems in organic synthesis, measuring their selectivity, especially enantioselectivity, remains tedious and time consuming. Recently, we developed a colorimetric screening method to measure the enantioselectivity of hydrolases. Here we apply this rapid screening method to map the substrate selectivity of four new hydrolases: lipases from the thermophilic Bacillus thermocatenulatus (DSM 730, BTL2) and a filamentous fungus Ophiostoma piliferum (NRRL 18917, OPL) and esterases from two bacteria, Pseudomonas fluorescens (SIK-W1, esterase I, PFE) and Streptomyces diastatochromogenes (Tue 20, SDE). We screened a general library of 29 substrates and a chiral library of 23 pairs of enantiomers. All four hydrolases catalysed the hydrolysis of unnatural substrates, but the two lipases accepted a broader range of substrates than the two esterases. As expected, the two lipases favoured more hydrophobic substrates, while the two esterases showed a preference for smaller substrates. Several moderately enantioselective reactions were identified for the solketal esters: BTL2, butyrate, E = 7.9 (R); octanoate, E = 4.9 (R) and 3-bromo-2-methyl propionate methyl esters, PFE, E = 12 (S); SDE, E = 5.6 (S). OPL showed low enantioselectivity toward all substrates tested. The current colorimetric screen could not measure the selectivity for several slow-reacting substrates. Traditional screening identified high enantioselectivity of BTL2 and PFE toward one of these slow substrates, 1-phenylethyl acetate (E>50).

Full text of DOI:10.1016/S0957-4166(01)00072-6

TETRAHEDRON:
ASYMMETRY
Pergamon
Tetrahedron: Asymmetry 12 (2001) 545–556
Mapping the substrate selectivity of new hydrolases using
colorimetric screening: lipases from Bacillus thermocatenulatus
and Ophiostoma piliferum, esterases from Pseudomonas
fluorescens and Streptomyces diastatochromogenes
Andrew Man Fai Liu,a,† Neil A. Somers,a,‡ Romas J. Kazlauskas,a,* Terry S. Brush,b,§
Frank Zocher,c,¶ Markus M. Enzelberger,c,ꢀꢀ Uwe T. Bornscheuer,c,** Geoff P. Horsman,a
Alessandra Mezzetti,a Claudia Schmidt-Dannertc,†† and Rolf D. Schmidc
aMcGill University, Department of Chemistry, 801 Sherbrooke St. W., Montre´al, Que´bec, Canada
bClariant Corporation, Biotechnology Research Division, 128 Spring St., Suite 1, Lexington, MA 02421, USA
cStuttgart University, Institute for Technical Biochemistry, Allmandring 31, D-70569 Stuttgart, Germany
Received 19 December 2000; accepted 2 January 2001
Abstract—Recent advances in biochemistry and molecular biology have simplified the discovery and preparation of new
hydrolases. Although these hydrolases might solve problems in organic synthesis, measuring their selectivity, especially enantiose-
lectivity, remains tedious and time consuming. Recently, we developed a colorimetric screening method to measure the
enantioselectivity of hydrolases. Here we apply this rapid screening method to map the substrate selectivity of four new
hydrolases: lipases from the thermophilic Bacillus thermocatenulatus (DSM 730, BTL2) and a filamentous fungus Ophiostoma
piliferum (NRRL 18917, OPL) and esterases from two bacteria, Pseudomonas fluorescens (SIK-W1, esterase I, PFE) and
Streptomyces diastatochromogenes (Tu¨ 20, SDE). We screened a general library of 29 substrates and a chiral library of 23 pairs
of enantiomers. All four hydrolases catalysed the hydrolysis of unnatural substrates, but the two lipases accepted a broader range
of substrates than the two esterases. As expected, the two lipases favoured more hydrophobic substrates, while the two esterases
showed a preference for smaller substrates. Several moderately enantioselective reactions were identified for the solketal esters:
BTL2, butyrate, E=7.9 (R); octanoate, E=4.9 (R) and 3-bromo-2-methyl propionate methyl esters, PFE, E=12 (S); SDE,
E=5.6 (S). OPL showed low enantioselectivity toward all substrates tested. The current colorimetric screen could not measure the
selectivity for several slow-reacting substrates. Traditional screening identified high enantioselectivity of BTL2 and PFE toward
one of these slow substrates, 1-phenylethyl acetate (E>50). © 2001 Elsevier Science Ltd. All rights reserved.
* Corresponding author. Tel.: +1 514 398 7228; fax: +1 514 398 3797;
1. Introduction
e-mail: romas.kazlauskas@mcgill.ca
Present address: Hong Kong University of Science and Technol-
†
Recently, we developed a rapid method for measuring
ogy, Department of Biology, Clear Water Bay, Kowloon, Hong
the stereoselectivity of hydrolases, called Quick E.1 In
Kong.
‡
Present address: ThermoGen Inc., 2225 West Harrison, Chicago,
this paper we apply this method to map the substrate
selectivity, especially the enantioselectivity, of four new
hydrolase enzymes: two lipases and two esterases. Their
biochemical properties suggest that they may be useful
for organic synthesis (Table 1). They tolerate wide pH
ranges and elevated temperatures. Three of the four
hydrolases have already been cloned and overexpressed
in E. coli. However, they have not yet been applied to
problems in organic synthesis because little is known
about their substrate selectivity.
IL 60612, USA.
§
Present address: Extreme Biotech Inc., 321 Fortune Blvd., Milford,
MA 01757, USA.
Present address: Aventis Pharma Deutschland Gmbh, Process
¶
Development Biochemistry, Biologika Nord
D 741, D-65926
Frankfurt/Main, Germany.
ꢀꢀ
Present address: California Institute of Technology, Department of
Chemistry and Chemical Engineering, Pasadena, CA 91125, USA.
** Present address: Ernst-Moritz-Arndt-University Greifswald, Insti-
tute of Chemistry & Biotechnology, Department of Technical
Chemistry & Biotechnology, Soldtmannstrasse 16, D-17487 Greif-
swald, Germany.
†† Present address: University of Minnesota, Department of Biochem-
istry, Molecular Biology and Biophysics, 256 Gortner Laboratory,
1479 Gortner Avenue, St. Paul, MN 55108-1022, USA.
Lipase from a thermophilic Bacillus, B. thermocatenula-
tus (DSM 730), has been cloned and overexpressed in
0957-4166/01/$ - see front matter © 2001 Elsevier Science Ltd. All rights reserved.
PII: S0957-4166(01)00072-6

546
A. M. F. Liu et al. / Tetrahedron: Asymmetry 12 (2001) 545–556
Table 1. Biochemical properties of esterases and lipases studied in this paper
Hydrolase
BTL2
OPL
PFE
SDE
Host
Bacillus thermocatenulanatus,
DSM 730
43
Ophiostoma piliferum,
NRRL 18917
60
Pseudomonas fluorescens,
SIK W1
30
Streptomyces
diastatochromogenes, Tu¨ 20
30.9
290
Q59837
E. coli
44°C
4.5–9.5
7.5
microorganism
Mol. wt. (kDa)
Aaa
417
Unknown
272
Sequenceb
Q59260
E. coli
60°C
9–10
Unknown
P22862
E. coli
45°C
5–10
c
Expression host
Optimal temp.
Stable at pH
Optimal pH
35°Cd
3–8
7
8–9
7.5–8
a Number of amino acids in the mature protein.
b TrEMBL or Swiss-Prot accession numbers for nucleic acid or amino acid sequences.
c Not genetically engineered. Isolated from natural source.
d Denatures starting above 35°C.
