Enantiomers of adrenaline-type amino alcohols by Burkholderia cepacia lipase-catalyzed asymmetric acylation

Article abstract of DOI:10.1016/j.tetasy.2004.10.027

The enantiomers of norphenylephrine and octopamine were prepared using lipase PS-catalyzed enantioselective acylation with butanoic anhydride. Burkholderia cepacia lipase-catalyzed acylation with butanoic anhydride was used for the preparation of pharmaceutically important norphenylephrine and octopamine enantiomers, all in 98% ee. Reactivity of the phenolic OH groups and easy racemization, especially in the case of octopamine, complicated optimization. For norphenylephrine, chemical N-acylation was fast and allowed the subsequent enzymatic benzylic acylation in situ. The preparation of the octopamine enantiomers became possible by using the N-Fmoc protected substrate and by the Candida antarctica lipase B-catalyzed deprotection of the OH groups before the N-deprotection was performed.

Full text of DOI:10.1016/j.tetasy.2004.10.027

Tetrahedron:
Asymmetry
Tetrahedron: Asymmetry 15 (2004) 3723–3729
Enantiomers of adrenaline-type amino alcohols by Burkholderia
cepacia lipase-catalyzed asymmetric acylation
Katri Lundell, Erja Katainen, Anu Kiviniemi and Liisa T. Kanerva*
Laboratory of Synthetic Drug Chemistry and Department of Chemistry, University of Turku, Lemminka¨isenkatu 5 C,
7th floor, FIN-20520 Turku, Finland
Received 15 July 2004; accepted 25 October 2004
Abstract—Burkholderia cepacia lipase-catalyzed acylation with butanoic anhydride was used for the preparation of pharmaceuti-
cally important norphenylephrine and octopamine enantiomers, all in 98% ee. Reactivity of the phenolic OH groups and easy race-
mization, especially in the case of octopamine, complicated optimization. For norphenylephrine, chemical N-acylation was fast and
allowed the subsequent enzymatic benzylic acylation in situ. The preparation of the octopamine enantiomers became possible by
using the N-Fmoc protected substrate and by the Candida antarctica lipase B-catalyzed deprotection of the OH groups before
the N-deprotection was performed.
ꢀ 2004 Elsevier Ltd. All rights reserved.
1. Introduction
the catecholic meta- and para-OH groups of 2-amino-
1-phenylethanol agonists has been shown to be key to
receptor activation rather than to affect the binding
process.3 When the biological activity of chiral mole-
cules is studied both the enantiomers in an enantiopure
form are needed.
(R)-(ꢀ)-Noradrenaline (R)-1 is a natural catecholamine,
which acts as an a-adrenoceptor agonist (Scheme 1,
X = Y = OH).1,2 The compound also shows activity at
b-receptors, structural factors determining the activity
in such a way that a-activity is dramatically decreased
and b-activity increased when the compound is N-sub-
stituted and the substituent is larger than methyl.
Importantly, 2-amino-1-phenylethanol analogues exert
their pharmacological activity through the stereoisomer,
which corresponds to the absolute configuration of (R)-
(ꢀ)-noradrenaline. In such a stereostructure, the three
important groups (the charged nitrogen, the phenyl
groupand the benzylic OH) are in the most favourable
orientation for interaction with a receptor. The task of
Synthetic methods to enantiopure adrenergic 2-amino-1-
phenylethanols include the transformations of optically
active cyanohydrins and b-azido alcohols.4–6 In these
cases, chemical, chemoenzymatic and purely biocatalytic
methods have afforded the target molecules. Resolution
of 2-amino-1-phenylethanols has also been used. Thus,
octopamine, as its salts with optically active acids, has
been resolved by fractional crystallization,7 while the
lipase-catalyzed kinetic resolution has served as one of
the most fascinating methods in the case of vari-
ous ring-substituted 2-amino-1-phenylethanols.8–11 We
previously used Burkholderia cepacia lipase (lipase PS)-
catalyzed asymmetric acylation of various 2-amino-1-
phenylethanols with an acid anhydride in toluene/THF
(3:1) as a highly effective method to pure enantiomers.10
The addition of THF was necessary in order to improve
the solubility of 2-amino-1-phenylethanols while the
ring substituents were typically unreactive under the
reaction conditions. Herein we report the preparation
HO
(R)
H
HO
H
Y
X
NH2
H2N
Y
X
(S)
1: X=Y=OH
2: X =H, Y=OH
3: X=OH, Y=H
Scheme 1.
