Acylation of β-Amino Esters and Hydrolysis of β-Amido Esters: Candida antarctica Lipase A as a Chemoselective Deprotection Catalyst

Article abstract of DOI:10.1002/cctc.201501381

N-Acylation by lipase A from Candida antarctica (CAL-A) in ethyl butanoate was applied to the kinetic resolution of tert-butyl esters of 3-amino-3-phenylpropanoic acid (E>100), 3-amino-4-methylpentanoic acid (E>100) and 3-aminobutanoic acid (E=60) on 1.0-2.0 m scale. With the N-acylated resolution products, the exceptional ability of CAL-A to hydrolyse amides and bulky tert-butyl esters was then studied. In all N-acylated tert-butyl esters, chemoselectivity favoured the amide bond cleavage. The tert-butyl ester bond was left intact with 3-amino-3-phenylpropanoate, whereas with 3-amino-4-methylpentanoate and 3-aminobutanoate the CAL-A-catalysed hydrolysis of tert-butyl ester followed the amide hydrolysis.

Full text of DOI:10.1002/cctc.201501381

DOI: 10.1002/cctc.201501381
Full Papers
Acylation of b-Amino Esters and Hydrolysis of b-Amido
Esters: Candida antarctica Lipase A as a Chemoselective
Deprotection Catalyst
Harri Mäenpää, Liisa T. Kanerva, and Arto Liljeblad*[a]
N-Acylation by lipase A from Candida antarctica (CAL-A) in
ethyl butanoate was applied to the kinetic resolution of tert-
butyl esters of 3-amino-3-phenylpropanoic acid (E>100), 3-
amino-4-methylpentanoic acid (E>100) and 3-aminobutanoic
acid (E=60) on 1.0–2.0m scale. With the N-acylated resolution
products, the exceptional ability of CAL-A to hydrolyse amides
and bulky tert-butyl esters was then studied. In all N-acylated
tert-butyl esters, chemoselectivity favoured the amide bond
cleavage. The tert-butyl ester bond was left intact with 3-
amino-3-phenylpropanoate, whereas with 3-amino-4-methyl-
pentanoate and 3-aminobutanoate the CAL-A-catalysed hy-
drolysis of tert-butyl ester followed the amide hydrolysis.
Introduction
The serine hydrolase superfamily includes enzymes, such as li-
pases (EC 3.1.1.3) and endopeptidases (EC 3.4.21.–), which are
pivotal for life. The natural function of lipases is in digestion
where they mainly hydrolyse ester bonds of triacylglycerols.
Endopeptidases, in turn, typically hydrolyse amide bonds in
various physiological processes like in cell signalling and de-
struction, virulence and immune response, in addition to their
function in digestion.[1] Apart from the different functions of
serine hydrolase enzymes, their catalytic triad, consisting of
Ser, His and Asp (or Glu) residues, provides them with the po-
tential to catalyse reactions by the formation and cleavage of
both ester and amide bonds through similar reaction mecha-
nisms. However, only restricted number of lipases (lipases from
Candida rugosa,[2a] Aspergillus niger,[2b] Rhizomucor miehei and
Pseudomonas stutzeri,[2c] Pseudomonas aerigunosa,[2d] and lipa-
ses A (CAL-A)[2e,f] and B (CAL-B)[2f–n] from Candida antarctica)
have been reported to cleave amide bonds, although usually
very slowly. The mechanistic difference between the two
enzyme types was previously proposed to reside in the transi-
tion-state-stabilising hydrogen bond at the amide hydrogen of
endopeptidases, which is lacking in lipases.[3]
Scheme 1) and the other enantiomer as the unreacted b-
amino ester enantiomer 1. In addition to N-acylations, the ki-
netic resolution of racemic b-lactams by enantioselective ring
hydrolysis and the hydrolysis of b-amino esters are well report-
ed in the presence of lipases.[6] As a mechanistic difference, the
b-amino acid derivative binds at different pockets of the
enzyme in N-acylation of a b-amino ester and hydrolysis of
a b-lactam or a b-amino ester. Thus, for lipase-catalysed N-acy-
lation, an achiral acyl donor (for instance ethyl butanoate) is
bound at the acyl binding site forming an acyl–enzyme inter-
mediate with the serine hydroxyl of the enzyme, and the b-
amino ester as an amine nucleophile reacts with the intermedi-
ate in the second mechanistic step. For the hydrolysis of a b-
lactam or a b-amino ester, these molecules themselves are
bound at the acyl binding site as acyl donors, and water acts
as a nucleophile in the second step.
We have long been interested in the exceptional properties
of CAL-A among lipases. CAL-A is a chemoselective N-acylation
catalyst, for instance, for the acylation of ethyl 3-aminobuta-
noate with carboxylic acid esters, whereas many other lipases
tend to form multiple products through both N-acylation and
interesterification.[7] As CAL-A has already shown interesting
amidase activity,[2e,f] the present paper focuses on attaining
a deeper insight into the utility of CAL-A on this activity. CAL-A
was previously shown to contain two active forms,[2e] the
native enzyme and the one possibly lacking the active site
flap[8] that usually restricts access to the active site. Both CAL-A
forms were shown to catalyse the amide hydrolysis of ethyl 3-
butanamidobutanoate.[2e] In addition, CAL-A-CLEA preparation
was previously used for deprotection of a number of racemic
arylethanamides where amide hydrolyses were completed in
4–6 days.[2f]
Enantiomers of b-amino acids are important structural units
of peptides, peptidomimetics and many biologically active nat-
ural products.[4] Among biocatalytic methods used for their
preparation, lipase-catalysed kinetic resolution of racemic b-
amino esters by N-acylation in organic solvents has been thor-
oughly studied and reviewed.[5] If successful, the method pro-
duces one enantiomer as an N-acyl ester 2 (protocol I,
[a] H. Mäenpää, Prof. L. T. Kanerva, Dr. A. Liljeblad
Institute of Biomedicine/Department of Pharmacology
Drug Development and Therapeutics
University of Turku
In the present work, we focus on the utility of CAL-A as
a commercial CAT#NZL-101-IMB preparation (herein CAL-A) in
the kinetic resolution of rac-1a–c (R1 =CH2CF3 or Et; R2 =tBu)
through N-acylation to learn about concentration effects on re-
FI-20014 (Finland)
E-mail: artlilje@utu.fi
The ORCID identification number(s) for the author(s) of this article can
be found under http://dx.doi.org/10.1002/cctc.201501381.
