Mapping the substrate selectivity and enantioselectivity of esterases from thermophiles

Article abstract of DOI:10.1016/j.tetasy.2004.07.044

To identify potential applications of nineteen esterases from thermophiles, we mapped their substrate selectivity and enantioselectivity using a library of 50 esters. We measured the selectivities colorimetrically using Quick E, which uses pH indicators to detect hydrolysis and a chromogenic reference compound as an internal control. The substrate selectivity mapping revealed one esterase, E018b, with a strong preference for acetyl esters (14- to 25-fold over hexanoate). The enantioselectivity mapping revealed a number of cases of high enantioselectivity. Thirteen of the 19 esterases showed moderate or better enantioselectivity (>19) toward 1-phenethyl butyrate favoring the (R)-enantiomer and two esterases (E008, E013) showed moderate or better enantioselectivity (>20) toward methyl 2-chloropropionate favoring the (S)-enantiomer. Three esterases (E001, E004, E005) showed high (>46) enantioselectivity toward menthyl acetate favoring the (R)-enantiomer. This rapid mapping of the selectivity simplifies the characterization of new enzymes.

Full text of DOI:10.1016/j.tetasy.2004.07.044

Tetrahedron:
Asymmetry
Tetrahedron: Asymmetry 15 (2004) 2991–3004
Mapping the substrate selectivity and enantioselectivity of
esterases from thermophiles
Neil A. Somers and Romas J. Kazlauskas*
´ ´
McGill University, Department of Chemistry, 801 Sherbrooke Street West, Montreal, Quebec, Canada H3A 2K6
Received 7 June 2004; accepted 13July 2004
Available online 12 September 2004
Abstract—To identify potential applications of nineteen esterases from thermophiles, we mapped their substrate selectivity and
enantioselectivity using a library of 50 esters. We measured the selectivities colorimetrically using Quick E, which uses pH indicators
to detect hydrolysis and a chromogenic reference compound as an internal control. The substrate selectivity mapping revealed one
esterase, E018b, with a strong preference for acetyl esters (14- to 25-fold over hexanoate). The enantioselectivity mapping revealed
a number of cases of high enantioselectivity. Thirteen of the 19 esterases showed moderate or better enantioselectivity (>19) toward
1-phenethyl butyrate favoring the (R)-enantiomer and two esterases (E008, E013) showed moderate or better enantioselectivity
(>20) toward methyl 2-chloropropionate favoring the (S)-enantiomer. Three esterases (E001, E004, E005) showed high (>46) enantio-
selectivity toward menthyl acetate favoring the (R)-enantiomer. This rapid mapping of the selectivity simplifies the characterization
of new enzymes.
ꢀ 2004 Elsevier Ltd. All rights reserved.
1. Introduction
ThermoGen.5 These esterases come from thermophiles
and are stable at high temperatures. Dimirjian et al.
have already mapped some of their substrate selectivity.6
Enzymes from thermophiles (microorganisms that grow
in unusual environments such as hot springs) may be
useful for chemical synthesis.1 For example, they allow
using higher temperatures where reactions are faster
and substrates are more soluble.2 In addition, these en-
zymes may also tolerate other unusual conditions such
as high concentrations of organic solvents. Although
thermophiles and other extremophiles are difficult to
grow in the laboratory, advances in molecular biology
permit researchers to produce extremophile enzymes in
microorganisms such as E. coli that grow more easily
in the laboratory.3
2. Results
2.1. Screening method
Preliminary screening measured the initial rate of
hydrolysis of different esters. The rate of reaction was
measured by monitoring the release of protons using a
pH indicator, 4-nitrophenol (Fig. 1a). The yellow color
of the phenoxide fades as the reaction proceeds and
reveals quantitatively the amount of ester hydrolyzed
as described previously.2
To apply these new enzymes in organic synthesis, one
needs some idea of their substrate range and selectivity.
Unfortunately, substrate mapping of enzymes is still a
slow and tedious process. Recently we developed a fast
colorimetric method for measuring the selectivity of
hydrolases.4 Herein, we apply this fast colorimetric
method to map the selectivity of nineteen esterases from
The second screening measured selectivity. Selectivity
measurements require a competition between sub-
strates.7 We used a reference compound, usually resoru-
fin acetate, as the competitive substrate (Fig. 1b).
Hydrolysis of resorufin acetate yields the pink resorufin.
We previously used this method to measure enantio-
selectivity (Quick E).2 In this paper we adapt this
method to also measure substrate selectivity (Quick S).
*
Corresponding author at present address: University of Minnesota,
Department of Biochemistry, Molecular Biology and Biophysics,
The Biotechnology Institute, 1479 Gortner Avenue, Saint Paul, MN
55108, USA. Tel.: +1 612 624 5904; fax: +1 612 625 5780; e-mail:
rjk@umn.edu
Accurate measurement of the selectivities requires that
the substrate and the reference compound react at
0957-4166/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetasy.2004.07.044

2992
N. A. Somers, R. J. Kazlauskas / Tetrahedron: Asymmetry 15 (2004) 2991–3004
probe acyl chain selectivity
O
O
hydrolase
(a)
+
R'OH
+
H
R'
vinyl esters
ethyl esters
methyl esters
BES, pH 7.2
R
O
R
O
O
O
O
O
NO2
R
O
R
O
R
O
yellow
HO
NO2
12, R = CH3
1, R = CH3
19, R = CH2Br
13, R = (CH2)2CH3
14, R = (CH2)4CH3
15, R = (CH2)6CH3
16, R = (CH2)8CH3
17, R = CF3
2, R = CH2CH3
3, R = (CH2)2CH3
4, R = (CH2)4CH3
5, R = (CH2)6CH3
6, R = (CH2)8CH3
7, R = (CH2)12CH3
8, R = (CH2)14CH3
9, R = (CH2)16CH3
10, R = Ph
20, R = CH2OH
colorless
21, R = CH(OH)C6H5
22, R = benzyl
23, R = p-C6H4CH2OH
O
O
(b)
+
R'OH
+
H
O
R'
R
O
hydrolase
R
O
18, R = CH2Br
31, R = CH2NH2 HCl
BES, pH 7.2
O
O
O
O
O
N
O
N
11, R = t-butyl
pink
O
probe alcohol moiety selectivity
+
+
1.1 H
O
butyrates
acetates
O
Figure 1. Measuring selectivity of hydrolases. (a) Estimated selectiv-
ities compared the initial rates of hydrolysis of esters measured
colorimetrically using 4-nitrophenol as a pH indicator. (b) Quantita-
tive measure of selectivity used resorufin esters as internal standards.
O
R
R
O
O
13, R = CH2CH3
3, R = CH=CH2
12, R = CH2CH3
24, R = (CH2)2CH3
25, R = (CH2)3CH3
26, R = Ph
30, R =
O
comparable rates to measure both reaction rates. When
the substrate hydrolysis was much slower than resorufin
acetate hydrolysis, we replaced resorufin acetate with
the slower reacting resorufin pivaloate or resorufin iso-
butyrate as reference compounds and extended the
measurement time. However, even after this change,
some substrates reacted too slowly for an accurate meas-
ure of selectivity using this Quick E/S method.
O
1, R = CH=CH2
27, R = C(CH3)=CH2
28, R = CH(CH3)CH2CH3
O
29, R =
O
O
Scheme 1. Esters used to survey the substrate selectivity of esterases.
Some substrates are listed more then once because they fit into more
than one category.
2.2. Survey of substrate selectivity
petitive experiment. We used resorufin esters as the com-
petitive substrate as explained above and normalized the
results to hexanoate, Table 2. Details for selected exper-
iments are given in Table 3 and Figure 2 shows a gray-
scale array of the selectivities toward vinyl esters. The
trends for the true selectivities were similar to the esti-
mated selectivities. Among the vinyl esters, the hexano-
ate or octanoate was the best substrate for all except
E018b, where the acetate was the best substrate.
E018b favored vinyl acetate 18-fold over vinyl hexano-
ate (1800/100, Table 2). The acetate was a very poor sub-
strate for many esterases (e.g., E006, E008, E011, E014).
For example, E006 favored vinyl hexanoate >700-fold
over vinyl acetate (100/0.13, Table 2).8
To survey the substrate selectivity of the thermophile
esterases, we first measured their ability to hydrolyze
31 different commercially available esters (Scheme 1,
Table 1). Good substrates showed a specific activity
>1lmol ester hydrolyzed/mg protein/min, while very
good substrates showed a specific activity >10. (The spe-
cific activities listed in the Tables are multiplied by 100
for convenience.) The best substrates were activated
esters, which are also chemically the most reactive in
the library: vinyl esters 1–11, ethyl trifluoroacetate 17,
and phenyl acetate 26. Among the vinyl and ethyl esters
with different chain lengths, most of the esterases appear
to favor the hexanoate or octanoate, while E018b
appears to favor acetate esters. The sterically hindered vi-
nyl pivaloate 11 was a poor substrate for all the esterases,
as was vinyl benzoate 10. Polar esters, for example, 20,
21, 30 were usually poor substrates. Ethyl butyrate 13
reacted much slower that the more hydrophobic tributy-
rin 29. This preference for hydrophobic substrates sug-
gests that these esterases may be related to lipases.
The hexanoate versus acetate selectivity varied for other
esters, but did not reverse selectivity. For the ethyl es-
ters, E018b favored ethyl acetate 21-fold over ethyl hex-
anoate (1400/100, Table 2), while E002 favored ethyl
hexanoate >1000-fold over ethyl acetate (100/<0.48,
Table 2). For the phenyl esters, E018b favored phenyl
acetate 25-fold over the hexanoate (2500/100, Table 2).
E011 favored phenyl hexanoate 47-fold over phenyl ace-
tate (100/2.1, Table 2). In comparison, this enzyme
showed much higher chain length selectivity for the vinyl
esters: it favored vinyl hexanoate 1100-fold over vinyl
acetate (100/0.089, Table 2). For the glycerol esters
E018b favored triacetin 19-fold over trihexanoin (hexa-
noic acid triglyceride, 2800/150, Table 2). Confirming
the selectivity of E018b for acetyl groups, E018b
only catalyzed the hydrolysis of acetyl esters, but not
Three pairs of esterases (E010/E011, E013/E014, E019/
E020) showed similar reaction rates with all substrates
and therefore might be very similar or even identical
enzymes.
2.3. Acyl chain length selectivity
We determined the true selectivity of the thermophile
esterases toward different acyl chain length using a com-

