On the stereoselectivity of asymmetric strecker synthesis in a cyclohexane system: Synthesis of optically active cis- and trans-1-amino-2- hydroxycyclohexane-1-carboxylic acids

Article abstract of DOI:10.1246/bcsj.79.768

Stereoselective synthesis of both optically active cis- and trans-1-amino-2-hydroxycyclohexane-1-carboxylic acids (Ahhs) 1a and 1b was accomplished by an asymmetric version of the Strecker synthesis. Stereochemical aspects of their chiral induction proces

Full text of DOI:10.1246/bcsj.79.768

768 Bull. Chem. Soc. Jpn. Vol. 79, No. 5, 768–774 (2006)
ꢀ 2006 The Chemical Society of Japan
On the Stereoselectivity of Asymmetric Strecker Synthesis
in a Cyclohexane System: Synthesis of Optically Active cis-
and trans-1-Amino-2-hydroxycyclohexane-1-carboxylic Acids
ꢀ
Taiki Umezawa, Masaki Kawasaki, Kosuke Namba, and Yasufumi Ohfune
Tetsuro Shinada, Tadashi Kawakami, Hiroshi Sakai, Hiromi Matsuda,
ꢀ
Graduate School of Science, Osaka City University, 3-3-138 Sugimoto, Sumiyoshi-ku, Osaka 558-8585
Received November 1, 2005; E-mail: Ohfune@sci.osaka-cu.ac.jp
Stereoselective synthesis of both optically active cis- and trans-1-amino-2-hydroxycyclohexane-1-carboxylic acids
(Ahhs) 1a and 1b was accomplished by an asymmetric version of the Strecker synthesis. Stereochemical aspects of their
chiral induction processes are investigated.
Functionalization of the hydroxy group of serine or threo-
nine in peptides and proteins is an important biological proc-
ess. For instance, their phosphorylation and glycosidation are
associated with cell differentiation and signal transduction.1
Furthermore, the hydroxy moiety plays a key role as a catalytic
site in a class of serine proteases.2 The above reactions are cat-
alyzed by an enzyme and performed in a spatially limited en-
zyme pocket. In this circumstance, the conformation around
the hydroxy group is fixed to facilitate these chemical modifi-
cations. Therefore, exploitation of the conformational demand
of serine or threonine is an important subject not only for the
elucidation of the enzyme reaction mechanisms at the molecu-
lar level, but also for the rational design of effective ligands.3
In this context, we have synthesized cis- and trans-1-amino-2-
hydroxycyclohexane-1-carboxylic acid (cis-Ahh 1a and trans-
Ahh 1b), which can be viewed as conformational variants of
serine or threonine.4 Their effect on the peptide conformation
and biological activity has been well documented by our recent
studies,5 i.e., incorporation of Ahh isomers into the Gly(2) res-
idue of leucine–enkephalin (Leu–Enk) revealed that (1) each
Ahh isomer restricted the local conformation of Leu–Enk to
a ꢀ-turn conformation and (2) the cis-Ahh isomer 2a, whose
hydroxy group is fixed to an equatorial orientation, exhibited
17 times more potent binding activity to ꢁ-opioid receptors
of cultured rat brain membrane than that of Leu–Enk, while
the potency of its trans-isomer 2b having an axial orientation
of the hydroxy group in the peptide was 1/100 of that of 2a
(Scheme 1).
The Ahh isomers 1 have been prepared from the 2-acyl-
oxycyclohexanone 3a possessing L-Phe as the acyloxy group
using an asymmetric version of the Strecker synthesis.4,6 The
efficiency of this method has been demonstrated by the synthe-
sis of a number of acyclic ꢀ-hydroxy-ꢂ,ꢂ-disubstituted ꢂ-
amino acids and natural products in view of its high enantio-
and diastereoselectivity, short-step conversion, and overall
yields.6 On the other hand, the synthesis of Ahh isomers in-
volves the cyanide addition to a fused-bicyclic ketimine inter-
mediate. This transformation was performed by the use of
NaCN in 2-PrOH from 3a without isolation of the ketimines
4a and 5a to give a mixture of the cis- and trans-ꢂ-amino ni-
triles 7a and 8a in 85% yield with moderate diastereoselectiv-
ity (cis/trans = 2–4:1). The major isomer was isolated by re-
crystallization and was converted to the cis-Ahh 1a. The minor
trans-isomer 1b was obtained by chromatographic separation
after conversion of the mixture to the corresponding ꢂ-imino
nitriles 9a and 10a, followed by its hydrolysis4 (Scheme 2).
However, in the previous synthesis, the stereochemical out-
come of the cyanide addition to the ketimine intermediate of
the cyclohexane system was not investigated in detail and sup-
ply of the trans-isomer 1b was required. In this report, we wish
to describe our extensive studies regarding (1) the effect of the
chirality transferring groups employing L-Phe-, L-Trp-, L-Leu-,
and L-t-Leu esters 3a–3d on the ratio of both the ketimines
(4 vs 5) and the ꢂ-amino nitriles (7 vs 8), (2) stereoselective
synthesis of 1b, (3) equilibrium experiments between 4 and
5, and (4) conformational analysis of 4 and 5.
H
HOOC
NH2
2S
OH
H
X
OH
H
1
N
2
O
R
Y
H
HN
H
H
N
Gly-Phe-Leu-OH
HO
H
O
H
Try NH
1a (1R, 2S)-cis-Ahh
O
O
NH2
NH Leu
X = OH, Y = H
1b (1R, 2R)-trans-Ahh
2a (IC50 = 0.01 nM)
2b (IC50 = 1.0 nM)
X = H, Y = OH
Scheme 1.
Published on the web May 12, 2006; DOI: 10.1246/bcsj.79.768

T. Shinada et al.
Bull. Chem. Soc. Jpn. Vol. 79, No. 5 (2006)
769
Bn
O
Bn
Bn
Bn
1) TFA
CH2Cl2
BocHN
O
O
O
O
O
TFA, NaCN
N
HN
N
O
6S
O
O
6R
2-PrOH
2) Na2CO3
MgSO4
CH3CN
6
(2RS)-3a
R:S=1:1
(6S)-4a
(6R)-5a
Imine-enamine equilibrium
Bn
Bn
Bn
O
O
O
HN
1
N
HN
NC
NC
NC
t-BuOCl
concd HCl
O
6
O
O
+
1a or 1b
6
Et3N
110 °C, 14h
cis-7a (2~4 : 1) trans-8a
(1S,6S)
(6R)-9a
(6S)-10a
(1S,6R)
Scheme 2.