of both the enantiomers of norphenylephrine
(X = H, Y = OH; Scheme 1) and octopamine
(X = OH, Y = H) with phenolic OH groups using the
2
3
*
Corresponding author. Tel.: +358 2333 6773; fax: +358 2333 7955;
e-mail: lkanerva@utu.fi
0957-4166/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetasy.2004.10.027

3724
K. Lundell et al. / Tetrahedron: Asymmetry 15 (2004) 3723–3729
OH
OCOPr
HO
H
Y
X
NH2
Y
X
NHR
Y
X
NHR
S-selective
rac-2 or rac-3
rac-4 or rac-5a-c
6 or 7a-c
Lipase PS
(P
rCO
)2O
or
gan
ic
solv
en
t
40 o
C
OH
OCOPr
Y
X
NHR
S-selective
Y
X
NHR
8 or 9a-c
10 or 11a-c
Scheme 2. Reagents and conditions: norphenylephrine derivatives: 4 and 6 (X = H, Y = OH, R = COPr); 8 and 10 (X = H, Y = OCOPr, R = COPr).
Octopamine derivatives: 5 and 7 (X = OH, Y = H); 9 and 11 (X = OCOPr, Y = H) with a (R = COPr), b (R = Boc) and c (R = Fmoc).
lipase PS-catalyzed acylation with butanoic anhydride.
The behaviour of the enantiomers for agonist promoted
human a2A-adrenoceptor activation has already been
published although the details for the preparation of
the enantiomers were left to wait a proper optimization.3
Difficulties arose from two facts. First, substrates 2 and
3 are trifunctional, with all functional groups tending to
react under acylation conditions (Scheme 2). A more
serious problem was connected to the easy racemization
of 3, in particular. Thus, attention had to be paid to pro-
tecting groups, which allow deprotection under condi-
tions where the amino alcohol is stereochemically stable.
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
2. Results and discussion
0
1
2
3
4
t/h
2.1. Asymmetric acylation of norphenylephrine and
octopamine derivatives
Figure 1. Progression curves for the lipase PS-catalyzed reaction
between 4 and butanoic anhydride in toluene THF (3:1) at 40ꢁC; the
disappearance of 4 (d) and the appearances of 6 (j), 8 (r) and 10 (m).
Chemo- and/or regioselectivities for the kinetic resolu-
tion of polyfunctional compounds are possible advan-
tages of lipase catalysis. In the best cases, these
selectivities can save time and chemicals by reducing
the number of synthetic steps because separate protec-
tion/deprotection steps become unnecessary. Based on
this reasoning, the lipase PS-catalyzed acylation of var-
ious 2-amino-1-phenylethanols with an acid anhydride
in toluene/THF (3:1) was previously studied.10 In this
method, highly efficient chemical N-acylation was fol-
lowed by highly enantioselective enzymatic O-acylation
in situ. Accordingly, enzymatic chemoselectivity had
no possibility in playing a role in the studied kinetic res-
olutions. Herein, compounds 2 and 4 and 3 and 5a–c
were subjected to the lipase PS-catalyzed acylation with
butanoic anhydride in toluene/THF (3:1). The phenolic
and benzylic OH groups were both susceptible to enzy-
matic acylation and the various acylation products were
all detected by the chiral GC method. As shown for the
acylations of 4 and 5c (Figs. 1 and 2) as examples for
general trends, the lipase showed relatively high chemo-
selectivity towards the benzylic OH, leading to just low
amounts of 8 and 9c (Ç) and also to those of 10 and 11c
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
0
5
10
15
20
25
t/h
Figure 2. Progression curves for the lipase PS-catalyzed reaction
between 5c and butanoic anhydride in toluene THF (3:1) at 40ꢁC; the
disappearance of 5c (d) and the appearances of 7c (j), 9c (r) and 11c
(m).

K. Lundell et al. / Tetrahedron: Asymmetry 15 (2004) 3723–3729
3725
(m). Except for negligible benzylic acylation of 4 (2.5%
in 16h), other chemical reactions were not detected.
ative effect of butanoic acid was at least partly ruled out
by replacing the anhydride with 2,2,2-trifluoroethyl but-
anoate and by observing that the reaction in tert-alcoh-
ols then similarly stopped (entry 3 compared to entries 4
or 5). The amount of acid seemed to affect enantiopuri-
ties throughout (entries 3–5). Racemization can be
understood through the resonance stabilized kinoidic
structure of the substrate.
