Total synthesis of crispine A enantiomers through a Burkholderia cepacia lipase-catalysed kinetic resolution

Article abstract of DOI:10.1016/j.tetasy.2011.06.026

Both enantiomers of the antitumour-active alkaloid crispine A (ee = 95%) were synthesised through the Burkholderia cepacia lipase-catalysed acylation of the primary hydroxy group of N-Boc-protected 1-(3- hydroxypropyl)-6,7- bis(methyloxy)-1,2,3,4-tetrahydroisoquinoline (±)-3 and the enantioselective hydrolysis of the corresponding O-decanoate (±)-4 [R = (CH₂)₈Me] with a remote, four-atom distant stereogenic centre. High enantioselectivities were observed for the (S)-selective O-acylation with vinyl decanoate in the presence of Et₃N and Na ₂SO₄ in t-BuOMe at 45 °C (E = 68), and for the (S)-selective hydrolysis with H₂O in t-BuOMe at 45 °C (E = 52). The enzymatic resolutions, performed in two steps, afforded the key alcohol and ester enantiomers with high enantiomeric excesses (ee≥94%). Ring-closure reactions of alcohol enantiomers (+)-3 and (-)-3 with thionyl chloride afforded the desired crispine A enantiomers (+)-1 and (-)-1 (ee≥95%).

Full text of DOI:10.1016/j.tetasy.2011.06.026

Tetrahedron: Asymmetry 22 (2011) 1255–1260
Contents lists available at ScienceDirect
Tetrahedron: Asymmetry
journal homepage: www.elsevier.com/locate/tetasy
Total synthesis of crispine A enantiomers through a Burkholderia cepacia
lipase-catalysed kinetic resolution
⇑
}
Eniko Forró, László Schönstein, Ferenc Fülöp
Institute of Pharmaceutical Chemistry, University of Szeged, H-6720 Szeged, Eötvös u 6, Hungary
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 15 June 2011
Accepted 28 June 2011
Both enantiomers of the antitumour-active alkaloid crispine A (ee = 95%) were synthesised through the
Burkholderia cepacia lipase-catalysed acylation of the primary hydroxy group of N-Boc-protected 1-(3-
hydroxypropyl)-6,7-bis(methyloxy)-1,2,3,4-tetrahydroisoquinoline ( )-3 and the enantioselective hydro-
lysis of the corresponding O-decanoate ( )-4 [R = (CH2)8Me] with a remote, four-atom distant stereogenic
centre. High enantioselectivities were observed for the (S)-selective O-acylation with vinyl decanoate in
the presence of Et3N and Na2SO4 in t-BuOMe at 45 °C (E = 68), and for the (S)-selective hydrolysis with
H2O in t-BuOMe at 45 °C (E = 52). The enzymatic resolutions, performed in two steps, afforded the key
alcohol and ester enantiomers with high enantiomeric excesses (ee P 94%). Ring-closure reactions of
alcohol enantiomers (+)-3 and (ꢀ)-3 with thionyl chloride afforded the desired crispine A enantiomers
(+)-1 and (ꢀ)-1 (ee P 95%).
Ó 2011 Elsevier Ltd. All rights reserved.
1. Introduction
6,7-dimethoxy)-1,2,3,4-tetrahydroisoquinoline (+)-3 and (ꢀ)-3,
key intermediates for the preparation of enantiomeric crispine
The plant Carduus crispus has long been used in Chinese folk
medicine for the treatment of colds, stomach problems and rheu-
matism. 8,9-Bis(methyloxy)-1,2,3,5,6,10b-hexa-hydropyrrolo[2,1-
a]isoquinoline 1 (crispine A) and another isoquinoline alkaloid 2,
containing a guanidyl group (crispine E) (Scheme 1), isolated in
2002 from C. crispus by Zhao and co-workers,1 display high biolog-
ical activity against SKOV3, KB and HeLa human cancer cell lines.
In recent years, extensive investigations have been carried out
on the chemistry of tetrahydroisoquinoline alkaloids,2 in view of
their potential pharmaceutical activity. Crispine A was obtained
by synthesis3 before its isolation from C. crispus, and both racemic4
and enantiopure2c,5 forms have been prepared. Crispine E, a basic
constituent of C. crispus, has also been studied because of its prom-
ising pharmaceutical activity.1,6 Itoh and co-workers reported the
synthesis of (ꢀ)-crispine E by using chiral auxiliaries,5b and the
enantioselective synthesis of the antipode (+)-crispine E, via
asymmetric transfer hydrogenation, used as a key step in the
synthesis, has recently been described.7 Almost all of the
approaches are based on an asymmetric synthetic route; only
Turner and co-workers reported a chemo-enzymatic synthesis of
A and E (Scheme 1). In spite of the low to moderate enantiose-
lectivities described in the literature for the lipase-catalysed
asymmetric acylations of primary alcohols,9 and especially those
with a remote stereogenic centre,10 we first planned acylation of
racemic N-Boc-protected 1-(3-hydroxypropyl)-6,7-dimethoxy-
1,2,3,4-tetrahydroisoquinoline ( )-3 in an organic medium. Re-
sults on the lipase-catalysed hydrolysis of O-acylated primary
alcohols with a remote stereogenic centre9a,10e,11 suggested the
possibility of the lipase-catalysed enantioselective hydrolysis of
racemic N-Boc-protected 1-(3-acyloxypropyl)-6,7-dimethoxy-
1,2,3,4-tetrahydroisoquinoline ( )-4.