Table 1. Formation of ꢂ-Amino Nitriles 7, 8, and 11 from ꢂ-Amino Acid Esters 3
1S
NC
condition A:
TFA, NaCN
2-PrOH
or
R
6
R
R
O
H
N
cis-7
(1S,6R)
O
O
O
N
HN
N
O
O
6S
(2RS)-3
R:S=1:1
R
O
O
O
6R
+
condition B:
ZnCl2, TMSCN
2-PrOH
CN
H
N
R
(6S)-4
6
(6R)-5
O
trans-8
+
(1S,6S)
(1R)-
isomers 11
Yields and ratios of 7:8:11 [Yield/%]
Substrate 3a–3d
R
Ratio of
ketimines 4:5
Condition A (TFA)
66:34:N.D. [96]aÞ
Condition B (ZnCl2)
41:59:N.D. [89]aÞ
a
Bn (Phe)
50:50
50:50
H2C
(Trp)
b
77:23:N.D. [56]aÞ
33:67:N.D. [48]aÞ
N
H
c
d
CH2i-Pr (Leu)
t-Bu (t-Leu)
50:50
20:80
39:42:19 [86]
16:82:2 [88]
18:71:11 [73]
5:90:5 [80]
1
a) (1R)-11a and -11b were not detected (N.D.) by H NMR.
from the Boc-t-Leu ester 3d; (2) the Boc-L-Phe and Boc-L-
Trp esters 3a and 3b bearing an aromatic ring afforded the
cis-ꢂ-amino nitriles 7a and 7b as the major products under
condition A; (3) preferential formation of the trans-ꢂ-amino
nitriles 8a–8d was observed under condition B. Among them,
the t-Leu ester 3d gave the highest trans-selectivity (trans-8d/
cis-7d = 18:1); and (4) the cyanide addition to the ketimines
having an alkyl side chain gave a small amount of the (1R)-iso-
mers 11c and 11d, while such isomers were not produced at all
upon cyanide addition to the ketimines possessing an aromatic
group derived from 3a and 3b.
Results and Discussion
Effects of the Chirality Transferring Group and Stereo-
selective Synthesis of trans-Ahh (1b). Upon the asymmetric
Strecker process for Ahh using L-Phe as the chirality transfer-
ring group, we initially presumed that the ketimine intermedi-
ates were under a rapid equilibrium between (6S)-4a and (6R)-
5a via the enamine 6a, where its composition would reflect the
ratio of the resulting ꢂ-amino nitriles 7a and 8a. In this study,
the intermediary ketimines were isolated and the ratios of the
1
mixture of (6S)-4 and (6R)-5 were analyzed by their H NMR
data. The cyanide addition was performed using two reac-
tion conditions: conditions A (NaCN, TFA) and B (TMSCN,
ZnCl2).7 These results are depicted in Table 1 and are summa-
rized by the following points: (1) 1H NMR analysis of the iso-
lated ketimines revealed that they consist of a 1:1 mixture of
the (6S)- and (6R)-isomers 4 and 5, indicating that each keti-
mine ratio is not in accord with that of the ꢂ-amino nitriles
(cis-7/trans-8), except for the case of the ketimine derived
The structures of the ꢂ-amino nitriles 7b and 7c and 8b and
8c were assigned as depicted in Table 1 by comparison of their
1H NMR data with those of 7a and 7d and 8a and 8d (vide
infra). The C3 proton of all of the cis-adducts 7 was observed
at a higher magnetic field than that of trans-8, and the C6 pro-
ton appeared in an opposite relationship (Table 2).
Upon oxidation of cis-7a to the corresponding ꢂ-imino
nitrile 9a, the use of DABCO/t-BuOCl (method A) was found

770 Bull. Chem. Soc. Jpn. Vol. 79, No. 5 (2006)
Asymmetric Strecker Synthesis
1
Table 2. Selected H NMR Data of ꢂ-Amino Nitriles 7 and 8aÞ
9
CN
NC
H
3
9
6
7
O
H
3
R
N
N
O
7
6
O
O
R
cis-7
trans-8
ꢂ-Amino nitriles
3-H
6-H
7-Hax
9-Hax
Phe-7a
Phe-8a
Trp-7b
Trp-8b
Leu-7c
Leu-8c
t-Leu-7d
t-Leu-8d
4.14 [ddd, 8.8, 4.9, 3.7]
4.22 [dt, 9.8, 3.9]
4.18 [dt, 8.7, 3.6]
4.31 [dd, 9.6, 3.0]
3.89 [dd, 8.8, 4.2]
4.06 [dd, 8.4, 4.0]
3.62 [s]
4.34 [ddd, 10.7, 4.8, 2.1]
3.96 [dd, 12.3, 4.3]
4.33 [dd, 10.2, 4.3]
3.98 [dd, 12.2, 4.1]
4.47 [dd, 8.8, 4.4]
4.10 [dd, 12.2, 4.1]
4.38 [dd, 10.0, 4.9]
4.07 [dd, 12.2, 4.4]
1.4–1.6 [m]
1.4–1.6 [m]
1.4–1.6 [m]
1.4–1.6 [m]
1.5–1.7 [m]
1.6–1.8 [m]
1.5–1.7 [m]
1.5–1.7 [m]
1.2–1.4 [m]
1.2–1.4 [m]
1.2–1.4 [m]
1.2–1.4 [m]
1.3–1.5 [m]
1.3–1.5 [m]
1.3–1.5 [m]
1.3–1.5 [m]
3.75 [s]
a) Chemical shifts and coupling constants are shown in ꢁ (ppm) and J (Hz), respectively.
t-Bu
HN
method A
R
O
DABCO, t-BuOCl
CH2Cl2, 0 °C
O
N
NC
NC
concd HCl
cis-1a (92% from 7a)
trans-1b (90% from 8d)
cis-7a or
trans-8d
OH
O
+
6
or
110 °C, 14h
method B:
O3, AcOEt, −78 °C
(6S)-amide
(6R)-9a (R = Bn)
(6S)-10d (R = t-Bu)
method A: (6R)-9a, 97%; (6S)-10d, 0%
method B: (6S)-10d, 59%; (6S)-amido, 34%
Scheme 3.
