Enzymatic preparation of (S)-3-amino-3-(o-tolyl)propanoic acid, a key intermediate for the construction of Cathepsin inhibitors

Article abstract of DOI:10.1016/j.molcatb.2013.04.001

Enantiomerically pure (S)-3-amino-3-(o-tolyl)propanoic acid [(S)-6], identified as the preferred enantiomeric form for the construction of novel β-amino acid derivatives as inhibitors of Cathepsin, was prepared through both indirect and direct enzymatic strategies. Resolution of hydroxymethylated β-lactam (±)-1 through Burkholderia cepacia lipase PSIM-catalysed R-selective butyrylation (E > 200) was first carried out in t-BuOMe. Treatment of the unreacted (S)-1 with 18% HCl then furnished the desired (S)-6·HCl. Next, Candida antarctica lipase B catalysed the ring cleavage of racemic 4-(o-tolyl)azetidin-2-one [(±)-2] with excellent R enantioselectivity (E > 200), either in t-BuOMe with added H₂O as nucleophile or in H₂O at 60 C. Hydrolysis of the less reactive β-lactam enantiomer [(S)-2] with 18% HCl afforded (S)-6·HCl. A direct enzymatic route to enantiomeric (S)-6 was finally optimized through the lipase PSIM-catalysed S-enantioselective (E > 200) hydrolysis of racemic ethyl 3-amino-3-(o-tolyl)propanoate [(±)-3] in t-BuOMe with added H ₂O at 45 C or in H₂O at 3 C.

Full text of DOI:10.1016/j.molcatb.2013.04.001

Journal of Molecular Catalysis B: Enzymatic 93 (2013) 8–14
Contents lists available at SciVerse ScienceDirect
Journal of Molecular Catalysis B: Enzymatic
journal homepage: www.elsevier.com/locate/molcatb
Enzymatic preparation of (S)-3-amino-3-(o-tolyl)propanoic acid, a key
intermediate for the construction of Cathepsin inhibitors
Eniko˝ Forró, Gábor Tasnádi, Ferenc Fülöp∗
Institute of Pharmaceutical Chemistry, University of Szeged, Eötvös u 6, H-6720 Szeged, Hungary
a r t i c l e i n f o
a b s t r a c t
Article history:
Enantiomerically pure (S)-3-amino-3-(o-tolyl)propanoic acid [(S)-6], identified as the preferred enan-
tiomeric form for the construction of novel ␤-amino acid derivatives as inhibitors of Cathepsin, was
prepared through both indirect and direct enzymatic strategies. Resolution of hydroxymethylated ␤-
lactam ( )-1 through Burkholderia cepacia lipase PSIM-catalysed R-selective butyrylation (E > 200) was
first carried out in t-BuOMe. Treatment of the unreacted (S)-1 with 18% HCl then furnished the desired (S)-
6·HCl. Next, Candida antarctica lipase B catalysed the ring cleavage of racemic 4-(o-tolyl)azetidin-2-one
[( )-2] with excellent R enantioselectivity (E > 200), either in t-BuOMe with added H2O as nucleophile
or in H2O at 60 ◦C. Hydrolysis of the less reactive ␤-lactam enantiomer [(S)-2] with 18% HCl afforded
(S)-6·HCl. A direct enzymatic route to enantiomeric (S)-6 was finally optimized through the lipase PSIM-
catalysed S-enantioselective (E > 200) hydrolysis of racemic ethyl 3-amino-3-(o-tolyl)propanoate [( )-3]
in t-BuOMe with added H2O at 45 ◦C or in H2O at 3 ◦C.
Received 11 March 2013
Received in revised form 2 April 2013
Accepted 2 April 2013
Available online 8 April 2013
Keywords:
␤-Amino acid
␤-Amino ester
CatA inhibitor
␤-Lactam
Lipase
© 2013 Elsevier B.V. All rights reserved.
1. Introduction
Pseudomonas cepacia lipase PS or lipase PSIM-catalysed enantio-
selective (E > 200) hydrolyses of ␤-amino esters with H2O in an
A number of chiral ␤-aryl-substituted ␤-amino acids have been
intensively investigated, due to their pharmaceutical importance
[1–4] and their utility in synthetic chemistry [5–7] and drug
research [8–10]. They are also components of naturally occurring
cyclic peptide astins with antitumour properties [11]. In recent
years, extensive investigations have been carried out on the enan-
tioselective syntheses of such ␤-amino acids, and as a consequence
a large number of new enzymatic methods have emerged [12–16].
As an example, a direct enzymatic method has been described for
the enantioselective (E > 200) ring cleavage of acyclic ␤-lactams
[17]. Aryl substituents with different electronic characters and in
different positions were selected with regard to the synthetic appli-
cability of the products. The method was later optimized and two
very efficient direct enzymatic methods were developed, e.g. lipase-
catalysed kinetic and sequential kinetic strategies for the synthesis
of (2R,3S)-3-amino-3-phenyl-2-hydroxypropanoic acid [18], which
is the key structure for the side-chain of Taxol® (paclitaxel) and its
analogue Taxotère® (docetaxel), currently among the most impor-
tant drugs in cancer chemotherapy [19].
organic solvent [20–22]. One of the enantiomers prepared, (R)-
3-amino-3-(2,4,5-trifluorophenyl)butanoic acid, the intermediate
of the new antidiabetic drug JanuviaTM (sitagliptin phosphate)
[23,24], was prepared for the first time through enzymatic reso-
lution [22].
By means of the above-mentioned direct enzymatic strate-
gies, enantiomers of 3-amino-3-phenyl- and 3-amino-3-(p-
tolyl)propanoic acids and their derivatives have been synthetized
via lipase PS-catalysed R-selective butyrylation (E ≥ 57) of the pri-
mary hydroxy group of the corresponding N-hydroxymethylated
␤-lactams, or hydrolysis (E ≥ 89) of the corresponding ester deriva-
tives, followed by ring opening to the desired ␤-amino ester or acid,
respectively [25].
A highly enantioselective (E > 200) synthesis of (S)-3-amino-
3-phenylpropanoic acid, a key intermediate for the preparation
of (S)-dapoxetine, was reported recently, through Carica papaya
lipase-catalysed alcoholysis of its racemic N-protected 2,2,2-
trifluoroethyl esters in an organic solvent [26]. It was found that the
enantioselectivity and reaction rate were significantly enhanced by
switching the methyl ester to an activated trifluoroethyl ester.