2. Results and discussion
2.1. Synthesis of ( )-3 and ( )-4
Racemic ( )-3 was prepared in a four-step process, according to
known literature methods.12 First, b-(3,4-dimethoxyphenyl)ethyl-
amine was reacted with c-butyrolactone, and the resulting amide
was subjected to Bischler–Napieralski cyclisation. Next, the
sodium borohydride reduction of the ring-closed 1-(3-hydroxypro-
pyl)-6,7-dimethoxy-3,4-dihydroisoquinoline afforded 1-(3-hydr-
oxypropyl)-6,7-dimethoxy-1,2,3,4-tetrahydroisoquinoline. Finally,
N-Boc protection with di-tert-butyl dicarbonate furnished the
desired ( )-3. Acylation10c of ( )-3 with acetic anhydride gave the
corresponding ( )-4, R = Me, while with decanoyl chloride resulted
in ( )-4, R = (CH2)8Me (see Section 4).
(+)-crispine
A
through the monoamine oxidase-catalysed
deracemisation of racemic crispine A.5d,8
Our aim was to develop new enzymatic strategies for the
synthesis of enantiopure N-Boc-protected 1-(3-hydroxypropyl)-
⇑
Corresponding author. Tel.: +36 62 54 5564; fax: +36 62 54 5705.
E-mail address: fulop@pharm.u-szeged.hu (F. Fülöp).
0957-4166/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetasy.2011.06.026

1256
E. Forró et al. / Tetrahedron: Asymmetry 22 (2011) 1255–1260
MeO
MeO
MeO
MeO
MeO
N
NBoc
MeO
NBoc
NBoc
H
OX
OH
3 4
(-)- and (+)-
1
1
crispine A (-)- or (+)-
( )-3
MeO
MeO
MeO
MeO
MeO
NH
NBoc
MeO
OX
( : X = H; : X = COR)
OCOR
2
3
(+)- and (-)-4
crispine E
( )-4
NH
3
4
H2N
NH
Scheme 1. Planned synthesis of enantiopure 3 and 4.
MeO
MeO
MeO
MeO
acyl donor
(S)
(R)
NBoc
NBoc
+
NBoc
MeO
MeO
lipase
organic solvent
OCOR
OH
OH
(+)-4
(-)-3
( )-3
Scheme 2. Enzymatic acylation of ( )-3.
Table 1
Conversion, enantiomeric excesses (ee) and enantioselectivities (E) of the acylation of ( )-3a
b
b
Entry
Enzyme
Acyl donor (equiv)
Solvent
t (h)
ees (%)
eep (%)
Conversion (%)
43
E
1
2
3
4
5
6
7
8
CAL-Ac
CAL-B
MeCO2CH@CH2 (1.1)
MeCO2CH@CH2 (1.1)
MeCO2CH@CH2 (1.1)
MeCO2CH@CH2 (1.1)
MeCO2CH@CH2 (1.1)
MeCO2CH@CH2 (1.1)
MeCH@CHCO2Me (1.1)
Me(CH2)2CO2CH@CH2 (1.1)
t-BuCO2CH@CH2 (1.1)
Me(CH2)8CO2CH@CH2 (1.1)
Me(CH2)8CO2CH@CH2 (1.1)
Me(CH2)8CO2CH@CH2 (1.1)[+Et3N + Na2SO4]
Me(CH2)8CO2CH@CH2 (3)[+Et3N + Na2SO4]
Me(CH2)8CO2CH@CH2 (4)[+Et3N + Na2SO4]
Me(CH2)8CO2CH@CH2 (1.1)
Me(CH2)8CO2CH@CH2 (1.1)
Me(CH2)8CO2CH@CH2 (1.1)
Me(CH2)10CO2CH@CH2 (1.1)
MeCO2CH@CH2 (1.1)
iPr2O
iPr2O
iPr2O
iPr2O
iPr2O
iPr2O
iPr2O
iPr2O
24
1
24
24
24
1
1
1
1
1
1
1
1
1
1
1
1
1
No reaction
rac
No reaction
No reaction
rac
1
Lipase AKc
Lipase AYc
PPL
No reaction
Lipase PSc
Lipase PSc
Lipase PSc
lipase PSc
Lipase PSc
Lipase PSc
Lipase PSc
Lipase PSc
Lipase PSc
Lipase PSc
Lipase PSc
Lipase PSc
lipase PSc
Lipase PS-IM
Lipase PS-IM
4
8
12
22
35
71
15
19
14
1.6
2.2
7
9
iPr2O
iPr2O
No reaction
10
11
12
13
14
15
16
17
18
19
20
9
10
15
19
33
26
5
85
95
95
96
95
88
67
10
10
14
17
26
23
7
13
43
45
59
53
20
5
t-BuOMe
t-BuOMe
t-BuOMe
t-BuOMe
n-Hexane
Toluene
MeCN
iPr2O
No reaction
8
7
26
86
14
92
7
33
22
14
1.4
30
iPr2O
t-BuOMe
1
1
Me(CH2)8CO2CH@CH2 (1.1)
a
b
c
0.0125 M substrate, 30 mg mLꢀ1 enzyme, 1 mL of organic solvent, acyl donor, 45 °C.
According to HPLC (see Section 4).
Contains 20% (w/w) lipase adsorbed on Celite in the presence of sucrose.
2.2. Enzymatic acylation of ( )-3
a distance of two atoms from the reaction centre,10b–e we at-
tempted the acetylation of ( )-3 with 1.1 equiv of vinyl acetate in
the presence of lipase PS [containing 20% (w/w) lipase adsorbed
on Celite in the presence of sucrose] in iPr2O at 45 °C (Scheme 2,
R = Me).
However, poor levels of enantioselectivity (E <2) were observed
(Table 1, entry 6). Accordingly, preliminary experiments were
needed to determine the optimum conditions for gram-scale
resolution. The effects of enzyme, acyl donor, solvent, temperature
and additive on the enantioselectivity were investigated (Table 1).