Table 1. Specific activities of thermophile esterases toward esters 1–31a
Substrate
Esterase
E010 E011
E001
E002
E003E004
6360
940730 2202107802601302017400114600
E005
E006
E007
E008
E009
340
E012
3.1
E013E014
3.0
E015
1.2
E016
89
E018b
63
E019
E020
1
150160
140120
2.312.0
5.5
5.1
150121010
20
1.1
2
11
37
140170350220
240180280230
66005014300110
400
370
3
860
1000
3000
1700
1600
0.53 16
0.06
300
790
760
960
150
550
410
1700
960
76
320
980
300
2.8
2000
3800
1800
1600
190
630
220
240
220
680
260
280
490550
2400
8080120
600
4
4900
2100
1800
2.6
1400
1600
<5
4100
1200
1500
510
160
10
48
50
9.3
2.0
3000
590
5
6700
190170330
4600
320
1200
15
6
0.55
0.430.88
0.13
5.7
27
4.8
7
13 1.1
0.61
2.1
0.930.45
7.6
0.44
0.81 .0
<2
22
3
21
5.6
0.68
1.0
8
0.19
<5
44
3.0
58
0.13
120
1.6
<0.1
1.4
<0.2
6.5
<0.2
230
27
<0.2
2.6
<0.1
2.7
0.10
1.7
<0.1
1.0
0.16
5.9
2.2
50
1.3
19
<0.2
59
<0.2
14
11041010
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
1.9
ND
2.1
8.2
ND
8.8
ND
96
44
0.52
ND
0.02
ND
0.19
ND
3.7
3.2
1.2
0.80
ND
0.71
ND
16
0.42
ND
0.40
ND
0.40
ND
5.1
4.0
20
1.1
ND
<1
0.58
ND
1.5
1.2
ND
3.3
ND
26
ND
2.8
ND
2.4
ND
120
ND
ND
1.1
ND
<2
ND
2.0
ND
3.5
ND
3.0
4.1
ND
1.1
ND
0.63
ND
ND
ND
51
ND
250160
ND
4.4
ND
ND
ND
19
ND
120
ND
190
ND
8319
ND
1500
16
ND
1.2
ND
1100
19
ND
2.6
ND
280240
ND
8.2
1205705090120
6.7
ND
ND
320
75
ND
1600
25
ND
1400
54
ND
1.7
ND
1.9
ND
8.8
15
ND
750
410
ND
4.1
6.6
ND
2.7
5.5
ND
1200
5.9
ND
ND
1.2
ND
0.35
3.1
2.31.4
920066010
7.6
19
8.8
66
4.8
1.0
2.5
4.9
9.9
<5
9.34.1
130.15
<2
1.9
0.27
<0.1
<0.1
0.16
4.397
<0.2
<0.1
<0.1
0.22
4.4
1.8
0.52 0 0.3<0.1
1.6
2.4
<2
1.30.732.6
7.8
1.9
1.8
1.6
0.25
0.12
0.51
0.730.50
1.30.82
1.9
0.15
<0.05
0.30
<5
0.11
<0.1
0.27
<0.1
<0.1
<0.1
<0.1
<0.1
<2
<0.2
<0.2
<5
<5
<0.2
0.21
<0.2
0.83
<0.2
0.26
0.04
<0.1
<2
<2
<0.1
<0.1
<0.1
0.14
<2
<2
<1
<1
<0.2
0.21
<0.2
0.28
24
<2
<2
6.1
0.45
0.69
0.27
0.3<0.1
0.19
0.22
<0.1
0.63<0.2
0.48
20
<0.1
0.25
<0.2
<2
2.8
<0.1
<0.1
<0.1
<0.1
0.35
210150320190150 250390
0.13<2
0.26
0.21
<1
0.24
0.47
<5
1
<2
230
48
0.49
1.6
0.85
2.0
1.6
2.8
1.5
0.54
0.85
0.5
<2
30
1.0
13
520490 350290057030
87
3024100129700
5.2
0.5
011010
0.66
3
<5
8
56150
0.38
69
0.16
0.92
<0.1
88
250320590
4.3
2.4
<2
37
2.0
0.19
0.73
<0.1
100
0.29
49
<2
75
12
3.1
3.3
2.1
0.52
<2
0.37
0.19
450
1.2
0.51
1400
1.2
720570 190290640370
600
3.5
3000
94
430
1.4
330280550130
3.4
1.6
250
1600
2.0
14
<0.1
58
<0.05
22
14
9.9
12
65
<0.1
3.5
<0.1
0.65
<0.1
3.4
6.0
1.0
<0.1
0.53
<0.1
3.2
<0.1
14
13
<0.2
<0.2
<0.2
<0.1
0.9
1.4
a In lmol ester hydrolyzed/min/mg protein · 100. Bold numbers highlight all values >100. ND = not determined.

2994
N. A. Somers, R. J. Kazlauskas / Tetrahedron: Asymmetry 15 (2004) 2991–3004
Table 2. Acyl chain length selectivity of thermophile and several acetyl selective esterases toward vinyl, phenyl, glyceryl, and ethyl esters
(hexanoate = 100)a
Esterase
Vinyl esters
Phenyl esters
C6
Glycerol esters
C4, 29 C6
Ethyl esters
C2, 1 C3, 2 C4, 3 C6, 4 C8, 5 C10, 6 C2, 26 C4
C8 C2
C2, 12
ND
C6, 14 C10, 16
E001
E002
E0030.97
E004
E005
E006
E007
E008
E009
E010
E011
E012
E.00133
E014
E015
E016
E018b
E019
E020
AcCEf
AcEf
0.45b
3.4
71
12
260
100
100
49
650
11c
9.5
5.9
6.0
3.3
3.0
5.2
12
47
30
100 160 <0.053b
100
100
100
100
100
100
100
72d
ND
100d
<0.053b 100d
ND
230d
55d
19d
300c
100 32
<0.075b
120d <0.48b
d
c
b
7.319
0.70
11
100
46
22
173100 180 <1.23
147 100 110 <0.015b
33d
6.6d ND
89d
40d
16d
77d
5.9c ND
2.1c ND
2.4c <0.76b
6.4c ND
42c
21d
0.18d
6.23d
0.13c
2.0d
0.56d
0.14d
0.38d
7.9
29
82
100
100
100
8.0
96
2.1d
48c
70c
c
ND
ND
ND
ND
100d
ND
ND
100d
ND
ND
ND
ND
ND
100d
ND
ND
ND
ND
ND
ND
ND
ND
23d
16
40
100 45
100 39
<0.050b
<0.20b
ND
ND
26
4.316
330
100
7.37.8
8.3100 10
100
18
100
18
18
100
64
13
141 100 240 <2.0b
ND
<0.20b
d
d
5.35.3100
3.1
2.3140 100 68
39
100 24
0.0010
<0.0020b 100
d
0.86
2.7
1.8
1.4
100
100
ND
ND
15d
9.8
14
8.5d
2.0
2.1
65
60
74
77
98
95
74
100 18
100 27
<0.016b
0.089d 0.08
13100
7.8
<0.0045b 100
d
0.11c
2.0
19
26
4.1c
d
6.7
3.3
100 110 0.17d
100
ND
ND
ND
ND
ND
420c
ND
ND
ND
ND
d
13100
13100
13
40
4.6
100 56
100 73<0.15
<0.0011b 100
ND
ND
d
b
0.010d 1.2
1.8
3.2
100
100
100
100d
100
100
ND
100d
0.46d
24d
3.3
25
18
12d
30c
100 44
100 81
310c 100c 48c 2800c
<0.015b
<0.065b
1.0c ND
9.8c ND
150d 1400c,e
5.1c ND
5.4c ND
ND ND
77
100
192
220c
230
220
23c
6.5
1800c 680c
74c
53
100c
100
110c
36c
2500c
3.4
1.7c
1.5d
6.2c
850c
13
54
50
100 23
100 29
<0.067b
0.010d
12
41
100
34c
3.2
ND
3500c
6.9c
370c
5.4c
140c
100c
100c
25c
ND ND ND ND
122c 100d 67d 730c
150c
160c
87d
ND
a True selectivities (estimated error limits are 10–30%) measured using resorufin acetate as the reference compound unless otherwise noted.
ND = not determined. Measurements using different reference compounds are scaled so that relative values for one enzyme and a substrate
series (e.g., vinyl esters) can be compared. Figure 2 comes from the data for vinyl esters in this table, scaled by the vinyl hexanoate activities listed in
Table 1.
b Substrate reacted much slower than all available resorufin esters. The value listed is an estimated upper limit.
c Resorufin isobutyrate as the reference compound.
d Resorufin pivaloate as the reference compound.
e Estimated error limits are larger, 50%.
f AcCE = acetylcholine esterase; AcE = acetyl esterase from orange peel.
Table 3. Representative experimental data for Quick S and Quick E measurements in Tables 2 and 9a
Ester
Substrate
Specific activityd
Referenceb
Specific activityd
0.22
0.3
Selectivitye
3.0
Quick E/Sf
Obsvd ratec
Obsvd ratec
AcE
1
ꢀ16
ꢀ15
ꢀ6
0.67
0.091
3.2
20
230.26
3
8.5 (1)
4.9 (S)
5.1 (S)
2.8 (S)
S)
4
5
E005
E011
E013(
E014
(R)-45
(S)-45
(R)-45
(S)-45
R)-45
(S)-45
(R)-45
(S)-45
0.48
0.47
0.33
0.18
0.11
0.10
02.173(3
6.7
33
ꢀ24
ꢀ34
ꢀ69
ꢀ38
ꢀ82
ꢀ24
ꢀ54
15
2
33
0.86
2.4
2.6
13
17
22
0.58
1.5
5.3
15
20
18
0.47
1.317
.6
0.139.9
a AcE is acetyl esterase from orange peel.
b Quick E measurements used resorufin acetate, while Quick S measurements resorufin pivaloate as reference compound.
c Observed rate is in mOD/min and is an average of three or four readings, typical errors are 10–20%. Negative values reflect the fading yellow color
of the pH indicator during the reaction, while positive values reflect the increase in the pink color of resorufin.
d In U/mg protein; reactions contained differing amounts of protein.
e Selectivity for the substrate versus the reference compound.
f The favored substrate or enantiomer is in parentheses.
butanoyl or other esters, in the enantiomer pair library
(see below). For the vinyl, phenyl, glyceryl, and ethyl es-
ters E018b favored acetyl esters 14- to 25-fold over hexa-
noyl esters.
from orange peel (AcE) and acetylcholine esterase from
electric eel (AcCE). AcE has been used several times in
organic synthesis for removal of acetyl groups.10 For
the vinyl esters, AcE favored an acetyl group 8.5-fold
over a hexanoyl group (850/100, Table 2), approxi-
mately a 2-fold lower selectivity than E018b, 18-fold,
Table 2. For the glycerol esters, AcE preferred the ace-
To compare E018b to other acetyl-selective enzymes, we
measured the chain length selectivity of acetyl esterase

