Communications
Table 1: Lipases and esterases in the desymmetrization of 1a.[a]
Table 2: Desymmetrization of allendiols.[a]
Entry
Diol
R
t [h]
Yield [%]
ee [%]
Entry
Enzyme
Conv. [%]
ee (2a) [%]
2a/3a
1
2
3
4
5
6
7
8
1a
1b
1c
1d
1e
1 f
1g
1h
1i
phenyl
24
96
48
48
30
32
36
32
46
36
96
95 (2a)
69 (2b)
90 (2c)
78 (2d)
93 (2e)
96 (2 f)
92 (2g)
93 (2h)
89 (2i)
98
76
98
84
99
99
99
99
2-methylphenyl
3-methylphenyl
4-methylphenyl
4-chlorophenyl
4-methoxyphenyl
3,4-methylendioxyphenyl
2-naphthyl
2-trimethylsilylethynyl
n-heptyl
cyclohexyl
1
2
3
porcine pancreatic lipase
a-chymotrypsin
cholesterol esterase
C. antarctica lipase B
P. cepacia lipase
P. fluorescens lipase
A. niger lipase
C. rugosa lipase
96
<1
36
94
78
98
1
1
2
4
2
98
n.d.
71
91
63
68
n.d.
39
25
9
98:2
n.d.
92:8
84:16
97:3
71:29
4:96
40:60
64:36
87:13
46:54
4[b]
5
6[c]
7
9
10
11
99
8
9
10
11
1j
1k
77 (2j)
<5 (2k)
92[b]
n.d.
M. javanicus lipase
P. stutzeri lipase
R. oryzae lipase
11
[a] Reaction conditions: allendiol 1 (0.2 mmol), vinyl butyrate (114 mL,
1.0 mmol), and PPL (20 mg) in 1,4-dioxane (1 mL) were incubated at
408C for 24–96 h. ee values were determined by chiral HPLC. n.d.=not
determined. [b] ee value determined by chiral HPLC after derivatization
to the corresponding 3,5-dinitrobenzoate.
[a] Reaction conditions: 1a (8.8 mg, 50 mmol), vinyl butyrate (13 mL,
100 mmol), and enzyme (2.5 mg) in 1,4-dioxane (1 mL) were incubated
for 72 h at 408C. Conversion and ee were determined by chiral HPLC.
n.d.=not determined. [b] 1 h incubation time. [c] 16 h incubation time.
selectivity over reference diol 1a, and monobutyrates 2e and
2 f were obtained in nearly enantiopure form and very high
yields (Table 2, entries 5 and 6). In the same way, allenic diols
bearing bicyclic (1g and 1h) or acetylenic substituents (1i)
were desymmetrized with a high degree of enantioselectivity
(Table 2, entries 7–9). While alkyl-substituted allenols
showed rather low selectivity in the kinetic resolution using
PPL,[6] in the case of allene desymmetrization even n-heptyl
derivative 1j was isolated with an enantiomeric excess of 92%
(Table 2, entry 10). Only cyclohexyl-substituted diol 1k did
not react under these conditions.
As none of the tested proteins exhibited inverse selectivity
compared to PPL (see Table 1), we were also interested in the
lipase-catalyzed hydrolysis of the prochiral dibutyrates 3 as an
enantiocomplementary approach yielding (S)-configured
monobutyrates 2. To our surprise, PPL-catalyzed saponifica-
tion of 3a in aqueous buffer with acetone as cosolvent at 408C
led to fast formation of (S)-2a, albeit at a low level of
enantioselectivity (Scheme 2). This result was in sharp con-
trast to previous studies, where we used similar conditions for
the hydrolysis of racemic allenyl butyrates with excellent
stereocontrol.[9] However, treatment of 3a with lipase from
porcine pancreas under non-aqueous conditions (1-butanol in
heptane) efficiently solved this selectivity issue, and highly
enantioenriched (S)-2a was isolated in good yield. Thus, a
single enzyme can be used to produce either (S)- or (R)-
configured, axially chiral allenic monobutyrates in high
enantiomeric purity by biocatalytic desymmetrization by
simply choosing between synthetic or digestive reaction
conditions.[15]
proteins. While a-CT gave only marginal conversion (Table 1,
entry 2), ChE catalyzed the transesterification with consid-
erable effect, although with inferior selectivity (Table 1,
entry 3). These results indicate that PPL does represent the
catalytically active protein; however, traces of ChE might be
responsible for a certain erosion of enantioselectivity.[13]
Furthermore, we also tested other commercially available
lipases for their ability to catalyze the transesterification of
1a. Among the proteins tested, only lipases from Candida
antarctica (lipase B), Pseudomonas cepacia, and Pseudomo-
nas fluorescens showed synthetically useful conversions
(Table 1, entries 4–6). The outstanding activity of Candida
antarctica lipase B (94% conversion after 60 min), however,
was qualified by a slightly lower enantioselectivity than PPL,
accompanied by a significant degree of overacylation. Other
lipases, namely from Aspergillus niger, Candida rugosa,
Mucor javanicus, Pseudomonas stutzeri, or Rhizopus oryzae,
did not reveal substantial catalytic activity (Table 1, entries 7–
11). In all cases, the (R)-enantiomer was formed preferen-
tially.[14]
To study the scope of the reaction, we conducted
desymmetrizations of a variety of substituted allendiols
through enzymatic transesterification. As the reaction pro-
ceeded relatively slowly under the reaction conditions used in
the enzyme screening (Table 1), substrate concentration,
enzyme loading, and excess of the acylating agent were
further increased. Under these modified conditions, the
desymmetrization of phenyl-substituted diol 1a reached full
conversion after 24 h, and monoester 2a was isolated in 95%
yield and 98% ee (Table 2, entry 1). Tolyl derivatives 1b–1d
showed an interesting phenomenon, as substitution in ortho
or para position of the aryl moiety led to a considerable
decrease in enantioselectivity (Table 2, entries 2 and 4), while
meta derivative 2c was isolated with an excellent enantio-
meric excess of 98% (Table 2, entry 3). In contrast, hetero-
substituents in para position of the arene led to improved
The spectrum of interesting reactions employing axially
chiral allenols is broad; hence, these compounds are increas-
ingly found as intermediates in the synthesis of complex
natural products.[6,16] Exemplarily, the synthetic potential of
the optically active monoesters formed in the allene desym-
metrization is demonstrated in the silver-mediated cyclo-
isomerization of piperonyl derivative 2g, where dihydrofuran
9732
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 9731 –9734