Amongst the lipases tested, CAL-A (Candida antarctica lipase A) (en-
try 1), lipase AK (entry 3), lipase AY (from Candida rugosa) (entry 4)
The enantioselectivities of the lipase-catalysed resolutions of
primary alcohols are well known to be considerably lower then
those of secondary alcohols. Few articles have dealt with the asym-
metric acylation of primary alcohols with an even more remotely
located stereocentre. The lipases most commonly used in those
reactions were lipase PS (from Burkholderia cepacia, formerly called
Pseudomonas cepacia lipase) and lipase AK (from Pseudomonas
fluorescens).9,10
Due to the earlier results on the lipase-catalysed enantioselec-
tive (E >200) acylation of alcohols with the stereogenic centre at

E. Forró et al. / Tetrahedron: Asymmetry 22 (2011) 1255–1260
1257
and PPL (porcine pancreatic lipase) (entry 5) did not catalyse the
reaction (no product was detected after 24 h). Acylation was very
fast with CAL-B (Candida antarctica lipase B) (ꢁ43% conversion
after 1 h), but no enantioselectivity was observed (entry 2). Lipase
PS-IM (from Burkholderia cepacia) gave an even faster reaction (33%
conversion after 1 h) than with lipase PS, but again with low
enantioselectivity (E <2) (entry 19).
When the reaction temperature was lowered (from 45 to 30 °C,
and then to 2 °C), the conversion decreased slightly (data not
shown), but no beneficial effect on the enantioselectivity was ob-
served (E <2). Accordingly, subsequent experiments were planned
at 45 °C.
However, there was basically no influence on the enantioselectiv-
ity (E ꢁ50) and for reasons of economy, 30 mg mLꢀ1 lipase PS
was finally chosen for the gram-scale resolution of ( )-3.
In view of the results of the optimisation study, we decided to
perform the gram-scale resolution of ( )-3 in the presence of
30 mg mLꢀ1 lipase PS, with 4 equiv of vinyl decanoate and catalytic
amounts of Et3N and Na2SO4, in t-BuOMe, at 45 °C in two steps. We
planned to stop the acylation at about 40% conversion (ee acylated
product ꢁ95%), and then perform the subsequent reaction until
about 60% conversion (ee unreacted alcohol ꢁ95%). The results
are shown in Table 3 and the Section 4.
In an effort to enhance the enantioselectivity, vinyl acetate was
replaced by other vinyl esters [vinyl butyrate (entry 8), vinyl piva-
late (entry 9), vinyl decanoate (entry 10), vinyl laurate (entry 18)
and i-propenyl acetate (entry 7)]. The best enantioselectivities (E
ꢁ14) were observed for the esters with longer carbon chains, and
vinyl decanoate was therefore chosen as the acyl donor for further
studies.
2.3. Enzymatic hydrolysis of ( )-4
The previous results on the lipase-catalysed deacylation of O-
acylated primary alcohols,9a,10e,11,13 prompted us to start our
experiments on the hydrolysis of racemic N-Boc-protected 1-(3-
acetyloxypropyl)-6,7-dimethoxy-1,2,3,4-tetrahydroisoquinoline
with lipase PS as catalyst and H2O as the nucleophile (Scheme 3,
R = Me).
When catalytic amounts of Et3N and Na2SO4 were added to the
reaction mixture, a slight increase in reaction rate was observed
(entry 12 vs entry 11). The reaction rate was further enhanced
when the lipase PS-catalysed acylation was performed with 3 or
even 4, instead of 1.1 equiv of vinyl decanoate (entries 12–14).
In order to improve the enantioselectivity, we also tested sev-
eral solvents, such as t-BuOMe, n-hexane, toluene and acetonitrile
(MeCN) (entries 11 and 15–17). Overall, n-hexane ensured a
slightly better combination of enantioselectivity and reaction rate
(entry 15) than did t-BuOMe (entry 11), but with regard to the
development of more environmentally friendly routes to enantio-
pure products, t-BuOMe was chosen as a green solvent for further
optimisation and gram-scale reactions.
Since a slow reaction with relatively low enantioselectivity (E
ꢁ5) was observed when the lipase PS-catalysed reaction was per-
formed in t-BuOMe (Table 2, entry 1), we conducted preliminary
screening experiments with enzymes (30 mg mLꢀ1
) including
CAL-A, CAL-B, lipase AY and PPL. Practically no reaction was ob-
served for the majority of the lipases tested (entries 4–6) even after
3 weeks; the only exception was CAL-A, which catalysed the
hydrolysis with a comparable reaction rate to that with lipase PS,
but with a lower E value (entry 3).
Our earlier observations on the enzymatic hydrolysis of N,O-
diacetyl derivatives of cyclic 1,3-amino alcohols10c led us to use
EtOH instead of H2O as a nucleophile for the lipase PS-catalysed
hydrolysis, but the reaction became slower and no beneficial
change in E (Table 2, entry 2) was observed.
Reactions with 50 or 70 mg mLꢀ1 instead of 30 mg mLꢀ1 of li-
pase PS in t-BuOMe in the presence of 1.1 equiv of vinyl decanoate
and catalytic amounts of Et3N and Na2SO4 at 45 °C exhibited
slightly higher reaction rates (after 1 h, conversions of 26% with
30 mg mLꢀ1; 28% with 50 mg mLꢀ1 and 32% with 70 mg mLꢀ1).
The size of the acyl moiety had a marked effect on both the
enantioselectivity and the reaction rate. For example, significantly
higher enantioselectivity and a faster reaction were observed for
MeO
MeO
MeO
MeO
MeO
H2O
(R)
(
)
S
NBoc
NBoc
NBoc
+
lipase
MeO
organic solvent
OCOR
OCOR
OH
( )-4
(-)-4
3
(+)-
Scheme 3. Enzymatic hydrolysis of ( )-4.