N. A. Somers, R. J. Kazlauskas / Tetrahedron: Asymmetry 15 (2004) 2991–3004
2995
AcO
E018b
O
E018b
40: 1
AcO
OMe
O
NH
t-Bu
O
hexanoyl
O
monoacetate monohexanoate of
O-acetyl-N-Boc L-serine
2,2-bis(4-hydroxyphenyl)propane
methyl ester
Figure 2. Acyl length selectivity of thermophile esterases toward vinyl
esters. Darker squares correspond to higher activity. Most esterases
favor either vinyl hexanoate or octanoate, but E018b and AcE (acetyl
esterase from orange peel) favor vinyl acetate. To show selectivity
clearly, the absolute enzyme activities were scaled as indicated. This
grayscale array representation was created from the data in Tables 1
and 2 as described by Reymond et al.9 AcCE = acetylcholine esterase.
Figure 3. Chemoselective hydrolysis of acetyl esters with E018b. (a)
Selective removal of an acetate ester with E018b and a hexanoate ester
with E015 in a mixed diester of a bisphenol. (b) Selective removal of
the acetyl group in a serine derivative with no detectable hydrolysis of
the methyl ester or the tert-butoxycarbonyl group.
rates of hydrolysis of the esterases using 21 pairs of
enantiomers (Scheme 2, Tables 4–7). The ratio of the
initial rates of the enantiomers is an estimated
enantioselectivity.
tate over the hexanoate by 8.4-fold (730/87, Table 2),
approximately 2-fold less then E018b, 19-fold. For the
phenyl esters, AcE favored an acetyl group 35-fold over
a hexanoyl group (3500/100, Table 2), slightly higher
than the selectivity of E018b, 25-fold. AcCE did not
show the expected acetyl preference with vinyl esters,
but favored vinyl hexanoate 16-fold over the acetate
(100/6.2, Table 2). Thus, E018b shows comparable or
higher acetyl selectivity than other acetyl-selective
enzymes.
Substrates with the chirality in the alcohol portion
(Tables 4, 5) usually reacted faster than substrates where
the chirality is in the carboxylic acid portion (Tables 6–
8). The best substrates––35, 36, and 38––were nonpolar
molecules with the stereocenter in the alcohol portion.
The poorest substrates––39, 40, 44, 46, 47, 50, 51, and
52––were polar molecules with the stereocenter in the
carboxylic acid portion. This observation is consistent
with the screening above which also showed that polar
esters were poor substrates.
A competition between vinyl butyrate and acetate mon-
itored by H NMR confirmed the high acetyl selectivity
1
of E018b. The resonances of both substrate vinyl esters,
and the products, butyric, and acetic acid, were moni-
tored over time and revealed a 17-fold acetyl preference,
Table 10 below. In comparison the Quick S measure-
ments showed a 24-fold acetyl preference (1800/74 in
Table 2). For comparision, acetyl esterase from orange peel
showed only a 4-fold preference for acetyl in this exper-
iment. This value is similar to the value of 5.9 (840/144
from Table 2) predicted by the Quick S measurements.
As a complementary enzyme, we tested E015, which
showed an 18-fold preference for butyryl over acetyl,
as compared to the 51-fold preference in Quick S meas-
urements (3.7/0.072 from Table 2). Small differences in
the experimental conditions, such as added detergent
Triton X-100 for the Quick S measurements, but not
the NMR measurements, may cause these differences
in selectivity.
The estimated enantioselectivities were below two in
most cases, but two substrates (38 and 45) showed high-
er estimated enantioselectivities with most thermophile
esterases. In addition, seven other substrates (34, 36,
42–44, 47, 50) showed estimated enantioselectivities
above two with several esterases. These esters were
screened further using the more accurate Quick E
method.
2.5. True enantioselectivity
We measured the true enantioselectivity for esters 34, 36,
38, 43, and 45 using a resorufin ester as a reference com-
pound as described above. When the slow enantiomer
reacted much slower than the reference compound, we
could only set a lower limit on the enantioselectivity.
For several other substrates (42, 44, 47, and 50), both
enantiomers reacted very slowly compared to the resoru-
fin esters used as reference compounds, so we could not
accurately measure enantioselectivity using Quick E.
Results are summarized in Table 9, detailed data for
selected measurements is in Table 3 above.
We also demonstrated the selective removal of an acetyl
group in a mixed diester of bisphenol A using E018b
(Fig. 3). HPLC analysis showed selective removal of
the acetyl group and a 40:1 ratio of the monohexanoate
over the mono acetate. This selectivity was slightly high-
er than that measured for the phenyl esters: phenyl ace-
tate, 26, versus phenyl hexanoate was 25:1, likely as a
result of the slightly different substrates. E018b also
selectively removed the acetyl ester in O-acetyl-N-Boc
L-serine methyl ester.
None of the esterases showed enantioselectivities above
five for 34, 2-methyl glycidyl 4-nitrobenzoate, an ester of
a primary alcohol. This is consistent with the screening
using estimated E (Table 4), where none of the esterases
showed an estimated enantioselectivity of above three.
2.4. Estimated enantioselectivity
To evaluate the potential of the thermophile esterases
for enantioselective reactions, we measured the initial
For menthyl acetate, 36, an acetate of a secondary alco-
hol, the estimated enantioselectivities (Table 5) were also