(S)-3-Amino-3-(o-tolyl)propanoic acid was very recently iden-
tified as the preferred enantiomeric form for the construction of
Cathepsin (CatHA) inhibitors (Fig. 1) [27]. The inhibition of CatHA
has potential beneficial effects in cardiovascular diseases. One of
the novel CatHA inhibitors is currently undergoing phase I clinical
trials as a drug candidate.
As another example, an efficient direct enzymatic strategy
of biologically relevant ␤-aryl-, ␤-heteroaryl- and ␤-arylalkyl-
substituted ␤-amino acid enantiomers has been devised through
∗
Corresponding author. Tel.: +36 62 545564; fax: +36 62 545705.
E-mail address: fulop@pharm.u-szeged.hu (F. Fülöp).
1381-1177/$ – see front matter © 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.molcatb.2013.04.001

E. Forró et al. / Journal of Molecular Catalysis B: Enzymatic 93 (2013) 8–14
9
O
O
O
COOH
O
N
N
N
N
vinyl butyrate
lipase
OCOPr
OH
OH
+
N
H
HO
N
O
F
K2CO3
(R)-5
18% HCl,
( )-1
N
(S)-1
18% HCl,
∇
∇
MeOH
O
COOH
COOH
OH
.
.
NH2 HCl
NH2 HCl
Fig. 1. Optimized CatHA inhibitor.
.
.
(S)-6 HCl
(R)-6 HCl
(R)-1
Substituted ␣- and ␤-phenylalanines, among them (S)-3-amino-
3-(o-tolyl)propanoic acid (merely mentioned, but without data),
have been synthetized with excellent enantiomeric excess values
through the phenylalanine aminomutase (from Taxus chinensis)-
catalysed enantioselective addition of NH3 to a wide range of
substituted cinnamic acids [28].
Our aim in the present work was to prepare (S)-3-amino-
3-(o-tolyl)propanoic acid, through different kind of enzymatic
strategies. We initially planned to carry out enzymatic acylation
of the primary OH group of racemic 1-(hydroxymethyl)-4-(o-
tolyl)azetidin-2-one [( )-1] (Fig. 2). The extensive investigations
of the lipase-catalysed ring opening of ␤-lactams [17,18] and
the enantioselective hydrolysis of ␤-amino esters [20–22] then
suggested the possibility of the lipase-catalysed enantioselective
ring cleavage of racemic 4-(o-tolyl)azetidin-2-one [( )-2] and the
hydrolysis of racemic ethyl 3-amino-3-(o-tolyl)propanoate [( )-3]
(Scheme 1).
Scheme 2. Lipase-catalysed O-acylation of ( )-1.
acylation of ( )-1 (0.05 M) in toluene by using lipase PSIM
(immobilized on diatomaceous earth; 30 mg mL−1) as catalyst, and
vinyl butyrate (2 equiv.) as acyl donor, carrying out the reaction at
25 ◦C (Scheme 2). The highly enantioselective (E > 200) reaction had
reached 42% conversion (eeS = 72%, eeP > 99%) after 6 h.
When the toluene was replaced by t-BuOMe, accepted as a
green solvent, a faster reaction was observed (conv. = 47% after 5 h,
eeS = 88%, eeP > 99%), while E remained excellent (E > 200).
A further increase in reaction rate was obtained on raising
the reaction temperature from 25 to 45 ◦C (conv. = 42%, eeS = 72%,
eeP > 99% after 1 h at 45 ◦C vs. conv. = 31%, eeS = 44%, eeP > 99% after
2 h at 25 ◦C).
The gram-scale resolution of ( )-1 was therefore carried out
with lipase PSIM and vinyl butyrate in t-BuOMe at 45 ◦C. The results
are summarized in Table 1 and Section 4.
The stereochemistry of the resulting ester and the unreacted
alcohol enantiomers was assigned on the basis of the similar
chromatographic behaviour of racemic 1 and the earlier-studied
1-(hydroxymethyl)-4-arylazetidin-2-one and 1-(hydroxymethyl)-
4-(p-tolyl)azetidin-2-one. The analysed chromatograms indicated
that the corresponding enantiomers of the lactams in question react
preferentially on lipase PS- or PSIM-catalysed O-butyrylation. Thus,
the lipase PSIM-catalysed acylation of ( )-1 displayed R selectivity.
The transformations involving the hydrolysis of (R)-5 with
K2CO3/MeOH furnished the corresponding alcohol (R)-1 (ee > 99%),
while the ring opening of (R)-5 and (S)-1 with 18% aqueous
HCl resulted in the corresponding enantiomeric ␤-amino acid
hydrochloride 6 (ee(R)-6 > 99%, ee(S)-6 = 99%,), (S)-6·HCl being the
desired product (Section 4).
2. Results and discussion
Racemic 4-(o-tolyl)azetidin-2-one [( )-2] was prepared from
1-methyl-2-vinylbenzene 4 by chlorosulfonyl isocyanate (CSI)
addition [25,29]. Transformation of ( )-2 with paraformaldehyde
under sonication [25] resulted in N-hydroxymethylated lactam ( )-
1, the starting substance for lipase-catalysed O-acylation. The ring
opening of ( )-1 with 22% ethanolic HCl and subsequent libera-
tion of the free base with aqueous KOH [20] afforded the desired
racemic ethyl 3-amino-3-(o-tolyl)propanoate [( )-3] (Section 4).
2.1. Lipase-catalysed acylation of racemic
1-(hydroxymethyl)-4-(o-tolyl)azetidin-2-one [( )-1]
As reported previously on the asymmetric acylation of N-
hydroxymethylated 4-phenyl- (E > 200) and 4-(p-tolyl)- (E = 57)
azetidin-2-ones by using lipase PS (immobilized on Celite) in
toluene [25], we conducted preliminary experiments for the
2.2. Lipase-catalysed ring cleavage of racemic
4-(o-tolyl)azetidin-2-one [( )-2]
On the basis of our earlier results on the enzymatic ring open-
ing of 4-aryl-substituted [17,18] and 4-arylalkyl-substituted [31]
␤-lactams, we started preliminary experiments relating to the ring
opening of ( )-2 (Scheme 3) with 0.5 equiv. of H2O in t-BuOMe,
with enzyme screening at 45 or 60 ◦C. The lipases tested were Lipo-
lase, Novozym 435, Chirazyme L-2 (all three contain lipase B from
Candida antarctica, but immobilized differently), PPL (porcine pan-
creatic lipase), Chirazyme L-5 (lipase A from Candida antarctica),
lipase AK (Pseudomonas fluorescens), lipase PSIM and lipase AY
O
O
COOEt
NH2
~
N
NH
OH
2
( )-
( )-1
( )-3
Fig. 2. Substrates ( )-1–( )-3 in the enzymatic strategies planned.