ChemCatChem 2016, 8, 1226 – 1232
1226
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Scheme 1. N-Acylation of b-amino esters 1a–c and hydrolysis of the amido esters 2a–c using CAL-A.
activity and enantioselectivity (Scheme 1, protocol I). As the ki-
netic resolution produces the reactive enantiomer as an amide
product 2, another focus has been on the study of the chemo-
and enantioselectivity of the CAL-A-catalysed hydrolysis of the
amido esters rac-2a–c (R2 =tBu) and the possibility to trans-
form the enzymatic N-acylation products (R or S)-2a–c (R2 =
A-catalysed hydrolysis. As our focus has been on the lipase-cat-
alysed chemo- and enantioselective formation and cleavage of
the amide bond rather than the cleavage of the ester bond by
hydrolysis, tert-butyl esters were used as substrates rac-1a–c.
tBu) into the corresponding enantiomers of 1a–c (R2 =tBu; Results and Discussion
protocol II). Therefore, three structurally different tert-butyl
N-Acylation of amino esters rac-1a–c (R2 =tBu)
esters of b3-amino esters rac-1a–c were prepared by using
known methods.[9a–c] Alkyl butanoates (PrCO2R1) were used as
acyl donors in N-acylations because the acyl-binding site of
CAL-A was previously shown to be a shallow binding pocket
favouring accommodation of long unbranched carbon chains
in the acyl part of the acyl donor.[10] Thereafter the b3-butana-
mido esters 2a–c (R2 =tBu) were prepared as racemates or
using CAL-A catalysis, followed by N-deacylation through CAL-
CAL-A is well-known for its ability to catalyse the chemoselec-
tive N-acylation of various b-amino esters in organic sol-
vents.[5,7b,d,11,12] Thus, we previously reported CAL-A adsorbed
on celite as an N-acylation catalyst in the kinetic resolution of
rac-1a–c (Scheme 1, R2 =Et) with 2,2,2-trifluoroethyl butanoate
in diisopropyl ether (DIPE; Table 1, entries 1–3).[7b] These reac-
tions were shown to proceed with relatively moderate enantio-
Table 1. CAL-A-catalysed N-acylation of rac-1a–c at 238C.
Entry
Substrate
c [molLÀ1
]
Enzyme
[mgmLÀ1
Acyl donor
t [h]
ees(1) [%]
eep(2) [%]
c [%]
E
]
[c]
[c]
[c]
[c]
[c]
[c]
[c]
[c]
1
2
3
4
5
6
7
8
1a (R2 =Et)[a]
1b (R2 =Et)[a]
1c (R2 =Et)[a]
1a (R2 =tBu)
1a (R2 =tBu)
1a (R2 =tBu)
1a (R2 =tBu)
1a (R2 =tBu)
1a (R2 =tBu)
1a (R2 =tBu)
1a (R2 =tBu)
1a (R2 =tBu)
1a (R2 =tBu)
1b (R2 =tBu)
1b (R2 =tBu)
1c (R2 =tBu)
1c (R2 =tBu)
0.10
0.10
0.10
0.050
0.10
0.50
1.0
40[b]
40[b]
40[b]
20
20
20
20
20
20
20
40
80
80
80
80
80
80
PrCO2CH2CF3
PrCO2CH2CF3
PrCO2CH2CF3
PrCO2CH2CF3
PrCO2CH2CF3
PrCO2CH2CF3
PrCO2CH2CF3
PrCO2CH2CF3
PrCO2Et[d]
1
2.5
25
0.75
2
2
98
99
–
90
87
–
52
53
50
51
52
50
51
56
51
48
51
51
50
50
47
54
38
75
106
6
>99 (R)
>99 (R)
98 (R)
>99 (R)
>99 (R)
>99 (R)
92 (R)
>99 (R)
>99 (R)
>99 (R)
>99 (R)
88 (R)
96 (S)
93 (S)
98 (S)
94 (S)
79 (S)
98 (S)
>99 (S)
97 (S)
96 (S)
99 (S)
>99 (S)
>99 (S)
84 (R)
93 (R)
>100
>100
>100
>100
43Æ3
>100
>100
>100
>100
>100
>100
>100
53
6
2.0
24
24
24
24
24
24
24
24
24
24
9
0.25
0.50
0.50
1.0
2.0
1.0
2.0
1.0
2.0
10
11
12
13
14
15
16
17
PrCO2Et[d]
PrCO2Et[d]
PrCO2Et[d]
PrCO2Et[d]
PrCO2Et[d]
PrCO2Et[d]
PrCO2Et[d]
99 (S)
56 (S)
PrCO2Et[d]
60Æ1
[a] Ref. [7b]. [b] CAL-A on celite. [c] 2 equiv. 2,2,2-trifluoroethyl butanoate in DIPE. [d] A solvent and an acyl donor.
ChemCatChem 2016, 8, 1226 – 1232
www.chemcatchem.org
1227
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
selectivity, strongly depending on the substrate structure, as Hydrolysis of amido esters rac-2a–c (R2 =tBu)
measured by the values of enantiomeric ratio (E).
Herein, rac-1a [0.050–2.0m, (R2 =tBu)] was subjected to As discussed in the Introduction, lipases can, in principle, hy-
CAL-A-catalysed N-acylation with 2,2,2-trifluoroethyl butanoate drolyse both ester and amide bonds, although natural chemo-
(2 equiv.) in DIPE (Table 1, entries 4–8). Elevation of the sub- selectivity should be in favour of ester hydrolysis. In accord-
strate concentration took place at the expense of the time ance with this, the hydrolysis of ethyl 3-butanamidobutanoate
needed to reach the theoretical 50% conversion of kinetic res- rac-2c (R2 =Et, X=Me, Scheme 1) was previously shown to
olution. On the other hand, the reactions at 0.050–1.0m sub- take place at the ester and amide functional groups with CAL-
strate concentrations in the presence of 20 mgmLÀ1 of CAL-A A in phosphate buffer (0.100m, pH 7.5), yielding amido acid
proceeded with excellent enantioselectivity (E>100), whereas (S)-3c through ester hydrolysis (step C), amino ester (R)-1c
the enantioselectivity dropped at 2.0m concentration (entry 8), through amide hydrolysis (step A) and amino acid (R)-4c
indicating an unfavourable solvent effect caused by high pro- through sequential hydrolysis (steps A+B).[2e] In the present
portions of 2,2,2-trifluoroethyl butanoate or 2,2,2-trifluoroetha- work, rac-2a–c as sterically hindered tert-butyl esters rather
nol produced by the acylation. Therefore, the medium 2,2,2-tri- than as n-alkyl esters were used as substrates for CAL-A. Al-
fluoroethyl butanoate in DIPE was replaced by ethyl butanoate though CAL-A is able to hydrolyse tert-butyl esters,[14] hydroly-
as an acyl donor and a solvent (entries 9–13). N-Acylation sis can be expected to be considerably reduced compared to
turned slow compared to the acylation with alkyl-activated the hydrolysis of the corresponding ethyl esters.