2996
N. A. Somers, R. J. Kazlauskas / Tetrahedron: Asymmetry 15 (2004) 2991–3004
E006 showed a lower limit of 16. Like other lipases
esters of chiral primary alcohols
and esterases, the thermophile esterases favored the
(R)-enantiomer of 38. These true enantioselectivities
measured by Quick E were usually higher than the esti-
mated E values in Table 5 above.
NO2
O
R
O
O
O
R
O
O
O
32, R = C3H7
solketal butyrate
35, R = C7H15
33, R = H
Quick E measurements with methyl 2,2-dimethyl-1,4-
dioxolane-4-carboxylate, 43, were difficult because the
reactions were slow. Quick E measurement only set
lower limits of >2.2 to >5.5 on the enantioselectivity
for five esterases. By comparison, estimated enantio-
selectivity values for these five esterases ranged from
4.1 to 8.9 (Table 7).
glycidyl 4-nitrobenzoate
34, R = Me
2-methylglycidyl 4-nitrobenzoate
solketal octanoate
esters of chiral secondary alcohols
O
AcO
AcO
O
Most of these esterases showed moderate or better enan-
tioselectivity toward methyl 2-chloropropionate, 45.
Esterase E008 showed the highest enantioselectivities,
21, and the lower limits on the enantioselectivity of est-
erases E010 and E011 were >6.2 and >15. As noted
above, these two esterases are very similar and may be
identical. All esterases favored the (S)-enantiomer.
38
37
36
1-phenylethyl
butyrate
neomenthyl
acetate
menthyl
acetate
esters of chiral carboxylic acids, stereocenter α to carbonyl
COOMe
COOMe
COOMe
OH
O
HO
O
methyl 3-hydroxy-2-
methylpropionate
39
To confirm these Quick E measurements, we also meas-
ured the enantioselectivity of several reactions using
scale-up reaction.11 In most cases, enantioselectivities
measured by Quick E and scale-up reactions agreed with
one another (Table 10).
methyl lactate
methyl 2,2-dimethyl-1,4-
dioxolane-4-carboxylate
43
42
COOMe
COOMe
COOMe
Br
HO
Cl
methyl 3-bromo-2-
methylpropionate
44
methyl
2-chloro-
propionate
45
For substrate 36, the enantioselectivities measured by
Quick E and by the endpoint (scale-up) method agreed.
The enantioselectivity of E004 was high using either
method––>81 using Quick E and >86 using the endpoint
method. Due to low reactivity, Quick E could only set
lower limit of enantioselectivity for esterases E001
(>3.2) and E005 (>2.0), but the endpoint method
revealed that the enantioselectivity was very high, >46
and >500, respectively (Table 10).
methyl mandelate
46
COOEt
COOMe
COOEt
NH
MeOOC
H2N
HN
O
Ph
N
alanine
H2N
methyl ester
tryphtophan
47
ethyl
N-benzylproline
ethyl ester
49
methyl ester
pyroglutamate
50
48
esters of chiral carboxylic acid
lactones
stereocenter β to carbonyl
O
For substrate 38, the enantioselectivities measured by
Quick E and the endpoint method agreed in seven out
of nine cases. The two exceptions were E002 and
E013, where the lower limits on the enantioselectivity
measured by Quick E were slightly higher than the end
point values. In the first case, the Quick E value was
>55, while the end point value was 46. In the second
case, the Quick E value was >38, while the end point
value was 25.
O
COOR
HO
O
O
HO
40, R = Me
CH2OH
dihydro-5-hydroxymethyl- pantolactone
methyl 3-hydroxybutyrate
41, R = Et
2(3H)-furanone
52
51
ethyl 3-hydroxybutyrate
Scheme 2. Enantiomer pairs used to survey the enantioselectivity of
esterases. Only one enantiomer is shown.
below three. However, Quick E identified one esterase,
E004, with high enantioselectivity (E >81), favoring
the (R) enantiomer. In addition, two other esterases,
E001 and E002, may also be highly enantioselective,
but our measured lower limits for the enantioselectivity
were only 3.2 and 8.9, respectively. The favored enanti-
omer was (R), as previously seen for esterases and
lipases.
3. Discussion
This pH-indicator based colorimetric screening allowed
the rapid testing of 19 esterases with more than 50 sub-
strates, representing more than 1000 selectivity measure-
ments. Researchers previously measured the chain
length selectivity of hydrolases by mixing substrates
and measuring the relative amounts of hydrolysis by
TLC or GC.12 Similarly, researchers have developed
GC, MS, and electrophoresis-based methods for meas-
uring enantioselectivity.13 The main advantages of the
pH-indicator based methods are simplicity and speed.
They use unmodified ester substrates, so that adding
more substrates to refine screening is easy. They do
For 1-phenylethyl butyrate, 38, another ester of a sec-
ondary alcohol, most of the esterases showed high enan-
tioselectivity (Table 9). Esterases E001, E002, E004,
E005, E008, E009, E010, E011, E013, E014, E015,
E019, and E020 showed enantioselectivities >19 and

N. A. Somers, R. J. Kazlauskas / Tetrahedron: Asymmetry 15 (2004) 2991–3004
2997
Table 4. Specific activities and estimated enantioselectivities of thermophile esterases toward esters of chiral primary alcoholsa
(R)-32
(S)-32
Est. Eb
(R)-35
(S)-35
Est. Eb
(R)-33
(S)-33
Est. Eb
(R)-34
(S)-34
Est. Eb
E001
E002
E00319
E004
E005
E006
E007
E008
E009
E010
E011
E012
E013
E014
E015
E016
E018b
E019
E020
9.4
7.35.6
7.1
1.3
1.3
200
250160
140
180160
300
100
4.0
140
1.4
1.6
1.5
1.1
1.8
1.7
1.6
1.6
1.6
1.4
1.6
1.0
1.1
1.1
1.2
12
12.5
8.5
1.0
1.1
16
16
36
40
85
2.6
2.4
7.8
18
15
1.3
92
131.4
34
7.2
27
1.7
53
96
27
5.8
20
1.4
1.3
1.3
1.4
1.4
1.4
1.0
1.0
1.1
1.3
1.2
34
1.0
1.0
1.2
1.0
1.4
1.2
1.3
1.2
1.2
1.4
1.2
1.2
68
44
1.3
2.2
2.7
2.9
1.0
1.2
1.1
1.1
1.5
1.0
1.0
1.0
4.4
1.8
0.32
4.2
170
61
6.9
2.5
0.44
6.4
150190
2.9
2.6
5.1
1.4
1.2
7.6
2.4
2.9
3.0
0.43
17
7.9
1.3
16
0.44
6.0
2.5
150
490
110
94
0.46
8.9
93
150110
2.8
2.2
300
81
360320
6.4
6.1
14
2.7
2.2
3.6
3.0
4.3
2.0
1.4
9.2
6.0
5.3
9.6
3.0
3.4
18
58
19
3.2
1.5
3.0
1.1
18
84
94
83
3.0
3.3
18
1.4
7.2
1.2
76
6.31.2
9.31.5
<1.5
2.9
230190
1.6
14
9357
1310
1.311
26
2.3
c
c
c
c
<1.5
3.1
<1.5
120
120
<1.5
76
<1.5
3.8
<1.5
3.1
<1.5
2.7
<1.5
4.8
1.1
1.0
1.6
1.5
1.3
1.8
4.5
4.5
85
6.1
4.8 .9 1.33 8.9
2.3
a Specific activity in lmol ester hydrolyzed/mg protein/min · 100. Rates P100 are bolded for emphasis.
b Estimated enantioselectivity is the ratio of the initial rates of hydrolysis of the two enantiomers (rate for fast enantiomer/rate for slow enantiomer).
c Could not be measured because the substrate did not react.
Table 5. Specific activities and estimated enantioselectivities of thermophile esterases toward esters of chiral secondary alcoholsa
(R)-36
(S)-36
Est. Eb
(R)-37
(S)-37
Est. Eb
(R)-38
(S)-38
Est. Eb
E001
E002
E003
E004
E005
E006
E007
E008
E009
E010
E011
E012
E01376
E014
E015
E016
E018b
E019
E020
0.82
0.54
4.2
1.8
1.4
7.0
120
2.2
2.6
1.6
2.0
0.52
0.08
3.5
0.40
0.08
3.2
1.324
1.1
1.1
1.3
1
2.0
110
37
11
21
5.1
6.6
5.7
5.6
28
62
3.8
4.9
160
0.531.3 2.4
0.45
0.22
150
0.21
0.19
1.1
4.0
0.3 13
0.54
0.26
200
1.2
1.2
1.0
1.0
1.1
1.2
1.2
0.27
0.14
0.32
0.65
0.28
0.14
8.1
0.25
0.14
0.41
1.32.0
0.32
0.19
7.8
1.1
1.0
1.3
5.0
1.2
260
900
16
0.27
0.235.2
57
18
4.6
6.7
11
930930
150130
180150
16
120
1.4
1.1
1.4
1.1
130
12
15
<2
8.6
>9.3
19
70
1.5
0.20
0.23 1.2
5.3 0.46
12
34
93
47
1.4
1.7
1.1
1.7
1.6
1.7
0.13
0.25
5.2
0.14
0.38
5.31.0
6.1
1.1
1.5
5.0
24
0.32
1.5
15
16
150
1.4
6.2
0.68
1.1
1.5
11
.5
23 3 6.4
6.9
1.1
1.0
1.1
1.31.31.0
7.5
9.2
0.44
0.64
0.20
0.31
0.19
0.34
0.45
0.63
17
15
a Specific activity in lmol ester hydrolyzed/mg protein/min · 100. Rates P100 are bolded for emphasis.
b Estimated enantioselectivity is the ratio of the initial rates of hydrolysis of the two enantiomers (rate for fast enantiomer/rate for slow enantiomer).
not require finding GC or other separation conditions
for each new substrate.
reaction avoid reaction in other parts of the molecule.14
This is particularly important in sugars, which contain
many difficult-to-distinguish alcohol residues. Previ-
ously enzymes have been discovered, which chemoselec-
tively hydrolyze phenylacetate,15 phenylalanine ester,16
or acetate14 esters. WaldmannÕs group17 used the acetyl
selectivity of acetyl esterase (AcE) to distinguish be-
tween acetyl groups in acetyl-protected carbohydrates
and nucleosides.18 Several other applications, for exam-
ple, the synthesis of a lipopeptide, relied on the selectiv-
ity of AcE for acetyl groups.19 For example, AcE
removed an acetyl protective group while leaving a
C-terminal peptide allyl ester and a palmitoyl thioester
This screening yielded three main conclusions. First, est-
erase pairs E010/E011, E013/E014, and E019/E020 are
likely identical. Second, esterase E018b is highly selec-
tive for acetyl esters, while the other esterases favor
hexanoyl or octanoyl acyl groups. Third, at least one
of the esterases shows high enantioselectivity toward
substrates, 36, 38, and 45.
Acetyl selectivity can be useful to protect or deprotect
substituents, and the mild conditions of an enzymatic