The lipase-catalyzed benzylic acylation of 4 and 5a–c
and that of 8 and 9a–c proceeded in a highly S-selective
manner while the acylation of the phenolic OH was not
enantioselective (Scheme 2). Accordingly, after long
enough reaction times, 8 and 9a–c could be obtained
as pure (R)-enantiomers after the (S)-enantiomers had
further reacted in the formation of (S)-10 and (S)-11a–
c, respectively. The phenolic acylation of (S)-6 and
(S)-7a–c was another route to the respective (S)-10
and (S)-11a–c. However, due to the high chemo- and
enantioselectivity of the present reactions towards benz-
ylic acylation, it is advantageous to stopthe reaction as
soon as the unreacted substrate enantiomer has had time
enough to become enantiomerically pure. This reason-
ing holds well for the kinetic resolution of rac-2 through
rac-4 in the presence of lipase PS preparation and buta-
noic anhydride in toluene/THF (3:1). The unreacted (R)-
4 (ee = 97%) and the produced (S)-6 (ee = 93%) were
obtained at 55% conversion (row 1, Table 1), allowing
the straightforward gram-scale preparation of the
enantiomers of 2 as described later. It is noteworthy that
the conversion of the reaction always means the total
reaction.
In order to prepare the enantiomers of 3, a separate N-
protection step was introduced through the synthesis of
rac-5b and rac-5c. For the lipase PS-catalyzed acylation
of the N-Boc-protected amino alcohol 5b in toluene/
THF (3:1), long reaction time (142h to reach 50% con-
version) gave a reason to turn to the N-Fmoc protected
amino alcohol 5c (entries 6 and 7, respectively; Table 1).
As a more important reason, the stereochemical stability
of the enantiomers under the subsequent deprotection
steps favoured the Fmoc protection. In order to get
the unreacted (R)-5c in an enantiopure form, there was
a need to conduct the reaction considerably over 50%
conversion. As a consequence, the reaction mixture con-
tained considerable amounts of (R)-9c in addition to
(R)-5c and (S)-7c (Fig. 2). The amount of (S)-11c always
remained negligible.
2.2. Deprotection and enantiopurity
The lipase PS-catalyzed acylation of rac-3 through rac-
5a with butanoic anhydride in organic solvents was
more complicated and the reactions tended to stop at
conversions shown in Table 1 (entries 2 and 3) with ee
values, which were not acceptable for pharmaceutical
studies. With acid anhydrides as acyl donors, large
amounts of acid were introduced into the reaction mix-
ture during the course of the reaction. However, the neg-
Due to easy racemization, conditions used for the re-
moval of the N-protective groups (PrCO, Boc and
Fmoc) of the present compounds are critical for enan-
tiopurity. For this purpose, (R)-4 and (R)-5a–c were
subjected to various conditions, which are typically used
in order to remove the protecting groups (Table 2).
Table 1. Lipase PS (50mg/mL)-catalyzed acylation of 4 and 5a–c (0.05M) with butanoic anhydride (0.2M) at 40ꢁC
Entry
Substrate
Solvent
Time (h)
Conversion (%)
Product
ee4 or 5/ee6 or 7 (%)
1
2
3
4
5
6
7
4a
5a
Toluene/THF (3:1)
Toluene/THF (3:1)
tert-Amyl alcohol
tert-Amyl alcohol
tert-Butyl alcohol
Toluene/THF (3:1)
Toluene/THF (3:1)
4
4
55
49
29
15
15
50
49
6
97/93
88/90
7a
7a
7a
7a
7b
7c
5a
5
rac/rac
17/>98
18/>98
>98/>98
64/93
5ab
5ab
5bc
5cc
3
24
142
4
a Butanoic anhydride 0.3M.
b 2,2,2-Trifluoroethyl butanoate as an acyl donor.
c Butanoic anhydride 0.1M.
Table 2. Effect of N-deprotection on enantiopurity
Entry
Compound
Method
H2SO4 (0.4–1M)/MeOH
ee(4 or 5a–c) (%)
ee(2 or 3) (%)
1
2
3
4
5
6
7
8
(R)-4
(R)-4
>98
>98
>98
>98
>98
>98
>98
>98
>98
>98
(1) HCl (0.54M)/MeOH, (2) NH3
(1) HCl (0.54M)/MeOH, (2) NH3
(1) HCl (0.54M)/MeOH, (2) NH3
BF3/Et2O
(R)-5a
(R)-5b
(R)-5b
(R)-5b
(R)-5b
(R)-5c
rac
rac
rac
HCl (1M)/AcOH
Iodotrimethylsilane
5% Piperidine/THF
rac
No reaction
>98

3726
K. Lundell et al. / Tetrahedron: Asymmetry 15 (2004) 3723–3729
Table 3. Screening of lipases (75mg/mL) and solvents for the alcoholysis of (S)-7c (0.05M) at 40ꢁC
Entry
Enzyme
Solvent
Nucleophile
t (h)
Yielda (%)
ee5c (%)
1
2
3
4
5
6
7
8
9
CAL-B
CRL
Toluene/THF (3:1)
Toluene/THF (3:1)
Toluene/THF (3:1)
THF
OctOH (0.15M)
OctOH (0.15M)
OctOH (0.15M)
OctOH (0.15M)
OctOH (0.15M)
EtOH
96
96
96
70
70
96
94
94
94
20
0
98
—
98
98
—
98
98
98
98
Lipase PS
CAL-B
CAL-B
CAL-B
CAL-B
CAL-B
CAL-B
16
6
H2O [5% (v/v)]/THF
EtOH
0
20
68
45
40
Toluene/THF (3:1)
THF
Toluene/THF (1:1)
2-PrOH (0.3M)
2-PrOH (10 M)
2-PrOH (6.5M)
a Relative proportion.