Table 2
Conversion, enantiomeric excess (ee) and enantioselectivity (E) of the hydrolysis of ( )-4a
b
b
Entry
Enzyme
R
Nucleophile (equiv)
Solvent
t (h)
ees (%)
eep (%)
Conversion (%)
E
1
2
3
4
5
6
7
8
Lipase PSc
Lipase PSc
CAL-Ac
Me
Me
Me
Me
Me
Me
(CH2)8Me
(CH2)8Me
(CH2)8Me
(CH2)8Me
(CH2)8Me
(CH2)8Me
(CH2)8Me
(CH2)8Me
(CH2)8Me
H2O (0.5)
EtOH (t-BuOMe:EtOH 24:1)
H2O (0.5)
H2O (0.5)
H2O (0.5)
H2O (0.5)
H2O (1)
H2O (1)
H2O (1)
H2O (0)
H2O (4)
t-BuOMe
t-BuOMe
t-BuOMe
t-BuOMe
t-BuOMe
t-BuOMe
t-BuOMe
iPr2O
n-hexane
t-BuOMe
t-BuOMe
t-BuOMe
t-BuOMe
t-BuOMe
t-BuOMe
50
50
50
11
3
3
rac
No reaction
No reaction
32
24
50
20
28
54
rac
10
2
65
63
26
rac
14
5
10
50
5.2
4.5
1.8
1
CAL-B
50
Lipase AYc
PPL
500
500
120
120
120
120
120
120
50
Lipase PSc
Lipase PSc
Lipase PSc
Lipase PSc
Lipase PSc
Lipase PS-IM
CAL-B
96
84
88
96
96
77
rac
72
45
25
22
36
17
23
41
71
12
4
67
15
27
60
64
13
1
9
10
11
12
13
14
15
H2O (1)
H2O (1)
H2O (1)
H2O (1)
Lipase AYc
Lipase AKc
120
120
7
3
a
b
c
0.0125 M substrate, 30 mg mLꢀ1 enzyme, 1 mL of t-BuOMe, 45 °C.
According to HPLC (see Section 4).
Contains 20% (w/w) lipase adsorbed on Celite in the presence of sucrose.

1258
E. Forró et al. / Tetrahedron: Asymmetry 22 (2011) 1255–1260
the hydrolysis of the corresponding decanoate [R = (CH2)8Me] than
for the acetate (R = Me) (Table 2, entry 7 vs entry 1).
Solvents such as iPr2O and n-hexane (Table 2, entries 8 and 9)
were also tested for the lipase PS (30 mg mLꢀ1) catalysed hydroly-
sis of ( )-4 [R = (CH2)8Me], but none of them ensured a better
combination of E or reaction rate than with t-BuOMe (Table 2,
entry 7).
In an attempt to increase the enantioselectivity further, lipase
PS-IM, CAL-B, lipase AY and lipase AK were also tested for the
hydrolysis of ( )-4 [R = (CH2)8Me]. Although all of these lipases
demonstrated similar enantioselectivities (Table 2, entries
12–15), none of them proved to be a better catalyst than lipase
PS (E ꢁ65).
When the reaction was performed with 75 mg mLꢀ1 instead of
30 mg mLꢀ1 of lipase PS in t-BuOMe in the presence of 1 equiv of
H2O at 45 °C, both the reactivity and the enantioselectivity
increased slightly (after 9 days: conversion = 39%, E = 60 for
In order to determine the effects of H2O in the reaction medium
on the enzymatic activity, ( )-4 [R = (CH2)8Me] was hydrolysed in
the presence of different amounts of water (Table 2, entries 7, 10
and 11). Neither the enantioselectivity nor the conversion was
affected when 1 or 4 equiv of water were used. Furthermore, the
hydrolysis took place with comparable E, but was slower without
the addition of any H2O (Table 2, entry 10) (the H2O present in
the enzyme preparation or in the solvent used was sufficient for
the hydrolysis).
30 mg mLꢀ1
,
and conversion = 44%, E = 71 for 75 mg mLꢀ1
enzyme). However, for reasons of economy, 30 mg mLꢀ1 lipase PS
was chosen for the preparative-scale resolutions of ( )-4
[R = (CH2)8Me].
The results of the above experiments suggested the preparative-
scale resolution of ( )-4 [R = (CH2)8Me] in the presence of lipase PS
(30 mg mLꢀ1), with 1 equiv of H2O, in t-BuOMe, at 45 °C, in two
steps. The results are presented in Table 3 and in the Section 4.
Table 3
Lipase PSa-catalysed preparative-scale resolutions of ( )-3 and ( )-4 [R = (CH2)8Me]a
½ ꢂ (CHCl3)
a 2D5
Time (h)
Conversion on workup (%)
Enantiomer
Yield (%)
Isomer
eeb (%)
( )-3
1
3.5
54
43
64
24
60
(S)-4 [R = (CH2)8Me]
(R)-3
(S)-3
31
31
21
28
(S)
(R)
(S)
(R)
94
95
94
96
+52 (c 0.2)
-60 (c 0.5)
+59 (c 0.265)
-58 (c 0.305)
( )-4 [R = (CH2)8Me]
192
(R)-4 [R = (CH2)8Me]
a
30 mg mLꢀ1 lipase PS [containing 20% (w/w) lipase adsorbed on Celite in the presence of sucrose], t-BuOMe, at 45 °C; 4 equiv of vinyl decanoate for the acylation of ( )-3
and 1 equiv of water for the hydrolysis of ( )-4 [R = (CH2)8Me].
b
According to HPLC (see Section 4).
4
4
K2CO3/MeOH
(-)- or (+)-
MeO
MeO
MeO
MeO
SOCl2 / CH2Cl2
5 M NaOH
(-)-3 or (+)- 3
(R)
(S)
or
N
N
3
4
Enzymatic resolutions of ( )- or ( )-
1
(+)-crispine A (+)-
(-)-crispine A (-)-1
Scheme 4. Syntheses of (ꢀ)-1 and (+)-1.