Table 6. Specific activities and estimated enantioselectivities of thermophile esterases toward esters of chiral carboxylic acids (stereocenter at the a-position)a
(R)-39
(S)-39
Est. Eb
(R)-42
(S)-42
Est. Eb
(R)-44
(S)-44
Est. Eb
(R)-45
(S)-45
Est. Eb
(R)-47
(S)-47
Est. Eb
(R)-50
(S)-50
Est. Eb
c
c
c
c
c
c
c
c
E001
E002
E0036.7
E004
E005
E006
E007
E008
E009
E010
E011
E012
<0.1
<0.05
<0.1
<0.05
0.60
0.45
0.51
0.30
1.2
1.6
0.38
0.19
0.18
0.053
2.1
2.1
2.3
50
20
8.5
3.4
<0.1
<0.05
<2.5
<2
<0.1
<0.05
<2.5
<2
<0.1
0.02
0.22
0.17
>2.2
8.5
c
100
38
c
5.4
1.2
15
15
1.0
<2.5
<2.5
11
13
3.4
2.4
4.37.2
<2
1.7
c
27
29
1.1
c
46
39
1.2
2.4
3.0
0.035
<0.1
<2
>2.3
2.1
31
<2
0.73>7.3
0.28
<0.1
<0.1
<0.06
<0.2
<0.2
<0.1
<0.05
29
<0.1
<0.1
<0.06
<0.2
<0.2
<0.1
<0.05
28
0.37
0.20
<0.06
0.42
1.0
0.16
0.12
<0.06
0.46
1.9
0.015
<0.1
<0.06
0.21
0.79
<0.1
<0.05
<2
56
120
4.9
1.0
7.3
180
2.5
45
2.2
6.1
<0.1
<0.1
<0.06
0.28
<0.1
<0.1
<0.06
0.66
<0.2
<0.1
<0.1
<0.06
<0.2
<0.2
<0.1
<0.05
<2
c
c
c
c
c
c
c
1.6
c
0.81
0.21
3.1
56
>2.8
c
c
<0.06
0.63
3.1
5.1
2.3
<0.06
<0.2
0.33
c
1.1
1.8
1.1
1.5
3.0
4.0
2.4
c
3.2
2.3<0.1
<0.2
<0.1
<0.05
<2
>1.6
c
c
c
c
c
c
c
c
c
c
c
c
0.19
0.050
34
0.21
0.080
37
<0.1
1.1
28
<0.1
<0.05
<2
c
c
c
c
c
c
c
c
<0.05
<2
<0.05
1.6
c
<0.05
<2
<0.05
1.1
c
1.1
c
<2
0.58
<2
1.5
2.8
7.5
28
E013<0.05
<0.05
<0.05
<0.05
0.095
37
<0.05
<0.05
0.25
34
<0.05
2.6
<0.05
<0.05
<0.05
<0.5
<1.5
<0.1
<0.2
<0.05
<0.05
<0.05
<0.5
<1.5
<0.1
<0.2
<0.05
<0.05
<0.05
1.7
c
c
c
E014
E015
E016
E018b
E019
E020
<0.05
<0.05
27
<0.05
<0.05
29
0.0230.015
1.5
5 0.3 0.97
2.9
9.6
<0.05
2.5
2.9
2.6
1.1
0.24
1.2
0.06
1.2
4.0
1.0
<0.05
<0.5
<1.5
1.1
c
>3.4
c
c
c
c
<1.5
0.10
0.64
<1.5
0.16
0.68
<1.5
0.25
0.41
<1.5
0.16
0.29
<1.5
<0.1
<0.2
<1.5
<0.1
<0.2
<1.5
0.74
1.7
<1.5
5.4
7.5
<1.5
<0.1
<0.2
c
c
c
c
1.6
1.1
1.6
1.6
7.3<0.1
4.5
<0.2
a Specific activity in lmol ester hydrolyzed/mg protein/min · 100. Rates P 100 are bolded for emphasis.
b Estimated enantioselectivity is the ratio of the initial rates of hydrolysis of the two enantiomers (rate for fast enantiomer/rate for slow enantiomer).
c Could not be measured because the substrate did not react.

N. A. Somers, R. J. Kazlauskas / Tetrahedron: Asymmetry 15 (2004) 2991–3004
2999
Table 7. Specific activities and estimated enantioselectivities of thermophile esterases toward esters of chiral carboxylic acids (stereocenter at the
a-position)a
(R)-43
(S)-43
8.1
431.2
1.5
Est. Eb
0.130.14
<0.05
(R)-46
(S)-46
Est. Eb
(R)-48
(S)-48
Est. Eb
(R)-49
(S)-49
Est. Eb
E001.2
E002
3
50
0.4
1.1
<0.05
<2.5
0.130.14
c
1.0
4 0.3 0.42
1.2
c
c
c
c
c
c
c
c
c
c
<0.05
<2.5
<2
<0.05
<2.5
<2
<0.05
<2.5
<2
<0.05
<2.5
<2
c
c
c
c
c
c
E0038.2
E004
E005
5.6
<2.5
<2
61
<2
52
1.2
1.2
<2
<2
<0.1
<0.1
<0.1
<0.1
<0.06
<0.2
0.26
<0.1
<0.1
<0.06
0.70
<0.1
<0.1
<0.06
1.1
<0.1
0.15
0.10
0.81
0.69
0.17
0.12
<2
<0.1
0.14
0.17
0.90
0.91
0.25
0.10
<2
E006
E007
0.75
0.11
<0.06
0.22
0.37
<0.1
<0.05
<2
6.6
c
1.1
1.7
1.1
1.3
1.4
<0.06
0.26
0.63
<0.1
<0.05
<2
<0.06
<0.2
0.33
E008
E009
1.2
1.7
1.5
1.1
1.3
c
0.80
0.17
0.85
0.25
c
E010
E011
E012
<0.1
<0.05
<2
<0.1
<0.05
<2
1.4
c
c
c
c
c
c
c
c
c
c
<0.05
<2
<0.05
<2
1.3
c
c
c
c
c
E013<0.05
E014
E015
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
<0.05
0.11
<0.05
0.20
0.131.3
<0.05
0.18
c
<0.05
<0.05
0.130.060
5.31.34.1
<1.5
2.1
1.2
<0.05
1.5
<0.05
1.2
1.1
E016
0.90
0.75
1.2
6.7
8.4
1.2
c
c
c
c
c
c
c
c
c
E018b
E019
E020
<1.5
0.10
<1.5
<0.1
<1.5
<0.1
<1.5
<0.1
<0.2
<1.5
<0.1
<0.2
<1.5
0.230.21
<0.2
<1.5
1.1
<0.2
0.90
1.30.25
8.9
5.4
<0.2
<0.2
a Specific activity in lmol ester hydrolyzed/mg protein/min · 100.
b Estimated enantioselectivity is the ratio of the initial rates of hydrolysis of the two enantiomers (rate for fast enantiomer/rate for slow enantiomer).
c Could not be measured because the substrate did not react.
Table 8. Specific activities and estimated enantioselectivities of thermophile esterases toward esters of chiral carboxylic acids (stereocenter at the
b-position) and lactonesa
(R)-40
(S)-40
Est. Eb
(R)-41
(S)-41
Est. Eb
(R)-51
(S)-51
Est. Eb
(R)-52
(S)-52
Est. Eb
c
c
c
c
c
E001
E002
<0.1
<0.05
<0.1
<0.05
0.16
<0.05
11
0.15
<0.05
9.4
1.1
c
<0.1
<0.05
10
<0.1
<0.05
11
<0.1
<0.05
7.7
<0.1
<0.05
11
c
c
E003<0.25
E004
E005
<0.25
1.2
1.4
1.1
c
1.4
c
41
39
1.1
c
24
33
<2
<2
<2
<2
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
<0.1
<0.1
<0.06
<0.2
0.30
0.24
<0.05
29
<0.1
<0.1
<0.06
<0.2
0.35
0.24
<0.05
31
0.15
<0.1
<0.06
0.27
0.34
<0.1
<0.05
21
0.22
<0.1
<0.06
0.25
0.47
<0.1
<0.05
28
1.4
c
<0.1
<0.1
<0.06
<0.2
<0.2
<0.1
<0.05
<2
<0.1
<0.1
<0.06
<0.2
<0.2
<0.1
<0.05
<2
<0.1
<0.1
<0.06
<0.2
<0.2
<0.1
<0.05
<2
<0.1
<0.1
<0.06
<0.2
<0.2
<0.1
<0.05
<2
c
c
c
E006
E007
c
E008
E009
1.1
1.4
1.2
1.0
c
E010
E011
c
c
E012
1.1
c
1.3
c
E013<0.05
E014
E015
<0.05
<0.05
<0.05
<0.05
24
<0.05
<0.05
<0.05
32
<0.05
<0.05
<0.05
1.0
<0.05
<0.05
<0.05
1.1
<0.05
<0.05
<0.05
1.0
<0.05
<0.05
<0.05
1.0
c
c
c
c
<0.05
<0.05
27
<0.05
<0.05
28
E016
1.0
c
1.3
c
1.0
c
1.0
c
E018b
E019
E020
<1.5
<0.1
<0.2
<1.5
<0.1
<0.2
<1.5
<0.1
<0.2
<1.5
<0.1
<0.2
<1.5
<0.1
<0.2
<1.5
<0.1
<0.2
<1.5
<0.1
<0.2
<1.5
<0.1
<0.2
c
c
c
c
c
c
c
c
a Specific activity in lmol ester hydrolyzed/mg protein/min · 100.
b Estimated enantioselectivity is the ratio of the initial rates of hydrolysis of the two enantiomers (rate for fast enantiomer/rate for slow enantiomer).
c Could not be measured because the substrate did not react.
untouched.20 However, AcE was not sufficiently acetyl
selective in the synthesis of O-sialyl-Lewis-X peptides
because AcE removed both the acetyl groups and
the C-terminal ester.21 Esterase E018b is often more
selective than AcE for acetyl groups and may solve these
synthetic problems. Additionally, more than 10 thermo-
phile hydrolases favor hexanoate or octanoate esters
over their shorter acyl chain analogues. This selectivity
provides a method for selectively hydrolysis of one type
of ester in the presence of the other.
Our enantioselectivity results are similar, but not identi-
cal, to those reported by ThermoGen.6 For example, in
the hydrolysis of 1-phenethyl acetate, Dimirjian et al. re-
ported enantioselectivities >50 for E002, E003, E005,
E009, and E020, but moderate enantioselectivities
(E = 8–10) for E004, E008, and E013. In agreement with
the ThermoGen group, we measured enantioselectivities
>50 for all five esterases, but found slightly higher results
for E008, and E013( E = 25–30). There are two likely rea-
sons for these differences. We tested the butyrate ester,