2,2,2-trifluoroethyl butanoate, and elevation on enzyme con-
The effect of pH (pH range 1.0–9.0) on enantioselectivity
tent was necessary with increasing content of rac-1a to reach and conversion for the hydrolysis of rac-2a (R2 =tBu) by CAL-A
50% conversion in 24 h. With the enzyme content of (100 mgmLÀ1
80 mgmLÀ1, even 2.0m substrate concentration gave excellent (0.100m; Table 2). At pH 1.0 the reaction gave rac-3a (R2 =tBu)
enantioselectivity after 24 h (E>100, entry 13, compared to
) was studied in suitable buffer solutions
E=43, entry 8). The change of rac-1a to aliphatic rac-1b and
rac-1c in ethyl butanoate resulted in reduction in reactivity
(entries 14–17), as shown in lower conversions after 24 h at
Table 2. CAL-A-catalysed (100 mgmLÀ1) hydrolysis of rac-2a (0.050m;
R2 =tBu) and formation of (S)-1a (R2 =tBu) in various buffers (0.100m; t=
24 h) at 478C.
2.0m substrate concentrations even with enzyme content
80 mgmLÀ1 (entries 15 and 17). Moreover, rac-1c was not as
good a substrate for CAL-A, as deduced from the moderate E
values 53 and 60 for the kinetic resolution of rac-1c (R2 =tBu;
Entry
pH
ees(2)
[%]
eep(1)
[%]
c
[%]
E
entries 16 and 17) in ethyl butanoate. However, even then the
enantioselectivities were clearly higher than with rac-1c (R2 =
Et) where E=6 was observed with 2,2,2-trifluoroethyl buta-
noate in DIPE (entry 3).
1[a]
2
3
4
5
6
7
8
1.0
2.0
3.0
4.0
5.0
6.6
7.5
9.0
rac
27 (R)
34 (R)
41 (R)
27 (R)
31 (R)
45 (R)
33 (R)
rac
95 (S)
83 (S)
87 (S)
80 (S)
88 (S)
93 (S)
96 (S)
40
22
29
32
25
26
33
26
–
51
15
21
12
21
43
68
For the separation of the enantiomers, one gram of rac-1a
(2.00m), rac-1b (1.00m) and rac-1c (2.00m) as tert-butyl esters
were resolved with CAL-A (80 mgmLÀ1) in ethyl butanoate,
yielding the unreacted 1 as the respective (R)-, (R)- and (S)-
enantiomers and the butanamide 2 as the respective (S)-, (S)-
and (R)-enantiomers as shown in Experimental Section. Nota-
bly, the spatial structures of the asymmetric centres are equiva-
lent, and the observed variations in R and S configurations in
[a] Non-enzymatic hydrolysis into 3a.
the series of resolution products 1a–c and 2a–c stem from the through acid-catalysed ester hydrolysis (Table 2, row 1). In-
CIP rules. The enantiomeric products were all obtained at ee> crease in pH resulted in conversions between 22 and 33 (rows
99%, except (R)-2c for which ee=71% was detected in accord- 2–8), the samples containing traces of amino acid (S)-4a (R2 =
ance with the moderate enantioselectivity (Table 1, entries 16 tBu) after one day. As for the sequential reactions, the E values
and 17).
cannot be considered as a measure of enantioselectivity. On
The absolute configurations are in accordance with the pre- the other hand, ee(R)-2a is lowest at pH 5.0 (row 5) and enantio-
vious studies for the N-acylation of rac-1a–c (R2 =Et) with selectivity is highest at pH 9.0 (row 8). Accordingly, ammonium
2,2,2-trifluoroethyl butanoate where (S)-1a (R2 =Et), (S)-1b acetate buffer (pH 9.0; 0.100m) was chosen for further studies.
(R2 =Et) and (R)-1c (R2 =Et) are the more reactive enantiomers The easy removal of ammonium acetate by evaporation was
in the presence of CAL-A on celite.[7b] The configurations are another reason to use this buffer.
further in accordance with the specific rotations given in litera-
ture as described in the Experimental Section.[13]
The hydrolysis of rac-2a (R2 =tBu; 0.050m) with CAL-A
(100 mgmLÀ1) in ammonium acetate buffer (pH 9.0; 0.100m)
chemoselectively produced (S)-1a, even though traces of 4a
were detected through steps A+B (Figure 1). 3a was not
formed at all. This result is a strong indication about high che-
moselectivity toward the amide hydrolysis of rac-2a. Notably,
ChemCatChem 2016, 8, 1226 – 1232
www.chemcatchem.org
1228
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Figure 1. CAL-A-catalysed (CAT#NZL-101-IMB; 100 mgmLÀ1) hydrolysis of
2
Figure 2. CAL-A-catalysed (CAT#NZL-101-IMB; 100 mgmLÀ1) hydrolysis of
&
*
rac-2a ( , 0.050m; R =tBu) and formation of the products (S)-1a ( ) and
2
!
!
&
(S)-4a ( ) in ammonium acetate buffer (0.100m, pH 9.0) at 478C; ee values
rac-2b ( , 0.050m; R =tBu) and formation of the products (S)-4b ( ) and
in parentheses.