E. coli.2 This lipase tolerates high temperatures, some
detergents and organic co-solvents. These properties
suggest that it may be a rugged catalyst for organic
synthesis. However, no information is available about
the stereoselectivity of this lipase. Researchers have
measured its apparent acyl chain length selectivity and
found that it has typical lipase behaviour. For 4-nitro-
phenyl esters, the C10 ester reacted fastest, but C12,
C14, and C16 also reacted at rates similar to C10. For
triglycerides, tributyrin reacted fastest (C4 acyl chain),
but triglycerides with C6, C8, C10, and C12 acyl chains
also reacted at similar rates to C4.
for an esterase, favours esters of shorter chain car-
boxylic acids (propanoates, butyrates) over the longer
chain caprylates (C8) or decanoates (C10). Aryl
acetates were good substrates, but not acetates of ben-
zyl alcohol or 2-phenylethanol. Esters of aromatic acids
(methyl benzoate, methyl salicylate) were also found to
be unreactive.
Previous work also identified several stereoselective
reactions catalysed by PFE.8 PFE showed high enan-
tioselectivity in the hydrolysis of 1-phenylethyl acetate
(E>50), but the closely related 1-phenylpropyl acetate
reacted slowly and with low enantioselectivity (E=2).
PFE catalysed the acylation of the primary alcohol
solketal (2,2-dimethyl-1,3-dioxolane-4-methanol) and
five closely related derivatives, but showed low enan-
tioselectivity (E=1.1–4.9). PFE also showed low enan-
tioselectivity in the hydrolysis of methyl 3-phenyl-
butyrate (E=3.5). The positional isomer ethyl 2-phenyl
butyrate did not react.
The second lipase comes from the filamentous fungus
Ophiostoma piliferum (NRRL 18917). The forestry
industry uses an albino strain of this fungus for two
reasons.3 First, it secretes a lipase that breaks down
pitch in wood, which minimizes fouling of equipment
during paper manufacture. Second, colonization by O.
piliferum excludes other fungi, including those that stain
wood dark blue, thereby reducing its value. Brush et
al. recently isolated a lipase from this strain of O.
piliferum.4 It has molecular weight of approximately
60 kDa, a pI of 3.9, a pH optimum near 7, and is stable
in solution up to 35°C. Surprisingly for a lipase,
butanoate esters of 4-nitrophenol reacted faster than
esters of longer chain fatty acids. (C4 esters reacted
approximately 20 times faster than C18 esters.) This
lipase may also be useful for organic synthesis.
Although Gao and Breuil isolated a lipase from
another Ophiostoma species, O. piceae,5 it differs signifi-
cantly from the O. piliferum lipase. The O. piceae lipase
has a molecular weight of only 35 kDa, a pH optimum
near 5, and temperature optimum near 30°C.
Streptomyces are Gram-positive soil bacteria, many of
which produce useful antibiotics. A number of strains,
especially plant pathogens, secrete esterases that
degrade the protective waxy coat of leaves. Wei et al.
solved the X-ray structure of an esterase from the plant
pathogen Streptomyces scabies.9 This structure showed
that the active site does not consist of the usual Ser-
His-Asp triad, but a Ser-His diad. In addition, the
protein fold differed slightly from the a,b-hydrolase
fold. It is not clear how these unusual features might
change the stereoselectivity. This paper examines the
stereoselectivity of a closely related esterase. Tesch et al.
isolated and cloned an esterase from Streptomyces dia-
statochromogenes (Tu¨ 20, SDE).10 Khalameyzer and
Bornscheuer have overexpressed this esterase in E.
coli.11 It shows a similar amino acid sequence (31%
identity in a region of 156 amino acids) to the S. scabies
esterase, including the active site residues. Consistent
with its designation as an esterase, it favoured esters of
shorter chain fatty acids over longer chains: 4-nitro-
phenyl octanoate (C8, rate=100), 4-nitrophenyl laurate
(C12, rate=10), 4-nitrophenyl palmitate (C16, rate=
2).10 Although SDE catalysed the hydrolysis of unnatu-
Researchers have isolated a number of different hydro-
lases from Pseudomonas fluorescens (SIK-W1). This
work focuses on an ‘arylesterase’ first isolated and
cloned by Choi et al.6 and later resequenced to correct
errors.7 The amino acid sequence of this esterase is
similar to a non-heme haloperoxidase and indeed this
esterase shows low haloperoxidase activity.7 Previous
substrate mapping of this esterase8 showed that it
accepts a range of unnatural substrates and, as expected

A. M. F. Liu et al. / Tetrahedron: Asymmetry 12 (2001) 545–556
547
Figure 1. Measuring selectivity of hydrolases. (a) For estimated selectivities, the initial rate of hydrolysis of different esters is
measured colorimetrically using 4-nitrophenol as a pH indicator. The intensity of the yellow colour (monitored at 404 nm)
diminishes as hydrolysis proceeds. (b) Quantitative measure of selectivity requires a competitive experiment. We used resorufin
acetate or another resorufin ester as the competitive substrate because its hydrolysis generates the easily measured resorufin anion
(monitored at 574 nm). The total amount of hydrolysis is measured using 4-nitrophenol as the pH indicator as in part a. Only
10% of resorufin is deprotonated to the anion at pH 7.2.
ral substrates, it showed an unusually low enantio-
selectivity toward both secondary alcohols tested: E=
3.3 for 1-phenylethyl acetate, E=1 for 1-phenylpropyl
acetate. SDE is available commercially from Ju¨lich
Fine Chemicals.12
2. Results
2.1. General ester library
The general ester library consisted of 29 commercially
available esters (Scheme 1). The esters were chosen to
identify acyl chain length preferences of the hydrolases
and also their ability to hydrolyse hindered or charged
substrates. Simple alkyl esters as well as activated esters
(vinyl and phenyl esters, esters with electron-withdraw
ing substituents in the acyl portion) were included to
test whether activated esters would react faster.
Our rapid screening method uses pH indicators to
monitor ester hydrolysis colorimetrically1 (Fig. 1). We
carry out the screening in 96-well plates where each well
contains a different substrate. Comparing the rates of
reaction of different substrates gives an estimated or
apparent selectivity, but not the true selectivity (Fig.
1a).‡‡ Indeed, we previously showed that such compari-
sons can be in error by more than a factor of 20
from the true selectivity.1 To measure the true selec-
tivity we added a resorufin ester as a reference com-
pound to reintroduce competition (Fig. 1b). This com-
petition between the substrate and the resorufin ester
accounts for differences in KM and yields the correct
selectivity.
The four hydrolases accepted most of the substrates in
the general ester library (Table 2), but the most polar
molecules ethyl glycine, 8, and methyl hydroxyacetate,
20, were poor substrates. The hydrolase with the lowest
overall activity was BTL2, which came from a ther-
mophilic microorganism. At room temperature, only 12
of the 29 esters in the general library and only four of
23 pairs in the chiral library (see below) reacted. Upon
increasing the temperature to 40°C, almost all of the
esters reacted: 26 of the 29 esters in the general library
and at least one ester in all of the 22 pairs in the chiral
library. For this reason, all data on BTL2 in this paper
refer to a reaction temperature of 40°C. Even at this
increased temperature, BTL2 was less active than the
other three hydrolases were at 25°C.