3000
N. A. Somers, R. J. Kazlauskas / Tetrahedron: Asymmetry 15 (2004) 2991–3004
Table 9. Enantioselectivity of thermophile esterases toward selected substratesa
34
36
38
43
45
E001
E002
E003ND
E004
E005
E006
E007
E008
E009
E010
E011
E012
E013ND
E014
E015
E016
E018b
E019
E020
3.4 (S)
3.6 (S)
>3.2 (1R)
>8.9 (1R)
>22 (R)
>55 (R)
R)
>4.7 (R)
ND
ND
9.7 (S)
7.0 (S)
6.4 (S)
7.2 (S)
4.9 (S)
10 (S)
4.2 (S)
21 (S)
6.9 (S)
>6.2 (S)
5.1 (S)
NR
ND
>19 (
ND
>81 (1R)
>2.0 (1R)
ND
>3.4 (R)
>31(R)
>16 (R)
>2.1 (R)
>33 (R)
110 (R)
>33 (R)
>66 (R)
2.8 (R)
R)
NR
ND
3.9 (S)
3.7 (S)
2.4 (S)
ND
>2.0 (R)
NR
ND
ND
ND
ND
ND
ND
ND
ND
NR
ND
ND
ND
ND
NR
NR
ND
8 (
>3
NR
NR
2.8 (S)
2.7 (S)
10 (S)
5.4 (S)
NR
ND
ND
ND
ND
ND
ND
ND
ND
>28 (R)
>150 (R)
12 (R)
ND
>5.5 (R)
NR
3.3 (S)
NR
ND
ND
>24 (R)
>50 (R)
>2.2 (R)
>2.3( R)
13 S)
11 (S)
(
3.1 (S)
a NR = not determined because the substrate did not react. ND = not determined because the estimated enantioselectivity was <2. Unless otherwise
noted, the true enantioselectivities were measured using resorufin esters as a reference compound and have an estimated error of 10–30%. When
the rate of the slow enantiomer was too slow to measure accurately, a lower limit for the enantioselectivity is given.
Table 10. Selectivity measurement using scale up reactions to confirm Quick E/S measurementsa
Substrate (s)
Enzyme
Estimated E/S
Quick E/S
Ees
Eep
End-point
Lit.b
36
36
E001
E004
2.2 (R)
2.0 (R)
2.4 (R)
5.1 (R)
R)
>3.2 (R)
>81 (R)
>2.0 (R)
>55 (R)
>19 (R)
>3.9 (R)
>3 R)
>2.1 (R)
>3 R)
110 (R)
>3 R)
>50 (R)
24 (1)
0.029
0.13>0.86
0.80
>0.96
>46 (R)
R)
NR
NR
>86 (
36
38
E005
E002
>0.99
0.68
>500 (R)
46 (R)
R)
NR
0.91
0.730.98
0.71
103( R)
98 (R)
9 (R)
101 (R)
20 (R)
7.7 (R)
88 (R)
10 (R)
57 (R)
NR
38
38
E0036.6 (
E004
E005
200 (
5.7 (R)
13( R)
5.2 (R)
4.6 (R)
6.7 (R)
R)
0.76
16 (R)
R)
38
38
01.2(3>0.95
0.10
30.29
>52 (
E007
E008
0.19
>0.9(2
>0.97
1.6 (R)
32 (R)
>80 (R)
R)
38
38
E009
E01312 (
E020
0.092
08.0(230.88
0.36
38
38
25 (
15 (R)
136 (1)
ND
>0.85
0.92c
0.35c
0.56c
>170 (R)
17 (1)
4.0 (1)
18 (3)
1, 3
1, 3
1, 3
E018b
AcE
E015
0.37c
5.9 (1)
28 (3)
0.69c
NR
NR
3.1 (3)
0.82c
a ND = not determined. Ees = enantiomeric excess of the substrate at the end of the reaction. Eep = enantiomeric excess of the product at the end of
the reaction.
b Enantioselectivity measured using 1-phenethyl acetate instead of 1-phenethyl butyrate, 38, from Ref. 6. NR = not reported. Based on NMR data.
c Excess in starting materials and products respectively at the end of the reaction.
while the ThermoGen group tested the acetate ester.
Hydrolysis of butyrates often shows higher selectivity
than hydrolysis of acetates, consistent with our generally
higher enantioselectivies.22 Second, all our reactions in-
cluded 7% acetonitrile while Dimirjian et al. did not
and organic co-solvents can change enantioselectivity.23
In many methods, the solution to measuring slow sub-
strates is to extend the reaction time or to add more en-
zyme. This solution does not work for Quick E methods
because the key measurement is the relative rate of
hydrolysis of the substrate and the reference compound.
An accurate relative rate for a slow substrate requires a
slow reference compound. The slowest-reacting resoru-
fin compounds we used were the isobutyrate and pivalo-
ate esters. For example, we could only place a lower
limit of 2 on the enantioselectivity of E005 toward men-
thyl acetate, 36. However, a scale-up reaction revealed
that the enantioselectivity was very high, E >100. Our
current research has discovered other slow reacting ref-
erence compounds that promise to minimize this limita-
tion in the future.
Using qualitative screening the ThermoGen group also
identified esterases E002, E003, E005, E006, E007, and
E012 as highly selective for the (R)-enantiomer of dioxol-
ane 43. We found very slow reaction for this substrate
and thus could not confirm these results. Esterases
E007 and E012 did not react, and it was only possible
to estimate a lower limit for the enantioselectivity for
five esterases.

