Fully enzymatic resolution of chiral amines: Acylation and deacylation in the presence of Candida antarctica lipase B

Article abstract of DOI:10.1002/adsc.200800091

A fully enzymatic methodology for the resolution of chiral amines has been demonstrated. Candida antarctica lipase B (CaLB)-catalyzed acylation with N-methyl-and N-phenylglycine, as well as analogues having the general formula R¹-X-CH₂CO₂R² (R¹ = Me, Ph; X = O, S) afforded the corresponding enantioenriched amides, which were subsequently enzymatically hydrolyzed. Surprisingly, CaLB also proved to be the catalyst of choice for this latter step. The heteroatom in the acyl donor profoundly influences both the enzymatic acylation and deacylation; the O-substituted reagents performed best with regard to enantioselectivity as well as reaction rate in synthesis and hydrolysis.

Full text of DOI:10.1002/adsc.200800091

FULL PAPERS
DOI: 10.1002/adsc.200800091
Fully Enzymatic Resolution of Chiral Amines: Acylation and
Deacylation in the Presence of Candida antarctica Lipase B
Hilda Ismail,a,b Rute Madeira Lau,a,c Fred van Rantwijk,a and Roger A. Sheldona,*
a
Laboratory of Biocatalysis and Organic Chemistry, Department of Biotechnology, Delft University of Technology,
Julianalaan 136, 2628 BL Delft, The Netherlands
Fax : (+31)-15-278-1415; e-mail: r.a.sheldon@tudelft.nl
Present address: Department of Pharmacochemistry, Faculty of Pharmacy, Gadjah Mada University, Sekip Utara,
b
Yogyakarta, Indonesia
c
Present address: DSM Anti-Infectives, Alexander Fleminglaan 1, 2613 AX, Delft, The Netherlands
Received: February12, 2008; Revised: May9, 2008; Published online: June 23, 2008
Abstract: A fullyenzymatic methodologyfor the res-
this latter step. The heteroatom in the acyl donor
olution of chiral amines has been demonstrated. profoundly influences both the enzymatic acylation
Candida antarctica lipase B (CaLB)-catalyzed acyla- and deacylation; the O-substituted reagents per-
tion with N-methyl- and N-phenylglycine, as well as formed best with regard to enantioselectivityas well
1
À À
analogues having the general formula
R
X
as reaction rate in synthesis and hydrolysis.
CH2CO2R2 (R1 =Me, Ph; X=O, S) afforded the cor-
responding enantioenriched amides, which were sub- Keywords: amines; Candida antarctica lipase B;
sequently enzymatically hydrolyzed. Surprisingly, enantioselective acylation; enzymatic deacylation;
CaLB also proved to be the catalyst of choice for enzyme catalysis
Introduction
Entirelyenzmyatic amine resolution procedures
employing acyl reagents based on (R)-phenylglycine
The lipase-catalyzed enantioselective acylation of derivatives have indeed been reported[5,6] but the
chiral amines is at the basis of a highlyefficient reso-
methodologysuffers from the drawback that the recy-
lution process that is now performed at a multi-thou- cling of the chiral acyl donor is by no means trivial.
sand ton per year scale.[1,2] The subsequent (chemical) We surmised that a non-chiral amino acid would be
deacylation step is much less developed and requires easier to recycle[7] and have now investigated the suit-
strong base at elevated temperatures,[2,3] because of abilityof resolution reagents based on N-methyl- and
the thermodynamic stability of the amide bond. Such
harsh reaction conditions are not compatible with
sensitive functional groups; moreover, theygive rise
to an undesirable salt waste stream. A completelyen-
zymatic procedure, employing enzymes in both acyla-
tion and deacylation steps (see Figure 1) would be
preferable. Enzymatic deacylation was, until recently,
not practical due to the low reactivityand modest sol-
ubilityof amides and a lack of suitable amidases.