‡‡ The substrate selectivity of an enzyme is the ratio of the specificity
constants (kcat/KM) for each substrate (Fersht, A. Structure and
Mechanism in Protein Science; Freeman: New York, 1999, pp.
116–117). By measuring initial rates of the substrates separately, we
eliminate competitive binding between the two substrates and there-
fore do NOT measure the true selectivity. For example, at saturat-
ing substrate concentrations, where [S]>KM, the relative initial rates
equal the relative kcat values; at partially saturating conditions,
where [S]<KM, the initial rates also depend on the KM values. Thus,
the ratio of separately measured initial rates ignores some or all of
the effect of KM on selectivity.
Surprisingly, the two lipases, BTL2 and OPL, catalysed
the hydrolysis of small activated and small unactivated

548
A. M. F. Liu et al. / Tetrahedron: Asymmetry 12 (2001) 545–556
Scheme 1. Esters used to survey the substrate selectivity of hydrolyses. Substrates that fit into two categories are listed twice.
EWG=electron-withdrawing group.
Table 2. Hydrolytic activity of the four hydrolases towards the general ester librarya
Substrate
BTL2
OPL
PFE
SDE
Nd
0.75
Nd
3.5
Nd
1300
120
0.66
2100
1600
650
340
270
18
5.1
B0.20
2.4
4.8
90
B0.20
B0.20
B0.20
B0.20
B0.20
B0.20
74
46
B0.20
1.9
1, Ethyl acetate
2, Ethyl butanoate
3, Ethyl hexanoate
4, Ethyl octanoate
2.9
3.9
1.1
1.3
0.72
2.9
0.60
B0.20
0.84
1.3
0.83
17
28
29
1.4
B0.20
2.1
B0.20
0.81
0.85
0.62
1.2
0.71
0.45
0.85
2.8
23
38
53
1000
110
37
Nd
3.2
Nd
13
Nd
2140
240
B0.20
2900
4600
1900
670
5, Ethyl decanoate
6, Ethyl trifluoroacetate
7, Ethyl bromoacetate
8, Ethyl glycine
110
2.3
9.8
110
250
910
1700
1300
27
9.6
250
14
35
1.0
B0.20
B0.20
33
12
3.0
130
6.6
1.2
9, Vinyl acetate
10, Vinyl propanoate
11, Vinyl butanoate
12, Vinyl hexanoate
13, Vinyl octanoate
14, Vinyl decanoate
15, Vinyl tetradecanoate
16, Vinyl hexadecanoate
17, Vinyl benzoate
18, Vinyl trimethylacetate
19, Methyl bromoacetate
20, Methyl hydroxyacetate
21, Methyl mandelate
22, Methyl benzoate
23, Methyl 4-(hydroxy-methyl)benzoate
24, Propyl acetate
1000
91
8.1
0.69
7.4
12
220
0.47
B0.20
B0.20
B0.20
1.0
0.52
290
170
25, Butyl acetate
26, Phenyl acetate
27, Isopropenyl acetate
28, Isobutyl acetate
29, Tributyrin
0.58
0.80
7.8
B0.20
6.5
640
a Activities are in units/mg of protein where a unit is 1 mmol of ester hydrolysed/minute. The detection limit was 0.20 unit/mg. Screening was done
at 25°C except for BTL2, which was screened at 40°C. Nd=not determined. To highlight the fast-reacting examples, all activities over 50 U/mg
are shown in bold.

A. M. F. Liu et al. / Tetrahedron: Asymmetry 12 (2001) 545–556
549
esters with similar rates. For example, ethyl acetate, 1,
ethyl trifluoroacetate, 6, and vinyl acetate, 9, reacted at
similar rates. On the other hand, larger activated esters
reacted significantly faster than larger unactivated
esters, especially in the OPL-mediated reactions. For
example, vinyl decanoate, 14, reacted 12–40 times faster
than the ethyl decanoate, 5. Similarly, with OPL, vinyl
benzoate, 17, reacted more than 1000 times faster than
methyl benzoate, 22. However, for BTL2, vinyl ben-
zoate, 17, reacted at a similar rate as methyl benzoate,
22. The similar reaction rate for activated and non-acti-
vated esters suggests that a non-chemical step (possibly
lid opening) limits the rate of reaction for BTL2.13 This
lid opening may be more difficult for small substrates
with the thermophile-derived BTL2.
group.§§ For the vinyl esters (Fig. 2) the lipases
favoured C8 or C10 acyl chains, while the esterases
favoured C2 or C3 chains. However, the preferred acyl
group for the esterases differed upon changing from
vinyl to ethyl esters. The esterases now preferred not
the butanoate chain, but the longer octanoate
analogues.
Previous work on p-nitrophenyl esters also showed a
different selectivity: the ideal for BTL2 was a C10–C16
chain, OPL favoured the butanoate analogues, whilst
SDE favoured the octanoate esters. For BTL2 and
triglycerides, the chain length preference shifted yet
again and tributyrin was the best substrate. Although
short length acyl chains were usually the ideal for
esterases, the lipases usually favoured longer chain
lengths; the true acyl chain selectivity cannot be mea-
sured by this non-competitive approach. The apparent
differences in chain length selectivity upon changing the
alcohol fragment of the ester probably reflect changes
in KM upon changing the alcohol portion of the ester.
For esterases PFE and SDE, activated esters reacted
faster than unactivated esters regardless of their size.
For example, ethyl trifluoroacetate, 6, and vinyl acet-
ate, 9, reacted more than 600 times faster than similar
unactivated esters like ethyl butanoate, 2, or propyl
acetate, 24. Similarly, vinyl octanoate, 13, reacted
approximately 70 times faster than the unactivated
ethyl octanoate, 4.
§§ The apparent chain length selectivity is the relative rate of reaction
for different esters in separately measured reactions. It is not the
true selectivity for the reasons detailed in the previous footnote.
The apparent acyl chain length selectivity of the hydro-
lases varied dramatically with the nature of the leaving
Figure 2. Initial rates of reaction of vinyl esters with the four hydrolases. The two lipases (BTL2 and OPL) both favour medium
chain length (C6 to C10), while the esterases (PFE and SDE) show preference for short chain acyl groups (C2 to C4). The initial
rates of hydrolysis for each ester were measured separately. For this reason the variations do not give the true selectivity, but only
the apparent selectivity.

550
A. M. F. Liu et al. / Tetrahedron: Asymmetry 12 (2001) 545–556
Scheme 2. Pure enantiomers used to survey the enantioselectivity of hydrolyses. For simplicity only the (R)-enantiomers are
shown, but both were screened.
2.2. Chiral ester library
resorufin ester accounts for differences in KM and yields
the correct enantioselectivity. For slow-reacting sub-
strates, we used slower-reacting reference compounds
such as resorufin pivalate or isobutyrate instead of the
faster-reacting acetate, which permits more accurate
measurement of the relative rate of hydrolysis of the
substrate and reference compound. However, for the
slowest reacting substrates, even these resorufin esters
reacted too rapidly. During the competitive experiment,
we observed only hydrolysis of the reference compound
and no detectable reaction of the substrate. In these
cases, we could not measure the enantioselectivity using
Quick E, but could only set a lower limit.
The chiral ester library contained 23 pairs of enan-
tiomers, as shown in Scheme 2. We grouped the sub-
strates according to the location of the stereocentre:
esters of chiral primary alcohols (five pairs), esters of
secondary alcohols (three pairs), esters of chiral car-
boxylic acids with a stereocentre at the a-position (ten
pairs), esters of chiral carboxylic acids with a stereocen-
tre at the b-position (two pairs), and lactones (three
pairs).