Table 3. Equilibrium Experiments of 4a and 5a in 2-Propa-
aÞ
to be superior to the previous Et3N/t-BuOCl oxidation (60–
90% yield)4,6 in view of its shortened reaction period (2 h)
and high and reproducible yield (97–100%).8 In particular, the
cis-Ahh 1a was obtained in 92% yield from 7a. On the other
hand, trans-8d having a sterically bulky t-Bu group was resist-
ant to the method A resulting in complete recovery of 8d. This
problem was overcome by the use of O3 oxidation (method B)9
to give a mixture of 10d and ꢂ-amido nitrile in excellent yield.
Hydrolysis of the mixture afforded the optically pure (1R,2R)-
trans-Ahh 1b in 90% yield from 8d. Thus, the structure of
trans-8d was unambiguously confirmed and the stereoselective
synthesis of the trans-Ahh 1b was accomplished by means of
L-t-Leu as the chiral auxiliary (Scheme 3).
nol-d8
Bn
Bn
Bn
O
O
O
Additive
N
N
N
6
O
O
O
2-Propanol-d8
+
D
D
H
2 : 1
(6S)-12
(6R)-13
(6S)-4a : (6R)-5a = 1 : 1
TFA or ZnCl2 (1.0 equiv), 2-PrOH
Ratio
12:13
d-Incorporation
at C-6/%
Additive
Time
None
72 h
2:1
2:1
2:1
—
100
100
100
0
Imine–Enamine Equilibrium and Conformational Anal-
ysis of Ketimines (6S)-4 and (6R)-5. We examined equilib-
rium experiments of the ketimines 4a and 5a using 2-propanol-
d8 in the presence or absence of the cyanide source (Table 3).
Treatment of a 1:1 mixture of 4a and 5a with TFA in 2-prop-
anol-d8 reached equilibrium within 7 min to give a 2:1 mixture
of mono deuterated (6S)-12 and (6R)-13. The deuterium incor-
poration was also effected by ZnCl2 within 2 h to give 12 and
13 in the same product ratio (2:1). On the other hand, the in-
corporation proceeded very slowly in the absence of additives.
No incorporation was observed in the presence of NaCN with-
out TFA or ZnCl2. Treatment of a 2:1 mixture of deuterated 12
and 13 with TFA in 2-PrOH afforded a 1:1 mixture of 4a and
5a. These results clearly indicated the existence of the equilib-
CF3CO2D (1.0 equiv)
ZnCl2 (1.0 equiv)
NaCN (1.0 equiv)
<7 min
2 h
72 h
a) All experiments were carried out at room temperature. A
1:1 mixture of 4a and 5a was used.
rium between 4a and 5a via the enamine 6a in 2-PrOH. The
equilibrium experiments in 2-PrOH using Phe-, Leu-, and t-
Leu ketimines 4a, 4c, and 4d and 5a, 5c, and 5d are summa-
rized in Table 4. The initial ratios of the ketimines 4a and
4d and 5a and 5d were retained or slightly changed by treat-
ment with TFA or ZnCl2. In the case of a 1:1 mixture of 4c
and 5c, the use of TFA did not affect the original ratio, but

T. Shinada et al.
Bull. Chem. Soc. Jpn. Vol. 79, No. 5 (2006)
is exposed from the ketimine ring (Fig. 1).
771
Table 4. Equilibrium Experiments of 4a, 4c, and 4d and 5a,
5c, and 5d in 2-PrOH
Proposed Mechanism. From the above experiments, it is
clearly understood that the facial selectivity (1R vs 1S) and the
cis/trans selectivity upon cyanide addition to the ketimines are
affected by the nature of the chirality transferring amino acids
(aromatic vs alkyl side chain) and the reaction conditions (con-
ditions A vs B). Regarding the diastereofacial selectivity, we
propose that the exclusive formation of the (1S)-ꢂ-amino ni-
trile isomers 7a and 7b and 8a and 8b from Phe-3a and Trp-
3b is attributed to the conformation of their ketimine inter-
mediates, 4a and 4b and 5a and 5b, where the aromatic group
of the ketimine intermediate shields the ꢂ-face from the attack
of the cyanide ion. A small amount of the (1R)-isomers 11c
(conditions A: 1S/1R = 81:19 and B: 89:11) and 11d (A:
98:2 and B: 95:5) obtained from 3c and 3d suggests that the
ꢂ-face of their ketimine intermediates, 4c and 4d and 5c and
5d, is not sufficiently shielded due to the exo orientation of
the alkyl group (Fig. 1). Regarding the cis/trans selectivity,
additional experiments were performed to see whether an equi-
librium exists between the ꢂ-amino nitriles 7a and 8a through
the ketimines 4a and 5a. This has been often observed in a
monocyclic ꢂ-amino nitrile system.4,6 However, the presence
of such an equilibrium in this case was ruled out by the fact
that treatment of the ꢂ-amino nitriles 7a and 8a with Na13CN/
TFA in 2-PrOH and Na12CN/TFA-d1 in 2-PrOH-d8 incorpo-
rated none of the 13CN or D into 7a and 8a, respectively
(Scheme 4). These results suggest that the cyanide addition
to the ketimine is a rate-determining step under condition A
since the equilibrium between the ketimines 4a and 5a reached
a plateau within 7 min and the overall reaction takes 3 h.
Therefore, the unusual stacking conformation of the ketimines
derived from Phe-3a and Trp-3b would sterically hinder an
axial attack of the cyanide ion to 5a and 5b to give cis-7a
and -7b as the major product, respectively (Scheme 3). On
the other hand, the cis/trans ratios of 7c/8c and 7d/8d having
an alkyl side chain were almost in accord with those of their
ketimines 4c/5c and 4d/5d, respectively. In these cases, the
equilibrium between ketimines was also fast (Table 4). Signif-
icant differences of their conformation from 7a and 7b and 8a
and 8b as well as stereoelectronic preference (axial attack)11
for the cyanide addition might reflect the product ratios.
