Acylase I in the alcoholysis of α-substituted dicarboxylic acid esters and derivatives: Enantio- and regioselectivity

Article abstract of DOI:10.1016/S0957-4166(00)00346-3

Enantio- and regioselective butanolyses of α-substituted dimethyl succinates (substituents: Me-, MeO₂CCH₂-, NH₂-, AcHN-, PrCOHN-, HO-, MeO-, PrO-, AcO-, PrCO₂-, HepCO₂-, Cl- and Br-) and glutarates (substituents: PrCONH- and CbzNH-) and that of methyl pyroglutamate with acylase I enzymes have been studied. Acylase I-catalyzed reactions were totally regioselective proceeding exclusively at the sterically more hindered methyl ester group α to the substituent. High enantioselectivities (E from 50 to >>100) were observed only for the substrates containing CONH functionality in the substituent although the C-N bond was unreactive. The nature of the substituent influenced which of the two enantiomers reacted faster. Copyright (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0957-4166(00)00346-3

TETRAHEDRON:
ASYMMETRY
Pergamon
Tetrahedron: Asymmetry 11 (2000) 3957–3966
Acylase I in the alcoholysis of a-substituted dicarboxylic
acid esters and derivatives: enantio- and regioselectivity
Arto Liljeblad, Jutta Lindborg and Liisa T. Kanerva*
Departments of Chemistry and Biomedicine, University of Turku, Lemminka¨isenkatu 2, FIN-20520 Turku, Finland
Received 27 July 2000; accepted 30 August 2000
Abstract
Enantio- and regioselective butanolyses of a-substituted dimethyl succinates (substituents: Meꢀ,
MeO2CCH2ꢀ, NH2ꢀ, AcHNꢀ, PrCOHNꢀ, HOꢀ, MeOꢀ, PrOꢀ, AcOꢀ, PrCO2ꢀ, HepCO2ꢀ, Clꢀ and Brꢀ)
and glutarates (substituents: PrCONHꢀ and CbzNHꢀ) and that of methyl pyroglutamate with acylase I
enzymes have been studied. Acylase I-catalyzed reactions were totally regioselective proceeding exclusively
at the sterically more hindered methyl ester group a to the substituent. High enantioselectivities (E from
50 to ꢀ100) were observed only for the substrates containing CONH functionality in the substituent
although the CꢀN bond was unreactive. The nature of the substituent influenced which of the two
enantiomers reacted faster. © 2000 Elsevier Science Ltd. All rights reserved.
1. Introduction
Acylases are readily available—although synthetically seldom studied—hydrolytic enzymes.
Especially interesting are the acylase I enzymes (N-acylamino acid amidohydrolases; EC
3.5.1.14) as they hydrolyze neutral aliphatic N-acyl a-amino acids, including many xenobiotic-
derived mercapturates (S-substituted N-acetyl- -cysteine conjugates) and nonproteinogenic
L
amino acids, to give the corresponding -amino acids (usually the (S) absolute configuration)
L
and fatty acids (Scheme 1).1–4 Accordingly, acylase I from Aspergillus melleus is reported to
catalyze the enantioselective hydrolysis of numerous natural as well as non-natural N-acyl
Scheme 1.
* Corresponding author. E-mail: lkanerva@utu.fi
0957-4166/00/$ - see front matter © 2000 Elsevier Science Ltd. All rights reserved.
PII: S0957-4166(00)00346-3

3958
A. Liljeblad et al. / Tetrahedron: Asymmetry 11 (2000) 3957–3966
amino acids, the terminal carboxylate group being important for productive binding and
turnover of the substrates.3 The highly enantioselective hydrolysis of methyl cis-3-
chloromethyl-2-tetrahydrofuran carboxylate reveals that the enzyme is capable of cleaving not
only CꢀN bonds but also ester bonds.5 In accordance with the CꢀO bond cleaving capability,
the acylase I enzymes from commercial Aspergillus species previously allowed the (S)-enan-
tioselective reactions of N-butanoylated dimethyl aspartate as novel alcoholysis reactions
(Scheme 2).6 These reactions are noteworthy as they proceed exclusively at the sterically more
hindered a-methyl ester function, leaving the more natural amide bond as well as the b-
methyl ester function unreacted. Also noteworthy is that the enzymes accept dimethyl aspar-
tate, a substrate for acylase II (aspartoacylase, EC 3.5.1.15):1,2 N-acyl derivatives of aspartic
acid were previously reported as non-substrates for hydrolysis with acylase I (Scheme 1).3
Acylase I from Aspergillus melleus is also useful for enantioselective acylations of primary
and secondary alcohols by vinyl esters.7–11
Scheme 2.