The first stage of screening measured the initial rate of
hydrolysis of each enantiomer separately to identify
which esters reacted. In addition, for those cases where
both enantiomers reacted, we also calculated the ratio
of initial rates as an estimated enantioselectivity. Note
that this ratio is NOT the true enantioselectivity, or
enantiomeric ratio, E, because we measured the reac-
tion rates of the pure enantiomers separately.
2.3. Enantioselectivity of BTL2
The chiral esters reacted more slowly than the esters in
the general library with all hydrolases (Table 3). For
BTL2 (40°C) at least one enantiomer reacted for all but
one of the enantiomer pairs, but the rates were much
slower. The fastest rates for BTL2 in the general library
were 17–29 units/mg protein, while the fastest rates in
the chiral library were 2.6–3.2 units/mg protein. BTL2
also catalysed the hydrolysis of chiral lactones, which
are often poor substrates for hydrolases.
In the second stage of screening, we measured the true
enantioselectivity using the colorimetric Quick
E
method.1 This method adds a reference compound, a
resorufin ester, to reintroduce competition (Fig. 1b).
This competition between the substrate and the

A. M. F. Liu et al. / Tetrahedron: Asymmetry 12 (2001) 545–556
551
Table 3. Hydrolytic activitya and estimated enantioselectivityb of the four hydrolases towards the chiral ester library
Substrate
BTL2
Est. E
OPL
Est. E
PFE
Est. E
SDE
Activity
Activity
Activity
Activity
Est. E
30R
30S
31R
31S
32R
32S
33R
33S
34R
34S
35R
35S
36R
36S
37R
37S
38R
38S
39R
39S
40R
40S
41R
41S
42R
42S
43R
43S
44R
44S
45R
45S
46R
46S
47R
47S
48R
48S
49R
49S
50R
50S
51R
51S
52R
52S
0.18
B0.045
2.6
\4.0 R
\4.0 R
3.2 R
3.2 R
5.2 R
5.2 R
1.2 R
1.2 R
Nr
Nr
Nr
Nr
2.4 S
2.4 S
1.4 S
1.4 S
1.7 R
1.7 R
1.0
150
52
2.9 R
2.9 R
1.4 R
1.4 R
\1.0 R
\1.0 R
1.3 S
1.3 S
1.3 S
1.3 S
6.9 R
6.9 R
2.3 S
2.3 S
Nr
0.50
0.48
1.6
1.8
B0.40
B0.40
B0.40
B0.40
Nt
Nt
0.52
B0.40
0.94
0.99
0.93
1.0
B0.40
B0.40
2.3
19
59
27
B0.40
B0.40
B0.40
B0.40
0.50
1.0
1.0
1.1 S
1.1 S
Nr
Nr
Nr
Nr
–
0.25
0.24
B0.20
B0.20
0.21
0.27
B0.20
0.23
Nt
Nt
0.22
B0.20
0.64
0.56
0.57
0.57
B0.20
B0.20
0.37
1.0
1.0
Nr
Nr
1.3 S
1.3 S
\1.1 S
\1.1 S
–
8400
6000
0.30
B0.30
0.31
0.40
16
21
73
11
6.4
0.8
0.23
0.045
0.14
0.11
B0.045
B0.045
B0.045
B0.045
0.099
0.23
0.18
0.25
0.73
0.43
1.7
–
–
\1.3 R
\1.3 R
1.1 S
1.1 S
1.1 S
1.1 S
Nr
\1.1 R
\1.1 R
1.1 R
1.1 R
1.0
1.0
Nr
Nr
13 S
13 S
Nr
Nr
Nr
Nr
Nr
Nr
1.4 S
1.4 S
Nr
Nr
Nr
Nr
Nr
Nr
Nr
Nr
Nr
Nr
Nr
Nr
Nr
Nr
Nr
15
B0.30
B0.30
B0.30
B0.30
B0.30
B0.30
28
28
28
25
Nr
Nr
Nr
Nr
Nr
8.3 S
8.3 S
2.2 R
2.2 R
Nr
Nr
Nr
Nr
1.7
3.0
3.2
1.1
1.0
Nr
4.9
1.1 S
1.1 S
1.3 S
1.3 S
2.0 R
2.0 R
1.4 R
1.4 R
2.8 R
2.8 R
1.1 R
1.1 R
2.1 S
2.1 S
2.6 S
2.6 S
1.6 R
1.6 R
1.4 R
1.4 R
1.4 R
1.4 R
1.0
1.0 S
1.0 S
1.0 R
1.0 R
Nr
B0.20
B0.20
B0.20
B0.20
B0.20
B0.20
0.21
1.5
0.87
0.44
0.90
0.65
2.5
0.89
0.65
0.57
0.31
0.63
0.20
0.51
1.1
0.69
0.48
0.35
1.0
0.72
0.93
0.97
0.51
0.50
B0.30
B0.30
2.7
Nr
10 S
10 S
Nr
1.7 S
1.7 S
Nr
27
0.86
0.29
B0.30
B0.30
0.63
0.39
B0.30
B0.30
B0.30
B0.30
34
32
36
58
B0.30
B0.30
B0.30
B0.30
1.8
B0.40
B0.40
0.44
B0.20
B0.20
B0.20
B0.20
B0.20
B0.20
B0.20
B0.20
B0.20
B0.20
B0.20
B0.20c
B0.20c
B0.20c
B0.20c
B0.20c
Nt
Nr
Nr
1.6 R
1.6 R
Nr
Nr
Nr
1.1 R
1.1 R
Nr
Nr
Nr
Nr
Nr
Nr
Nr
Nr
Nr
Nr
Nr
Nr
–
–
0.40
B0.40
B0.40
B0.40
B0.40
B0.40
B0.40
B0.40
B0.40c
B0.40c
B0.40c
B0.40c
B0.40c
Nt
Nr
1.1 R
1.1 R
1.6 R
1.4 R
Nr
Nr
Nr
Nr
1.6 R
1.6 R
1.0
1.0
1.0
Nr
–
–
1.1
Nt
Nt
a Activities are in units/mg of protein where a unit is 1 mmol of ester hydrolysed/min. To highlight the fast-reacting examples, all activities over
50 U/mg are shown in bold. Reaction conditions: 1 mM substrate, 7 vol% acetonitrile, 0.45 mM 4-nitrophenol, 1.0 mM BES at pH 7.20, 25°C
except 40°C for BTL2. Nt=not tested, Nr=substrate did not react so no ratio can be calculated.
b Estimated E is the rate of the faster enantiomer over the slower one. The configuration of the fast-reacting enantiomer is shown in italics
following the value.
Quick E measurements on BTL2 showed moderate
enantioselectivity toward solketal butanoate (E=7.9)
(Table 4). The true enantioselectivity of BTL2 toward
the other substrates could not be measured with Quick
E because they reacted too slowly compared to
resorufin esters. We used three different reference com-
pounds: resorufin acetate, resorufin isobutyrate, and
resorufin pivalate. The increasing steric bulk of the acyl
group slowed down the relative rate of reaction, allow-
The first stage of screening suggested that BTL2 may be
enantioselective toward solketal butyrate, 30 (estimated
E>4.0 favouring the (R)-enantiomer), and glycidyl 4-
nitrobenzoate, 32 (estimated E=5.2 favouring the (R)-
enantiomer). Glycidyl 2-methyl 4-nitrobenzoate, 33, an
ester similar to 32, showed a low estimated E. Several
other esters showed low estimated enantioselectivity
(ꢀ2.5): menthyl acetate, 36, methyl mandelate, 44, and
N-benzyl-proline ethyl ester, 47.