MeO
MeO
N
(
)-1
MeO
MeO
N
1
(-)-
MeO
MeO
N
(+)-1
Figure 1. Enantioseparation of ( )-1 [Chiralcel OD-H column, eluent: n-hexane: (Et2NH 0.1% v/v):iPA (95:5), flow rate: 0.5 mL minꢀ1, detection at 260 nm].

E. Forró et al. / Tetrahedron: Asymmetry 22 (2011) 1255–1260
1259
2.4. Synthesis of (ꢀ)-1 and (+)-1
ganism and adsorbed on a macroporous resin), lipase PS (Pseudo-
monas cepacia) and PPL (porcine pancreas lipase type II) were
from Sigma, and Chyrazyme L-5 (lipase A from Candida antarctica)
was from Novo Nordisk. Before use, lipase PS, lipase AK, lipase AY
and CAL-A (5 g) were dissolved in a tris-HCl buffer (0.02 M; pH 7.8)
in the presence of sucrose (3 g), followed by adsorption on Celite
(17 g). The lipase preparation thus obtained contained 20% (w/w)
lipase. The solvents were of the highest analytical grade.
In a typical small-scale experiment, ( )-3 or ( )-4 (0.025 M solu-
tion) in an organic solvent (1 mL) was added to the enzyme tested
(30, 50 or 75 mg mLꢀ1). The acyl donor (1.1, 3 or 4 equiv) or H2O
(0.5, 1 or 4 equiv) was added, followed in some cases by catalytic
amounts of Et3N and Na2SO4. The mixture was shaken at 2, 30,
45 or 60 °C.
The O-acylated enantiomers (+)-4 [R = (CH2)8Me] and (ꢀ)-4
[R = (CH2)8Me] were hydrolysed to the corresponding alcohols
(+)-3 and (ꢀ)-3 in K2CO3/MeOH, at room temperature, without a
loss in enantiopurity. The enantiopure (+)-3 and (ꢀ)-3 were then
converted to the corresponding enantiopure crispine A (Scheme 4).
In contrast with the literature cyclisation procedure,2c (+)-3 and
(ꢀ)-3 did not need deprotection. Instead they underwent cyclisa-
tion in the presence of thionyl chloride to furnish (+)-1 and (ꢀ)-1
(ee P 94%) in 1 step in relatively high yields (Section 4).
It should be mentioned that (+)-3 and (ꢀ)-3 can also serve as
starting enantiomers for the synthesis of enantiomerically pure cri-
spine E. As an example, Itoh and co-workers5b described the syn-
thesis of (ꢀ)-crispine E (ꢀ)-2 from aminoalcohol (+)-3 through a
Mitsunobu reaction, followed by removal of the Boc groups.
In order to determine the exact ee values for the crispine A
enantiomers prepared, a new direct HPLC method was devised
(Fig. 1).
The ee values for the amino alcohol, amino ester and crispine A
enantiomers produced were determined by HPLC [Chiralpak IA col-
umn, eluent: n-hexane:iPA (85:15), flow rate: 0.5 mL minꢀ1, detec-
tion at 260 nm]; retention times (min) for (ꢀ)-3: 25.61, for (+)-3:
41.61, for (ꢀ)-4 [R = (CH2)8Me]: 12.43 and for (+)-4 [R = (CH2)8Me]:
15.6, and [Chiralcel OD-H column, eluent: n-hexane (Et2NH 0.1% v/
v):iPA (95:5), flow rate: 0.5 mL minꢀ1, detection at 260 nm] for
(ꢀ)-1: 33.61 and for (+)-1: 49.84.
2.5. Absolute configurations
The stereochemistry of the crispine A enantiomers was con-
Optical rotations were measured on a Perkin-Elmer 341 polar-
imeter. The results of elemental analyses (CHNS) corresponded clo-
sely (within 0.3%) with the calculated data in all cases.
firmed by comparing the
a
values with the literature data1,5a (see
Section 4). Thus, both the lipase PS-catalysed acylation of ( )-3
and the hydrolysis of ( )-4 [R = (CH2)8Me] displayed (S)-selectivity.
4.2. Synthesis of ( )-4 [R = (CH2)8Me]
3. Conclusions
Alcohol ( )-3 (500 mg, 1.42 mmol) was dissolved in CHCl3
(50 mL). Decanoyl chloride (297.5 mg, 1.56 mmol) was then added.
The solution was stirred at room temperature for overnight. After
evaporation, the residue was chromatographed on silica. Elution
with AcOEt:hexane (1:5) afforded ( )-4 [R = (CH2)8Me] as a light-
yellow oil (589 mg, 82%). The 1H NMR (400 MHz, CDCl3, 50 °C,
TMS) spectroscopic data for ( )-4 [R = (CH2)8Me]: d (ppm) = 0.91
(3H, t, J = 6.4 Hz, CH2CH3), 1.3 (16H, br s), 1.51 (9H, s, C(CH3)3),
1.61–1.84 (4H, m, CH2CH2CH2OH), 2.62–2.66 (1H, m, ArCH2),
2.81–2.9 (1H, m, ArCH2), 3.2 (1H, br s, CH2N), 3.86 (3H, s, OCH3),
3.88 (3H, s, OCH3), 4.11–4.19 (3H, m, CH2OCOCH2 and CH2N over-
lapping), 5.1 (1H, br s, CHN), 6.61 (2H, s, Ar). Anal. Calcld for
A new total synthesis of crispine A enantiomers has been devel-
oped via the Burkholderia cepacia lipase-catalysed acylation of the
primary hydroxy group of N-Boc-protected 1-(3-hydroxypropyl)-
6,7-bis(methyloxy)-1,2,3,4-tetrahydroisoquinoline ( )-3 or enan-
tioselective hydrolysis of the corresponding O-decanoate ( )-4
[R = (CH2)8Me], with a remote, four-atom distant stereogenic cen-
tre. The (S)-selective acylation of ( )-3 was performed in two steps
with vinyl decanoate (0.5 equiv), using lipase PS in t-BuOMe, at
45 °C (E = 68), while the lipase PS-catalysed S-selective two-step
hydrolysis of ( )-4 [R = (CH2)8Me] required 1 equiv of H2O as
nucleophile, and the reaction was performed in t-BuOMe, at
C29H47NO6: C, 68.88; H, 9.37; N, 2.77. Found: C, 68.59; H, 9.44;
45 °C (E = 52). The enantiomers of
(ee P 94%) were isolated in good yields (P21%), with separation
by column chromatography. Ester enantiomers (+)-4
3 and 4 [R = (CH2)8Me]
N, 2.60.