Regarding the trans-selectivity observed under condition B,
ZnCl2 plays an important role. In particular, the (6R)-keti-
mines 5c and 5d were preferentially formed under the equi-
librium conditions using ZnCl2 (Table 4). The ratios of the
ketimines and the resultant ꢂ-amino nitriles were almost equal.
As shown in Scheme 5, ZnCl2 would contribute to stabiliza-
R
R
A) TFA, 2-PrOH
O
O
N
N
or
4a,c,d + 5a,c,d
O
O
+
B) ZnCl2, 2-PrOH
(6S)-4a,c,d
(6R)-5a,c,d
Ratio after
equilibrium (4:5)aÞ
Initial ratio
Substrate
R
4:5
A (TFA) B (ZnCl2)
4a + 5a (Phe)
4c + 5c (Leu)
4d + 4d (t-Leu)
Bn
CH2i-Pr
t-Bu
50:50
50:50
21:79
50:50
50:50
17:83
50:50
29:71
13:87
a) Reactions were carried out at room temperature for 7 min
under the condition A and for 2 h under the condition B.
ZnCl2 gave trans-5c as the major isomer (4c/5c = 29:71).
With the above results in hand, we next inspected the
1H NMR data of the ketimines 4 and 5 based on coupling con-
stants and NOESY experiments (Table 5). In all cases, signals
corresponding to the enamine 6 were not observed. Large J
values (10–12 Hz) between 6-H and 7-Hax were observed in
all (6S)- and (6R)-ketimines 4 and 5, indicating that these pro-
tons are in an antiperiplanar relationship. In comparison of the
homoallylic coupling constants (4J3-H{6-H) of the ketimines 4
and 5 with those reported of 14,10 the ketimine having a small-
4
er JH{H value (ꢁ1:8 Hz) was assigned to (6R)-5, whereas the
larger J value (>3:1 Hz) was assigned to (6S)-4. These J val-
ues together with NOESY experiments suggested that the ke-
timines 4 and 5 possess a boat-like conformation10 with the
amino acid side chain oriented to a pseudoaxial position as
shown in Fig. 1. In the case of the ketimines having an aromat-
ic side chain, an unusual high field shift of 6-H for 5a and 5b,
7-Hax for 4a and 4b, and 9-Hax for 4b was observed probably
due to anisotropic effects by the aromatic group. In addition,
the J values between the benzyl methylene group and 3-H
0
(J3{1 H ¼ 4:5, 4.5 Hz) suggested the local conformation to have
a bisect form. These observations led 4a and 4b and 5a and 5b
to have a stacking conformation where the ꢂ-face is complete-
ly shielded (Fig. 1). In contrast to 4a and 4b, none of the pro-
tons that shift to a higher magnetic field were observed in the
case of the ketimines 4c and 4d and 5c and 5d derived from
Leu and t-Leu. The dihedral angle between 2-H and 10-Hs
0
(J3{1 H ¼ 5:7, 11.0 Hz) according to the Karplus equation indi-
cated that the alkyl side chain of the Leu–ketimines 4c and 5c
1
Table 5. Selected H NMR Data of Ketimines 4 and 5aÞ
Ketimines
3-H
6-H
7-Hax
9-Hax
Phe-4a
Phe-5a
Trp-4b
Trp-5b
Leu-4c
Leu-5c
t-Leu-4d
t-Leu-5d
4.57 [ddt, 4.6, 4.3, 3.2]
4.74 [tt, 4.5, 1.8]
4.61 [tt, 4.3, 3.5]
4.77 [ddt, 4.3, 4.1, 1.5]
4.01 [ddt, 11.0, 5.7, 3.5]
4.30 [ddt, 7.8, 6.3, 1.5]
3.85 [t, 3.1]
4.57 [ddd, 12.3, 5.9, 3.2]
3.26 [ddd, 12.3, 6.3, 1.8]
4.47 [ddd, 12.3, 5.7, 3.5]
3.24 [ddd, 12.1, 6.4, 1.5]
4.77 [ddd, 12.9, 6.9, 3.5]
4.80 [ddd, 11.0, 7.3, 1.5]
4.80 [ddd, 14.0, 5.1, 3.1]
4.78 [ddd, 10.2, 6.7, 1.4]
À0:02 [qd, 12.3, 3.9]
1.2–1.4 [m]
1.2–1.4 [m]
0.23 [qt, 13.5, 4.2]
1.2–1.4 [m]
1.4–1.7 [m]
1.4–1.7 [m]
1.4–1.7 [m]
1.4–1.7 [m]
1.2–1.4 [m]
À0:21 [qd, 12.3, 4.0]
1.2–1.4 [m]
1.4–1.7 [m]
1.4–1.7 [m]
1.4–1.7 [m]
1.4–1.7 [m]
4.10 [t, 1.4]
a) Chemical shifts and coupling constants are shown in ꢁ (ppm) and J (Hz), respectively.

772 Bull. Chem. Soc. Jpn. Vol. 79, No. 5 (2006)
Asymmetric Strecker Synthesis
1.2-1.4 [m]
δ
δ
4.57
H
[ddd, J = 12.3, 5.9, 3.2 Hz]
N
H
H
1.8 Hz
12.3 Hz
H
H
9
3.2 Hz
12.3 Hz
O
O
7
6
N
O
3
H
H
H
H
δ
-0.02
O
δ
3.26
[qd, J = 12.3, 3.9 Hz]
H
1'
[ddd, J = 12.3, 6.3, 1.8 Hz]
4.5 Hz
O
4.5 Hz
4.3 Hz
4.6 Hz
H
H
H
H
H
H
O
N
N
Ph
Ph
(6S)-Phe-4a
4.61 [m]
(6R)-Phe-5a
4.77
δ
δ
[ddd, J = 12.9, 6.9, 3.5 Hz]
δ
4.38
H
H
[dd, J = 17, 2.2 Hz]
3.5 Hz
N
H
H
12.9Hz
7
6
H
N
O
O
O
3
H
H
H
δ
1.4-1.7 [m]
H
δ
3.53
O
[dd, J = 17, 1.5 Hz]
H
5 Hz
5 Hz
H
5.7 Hz
H
11.0 Hz
H
H
H
O
O
N
Ph
N
H
14
(6S)-Leu-imine 4c
Fig. 1. Proposed conformation of Phe-4a, Phe-5a, Leu-4d, and monocyclic ketimine 14.