The mechanism of acylase I-catalyzed acyl transfers is not known. Some mechanistic
features can be proposed on the basis of the mechanisms of Zn2+-containing carboxypeptid-
ases.12 The enzymes such as zinc metalloproteins utilize Zn2+ as a cofactor and a Lewis acid.
Thus, for the hydrolysis of N-acyl a-amino acids (Scheme 1) the metal cation is expected to
coordinate to nucleophilic water, making it more acidic, and/or to the carbonyl group of the
scissile CꢀN bond, increasing its susceptibility toward nucleophilic attack. For the hydrolysis
of alkyl esters by acylase I, the ester carbonyl can coordinate in the place of the amide
carbonyl. For alcoholysis, the alcohol as a nucleophile takes the place of water.
The existing data clearly show that the acylase I enzymes are capable of acting on the
CꢀO bonds of carboxylic acid esters and thus function chemo- and regioselectively recogniz-
ing chirality at the acid part of their substrates.5–11 This property made us interested in
widening the scope of dicarboxylic acid substrates for acylase I-catalyzed alcoholyses in neat
alcohols. Thus, the regio- and enantioselective butanolyses of a-substituted dimethyl succi-
nates and glutarates 1a–15a have now been studied paying attention to the nature of the
a-substituent Y (Scheme 2; Table 1). Methyl pyroglutamate 16a has served as a cyclic
bifunctional substrate (Scheme 3).

A. Liljeblad et al. / Tetrahedron: Asymmetry 11 (2000) 3957–3966
3959
Table 1
Acylase I-catalyzed butanolysis of a-substituted dimethyl succinates and glutarates (t=22 h) at room
temperature
Reactive
Acylase I from
Acylase I from
Acylase I on Eupergit
enantiomer
Aspergillus genus
Aspergillus melleus
C
Compound
Y
c (%)
E
c (%)
E
c (%)
E
1a (n=1) Me
2a (n=1) MeO2CCH2
3a (n=1) H2N
R
–
1
1
35
16
13
–
–
8
4
1
4
11
2
23
13
2
–
5
4
2
2
47
20
2
–
13
9
5
10
85
46
51
44
49
37
29
7
24
24
7
45
38
2
–
7
14
a
S
S
S
S
S
S
R
R
S
S
S
R
R
4a (n=1) AcNH
5a (n=1)b PrCOHN
6a (n=2) PrCONH
7a (n=2) CbzNH
8a (n=1) HO
9a (n=1) MeO
10a (n=1) PrO
11a (n=1) AcO
12a (n=1) PrCO2
13a (n=1) HepCO2
14a (n=1) Cl
15 23
\100
50
–
–
3
5
4
3
–
2
2
3
–
–
15
8
1
7
13
2
27
22
–
–
7
7
4
3
–
2
2
3
\100
ꢀ100
3
17
9
2
7
3
3
4
c
c
15a (n=1) Br
a Both the substrate and the reaction product are prochiral.
b Ref. 6.
c Mixture of products observed.
Scheme 3.
2. Results and discussion
The results for the acylase I-catalyzed butanolyses of a-substituted dimethyl succinates 1a–5a,
8a–15a and glutarates 6a and 7a by fungal acylase I enzymes are given in Table 1. Reactivities
(conversion at a certain time) of the three enzyme preparations clearly differ from each other,
highest reactivity always being observed when the enzyme on Eupergit C is used. There is hardly
any reaction with Y=Me or CH2CO2Me (compounds 1a and 2a). In contrast, the substrates
3a–9a, 11a, 12a, 14a and 15a show from reasonable to high reactivities. In these cases Y is
attached to the succinate or glutarate backbone through an electronegative oxygen, nitrogen or
halogen atom which all contain free electron pairs. If electronic properties alone are important
for reactivity it is difficult to explain why compounds 10a and 13a react slowly.