552
A. M. F. Liu et al. / Tetrahedron: Asymmetry 12 (2001) 545–556
Table 4. Enantioselectivity of the hydrolases towards selected chiral substrates measured by Quick E
Hydrolase
Substrate
Est. Ea
Rate subs.b
Rate ref.c
Ref.d
Quick E
R
S
R
S
BTL2
BTL2
OPL
OPL
OPL
OPL
OPL
PFE
PFE
SDE
30
31
31
34
35
36
40
39
40
39
3.3 (R)
3.2 (R)
1.4 (R)
1.3 (S)
6.9 (R)
2.3 (S)
1.0 (S)
8.3 (S)
2.2 (R)
13 (S)
12
78
15
15
44
0.064
5.7
16
57
0.9
4.0
26
12
21
30
0.087
8.4
19
27
8.2
1.7
26
140
0.37
0.19
17
0.23
4.3
4.0
1.5
2.3
42
130
0.39
0.11
18
0.39
4.3
3.3
2.4
i-But
Ac
Ac
Pv
Pv
i-But
Pv
Ac
Pv
Pv
7.9 (R)e
4.9 (R)
1.3 (R)
1.5 (S)
1.5 (R)
1.4 (S)
1.5 (S)
12 (S)f
1.8 (R)
5.6 (S)
a Estimated enantioselectivity from Table 3. This is the ratio of the initial rates of hydrolysis of the two enantiomers measured without
competition.
b Initial rate of hydrolysis of the substrate in units/mg protein in the presence of resorufin ester.
c Initial rate of hydrolysis of the resorufin ester in units/mg protein in the presence of substrate.
d Different reference compounds were used: Ac=resorufin acetate, i-But=resorufin isobutyrate, Pv=resorufin pivalate.
e The concentration of the slow-reacting (S)-enantiomer was 2 mM in the assay to improve accuracy.
f The concentration of the slow-reacting (R)-enantiomer was 10 mM to improve the accuracy. Also, the BES buffer concentration was 5 mM, also
to increase accuracy. A check using the endpoint method yielded an enantioselectivity of 13 2 (mean and standard deviation for four
measurements). Representative data: 35.5% conversion, 43.6% e.e. for remaining starting material, 79.3% e.e. for product; these data correspond
to E=13.2.
ing us to measure the enantioselectivity of slower sub-
strates. For BTL2, the pivalate did not react at all, so
the slowest available reference compound was resorufin
isobutyrate.
Changing the organic co-solvent increased the rate of
BTL2-catalysed hydrolysis, but to a lesser extent. The
screening solution contained 7 vol% acetonitrile to dis-
solve the substrate and, when needed, the reference
compound. Replacing acetonitrile with iso-propanol
increased the rate of the BTL2-catalysed hydrolysis of
tributyrin 1.6-fold. Several other solvents showed more
modest changes; with acetone a 1.1-fold rate increase
was seen, whilst with tetrahydrofuran led to a 1.3-fold
increase in the rate.
Since the earlier general ester screening showed that
BTL2 prefers longer acyl chains, we prepared solketal
octanoate, 31, and added it to the chiral library. As
expected, the rate of reaction increased more than
10-fold: 2.6 units/mg protein for the octanoate as com-
pared to 0.18 for the butyrate. The enantioselectivity
decreased slightly with E=4.9 for the octanoate as
compared to 7.9 for the butyrate. We confirmed the
accuracy of this result using the endpoint method,
detailed in Table 5. A small-scale resolution of racemic
solketal octanoate yielded the remaining starting mate-
rial with 40.1% e.e. and product with 53% e.e. at 43.1%
conversion corresponding to E=4.5. The mean of five
measurements was 4.4 with a standard deviation of 0.4,
in good agreement with the Quick E measurement.
Although the first stage screening showed no reaction
for 1-phenylethyl butyrate, 35, many other lipases
showed high enantioselectivity toward 1-phenylethyl
esters. For this reason, we measured the enantioselectiv-
ity of BTL2 toward 1-phenylethyl esters using tradi-
tional methods. Although the reactions were very slow
(ꢀ105–106 times slower than tributyrin), BTL2 did
catalyse both hydrolysis of the 1-phenylethyl acetate
and acylation of 1-phenylethanol with vinyl acetate. In
Table 5. Enantioselectivity of BTL2 toward solketal, 1-phenylethanol, and 1-phenylpropanol
Reactants
Conv. (%)
% e.e. S
% e.e. P
E
Time (h)
Solketal octanoate/H2O
43
43
43
21
41
40
74
75
26
69
53
100
100
100
100
4.490.4a
\100
\100
\100
\100
1b
92
91
91
93
1-Phenylethanol/vinyl acetate
1-Phenylethanol/vinyl butyrate
1-Phenylpropanol/vinyl acetate
1-Phenylethyl acetate/H2O
a Mean and standard deviation for five measurements. Data are for one of these measurements.
b Different amounts of enzyme were used so the times cannot be directly compared.

A. M. F. Liu et al. / Tetrahedron: Asymmetry 12 (2001) 545–556
553
both cases the enantioselectivity was excellent, E>100
(Table 5). BTL2 catalysed the acylation of 1-phenyl-
propanol with similar rates and with excellent
enantioselectivity.
Methyl 3-bromo-2-methyl propionate, 39, showed the
highest estimated enantioselectivity: 8.3 for PFE and 13
for SDE, both favouring the (S)-enantiomer. The true
enantioselectivity was a moderate 12 for PFE and 5.6
for SDE. Small-scale resolutions of racemic 39 with
PFE, where the enantiomeric purity of the remaining
starting material and product were measured by gas
chromatography, showed an enantioselectivity of 13 2,
in good agreement with the value measured by Quick
E.
BTL2 catalysed hydrolysis of one b-lactone and two
d-lactones, but with low estimated enantioselectivity
(est. E=1.1–1.4). We checked the enantioselectivity of
BTL2 toward several g- and d-lactones using traditional
screening, but the enantioselectivity was also low. For
the g-lactones, the enantiomeric purity of the unreacted
lactone was as follows (conversion not determined):
g-valerolactone no reaction, g-octalactone 2% e.e., g-
nonalactone 2% e.e., g-decalactone 4% e.e., g-unde-
calactone 7% e.e., g-dodecalactone 12% e.e. For the
d-lactones the enantiomeric purity of the unreacted
lactone was as follows (50% conversion): d-octalactone
0% e.e., d-nonalactone 0% e.e., d-decalactone 5% e.e.,
d-undecalactone 5% e.e., d-dodecalactone 10% e.e.
Enantioselectivity in the BTL2-catalysed hydrolysis of
lactones was found to be so low as to be of no practical
use.