Cl
Zn
CN
CN
N
Cl
H
R
Cl
Zn
Bn
H
H
O
O
N
HN
H
N
HN
N
O
Cl
R
O
O
O
O
O
O
O
O
R
O
(6R)-5
(6S)-4
TFA, NaCN
equilibrium > 7 min
faster
(6S)-Phe-4a
R = CH2CHMe2
or t-Bu
(6R)-Phe-5a
CN
CN
CN
NC
slower
CN
CN
NC
Addition of cyanide ion
~3 h for completion of the reaction
H
O
N
R
N
H
CN
O
H
O
Bn
N
O
H
O
N
O
R
cis-7
trans-8
O
O
Bn
trans-Phe-8a
cis-Phe-7a
Scheme 5.
TFA, Na13CN or
No incorporation of 13CN or D into 7a and 8a
7a + 8a
Conclusion
TFA-d1, NaCN
We have examined the effects of the chiral auxiliary for the
stereoselective synthesis of 1a and 1b through exploitation of
intermediary ketimine equilibrium and conformations under
different reaction conditions. From these experiments, the ster-
eochemical outcome of cyanide addition to a bicyclo[4.4.0]-
azaoxaoctane system to form the corresponding ꢂ-amino ni-
trile was clearly understood. On the basis of these studies,
we have established a stereoselective route to access each
Scheme 4.
tion of the (6R)-ketimine formation through chelation to the
imino group, although its effect on the ketimine conformation
can not be verified at this stage. It is noted that dehydration of
3d having a t-Leu ester gave (6R)-5d as the major ketimine
isomer. The bulky t-Bu group would restrict the formation of
(6S)-4d due to steric reasons.

T. Shinada et al.
Bull. Chem. Soc. Jpn. Vol. 79, No. 5 (2006)
773
dd, J ¼ 13:3, 4.3 Hz), 4.57 (1/2H, ddt, J ¼ 4:6, 4.3, 3.2 Hz), 4.57
(1/2H, ddd, J ¼ 12:3, 5.9, 3.2 Hz), 4.74 (1/2H, tt, J ¼ 4:5, 1.8
Hz), 7.07–7.28 (5H, m). 1H NMR (400 MHz, C6D6) ꢁ ꢃ0:01
(1/2H, qd, J ¼ 12:3, 3.7 Hz), 0.52–1.67 (7H, m), 2.37 (1/2H,
m), 2.49 (1/2H, m), 3.08 (1/2H, dd, J ¼ 13:1, 4.5 Hz), 3.17
(1/2H, ddd, J ¼ 12:3, 6.3, 1.8 Hz), 3.23 (1/2H, dd, J ¼ 13:2,
4.3 Hz), 3.37 (1/2H, dd, J ¼ 13:1, 4.5 Hz), 3.52 (1/2H, dd,
J ¼ 13:2, 4.6 Hz), 3.84 (1/2H, ddd, J ¼ 12:3, 5.9, 3.2 Hz), 4.38
(1/2H, ddt, J ¼ 4:6, 4.3, 3.2 Hz), 4.72 (1/2H, tt, J ¼ 4:5,
1.8 Hz), 7.00–7.28 (5H, m).
(2S,4aR)- and (2S,4aS)-2-Benzyl-4a-deuterio-3,4a,5,6,7,8-
hexahydro-2H-4-oxaquinolin-3-ones (12) and (13) (12:13 =
2:1): 1H NMR (400 MHz, 2-propanol-d8) ꢁ ꢃ0:16 (1/2H, qd,
J ¼ 12:3, 4.1 Hz), 1.02–2.15 (7H, m), 2.42 (1/2H, m), 2.54 (1/
2H, m), 3.02 (1/2H, dd, J ¼ 13:2, 4.6 Hz), 3.13 (1/2H, dd, J ¼
13:2, 4.6 Hz), 3.20–3.26 (1H, m), 3.32 (1/2H, dd, J ¼ 13:2, 4.1
Hz), 4.43 (1/2H, ddt, J ¼ 4:6, 4.1, 3.2 Hz), 4.58 (1/2H, m), 4.67
(1/2H, ddd, J ¼ 12:3, 5.5, 3.2 Hz), 6.94–7.18 (5H, m).
cis- and trans-isomer of the Ahh 1a and 1b using an asymmet-
ric version of the Strecker synthesis. Extension of these results
for the synthesis of further substituted Ahh derivatives and
their incorporation to peptides is in progress.
Experimental
General. All reagents and solvents were purchased from ei-
ther Aldrich Chemical Company, Inc.; Merck & Co., Inc.; Nacalai
Tesque Company, Ltd.; Peptide Institute; Tokyo Kasei Kogyo
Co., Ltd.; or Wako Pure Chemical Industries, Ltd.; and used with-
out further purification unless otherwise indicated. Tetrahydro-
furan (THF) was distilled under an argon atmosphere from so-
dium/benzophenone ketyl. Dichloromethane (CH2Cl2) was dis-
tilled from diphosphorus pentaoxide (P2O5). Methanol (MeOH)
was distilled from magnesium turnings and iodine. Acetonitrile
(CH3CN) was distilled from calcium hydride (CaH2). DMSO
was distilled under reduced pressure from CaH2. Diethyl ether
(Et2O) and N,N-dimethylformamide (DMF) of anhydrous grade
were used. 1H NMR spectra were recorded on either a JEOL JNM-
LA300 (300 MHz), JEOL JNM-LA400 (400 MHz), or Varian Uni-
ty plus 500 spectrometer. Chemical shifts of 1H NMR were report-
ed in parts per million (ppm, ꢁ) relative to tetramethylsilane (ꢁ ¼
0:00) in CDCl3 or HDO (ꢁ ¼ 4:80) in D2O, or CD2HOD (ꢁ ¼
3:30) in CD3OD. 13C NMR spectra were recorded on a JEOL
JNM-LA300 (75 MHz) or JEOL JNM-LA400 (100 MHz) spec-
trometer. Chemical shifts of 13C NMR were reported in ppm (ꢁ)
relative to CHCl3 (ꢁ ¼ 77:0) in CDCl3, CH3OH (ꢁ ¼ 49:0) in
D2O or CH3OH (ꢁ ¼ 49:0) in CD3OD, or DMSO (ꢁ ¼ 39:5) in
DMSO-d6. Low resolution mass spectra (LRMS) and high resolu-
tion mass spectra (HRMS) were obtained on a JEOL JMS-AX500
for fast atom bombardment ionization (FAB), chemical ionization
(CI), or electron ionization (EI). Elemental analyses were record-
ed on a Perkin-Elmer 240C. All reactions were monitored by thin-
layer chromatography (TLC), which was performed with precoat-
ed plates (silica gel 60 F-254, 0.25 mm layer thickness, manufac-
tured by Merck). TLC visualization was accomplished using UV
lump (254 nm) or a charring solution (ethanoic p-anisaldehyde,
ethanoic molybdophosphoric acid, aqueous potassium permanga-
nate, and butanoic ninhydrin). Daisogel IR-60 1002W(40/63 mm)
was used for flash column chromatography on silica gel. Reversed
phase chromatography was performed on a Cosmosilꢁ 140C18-
PREP. The esters 3a–3d were prepared according to the proce-
dures reported.6b
General Procedure for Formation of Ketimines 4a–4d and
5a–5d from Esters 3a–3d. To a solution of the ester 3 (0.5
mmol) in CH2Cl2 was added TFA (1 mL) at 0 ꢂC with stirring.