Acylase I-catalyzed reactions of compounds 1a–15a with butanol all proceed exclusively at the
sterically more hindered a-ester group of a molecule even though the presence of the b- or
g-methyl ester and amide functions (Scheme 2). There is no sign of the formation of the dibutyl
esters. The strict chemo- and regioselectivity was confirmed by preparing mixed butyl methyl
esters from the corresponding commercial half esters and confirming the identities of the GLC

3960
A. Liljeblad et al. / Tetrahedron: Asymmetry 11 (2000) 3957–3966
chromatograms with those obtained from enzymatic reactions. In the case of compound 12a,
however, the reaction proceeds at the a-butanoyl group in addition to the a-methyl ester group,
leading to the mixture of products. This is taken to be indicative of at least two different ways
of substrate binding at the active site of the enzyme. The corresponding acyl transfers from the
O-acylated dimethyl malates 11a and 13a are not observed. This is in accordance with the previous
observations that butanoate is favored over acetate, hexanoate or higher carboxylates for the
acylase I-catalyzed reactions of vinyl esters with alcohols.9,11
The enantioselectivities of the acylase I enzymes for substrates 1a–15a are generally from
negligible to moderate except when N-substituted amino acid esters 5a–7a serve as substrates
(Table 1). Clearly, acylase I on Eupergit C best suits the enantioselective butanolyses in neat
butanol. Thus, the resolutions of N-butanoylated esters 5a and 6a result in excellent enantiose-
lectivities, the reaction in the case of 7a being practically completely enantiospecific (both the
unreacted (R)-enantiomer and the obtained a-butyl b-methyl ester with the (S) absolute
configuration are obtained with >99% ee at 50% conversion). Moderate enantioselectivity is
observed for the N-acetylated compound 4a.
It is clear according to the results in Table 1 that good reactivity does not necessarily correlate
with good enantioselectivity (E values) for the acylase I-catalyzed alcoholyses of compounds 3a–9a,
11a, 12a, 14a and 15a. Noteworthy amongst the results is the dependence of the observed
enantiopreference (which enantiomer reacts faster) on the structure of Y. Thus, the (R)-enantiomer
reacts faster in the case of halogen, alkoxy and methyl substituted compounds 1a, 9a, 10a, 14a
and 15a, while the (S)-enantiopreference is observed in all the other cases. Two different binding
possibilities at the active site of the enzyme are proposed, leading from excellent (S)-selectivity
(Eꢀ100) in the case of N-substituted glutarate 7a to acceptable (R)-selectivity (E=17) in the case
of 9a and finally to good (R)-enantioselectivity (E=35) in the case of pyroglutamate 16a as shown
in Tables 2 and 3 (Scheme 3). As a possible binding mode, the change from S to R takes place
when the substituent Y competes from the subsite which in the case of (S)-selectivity tends to
be occupied by the succinate or glutarate backbone and vice versa.
Table 2
Acylase I (on Eupergit C)-catalyzed butanolysis of glutamic acid derivatives at room temperature
Compound
Time (h)
c (%)
eeS
eeP
6a
7a
16a
66
24
20
50
50
53
96
\99
92
95
\99
83
Table 3
Acylase I (on Eupergit C)-catalyzed alcoholysis of methyl pyroglutamate 16a in neat alcohols (t=4 h)
Alcohol
Temperature (°C)
c (%)
E
EtOH
PrOH
25
25
8
25
42
25
25
16
44
36
40
25
33
28
1998
3499
6395
3494
2092
4093
3597
BuOH
BuOH
BuOH
Pentanol
Hexanol

A. Liljeblad et al. / Tetrahedron: Asymmetry 11 (2000) 3957–3966
3961
Finally, the acylase I-catalyzed alcoholysis of methyl pyroglutamate 16a was studied in neat
alcohols (Table 3, Scheme 3). Compared to the open-chain structures of glutamic acid deriva-
tives 6a and 7a the cyclic structure now makes the molecule less flexible. The present results
clearly indicate that the butanolysis of 16a proceeds smoothly, the time needed to obtain 50%
conversion being about the same as in the case of 7a (Tables 1 and 2). Clearly, reduced
enantioselectivity is observed in the case of 16a compared with 7a. Efforts to enhance enzymatic
enantioselectivity by performing the alcoholysis in different alcohols were unsuccessful (Table 3).
However, a considerable temperature effect on E is evident in favor of low temperatures. The
more reactive enantiomer of 16a is the (R)-counterpart, i.e. an opposite enantiodiscrimination is
observed compared to the open-chain derivatives 6a and 7a. This is in accordance with the
expectation that the cyclic part of the molecule (ꢀNHCOCH2CH2ꢀ) is bound to the subsite
where the glutarate backbone of 6a or 7a is bound in the case of (S) stereoselectivity.