In addition, the estimated enantioselectivity was >1.1
for 2-methylglycidyl 4-nitrobenzoate, 33, with SDE,
>1.1 for 1-phenylethyl butyrate, 35, with both PFE and
SDE, and 2.2 for methyl 2-chloropropionate, 40, with
PFE. The true enantioselectivity of PFE toward 40 was
only 1.8. The true enantioselectivities for 2-methylgly-
cidyl 4-nitrobenzoate, 33, and 1-phenylethyl butyrate,
35, could not be measured because they reacted too
slowly compared to the reference compounds. How-
ever, previous work using traditional methods mea-
sured an enantioselectivity of these esterases toward a
related compound, 1-phenylethyl butyrate. PFE showed
an enantioselectivity >100,8 while SDE showed an
enantioselectivity of only 3.3.11
2.4. Enantioselectivity of OPL
In the OPL-catalysed hydrolysis of a more limited
range of chiral esters, at least one enantiomer reacted
for only 14 of the 23 pairs. By far the fastest substrate
was solketal octanoate, 31. Most of the esters in the
chiral library were acetates and butyrates. Screening
with the general ester library showed these are poor
substrates for this lipase.
3. Discussion
Mapping the substrate selectivity of hydrolases using
the Quick E screening was dramatically faster than
traditional methods. Once the library was assembled,
both the general and chiral library could be screened
for all four hydrolases in one or two afternoons. Using
traditional methods, such a mapping would require at
least one month. There are at least four reasons for this
dramatic increase in throughput. Most importantly, the
Quick E screening eliminates the need to measure enan-
tiomeric purity. Measuring the enantiomeric purity for
23 different substrates and products would require at
the very least several different expensive chiral HPLC
and GC columns and each measurement would require
at least one hour; secondly, Quick E screening measures
only initial rates (>5% conversion) so there is no need
to wait for the reaction to reach 30 or 40% conversion;
thirdly, the colorimetric method allows 96 reactions to
be followed simultaneously using a microplate reader.
Finally, the small scale of the reaction (typically 0.1
mmol) allows the use of more enzyme per mole of
substrate and thus the reaction is faster.
OPL showed the highest estimated enantioselectivities
toward 1-phenylethyl butyrate, 35 (estimated E=6.9
favouring the (R)-enantiomer), and methyl lactate, 43
(estimated E=10 favouring the (S)-enantiomer). OPL
showed low estimated Es toward solketal butyrate, 30
(estimated E=2.9), and menthyl acetate, 36 (estimated
E=2.3). True enantioselectivities measured by Quick E
were low, all <2: solketal octanoate, 31, glycidyl
butyrate, 34, 1-phenylethyl butyrate, 35, menthyl ace-
tate, 36, and methyl 2-chloropropionate, 40. The true
enantioselectivities for solketal butyrate, 30, and methyl
lactate, 43, could not be measured because they reacted
too slowly compared to the reference compounds.
Thus, OPL showed low enantioselectivity toward all the
substrates that we tested.
2.5. Enantioselectivity of PFE and SDE
PFE and SDE catalysed the hydrolysis of a limited
number of esters in the chiral ester library. For PFE at
least one enantiomer reacted for only nine of the 21
pairs and for SDE only eight of the 21 pairs. (Two
esters were not tested with the esterases.) Neither PFE
nor SDE catalysed hydrolysis of chiral lactones or most
of the chiral carboxylic derivatives. PFE did not
catalyse hydrolysis of the two 3-hydroxy butyrate esters
48 and 49. Bornscheuer et al. also found that PFE did
not catalyse the hydrolysis of a related 3-hydroxy ester,
but used directed evolution to find a PFE mutant that
did.14
The only limitation of substrate mapping with Quick E
is that the true enantioselectivity for very slowly react-
ing substrates cannot be measured. This arises from the
relative reactivity of the substrate and the resorufin
ester used as a reference compound. For poorly reactive
substrates, only the resorufin ester reacted; thus, we
could not measure the relative rate. Utilizing several
different resorufin esters—acetate, isobutyrate, and
pivalate—minimized this problem. The increasing steric
bulk of the acyl group slowed down the reaction rate,

554
A. M. F. Liu et al. / Tetrahedron: Asymmetry 12 (2001) 545–556
Table 6. Summary of substrate selectivity
(R)-enantiomer. By increasing the length of the acyl
chain to octanoate, the rate increased 10-fold with
similar enantioselectivity, E=4.9. Other ways to
increase the reaction rate were to increase the tempera-
ture or to change the co-solvent from acetonitrile to
iso-propanol. Although this enantioselectivity is too
low for preparative use, it represents a good starting
point for directed evolution. Indeed, directed evolution
has yielded several more enantioselective mutants.17
Enzyme
BTL2
Information
Low activity at room temperature
For small esters, both activated and unactivated
esters react at similar rates
Favours longer acyl chains, e.g. n-octanoate and
n-decanoate vinyl esters
High enantioselectivity toward 1-phenylethanol and
1-phenylpropanol, but slow reaction
OPL showed good activity toward a wide range of
unnatural substrates, but low enantioselectivity. The
use of acetonitrile co-solvent in our assay may have
reduced the enantioselectivity of OPL, since OPL often
denatures in the presence of organic co-solvents. One
possible application of this lipase could be the deprotec-
tion of synthetic intermediates under very mild
conditions.
Moderate enantioselectivity toward primary alcohols,
e.g. solketal butanoate and octanoate
Catalyses hydrolysis of lactones, but with low
enantioselectivity
High activity toward hydrophobic substrates
Strongly favours long acyl chains, e.g. n-octanoate
and n-decanoate vinyl esters
OPL
PFE
Low enantioselectivity
May show enantioselectivity toward methyl lactate
Favours vinyl esters and medium chains (some
preference to C4) and activated esters (vinyl, a-halo)
High enantioselectivity toward 1-phenylethanol
Moderate enantioselectivity toward methyl
(S)-3-bromo-2-methylpropionate and methyl
(R)-2-chloropropionate
Favours vinyl esters and long chains (some preference
to C6) and activated esters (vinyl, a-halo)
Moderate enantioselectivity toward methyl
(S)-3-bromo-2-methylpropionate
Both PFE and SDE showed moderate enantioselectivity
toward methyl 3-bromo-2-methylpropanoate. Synthetic
chemists have used this important chiral synthon to
prepare retroviral protease inhibitors,18 unnatural
amino acids,19 cyclooxygenase inhibitors,20 captopril (a
treatment for high blood pressure)21 and alkaloids.22
Although this moderate enantioselectivity is insufficient
for preparative use, we are currently increasing the
enantioselectivity using directed evolution. Other
hydrolases also show moderate enantioselectivity
toward this acid.21 Henke and Bornscheuer increased
the enantioselectivity of PFE toward chiral 3-phenylbut-
anoic acid using directed evolution.23
SDE
so pivalate often served as a reference compound for
substrates with low reactivity. To further extend the
range of Quick E, a wider range of reference com-
pounds is required, especially deactivated ester sub-
strates. Several new reference compounds are currently
being evaluated.15 Otherwise, the mapping gave a good
overview of the selectivity of each hydrolase and iden-
tified several interesting, though not yet useful, reac-
tions. An alternative is to identify which substrates
react using pH indicators coupled with standard GC
methods to measure the actual enantioselectivity.16
In spite of the unusual structure of the active site of
SDE, its stereoselectivity was similar to that of PFE.
The only major difference between PFE and SDE was
in the enantioselectivity toward 1-phenylethyl acetate,
where the PFE mediated reaction was highly enantiose-
lective, whilst SDE showed low enantioselectivity for
this substrate.
4. Experimental
4.1. General
A summary of the substrate mapping (Table 6) pro-
vides useful guidelines for applications of these hydro-
lases. As expected, the lipases usually favoured
compounds carrying longer chain acyl groups, while the
esterases preferred esters with short chain acyl groups.