Hydrolysis or methanolysis in the presence of HCl is
commonly used for splitting amide bonds. The method
worked well in the case of compound (R)-4 (entries 1
and 2). However, (R)-5a and (R)-5b were unable to with-
stand acidic reaction conditions, thus leading to total
racemization (entries 3–6). An explanation is the reso-
nance stabilization through the kinoidic intermediate
of octopamine-like compounds under acidic conditions.
Fmoc is typically removed under basic conditions.
Accordingly, the deprotection of (R)-5c easily proceeded
using 5% piperidine in THF, leading to (R)-3 without
any racemization (entry 8).
2.3. Gram-scale preparation of the enantiomers of 2 and 3
The results in Tables 1 and 2 clearly indicate that the
previously developed acylation method with an acid
anhydride in toluene/THF (3:1) well suits the kinetic res-
olution of rac-2 with butanoic anhydride and the lipase
PS preparation. A gram-scale resolution was performed
as described in the Experimental section (Scheme 3). The
enantiopure compounds (R)-4 and (S)-6 were isolated
close to quantitative chemical yields. Deprotection with
HCl/MeOH followed by NH3-bubbling proceeded well
without racemization in the formation of enantiopure
(R)-2 and (S)-2.
Removal of an ester groupfrom ( S)-7c with saponifica-
tion using NaOH or K2CO3 caused racemization. This
was disappointing because the removal of the Fmoc
had proceeded at the same time. The only feasible
way for the deprotection proved to be enzymatic alco-
holysis. For this purpose, three lipases were first
screened for the octanolysis of (S)-7c in toluene/THF
(3:1) (entries 1–3, Table 3). Candida rugosa lipase
(CRL)-catalyzed alcoholysis did not proceed at all while
lipase PS and CAL-B gave slow reactions. No improve-
ment was seen when the composition of the solvent was
changed (entries 4 and 5). The reaction in neat ethanol
was also slow (entry 6). Finally, 2-propanolysis in tolu-
ene/THF (3:1) was exploited in order to split the ester
bond from 7c. Phenolic and benzylic esters reacted
under the same conditions with the phenolic ester bond
being more alert to enzymatic alcoholysis than the
benzylic one.
The gram-scale kinetic resolution of rac-5c was per-
formed as described in the Section 4 (Scheme 4). The
lipase PS-catalyzed acylation led to the resolved mixture
containing the unreacted (R)-5c and the produced (S)-
7c, (R)-9c and (S)-11c at 67% conversion. The products
were all separated by column chromatography. (R)-5c
and (R)-9c, on one hand, and (S)-7c and (S)-11c, on
the other hand, can be joined if desired to produce
(R)- and (S)-3, respectively. Herein, the ester group of
(S)-7c was first removed by CAL-B-catalyzed 2-propan-
olysis in toluene/THF. The reaction was slower than
when performed on a small scale and 90% conversion
was reached in 16 days. Long reaction times are evi-
dently due to the equilibrium nature of the enzymatic
alcoholysis. Afterwards, the Fmoc-groupof ( R)-5c and
(S)-5c was removed by piperidine treatment leading to
enantiopure (R)-3 and (S)-3, respectively.
OH
OH
HO
H
PrOCO
PrOCHN
H
HO
NHCOPr
OH
HO
NHCOPr
(PrC
O)2O
HO
NH2
+
(PrC
O)2O
Lipase PS
(S)-6
(R)-4
Toluene/THF
Toluene/THF
rac-2
rac-4
40o
C
40 o
C
HCl/MeOH
NH3
HCl/MeOH
NH3
HO
H
HO
H
HO
NH2
H2N
OH
(R)-2
(S)-2
ee 98%
ee 98%
Scheme 3.

K. Lundell et al. / Tetrahedron: Asymmetry 15 (2004) 3723–3729
3727
OH
HO
H
PrOCO
FmocHN
H
NHFmoc
NHFmoc
(PrC
O)2O
(S)-11c
+
(R)-9c
+
+
Lipase PS
Tol
40 o
uen
e/TH
rac-5c
F
HO
HO
(R)-5c
(S)-7c
OH
C
2-PrOH
CAL-B
Toluene/THF
5% Piperidine
THF
HO
H
HO
H
FmocHN
NH2
(S)-5c
OH
(R)-3
ee >98%
HO
5% Piperidine
THF
HO
H
H2N
(S)-3
OH
ee 98%
Scheme 4.