*
(S)-1b ( ) in ammonium acetate buffer (0.100m, pH 9.0); ee values in paren-
theses.
the unbranched butanoyl part of 2a is bound into the acyl
binding pocket of CAL-A for amide hydrolysis, whereas for
ester hydrolysis the acyl part of 2a might be large and too
polar for accommodating in the pocket. In accordance with
this, the fact that amino ester (S)-1a hardly hydrolysed into
amino acid (S)-4a indicates that the phenyl group of the sub-
strate is not accepted into the acyl binding pocket. As an un-
expected drawback, the reaction tended to stop at approxi-
mately 30% conversion after 4 days. If the substrate concentra-
tion was lowered to 0.020m under the otherwise same condi-
tions, the reaction proceeded slightly further (ꢀ35% conver-
sion). If the initial reaction mixture contained both rac-1a (R2 =
tBu; 0.050m) and rac-2a (R2 =tBu; 0.050m), the hydrolysis of
rac-2a (R2 =tBu) ceased (c=13% instead of 20% after 24 h in
Figure 1), proposing product inhibition behind the phenomen-
on. Accordingly, (S)-1a is expected to have high affinity to the
nucleophile binding site in aqueous medium, preventing bind-
ing of (S)-2a.
Figure 3. CAL-A-catalysed (CAT#NZL-101-IMB; 100 mgmLÀ1) hydrolysis of
2
!
&
*
)
rac-2c ( , 0.050m; R =tBu) and formation of the products (R)-4c ( ), 1c (
~
and (R)-3c ( ) in ammonium acetate buffer (0.100m, pH 9.0); ee values in
parentheses.
Applicability of the reaction was studied by subjecting (S)-
2a (0.55 g, 1.89 mmol, ee>99%; 0.050m), obtained from the
N-acylation reaction described above, to hydrolysis in ammoni-
um acetate buffer (0.100m, pH 9.0; 38 mL) at 478C. In accord-
ance with analytical-scale reactions, amino ester (S)-1a (0.16 g;
0.72 mmol; 38%; ee>99%) and the unreacted (S)-2a (0.28 g;
0.96 mmol; ee=98%) were isolated after 3 days.
The hydrolysis of rac-2b (R2 =tBu; 0.050m) by CAL-A
(100 mgmLÀ1) in ammonium acetate buffer (pH 9.0; 0.100m)
chemoselectively yielded the amino ester (S)-1b, which imme-
diately reacted into the amino acid (S)-4b with ee>99%
(Figure 2; Scheme 1, steps A+B). The concentration of (S)-1b
stayed minimal all the time. Accordingly, CAL-A enabled se-
quential deprotection by hydrolysis of both amide and tert-
butyl groups in the formation of (S)-4b (R2 =tBu).
products were detected, albeit amido ester (R)-3c only in low
amount (<5%). (R)-3c did not react further as tested separate-
ly with rac-3c. Enantioselectivity for the hydrolysis of rac-2c
was low as concluded from ee(S)–2c =96% at 80% conversion. If
(R)-2c (0.64 g, 2.80 mmol, ee 71%; 0.050m), obtained from the
gram-scale N-acylation described above, was hydrolysed by
CAL-A (100 mgmLÀ1) in ammonium acetate buffer (0.100m,
pH 9.0; 55.8 mL) at 478C, (R)-4c (0.18 g; 1.8 mmol; 64%; ee
96%) and the unreacted (S)-2c (0.14 g; 0.61 mmol; ee 4%)
were isolated after two days. Formations of 1c and 3 were so
low that they were not able to isolate.
Overall, the obtained results suggest that CAL-A cleaves, al-
though slowly, the amide over the tert-butyl ester of b-amido
esters in highly chemo- and enantioselective fashion under
mild conditions.
Hydrolysis of rac-2c (R2 =tBu; 0.050m) with CAL-A
(100 mgmLÀ1) in ammonium acetate buffer (pH 9.0; 0.100m)
was faster than with the other substrates, evidently leading to
the total hydrolysis of the substrate in the formation of 4c at
reaction times long enough (Figure 3). All possible hydrolysis
ChemCatChem 2016, 8, 1226 – 1232
www.chemcatchem.org
1229
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
flame ionisation detection (Agilent 6850) using VARIAN CP Chirasil-
Dex CB FAST GC (CP99927; 15 m”0.15 mm”0.15 mm) and Varian
WCOT CP Chirasil-Dex CB columns (CP7050215; 25 m”0.25 mm”
0.25 mm). The whole reaction volume (1.00 mL) was used for analy-
sis at each time interval. The enzyme used was carefully washed
with dichloromethane. Water and dichloromethane were evaporat-
ed from the samples, and the products were derivatised into
amido esters by adding acetic anhydride (80.0 mL) and after 10 min
MeOH (320 mL). The derivatisation reaction was let to proceed for
at least 2 h before injecting a sample into GC for analysis. The rela-
tive proportions of the four compounds were then determined
from the peak areas by taking into account the peak responses,
which were determined by making a solution containing all the
four compounds in equal amounts.
Conclusions
We have studied kinetic resolutions based on the N-acylation
of tert-butyl esters of 3-amino-3-phenylpropanoic acid (rac-1a),
3-amino-4-methylpentanoic acid (rac-1b) and 3-aminobutanoic
acid (rac-1c) over the concentration range 0.10–2.0m with
2,2,2-trifluoroethyl butanoate in diisopropyl ether and with
ethyl butanoate as an acyl donor and a solvent. We have
shown that even 2.0m substrate concentrations are possible
for the enantioselective N-acylation of the substrates, produc-
ing butanamido esters 2a–c with (S)-, (S)- and (R)-configura-
tions, respectively.
Kinetic resolutions based on the hydrolysis of tert-butyl
esters of rac-2a–c (0.050m) by CAL-A in ammonium acetate
buffer (0.100m, pH 9.0) were shown to depend strongly on the
substrate structure. The formation of the amido acid 3c was
detected only with 2c, and even then chemoselectivity strong-
ly favoured amide hydrolysis over that of tert-butyl ester. On
the other hand, merely the hydrolysis of rac-2a allowed the
preparation of the amino ester (S)-1a in which the tert-butyl
ester remained intact. Accordingly, hydrolysis of rac-2b gave
the corresponding amino acid (S)-4b (ee>99%) in a highly
enantioselective manner, and hydrolysis of rac-2c enantiomeri-
cally enriched (R)-4c (ee 41%) with 75% yield after 4 days.