It may be useful to use longer chain acyl donors, e.g.
vinyl octanoate, instead of shorter chain acyl donors
such as vinyl acetate. The apparent acyl chain length
selectivity changed dramatically with the nature of the
leaving group. This inconsistency is most likely due to
the fact that we measure not the true selectivity with a
competitive experiment, but an apparent (or estimated)
selectivity using separately measured rates. Establishing
the true acyl chain selectivity of these lipases could be
accomplished using an appropriate reference com-
pound, but was not a goal of this research.
Unless otherwise noted BES (N,N-bis[2-hydroxyethyl]-
2-aminoethanesulfonic acid) buffer (1.0 mM, pH 7.20)
was used for screening.
4.2. Enzyme solutions
Lipase from O. piliferum (NRRL 18917) was prepared
as previously described4 and diluted 10-fold with BES
buffer for screening. SDE11 and PFE8 were prepared as
previously described and lyophilized. The lyophilized
powder was dissolved in BES buffer (7 mg powder/mL,
1 mM BES) and the pH was readjusted to 7.2. Protein
assay (BioRad dye binding assay calibrated with bovine
serum albumin) showed a protein concentration of 5
mg protein/mL. A 1:10 dilution of this solution was
used for both Quick E measurement and the scale-up
reactions. BTL2 was expressed in E. coli JM105 as
described previously.2 After growing at 30°C to an
BTL2 catalysed the slow hydrolysis of solketal butyrate
with moderate enantioselectivity, E=7.9, favouring the

A. M. F. Liu et al. / Tetrahedron: Asymmetry 12 (2001) 545–556
555
OD600 of 1.0, the culture was heated to 42°C for 3 h to
induce lipase expression. (A heat shock promoter con-
trolled the lipase expression.) The culture was cen-
trifuged (4°C, 6000×g, 15 min) and the cells were
washed twice with BES buffer (1 mM, pH 7.20). The
bacterial pellet was resuspended in BES buffer (5 mL),
sonicated (2 min on ice), and centrifuged to remove cell
debris (4°C, 6000×g, 10 min). The supernatant was
adjusted to pH 7.20 and used in screening.
base indicated 25–75% conversion, the mixture was
extracted with ethyl ether. The extracts were dried,
concentrated and analysed by gas chromatography on
the column noted above. Enantiomers of both solketal
(h=1.03, kR=6.47, kS=6.64) and solketal octanoate
(h=1.03, kS=24.2, kR=24.8) were separated using the
following temperature program: initially 20 min at
100°C, then increased to 150°C over 10 min, then held
at 150°C for 30 min. Data are shown in Table 5.
4.3. Ester library
4.7. Enantioselectivity of PFE toward methyl 3-bromo-
2-methylpropionate 39
Unless otherwise noted esters for the substrate library
were purchased from Aldrich (Oakville, ON) or Fluka
(Oakville, ON). Resorufin esters, solketal butyrates 30
and 31 and 1-phenylethyl butyrate 45 were available
from previous work.1 Stock solutions of ester (100 mM
in acetonitrile) were stored at −20°C. These stock solu-
tions were diluted to 14.3 mM for preparing the screen-
ing solution.
The enzyme solution (100 mL in 1 mM BES, pH 7.2)
was added to a solution of ( )-methyl 3-bromo-2-
methylpropionate (5.0 mM) in BES buffer (1.0 mM, pH
7.2, 10 mL) containing 7 vol% acetonitrile and Triton
X-100 (0.3 mM). The mixture was stirred rapidly and
the pH was maintained at 7.2 by the addition of NaOH
(0.10 M) controlled by a pHstat. The reaction vessel
was sealed with parafilm to minimize evaporation of
acetonitrile. When the consumption of base indicated
40% conversion, the mixture was extracted with ethyl
acetate to remove the unreacted ester. The extracts were
dried, concentrated, dissolved in acetonitrile and
analysed by gas chromatography on the column noted
above: isothermal at 95°C, kS=11.8, kR=12.3. The
aqueous phase containing the product acid was aci-
dified to pH 1 or 2 with conc. HCl and extracted with
ethyl acetate. The extracts were dried and concentrated
by rotary evaporation. The residue was dissolved in
methanol (1–2 mL containing 5–6 drops of conc.
H2SO4) and refluxed under nitrogen for 1–2 h. The
reaction was then neutralized with NaOH (1N) to pH
6–7 and ethyl actetate (7 mL) and water (5 mL) were
added, and the phases separated. The organic layer was
washed with water (5 mL), which was then added to the
aqueous fraction. The aqueous fraction was washed
with ethyl acetate (3×3 mL) and the washes were added
to the organic fraction, which was then dried over
MgSO4, and concentrated. The resulting methyl ester
was dissolved in acetonitrile and analysed by gas chro-
matography as above. Data are shown in the footnotes
of Table 4.
4.4. Screening
Rates of ester hydrolysis both with and without a
reference compound were measured colorimetrically as
described previously.1 The reaction mixtures contained
1.0 mM substrate, but the buffer concentration was
reduced from the recommended 5.0 mM BES to 1.0
mM BES to increase the sensitivity of the assay, while
compromising some accuracy.
4.5. (R)- and (S)-solketal octanoate
Octanoyl chloride (13 mmol, 2.1 mL) was added drop-
wise to a cold (ice bath), stirred solution of (R)- or
(S)-solketal (5.0 mmol, 660 mg) and pyridine (9.0
mmol, 750 mL) in anhydrous diethyl ether (20 mL).
After stirring for an additional 2 h, TLC (silica gel, 8:2
hexane/ethyl acetate) showed no remaining solketal.
The mixture was washed with water, saturated sodium
bicarbonate, and brine and dried over magnesium sul-
fate. Flash chromatography (silica gel, 8:2 hexane/ethyl
acetate) yielded a clear oil: 63–72% yield. 1H NMR
(CDCl3, 400 MHz) l 0.86 (t, 3H, octanoyl CH3), 1.28
(m, 8H, octanoyl CH2), 1.36 (s, 3H, solketal CH3), 1.42
(s, 3H, solketal CH3), 1.61 (m, 2H, octanoyl CH2), 2.33
(t, 2H, octanoyl CH2), 3.72 (m, 1H, 1H of CH2),
4.0–4.27 (m, 3H, 1H of CH2, CH2), 4.30 (m, 1H, CH);
MS (CI, NH3) m/z: 259 (M+H+, 5); 243 (81); 201 (50);
127 (47); 101 (100); 82.9 (36); 72 (42). The enantiomeric
purities were ]98% e.e. as determined by GC (Chiral-
sil-DEX CB, 25 m×0.25 mm, Chrompack, Raritan, NJ)
120°C, kR=36.2, kS=36.5 and h=1.01.