Interestingly, when the N-acetylated pyroglutamic acid was previously subjected to traditional
hydrolysis (Scheme 1) the substrate was practically unreactive.3
In the previous work, the gram-scale butanolysis of N-butanoyl dimethyl aspartate led to the
separation of the two enantiomers in a single resolution step by acylase I from Aspergillus
melleus.6 The gram-scale resolution of 6a has now been performed using acylase I on Eupergit
C (see Section 4).
3. Conclusions
Butanolyses of a-substituted succinic and glutaric acid dimethyl esters 1a–15a and that of
methyl pyroglutamate 16a in the presence of acylase I enzymes from Aspergillus melleus, genus
and sp. on Eupergit C were studied (Tables 1–3; Schemes 2 and 3). The excellent regioselectivity
of the enzymes makes them potential catalysts when the reactions at the sterically more hindered
a-methyl ester group at the stereocenter of polyfunctional substrates are needed.
Enantioselectivity of the acylase I enzymes strongly depends on the structure of a substrate.
Thus, acylase I on Eupergit C, in particular, is a fascinating resolution catalyst especially for the
highly (S)-enantioselective butanolyses of N-substituted aspartic and glutamic acid esters 5a–7a.
Accordingly, the gram-scale resolution of N-butanoylated dimethyl glutamate 6a with butanol
allowed the simultaneous preparation of the unreacted (R)-6a (ee 91%) and the corresponding
a-butyl b-methyl (S)-ester (S)-6b (ee 99%) at 48% conversion. Cyclic pyroglutamate 16a also
serves as a resolvable substrate, leading to the formation of the (R)-product. a-Substituted
carboxylic acid esters are potential substrates for enantioselective alcoholyses by acylase I
enzymes, but as is the case with many other hydrolytic enzymes,13 enantioselectivity is strongly
dependent on the structure of a chiral acid component.
4. Experimental
4.1. Materials
Acylase I from Aspergillus melleus was obtained from Sigma (Deisenhofen, Germany).
Acylase I from Aspergillus genus and from Aspergillus on Eupergit C (388.4 U/g) were the
products of Tokyo Kasei (Tokyo, Japan) and Fluka (Buchs, Switzerland), respectively. The
origin of the Aspergillus enzyme immobilized on Eupergit C is not reported by the producer.

3962
A. Liljeblad et al. / Tetrahedron: Asymmetry 11 (2000) 3957–3966
The esters of malic acid were products of Tokyo Kasei except for (S)-dimethyl malate which
was obtained from Aldrich. The esters of aspartic and (S)-glutamic acid hydrochloride,
N-benzyloxycarbonyl-(S)-glutamic acid and its a-methyl ester as well as (S)-bromo- and
(R)-chlorosuccinic acids were purchased from Sigma. Glutamic acid monohydrate, chloro- and
bromosuccinic acids and dimethyl methylsuccinate were products of Aldrich. (R)-Glutamic acid
and (S)-glutamic acid g-methyl ester were from Acros. (R)-3-Methylsuccinic acid was obtained
from Tokyo Kasei. The solvents were of the highest analytical grade and obtained from Lab
Scan Ltd or Aldrich.
4.2. Methods
The progress of the reactions was followed by taking samples (100 ml) at intervals, filtering off
the enzyme and analyzing the samples by GLC on Chrompack CP-Chirasil-DEX CB (com-
pounds 3a and 8a), Chrompack CP-Chirasil- -Valine (compounds 4a–6a and 16a) and Astec
L
Chiraldex G-TA capillary columns (compounds 1a, 2a, 9a–15a). Compound 7a was analyzed by
HPLC on a Chirasil OD column. Conversion was calculated according to the equation
c=eeS/(eeS+eeP) or by using dihexyl ether as an internal standard. The determination of E was
based on the equation E=ln[(1−c)(1−eeS)]/ln[(1−c)(1+eeS)].14 Using linear regression E is
achieved as the slope of a line. 1H NMR spectra were measured in CDCl3 on a Jeol Lambda 400
or Bruker 200 (compounds 6a, 7a, 10a, 11a and 13a–16a) Spectrometer, tetramethylsilane being
the internal standard. MS-spectra were recorded on a VG Analytical 7070E instrument equipped
with a VAXstation 3100 M76 computer. Optical rotations were measured using a Jasco DIP-360
polarimeter. Elemental analyses were performed using a Perkin–Elmer CHNS-2400 Ser II
Elemental Analyzer.