4.3. Gram-scale resolution of ( )-3
[R = (CH2)8Me] and (ꢀ)-4 [R = (CH2)8Me] were hydrolysed to the
corresponding alcohols (+)-3 and (ꢀ)-3 in K2CO3/MeOH without a
loss of enantiopurity. A direct ring-closing reaction of (+)-3 and
(ꢀ)-3 without deprotection resulted in the desired enantiomers
of crispine A (ee P 95%).
It should be mentioned that, although crispine A enantiomers
have previously been prepared by asymmetric routes, to the best
of our knowledge they have not yet been obtained via a strategy
involving an enzyme-catalysed kinetic resolution. In parallel, a
new HPLC method has been developed for the enantioseparation
of racemic crispine A on a Chiralcel OD-H column.
Racemic 3 (500 mg, 1.42 mmol) was dissolved in t-BuOMe
(40 mL). Lipase PS (1.2 g, 30 mg mLꢀ1), vinyl decanoate (0.65 mL,
2.84 mmol), Et3N (14.37 mg, 0.142 mmol) and Na2SO4 (catalytic
amount) were added and the mixture was shaken in an incubator
shaker at 45 °C for 1 h. The enzyme was filtered off at 43% conver-
sion (eep = 94%), and the solvent was evaporated off. The residue
was chromatographed on silica, with elution with n-hexane:EtOAc
(5:1) affording the ester (S)-4 [R = (CH2)8Me] {223 mg, 31%;
½
a 2D5
ꢂ
¼ þ52 (c 0.2, CHCl3); a light-yellow oil; ee = 94%} and the
enantiomerically enriched alcohol (244 mg, 0.69 mmol;
ee = 70%). In order to obtain with high ee, the alcohol
3
3
(ee3 = 70%) (244 mg) was subjected to a second enzymatic resolu-
tion in t-BuOMe (40 mL), using lipase PS (400 mg, 10 mg mLꢀ1), vi-
nyl decanoate (0.65 mL, 2.84 mmol), Et3N (7.2 mg) and a catalytic
amount of Na2SO4 at 45 °C. The reaction was stopped after 3.5 h
(eeS = 95%). After a second column chromatography, the desired
4. Experimental
4.1. Materials and methods
(R)-3 {156 mg, 31%; ½a D25
¼ ꢀ60 (c 0.5, CHCl3); a light-yellow oil,
ꢂ
Lipase PS-IM (from Burkholderia cepacia, immobilised on diato-
maceous earth) was a gift from Amano Enzyme Europe Ltd, Lipase
AK (Pseudomonas fluorescens) was from Amano Pharmaceuticals,
Lipolase (lipase B from Candida antarctica, produced by submerged
fermentation of a genetically modified Aspergillus oryzae microor-
ee = 95%} was obtained.
The 1H NMR (400 MHz, CDCl3, 50 °C, TMS) spectroscopic data
for (R)-3: d (ppm) = 1.51 [9H, s, C(CH3)3], 1.66–1.87 (4H, m,
CH2CH2CH2OH), 2.66 (1H, dt, J = 15.6, 3.7 Hz, ArCH2), 2.81–2.89

1260
E. Forró et al. / Tetrahedron: Asymmetry 22 (2011) 1255–1260
(1H, m, ArCH2), 3.2 (1H, br s, CH2N), 3.72–3.78 (2H, m, CH2OH),
3.86 (3H, s, OCH3), 3.88 (3H, s, OCH3), 4.15 (1H, br s, CH2N), 5.15
(1H, br s, CHN), 6.6 (1H, s, Ar), 6.64 (1H, s, Ar). Anal. Calcd for
88–89 °C, lit.5i mp 86–88 °C; ½a 2D5
ꢂ
¼ ꢀ61 (c 0.28, CHCl3), lit.5i
½
a 2D3
ꢂ
¼ ꢀ95 (c 1.3, CHCl3), ee = 96%}, as a colourless solid.
The 1H NMR (400 MHz, CDCl3, 25 °C, TMS) spectroscopic data
C
19H29NO5: C, 64.93; H, 8.32; N, 3.99. Found: C, 64.82; H, 8.03;
N, 4.21.
The 1H NMR (400 MHz, CDCl3, 50 °C, TMS) spectroscopic data
for (R)-1 were similar to those for (S)-1: d (ppm) = 1.69–1.95 (3H,
m, CH2CH2CH2N), 2.56–2.58 (1H, m, CH2CH2N), 2.56–2.65 (2H, m,
CH2N), 2.69–2.77 (1H, m, ArCH2), 3.05–3.22 (3H, m, CH2N and
ArCH2 overlapping), 3.41–3.43 (1H, m, CHN), 3.86 (3H, s, OCH3),
3.88 (3H, s, OCH3), 6.6 (1H, s, Ar), 6.64 (1H, s, Ar). Anal. Calcd for
for (S)-4 [R = (CH2)8Me] were similar to those for ( )-4
[R = (CH2)8Me]. Analysis: found for (S)-4 [R = (CH2)8Me]: C, 68.88;
H, 9.37; N, 2.77.