Amide hydrolysis in the presence of Candida antarcti-
ca lipase B (CaLB), for example, was slow even with
large amounts of enzyme.[4] We recentlyreported that
hydrolytic cleavage of the (R)-phenylglycinyl group
readilytakes place in the presence of the penicillin
acylase from Alcaligenes faecalis.[5] The propensityof
the phenylglycinyl group to hydrolytic cleavage has
been ascribed to the zwitterionic nature of the phe-
nylglycine hydrolysis product, which renders the hy-
drolysis energetically favourable.
Figure 1. General scheme of fullyenzymatic amine resolu-
tion.
Adv. Synth. Catal. 2008, 350, 1511 – 1516
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1511

Hilda Ismail et al.
FULL PAPERS
N-phenylglycine. For comparison, we have also stud-
ied acyl donors bearing ether, thioether and methyl-
ene moieties in the b-position (see Figure 2).
^
Figure 3. Effects of the acyl donor on the acylation of 2; ( )
~
&
*
3a, X=NH; ( ), 3b, X=O); ( ), 3c, X=S; ( ), 3d, X=
CH2.
donor.[15] This rate increase has originallybeen attrib-
uted to an inductive effect of the b-oxygen atom,[15]
but Cammenberg et al. recentlyconcluded that it
should rather be ascribed to hydrogen bonding of the
amino group in the nucleophile to the acyl-chain
Figure 2. CaLB-catalyzed amine resolution with acyl re-
agents CH3 X CH2CO2R .
2
À À
The acylation biocatalyst that we employed was oxygen.[16] The lower reactivityof the S-substituted
CaLB, which has an excellent record as amine resolu- acyl donor 3c, compared with 3b, supports the hydro-
tion catalyst[8,9] and readilytolerates non-natural reac-
gen bond hypothesis as S is a weaker H-bond accep-
tants and reaction conditions.[10] The funnel-like active tor than O.[17] It also refutes the electronegativityhy-
[11]
site of CaLB is known to be stericallyrestricted;
hence, it maybe expected that the substituent in the
pothesis, which would predict 3c to be more reactive
than 3b because S is slightlymore electronegative
acyl donor, methyl or phenyl, will affect the enantior- than O.[18]
ecognition.[12,13] Selecting a biocatalyst for the hydrol-
With X=NH (3a) the issue becomes confused,
ysis step proved to be much less straightforward, due however, as the accelerating effect is much less than
[14]
to the strict selectivityof penicillin acylase,
as well that of oxygen or sulfur, although nitrogen is a much
as other amide-hydrolysing enzymes. We will describe stronger H-bond acceptor than oxygen, as judged
how this problem was satisfactorilysolved eventually.
from solvatochromic effects[17] and Kamlet–Taft b pa-
rameters.[19] The electronegativityhypothesis, on the
other hand, would predict 3a to react slower than 3d,
because nitrogen is electron-releasing relative to
carbon.[18] The subject obviouslyrequires further
studybut it would seem, on the basis of the present
results, that the overall effects of b-hetero atoms
Results and Discussion
Amine Acylation
We selected 2-heptylamine (1) and 1-phenylethyl- cannot be attributed to one single cause but rather
amine (2) as suitable models for aliphatic and arylalk- are a combination of H-bond formation, inductive ef-
yl amines. These were subjected to enantioselective fects and, presumably, steric interactions.
acylation by acyl donor reagents with the general for- Good to excellent enantioselectivities were ob-
ACHTREUNG
2
2
À À
mula CH3 X CH2CO2R (R =methyl or ethyl; X= tained with all acyl donors. Amine 2 was resolved
NH, O, S or CH2, 3a–d, see Figure 2) in the presence with an enantiomeric ratio (E)>100 with all acyl
of Novozyme 435 (immobilized CaLB). All reactions donors, while with 1 E ranged from 51 (X=O) to
were performed with strict exclusion of water; zeolite >100 (X=CH2). Remarkably, the enantiomeric ratio
KA and CaA were added to absorb anytraces of
moisture as well as the liberated alcohol.