4.8. Enantioselectivity of BTL2 toward 1-phenylethanol
and 1-phenylpropanol
The hydrolysis of racemic 1-phenylethyl acetate (0.3
mmol) was performed in sodium phosphate buffer (1
mL, 50 mM, pH 7.5) and toluene (5 mL). Acylation of
1-phenylethanol and 1-phenylpropanol (0.3 mmol) was
performed using vinyl acetate or vinyl butyrate (1
mmol) in toluene (5 mL). Both reactions were accom-
plished at 40°C using 1000 U (based on pHstat assay
using tributyrin) crude extract of BTL. Samples from
the reaction mixture were centrifuged and, for hydroly-
sis experiments, extracted with methylene chloride (300
mL). The organic phase was analysed by gas chro-
matography (130°C isothermal, 60 kPa pressure, FS-
Cyclodex b-I/P CS-Fused silica capillary column, 50
m×0.25 mm, CS-Chromatographie Service GmbH,
Langerwehe, Germany) to give the enantiomeric excess
4.6. Enantioselectivity of BTL2 toward solketal
octanoate 31
A BTL2 solution was added to a suspension of ( )-
solketal n-octanoate (13.5 mg, 0.0522 mmol) in BES
buffer (50 mL, 10.0 mM, pH 7.2) containing 7.0 vol%
acetonitrile. The mixture was stirred rapidly and the pH
was maintained at 7.2 by the addition of NaOH (0.10
M) controlled by a pHstat. When the consumption of

556
A. M. F. Liu et al. / Tetrahedron: Asymmetry 12 (2001) 545–556
of both remaining starting material and product. The
enantioselectivity was calculated as described by Chen
et al.24 Absolute configurations were assigned by com-
parison with pure enantiomers.
5. Gao, Y.; Breuil, C. World J. Microbiol. Biotechnol. 1998,
14, 421–429.
6. Choi, K. D.; Jeohn, G. H.; Rhee, J. S.; Yoo, O. J. Agric.
Biol. Chem. 1990, 54, 2039–2045.
7. Pelletier, I.; Altenbuchner, J. Microbiology 1995, 141,
459–468.
8. (a) Krebsfa¨nger, N.; Schierholz, K.; Bornscheuer, U. T.
J. Biotechnol. 1998, 60, 105–112; (b) Krebsfa¨nger, N.;
Zocher, F.; Altenbuchner, J.; Bornscheuer, U. T. Enzym.
Microb. Technol. 1998, 22, 641–646; (c) Zocher, F.; Kreb-
sfa¨nger, N.; Yoo, O. J.; Bornscheuer, U. T. J. Mol. Catal.
B Enzym. 1998, 5, 199–202.
9. Wei, Y.; Schottel, J. L.; Derewenda, U.; Swenson, L.;
Patkar, S.; Derewenda, Z. S. Nat. Struct. Biol. 1995, 2,
218–223.
4.9. Enantioselectivity of BTL2 toward g- and d-
lactones
The hydrolysis was carried out in phosphate buffer (0.1
M, pH 7) for 48 h and the enantiomeric excess of the
remaining substrate was measured by gas chromatogra-
phy on
a
chiral stationary phase as described
previously.25
Acknowledgements
10. Tesch, C.; Nikoleit, K.; Gnau, V.; Go¨tz, F.; Bormann, C.
J. Bacteriol. 1996, 178, 1858–1865.
11. Khalameyzer, V.; Bornscheuer, U. T. Biotechnol. Lett.
We thank NSERC (Canada) for financial support. We
thank Professor Ted A. Meighen (McGill Biochemistry)
for cultures of E. coli JM105 and DH5a, Mr. Nadim
Saadeh for running the mass spectra, Dr. Lana Janes
for helpful discussions and Ms. Ebru Togan-Tekin for
help with the scale up reaction of BTL2.
1999, 21, 101–104.
12. http://www.juelich-chemicals.de
13. Review: Verger, R. Trends Biotechnol. 1997, 15, 32–38.
14. (a) Bornscheuer, U. T.; Altenbuchner, J.; Meyer, H. H.
Bioorg. Med. Chem. 1998, 58, 554–559; (b) Bornscheuer,
U. T.; Altenbuchner, J.; Meyer, H. H. Bioorg. Med.
Chem. 1999, 7, 2169–2174.
15. Bonafos, A.; Rix, P.-E.; Kazlauskas, R. J. et al., unpub-
References
lished results.
16. Baumann, M.; Hauer, B. H.; Bornscheuer, U. T. Tetra-
hedron: Asymmetry 2000, 11, 4781–4790.
1. (a) Janes, L. E.; Kazlauskas, R. J. J. Org. Chem. 1997,
62, 4560–4561; (b) Janes, L. E.; Lo¨wendahl, A. C.;
Kazlauskas, R. J. Chem. Eur. J. 1998, 4, 2317–2324; (c)
Janes, L. E.; Cimpoia, A.; Kazlauskas, R. J. J. Org.
Chem. 1999, 64, 9019–9029; (d) Janes, L. E.; Lo¨wendahl,
A. C.; Kazlauskas, R. J. PCT/US99/14448; (e) Janes, L.
E.; Lo¨wendahl, A. C.; Kazlauskas, R. J., manuscript in
preparation.
2. (a) Schmidt-Dannert, C.; Ru´a, M. L.; Atomi, H.; Schmid,
R. D. Biochim. Biophys. Acta 1996, 1301, 105–114; (b)
Ru´a, M. L.; Schmidt-Dannert, C.; Wahl, S.; Sprauer, A.;
Schmid, R. D. J. Biotechnol. 1997, 56, 89–102; (c)
Schmidt-Dannert, C.; Ru´a, M. L.; Schmid, R. D.
Biochem. Soc. Trans. 1997, 25, 178–182; (d) Ru´a, M. L.;
Atomi, H.; Schmidt-Dannert, C.; Schmid, R. D. Appl.
Microbiol. Biotechnol. 1998, 49, 405–410.
17. Liu, A. M. F.; Mezzetti, A.; Guerard, C.; Togan-Tekin,
E.; Kazlauskas, R. J. et al., unpublished results.
18. Sikorski, J. A.; Getman, D. P.; Decrescenzo, G. A.;
Devadas, B.; Freskos, J. N.; Lu, H.-T.; Mcdonald, J. J.
1997 WO 9718205; Chem. Abstr. 127, 65754.
19. (a) Fadel, A.; Khesrani, A. Tetrahedron: Asymmetry
1998, 9, 305–320; (b) Atlan, V.; Racouchot, S.; Rubin,
M.; Bremer, C.; Ollivier, J.; De Meijere, A.; Salaun, J.
Tetrahedron: Asymmetry 1998, 9, 1131–1135.
20. Tait, A.; Colorni, E.; Di Bella, M. Tetrahedron: Asymme-
try 1996, 7, 2703–2706.
21. Elferink, V. H. M.; Kierkels, J. G. T.; Kloosterman, M.;
Roskam, J. H. 1990 EP 369553; Chem. Abstr. 114, 41029.
22. Czeskis, B. A.; Moissenkov, A. M. J. Chem. Soc., Perkin
Trans. 1 1989, 1353–1354.
3. (a) Farrell, R. L.; Blanchette, R. A.; Brush, T. S.; Hadar,
Y.; Iverson, S.; Krisa, K.; Wendler, P. A.; Zimmerman,
W. J. Biotechnol. 1993, 30, 115–122; (b) Brush, T. S.;
Farrell, R. L.; Ho, C. TAPPI J. 1994, 77, 155–160.
4. Brush, T. S.; Chapman, R.; Kurzman, R.; Williams, D.
P. Bioorg. Med. Chem. 1999, 7, 2131–2138.
23. Henke, E.; Bornscheuer, U. T. Biol. Chem. 1999, 380,
1029–1033.
24. Chen, C. S.; Fujimoto, Y.; Girdaukas, G.; Sih, C. J. J.
Am. Chem. Soc. 1982, 104, 7294–7299.
25. Enzelberger, M. M.; Bornscheuer, U. T.; Gatfield, I.;
Schmid, R. D. J. Biotechnol. 1997, 56, 129–133.
.
.