O
O
COOEt
N
NH
OH
1. CH2O
NH2
1. CSI
2. ultrasound
2. Na2SO3
4
( )-2
( )-1
( )-3
1. 22% HCl/EtOH
2. 5% KOH
Scheme 1. Synthesis of ( )-1-( )-3.

10
E. Forró et al. / Journal of Molecular Catalysis B: Enzymatic 93 (2013) 8–14
Table 1
Gram-scale resolutions of ( )-1a, ( )-2b and ( )-3.c
Substrate
Time (h)
Conv. (%)b
E
Resulting enantiomers
Unreacted enantiomers
Yield (%)
Isomer
ee (%)
[˛]2D5
Yield (%)
Isomer
ee (%)
[˛]2D5
(
(
(
)-1
)-2
)-3
2
105
7
50d
51
50
>200
>200
>200
49
46
48
(R)-5
(R)-6
(S)-6
>99d
95h
+118.1e
+24.5i
−25k
40
44
48
(S)-1
(S)-2
(R)-3
>99f
>99f
99f
−292.7g
−285.1j
+20.1l
98h
a
b
c
d
e
f
Lipase PSIM (30 mg mL−1)-catalysed O-acylation of ( )-1 (0.05 M substrate) in t-BuOMe and vinyl butyrate (2 equiv.) at 45 ◦C.
CAL-B (30 mg mL−1)-catalysed ring cleavage of ( )-2 (0.05 M substrate) in t-BuOMe and H2O (0.5 equiv.) at 60 ◦C.
Lipase PSIM (30 mg mL−1)-catalysed hydrolysis of ( )-3 (0.05 M substrate) in t-BuOMe and H2O (10 equiv.) at 45 ◦C.
Determined by GC, using n-heptadecane as an internal standard.
c = 0.31; CHCl3.
According to GC.
c = 0.32; CHCl3.
According to GC after double derivatization [30].
c = 0.38; H2O.
c = 0.33; EtOH.
c = 0.15; H2O.
g
h
i
j
k
l
c = 0.40; EtOH.
Table 2
Conversion and enantioselectivity of ring cleavage of ( )-2.a
b
c
Entry
1
Enzyme (mg mL−1
Lipolase (30)
)
Time (h)
Added H2O (equiv.)
0.5
Conv. (%)
eeS (%)
eeP (%)
E
2
4
26
20
31
50
25
45
>99
>99
>99
>99
>200
>200
>200
2
3
4
5
Lipolase (10)
Lipolase (50)
Lipolase (75)
Lipolase (30)
2
2
2
4
32
4
4
4
2
26
2
0.5
0.5
0.5
0
5
26
31
26
50
32
26
16
26
51
5
5
34
64
35
98
46
34
19
34
88
24
>99
>99
>99
>99
>99
>99
>99
>99
>99
83
>200
>200
>200
>200
>200
>200
>200
>200
>200
31
6
7
8
9
Lipolase (30)
Lipolase (30)
Lipolase (30)
1
5
10
0.5
Novozym 435 (30)
10
Chirazyme L-2 (30)
0.5
21
2
a
0.05 M substrate in t-BuOMe at 60 ◦C.
According to GC.
According to GC after double derivatization [30].
b
c
(Candida rugosa) (Table 2). Not even traces of the ring-opened ␤-
amino acid were detected in the presence of PPL, Chirazyme L-5,
lipase AK or lipase AY after 2 h at 45 ◦C (data not shown), while a
very minor amount of ␤-amino acid was detected, but with low
ee (entry 10) in the presence of Chirazyme L-2, a carrier-fixed
lipase B from Candida antarctica. Novozym 435, immobilized on
macroporous poly(acrylic) beads, demonstrated excellent enan-
tioselectivity (E > 200, conv. 26% after 2 h) but with the progress
of the reaction this decreased considerably (E = 31, conv. 51% after
26 h) (entry 9). Lipolase, produced by the submerged fermenta-
tion of a genetically modified Aspergillus oryzae microorganism and
adsorbed on a macroporous resin, proved to be the best enzyme
(E > 200, conv. 50% after 26 h) (entry 1), and it was therefore
chosen as the enzyme for further studies and for the preparative-
scale resolution.
Next we analysed the effect of the added H2O on the enan-
tiodiscrimination and the reaction rate (entries 1 and 5–8). The
catalytic activity of the tested Lipolase was progressively lowered
on increase of the amount of H2O up to 1 equivalent, though the
enantioselectivity was apparently not affected (entries 7 and 8 vs.
entries 1 and 6). Furthermore, the degree of ring opening of ( )-2
in the presence of Lipolase was complete even without the addi-
tion of any H2O (entry 5), because the H2O present in the enzyme
preparation (<5%) or in the solvent used (<0.1%) was sufficient to
achieve the ␤-lactam ring opening.
In order to increase the reaction rate, the ring-opening reaction
was performed with 50 or 75 mg mL−1 instead of 30 mg mL−1 of
Lipolase in t-BuOMe in the presence of H2O (0.5 equiv.) at 60 ◦C.
As the amount of enzyme was increased, the necessary reaction
time (the time needed to reach 50% conversion) became shorter
and similarly good results (E > 200) were obtained (entries 3 and 4
vs. entry 1). The use of only 10 mg mL−1 of enzyme (entry 2) led to
the reaction proceeding relatively slowly, with a conversion of only
5% after 2 h.
O
O
COOH
NH2
H2O
lipase
NH
NH
+
( )-2
COOH
(R)-6
(S)-2
,
l
C
H
%
In spite of the fact that the optimum reaction rate was obtained
with 75 mg mL−1 of enzyme, for reasons of economy the gram-scale
resolution of ( )-2 was performed with 30 mg mL−1 of Lipolase.
When the ring-opening reaction was performed in distilled H2O
(pH 6.8), a good reaction rate (conv. = 24% after 2 h) and excellent
enantioselectivity (> 200) were observed (Table 3, entry 1). Several
8
1
.
NH2 HCl
.
(S)-6 HCl
Scheme 3. Lipase-catalysed ring cleavage of ( )-2.

E. Forró et al. / Journal of Molecular Catalysis B: Enzymatic 93 (2013) 8–14
11
Table 3
COOEt
COOH
NH2
COOEt
NH2
Conversion and enantioselectivity of ring cleavage of ( )-2.a
H2O
lipase
b
c
+
Entry
Buffer
Conv. (%)
eeS (%)
eeP (%)
E
NH2
1
2
3
4
5
6
7
Dist. H2O (pH 6.8)
Tap H2O (pH 7.4)
Tris–HCl (pH 7.8)
Phosphate (pH 7.0)
Malonate (pH 6.0)
Citrate (pH 5.5)
24
33
33
36
28
29
28
31
49
49
55
35
30
34
>99
>99
>99
97
88
75
>200
>200
>200
172
22
( )-3
(S)-6
(R)-3
18% HCl
COOH
18% HCl,
COOH
9
24
Malate (pH 5.0)
89
.
NH2 . HCl
NH2 HCl
a
0.05 M substrate in buffer.
According to GC.
According to GC after double derivatization [30].
b
c
.
.
(R)-6 HCl
(S)-6 HCl
Scheme 4. Lipase-catalysed hydrolysis of ( )-3.
buffers were also tested (entries 2–7) and it was finally concluded
that a pH close to neutral or basic (entries 1–4) was accompanied
by excellent values of E, whereas acidic pH caused a significant drop
in E (entries 5–7).
at 25 or 60 ◦C with enzyme screening. The lipases tested were Chi-
razyme L-5, Lipolase, PPL, lipase AK, lipase PSIM and lipase AY
(Table 4). No enantioselectivity was detected in the presence of
Chirazyme L-5 or lipase AY (entries 1 and 20), while only moderate
results (6 ≤ E ≤ 16) were obtained with Lipolase and PPL (entries 2
and 3). Lipase AK (entry 4) and Lipase PSIM (entry 9) proved to be
the best enzymes (conv. ∼43% after 63 h, E ≥ 87).
In view of the preliminary results, the gram-scale resolution of
(
)-2 was performed with 0.5 equiv. of H2O in the presence of Lipo-
lase (30 mg mL−1) in t-BuOMe at 60 ◦C. The results are presented in
Table 1. When the unreacted lactam (S)-2 was hydrolysed with 18%
HCl, the desired amino acid hydrochloride, (S)-6·HCl, was formed.
Absolute configurations were proved by comparing the ␣-
value of the ring-opened amino acid hydrochloride {[˛]2D5 = −6.1
(c = 0.35; H2O)} obtained from enantiomeric 2, the unreacted
lactam enantiomer obtained in the enzymatic ring cleavage of
As the temperature of the lipase AK-catalysed hydrolysis of ( )-
3 was raised from 25 to 45 ◦C, the reaction rate rose considerably,
while E also increased (entry 5 vs. 4). When lipase PSIM was used
at 45 instead of 25 ◦C, the reaction became least 10 times faster,
without a drop in E (entry 10 vs. 9).
At the beginning of the lipase AK-catalysed hydrolysis of ( )-3,
E was <200 (entry 4), and several solvents were therefore tested to
study the effect of the solvent in the lipase AK-catalysed hydrolysis
of ( )-3 at 25 ◦C (entries 6–8). The reaction proceeded considerably
more slowly in all of the solvents tested than in t-BuOMe (entry 4),
though all afforded high enantioselectivities (E > 200).
(
)-2 (Scheme 3), with the data obtained for (S)-3-amino-3-(o-
tolyl)propanoic acid hydrochloride {[˛]2D5 = −6.4 (c = 0.5; H2O},
described in section 2.1 (Section 4); thus, the Lipolase-catalysed
hydrolysis of ( )-2 displayed R selectivity.
2.3. Lipase-catalysed hydrolysis of racemic ethyl
3-amino-3-(o-tolyl)propanoate [( )-3]
As established previously for the enantioselective hydrolyses of
acyclic ␤-amino esters by using lipase PS or PSIM in an organic sol-
vent [20–22], we started preliminary experiments relating to the
hydrolysis of ( )-3 (Scheme 4) with 0.5 equiv. of H2O in t-BuOMe
Further increases in the lipase PSIM-catalysed reaction rate
were achieved by increasing the quantity of enzyme from 30 to 50,
and then to 75 mg mL−1 (entries 10–12). Although the optimum
E and reaction rate were observed with 75 mg mL−1 of enzyme,
Table 4
Conversion and enantioselectivity of the hydrolysis of ( )-3.a
b
c
Entry
Enzyme (mg mL−1
)
Time (h)
Temp. (◦C)
Solvent (1 mL)
Added H2O (equiv.)
Conv. (%)
eeS (%)
eeP (%)
E
1
2
3
4
Chirazyme L-5d (30)
Lipolase (30)
103
100
103
49
63
102
49
63
63
63
63
5
5
1
5
1
1
1
1
25
60
25
25
t-BuOMe
t-BuOMe
t-BuOMe
t-BuOMe
0.5
0.5
0.5
0.5
No selectivity
86
11
45
74
89
69
19
20
22
66
78
87
51
96
26
69
11
32
44
48
42
16
17
18
41
44
47
34
49
21
39
41
50
50
50
49
38
87
95
95
96
6
16
61
87
147
102
PPL (30)
Lipase AKd (30)
5
6
7
8
Lipase AKd (30)
Lipase AKd (30)
Lipase AKd (30)
Lipase AKd (30)
Lipase PSIM (30)
Lipase PSIM (30)
Lipase PSIM (50)
Lipase PSIM (75)
45
25
25
25
25
45
45
45
t-BuOMe
EtOAc
EtOH
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
96
>99
>99
>99
>99
>99
99
99
98
99
99
>200
>200
>200
>200
>200
>200
>200
>200
>200
>200
>200
>200
>200
>200
>200
Me2CO
9
t-BuOMe
t-BuOMe
t-BuOMe
t-BuOMe
10
11
12
13
14
15
16
17
18
19
20
Lipase PSIM (30)
Lipase PSIM (30)
Lipase PSIM (30)
Lipase PSIM (30)
Lipase PSIM (30)
Lipase PSIM (30)
Lipase PSIM (30)
Lipase AYd (30)
45
45
45
45
45
25
3
t-BuOMe
t-BuOMe
t-BuOMe
t-BuOMe
H2O
H2O
H2O
t-BuOMe
0
1
5
63
68
99
99
99
97
99
10
>99
>99
>99
>99
0.05
0.05
0.05
103
25
0.5
No selectivity
a
0.05 M substrate.
According to GC.
According to GC after double derivatization [30].
b
c
d
Contains 20% (w/w) of lipase adsorbed on Celite in the presence of sucrose.

12
E. Forró et al. / Journal of Molecular Catalysis B: Enzymatic 93 (2013) 8–14
further preliminary experiments were performed with 30 mg mL−1
of enzyme.
4. Experimental
4.1. Materials and methods
Finally, a set of preliminary experiments were performed at
45 ◦C with different amounts of added H2O (from 0 to 10 equiv.)
(entries 13–16). As the H2O content was increased, the reaction
became increasingly faster up to the point when the reaction was
carried out in distilled H2O, when a conversion of 50% was reached
after only 5 min (entry 17). For practical reasons, primarily in order
to attempt to decrease the reaction rate, the reaction tempera-
ture was decreased to 25, and then to 3 ◦C (entries 18 and 19),
but no change was observed in the reaction rate, while the E
value remained excellent. This latest result gives rise to the pos-
sibility of a very efficient hydrolysis of ( )-3 with high E (>200)
and a very attractive reaction rate (conv. = 50% after only 3 min)
in H2O.
Lipolase (lipase B from Candida antarctica), produced by the
submerged fermentation of a genetically modified Aspergillus
oryzae microorganism and adsorbed on a macroporous resin
(Catalogue no. L4777), and porcine pancreatic lipase (PPL) were
from Sigma–Aldrich. Chirazyme L-2 (a carrier-fixed lipase B from
Candida antarctica) and Chirazyme L-5 (lipase A from Candida
antarctica) were purchased from Roche Diagnostics Corporation,
while Novozym 435 as an immobilized preparation was from
Novo Nordisk. Lipase PSIM (Burkholderia cepacia, immobilized on
diatomaceous earth) was a kind gift from Amano Enzyme Europe
Ltd. Lipase AK (Pseudomonas fluorescens) was from Amano Pharma-
ceuticals, and lipase AY (Candida rugosa) and vinyl butyrate were
from Fluka. Before use, lipase AK or lipase AY or Chirazyme L-5 (5 g)
was dissolved in Tris–HCl buffer (0.02 M; pH 7.8) in the presence of
sucrose (3 g), followed by adsorption on Celite (17 g) (Sigma). The
lipase preparation thus obtained contained 20% (w/w) of lipase. The
solvents were of the highest analytical grade.
In a typical small-scale experiment, racemic substrate (0.05 M
solution) in an organic solvent or organic solvent–H2O mixture or
H2O or buffer solution (1 mL) was added to the lipase tested (10,
30, 50 or 75 mg mL−1). Vinyl butyrate (2 equiv.) or H2O (0, 0.5, 1,
5 or 10 equiv.) was next added. The mixture was shaken at 3, 25,
45 or 60 ◦C. The progress of the reaction was followed by taking
samples from the reaction mixture at intervals and analysing them
by GC. The ee values for the unreacted substrate enantiomers were
determined by GC on a Chrompack Chiralsil-Dex CB column (25 m)
{120 ◦C for 7 min → 180 ◦C (temperature rise 10 ◦C min−1), 140 kPa,
retention times (min), (S)-1: 25.34 [antipode (R)-1: 24.55]; 190 ◦C,
140 kPa, retention times (min), (S)-2: 9.1 [antipode (R)-2.: 8.80]} or
on a Chirasil L-Val column (30 m) {150 ◦C, 72 kPa, retention times
(min), (R)-3: 26.51 [antipode (S)-3: 26.74]}. The ee values for the
product amino acid 6 were determined on the earlier-mentioned
Chirasil L-Val column after a simple and rapid double derivatiza-
tion [30] with (i) CH2N2 (caution! derivatization with diazomethane
should be performed under a well-working hood); (ii) Ac2O in the
presence of 4-dimethylaminopyridine and pyridine {150 ◦C, 72 kPa,
retention times (min), (R)-6: 19.92 and (S)-6: 20.68}, while the ee
values for the ester 5 produced (during the preliminary exper-
iments) were calculated by using n-heptadecane as an internal
standard.
The preparative-scale resolution of ( )-3 was performed with
lipase PSIM and 5 equiv. of H2O in t-BuOMe at 45 ◦C. The results are
summarized in Table 1 and Section 4. When (S)-6 was hydrolysed
with 18% HCl, the desired amino acid hydrochloride, (S)-6·HCl, was
formed.
The absolute configurations were proved by comparing the
␣-value of the amino acid hydrochloride {[˛]2D5 = −6.4 (c = 0.55;
H2O)} prepared from enantiomer 6, the amino acid enantiomer
obtained in the enzymatic hydrolysis of
( )-3 (Scheme 4),
with the data obtained for (S)-3-amino-3-(o-tolyl)propanoic acid
hydrochloride {[˛]2D5 = −6.4 (c = 0.5; H2O}, described in section 2.1
(Section 4); thus, lipase PSIM catalysed the hydrolysis of ( )-3 with
S selectivity.
3. Conclusions
Indirect and direct enzymatic strategies have been given for the
synthesis of (S)-3-amino-3-(o-tolyl)propanoic acid [(S)-6], a key
intermediate for the construction of novel ␤-amino acid derivatives
recognized as inhibitors of Cathepsin.
Burkholderia cepacia lipase PSIM catalysed the R-acylation of
(
)-1 with vinyl butyrate (2 equiv.) in t-BuOMe, at 45 ◦C with
excellent enantioselectivity (E > 200) in spite of the relatively large
distance between the reaction site and the asymmetric centre.