The nature of the moietyin the b position pro-
foundlyaffected the course of the acylation, as would
decreased in the order X=CH2 >NH>S>O with
this latter amine.
Kinetic resolution using the stericallymore de-
manding acyl donors 6a–d, with the general formula
be expected.[15,16] The acylation rate increased in the C6H5 X CH2CO2R (X=NH, O, S, CH2; R =CH3,
order X=CH2 <NH<S<O, as is observed, for exam- C2H5; see Figure 4) was investigated to assess any
ple, in the resolution of 2 (Figure 3), which reconfirms steric effects of the donor. These acyl reagents were
the accelerating effect of a b-oxygen atom in the acyl much less active in the acylation of 1 and 2 than their
2
2
À À
1512
asc.wiley-vch.de
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 1511 – 1516

FullyEnzymatic Resolution of Chiral Amines
FULL PAPERS
amino-substituted donor 6a was much less active than
the isomeric (R)-phenylglycine esters[6] under compa-
rable conditions, contraryto what would be expected
on account of branching effects.
The enantiomer discrimination by 6 was unsatisfac-
toryin general, with the exception of oxygen-substi-
tuted 6b, which gave excellent results. In conclusion,
so far, the acyl donors of choice with regards to rate
and enantioselectivityare the O-substituted acly
donors 3b and 6b.
Enzymatic Deacylation
Figure 4. CaLB-catalyzed resolution with acyl reagents
C6H5 X CH2CO2R .
Efficient enzymatic deacylation of the enantiomeri-
callyenriched amides 4, 5, 7, and 8 (see Figure 5)
proved to be far from trivial. Hydrolysis in the pres-
2
À À
methyl counterparts, which could be explained by ence of A. faecalis penicillin G acylase, the favoured
steric hindrance and/or their lower hydrogen bond-ac- catalyst for removing the (R)-phenylglycyl group,[6]
cepting capabilitydue to the electron-withdrawing
was quite sluggish (<10% conversion in 4 d), as
phenyl ring.[17] Acyl donor 6b (X=O, see Table 2) was would be expected in view of the enzymeꢁs specificity
an exception as it reacted hardlyslower than 3b. It is for phenylacetic acid derivatives. Other amide hydro-
worth noting here that a phenyl-substituted oxygen is lyzing enzymes (thermolysin, amino acylase I from
a much weaker hydrogen-bond acceptor than a Aspergillus melleus or porcine kidney) also failed to
methyl-substituted one but also much more electron- hydrolyze these amides at an appreciable rate.
withdrawing.[18] Finally, it is remarkable that the
2
Table 1. Kinetic resolution[a] of 1 and 2 with acyl donors CH3 X CH2CO2R .
À À
Amine
Acyl donor
X
Amide
vinit [%][b]
Time [h]
Conv. [%]
eeamine [%]
E
1
2
1
2
1
2
1
2
3a
3a
3b
3b
3c
3c
3d
3d
NH
NH
O
O
S
S
CH2
CH2
4a
5a
4b
5b
4c
5c
4d
5d
100
54
63
128
5
19
24
48
216
216
51
45
50
50
50
48
42
40
97
80
91
99
91
91
76
73
94
>99
51
>99
80
>99
>99
>99
>390
>350
167
141
11
6
[a]
Reaction conditions: amine (5 mmol), acyl donor (3 mmol), Novozym 435 (100 mg) in 1,2-dimethoxyethane (5 mL), zeo-
lite KA and CaA (150 mg each) at 408C.
Initial rate (conv./time) relative to the acylation of 1 by 3a.
[b]
Table 2. Kinetic resolution[a] of 1 and 2 with acyl donors C6H5-X-CH2CO2R2.
Amine
Acyl donor
X
Amide
vinit [%][b]
Time [h]
Conv. [%]
eeamine [%]
E
1
2
1
2
1
2
1
2
6a
6a
6b
6b
6c
6c
6d
6d
NH
NH
O
O
S
S
CH2
CH2
7a
8a
7b
8b
7c
8c
7d
8d
16
12
>120
>120
22
19
3
1
216
118
7
48
51
50
50
48
43
8
63
54
99
98
68
68
n.d.
n.d.
11
12
>99
>99
14
23
n.d.
n.d.
24
200
200
312
312
7
[a]
Reaction conditions: amine (5 mmol), acyl donor (3 mmol), Novozym 435 (100 mg) in 1,2-dimethoxyethane (5 mL), zeo-
lite KA and CaA (150 mg each) at 408C.
Initial rate (conv./time) relative to the acylation of 1 by 3a.
[b]
Adv. Synth. Catal. 2008, 350, 1511 – 1516
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1513

Hilda Ismail et al.
Table 4. Preparative-scale hydrolysis of 5b and 5c.[a]
FULL PAPERS
Amide
Time [h]
Conversion [%]
408C 608C
(R)-5b
(R)-5c
48
72
48
72
100
100
94
87
90
48
67
100
[a]
Reaction conditions: amide (0.3 mmol) in 0.1M phos-
phate buffer pH 7.0 (3 mL), CaLB CLEA (90 mg, 600
LU) at 408C.
Figure 5. Hydrolysis of the enantiomerically pure amides 4,
5, 7 and 8 (full structures are given in Figure 2 and
Figure 4).
the oxygen in 4b and 5b has a much more profound
effect on the hydrolysis than the formation of a zwit-
terionic hydrolysis product from the N-analogues 4a
and 5a (Table 4). Hence, acyl donors with X=O are
the reactants of choice with regards to acylation as
well as deacylation.
Table 3. Deacylation of amides 4, 5, 7 and 8 in the presence
of CaLB.[a]
Amide
X
R1
Time [h]
Conv. [%]
(R)-4a
(R)-5a
(R)-4b
(R)-5b
(R)-4c
(R)-5c
(R)-7a
(R)-8a
(R)-7b
(R)-8b
(R)-7c
(R)-8c
NH
NH
O
O
S
C5H11
C6H5
C5H11
C6H5
C5H11
C6H5
C5H11
C6H5
C5H11
C6H5
C5H11
C6H5
147
120
120
120
120
120
90
120
120
120
120
144
24
14
74
100
36
85
30
3
21
42
50
39
Encouraged bythese results, we next performed
the hydrolysis of 5b and 5c at a synthetically relevant
concentration.
A cross-linked enzyme aggregate
(CLEA) of CaLB[25] was employed as the biocatalyst,
for easyrecovering and recycling. At 60 8C the reac-
tion unexpectedlycame to a standstill at incomplete
conversion. It would seem that thermal deactivation
of the CLEA is significant at 608C, which is presuma-
blycaused bydissociation of the cross-links as CaLB
is intrinsicallystable up to 70 8C. Complete hydrolysis
could be accomplished, in contrast, at 408C. The
rather high catalyst usage is still an obstacle to practi-
cal application, but it is relevant to note here that the
biocatalyst is recyclable. Presumably, improved hy-
drolysis rates could result from appropriate mutagen-
esis.[26]
S
NH
NH
O
O
S
S
[a]
Reaction conditions: amide (5 mg, 20–27 mmol) in 0.1M
phosphate buffer pH 7.0 (3.5 mL), Novozym 435 (30 mg,
135 LU), 708C.
Persistent literature reports on amidase activityof
CaLB[4,20–23] prompted us to applythis latter enzyme
Hydrolysis of racemic 5b in the presence of CaLB
showed negligible enantiodiscrimination, which indi-
to the hydrolysis of these amides. The highest rate cates that the rate-determining steps in synthesis and
was observed at 708C with Novozym 435 but even hydrolysis are different,[27] and the issue was not pur-
then 4d, 5d, 7d and 8d, as well as 8a, reacted with ex- sued anyfurther. Enantiomericallypure ( R)-1 and
treme sluggishness. All of the other amides could be (R)-2 were obtained, nevertheless, on account of the
hydrolyzed at an appreciable rate (see Table 3).
A considerable effect of the molecule size on the
good enantioselectivities in the acylations.
Furthermore, the hydrolysis of the phenoxy-substi-
reaction rate was observed as the phenyl-substituted tuted amides (7b and 8b) using penicillin V acylase
amides 7 and 8 were hydrolyzed much slower than was investigated, as this latter enzyme is known to
the amides derived from CH3 X CH2CO2H (4 and have a high affinity for the phenoxyacetyl moiety.[28]
À À
5). A somewhat similar chain length effect was re- From tests with three penicillin V acylases from differ-
centlyreported byTorres-Gavilµn et al. [24] The hydrol- ent sources and formulations (Table 5) we found Sem-
ysis rate of 4 and 5 increased in the order of X= acylase (a commercial preparation from Novozymes)
NH<S<O, as was observed in the acylation step, to be the best catalyst, as almost 100% amide was hy-
stronglyindicating that similar substituent effects are
apparent here. Amides bearing an oxygen (4b and 5b)
were hydrolyzed the fastest, but still 700 times slower
drolyzed within 120 h.
than synthesis, resulting in 100% conversion of 5b Conclusions
within 120 h.
In conclusion, it would seem that the hydrogen Proof of principle of fullyenzymatic resolution of two
bond-accepting and electron demanding character of chiral amines, 2-heptylamine and 2-phenylethylamine,
1514
asc.wiley-vch.de
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 1511 – 1516

FullyEnzymatic Resolution of Chiral Amines
FULL PAPERS
Table 5. Hydrolysis of 7b and 8b catalyzed by penicillin V
CaA (powder, 150 mg each) and 1,3-dimethoxybenzene
(150 mL, internal standard) were added. The reaction mix-
ture was incubated at 408C; samples (100 mL) were with-
drawn at regular intervals to monitor the reaction over time.
Initial rates (conversion/time) relative to the acylation of 1
by 3a are given in Table 1 and Table 2.
The reaction was stopped at 50% amine conversion byre-
moval of the enzyme and the molecular sieves. The unreact-
ed amine was isolated byextraction with dilute sodium hy-
droxide solution (pH 8.5). The organic fraction was concen-
trated under vacuum and pure amide was obtained byre-
crystallization from n-hexane.
Enzymatic hydrolysis of the amides: Preliminary hydroly-
sis experiments in the presence of CaLB (see Table 3) were
performed with amide (5 mg, 20–27 mmol) in 3.5 mL 0.1M
phosphate buffer pH 7, in the presence of immobilized
CaLB (Novozym 435, 30 mg, 135 LU) at 708C. The prepara-
tive experiments were performed with 5b or 5c (0.1M) and
CaLB CLEA (90 mg, 600 LU) in 3 mL 0.1M phosphate
buffer pH 7 at 408C.
acylase.[a]
Amide Pen V acylase
PVU[b] Time [h] Conv. [%]
(R)-7b Semacylase
Cryptocossus sp.
Biocat V
(R)-8b Semacylase
Cryptocossus sp.
Biocat V
6
4
6
6
4
6
120
120
120
75
120
120
98
17
14
95
21
52
[a]
Reaction conditions: amide (20 mmol) in 0.1M phosphate
buffer pH 7.0 (3.5 mL), penicillin V acylase at 408C.
Penicillin V acylase units, see Experimental Section.
[b]
has been delivered, using acyl donor reagents with
NH, O and S moieties in the b-position and CaLB as
acylation and deacylation catalyst. The preferred acyl
reagent was methyl 2-methoxyacetate, on account of
its high rate in acylation and deacylation. Efficient
resolution of these amines was also accomplished
upon acylation with methyl 2-phenoxyacetate in the
Amides 7b and 8b (5 mg, 20 mmol) were subjected to hy-
drolysis in the presence of penicillin acylase V (30 mg) in
0.1M phosphate buffer pH 7 (3.5 mL) at 408C.
presence of CaLB and deacylation catalyzed by peni- Synthesis of Acyl Donors
cillin V acylase.
Phenylpropionate ethyl ester: A solution of 3-phenylpro-
pionyl chloride (10 g, 60 mmol) in 50 mL of absolute ethanol
and 4.6 g (60 mmol) of pyridine was stirred overnight at
room temperature. The excess of ethanol was removed
under vacuum. Ethyl acetate (20 mL) was added and the re-
sulting solution was washed with cold water. The organic
layer was dried over Na2SO4. Ethyl 3-phenylpropionate was
obtained as a colorless oil; yield: 6.136 g (57.4%); GC-MS:
MW=178; 1H NMR (300 MHz, CDCl3): d=1.21 (t, 3H,
OCH2CH3), 2.62 (t, 2H, CH2CH2CO), 2.93 (t, 2H,
CCH2CH2), 4.12 (q, 2H, OCH2CH3), 7.29–7.15 (m, 5H, aro-
matic).
Phenylthioacetate methyl ester: Thiophenol (10 g, 0.09
mol) was dissolved in a mixture of 50 mL drymethanol and
10 mL solution of sodium methoxide in methanol [30%,
(wt)]. Methyl chloroacetate (16.6 g, 0.15 mol) was added
slowlyand the resulting mixture was stirred overnight at
room temperature. The excess of methanol was removed
under vacuum and dichloromethane (50 mL) was added.
The resulting solution was washed with cold water and dried
over Na2SO4. Kugelrohr distillation afforded an unpleasantly
smelling oil; yield: 7.5 g; GC-MS: MW=182; 1H NMR
(300 MHz, CDCl3): d=3.64 (s, 3H, OCH3), 3.7 (s, 2H,
SCH2CO), 7.42–7.21 (m, 5H, aromatic).
Experimental Section
Materials
Novozym 435 (immobilized Candida antarctica lipase B,
CaLB, 4500 LU gÀ1) was kindlydonated byNovozmy es
(Bagsvaerd, Denmark). A cross-linked enzyme aggregate
(CLEA) of CaLB was donated byCLEA Technologies
(Delft, The Netherlands). Penicillin G acylase from E. coli
(PGA-300) was obtained from Roche Diagnostics (Penz-
berg, Germany) as a gift. Penicillin G acylase from Alcali-
genes faecalis was obtained from the fermentation of re-
combinant E. coli cells.[29] Semacylase (190 PVU gÀ1) was
kindlydonated byNovozmyes (Bagsværd, Denmark),
Biocat V (190 PVU gÀ1) was purchased from Biochemie,
penicillin V acylase from Cryptococcus sp. (126 PVU gÀ1)
was received from the Academyof Sciences of the Czech
Republic (Prague, The Czech Republic) as a gift. All other
compounds were purchased from ACROS and Sigma or
synthesized as described below.
Enzyme Assays
Analytical Methods
One unit (LU) of lipase will liberate one mmol per min of
butyric acid from 8% tributyrin at pH 7.5 and 408C. One
unit (PVU) of penicillin V acylase will liberate one mmol per
min of phenoxyacetic acid from 2% penicillin V at pH 8 and
408C.
The progress of the acylation reactions was monitored by
gas chromatographyusing a CP Sil 5 CB column (50 m”
0.53 mm) at programmed temperature from 50–2508C. The
chiral analysis of the unconverted amine was performed by
derivatization with trifluoroacetic anhydride at 808C (2-hep-
tylamine) or 1008C (1-phenylethylamine) followed by chiral
GC on an Astec b-PH GC column (0.35 mm”30 m).