Thus, the present results show that CAL-A is able to catalyse
not only the hydrolysis of amide bonds but also the hydrolysis
of tert-butyl esters of b-amino esters. These abilities make CAL-
A unique not only among lipases, but among enzymes in gen-
eral. These unusual reaction types can be further utilised in the
removal of protective groups of this type of substrates. Slow
reaction time may be overcome by enzyme engineering in the
future studies.
In analytical-scale N-acylations, rac-1a–c (0.050–2.0m) was dis-
solved into the solvent followed by addition of acyl donor and
enzyme. Reactions proceeded under shaking (170 rpm) at 238C.
Samples (0.1 mL) were taken to follow the reaction by filtering off
the enzyme and derivatising the free amine with acetic anhydride
(10 mL). The samples were diluted with dichloromethane and ana-
lysed by GC.
In the synthesis of the racemic starting materials, thin-layer chro-
matography (TLC) was performed by using Merck Kieselgel 60F254
sheets. Spots were visualised by treatment with 5% ethanolic
phosphomolybdic acid solution and heating.
Determination of absolute configurations
(R)-1a (R2 =tBu): CAL-A-catalysed N-acylation of rac-1a yielded (R)-
25
1a (ee>99%) with [a]D +19.6 (c=1.00, CHCl3). Literature value
for (S)-1a is À21.0 (c=1.00, 208C, CHCl3)[13a] and +20.0 (c=0.71,
208C, CHCl3)[13b] for (R)-1a (ee 99%).
(R)-1b (R2 =tBu): CAL-A-catalysed N-acylation of rac-1b yielded (R)-
22
1b (ee>99%) with [a]D +23.4 (c=0.46, CHCl3). Literature value
for (S)-1b is À26.4 (c=0.67, 198C, CHCl3)[13c] and +19.1 (c=0.68,
258C, CHCl3)[13b] for (R)-1b.
(S)-1c (R2 =tBu): Hydrolysis of the N-acylation product (R)-2c yield-
Experimental Section
22
ed amino acid (R)-4c (ee 96%) with [a]D À35.9 (c=0.50, H2O). Lit-
Materials and Methods
erature value for (S)-4c is +32.0 (c=0.6, 258C, H2O).[13d] Thus, the
N-acylation products were (S)-1c and (R)-2c.
Reagents were purchased from Acros, Aldrich, Fluka, J. T. Baker,
Lancaster or Merck and used as received. Lipase A from C. antarcti-
ca (NZL-101-IMB) immobilised on a resin was a product of Biocata-
lytics (now Codexis). Lipase B from C. antarctica (CAL-B, Novozym
435) was obtained from Novozymes.
Synthesis of the racemic starting materials
b-Amino acid rac-4c was commercially available. rac-4a and -4b
were prepared according to the literature, that is, ammonium ace-
tate, malonic acid and an aldehyde (benzaldehyde or isobutyralde-
hyde, respectively) were used as starting materials in the formation
of b-amino acids.[9a] rac-1a (R2 =Et) was further prepared from the
corresponding b-amino acid using SOCl2/EtOH, and rac-1a–c (R2 =
tBu) using tBuOAc/HClO4.[9b,c] The N-acyl group was attached to
amino esters rac-1a–c from the corresponding acyl chloride or an-
hydride to yield rac-2a–c.
1
The H and 13C NMR spectra were recorded on a Bruker spectrome-
ter operating at 500 MHz with tetramethylsilane as an internal
standard. High-resolution mass spectrometry was performed in
ESI+ mode with Bruker Avance microOTOF-Q quadrupole-TOF
spectrometer. Melting points were recorded with a Gallenkamp ap-
paratus. The optical rotations were determined with PerkinElmer
341 polarimeter, and [a]D values were given in units of
10À1 degcmÀ2 gÀ1. The determination of E for the N-acylation reac-
tions was based on the equation E=ln[(1Àc)(1ÀeeS)]/ln[(1Àc)(1+
eeS)] with c=eeS/(eeS +eeP) using linear regression (E as the slope
of the line ln[(1Àc)(1ÀeeS)] versus ln[(1Àc)(1+eeS)]).[15] Here eeS
and eeP denote enantiomeric excesses of the substrate and the
product, respectively.
rac-4a
(3-amino-3-phenylpropanoic
acid):
Benzaldehyde
(90.0 mmol, 9.55 g, 9.15 mL), malonic acid (90.9 mmol, 9.43 g) and
ammonium acetate (121 mmol, 9.30 g) were dissolved in ethanol
(150 mL), and the mixture was kept at for 5 h. rac-4a was filtered,
washed with ethanol and separated as white crystals (38.7 mmol,
6.40 g, 43%; Rf =0.7 (80% EtOH, 20% H2O); m.p. 222–2238C).
1H NMR (500 MHz, D2O): d=2.71 (dd, 1H, J=16 Hz, 6.5 Hz,
ArCHCH2), 2.81 (dd, 1H, J=16 Hz, 8.5 Hz, ArCHCH2), 4.54 (m, 1H,
Analytical-scale hydrolyses of rac-2a–c into 1a–c, 3a–c and 4a–
c took place in New Brunswick Scientific G24-incubator (478C). The
reaction samples were analysed by gas chromatography with
ChemCatChem 2016, 8, 1226 – 1232
www.chemcatchem.org
1230
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
ArCH), 4.70 (s, 2H, ArCHNH2), 4.70 (s, 1H, COOH), 7.34–7.40 ppm
(m, 5H, C6H5); 13C NMR (125 MHz, D2O) d=40.41, 52.69, 126.87,
129.21, 129.24, 135.96, 177.19 ppm; MS–ESI+: m/z 165.0791,
C9H11NO2 requires 165.0789.