Treatment of the unreacted alcohol [(S)-1] (ee > 99%) with 18% HCl
furnished the desired (S)-6·HCl (ee > 99%). The excellently appli-
cable Candida antarctica lipase B (Lipolase)-catalysed R-selective
ring cleavage of ( )-2 with added H2O (0.5 equiv.) in t-BuOMe
at 60 ◦C (E > 200), followed by ring opening of the unreacted lac-
tam (S)-2 with 18% HCl, similarly afforded the desired (S)-6·HCl
(ee > 99%). A direct enzymatic strategy (E > 200) was also devised
through the lipase PSIM-catalysed S-selective hydrolysis of amino
ester ( )-3 with H2O (5 equiv.) in t-BuOMe at 45 ◦C or in H2O at
3 ◦C, which afforded (S)-6 with high ee (98%) and in good yield
(48%). Although (S)-6 (merely mentioned, but without data) has
been prepared by an asymmetric route [28], to the best of our
knowledge it has not yet been obtained by lipase-catalysed kinetic
resolution.
Optical rotations were measured with a Perkin-Elmer 341
polarimeter. 1H NMR spectra were recorded on a Bruker Avance
DRX 400 spectrometer. Melting points were determined on a Kofler
apparatus.
4.2. Synthesis of racemic 4-(o-tolyl)azetidin-2-one [( )-2]
A solution of CSI (2.22 mL, 25.4 mmol) in toluene (25 mL) was
added dropwise to a stirred solution of 1-methyl-2-vinylbenzene
4 (3 g, 25.4 mmol) in toluene (25 mL) at 0 ◦C. The reaction mixture
was stirred at room temperature for 22 h, and the resulting liq-
uid was then added dropwise to a vigorously stirred solution of
Na2SO3 (0.33 g) and K2CO3 (6.90 g) in H2O (50 mL). After stirring of
the mixture for 3 h, the organic layer was separated and the aque-
ous phase was extracted with Et2O. The combined organic layers
were dried (Na2SO4), filtered and concentrated. The resulting white
solid, racemic 2 (3.66 g, 89%), was recrystallized from t-BuOMe (mp
101–102 ◦C).
The advantages of the enzymatic ring cleavage of lac-
tam
( )-2 and the hydrolysis of amino ester ( )-3 in an
organic solvent are the fact that the amino group need
not necessarily be protected, and the products can be easily
separated.
One common advantage of the use of lipases is that they are
commercially available, stable, demonstrate high enantioselectiv-
ity and can be applied on an industrial scale [32,33].
In parallel, a new GC method has been developed for the enan-
tioseparation of racemic 3-amino-3-(o-tolyl)propanoic acid [6] on a
CP-Chirasil-Dex CB column after a simple and rapid double deriva-
tization (esterification followed by N-acylation).
The 1H NMR (400 MHz, CDCl3, 25 ◦C, TMS) ı (ppm) data for ( )-
2: 2.33 (3 H, s, CH3), 2.78–2.82 (1 H, dd, J = 2.5, 14.6, CHAH) 3.48–3.54
(1H, m, CHBH), 4.93–4.95 (1H, dd, J = 2.6; 7.7, CH), 6.09 (1H, bs, NH)

E. Forró et al. / Journal of Molecular Catalysis B: Enzymatic 93 (2013) 8–14
13
7.20–7.45 (4H, m, Ph). Anal. calcd for C10H11NO: C, 74.51; H, 6.88;
N, 8.69. Anal. found: C, 74.70; H, 6.87; N, 8.66.
Ph). Anal. calcd for C15H19NO3: C, 68.94; H, 7.33; N, 5.36. Anal.
found: C, 69.05; H, 7.33; N, 5.30.
The 1H NMR (400 MHz, D2O) ı (ppm) data for (R)-6·HCl were
similar to those for (S)-6·HCl: 2.48 (3H, s, CH3), 3.11–3.27 (2H, m,
CH2), 5.11–5.14 (1H, m, CH), 7.39–7.52 (4H, m, Ph). Anal. calcd for
4.3. Synthesis of racemic
1-(hydroxymethyl)-4-(o-tolyl)azetidin-2-one [( )-1]
C
10H14ClNO2: C, 55.69; H, 6.54; N, 6.49. Anal. found for (R)-6·HCl: C,
(
)-2 (1 g, 6.2 mmol) was dissolved in THF (15 mL).
55.49; H, 6.54; N, 6.35. Anal. found for (S)-6·HCl: C, 55.55; H, 6.48;
Paraformaldehyde (0.19 g, 6.5 mmol), K2CO3 (0.1 g) and H2O
(0.38 mL) were added and the solution was sonicated for 5 h. The
solvent was next evaporated off and the residue was dissolved in
EtOAc (30 mL). The solution was dried (Na2SO4) and then concen-
trated. The residue was recrystallized from t-BuOMe to afford a
white crystalline product, ( )-1 (1.11 g, 93%; mp 88–89 ◦C).
The 1H NMR (400 MHz, CDCl3, 25 ◦C, TMS) ı (ppm) data for ( )-
1: 2.37 (3H, s, CH3), 2.77–2.81 (1H, dd, J = 2.7; 14.8, CHAH), 3.35–3.40
(1H, dd, J = 5.4; 14.8, CHBH), 3.89–3.90 (1H, bs, OH), 4.34–4.37 (1H,
d, 11.6 CHN), 5.12–5.17 (2H, m, CH2OH), 7.23–7.37 (4H, m, Ph).
Anal. calcd for C11H13NO2: C, 69.09; H, 6.85; N, 7.32. Anal. found:
C, 68.95; H, 6.88; N, 7.29.
N, 6.50.
4.6. Lipolase-catalysed ring cleavage of ( )-2
Racemic ˇ-lactam 2 (0.5 g, 3.10 mmol) was dissolved in t-BuOMe
(40 mL). Lipolase (1.2 g, 30 mg mL−1) and H2O (280 ␮L, 15.5 mmol)
were added, and the mixture was shaken in an incubator shaker
at 60 ◦C for 105 h. The reaction was stopped by filtering off the
enzyme at 51% conversion. The solvent was evaporated off and (S)-2
crystallized out from the residue [211 mg, 42%; recrystallized from
t-BuOMe {[˛]2D5 = −285.1 (c = 0.33; EtOH); mp 79–81 ◦C, ee > 99%}.
The filtered-off enzyme was washed with distilled H2O (3 × 15 mL),
and the H2O was evaporated off, yielding the crystalline ␤-amino
acid (R)-6 [248 mg, 46%; recrystallized from H2O and Me2CO;
[˛]2D5 = +24.5 (c = 0.38; H2O); mp 236–237 ◦C; ee = 95%]. When
(S)-2 (100 mg) was refluxed with 18% HCl (5 mL), (S)-6·HCl was
obtained [105 mg, 77%; [˛]2D5 = −6.4 (c = 0.5; H2O); mp 185–188,
ee = 99%].
4.4. Synthesis of racemic ethyl 3-amino-3-(o-tolyl)propanoate
[( )-3]
When ( )-2 (1 g) was refluxed with 22% HCl/EtOH (30 mL)
for 4 h, followed by solvent evaporation ( )-3·HCl was obtained.
Liberation of the free base with aqueous KOH (5%) afforded the
desired ethyl 3-amino-3-(o-tolyl)propanoate [( )-3] (1.19 g, 89%)
as a colourless oil.
The 1H NMR (400 MHz, CDCl3, 25 ◦C, TMS) ı (ppm) data for (S)-2
were similar to those for ( )-2. Anal. found: C, 74.38; H, 6.81; N,
8.60.
The 1H NMR (400 MHz, DMSO) ı (ppm) data for ( )-3: 1.08–1.13
(3H, t, J = 7, CH3), 2.26 (3H, s, CH3), 2.43–2.51 (2H, m, CH2CH3),
3.94–4.00 (2H, m, CH2COOEt), 4.37–4.41 (1H, dd, J = 5.7; 8.2, CH),
7.05–7.41 (4H, m, Ph). Anal. calcd for C12H17NO2: C, 69.54; H, 8.27;
N, 6.76. Anal. found: C, 69.66; H, 8.18; N, 6.87.
The 1H NMR (400 MHz, D2O) ı (ppm) data for (R)-6: 2.47 (3H,
s, CH3), 2.82–2.96 (2H, m, CH2), 4.98–5.02 (1H, dd, J = 6.3; 8.0, CH),
7.39–7.52 (4H, m, Ph). Anal. calcd for C10H13NO2: C, 67.02; H, 7.31;
N, 7.82. Anal. found: C, 66.84; H, 7.27; N, 7.90.
4.7. Lipase PSIM-catalysed hydrolysis of ( )-3
4.5. Lipase PSIM-catalysed acylation of ( )-1
Racemic 3 (0.5 g, 2.41 mmol) was dissolved in t-BuOMe (40 mL).
Lipase PSIM (1.2 g, 30 mg mL−1) and H2O (430 ␮L, 24 mmol) were
added and the mixture was shaken in an incubator shaker at
45 ◦C for 7 h. The reaction was stopped by filtering off the enzyme
at 50% conversion. The solvent was evaporated off, resulting in
the unreacted (R)-3 [240 mg, 48%; [˛]2D5 = +20.1 (c = 0.40; EtOH);
ee = 99%]. The filtered-off enzyme was washed with distilled H2O
(3 × 15 mL), and the H2O was evaporated off, yielding the crys-
talline amino acid (S)-6 {200 mg, 48%; [˛]2D5 = −25 (c = 0.15; H2O),
mp 235–237 ◦C, recrystallized from H2O, ee = 98%}. Treatment of
(S)-6 or (R)-3 (100 mg) with 18% HCl (5 mL) resulted in (S)-6·HCl
{100 mg, 83%; [˛]2D5 = −6.4 (c = 0.55; H2O), ee = 99%} or (R)-6·HCl
{96 mg, 92%; [˛]2D5 = +6.4 (c = 0.3; H2O, ee = 99%}.
Racemic 1 (0.5 g, 2.61 mmol) and vinyl butyrate (0.42 mL,
5.22 mmol) in t-BuOMe (30 mL) were added to lipase PSIM (0.9 g,
30 mg mL−1) and the mixture was shaken at 45 ◦C for 2 h. The reac-
tion was stopped by filtering off the enzyme at 50% conversion
{ee-[(R)-5] > 99%, ee-[(S)-1] > 99%}. After the solvent had been evap-
orated off, separation by column chromatography [EtOAc:hexane
(3:1)] furnished the unreacted (S)-1 {0.2 g, 40%; [˛]2D5 = −292.7
(c = 0.32; CHCl3); mp 66–68 ◦C, recrystallized from t-BuOMe);
ee > 99%} and the ester (R)-5 {0.33 g, 49%; [˛]2D5 = +118.1 (c = 0.31;
CHCl3); ee > 99%} as a pale-yellow oil. When (S)-1 or (R)-5 (100 mg)
was refluxed with 18% HCl (5 mL), (S)-6·HCl {110 mg, 97%; [˛]D25
=
−6.1 (c = 0.35; H2O), mp 180–185 ◦C (recrystallized from EtOH and
Et2O); ee = 99%} or (R)-6·HCl {104 mg, 92%; [˛]25 = +6.1 (c = 0.3;
The 1H NMR (400 MHz, CDCl3, 25 ◦C, TMS) ı (ppm) data for (R)-
3 were similar to those for ( )-3. Anal. found: C, 69.50; H, 8.42; N,
6.75.
D
H2O, mp 182–186 ◦C (recrystallized from EtOH and Et2O); ee = 99%}
was obtained.
A
mixture of (R)-5 (0.1 g, 0.38 mmol) and K2CO3 (0.14 g,
The 1H NMR (400 MHz, D2O) ı (ppm) data for (S)-6 were similar
to those for (R)-6. Anal. found for: C, 67.02; H, 7.31; N, 7.82.
0.96 mmol) in MeOH (15 mL) was stirred for 3 h at 45 ◦C. After evap-
oration, the residue was dissolved in H2O (20 mL) and extracted
with Et2O. The organic phase was dried (Na2SO4), filtered and evap-
orated. The product, (R)-1 {0.064 g, 87%; [˛]2D5 = +288.9 (c = 0.3,
CHCl3); ee = 98%}, was recrystallized from t-BuOMe (mp 66–69 ◦C).
The 1H NMR (400 MHz, CDCl3, 25 ◦C, TMS) ı (ppm) data for (S)-1
were similar to those for (R)-1 and ( )-1. Anal. found for (S)-1: C,
69.00; H, 6.79; N, 7.20. Anal. found for (R)-1: C, 69.13; H, 6.99; N,
7.31.
Acknowledgements
The authors acknowledge the receipt of grants OTKA No. K-
75433, K-108943 and TÁMOP-4.2.2/A-11/1/KONV-2012-0052 for
financial support.
The 1H NMR (400 MHz, CDCl3, 25 ◦C, TMS) ı (ppm) data for (R)-
5: 2.37 (3H, t, J = 7.3, CH2CH2CH3), 1.59–1.67 (2H, m, CH2CH2CH3),
2.27–2.31 (2H, m, CH2CH2CH3), 2.38 (1H, s CH3), 2.82–2.86 (1H, dd,
J = 2.7; 15, CHAH), 3.46–3.52 (1H, dd, J = 5.6; 15, CHBH), 4.96–5.04
(2H, m, CH2OCO), 5.42–5.45 (1H, d, 11.3 CHN), 7.23–7.32 (4H, m,
References
[1] H.H. Wasserman, H. Matsuyama, R.P. Robinson, Tetrahedron 58 (2002)
7177–7190.
[2] F. Fülöp, Chem. Rev. 101 (2001) 2181–2204.
[3] F. Fülöp, T.A. Martinek, G.K. Tóth, Chem. Soc. Rev. 35 (2006) 323–334.

14
E. Forró et al. / Journal of Molecular Catalysis B: Enzymatic 93 (2013) 8–14
[4] F. Fülöp, T.A. Martinek, Chem. Soc. Rev. 41 (2012) 687–702.
[5] O. Renault, J. Guillon, P. Dallemagne, S. Rault, Tetrahedron Lett. 41 (2000)
681–683.
[6] N. Leflemme, P. Dallemagne, S. Rault, Tetrahedron Lett. 45 (2004) 1503–1505.
[7] E.E. Baquero, W.H. James, S.H. Choi, S.H. Gellman, T.S. Zwier, J. Am. Chem. Soc.
130 (2008) 4795–4807.
[8] R. Juaristi, V.A. Soloshonok (Eds.), Enantioselective Synthesis of ␤-amino Acids,
2nd ed., Wiley-VCH, New York, 2005.
[9] T. Mukai, N. Suganuma, K. Soejima, J. Sasaki, F. Yamamoto, M. Maeda, Chem.
Pharm. Bull. 56 (2008) 260–265.
[10] S. Wisén, J. Androsavich, B.G. Evans, L. Chang, J.E. Gestwicki, Bioorg. Med. Chem.
Lett. 18 (2008) 60–65.
[11] S. Kosemura, T. Ogawa, K. Totsuka, Tetrahedron Lett. 34 (1993) 1291–1294.
[12] E. Forró, F. Fülöp, Mini Rev. Org. Chem. 1 (2004) 93–102.
[13] A. Liljeblad, L.T. Kanerva, Tetrahedron 62 (2006) 5831–5854.
[14] L. Kiss, E. Forró, F. Fülöp, in: A.B. Hughes (Ed.), Amino Acids, Peptides and Pro-
teins in Organic Chemistry, vol. 1, Wiley-VCH, Weinheim, 2009, pp. 367–409.
[15] E. Busto, V. Gotor-Fernández, V. Gotor, Chem. Rev. 111 (2011) 3998–4035.
[16] E. Forró, F. Fülöp, Curr. Med. Chem. 19 (2012) 6178–6187.
[17] E. Forró, T. Paál, G. Tasnádi, F. Fülöp, Adv. Synth. Catal. 348 (2006) 917–923.
[18] E. Forró, F. Fülöp, Eur. J. Org. Chem. (2010) 3074–3079.
[19] D.G.I. Kingston, D.J. Newman, Curr. Opin. Drug Disc. Dev. 10 (2007) 130–144.
[20] G. Tasnádi, E. Forró, F. Fülöp, Tetrahedron: Asymmetry 19 (2008) 2072–2077.
[21] G. Tasnádi, E. Forró, F. Fülöp, Tetrahedron: Asymmetry 20 (2009) 1771–1777.
[22] G. Tasnádi, E. Forró, F. Fülöp, Org. Biomol. Chem. 8 (2010) 886–895.
[23] D. Kim, J.E. Kowalchick, L.L. Brockunier, E.R. Parmee, G.J. Eiermann, M.H. Fisher,
H. He, B. Leiting, K. Lyons, G. Scapin, S.B. Patel, A. Petrov, K.D. Pryor, R.S. Roy,
J.K. Wu, X. Zhang, M.J. Wyvratt, B.B. Zhang, L. Zhu, N.A. Thornberry, A.E. Weber,
J. Med. Chem. 51 (2008) 589–602.
[24] N.A. Thornberry, A.E. Weber, Curr. Top. Med. Chem. 7 (2007) 557–568.
[25] E. Forró, F. Fülöp, Tetrahedron: Asymmetry 12 (2001) 2351–2358.
[26] P. You, J. Qiu, E. Su, D. Wei, Eur. J. Org. Chem. (2013) 557–565.
[27] S. Ruf, C. Buning, H. Schreuder, G. Horstick, W. Linz, T. Olpp, J. Pernerstorfer, K.
Hiss, K. Kroll, A. Kannt, M. Kohlmann, D. Linz, T. Hübschle, H. Rütten, K. Wirth,
T. Schmidt, T. Sadowski, J. Med. Chem. 55 (2012) 7636–7649.
[28] W. Szymanski, B. Wu, B. Weiner, S. de Wildeman, B.L. Feringa, D.B. Janssen, J.
Org. Chem. 74 (2009) 9152–9157.
[29] F. Hollywood, H. Suschitzky, Synthesis (1982) 662–665.
[30] E. Forró, J. Chromatogr. A 1216 (2009) 1025–1029.
[31] G. Tasnádi, E. Forró, F. Fülöp, Tetrahedron: Asymmetry 18 (2007) 2841–2844.
[32] R. Vaidyanathan, L. Hesmondhalgh, S. Hu, Org. Process Res. Dev. 11 (2007)
903–906.
[33] S.P. Allwein, R.C. Roemmele, J.J. Haley Jr., D.R. Mowrey, D.E. Petrillo, J.J. Reif, D.E.
Gingrich, R.P. Bakale, Org. Proc. Res. Dev. 16 (2012) 148–155.