3. Conclusions
(Novozym 435, CAL-B) was a product from Roche.
Lipase from C. rugosa (CRL) was obtained from Sigma.
Before use, lipase PS was adsorbed on Celiteꢂ in the
presence of sucrose and Tris–HCl buffer (20mM,
pH7.9). The enzyme preparation thus obtained con-
tained 38% (w/w) of the lipase and Celiteꢂ and 23%
(w/w) of sucrose.12 Norphenylephrine and octopamine
hydrochlorides, 9-fluorenylchloroformate (FmocCl),
di-tert-butyl dicarbonate (Boc2O) and butanoic anhy-
dride were purchased from Aldrich. Solvents used were
of the highest analytical grade and obtained from Merck
or Lab Scan Ltd.
The acylations of N-protected norphenylephrine and
octopamine derivatives 4 and 5a–c with butanoic
anhydride were studied in the presence of the lipase PS
preparation in toluene/THF (Scheme 2, Table 1). The
non-enantioselective acylations at the phenolic OH
groups were observed with both substrates although
chemoselectivity strongly favoured the highly enantio-
selective benzylic acylation. The enantiomers of norphen-
ylephrine 2 were prepared starting with the chemical
N-acylation of rac-2 and continuing with enzymatic
acylation in situ (Scheme 3). The deprotection of the
N,O-diacylated (S)-product 6 and that of the unreacted
N-acylated R-enantiomer 4 was accomplished by reflux-
ing in HCl/MeOH and finally by bubbling NH3. It was
not possible to use the same method for the enzymatic
kinetic resolution of rac-3 because the reactions had a
tendency to stopat early conversions with the enantio-
mers becoming racemized under the deprotection condi-
tions. Instead, the N-Fmoc-protected substrate was used
for lipase PS-catalyzed acylation with butanoic anhy-
dride (Scheme 4). The resolution stepresulted in 98%
ee at 67% conversion for the two main enantiomers
(R)-5c and (S)-7c. It occurred that in order to prevent
racemization, the O-deprotection through 2-propano-
lysis by CAL-B had to be done before the N-Fmoc
group was removed by the piperidine treatment.
Norphenylephrine rac-2 was prepared by bubbling
ammonia through its hydrochloride solution in chloro-
form/ethanol (3:1). HCl was removed from commercial
octopamine hydrochloride using Amberlite IRA-401
anion exchange resin in methanol. N-Boc-protected 5b
was prepared from 3 with di-tert-butyl dicarbonate
and triethylamine in methanol. N-Fmoc-protected 5c
was obtained using 9-fluorenylchloroformate and diiso-
propyl amine in water/dioxane. Amides 4 and 5a were
obtained as a result of the reactions of 2 and 3 with but-
anoic anhydride while the amides in this solution were
used in the enzymatic O-acylations.
In a typical lipase-catalyzed acylation reaction, butanoic
anhydride (0.2 or 0.1M) or 2,2,2-trifluoroethyl butano-
ate (0.2M) was added into one of the amino alcohol
derivatives 4 or 5a–c (0.05M) in an organic solvent
(2mL) and the enzyme preparation (50mg/mL) was
added to start the enzymatic reaction at 40ꢁC. The pro-
gress of the reactions was followed by taking samples
(100lL) at intervals, filtering off the enzyme and analyz-
ing the sample by GC on a Chrompack CP-Chirasil-L-
valine capillary column after the derivatization of the
free alcohol groups with propionic anhydride. The
4. Experimental
4.1. Materials and methods
Lipase from B. cepacia (previously Pseudomonas cepacia
lipase; lipase PS) was purchased from Amano Pharma-
ceutical Co., Ltd (Nagoya, Japan). Chirazyme L2

3728
K. Lundell et al. / Tetrahedron: Asymmetry 15 (2004) 3723–3729
absolute configurations of octopamine enantiomers 3
were determined by comparing the specific rotations to
those in the literature.7 In the case of norphenylephrine
enantiomers 2, the (S)-enantiomer was expected to react
as in the case of octopamine and other 2-amino-1-phen-
ylethanols studied in our previous works.8–10 1H NMR
spectra were recorded on a Bruker 200 or 400 spectro-
meter, with tetramethylsilane as an internal standard.
MS-spectra were recorded on a VG analytical 7070E
instrument equipped with VAXstation 3100 M76
computer. Optical rotations were measured using a
Jasco DIP-360 polarimeter, the [a]D values being in units
column chromatography (silica, methanol/dichloro-
methane (1:1)). Enantiopure (R)-2 {0.280g, 1.3mmol,
20
D
ee 98%, ½aŠ ¼ ꢀ1:7 (c 5.8, MeOH)} or (S)-2 {0.300g,
20
1.0mmol, ee 98%, ½aŠ ¼ þ1:7 (c 5.6, MeOH)} was
D
obtained.
1
(R)-2: H NMR (DMSO, 25ꢁC): d (ppm) 2.75 (m, 2H,
CH2NH2), 3.10 (s, 1H, CHOH), 4.25 (dd, 1H,
CH(OH)CH2), 6.51–6.65 (m, 3H, C6H4) and 7.01 (t,
1H, C6H4). 13C NMR (DMSO, 25ꢁC): d (ppm) 47.1,
71.4, 115.9, 119.7, 123.8, 129.5, 146.2 and 150.8. Mass
spectrum M+ (calculated for C8H11NO2) = 153.0792
(153.0790).
of 10ꢀ1 degcm2 gꢀ1
.
4.2. Gram-scale resolution of rac-2
4.4. Gram-scale resolution of rac-5c
rac-2 (1.03g, 6.7mmol) was dissolved in toluene/THF
((3:1), 142mL) after which butanoic anhydride
(4.6mL, 28mmol) was added. After the formation of
rac-4 (5min), the enzyme preparation (7.1g) was added
and the reaction allowed to proceed to 52% conversion
in 3h at 40ꢁC. The enzyme was filtered off and the
solvent evaporated. Purification by column chromato-
graphy [silica, ethyl acetate/hexane (7:3)] yielded
Lipase PS preparation (5.35g) was added into the solu-
tion of rac-5c (2.00g, 5.32mmol) and butanoic anhy-
dride (1.75mL, 10.7mmol) in toluene/THF [107mL,
(3:1)] at 40ꢁC. After 24h, the enzyme was filtered off
at 67% conversion. Purification by column chromato-
graphy (silica, ethylacetate/hexane (1:1)) yielded the unre-
20
D
1.0, MeOH)} and the produced (S)-7c {0.79g,
acted (R)-5c {0.71g, 1.89mmol, ee 96%, ½aŠ ¼ þ4:5 (c
20
the unreacted (R)-4 {0.692g, 3.1mmol, ee >98%,
1.77mmol, ee 98%, ½aŠ ¼ þ47:5 (c 1.0, MeOH)}, (R)-
D
20
D
20
D
MeOH)} and (S)-11c {0.52g, 1.00mmol, ee >98%,
½aŠ ¼ þ5:0 (c 2.0, MeOH)]} and the produced
9c {[0.26g, 0.59mmol, ee >98%, ½aŠ ¼ þ9:7 (c 1.0,
20
20
(S)-6 0.821g, 2.8mmol, ee 98%, ½aŠ ¼ þ46:8 (c 1.0,
D
MeOH)]. The proportions of 8 and 10 were negligible
and the products not separated.
½aŠ ¼ þ31:0 (c 1.0, MeOH)}.
D
(R)-5c: 1H NMR (DMSO, 25ꢁC): 3.09 (m, 2H,
CHCH2NHFmoc), 4.23 (m, 2H, NHCO2CH2CH),
4.49 (d, 1H, CHOH), 5.23 (t, 1H, CH(OH)CH2), 6.68
(d, 2H, J = 8.3, C6H4), 7.10 (d, 2H, J = 8.3, C6H4),
7.28–7.42 (tt, 2 · 2H, Fmoc), 7.85 (dd, 2 · 2H, Fmoc)
and 9.27 (s, 1H, ArOH). 13C NMR (DMSO, 25ꢁC):
48.7, 70.5, 71.7, 109.6, 114.7, 120.0, 121.3, 127.1,
127.2, 128.9, 134.0, 137.4, 138.3, 142.6, 143.0, 156.4
and 157.9. Mass spectrum M+ (calculated for
C23H21NO4) = 375.1481 (375.1471).
1
(R)-4: H NMR (CDCl3, 25ꢁC): d (ppm) 0.83 (t, 3H,
J = 7.4, CH3CH2), 1.49 (m, 2H, J = 7.4, CH3CH2CH2),
2.05 (t, 2H, CH3CH2CH2CO), 3.05 and 3.27 (m, 2H,
CHCH2NHCOPr), 5.37 (d, 1H, CHCH2NHCOPr),
6.63 (dd, 1H, C6H4), 6.73 (d, 1H, J = 7.8, C6H4), 6.76
(s, 1H, C6H4), 7.10 (t, 1H, J = 7.8, C6H4), 7.83 (t, 1H,
CH2NHCO) and 9.29 (s, 1H, ArOH). 13C NMR
(CDCl3, 25ꢁC): (ppm) 13.7, 18.7, 37.3, 46.9, 71.5,
112.9, 113.9, 116.6, 128.9, 145.4, 157.2 and 172.3. Mass
spectrum M+ (calculated for C12H17NO3) = 223.1212
(223.1208).
(S)-7c: 1H NMR (DMSO, 25ꢁC): 0.84 (t, 3H, CH3CH2,
J = 7.4), 1.52 (m, 2H, CH3CH2CH2, J = 7.4), 2.22 (m,
2H, CH3CH2CH2CO), 3.37 (m, 2H, CHCH2NHFmoc),
4.23 (d, 1H, NHCO2CH2CH), 4.26 (m, 2H,
NHCO2CH2CH), 5.68 (dd, 1H, CH2CHOCOPr), 6.68
(d, 2H, J = 8.37, C6H4), 7.11 (d, 2H, J = 8.37, C6H4),
7.31 (t, 2H, Fmoc), 7.41 (t, 2H, Fmoc), 7.55 (m, 1H,
CH2NHCO), 7.65 (m, 2H, Fmoc), 7.88 (d, 2H, Fmoc)
and 9.52 (s, 1H, ArOH). 13C NMR (DMSO, 25ꢁC):
13.8, 18.4, 36.1, 45.8, 47.1, 65.9, 74.0, 115.6, 120.6,
125.6, 127.5, 128.1, 128.3, 129.1, 141.2, 144.3, 156.7,
157.7 and 172.6. Mass spectrum M+ (calculated for
C27H27NO5) = 445.1898 (445.1889).
(S)-6: 1H NMR (CDCl3, 25ꢁC): d (ppm) 0.84 (tt, 2 · 3H,
J = 7.4, CH3CH2), 1.43–1.59 (qq, 2 · 2H, J = 7.4,
CH3CH2CH2), 2.05 and 2.31 (tt, 2 · 2H, J = 8.5,
CH3CH2CH2CO), 3.27–3.41 (m, 2H, CHCH2NH-
COPr), 5.66 (dd, 1H, J = 4.2, CHCH2NHCOPr), 6.69
(t, 3H, C6H4), 7.15 (t, 1H, C6H4), 7.99 (t, 1H,
CH2NHCO) and 9.46 (s, 1H, ArOH). 13C NMR
(CDCl3, 25ꢁC): d (ppm) 13.9, 14.0, 18.4, 19.1, 36.0,
37.7, 44.2, 74.0, 113.4, 115.3, 117.1, 129.9, 140.5,
157.8, 172.5 and 172.7. Mass spectrum M+ (calculated
for C16H23NO4) = 293.1636 (293.1627).
1
4.3. Deprotection of (R)-4and ( S)-6
(R)-9c: H NMR (DMSO, 25ꢁC): 0.87 (t, 3H, J = 7.1,
CH3CH2), 1.65 (m, 2H, J = 7.1, CH3CH2CH2), 2.25
(m, 2H, CH3CH2CH2CO), 3.37 (m, 2H, CHCH2NHF-
moc), 4.20 (m, 2H, NHCO2CH2CH), 4.23 (d, 1H,
NHCO2CH2CH), 5.50 (d, 1H, CH(OH)CH2), 7.10 (d,
2H, C6H4), 7.30–7.43 (m, 6H, C6H4 and Fmoc),
7.61 (t, 1H, CH2NHCO), 7.65 (d, 2H, Fmoc) and
7.88 (d, 2H, Fmoc). 13C NMR (DMSO, 25ꢁC): 13.3,
17.9, 32.2, 45.8, 48.3, 59.5, 71.4, 107.5, 117.7, 120.0,
(R)-4 (0.420g, 1.9mmol) or (S)-6 (0.600g, 2.0mmol) was
dissolved in HCl/methanol (10mL) and the mixture ref-
luxed overnight. Methanol was evaporated, the residue
dissolved in chloroform/ethanol (3:1) and ammonia
bubbled through the solution. The immediately
formed white precipitate (NH4Cl) was filtered off. The
solution was evaporated and the residue purified by

K. Lundell et al. / Tetrahedron: Asymmetry 15 (2004) 3723–3729
3729
127.5, 136.7, 138.8, 143.9, 149.4 and 171.7. Mass
spectrum M+ (calculated for C27H27NO5) = 445.1897
(445.1889).
CH(OH)CH2), 6.77 (d, 2H, J = 8.5, C6H4) and 7.17
(d, 2H, J = 8.5, C6H4). 13C NMR (DMSO, 25ꢁC): d
(ppm) 49.0, 71.8, 120.5, 121.9, 127.6, 129.4, 137.9
and 149.8. Mass spectrum M+ (calculated for
C8H11NO2) = 153.0787 (153.0790).
(S)-11c: 1H NMR (DMSO, 25ꢁC): 0.84 (m, 2 · 3H,
J = 7.1, CH3CH2), 1.51 and 1.63 (m, 2 · 2H, J = 7.1,
CH3CH2CH2), 2.36 (m, 2 · 2H, J = 7.1, CH3CH2CH2-
CO), 3.34 (m, 2H, CHCH2NHFmoc), 4.20 (m, 2H,
NHCO2CH2CH), 4.23 (d, 1H, NHCO2CH2CH), 5.77
(m, 1H, CH2CHOCOPr), 7.10 (d, 2H, C6H4), 7.30–
7.43 (m, 6H, C6H4 and Fmoc), 7.65 (d, 2H, Fmoc)
and 7.88 (d, 2H, Fmoc). 13C NMR (DMSO, 25ꢁC):
13.8, 18.3, 35.7, 36.0, 45.8, 47.1, 66.0, 73.6, 120.6,
122.3, 125.6, 127.5, 136.3, 141.2, 156.6, 172.1 and
Acknowledgements
The authors thank the Academy of Finland for financial
support (grant 75267 to L.K.).
172.6.
Mass
C31H33NO6) = 515.2329 (515.2308).
spectrum
M+
(calculated
for
References
1. Ruffolo, R. R., Jr. Tetrahedron 1991, 47, 9953–9980.
2. Foye, W. O.; Lemke, T. L.; Williams, D. A. In Principles
in Medicinal Chemistry, 4th ed.; Lippincott Williams &
Wilkins: Philadelphia, 1995; pp 345–365.
4.5. Deprotection of (R)-5c and (S)-7c
Compound (S)-7c (0.79g, 1.77mmol, ee 98%) and 2-pro-
panol (0.83mL) were dissolved in 35.3mL of toluene/
THF (3:1) followed by the addition of CAL-B (2.65g,
75mg/mL). The mixture was shaken at 40ꢁC for 16 days
in order to reach 90% conversion. The enzyme was fil-
tered off and the solvent evaporated. Purification by col-
umn chromatography [silica, ethylacetate/hexane (1:1)]
´
3. Nyro¨nen, T.; Pihlavisto, M.; Peltonen, J. M.; Hoffren,
A.-M.; Varis, M.; Salminen, T.; Wurster, S.; Marjama¨ki,
A.; Kanerva, L.; Katainen, E.; Laaksonen, L.; Savola,
J.-M.; Scheinin, M.; Johnson, M. S. Mol. Pharmacol.
2001, 59, 1343–1354.
4. Cho, B. T.; Kang, S. K.; Shin, S. H. Tetrahedron:
Asymmetry 2002, 13, 1209–1217.
followed by crystallization from hexane yielded (S)-5c
5. Fechter, M. H.; Griengel, H. In Enzyme Catalysis in
Organic Synthesis; Drauz, K., Waldmann, H., Eds.; Wiley-
VCH GmbH: Weinheim, 2002; Vol. II, pp 974–989.
6. North, M. Tetrahedron: Asymmetry 2003, 14, 147–176.
7. Kappe, T.; Armstrong, M. D. J. Med. Chem. 1964, 7, 569–
571.
8. Kanerva, L. T.; Rahiala, K.; Va¨nttinen, E. J. Chem. Soc.,
Perkin Trans. 1 1992, 1759–1762.
9. Va¨nttinen, E.; Kanerva, L. T. Tetrahedron: Asymmetry
1995, 6, 1779–1786.
10. Lundell, K.; Kanerva, L. T. Tetrahedron: Asymmetry
1995, 6, 2282–2286.
11. Izumi, T.; Fukaya, K. Bull. Chem. Soc. Jpn. 1993, 66,
1216–1221.
12. Kanerva, L. T.; Sundholm, O. J. Chem. Soc., Perkin
Trans. 1 1993, 2407–2410.
20
D
{0.36g, 0.95mmol, ee >98%, ½aŠ ꢀ4.9 (c 1.0, MeOH)}.
One of the compounds (R)-5c (0.50g, 1.34mmol,
ee 96%) or (S)-5c (0.30g, 0.80mmol, ee 98%) was
dissolved in 5% piperidine in THF and the solution
stirred for 2h. Purification by column chromatography
[silica, dichloromethane/methanol (1:1)] yielded (R)-3
20
D
{0.32g, 2.1mmol, ee >98%, ½aŠ ¼ ꢀ31:6 (c 1.0, MeOH),
20
25
lit.7 ½aŠ ¼ ꢀ37:4 (c 1.0, H2O)} or (S)-3 {0.12g,
D
0.80mmol, ee 98% and ½aŠ ¼ þ35:4 (c 1.0, MeOH),
D
25
D
lit.7 ½aŠ ¼ þ37:2 (c 1.0, H2O)}.
(R)-3: 1H NMR (DMSO, 25ꢁC): 2.64–2.86 (m, 2H,
CHCH2NH2), 3.18 (s, 1H, CHOH), 4.59 (m, 1H,