silica gel (90% petroleum ether, 10% acetone).The product rac-2a
(R1 =tBu; 2.2 mmol, 0.65 g, 99%) was a yellow, crystalline sub-
stance [Rf =0.1 (90% petroleum ether, 10% acetone); m.p. 69–
708C]. 1H NMR (500 MHz, CDCl3): d=0.98 (t, 3H, J=7.5 Hz,
NHCOCH2CH2CH3), 1.33 (s, 9H, COOC(CH3)3), 1.70 (dt, 2H, J=
22.5 Hz, 7.5 Hz, NHCOCH2CH2), 2.23 (t, 2H, J=8 Hz, NHCOCH2), 2.75
(dd, 2H, J=16 Hz, 6 Hz, ArCHCH2), 2.84 (dd, 2H, J=16 Hz, 6 Hz,
ArCHCH2), 5.43 (m, 1H, ArCH), 6.72 (d, 1H, J=7.5, NH), 7.26–
7.35 ppm (m, 5H, C6H5); 13C NMR (125 MHz, CDCl3) d=13.78, 19.17,
27.90, 38.74, 41.17, 49.54, 81.40, 126.31, 127.44, 128.55, 140.76,
170.72, 172.27 ppm; MS-ESI+: m/z 291.1829, C17H25NO3 requires
291.1834.
rac-4b (3-amino-4-methylpentanoic acid): rac-4b was prepared as
rac-4a, except that after evaporation of ethanol, the crude product
was re-dissolved into ethanol (50.0 mL) followed by refluxing. Addi-
tion of acetone (50.0 mL) precipitated the product. The mixture
was let cool to room temperature, and the product was filtered
and washed with acetone. rac-4b was obtained as white crystals
(48.8 mmol, 6.40 mg, 54%; Rf =0.5 (90% EtOH, 10% H2O), m.p.
202–2038C). 1H NMR (500 MHz, D2O): d=0.89 (d, 3H, J=7 Hz,
(CH3)2CH), 0.91 (d, 3H, J=7 Hz, (CH3)2CH), 1.85 (m, 1H, (CH3)2CH),
2.31 (dd, 1H, J=16 Hz, 9.5 Hz, CHNH2CH2), 2.48 (dd, 1H, J=16 Hz,
4 Hz, (CHNH2CH2), 3.24 (m, 1H, (CH3)2CHCH), 4.70 (s, 1H, COOH),
4,70 ppm (s, 2H, NH2); 13C NMR (125 MHz, D2O) d=17.21, 17.38,
29.98, 35.92, 54.81, 178.51 ppm; MS–ESI+: m/z 131.0945, C6H13NO2
requires 131.0946.
rac-2b (R2 =tBu; 3-butanamido-4-methylpentanoic acid tert-butyl
ester): rac-1b (R2 =tBu; 2.67 mmol, 0.50 g) was used as the starting
material, and the synthesis took place as with rac-2a (R2 =tBu),
except that the eluent for column chromatography was 90% di-
chloromethane, 10% methanol. The product rac-2b (R2 =tBu;
2.64 mmol, 0.68 g, 99%) was a white crystalline substance [Rf =0.9
(90% dichloromethane, 10% methanol), m.p. 46–478C]. 1H NMR
(500 MHz, CDCl3): d=0.93–1.01 (m, 3H, NHCOCH2CH2CH3), 0.93–
1.01 (m, 6H, (CH3)2CH), 1.46 (s, 9H, COOC(CH3)3), 1.69 (m, 2H,
NHCOCH2CH2), 1.84 (m, 1H, (CH3)2CH), 2.20 (m, 2H, NHCOCH2), 2.45
(m, 2H, CHNHCH2), 4.06 (m, 1H, (CH3)2CHCHNH), 6.21 ppm (s, 1H,
NH). 13C NMR (125 MHz, CDCl3) d=13.78, 19.07, 19.28, 19.30, 28.03,
31.60, 37.73, 38.94, 51.51, 171.65, 172.44 ppm; MS-ESI+: m/z
257.1986, C14H27NO3 requires 257.1991.
rac-1a (R2 =tBu; 3-amino-3-phenylpropanoic acid tert-butyl ester):
tert-Butyl acetate (80.0 mL) was added on rac-4a (18.2 mmol,
3.00 g). After mixing for 5 min, perchloric acid (27.3 mmol, 2.74 g,
2.35 mL) was added, and the reaction mixture was stirred for 15 h
at RT. The product was extracted by water (4”50 mL) and 0.5m
HCl (50.0 mL). The aqueous phases were combined, and the pH
was adjusted to 8.00 with 2m NaOH. The product was extracted
with dichloromethane (4”50.0 mL). The combined phases were
washed with water (50 mL) and saturated NaCl (50 mL). The organ-
ic phase was dried with sodium sulfate. The product (10.7 mmol,
2.36 mg, 59%) was obtained as slightly yellow oily liquid after fil-
rac-2c (R2 =tBu; 3-butanamidobutanoic acid tert-butyl ester): rac-
1c (R2 =tBu; 3.1 mmol, 0.50 g) was used as the starting material,
and the synthesis took place as with rac-2a (R2 =tBu). The product
rac-2c (R2 =tBu; 3.1 mmol, 0.72 g, 99%) was a greyish liquid [Rf =
0.2 (90% petroleum ether, 10% acetone)]. 1H NMR (500 MHz,
CDCl3): d=0.95 (t, 3H, J=7.5 Hz, NHCOCH2CH2CH3), 1.22 (d, 3H,
J=7.0 Hz, CH3CHNH2), 1.47 (s, 9H, COOC(CH3)3), 1.66 (dt, 2H, J=
22.5, 7.5 Hz, NHCOCH2CH2), 2.15 (t, 2H, J=7.5 Hz, NHCOCH2), 2.43
(m, 2H, CH3CHNH2CH2), 4.34 (m, 1H, CH3CHNH2), 6.29 ppm (s, 1H,
NH); 13C NMR (125 MHz, CDCl3) d=13.69, 19.17, 20.00, 28.08, 38.78,
41.15, 42.07, 81.17, 171.24, 172.26 ppm; MS-ESI+: m/z 229.1673,
C12H23NO3 requires 229.1678.