The mixture was stirred at room temperature for 30 min and con-
centrated in vacuo. The residue was diluted with toluene and con-
centrated in vacuo. To a solution of the residue in acetonitrile
(2 mL) was added Na2CO3 (263 mg, 2.5 mmol) and anhydrous
MgSO4 (600 mg, 5 mmol) with stirring. The mixture was filtered
and the filtrate was concentrated in vacuo to give a mixture of
the ketimines 4 and 5 as a pale yellow oil in a ratio as shown in
Table 1. Without further purification, the imine was used for the
cyanide addition reactions under either condition A or B and for
the equilibrium experiments.
(2S,4aR)- and (2S,4aS)-2-Benzyl-3,4a,5,6,7,8-hexahydro-2H-
4-oxaquinolin-3-ones (4a) and (5a) (4a:5a = 1:1): 1H NMR
(400 MHz, CDCl3) ꢁ ꢃ0:02 (1/2H, qd, J ¼ 12:3, 3.9 Hz), 1.20–
2.18 (7H, m), 3.50 (1/2H, ddt, J ¼ 13:5, 4.3, 1.7 Hz), 2.59 (1/2H,
ddt, J ¼ 14:8, 4.2, 1.7 Hz), 3.15 (1/2H, dd, J ¼ 13:3, 4.5 Hz),
3.24–3.28 (1H, m), 3.36 (1/2H, dd, J ¼ 13:3, 4.5 Hz), 3.45 (1/2H,
(2S,4aR)- and (2S,4aS)-2-(Indol-3-ylmethyl)-3,4a,5,6,7,8-
hexahydro-2H-4-oxaquinolin-3-ones (4b) and (5b) (4b:5b =
1:1): 1H NMR (400 MHz, CDCl3) ꢁ ꢃ0:21 (1/2H, qd, J ¼ 12:3,
4.0 Hz), 0.23 (1/2H, qt, J ¼ 13:5, 4.2 Hz), 1.00–2.50 (8H, m),
3.24 (1/2H, ddd, J ¼ 12:1, 6.4, 1.5 Hz), 3.37 (1/2H, dd, J ¼ 14:0,
4.1 Hz), 3.49–3.53 (1H, m), 3.57 (1/2H, dd, J ¼ 14:0, 4.3 Hz),
4.47 (1/2H, ddd, J ¼ 12:3, 5.7, 3.5 Hz), 4.61 (1H, tt, J ¼ 4:3,
3.5 Hz), 4.77 (1/2H, ddt, J ¼ 4:3, 4.1, 1.5 Hz), 6.98–7.34 (4H,
m), 7.54 (1/2H, d, J ¼ 8:1 Hz), 7.66 (1/2H, d, J ¼ 8:1 Hz), 8.16
(1/2H, s), 8.23 (1/2H, s). 1H NMR (400 MHz, C6D6) ꢁ ꢃ0:32
(1/2H, qd, J ¼ 12:2, 3.7 Hz), 0.08 (1/2H, qt, J ¼ 13:6, 4.3 Hz),
0.27–1.64 (7H, m), 2.22 (1/2H, dquint, J ¼ 13:5, 2.1 Hz), 2.40
(1/2H, dquint, J ¼ 14:4, 2.1 Hz), 3.05 (1/2H, ddd, J ¼ 12:1, 6.2,
1.8 Hz), 3.33 (1/2H, dd, J ¼ 14:2, 4.2 Hz), 3.48 (1/2H, dd, J ¼
14:1, 4.3 Hz), 3.53 (1H, dd, J ¼ 14:2, 4.2 Hz), 3.63 (1/2H, dd,
J ¼ 14:1, 4.3 Hz), 3.72 (1/2H, ddd, J ¼ 12:2, 5.6, 3.3 Hz), 4.43
(1/2H, tt, J ¼ 4:3, 3.3 Hz), 4.76 (1/2H, tt, J ¼ 4:2, 1.8 Hz), 6.62–
7.86 (6H, m).
(2S,4aR)- and (2S,4aS)-2-(2-Methylpropyl)-3,4a,5,6,7,8-
hexahydro-2H-4-oxaquinolin-3-ones (4c) and (5c) (4c:5c =
1:1): 1H NMR (400 MHz, CDCl3) ꢁ 0.91 (3/2H, d, J ¼ 3:5 Hz),
0.92 (3/2H, d, J ¼ 3:5 Hz), 0.92 (3/2H, d, J ¼ 1:1 Hz), 0.94 (3/
2H, d, J ¼ 1:1 Hz), 1.39–2.63 (11H, m), 4.01 (1/2H, ddt, J ¼
11:0, 5.7, 3.5 Hz), 4.30 (1/2H, ddt, J ¼ 7:8, 6.3, 1.5 Hz), 4.77
(1/2H, ddd, J ¼ 12:9, 6.9, 3.5 Hz), 4.80 (1/2H, ddd, J ¼ 11:0,
7.3, 1.5 Hz).
(2S,4aR)- and (2S,4aS)-2-(t-Butyl)-3,4a,5,6,7,8-hexahydro-
2H-4-oxaquinolin-3-ones (4d) and (5d) (4d:5d = 21:79):
1H NMR (400 MHz, CDCl3) ꢁ 1.04 (27/10H, s), 1.10 (63/10H, s),
1.45–2.72 (8H, m), 3.85 (3/10H, t, J ¼ 3:1 Hz), 4.10 (7/10H, t,
J ¼ 1:4 Hz), 4.78 (7/10H, ddd, J ¼ 10:2, 6.7, 1.4 Hz), 4.80 (3/
10H, ddd, J ¼ 14:0, 5.1, 3.1 Hz).
General Procedure for the Cyanide Addition to a Mixture
of Ketimines 4a–4d and 5a–5d: Condition A. To a solution
of a mixture of the ketimines 4 and 5 (0.5 mmol) in 2-PrOH
(2 mL) was added TFA (39 mL, 0.5 mmol) at room temperature.
After stirring for 5 min, NaCN (49 mg, 1 mmol) was added to
the reaction mixture. The mixture was stirred for 2 h at room tem-
perature, and then diluted with AcOEt and brine. The organic lay-
er was washed with H2O, dried over anhydrous MgSO4, and con-
centrated in vacuo. The residue was purified by flash column chro-
matography to give a mixture of the ꢂ-amino nitriles 7 and 8. The
yields and the ratios of the products are shown in Table 1. The
1H NMR data of 7a and 8a were identical with those reported.6d

774 Bull. Chem. Soc. Jpn. Vol. 79, No. 5 (2006)
Asymmetric Strecker Synthesis
Condition B. To a solution of a mixture of the ketimines 4
and 5 (0.5 mmol) in 2-PrOH (2 mL) was added ZnCl2 (0.5 mL
in THF solution, 0.5 mmol) at room temperature. The mixture
was stirred for 5 min, and then TMSCN (0.2 mL, 1.5 mmol) was
added. The mixture was stirred for 2 h at room temperature, and
diluted with AcOEt and brine. The organic layer was washed with
H2O, dried over anhydrous MgSO4, and concentrated in vacuo.
The residue was purified by flash column chromatography to give
a mixture of the ꢂ-amino nitriles 7 and 8. The yields and the ratios
of the products are shown in Table 1.
(1S,3S,6R)- and (1S,3S,6S)-2-Aza-3-(indol-3-ylmethyl)-5-
oxa-4-oxobicyclo[4.4.0]decane-1-carbonitriles (7b) and (8b)
(7b:8b = 1:2): 1H NMR (400 MHz, CDCl3) ꢁ 1.21–2.07 (8H,
m), 3.16 (1/3H, dd, J ¼ 14:4, 9.6 Hz), 3.19 (2/3H, dd, J ¼ 14:5,
8.7 Hz), 3.55 (2/3H, dd, J ¼ 14:5, 3.6 Hz), 3.62 (1/3H, ddd, J ¼
14:4, 3.0, 0.8 Hz), 3.98 (1/3H, dd, J ¼ 12:2, 4.1 Hz), 4.18 (2/3H,
dt, J ¼ 8:7, 3.6 Hz), 4.31 (1/3H, dd, J ¼ 9:6, 3.0 Hz), 4.33 (2/3H,
dd, J ¼ 10:2, 4.3 Hz), 7.10–7.40 (4H, m), 7.66 (2/3H, d, J ¼ 8:0
Hz), 7.67 (1/3H, d, J ¼ 7:8 Hz), 8.22 (1/3H, brs), 8.23 (2/3H,
brs). HRMS (CIþ) m=z calcd for C17H19N2O2 ðM ꢃ CNÞþ
283.1446, found 283.1443.
(1S,3S,6R)- and (1S,3S,6S)-2-Aza-3-(2-methylpropyl)-5-
oxa-4-oxobicyclo[4.4.0]decane-1-carbonitriles (7c) and (8c)
(7c:8c = 18:71): 1H NMR (400 MHz, CDCl3) ꢁ 0.95 (12/5H,
d, J ¼ 6:6 Hz), 0.96 (3/5H, d, J ¼ 6:5 Hz), 0.97 (12/5H, d, J ¼
6:6 Hz), 0.98 (3/5H, d, J ¼ 6:5 Hz), 1.38–2.17 (11H, m), 3.89
(1/5H, dd, J ¼ 8:8, 4.2 Hz), 4.06 (4/5H, dd, J ¼ 8:4, 4.0 Hz),
4.10 (4/5H, dd, J ¼ 12:2, 4.1 Hz), 4.47 (1/5H, dd, J ¼ 8:8,
4.4 Hz); HRMS (CIþ) m=z calcd for C13H21N2O2 ðM þ HÞþ
237.1603, found 237.1604.
CHCl3); FTIR (neat) 2952, 2867, 2245, 1747, 1620, 1456, 1360
cmꢃ1 1H NMR (400 MHz, CDCl3) ꢁ 1.54–2.01 (6H, m), 2.20
;
(1H, m), 2.63 (1H, ddd, J ¼ 13:6, 5.1, 2.9 Hz), 3.94 (1H, dd, J ¼
11:9, 4.3 Hz); 13C NMR (100 MHz, CDCl3) ꢁ 172.9, 153.7, 115.9,
79.6, 59.5, 39.5, 35.8, 28.7, 22.9, 22.3; HRMS (EI) m=z calcd for
C13H18N2O2 (M)þ 234.1368, found 234.1341.
N-((1S,2R)-1-Cyano-2-hydroxycyclohexyl)-2,2-dimethylpro-
25
panamide: ½ꢂꢄ
þ45:8 (c 0.48, CHCl3); FTIR (neat) 3408,
D
3345, 2930, 2245, 1647, 1558, 1539, 1456 cmꢃ1 1H NMR (400
;
MHz, CDCl3) ꢁ 1.27–1.47 (6H, m), 2.07 (1H, m), 2.82 (1H, m),
3.49 (1H, brs), 3.63 (1H, dd, J ¼ 10:8, 3.5 Hz), 6.17 (1H, s);
13C NMR (100 MHz, CDCl3) ꢁ 179.5, 117.6, 74.4, 58.9, 39.2,
34.3, 31.7, 27.4, 23.4, 21.8; HRMS (EI) m=z calcd for C12H20-
N2O2 (M)þ 224.1525, found 224.1536.