4.3. Synthesis of racemic starting materials
Racemates 1a, 2a and 8a were commercially available. Dimethyl aspartate 3a was prepared by
bubbling ammonia through the hydrochloride solution in chloroform. Dimethyl glutamate was
prepared from glutamic acid monohydrate and methanol with the aid of thionyl chloride.
Compounds 4a–6a and 11a–13a were prepared by normal procedures from dimethyl malate,
aspartate and glutamate and the corresponding acid anhydride (1.2 equiv.) or acid chloride
(1.5–2.0 equiv.) in the presence of DMAP (N,N-dimethylaminopyridine; 0.05–0.1 equiv.) and
triethyl amine (3–5 equiv.). The reaction between dimethyl glutamate (2.03 g, 11.6 mmol) and
benzyl chloroformate [7.3 ml (50% solution in toluene), 22 mmol] in toluene (90 ml) in the
presence of triethyl amine (4.0 ml, 28.9 mmol) resulted in the N-benzyloxycarbonyl (Cbz)
protection in the formation of 7a (2.41 g, 7.79 mmol). Methoxy substituted compound 9a (0.54
g, 3.1 mmol after 42 h) was synthesized from dimethyl malate (1.00 g, 6.2 mmol) and
iodomethane (1.16 ml, 18.5 mmol) in the presence of Ag2O (1.43 g, 6.2 mmol) in diethyl ether.15
Compound 10a (1.11 g, 5.4 mmol after 150 h) was synthesized in a similar manner from
dimethyl malate (1.00 g, 6.2 mmol), propyl iodide (2.41 ml, 24.7 mmol) and Ag2O (4.29 g, 18.5
mmol). Halogenated compounds 14a and 15a (yields 71 and 88% after 48 and 72 h, respectively)
were prepared from the corresponding acids (1 equiv.) by refluxing in methanol in the presence
of p-toluenesulfonic acid (1 equiv.). The above compounds were purified by column chromatog-
raphy using acetone/petroleum ether (1/9–3/7) as an eluent. As a modification of the known
method,16 methyl pyroglutamate 16a was prepared by heating dimethyl glutamate in an oil bath

A. Liljeblad et al. / Tetrahedron: Asymmetry 11 (2000) 3957–3966
3963
(3 h, 110°C) followed by chromatography with silica using elution with acetone/petroleum ether
(1/1) for the purification of the product.
4.4. Spectroscopic and analytical data
1
rac-(4a): H NMR: l (ppm) 2.04 (s, 3H, CH3CONH), 2.95 (2×q, 2H, CH2CO2Me), 3.70 (s,
3H, CO2CH3), 3.77 (s, 3H, CO2CH3), 4.86 (m, 1H, CH), 6.54 (s, 1H, NH). 13C NMR: l (ppm)
23.0 (CH3CO), 36.0 (CH2CO2Me), 48.4 (CH), 52.0 (CO2CH3), 52.7 (CO2CH3), 169.8 (CONH),
171.2 (CO2Me), 171.5 (CO2Me). Elemental analysis: obs. C, 46.93%; H, 6.32%; N, 6.78%. Calcd
C, 47.29%; H, 6.45%; N, 6.89% for the C8H13NO5.
1
rac-(6a): H NMR: l (ppm) 0.95 (t, 3H, CH3(CH2)2), 1.68 (m, 2H, CH2CH3), 2.01 (m, 2H,
CH2CH2CO2Me), 2.18 (m, 2H, CH2CONH), 2.38 (m, 2H, CH2CO2Me), 3.68 (s, 3H, CO2CH3),
3.75 (s, 3H, CO2CH3), 4.65 (m, 1H, CH), 6.32 (d, 1H, NH). 13C NMR: l (ppm) 13.6 (CH3),
18.9 (CH2CH3), 27.2 (CH2CH2CO2Me), 30.0 (CH2CO2Me), 38.3 (CH2CONH), 51.4 (CO2CH3),
51.8 (CO2CH3), 52.4 (CH), 172.4 (CO2), 173.0 (CO2), 173.3 (CONH). Mass spectrum: M+=
245.
1
rac-(7a): H NMR: l (ppm) 2.00 and 2.20 (2×m, 2H, CH2CH2CO2Me), 2.42 (td, 2H,
CH2CO2Me), 3.56 (s, 3H, CO2CH3), 3.74 (s, 3H, CO2CH3), 4.41 (m, 1H, CH), 5.10 (s, 2H,
PhCH2), 5.46 (d, 1H, NH), 7.35 (s, 5H, ArH). 13C NMR: l (ppm) 27.6 (CH2CH2CO2Me), 29.9
(CH2CO2Me), 51.8 (CO2CH3), 52.5 (CO2CH3), 53.3 (CHNH), 67.0 (NHCO2CH2), 128.0–128.5
(5C, ArꢀC), 136.1 (NHCO2CH2C), 155.9 (CONH), 172.2 (CO2), 173.0 (CO2). Mass spectrum:
M+=309.