N. A. Somers, R. J. Kazlauskas / Tetrahedron: Asymmetry 15 (2004) 2991–3004
3001
The thermophile esterases resemble lipases. Like lipases,
they favor hydrophobic substrates over polar substrates.
They also tolerate a wide range of alcohol moieties in
the ester, but favor straight chain acyl groups similar
to lipases. The high selectivity of all the Thermophilic
esterases, but E018b, for hexanoyl or octanoyl esters
suggests that future screening with these enzymes should
use hexanoyl or octanoyl esters, not acetyl or other es-
ters. However, unlike lipases, there is no evidence that
thermophile esterases show interfacial activation.
2H, CH2), 2.6–2.7 (t, 2H, octanoyl CH2), 7.1–7.4
(m, 5H, aromatic). MS m/z: 220 (MÅ+, 25); 127 (92);
101 (9); 94 (89); 60 (37); 57 (100); 29 (7.5).
4.5. Monohexanoate of 2,2-bis(4-hydroxyphenyl)propane
Monohexanoate of 2,2-bis(4-hydroxyphenyl)propane
was prepared as above for phenyl butyrate, yellowish
oil: 3.7g (77%). Rf = 0.18 (9:1 hexane/ethyl acetate);
1H NMR d 0.9 (t, 3H, hexanoyl CH3) 1.3(apparent s,
6H, CH2), 1.6 (s, 6H, propane CH3), 2.0 (t, 2H, CH2),
6.7–7.2 (m, 8H, aromatic). MS m/z: 327 (MÅ+, 31); 326
(32); 213 (100); 135 (17); 119 (20); 99 (6.8); 91 (6.7), 71
(9.2), 43(18).
4. Experimental
4.1. General
4.6. Monoacetate of 2,2-bis(4-hydroxyphenyl)propane
Unless otherwise noted, esters, reagents, and solvents
were purchased from commercial sources. Enzyme solu-
tions were prepared by dissolving the powder in BES
(N,N-bis[2-hydroxyethyl]-2-aminoethanesulfonic acid,
5.0mM, pH7.20) buffer at a concentration of 0.5–
2.0mg solid/mL. The solid was 2–60% protein as deter-
mined by the BioRad dye binding assay. Solutions were
stored at ꢀ20ꢁC. Solketal butyrate and resorufin acetate
were prepared as described previously.4 Solketal octano-
ate, 1-phenethyl butyrate, and other resorufin esters
Monoacetate of 2,2-bis(4-hydroxyphenyl)propane was
prepared as N,N-dimethylaminopyridine and acetic
anhydride and worked up as above for phenyl butyrate,
colorless oil: 0.58g (24%). Rf = 0.26 (9:1 hexane/ethyl
1
acetate); H NMR d 1.6 (s, 6H, propane CH3), 2.3(s,
3H, acetyl CH3), 6.7–7.2 (m, 8H, aromatic). MS m/z:
271 (MÅ+, 4.3); 255 (6.9); 228 (925); 213 (100); 135 (15);
119 (16); 91 (6.0); 74 (67); 59 (47); 45 (26); 31 (31).
1
were prepared similarly. H NMR spectra were run in
4.7. Monohexanoate monoacetate of 2,2-bis(4-hydroxy-
phenyl)propane
deuteriochloroform at 200MHz. Low resolution MS
spectra were obtained by direct inlet using electron
ionization.
Monohexanoate monoacetate of 2,2-bis(4-hydroxyphen-
yl)propane was prepared from the monohexanoate as
above for the monoacetate, colorless oil, 1.3g (95%).
1H NMR d 0.9 (t, 3H, hexanoyl CH3) 1.4 (apparent s,
4H, hexanoyl CH2), 1.7 (s, 6H, propane CH3), 1.8 (t,
2H, hexanoyl CH2), 2.3(s, H3 , acetyl CH 3), 2.6 (t,
2H, hexanoyl CH2), 7.0–7.2 (m, 8H, aromatic). MS
m/z: 368 (MÅ+, 26); 270 (100); 255 (38); 228 (100);
213(100); 135 (19); 119 (21); 99 (10); 91 (7.6); 71 (18);
55 (9.2); 43(40).
4.2. Phenyl butyrate, 26
Butanoyl chloride (4.8mL, 42mmol) was added drop
wise to a flame-dried, round-bottomed flask containing
a solution of phenol (2.0g, 21mmol), and pyridine
(2.5mL, 32mmol) in dry ether (30mL) under nitrogen
and cooled in an ice bath. After stirring the resulting
suspension for 2h, it was washed with water
(2 · 30mL), saturated sodium bicarbonate (2 · 40mL),
HCl (2 · 30mL, 0.2N), and saturated NaCl
(2 · 30mL). The organic phase was dried over magne-
sium sulfate, concentrated by rotary evaporation, and
purified by column chromatography on silica gel (4:1:2
hexanes–ethyl acetate–chloroform, Rf = 0.67) yielding
4.8. Initial screening (estimated selectivity)
To each well of a 96-well polystyrene microplate was
added substrate solution (7.0lL of a 14.3mM solution
in acetonitrile), 4-nitrophenol solution (23lL of a
1.9mM in BES buffer (1.0mM, pH7.20) and BES buffer
(65lL, 1.0mM, pH7.20). This mixture yielded final con-
centrations of 1.0mM substrate, 0.437mM 4-nitrophe-
nol, 1.13mM BES, and 7.0vol% acetonitrile. Esterase
solution (5lL in 5mM BES, typically 1mg solid/mL)
was added to each well. This mixture yielded final con-
centrations of 1.0mM substrate, 0.437mM 4-nitrophe-
nol, 1.13mM BES, and 7.0vol% acetonitrile. The
microplate was placed in the microplate reader, shaken
for 5s and the decrease in absorbance at 404nm was
measured as often as permitted by the microplate reader,
typically every 10s. Data was collected for 20min, at
25ꢁC, in triplicate and was averaged. For slow reactions,
a more concentrated esterase solution was used so that
the absorbance change was >2mA/min, while for fast
reactions a less concentrated esterase solution was used
so that the absorbance change was <200mA/min. The
reaction rates were calculated using Eq. 1 below.
1
a colorless oil: 1.7g (49%). H NMR d 1.0–1.1 (t, 3H,
CH3) 1.8–1.9 (m, 2H, CH2), 2.5–2.6 (t, 2H, CH2), 7.1–
7.4 (m, 5H, aromatic). MS m/z: 164 (MÅ+, 27); 94
(100); 77 (5.4); 71 (46); 43(27).
4.3. Phenyl hexanoate
Phenyl hexanoate was prepared as above for the buty-
1
rate, colorless oil: 3.2g (79%). Rf = 0.71; H NMR d
0.9–1.0 (t, 3H, CH3) 1.4–1.8 (m, 6H, CH2), 2.5–2.6 (t,
2H, CH2), 7.1–7.4 (m, 5H, aromatic). MS m/z: 192
(MÅ+, 28); 99 (77); 94 (100); 71 (38); 43 (47).
4.4. Phenyl octanoate
Phenyl octanoate was prepared as above for the buty-
1
rate, colorless oil: 4.7g (98%). Rf = 0.68; H NMR d
1.0 (s, 3H, CH3) 1.4–1.9 (m, 8H, CH2), 2.4–2.5 (m,

3002
N. A. Somers, R. J. Kazlauskas / Tetrahedron: Asymmetry 15 (2004) 2991–3004
ꢀ
ꢁ
dA404=dt
De404 Â l
½bufferŠ
For the reaction conditions above, these equations
simplify to Eqs. 2a and 3a.
rateðlmol= minÞ ¼
Â
þ 1
½indicatorŠ
 V  106
ð1Þ
dA574
rateref ðlmol= minÞ ¼
 0:02269
ð2aÞ
dt
In this equation, dA404/dt is the absorbance decrease at
404nm per minute, De404 is the difference in extinction
co-efficients for the protonated and unprotonated forms
of the indicator (17,300Mꢀ1 cmꢀ1), l is the path length
(0.2919cm for a 100lL reaction volume; in microplates,
the path length varies with reaction volume), and V is
the reaction volume in liters. For the reaction conditions
above, this equation simplifies to Eq. 1a.
ꢂ
ꢃ
dA404
dt
ratesubðlmol= minÞ ¼
 0:07411
ꢀ 1:1½rateref Š
ð3aÞ
The enantioselectivity was calculated from two measure-
ments (one for each enantiomers) and adjusted for the
concentration of substrate and reference compound in
each measurement as shown below in Eq. 4.
dA404
dt
rateðlmol= minÞ ¼
 0:07101
ð1aÞ
rateR
½refRŠ rateref S
½SŠ
Quick E ¼
Â
Â
Â
ð4Þ
The observed rates were divided by the protein amount
in the well to give the values in the Tables. Blanks con-
tained either enzyme, but no substrate or substrate, but
no enzyme.
rateref R
½RŠ
rateS
½refSŠ
In this equation, rateR represent the rate of hydrolysis of
the (R)-enantiomer, rateref R is the rate of hydrolysis of
the reference compound in the presence of the (R)-
enantiomer, [refR] is the concentration of the reference
compound during the measurement for the (R)-enantio-
mer, and [R] is the concentration of the (R)-enantiomer.
4.9. True selectivity and enantioselectivity
A buffer/indicator solution was prepared by mixing
4-nitrophenol solution (1.2mL of a 1.8mM solution
in 1.0mM BES containing 0.33mM Triton X-100,
pH7.2), BES buffer (3.3mL of a 1.0mM solution con-
taining 0.33mM Triton X-100, pH7.20), and acetonit-
rile (65lL). Substrate solution (e.g., 35lL of a
150mM vinyl pivaloate in acetonitrile) and resorufin es-
ter solution (e.g., 260lL of 2.0mM resorufin pivaloate
in acetonitrile) were added dropwise with continuous
vortexing to form a clear emulsion that was stable for
at least several hours. Final concentrations in the well
were 1.03mM substrate (vinyl pivaloate), 0.102mM
resorufin pivaloate, 0.423mM pNP, 1.16mM BES,
0.29mM Triton X-100, and 7.05vol% acetonitrile. This
solution was pipetted into a 96-well polystyrene micro-
plate (100lL/well). Esterase solution (5lL in 5mM
BES, typically 1mg solid/mL) was added to each well
and the microplate was placed in the microplate reader,
shaken for 5s and the decrease in absorbance at 404nm
and the increase in absorbance at 574nm were measured
as often as permitted by the microplate reader, typically
every 11s. Data was collected for 20min, at 25ꢁC, in
triplicate and was averaged. The rate of hydrolysis of
the reference compound and the substrate were calcu-
lated using Eqs. 3and 4 below using the initial, linear
parts of the curve.
4.10. Use of different reference compounds
In some cases we used a faster reacting reference com-
pound (e.g., resorufin acetate) with the faster reacting
enantiomer and a slower reacting reference compound
(e.g., resorufin pivaloate) with the slower reacting
enantiomer. Such measurements were corrected for the
selectivity of the enzyme for the two reference com-
pounds, Sres 1/res 2. This selectivity was measured using
a substrate of intermediate reaction rate. In this case
the substrate serves as the reference compound. The
selectivity was calculated using Eq. 5 below where rate-
is the observed rate of hydrolysis of resorufin ester
res 1
1, ratesub res 1 is the observed rate of hydrolysis of the
substrate when measured with resorufin ester 1 and the
brackets indicate concentrations.
rateres 1
½subŠ
rateres 2
½subŠ
S
res
res
1
2
¼
Â
Â
Â
ð5Þ
ratesub res 1 ½res 1Š ratesub res 2 ½res 2Š
4.11. Acyl chain length selectivity using the endpoint
method
The selectivity of E018b for vinyl acetate versus vinyl
butyrate was also measured by H NMR. Vinyl acetate
1
dA574=dt
De574 Â l
rateref ðlmol= minÞ ¼
 V  106
ð2Þ
(0.92lL, 10lmol), vinyl butyrate (1.3lL, 10lmol) were
dissolved D2O (1.0mL containing BES (1.0mM,
pH7.20) and 7vol% CD3CN). Enzyme solution (10lL)
was added and the resonances corresponding to vinyl
acetate (CH3, d 2.0), vinyl butyrate (CH2, d 2.3), acetic
acid (CH3, d 1.9), and butyric acid (CH2, d 2.2) were
ꢂ
ꢀ
ꢁ
dA404=dt
De404 Â l
½bufferŠ
ratesubðlmol= minÞ ¼
Â
þ 1
½indicatorŠ
ꢃ
1
monitored by H NMR (500MHz). Data in Table 10.
ÂV  106 ꢀ 1:1½rateref Š
ð3Þ
4.12. Selective hydrolysis of acetyl ester in monohexano-
ate monoacetate of 2,2-bis(4-hydroxyphenyl)propane
The symbols are the same as those defined for Eqs. 1
and 2 above, while De574 is the difference in extinction
co-efficients at 574nm for resorufin and the resorufin
ester (15,100Mꢀ1 cmꢀ1).
Esterase E018b (1mL) was added to a solution of mono-
hexanoate monoacetate of 2,2-bis(4-hydroxyphenyl)pro-