C14H19NO2: C, 72.07; H, 8.21; N, 6.00. Found for (R)-1: C, 71.88;
H, 8.29; N, 6.09. Found for (S)-1: C, 72.00; H, 8.28; N, 6.02.
4.4. Gram-scale resolution of ( )-4 [R = (CH2)8Me]
Acknowledgements
Racemic 4 [R = (CH2)8Me] (400 mg, 0.79 mmol) was dissolved in
t-BuOMe (40 mL). Lipase PS (1.2 g, 30 mg mLꢀ1) and water (14
lL,
We are grateful to the Hungarian Research Foundation (Grants
K 71938 and T 049407), and a Bolyai Fellowship for E.F. Thanks are
due to Mrs. Judit Árva for her technical assistance. We would also
like to express thanks to TÁMOP (4.2.1/B-09/1/KONV-2010-0005).
0.79 mmol) were added and the mixture was shaken at 45 °C for
54 h. The enzyme was filtered off at 24% conversion (eep = 95%).
Column chromatography under the above-mentioned conditions
afforded (S)-3 [59 mg, 21%; ½a D25
¼ þ59 (c 0.265, CHCl3); a light-
ꢂ
References
yellow oil; ee = 95%] and enantiomerically enriched 4 (259 mg,
0.51 mmol; ee = 30%). The ester (ee4 = 30%) was subjected to a sec-
ond enzymatic hydrolysis in t-BuOMe (40 mL) with lipase PS
1. Zhang, Q.; Tu, G.; Zhao, Y.; Cheng, T. Tetrahedron 2002, 58, 6795–6798.
2. (a) Rozwadowska, M. D. Heterocycles 1994, 39, 903–931; (b) Chrzanowska, M.;
Rozwadowska, M. D. Chem. Rev. 2004, 104, 3341–3370; (c) Miyazaki, M.; Ando,
N.; Sugai, K.; Seito, Y.; Fukuoka, H.; Kanemitsu, T.; Nagata, K.; Odanaka, Y.;
Nakamura, K. T.; Itoh, T. J. Org. Chem. 2011, 76, 534–542.
3. Orito, K.; Matsuzaki, T.; Suginome, H. Heterocycles 1988, 27, 2403–2412.
4. (a) Knölker, H.-J.; Agarwal, S. Tetrahedron Lett. 2005, 46, 1173–1175; (b) King, F.
D. Tetrahedron 2007, 63, 2053–2056; (c) Chiou, W.-H.; Lin, G.-H.; Hsu, C.-C.;
ChaterpaulI, S. J.; Ojima, I. Org. Lett. 2009, 11, 2659–2662; (d) Coldham, I.; Jana,
S.; Watson, L.; Martin, N. G. Org. Biomol. Chem. 2009, 7, 1674–1679; (e) Yioti, E.
G.; Mati, I. K.; Arvanitidis, A. G.; Massen, Z. S.; Alexandraki, E. S.; Gallos, J. K.
Synthesis 2011, 142–146.
5. (a) Szawkalo, J.; Zawadzka, A.; Wojtasiewicz, K.; Leniewski, A.; Maurin, J. K.;
Drabowicz, J.; Czarnocki, Z. Tetrahedron: Asymmetry 2005, 16, 3619–3621; (b)
Kanemitsu, T.; Yamashita, Y.; Nagata, K.; Itoh, T. Heterocycles 2007, 74, 199–
203; (c) Szawkalo, J.; Czarnocki, S. J.; Zawadzka, A.; Wojtasiewicz, K.;
Leniewski, A.; Maurin, J. K.; Czarnocki, Z.; Drabowicz, J. Tetrahedron:
Asymmetry 2007, 18, 406–413; (d) Bailey, K. R.; Ellis, A. J.; Reiss, R.; Snape, T.
J.; Turner, N. J. Chem. Commun. 2007, 3640–3642; (e) Allin, S. M.; Gaskell, S. N.;
Towler, J. M. R.; Page, P. C. B.; Saha, B.; McKenzie, M. J.; Martin, W. P. J. Org.
Chem. 2007, 72, 8972–8975; (f) Hou, G.-H.; Xie, J.-H.; Yan, P.-C.; Zhou, K.-L. J.
Am. Chem. Soc. 2009, 131, 1366–1367; (g) Adams, H.; Elsunaki, T. M.; Ojea-
Jiménez, I.; Jones, S.; Meijer, A. J. H. M. J. Org. Chem. 2010, 75, 6252–6262; (h)
Amat, M.; Elias, V.; Llor, N.; Subrizi, F.; Molins, E.; Bosch, J. Eur. J. Org. Chem.
2010, 4017–4026; (i) Gurram, M.; Gyimothy, B.; Wang, R.-F.; Lam, S.-Q.;
Ahmed, F.; Herr, R. J. J. Org. Chem. 2011, 76, 1605–1613.
(300 mg) and water (11 lL, 0.61 mmol), at 45 °C. The reaction
was stopped after 192 h (ees = 96%). By using the separation proce-
dure described above, the desired (R)-4 [R = (CH2)8Me] {110 mg,
28%); ½a 2D5
¼ ꢀ58 (c 0.305, CHCl3); a light-yellow oil, ee = 96%}
ꢂ
was obtained.
The 1H NMR (400 MHz, CDCl3, 50 °C, TMS) spectroscopic data
for (S)-3 were similar to those for (R)-3. Analysis: found for (S)-3:
C, 64.77; H, 8.28; N, 4.11.
The 1H NMR (400 MHz, CDCl3, 50 °C, TMS) spectroscopic data
for (R)-4 [R = (CH2)8Me] were similar to those for (S)-4
[R = (CH2)8Me] and ( )-4 [R = (CH2)8Me]. Analysis: found for (R)-4
[R = (CH2)8Me]: C, 68.71; H, 9.26; N, 2.62.