1
rac-(9a): H NMR: l (ppm) 2.78 (2×q, 2H, CH2CO2Me), 3.49 (s, 3H, CH3O), 3.72 (s, 3H,
CO2CH3), 3.79 (s, 3H, CO2CH3), 4.21 (dd, 1H, CH). 13C NMR: l (ppm) 37.5 (CH2CO2Me),
51.9 (CO2CH3), 52.1 (CO2CH3), 58.7 (CH3O), 76.6 (CH), 170.4 (CO2), 171.6 (CO2). Elemental
analysis: obs. C, 47.70%; H, 6.87%. Calcd C, 47.72%; H, 6.87% for the C7H12O5.
1
rac-(10a): H NMR: l (ppm) 0.90 (t, 3H, CH3CH2), 1.59 (m, 2H, CH3CH2), 2.75 (2×q, 2H,
CH2CO2Me), 3.50 (2×m, 2H, CH2O), 3.71 (s, 3H, CO2CH3), 3.77 (s, 3H, CO2CH3), 4.21 (dd,
1H, CH). 13C NMR: l (ppm) 10.3 (CH3CH2), 22.8 (CH2CH3), 37.8 (CH2CO2Me), 51.9
(CO2CH3), 52.1 (CO2CH3), 73.1 (CH2O), 75.3 (CH), 170.6 (CO2), 172.1 (CO2). M+=204.
1
rac-(11a): H NMR: l 2.15 (s, 3H, CH3CO2CH), 2.89 (d, 2H, CH2CO2Me), 3.73 (s, 3H,
CO2CH3), 3.78 (s, 3H, CO2CH3), 5.48 (t, 1H, CH). 13C NMR: l (ppm) 20.7 (CH3CO), 36.0
(CH2CO2Me), 52.3 (CO2CH3), 52.9 (CO2CH3), 68.3 (CH), 169.5 (CO2CH), 169.8 (CO2Me),
170.1 (CO2Me). Elemental analysis: obs. C, 47.09%; H, 6.03%. Calcd C, 47.06%; H, 5.92% for
the C8H12O6.
1
rac-(13a): H NMR: l (ppm) 0.88 (t, 3H, CH3(CH2)4), 1.29 (m, 4×2H, CH2), 1.64 (m, 2H,
CH2CH2CO2), 2.39 (t, 2H, CH2CH2CO2), 2.89 (2×q, 2H, CH2CO2Me), 3.72 (s, 3H, CO2CH3),
3.77 (s, 3H, CO2CH3), 5.49 (t, 1H, CH). 13C NMR: l (ppm) 14.0 (CH3CH2), 22.5 (CH3CH2),
24.7 (CH2CH2CO2), 28.8 (CH2(CH2)3CO2 and CH3(CH2)3CH2), 31.6 (CH3CH2CH2), 33.8
(CH2CO2CH), 35.9 (CH2CO2Me), 52.0 (CO2CH3), 52.5 (CO2CH3), 68.0 (CH), 169.4 (CO2CH)
169.5 (CO2Me), 172.7 (CO2Me). Elemental analysis: obs. C, 59.37%; H, 8.51%. Calcd C,
58.32%; H, 8.39% for the C14H24O6.
1
rac-(14a): H NMR: l (ppm) 3.05 (2×q, 2H, CH2CO2Me), 3.73 (s, 3H, CO2CH3), 3.82 (s, 3H,
CO2CH3), 4.66 (dd, 1H, CH). 13C NMR: l (ppm) 39.3 (CH2CO2Me), 51.1 (CO2CH3), 52.2
(CO2CH3), 53.2 (CH), 169.7 (CO2), 173.1 (CO2). Elemental analysis: obs. C, 39.90%; H, 5.38%.
Calcd C, 39.91%; H, 5.02% for the C6H9ClO4.