Enzymatic Reactions
Lipase-catalyzed acylation of amines: Acyl donor (3 mmol)
and amine (5 mmol) were dissolved in 5 mL of 1,2-dime-
thoxyethane. Novozym 435 (100 mg) and zeolite NaA and
The hydrolysis of 5 and 8 was monitored byHPLC on a
Waters 8”100 mm, 6 m Symmetry C18 cartridge contained
in
a Waters RCM 8”10 compression module, eluant
Adv. Synth. Catal. 2008, 350, 1511 – 1516
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1515

Hilda Ismail et al.
FULL PAPERS
MeOH-H2O [65:35 (v/v), SDS 1 gLÀ1 at pH 3.5 (KH2PO4
1 gLÀ1], 1 mLminÀ1, UV detection at 215 nm.
[6] H. Ismail, R. Madeira Lau, F. van Rantwijk, R. A. Shel-
don, Green Chem. 2008, 10, 415–418.
The analysis of the aliphatic amines liberated in the hy-
drolysis of 4 and 7 was performed byallowing a precise
amount of reaction mixture and a solution of 2-pentylamine
to react for 1 min with a commercial solution of o-phthal-
dialdehyde in the presence of 2-mercaptoethanol (1 mL).
Formation of 1 was quantified byHPLC on a Waters 8”
100 mm, 6 m Symmetry C18 cartridge contained in a Waters
RCM 8”10 compression module, eluant MeOH-H2O 80:20
(v/v) at 1 mLminÀ1, UV detection at 340 nm.
[7] M. A. Wegman, J. M. Elzinga, E. Neeleman, F. van
Rantwijk, R. A. Sheldon, Green Chem. 2001, 3, 61–64.
[8] F. van Rantwijk, R. A. Sheldon, Tetrahedron 2004, 60,
501–519.
[9] M. S. De Castro, J. V. S. Gago, Tetrahedron 1998, 54,
2877–2892.
[10] O. Kirk, M. W. Christensen, Org. Process Res. Dev.
2002, 6, 446–451.
[11] J. Pleiss, M. Fischer, R. D. Schmid, Chem. Phys. Lipids
The enantiomeric excess of 1 was measured byextracting
the amine at pH 11 into hexane, derivatization with tri-
fluoroacetic anhydride and GC on a Chiraldex 0.25 mm”
30 m b-PH GC column. The enantiomeric purityof 2 was
determined bychiral HPLC on a Daicel 4”150 mm, 5 mm
Crownpack CR+ column, eluant aqueous 0.1M HClO4,
pH 1.5 at 0.6 mLminÀ1, UV detection at 215 nm.
1998, 93, 67–80.
[12] B. H. Hoff, H. W. Anthonsen, T. Anthonsen, Tetrahe-
dron: Asymmetry 1996, 7, 3187–3192.
[13] J. Ottoson, K. Hult, J. Mol. Cat. B: Enzym. 2001, 11,
1025–1028.
[14] H. T. Huang, T. A. Seto, G. M. Shull, Appl. Microbiol.
1963, 11, 1–6.
[15] F. Balkenhohl, K. Ditrich, B. Hauer, W. Ladner, J.
Prakt. Chem. 1997, 339, 381–384.
[16] M. Cammenberg, K. Hult, S. Park, ChemBioChem
2006, 7, 1745–1749.
[17] C. Laurence, P. Nicolet, M. Helbert, J. Chem. Soc.
Acknowledgements
Generous donations of enzymes by Novozymes, CLEA Tech-
nologies, Roche Diagnostics and the Academy of Sciences of
The Czech Republic are gratefully acknowledged. Thanks are
due to Prof. Dr. Kasche (Technische Universität Harburg,
Germany) for help with the recombinant expression of the
penicillin acylase from A. faecalis. Experimental assistance
by Dr. M. H. A. Janssen (CLEA Technologies) is gratefully
acknowledged. Dr. Ismail wishes to thank the Gadjah Mada
University (Yogyakarta, Indonesia) for leave of absence. Fi-
nancial support for H. Ismail by the QUE Project of the Fac-
ulty of Pharmacy, Gadjah Mada University, Yogyakarta, In-
donesia and by the Faculty for the Future Program of the
Schlumberger Foundation is gratefully acknowledged.
Perkin Trans. 2 1986, 1081–1090.
[18] C. Hansch, A. Leo, R. W. Taft, Chem. Rev. 1991, 91,
185–195.
[19] M. J. Kamlet, J.-L. M. Abboud, M. H. Abraham, R. W.
Taft, J. Org. Chem. 1983, 48, 2877–2887.
[20] H. Smidt, A. Fischer, R. D. Schmid, Biotechn. Techn.
1996, 10, 335–338.
[21] D. Reyes Duarte, E. Castillo, E. Barzana, A. Lopez-
Munguia, Biotechnol. Lett. 2000, 22, 1811–1814.
[22] T. Waggeg, M. M. Enzelberger, U. T. Bornschener,
R. D. Schmidt, J. Biotechnol. 1998, 61, 75–78.
[23] See also: C. Simons, J. G. E. van Leeuwen, R. Stemmer,
I. W. C. E. Arends, T. Maschmeyer, U. Hanefeld, R. A.
Sheldon, Adv. Synth. Catal. submitted.
[24] A. Torres-Gavilµn, E. Castillo, A. L. Munguía, J. Mol.
Cat. B: Enzym 2006, 41, 136–140.
References
[25] P. Lopez-Serrano, L. Cao, F. van Rantwijk, R. A. Shel-
don, Biotechnol. Lett. 2002, 24, 1379–1383.
[26] R. Fujii, Y. Nakagawa, J. Hiratake, A. Sogabe, K.
Sakata, Protein Eng. Des. Sel. 2005, 16, 93–101.
[27] See also: M. C. de Zoete, A. C. Kock-van Dalen, F. van
Rantwijk, R. A. Sheldon, J. Chem. Soc. Chem.
Commun. 1993, 1831–1832.
[28] H. Vanderhaeghe, M. Claesen, A. Vlietinck, G. Par-
mentier, Appl. Microbiol. 1968, 16, 1557–1563.
[29] a) Z. Ignatova, S. Stoeva, B. Galunsky, C. Hçrnle, A.
Nurk, E. Piotraschke, W. Voelter, V. Kasche, Biotech-
nol. Lett. 1998, 20, 977–982; b) P. M. Deak, S. Lutz-
Wahl, H. Bothe, L. Fischer, Biotechnol. Lett. 2003, 25,
397–400.
[1] G. Hieber, K. Ditrich, Chimica Oggi 2001, 16, 16–20.
[2] M. Breuer, K. Ditrich, T. Habicher, B. Hauer, M. Keß-
eler, Stürmer, T. Zelinski, Angew. Chem. 2004, 116,
806–843; Angew. Chem. Int. Ed. 2004, 43, 788–824.
[3] K. Ditrich, F. Balkenhohl, W. Ladner, (BASF), PCT
Int. Appl. WO 97/10201, 1997; Chem. Abstr. 1997, 126,
277259f.
[4] A. Torres-Gavilµn, J. Escalante, I. Regla, A. López-
Munguía, E. Castillo, Tetrahedron: Asymmetry 2007,
18, 2621–2624.
[5] D. T. Guranda, L. M. van Langen, F. van Rantwijk,
ˇ
R. A. Sheldon, V. K. Svedas, Tetrahedron: Asymmetry
2001, 12, 1645–1650.
1516
asc.wiley-vch.de
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 1511 – 1516