1
tration and evaporation [Rf =0,5 (90% DCM, 10% MeOH)]. H NMR
(500 MHz, CDCl3): d=1.42 (s, 9H, COOC(CH3)3), 2.70 (m, 2H,
ArCHCH2), 3.26 (s, 2H, NH2), 4.42 (m, 1H, ArCH), 7.08–7.26 ppm (m,
5H, C6H5); 13C NMR (125 MHz, CDCl3) d=28.05, 44.27, 52.72, 80.98,
126.57, 127.66, 128.64, 143.04, 170.89 ppm; MS–ESI+: m/z
199.1567, C13H20NO2 requires 199.1596.
rac-1b (R2 =tBu; 3-amino-4-methylpentanoic acid tert-butyl ester)
was prepared similarly to rac-1a (R2 =tBu). The obtained product
(11.2 mmol, 2.09 g, 49%) was a bright oily liquid [Rf =0.5 (90%
1
DCM, 10% MeOH]. H NMR (500 MHz, CDCl3): d=0.94 (q, 6H, J=
3,5 Hz, (CH3)2CH), 1.47 (s, 9H, COOC(CH3)3), 1.65 (m, 1H, (CH3)2CH),
1.79 (s, 2H, NH2), 2,18 (dd, 1H, J=15.5 Hz, 10 Hz, CH3CHNH2CH2),
2.40 (dd, 1H, J=15.5 Hz, 4 Hz, CH3CHNH2CH2), 3.01 ppm (m, 1H,
(CH3)2CHCHNH2); 13C NMR (125 MHz, CDCl3) d=17.83, 18.83, 28.14,
33.21, 40.76, 53.66, 80.51, 172.41 ppm; MS–ESI+: m/z 187.1573,
C10H21NO2 requires 187.1572.
Preparative-scale N-acylation and hydrolysis
rac-1a (R2 =tBu; 1.00 g, 4.52 mmol) was dissolved into ethyl buta-
noate (2.26 mL) to obtain a 2.0m solution. Addition of CAL-A-prep-
aration (CAT#NZL-101-IMB; 80.0 mgmLÀ1) started the reaction. The
reaction mixture was shaken at RT and stopped at 50% conversion
rac-1c (R2 =tBu; 3-aminobutanoic acid tert-butyl ester) was pre-
pared similarly to rac-1a (R2 =tBu). The obtained product
(14.8 mmol, 2.36 mg, 51%) was a yellowish oily liquid [Rf =0.4
after 24 h. (S)-2a (0.62 g, 94%, ee>99%) and (R)-1a [0.47 g, 94%,
25
ee>99%, ½aŠ +19,6 (c=1.00, CHCl3)] were separated by column
D
chromatography using dichloromethane/methanol (9:1) as an
eluent.
1
(90% dichloromethane, 10% MeOH]. H NMR (500 MHz, CDCl3): d=
1.12 (d, 3H, J=6.5 Hz, CH3CHNH2), 1.46 (s, 9H, COOC(CH3)3), 1.83 (s,
2H, CH3CHNH2), 2.23 (dd, 1H, J=15.5 Hz, 8.0 Hz, CHNH2CH2), 2.33
(dd, 1H, J=15.5 Hz, 4.5 Hz, (CHNH2CH2), 3.34 ppm (m, 1H,
CH3CHNH2); 13C NMR (125 MHz, CDCl3) d=23.30, 28.13, 44.20,
45.39, 80.51, 171.77 ppm; MS–ESI+: m/z 159.1253, C8H17NO2 re-
quires 159.1259.
(S)-2a (R2 =tBu; 0.55 g, 1.89 mmol, ee>99%) was further hydro-
lysed by CAL-A (CAT#NZL-101-IMB; 100 mgmLÀ1) in 0.050m ammo-
nium acetate buffer (0.100m, pH 9.0, 38.0 mL) at 478C. The reaction
was stopped after three days. Purification by column chromatogra-
phy using dichloromethane/methanol (95:5) as an eluent yielded
22
(S)-2a (0.28 g, ee=98%, ½aŠ À48.2 (c=0.55, CHCl3)) and (S)-1a
D
25
rac-2a (R2 =tBu; 3-butanamido-3-phenylpropanoic acid tert-butyl
ester): The reaction mixture of rac-1a (R2 =tBu) (2.3 mmol, 0.50 g),
triethylamine (6.78 mmol, 0.45 g, 0.95 mL) and butanoic anhydride
(2.94 mmol, 0.45 g, 0.47 mL) was stirred for 60 min. The reaction
was stopped by adding methanol (30.0 mL) and mixing for 60 min.
The crude product was purified by column chromatography on
[0.16 g, 38%, ee>99%, ½aŠ À20.0 (c=1.00, CHCl3)].
D
rac-1b (R2 =tBu; 1.00 g, 5.34 mmol) was dissolved into ethyl buta-
noate (5.34 mL) to obtain a 1.0m solution. Addition of CAL-A-prep-
aration (CAT#NZL-101-IMB; 80.0 mgmLÀ1) started the reaction. The
reaction mixture was shaken at RT and stopped at 50% conversion
after 24 h. (S)-2b (0.64 g, 93%, ee>99%) and (R)-1b [0.45 g, 90%,
ChemCatChem 2016, 8, 1226 – 1232
www.chemcatchem.org
1231
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
22
D
geeva, V. V. Mozhaev, J. O. Rich, Y. L. Khmelnitsky, Biotechnol. Lett. 2000,
22, 1419–1422; k) D. R. Duarte, E. L. Castillo, E. Bµrzana, A. López-Mun-
guia, Biotechnol. Lett. 2000, 22, 1811–1814; l) A. Torres-Gavilµn, J. Esca-
lante, I. Regla, A. López-Munguía, E. Castillo, Tetrahedron: Asymmetry
2007, 18, 2621–2624; m) H. Ismail, R. M. Lau, F. van Rantwijk, R. A. Shel-
don, Adv. Synth. Catal. 2008, 350, 1511–1516; n) A. Radu, M. E. Moisa˘,
M. I. Tos¸a, N. Dima, V. Zaharia, F. D. Irimie, J. Mol. Catal. C 2014, 107,
114–119.
ee>99%, ½aŠ +23.4 (c=0.46, CHCl3)] were separated by column
chromatography using dichloromethane/methanol (9:1) as an
eluent.
rac-1c (R2 =tBu; 1.00 g, 6.28 mmol) was dissolved into ethyl buta-
noate (6.28 mL) to obtain a 2.0m solution. Addition of CAL-A-prep-
aration (CAT#NZL-101-IMB; 80.0 mgmLÀ1) started the reaction. The
reaction mixture was shaken in an incubator (t=478C) and
[3] a) R. Kourist, S. Bartsch, L. Fransson, K. Hult, U. T. Bornscheuer, ChemBio-
Chem 2008, 9, 67–69; b) P.-O. SyrØn, K. Hult, ChemCatChem 2011, 3,
853–860.