This study was supported by a grant for ‘‘The Research for
the Future Program (99L01204)’’ and Scientific Research
(No. 16201045) from the Japan Society for the Promotion of
Science (JSPS), and for Scientific Research (No. 16073214)
from the Ministry of Education, Culture, Sports, Science and
Technology, Japan. T.S. is also grateful for a SUNBOR Grant
from The Suntory Institute for Bioorganic Research.
References
1
G. M. Cooper, The Cell: A Molecular Approach, Oxford
University Press, Oxford, 1997.
C. Walsh, Enzymatic Reaction Mechanisms, W. H.
Freeman & Co., San Francisco, 1979.
a) M. Stasiak, W. M. Wolf, M. T. Leplawy, J. Pept. Sci.
1998, 4, 46. b) D. Obrecht, M. Altorfer, C. Lehmann, P.
Schonholzer, K. Muller, J. Org. Chem. 1996, 61, 4080. c) V.
2
3
(1S,3S,6R)- and (1S,3S,6S)-2-Aza-3-(t-butyl)-5-oxa-4-oxo-
bicyclo[4.4.0]decane-1-carbonitriles (7d) and (8d) (7d:8d =
¨
¨
5:90):
1H NMR (400 MHz, CDCl3) ꢁ 1.06 (15/2H, s), 1.12
Pavon, B. Di Blasio, A. Lombardi, O. Maglio, C. Isernia, C.
Pedone, E. Benedetti, E. Altmann, M. Mutter, Int. J. Pept. Protein
Res. 1993, 41, 15. d) H. Mickos, B. Sundberg, B. Luning, Acta
Chem. Scand. 1992, 46, 989. e) E. Altmann, K.-H. Altmann, M.
Mutter, Angew. Chem., Int. Ed. Engl. 1988, 27, 858. f) E. H.
Flynn, J. W. Hinman, E. L. Caron, D. O. Woolf, Jr., J. Am. Chem.
Soc. 1953, 75, 5867.
(3/2H, s), 1.21–2.15 (8H, m), 3.62 (1/6H, s), 3.75 (5/6H, s), 4.07
(5/6H, dd, J ¼ 12:2, 4.4 Hz), 4.38 (1/6H, dd, J ¼ 10:0, 4.9 Hz);
HRMS (CIþ) m=z calcd for C13H21N2O2 ðM þ HÞþ 237.1603,
found 237.1609.
Oxidation of ꢀ-Amino Nitrile 7a Using Method A. To a
solution of 7a (1.5 g, 5.57 mmol) in dichloromethane (7 mL) was
added t-BuOCl (0.8 mL, 6.71 mmol) at 0 ꢂC. After stirring the
mixture at 0 ꢂC for 30 min, DABCO (1.5 g, 13.4 mmol) was added.
The reaction mixture was stirred at the same temperature for
30 min, quenched by the addition of water, and extracted with eth-
yl acetate. The organic layer was washed with 1 M HCl, water,
brine, dried over anhydrous MgSO4, and then concentrated in
vacuo. The residue was purified by flash column chromatography
to give 9a (1.47 g, 97%), whose spectroscopic data were identical
to those reported.4,6
Oxidation of ꢀ-Amino Nitrile 8d Using Method B. Ozone
was bubbled through a solution of 8d (24 mg, 0.10 mmol) in ethyl
acetate (7 mL) at ꢃ78 ꢂC for 15–30 min (flow rate of O2: 150
NL hꢃ1, which corresponded to 3 g hꢃ1 of O3). To the mixture
was added dimethyl sulfide (1 mL). The mixture was warmed to
room temperature, and concentrated in vacuo. The residue was
purified by flash column chromatography on silica gel (elution
with hexane/EtOAc = 4:1, then 2:1) to give 10d (14 mg, 59%)
and ꢂ-amido nitrile (8 mg, 34%) as a colorless amorphous pow-
der. The mixture without separation was converted to 1b accord-
ing to the reported procedure.9
4
Y. Ohfune, K. Namba, I. Takada, T. Kan, M. Horikawa,
T. Nakajima, Chirality 1997, 9, 459.
M. Horikawa, Y. Shigeri, N. Yumoto, S. Yoshikawa, T.
Nakajima, Y. Ohfune, Bioorg. Med. Chem. Lett. 1998, 8, 2027.
a) S.-H. Moon, Y. Ohfune, J. Am. Chem. Soc. 1994, 116,
5
6
7405. b) Y. Ohfune, T. Shinada, Bull. Chem. Soc. Jpn. 2003,
76, 1115. c) Y. Ohfune, T. Shinada, Eur. J. Org. Chem. 2005,
5127.
7
H. Kunz, W. Sager, W. Pfrengle, D. Schanzenbach, Tetra-
hedron Lett. 1988, 29, 4397.
Method A was also effective for the oxidation of a mixture
of 7c and 8c to give a mixture of 9c and 10c in quantitative yield.
a) K. Namba, M. Kawasaki, S. Iwama, I. Takada, M.
8
9
Izumida, T. Shinada, Y. Ohfune, Tetrahedron Lett. 2001, 42,
3733. b) K. Namba, T. Shinada, T. Teramoto, Y. Ohfune, J. Am.
Chem. Soc. 2000, 122, 10708.
10 a) G. Schulz, W. Steglich, Chem. Ber. 1977, 110, 3615.
b) M. Sato, N. Kitazawa, S. Nagashima, K. Kano, C. Kaneko,
Chem. Lett. 1992, 485.
11 P. Delongchamps, Stereoelectronic Effects in Organic
Chemistry, Pergamon Press, Oxford, 1983, pp. 209–290.
(8aS,4aR)-2-(t-Butyl)-3-oxo-4a,5,6,7,8,8a-hexahydro-3H-4-
28
oxaquinolin-8a-carbonitrile (10d):
½ꢂꢄ
ꢃ168:5 (c 1.06,
D