3964
A. Liljeblad et al. / Tetrahedron: Asymmetry 11 (2000) 3957–3966
1
rac-(15a): H NMR: l (ppm) 3.10 (2×q, 2H, CH2CO2Me), 3.69 (s, 3H, CO2CH3), 3.78 (s,
3H, CO2CH3), 4.55 (dd, 1H, CH). 13C NMR: l (ppm) 37.7 (CH2CO2Me), 39.5 (CH), 52.1
(CO2CH3), 53.2 (CO2CH3), 169.5 (CO2), 170.0 (CO2). Elemental analysis: obs. C, 33.12%; H,
4.22%. Calcd C, 32.02%; H, 4.03% for the C6H9BrO4.
1
rac-(16a): H NMR: l (ppm) 2.26 (m, 2H, CH2CONH), 2.41 (m, 2H, CH2CH), 3.78 (s,
3H, CO2CH3), 4.28 (t, 1H, CH), 6.61 (s, 1H, NH). Mass spectrum: [M−H]+=142.
rac-(5a) and -(12a) were described in the previous paper.6
4.5. Enzymatic resolution
The resolutions were typically performed as small-scale experiments where one of the
substrates 1a–16a (0.1 M) was dissolved in butanol (2 ml). The enzyme preparation (75
mg/ml) was added in order to start the reaction and the reaction was shaken at room
temperature (25°C) or at another given temperature.
For gram-scale resolution, acylase I on Eupergit C (4.6 g) was added on 6a (1.51 g, 6.2
mmol) in butanol (62 ml). After 47 h the enzyme was filtered off at 48% con-
version. Purification by column chromatography (acetone:petroleum ether 3:7) yielded the
unreacted dimethyl ester (R)-6a (0.74 g, 2.6 mmol, ee 91%, [h]2D0 +26.1 (c=1, MeOH)) and
the corresponding (S)-a-butyl g-methyl ester (S)-6b (0.77 g, 2.7 mmol, ee 99%, [h]2D0 −23.6
(c=1, MeOH)). The spectroscopic data for (R)-6a are in accordance with those for racemic
6a.
1
(S)-6b: H NMR: l (ppm) 0.94, 0.95 (m, 6H, 2×CH3CH2), 1.40 (m, 2H, CH2CH3), 1.65
(m, 4H, CH2CH2CH3, CH2CH3), 2.00, 2.20 (2×m, 2H, CH2CH2CO2), 2.23 (m, 2H,
CH2CONH), 2.38 (m, 2H, CH2CO2), 3.68 (s, 3H, CH3), 4.14 (t, 2H, CO2CH2), 4.64 (m, 1H,
CH), 6.32 (d, 1H, NH). 13C NMR:
l
(ppm) 13.6 (CH3CH2CH2CONH), 13.8
(CH3CH2CH2CH2), 18.9 (CH2CH2CONH), 27.4 (CO2CH2CH2CH2), 30.0 (CH2CH2CO2), 30.4
(CH2CO2), 34.8 (CO2CH2CH2), 38.3 (CH2CONH), 51.5 (CO2CH3), 62.5 (CH), 65.4
(CO2CH2), 172.0 (CO2), 172.9 (CO2), 173.3 (CONH). M+=287.
4.6. Determination of absolute configurations
Absolute configurations of 3a–6a and 8a–13a were determined by synthesizing chemically
the (S)-enantiomers from commercial (S)-dimethyl malate and (S)-dimethyl aspartate and
glutamate hydrochlorides by the above-mentioned procedures. Compound (S)-7a was
obtained from N-Cbz-(S)-glutamic acid and methanol by the normal DCC (N,N-dicyclo-
hexylcarbodiimide) coupling procedure in the presence of DMAP. The enantiomers of 14a
and 15a were prepared from (R)-chloro- and (S)-bromosuccinic acids. The absolute configu-
ration of 1a was determined simultaneously with regioselectivity by preparing a-butyl b-
methyl (R)-2-methylsuccinate from the commercial b-methyl (R)-succinate by DCC and
DMAP in butanol. The retention times of the enantiomeric peaks of racemic and enantio-
pure compounds in the chromatogram were then compared for recognizing the less reactive
dimethyl ester enantiomer in the resolution mixtures. The commercial enantiomers gave a
sign about the antibody helping in identifying the two peaks for a racemate.

A. Liljeblad et al. / Tetrahedron: Asymmetry 11 (2000) 3957–3966
3965
4.7. Determination of enzymatic regioselectivity
The mixtures of dimethyl ester, methyl butyl regioisomers and dibutyl ester were obtained
when the corresponding dimethyl ester was dissolved in butanol in the presence of sulfuric acid
as a catalyst and the mixture was refluxed for 0.5–2 h. Analysis by GLC showed the retention
times for the enantiomers of the four possible esters in the reaction mixture. Regioselectivity of
the enzymatic reactions was then determined by preparing chemically the other one of the two
possible regioisomers from the commercial half ester as shown below. The place of the peak in
the chromatogram was then compared to that of the enzymatic reaction. In the case of 7a,
1
regioselectivity is based on the place of the CO2CH2CH2CH2CH3 protons in H NMR, giving
the triplets at 4.14 ppm for the a-butyl protons and at 4.06 for the g-butyl protons. The
synthesis products were not usually isolated.
Mixed a-butyl esters were prepared from the respective commercial b- or g-methyl ester of
(R)-2-methyl succinic acid and (S)-aspartic, (R)-glutamic or (S)-glutamic acid hydrochlorides by
the DCC coupling method in butanol followed by N-acylation with an acid chloride when
necessary.6 Mixed b- and g-butyl esters of a-methyl (S)-malic and N-Cbz-(S)-glutamic acids
were similarly prepared. For that purpose, a-methyl malic acid was synthesized by the known
method.17 b-Butyl a-methyl malate was transformed to the 2-methoxy analogue of 9a by
iodomethane using Ag2O as a catalyst as above and to the 2-bromo analogue of 15a with
phosphorus tribromide in CH2Cl2.15 For the reactions of 10a and 14a the same enzymatic
regioselectivity was expected as for those of 9a and 15a, respectively. Regioselectivity in the case
of 2a was detected straight by the chiral GLC method. The two possible monobutyl regioiso-
mers differ distinctively from each other. One is prochiral consisting of one peak in the
chromatogram and the other one consists of two enantiomers that separate well in the column.
Acknowledgements
The authors thank the Technology Development Center in Finland (TEKES) for financial
support and Dr. Reijo Aksela and Kemira Agro for their interest and support.
References
1. Anders, M. W.; Dekant, W. Adv. Pharm. 1994, 27, 431–448.
2. Uttamsingh, V.; Anders, M. W. Chem. Res. Toxicol. 1999, 12, 937–942.
3. Chenault, H. K.; Dahmer, J.; Whitesides, G. M. J. Am. Chem. Soc. 1989, 111, 6354–6364.
4. Bommarius, A. S.; Drauz, K.; Klenk, H.; Wandrey, C. Ann. NY Acad. Sci. 1992, 672, 126–136.
5. Udding, J. H.; Fraanje, J.; Goubitz, K.; Hiemstra, H.; Speckamp, W. N.; Kaptein, B.; Schoemaker, H. E.;
Kamphuis, J. Tetrahedron: Asymmetry 1993, 4, 425–432.
6. Liljeblad, A.; Kanerva, L. T. Tetrahedron: Asymmetry 1999, 10, 4405–4415.
7. Herrado´n, B.; Valverde, S. Synlett 1995, 599–602.
8. Ors, M.; Morcuende, A.; Jime´nez-Vacas, M. I.; Valverde, S.; Herrado´n, B. Synlett 1996, 449–451.
9. Faraldos, J.; Arroyo, E.; Herrado´n, B. Synlett 1997, 367–370.
10. Noheda, P.; Garc´ıa, G.; Pozuelo, M. C.; Herrado´n, B. Tetrahedron: Asymmetry 1996, 7, 2801–2804.
11. Bakker, M.; Spruijt, A. S.; van Rantwijk, F.; Sheldon, R. A. Tetrahedron: Asymmetry 2000, 11, 1801–1808.
12. Silverman, R. B. The Organic Chemistry of Enzyme-Catalyzed Reactions; Academic Press: San Diego, 2000; pp.
55–60.

3966
A. Liljeblad et al. / Tetrahedron: Asymmetry 11 (2000) 3957–3966
13. Bornscheuer, U. T.; Kazlauskas, R. J. Hydrolases in Organic Synthesis; Wiley-VCH: Weinheim, 1999; pp.
107–117.
14. Chen, C.-S.; Fujimoto, Y.; Girdaukas, G.; Sih, C. J. J. Am. Chem. Soc. 1982, 104, 7294–7299.
15. Bouzide, A.; Sauve´, G. Tetrahedron Lett. 1997, 38, 5945–5948.
16. Adkins, H.; Billica, H. R. J. Am. Chem. Soc. 1948, 70, 3121–3125.
17. Cammas, S.; Renard, I.; Boutault, K.; Gue´rin, P. Tetrahedron: Asymmetry 1993, 4, 1925–1930.
.