4.5. Deacylation of (S)-4 [R = (CH2)8Me] and (R)-4 [R = (CH2)8Me]
A mixture of (S)-4 [R = (CH2)8Me] (210 mg, 0.415 mmol) and
K2CO3 (0.14 g, 0.96 mmol) in MeOH (15 mL) was refluxed for 8 h.
After evaporation, the residue was dissolved in H2O (20 mL) and
extracted with Et2O. The organic phase was dried (Na2SO4), filtered
6. Xie, W. D.; Li, P. L.; Jia, Z. J. Pharmazie 2005, 60, 233–236.
7. Czarnocki, S. J.; Wojtasiewicz, K.; Józ´wiak, A. P.; Maurin, J. K.; Czarnocki, Z.;
and evaporated. The product (S)-3 {[0.13 g, 89%; ½a D25
¼ þ60 (c
ꢂ
Drabowicz, J. Tetrahedron 2008, 64, 3176–3182.
0.31, CHCl3); ee = 94%} was obtained as a light-yellow oil. Similarly,
8. Ellis, A. J.; Reiss, R.; Snape, T. J.; Turner, N. J. In Practical Methods for Biocatalysis
and Biotransformations; Whittall, J., Sutton, P. W., Eds.; John Wiley and Sons,
2009; pp 319–323.
(R)-4 [R = (CH2)8Me] (0.14 g, 0.96 mmol) afforded (R)-3 {[0.082 g,
84%; ½a 2D5
¼ ꢀ60 (c 0.2, CHCl3); ee = 96%}.
ꢂ
9. (a) Ghanem, A.; Aboul-Enein, H. Y. Tetrahedron: Asymmetry 2004, 15, 3331–
3351; (b) Bornscheuer, U. T.; Kazlauskas, R. J. Hydrolases in Organic Synthesis;
Wiley-VCH, 2006. pp 106–112; (c) Högberg, H.-E. Exploiting Enantioselectivity
of Hydrolases in Organic Solvents Selectivity. In Organic Synthesis with Enzymes
in Non-aqueous Media; Carrea, G., Riva, S., Eds.; Wiley-VCH: Weinheim, 2008;
pp 75–86. Chapter 4; (d) Chenevert, R.; Morin, P.; Peichat, N. Transesterification
and Hydrolysis of Carboxylic Acid Derivatives, Alcohols, and Epoxides. In
Asymmetric Organic Synthesis with Enzymes; Gotor, V., Alfonso, I., García-
Urdiales, E., Eds.; Wiley-VCH, 2008; pp 133–171.
10. (a) Jouglet, B.; Rousseau, G. Tetrahedron Lett. 1993, 34, 2307–2310; (b) Csomós,
P.; Fülöp, F.; Kanerva, L. T. Tetrahedron: Asymmetry 1996, 7, 1789–1796; (c)
Kámán, J.; Forró, E.; Fülöp, F. Tetrahedron: Asymmetry 2001, 12, 1881–1886; (d)
Forró, E.; Árva, J.; Fülöp, F. Tetrahedron: Asymmetry 2001, 12, 643–649; (e)
Forró, E.; Fülöp, F. Tetrahedron: Asymmetry 2001, 12, 2351–2358; (f) Isleyen, A.;
Tanyeli, C.; Dogan, Ö. Tetrahedron: Asymmetry 2006, 17, 1561–1567; (g)
Valente, E.; Gomes, J. R. B.; Moreira, R.; Iley, J. J. Org. Chem. 2004, 69, 3359–
3367; (h) Sabbani, S.; Hedenström, E.; Andersson, J. Tetrahedron: Asymmetry
2007, 18, 1712–1720.
4.6. Synthesis of crispine A enantiomers (R)-1 and (S)-1
Alcohol (ꢀ)-3 or (+)-3 (100 mg, 0.28 mmol) was dissolved in
CH2Cl2 (7 mL). Thionyl chloride (100 mg, 0.84 mmol) was added.
After stirring for 2 h at 40 °C, the reaction mixture was cooled to
below 20 °C and a solution of NaOH (5 N) was added until pH
P9. After a further stirring for 1 h at room temperature, the reac-
tion was quenched with 10 mL water. The CH2Cl2 layer was dried
over anhydrous Na2SO4 and evaporated. The oily residue was puri-
fied by column chromatography, with CHCl3:MeOH (9:1) as eluent.
The product was crystallised from Et2O as the desired crispine A
enantiomer (R)-1 {55 mg (83%); mp 87–89 °C, lit.1 mp 88–89 °C;
lit.5i mp 86–88 °C; ½a 2D5
ꢂ
¼ þ82 (c 0.255, MeOH), lit.5e
½
a 2D5
ꢂ
¼
11. Rosen, C. T.; Haufe, G. Tetrahedron: Asymmetry 2002, 13, 1397–1405.
12. (a) Lázár, L.; Fülöp, F.; Bernáth, G.; Mattinen, J. Org. Prep. Proc. 1993, 25, 91–97;
(b) Campbell, J. A.; Lee, W. K.; Rapoport, H. J. Org. Chem. 1995, 60, 4602–4616.
13. Nagai, H.; Shiozawa, T.; Achiwa, K.; Terao, Y. Chem. Pharm. Bull. 1993, 41, 1933–
1938.
þ43:9 (MeOH), lit.1
½
a 2D5
ꢂ
¼ þ91:0 (MeOH), ½a D25
¼ þ60 (c 0.255,
ꢂ
CHCl3), lit.5a
½
a 2D3
ꢂ
¼ þ100:4 (c 1, CHCl3), lit.5i
½
a 2D3
ꢂ
¼ þ96:9 (c 1.1,
CHCl3), ee = 95%} or (S)-1 {52 mg, 78 %; mp 87–89 °C, lit.1,5e mp