N. A. Somers, R. J. Kazlauskas / Tetrahedron: Asymmetry 15 (2004) 2991–3004
3003
Weber, M. J.; Johnson, S. P.; Vonstein, V.; Casadaban,
M. J.; Demirjian, D. C. Bio/Technology 1995, 13, 271–275.
4. Janes, L. E.; Kazlauskas, R. J. J. Org. Chem. 1997, 62,
4560–4561; Janes, L. E.; Lo¨wendahl, A. C.; Kazlauskas,
R. J. Chem. Eur. J. 1998, 4, 2324–2331; Janes, L. E.;
Cimpoia, A.; Kazlauskas, R. J. J. Org. Chem. 1999, 64,
9019–9029.
5. ThermoGen Inc. is now part of deCODE genetics,
Sturlugata 8. IS-101 Reykjavik, Iceland. deCODE cur-
rently does not sell the ThermoGen esterases, but may do
so upon request.
pane (0.24g, 0.67mmol) in BES (8.3mL of a 1.0mM
solution) and acetonitrile (700lL). The pH of the solu-
tion was maintained at 7.20 with a pHstat, which con-
trolled the addition of NaOH (0.1N). When the
consumption of base indicated ꢁ40% conversion, the
reaction mixture was extracted with ether (2 · 10mL).
The organic phase was concentrated by rotary evapora-
tion, the residue was dissolved in acetonitrile (1mL),
and analyzed by high performance liquid chromatog-
raphy on a reversed phase column (Zorbax C8,
4.6mm · 25mm) eluted with 0.60mL/min using 50/50
water/methanol at 25ꢁC. The detector was set at
254nm. kAc = 9.6 for the monoacetate of 2,2-bis(4-
hydroxyphenyl)propane; kHx = 15.3for the monohex-
anoate of 2,2-bis(4-hydroxyphenyl)propane. The molar
extinction co-efficients for the two monoesters were
assumed to be equal.
´
6. Demirjian, D. C.; Shah, P. C.; Morıs-Varas, F. Top. Curr.
Chem. 1999, 200, 1–29, Enantiomeric ratios were calcu-
lated from data in this paper.
7. The selectivity of an enzyme is determined by both the
relative kcat (rate of reaction) and relative KM (approxi-
mately, the binding of the substrate). For example,
imagine two substrates that react at the same rate, but
one binds better. In a competitive experiment, the better
binding one will selectively react, but if the rates are
measured separately, both will react at similar rates
(assuming the concentration of the substrate is above
KM). For this reason, the ratio of separately measured
initial rates does not measure the true selectivity of the
enzyme.
8. An alternate explanation for the long chain preference is
the formation of micelles and interfacial activation, as
observed for lipases. This hypothesis is unlikely for two
reasons. First, not all of the longer chain esters react
faster; for example vinyl decanoate reacts slowly. Second,
all reactions contain the surfactant Triton X-100, which
causes micelle formation regardless of the substrate added.
9. Wahler, D.; Badalassi, F.; Crotti, P.; Reymond, J.-L.
Chem. Eur. J. 2002, 8, 3211–3228.
4.13. True enantioselectivity using the endpoint method
Hydrolase solution (1mL) was added to a solution of
racemic ester (e.g., 110lL, 51lmol of menthyl acetate)
in BES (8.1mL of a 1.0mM solution, pH7.20) contain-
ing 7vol% acetonitrile (700lL). The pH of the solution
was maintained at 7.20 with a pHstat, which controlled
the addition of NaOH (0.1N). When the consumption
of base indicated ꢁ40% conversion, the reaction mixture
was extracted with ether (2 · 10mL). The organic phase
was concentrated by rotary evaporation, the residue was
dissolved in ethyl acetate (1mL), and analyzed by gas
chromatography on a Chirasil-DEX CB capillary col-
umn (Chrompack, Raritan, NJ) at 120ꢁC: kR = 7.15,
10. (a) Waldmann, H.; Heuser, A. Bioorg. Med. Chem. 1994,
2, 477–482; (b) Waldmann, H.; Heuser, A.; Reidel, A.
Synlett 1994, 65–67.
11. Chen, C. S.; Fujimoto, Y.; Girdaukas, G.; Sih, C. J. J. Am.
Chem. Soc. 1982, 104, 7294–7299.
kS = 6.95, and a = 1.03for menthyl acetate,
42,
kR = 12.1, kS = 11.2, and a = 1.08 for 1-phenethyl buty-
rate, 41. Data are in Table 9.
12. (a) Berger, M.; Schneider, M. P. Biotechnol. Lett. 1991, 13,
641–645; (b) Sugiura, M.; Isobe, M. Chem. Pharm. Bull.
1975, 23, 1226–1230.
Acknowledgements
13. For example: Reetz, M. T.; Kuhling, K. M.; Wilensek, S.;
¨
Husmann, H.; Ha¨usig, U. W.; Hermes, M. Catal. Today
We thank Andrew Man Fai Liu for setting up the ester
library, Kimberly Bull for the initial screening, Siao Li
Chuah for additional screening, David Dietrich and
Harjap Grewal for confirming the acetyl selectivity of
E018b, Krista L. Morley for rechecking some of the
selectivity data, David Demirjian (Zuchem, formerly
from ThermoGen, Inc., Chicago, IL) for the generous
gift of esterases and the Natural Sciences and Engineer-
ing Research Council of Canada for financial support.
2001, 67, 389–396; Reetz, M. T.; Kuhling, K. M.; Deege,
¨
A.; Hinrichs, H.; Belder, D. Angew. Chem., Int. Ed. 2000,
39, 3891–3895; Reetz, M. T.; Becker, M. H.; Klein, H.-W.;
Sto¨ckigt, D. Angew. Chem., Int. Ed. 1999, 38, 1758–1761.
14. Schelhaas, M.; Waldmann, H. Angew. Chem., Int. Ed.
1996, 35, 2056–2083.
15. Waldmann, H.; Braun, P.; Kunz, H. Biomed. Biochim.
Acta 1991, 50, 243–248.
16. Holla, E. W. J. Carbohydr. Chem. 1990, 9, 113–119;
Tamura, M.; Kinomura, K.; Tada, M.; Nakatsuka, T.;
Okai, H.; Fukui, S. Agric. Biol. Chem. 1985, 49,
2011–2023.
References
17. Removal of protecting groups from nucleoside diphos-
phate and triphosphate sugars under mild conditions:
Oehrlein, R.; Baisch, G. WO 9624683A1 960815; Chem.
Abstr. 1996, 125, 219752.
18. Waldmann, H.; Heuser, A. Bioorg. Med. Chem. 1994, 2,
477–482; Waldmann, H.; Heuser, A.; Reidel, A. Synlett
1994, 65–67.
19. See also selective removal of acetyl group from cephalo-
sporins: Hall, D. A.; Heiney, R. E. EP 109300 A2 840523;
Chem. Abstr. 1994, 101, 130512.
20. Na¨gele, E.; Schelhaas, M.; Kuder, N.; Waldmann, H.
J. Am. Chem. Soc. 1998, 120, 6889–6902.
1. Reviews: Hough, D. W.; Danson, M. J. Curr. Opin. Chem.
Biol. 1999, 3, 39–46; Sellek, G. A.; Chaudhuri, J. B.
Enzym. Microb. Technol. 1999, 25, 471–482; Bruins, M. E.;
Janssen, A. E. M.; Boom, R. M. Appl. Biochem. Biotech-
nol. 2001, 90, 55–186.
2. Taylor, I. N.; Brown, R. C.; Bycroft, M.; King, G.;
Littlechild, J. A.; Lloyd, M. C.; Praquin, C.; Toogood, H.
S.; Taylor, S. J. C. Biochem. Soc. Trans. 2004, 32, 290–292.
3. For example, Vonstein, V.; Johnson, S. P.; Yu, H.;
Casadaban, M. J.; Pagratis, N. C.; Weber, J. M.;
Demirjian, D. C. J. Bacteriol. 1995, 177, 4540–4543;

3004
N. A. Somers, R. J. Kazlauskas / Tetrahedron: Asymmetry 15 (2004) 2991–3004
21. Seitz, O.; Wong, C.-H. J. Am. Chem. Soc. 1997, 119,
8766–8776.
N.; Geymayer, P.; Stadler, P.; Faber, K.; Griengl, H.
Tetrahedron: Asymmetry 1990, 1, 111–118.
22. Cotteril, I. C.; Macfarlane, E. L. A.; Roberts, S. M. J.
Chem. Soc., Perkin Trans. 1 1988, 3387–3389; Klempier,
23. For example, Terradas, F.; Teston-Henry, M.; Fitzpatrick, P.
A.; Klibanov, A. M. J. Am. Chem. Soc. 1993, 115, 390–396.