[4] D. Seebach, A. K. Beck, D. J. Bierbaum, Chem. Biodiversity 2004, 1, 1111–
1239.
stopped at 58% conversion after 29 h. (R)-2c (0.74 g, 88%, ee 71%)
22
D
and (S)-1c as butanoate salt (0.20 g, 48%, ee>99%, ½aŠ +15.3
(c=0.50, CHCl3)) were separated by column chromatography using
dichloromethane/methanol (9:1) as an eluent.
[5] A. Liljeblad, L. T. Kanerva, Tetrahedron 2006, 62, 5831–5854.
[6] E. Forró, F. Fülçp, Curr. Med. Chem. 2012, 19, 6178–6187.
[7] a) S. Gedey, A. Liljeblad, F. Fülçp, L. T. Kanerva, Tetrahedron: Asymmetry
1999, 10, 2573–2581; b) S. Gedey, A. Liljeblad, L. Lµzµr, F. Fülçp, L. T. Ka-
nerva, Tetrahedron: Asymmetry 2001, 12, 105–110; c) X.-G. Li, L. T. Kaner-
va, Org. Lett. 2006, 8, 5593–5596.
[8] D. J. Ericsson, A. Kasrayan, P. Johansson, T. Bergfors, A. G. Sandstrçm, J.-
E. Bäckvall, S. L. Mowbray, J. Mol. Biol. 2008, 376, 109–119.
[9] a) L. Lµzµr, T. Martinek, G. Bernµth, F. Fülçp, Synth. Commun. 1998, 28,
219; b) J. Hu, M. J. Miller, J. Am. Chem. Soc. 1997, 119, 3462; c) H. Chen,
Y. Feng, Z. Xu, T. Ye, Tetrahedron 2005, 61, 11132.
(R)-2c (R2 =tBu; 0.64 g, 2.80 mmol, ee 71%) was further hydrolysed
by CAL-A (CAT#NZL-101-IMB; 100 mgmLÀ1) in 0.050m ammonium
acetate buffer (0.100m, pH 9.0, 55.8 mL) at 478C. The reaction was
stopped after two days. Purification twice by column chromatogra-
phy using dichloromethane:methanol (9:1) and ethanol:water (9:1)
22
D
as eluents yielded (R)-4c [0.18 g, 1.8 mmol, 64%, ee 96%, ½aŠ
22
D
À35.9 (c=0.50, H2O)] and (S)-2c [0.14 g, 0.61 mmol, ee=4%, ½aŠ
À1.45 (c=0.50, CHCl3)].
The spectral data of the isolated products in the enzymatic reac-
tions were identical to those of the racemic starting materials.
[10] a) A. G. Sandstrçm, Y. Wikmark, K. Engstrçm, J. NyhlØn, J.-E. Bäckvall,
Proc. Natl. Acad. Sci. USA 2012, 109, 78–83; b) M. P. Frushicheva, A. War-
shel, ChemBioChem 2012, 13, 215–223.
Keywords: acylation · chemoselectivity · enzyme catalysis ·
kinetic resolution · hydrolysis
[11] M. Solymµr, A. Liljeblad, L. Lµzµr, F. Fülçp, L. T. Kanerva, Tetrahedron:
Asymmetry 2002, 13, 1923–1928.
[12] X.-G. Li, L. T. Kanerva, Tetrahedron: Asymmetry 2005, 16, 1709–1714.
[13] a) S. G. Davies, N. M. Garrido, D. Kruchinin, O. Ichihara, L. J. Kotchie, P. D.
Price, A. J. P. Mortimer, A. J. Russell, A. D. Smith, Tetrahedron: Asymmetry
2006, 17, 1793–1811; b) T. Sakai, H. Doi, K. Tomioka, Tetrahedron 2006,
62, 8351–8359; c) M. Node, D. Hashimoto, T. Katoh, S. Ochi, M. Ozeki, T.
Watanabe, T. Kajimoto, Org. Lett. 2008, 10, 2653–2656; d) S. G. Davies,
A. W. Mulvaney, A. J. Russell, A. D. Smith, Tetrahedron: Asymmetry 2007,
18, 1554–1566.
[14] E. Henke, J. Pleiss, U. T. Bornscheuer, Angew. Chem. Int. Ed. 2002, 41,
3211–3213; Angew. Chem. 2002, 114, 3338–3341.
[15] C.-S. Chen, Y. Fujimoto, G. Girdaukas, C. J. Sih, J. Am. Chem. Soc. 1982,
104, 7294–7299.
[1] A. Fersht, Structure and mechanism in protein science, Vol. 3, W. H. Free-
man, New York, 1998, pp. 472–493.
[2] a) V. Gotor, R. Brieva, C. Gonzalez, F. Rebolledo, Tetrahedron 1991, 47,
9207–9214; b) B. Wang, Y. Liu, D. Zhang, Y. Feng, J. Li, Tetrahedron:
Asymmetry 2012, 23, 1338–1342; c) C. Simons, J. G. E. van Leeuwen, R.
Stemmer, I. W. C. E. Arends, T. Maschmeyer, R. A. Sheldon, U. Hanefeld, J.
Mol. Catal. B 2008, 54, 67–71; d) R. Fujii, Y. Nakagawa, J. Hiratake, A.
Sogabe, K. Sakata, Protein Eng. Des. Sel. 2005, 18, 93–101; e) A. Liljeblad,
P. Kallio, M. Vainio, J. Niemi, L. T. Kanerva, Org. Biomol. Chem. 2010, 8,
886–895; f) J. Brem, L.-C. Bencze, A. Liljeblad, M. C. Turcu, C. Paizs, F.-D.
Irimie, L. T. Kanerva, Eur. J. Org. Chem. 2012, 3288–3294; g) H. Smidt, A.
Fischer, P. Fischer, R. D. Schmidt, U. Stelzer (Bayer AG, Leverkusen)
DE 19507217, 1996; h) H. Smidt, A. Fischer, P. Fischer, R. D. Schmidt, Bio-
technol. Tech. 1996, 10, 335–338; i) T. Wagegg, M. M. Enzelberger, U. T.
Bornscheuer, R. D. Schmid, J. Biotechnol. 1998, 61, 75–78; j) M. V. Ser-
Received: December 18, 2015
Published online on February 18, 2016
ChemCatChem 2016, 8, 1226 – 1232
www.chemcatchem.org
1232
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim