Four Stereoisomers of 2-Aminomethyl-1-cyclopropanecarboxylic Acid: Synthesis and biological evaluation

Article abstract of DOI:10.1246/bcsj.20190168

Here, we report a practical method for asymmetric synthesis of cyclopropane-fused GABA analogs. Starting from 2-furaldehyde, the cis-isomer (CAMP) was synthesized over 10 steps; (¹)- and (+)-CAMP￠HCl were synthesized by employing d- and l-menth

Full text of DOI:10.1246/bcsj.20190168

Advance Publication Cover Page
Four Stereoisomers of 2–Aminomethyl–1–cyclopropanecarboxylic Acid: Synthesis
and Biological Evaluation
Masato Oikawa,* Yuka Sugeno, Hideyuki Tukada, Yuichi Takasaki,
Satoshi Takamizawa, and Raku Irie
Advance Publication on the web August 2, 2019
doi:10.1246/bcsj.20190168
© 2019 The Chemical Society of Japan
Advance Publication is a service for online publication of manuscripts prior to releasing fully edited, printed versions. Entire
manuscripts and a portion of the graphical abstract can be released on the web as soon as the submission is accepted. Note that the
Chemical Society of Japan bears no responsibility for issues resulting from the use of information taken from unedited, Advance
Publication manuscripts.

Four Stereoisomers of 2–Aminomethyl–1–cyclopropanecarboxylic Acid: Synthesis
and Biological Evaluation#
Masato Oikawa,* Yuka Sugeno, Hideyuki Tukada, Yuichi Takasaki, Satoshi Takamizawa, Raku Irie
Yokohama City University, Seto 22–2, Kanazawa-ku, Yokohama 236–0027, Japan
E-mail: moikawa@yokohama-cu.ac.jp
Abstract
Here, we report practical method for asymmetric
synthesis of cyclopropane-fused GABA analogs. Starting from
2–furaldehyde, the cis–isomer (CAMP) was synthesized over
10 steps; (–)– and (+)–CAMP•HCl were synthesized by
employing d– and l–menthol as the chiral auxiliary for total
2.5% and 1.3% yields, respectively. On the other hand, the
trans–isomer (TAMP) was elaborated via double asymmetric
induction, i.e. organocatalytic asymmetric cyclopropanation on
chiral substrate. Thus, starting from l– and d–menthyl acrylate,
in combination with quinidine-derived and quinine-derived
organocatalysts, (–)– and (+)–TAMP•HCl were synthesized in
total 6.6% and 3.7% yields, respectively, over 8 steps each.
Configurational analysis of the synthetic intermediates based
on 13C NMR is also reported. Preliminary oncological assays
showed the weak but specific activities of CAMP and TAMP as
the molecular basis of GABA analogs, which are still left
unexplored.
Keywords: Amino acids, Organocatalysis, Cytotoxicity
Introduction
Diastereomeric cis–2–aminomethyl–1–cyclopropanecarbo
xylic acid (CAMP) and the trans–congener (TAMP), and the
enantiomers (Figure 1, 1–4), are conformationally restricted
analogs of 4–aminobutyric (γ–aminobutyric) acid (GABA),
which is one of the neurotransmitters in the mammalian central
nervous system.1 The neuropharmacology of four analogs 1–4
bearing three-membered ring has been well studied to date.2
In 2004, Sakai and co–workers have reported isolation of
N–methylated cyclopropane GABA analog with more
substituents, dysibetaine CPa (DBCPa) and CPb (DBCPb)
(Figure 1), from Micronesian marine sponge.3 Since
neuroactivities of DBCPa/b could not be investigated due to the
limited amount from the marine sponge, we have carried out
chemical synthesis and the results have been already reported.4
We also decided to synthesize CAMP and TAMP as control
compounds in the neuropharmacological evaluation of
DBCPa/b.5 Since DBCPa/b were later found, unfortunately, to
be neurologically inactive from mice in vivo assays,4a we have
performed other assays related to oncology on CAMP and
TAMP, which, to the best of our knowledge, have not yet been
reported. Herein, we report, in full detail, our studies on
asymmetric synthesis and the oncological evaluation of four
stereoisomers of 2–aminomethyl–1–cyclopropanecarboxylic
acid; CAMP and TAMP.5 Since the method for synthesizing
more than subgram quantities of enantiomerically pure
specimens of CAMP and TAMP seemed to be not established
well, we first studied facile and promising synthesis route for
these analogs, as follows.5-6
Figure 1. GABA and the analogs conformationally restricted
by cyclopropane ring
Results and Discussion
Synthesis of (+)– and (–)–cis–2–aminomethyl–1–cycl
opropanecarboxylic acid (CAMP)
Several methods, reported to date for enantioselective
synthesis
of
CAMP,
employ
chiral
chiral
auxiliary
resolution,8
(1–phenylethan–1–amine),7
polymer-supported reagents,9 and radical reaction.10 Meanwhile,
our strategies toward asymmetric synthesis of (+)– and
(–)–CAMP were based on chiral auxiliary-assisted organic
synthesis using menthol.6a In 1994, Feringa’s group reported
diastereoselective synthesis of cyclopropane 8 using l–menthol
as a chiral auxiliary (Scheme 1).11
Scheme 1. Chiral auxiliary-guided cyclopropanation (Feringa,
1994)11

conformational search with MMFF94,16 2) structural
optimizations with HF/3–21G level, 3) energy calculations at
ωB97X–D/6–31G* level, 4) structural optimizations with
ωB97X–D/6–31G* level, 5) energy calculations at
ωB97X–V/6–311+G(2df,2p)[6–311+G*] level with fixing the
geometries to generate Boltzmann distribution, 6) empirically
corrected calculations of the 13C NMR chemical shifts at
ωB97X-D/6-31G* level with ωB97X–D/6–31G* model,17 and
7) correction of the 13C NMR shift values based on the
Boltzmann weighting. Parameters regarding solvents (C6D6)
were not added in these calculations.
The experimental 13C values for 7D (7Dobs), and the
calculated 13C values for 7D (7Dcalc) and 7D’ (7D’calc), are
summarized in Table 1. It was found that calculated 13C values
for 7D (7Dcalc) agree with the experimental values (7Dobs); the
root mean square deviation (RMS) for 7Dcalc is 0.97 ppm. On
the other hand, RMS values for 7D’calc is higher (2.60 ppm).
Furthermore, analysis using DP4 probability statistics14a,18 of
the calculated 13C shifts with the experimental data showed that
the probability ratio for 7Dcalc/7D’calc was 100.0%:0.0%. The
13C-based configurational analysis for 7D was consistent with
the NOESY cross peaks observed at Hc/He (see Scheme 2 and
the Supporting Information).
Since the procedure described11 seemed to be not
established well, we decided to improve their cyclopropane
formation method for synthesis of highly enantiopure
cis-disubstituted
cyclopropane.
Eventually,
large-scale
synthesis of 8 and the enantiomer, followed by 6-step
transformation, was successfully developed for the synthesis of
both enantiomers of CAMP (1, 2) as follows.6a
The starting material, (5S)–(d–menthyloxy)–2(5H)–furan
one (6D),12 was readily prepared by rose bengal-sensitized
photooxidation of 2–furaldehyde (5), followed by
CSA-catalyzed condensation with d–menthol (Scheme 2).13
A
diastereomeric mixture of 5–(d–menthyloxy)–2(5H)–furanones
(6D and the isomer) was formed in 1:1 ratio.
a
Diastereomerically pure 6D was obtained by recrystallization
from petroleum ether, in 20% yield over 2 steps from
2–furaldehyde (5). Treatment of 6D with diazomethane in Et2O
at –40 °C for two days gave diazene 7D and the minor isomer
7D’ (for the structure, see Figure 2) in a ratio of 75:25, as
judged by 1H NMR analysis of the crude 1,3–dipolar
cycloaddition product.
Table 1. Comparison of experimental 13C shifts of
trans–diazene 7D, and calculated 13C shifts for trans–diazene
7D and cis–diazene 7D’
Calculated c
Experiment
Position
al for 7D b
a
trans–diazene
cis–diazene
(7Dobs
)
7D (7Dcalc
)
7D’ (7D’calc
94.6
38.1
96.8
169.3
78.7
74.9
30.0
23.3
32.8
46.6
37.6
22.2
25.6
20.4
15.5
2.60
0.0
)
a
b
c
d
e
f
g
h
i
j
k
l
m
n
o
93.5
39.9
94.3
40.2
103.2
168.0
82.8
76.6
30.1
24.3
32.8
46.6
38.6
22.2
27.0
20.4
16.6
0.97
100.0
104.3
167.0
83.1
77.7
31.3
23.2
34.3
48.1
39.1
22.3
25.7
21.0
15.9
Scheme 2. Preparation of diazene 7D
Figure 2. A set of diastereomers obtained in the cycloaddition.
The configurations were analyzed by the NMR calculations
(see text)
RMS / ppm
DP4 / %
Since diazene products, especially the minor isomer 7D’,
were unstable under both acidic and alkaline conditions,
purification by flash column chromatography using neutral
silica gel 60N was carefully performed to give pure diazene 7D
in 66% isolated yield. The minor diastereomer 7D’ (for the
structure, see Figure 2) was not isolated in pure form, and
hence was not used in this synthetic study. The stereochemistry
of the major isomer, diazene 7D, was first speculated to be
“trans” from NOESY cross peaks (see Scheme 2 and the
Supporting Information). The solid poof was obtained by
comparison of the 13C NMR chemical shift values with those
expected by density functional theory (DFT) calculation.14 The
calculations were performed with Spartan ’18 (Wavefunction,
Irvine, CA, U.S.A.).15 The theoretical 13C NMR shifts for
trans–diazene 7D and cis–diazene 7D’ (Figure 2) were
independently obtained by the calculation sequence; 1)
a For numbering, see Figure 2.
Experimental 13C NMR data were collected at 100 MHz in
b
C6D6.
Calculated 13C NMR data were obtained employing
ωB97X–V/6–311+G(2df,2p)[6–311+G*]//ωB97X–D/6–31G*
model. For detail, see text.
c
As shown in Table 2, the configurational analysis was
also conducted on the experimental 13C values for the
diastereomer 7D’ (7D’obs). The RMS and DP4 analyses have
assigned that the stereochemistry of the minor diazene 7D’ is
“cis” but not “trans”.
In total, the major diazene 7D and the minor diazene 7D’
were concluded to be the trans and the cis isomers, respectively,
from the 13C NMR analyses.14b,19 The assignments were finally
confirmed by leading to (–)–CAMP (see below).

Table 2. Comparison of experimental 13C shifts of cis–diazene
7D’, and calculated 13C shifts for trans–diazene 7D and
cis–diazene 7D’
decomposition of the cyclopropane ring under hydrogenation
conditions. After screening various conditions for the
reduction,20 the stepwise reduction protocol was finally found
to be fruitful. Thus, esterification of 9D with TMSCHN2 gave
methyl ester 10D (94% yield), whose C=N bond was then
reduced by NaBH3CN in AcOH to afford benzyloxyamine 11D
in 44% yield. Protection of amine 11D with Boc group (Boc2O,
Et3N, MeOH) gave 12D in 89% yield. To cleave the N–O bond,
12D was treated with freshly prepared Raney–Ni (W–7) in
EtOH21 to give desired 13D with concomitant
transesterification in 85% yield. The transesterification would
be induced by residual sodium hydroxide, which had been used
for preparation of Raney–Ni (W–7). Global deprotection under
acidic conditions (6 M HCl, 90 °C) successfully delivered
desired (–)–CAMP•HCl (1•HCl) in quantitative yield.
Spectroscopic data were in good agreement with those
reported.7-10 In total, starting from 2–furaldehyde (5), the
synthesis of (–)–CAMP was successfully accomplished in 2.5%
yield over 10 steps.
Calculated c
Experiment
Position
al for 7D’ b
a
trans–diazene
cis–diazene
(7D’obs
)
7D (7Dcalc
94.3
40.2
103.2
168.0
82.8
76.6
30.1
24.3
32.8
46.6
38.6
22.2
27.0
20.4
16.6
2.86
0.0
)
7D’ (7D’calc
)
a
b
c
d
e
f
g
h
i
j
k
l
m
n
o
93.5
39.5
93.7
168.0
78.9
77.0
31.2
23.1
34.3
47.8
36.0
22.3
25.5
20.8
15.6
94.6
38.1
96.8
169.3
78.7
74.9
30.0
23.3
32.8
46.6
37.6
22.2
25.6
20.4
15.5
1.34
100.0
(+)–CAMP was also synthesized employing l–menthol as
the chiral auxiliary (Scheme 4). It should be noted that, in the
synthesis of (+)–CAMP, reduction of the N–O bond of 12L
(Raney–Ni
(W–7),
EtOH)
did
not
accompany
transesterification, using strictly neutralized Raney–Ni
prepared carefully. The synthesis of more than 30 mg of
(+)–CAMP•HCl (2•HCl) was thus accomplished in 1.3% yield
over 10 steps. The spectroscopic data were identical with those
reported.7-10
RMS / ppm
DP4 / %
a For numbering, see Figure 2.
b
Experimental 13C NMR data were collected at 100 MHz in
C6D6.
c
Calculated 13C NMR data were obtained employing
ωB97X–V/6–311+G(2df,2p)[6–311+G*]//ωB97X–D/6–31G*
model. For detail, see text.
Scheme 4. Synthesis of (+)–CAMP•HCl (2•HCl)
Synthesis of (+)– and (–)–trans–2–aminomethyl–1–c
yclopropanecarboxylic acid (TAMP)
Three works are known for the synthesis of
enantiomerically pure TAMP.8,22 Our strategies toward
synthesis of (+)– and (–)–TAMP are based on a combination of
chiral auxiliary-guided synthesis and organocatalyst-mediated
asymmetric cyclopropanation.6b In 200323 and 2004,24 Gaunt’s
group
reported
enantioselective
organocatalytic
cyclopropanation via ammonium ylides derived from 15 and 16
(Figure 3). We decided to employ their method, also using
menthol as a chiral auxiliary, for construction of highly
enantiopure trans-disubstituted cyclopropane.6b
Scheme 3. Synthesis of (–)–CAMP•HCl (1•HCl)
Next, as shown in Scheme 3, 1,3–elimination of nitrogen
(N2) from diazene 7D gave cyclopropane 8D in 71% yield upon
photoirradiation (high-pressure mercury lamp, benzophenone,
benzene). To introduce a nitrogen functionality of CAMP,
d–menthyl acetal 8D was treated with benzyloxyamine to give
oxime ether 9D in 88% yield. Here, several hydrogenation
methods, using 10% Pd/C, Raney–Ni (W–7), and PtO2, were
attempted to directly reduce oxime ether group to amine.
However, most of them were unsuccessful because of the
Figure 3. Amine catalysts for formation of ammonium ylide in
the cyclopropanation

First, effects of four amine catalysts including BnEt3NCl,
1,4–diazabicyclo[2.2.2]octane (DABCO, 14), 15,23 and 1623
(Figure 3), on the diastereoselectivity in the cyclopropanation
of l–menthyl acrylate (17L) with tert–butyl bromoacetate, were
studied. l–Menthyl acrylate (17L) was prepared by
esterification of l–menthol with acryloyl chloride in 94%
yield.25 As shown in Table 3, cyclopropanation using BnEt3NCl
as a catalyst did not proceed after 31 h (entry 1).
reaction afforded trans–cyclopropane in 34% yield (63% based
on recovered starting material 17L) with higher diastereomer
ratio of 7:93 (entry 5). Finally, the cyclopropanation yield with
catalyst 16 was improved to 55% by the slow mixing procedure
(entry 6), as also observed in the cases where quinidine
derivative 15 was employed (entries 3 and 4). From these
observations and the empirical rule in cyclopropanation
described by Gaunt,24 the reaction of l–menthyl acrylate (17L)
using quinidine derivative 15 (entries 3 and 4) is mismatched
asymmetric cyclopropanation, whereas the reaction of 17L
employing quinine derivative 16 (entries 5 and 6) is likely
stereochemically matched (double asymmetric induction effect).
Since sterically demanding acrylate has been reported to be
poorly reactive in the ammonium ylide-mediated
cyclopropanation,24 the low isolated yield observed in Table 3
was likely attributed to the menthyl group. To our
disappointment, however, all attempts to reduce the undesirable
steric repulsion between the acrylate and the ammonium ylide
were unfruitful; for example, the use of sterically less
demanding benzyl bromoacetate in combination with the
catalyst 15 resulted in much lower yield (18%) than that shown
in Table 3 (entry 3), since hydrolytic decomposition of the
benzyl ester also took place under such harsh alkaline
conditions (Cs2CO3, MeCN, 80 °C) (data not shown). Our
efforts still continue to improve the isolation yield in the
asymmetric cyclopropanation.
Table 3. Asymmetric cyclopropanation on chiral acrylate ester
mediated by four amine catalysts
Here, we next show the synthesis of (–)–TAMP from
trans–cyclopropane 18L, which was obtained as a mixture with
19L (18L/19L = 75:25, Table 3, entries 3, 4). We anticipated
practical separation of the diastereomers would be possible
later in the synthesis, and, hence, decided to start the synthesis
toward (–)–TAMP using the cyclopropane 18L of 50% de, as
shown in Scheme 5. Thus, the tert–butyl ester 18L (50% de)
was removed by TFA to give carboxylic acid 20L, which, in
turn, was reduced to alcohol 21L (BH3•THF, or ClCOOMe
followed by NaBH426). The alcohol 21L was still an
inseparable mixture with the (S,S)–diastereomer. To introduce
nitrogen functionality of TAMP, alcohol 21L was then
transformed into phthalimide 22L (42% de) under Mitsunobu
conditions (phthalimide, DEAD, PPh3, benzene) in 93%
yield.27 Here, the phthalimide 22L was found to be a crystalline
solid and was successfully purified by recrystallization from
hot hexane to be >99.5% de.
With diastereomerically pure, fully protected (–)–TAMP
22L in hand, protecting group manipulation was then
conducted. The N–phthaloyl group was first removed by
hydrazine monohydrate in EtOH to give pure N–Boc amine
24L in 76% yield after reprotection (Boc2O, Et3N), ready for
the final deprotection. Finally, acidic hydrolysis by 6 M HCl at
60 °C successfully delivered the desired (–)–TAMP•HCl
(3•HCl) quantitatively. Spectroscopic data including the [α]D
value were in good agreement with those reported.8,22b In total,
the synthesis of (–)–TAMP•HCl (3•HCl) was accomplished in
6.6% yield over 8 steps from l–menthol.
Amine
catalyst
18L/19L b
(R,R)/(S,S)
Entry
Yield a
0%
BnEt3NCl
(1 equiv)
DABCO 14
(1 equiv)
quinidine
derivative 15 42% (89%)
(0.2 equiv)
quinidine
derivative 15 69% (99%)
(0.2 equiv)
quinine
derivative 16 34% (71%)
(0.2 equiv)
1
2
–
35%
56:44
3
4 c
5
75:25
75:25
7:93
quinine
derivative 16 55% (71%)
(0.2 equiv)
6 c
7:93
a
In parentheses are the yield based on recovered starting
material 17L.
b
Diastereoselectivity determined by chiral HPLC analysis
(CHIRALPAK ID–3, 0.46 × 25 cm, H2O/CH3CN = 40:60).
c Reaction conducted by slow addition procedure.
In the presence of 1.0 equiv of DABCO (14, entry 2),23
tert–butyl bromoacetate reacted with l–menthyl acrylate (17L)
to form trans–cyclopropanes 18L and 19L in 35% yield. The
diastereomeric ratio was determined to be 18L/19L = 56:44 by
chiral HPLC analysis (CHIRALPAK ID–3, 0.46 × 25 cm,
H2O/CH3CN = 40:60). The structures were determined later by
X–ray crystallographic analysis of the derivative 22D (see
below), and also by leading to (+)– and (–)–TAMP finally.
Changing the cyclopropanation catalyst to quinidine derivative
15 (0.2 equiv) gave trans–cyclopropane in 42% yield (89%
based on recovered starting material 17L) with the
diastereomer ratio of 18L/19L = 75:25 (entry 3). Furthermore,
slow mixing procedure24 greatly improved the yield to 69%
without loss of the diastereoselectivity (entry 4). We also found
that the use of quinine-derived catalyst 16 (0.2 equiv) in the

phthalimide 22D (>99.5% de) after recrystallization. Here,
structural analysis of 22D was conducted by X–ray
crystallography (Figure 4) to confirm the structure. Gaunt’s
empirical rule for the stereoselectivity in the two chiral
alkaloid-catalyzed cyclopropanations was thus found to be
applicable to also the cases where chiral substrates are
employed (Table 3, entries 3–6). The synthesis of
(+)–TAMP•HCl (4•HCl) was finally accomplished in 3.7%
yield over 8 steps starting from d–menthol. The procedure
established in this study has successfully allowed to synthesize
253 mg of (+)–TAMP•HCl (4•HCl).
Scheme 5. Synthesis of (–)–TAMP•HCl (3•HCl)
Figure 4. Molecular structure of 22D determined by single
crystal X–ray diffraction measurement at 183 K. ORTEP
representation is drawn at 50% probability level for the
ellipsoid.
tBuO
Br
(S)
(S)
O
O
O
tBuO2C
D
D
O
O
Oncological
activity
of
four
isomers
for
16
quinine derivative
Cs2CO3, CH3CN
80 °C
2–aminomethyl–1–cyclopropanecarboxylic acid
17D
18D
(S,S)-isomer
(54% de)
46%
GABA is one of the neurotransmitters, and neuroactivity
of the cyclopropane analogs 1–4 has been well investigated;
(–)–CAMP (1) and (+)–CAMP (2) were characterized as an
antagonist and a full agonist for GABAc receptor, respectively,
whereas (–)–TAMP (3) is the partial agonist.2b In this study,
with all four stereoisomers in hand, we evaluated cytotoxicity
and other in vitro activities as a preliminary study toward
oncological application of GABA analogs. While GABAergic
system gene expression has been previously studied in patients
with neuroblastoma,28 no structure–activity relationships (SAR)
study has been reported on GABA analogs yet. Results for our
preliminary evaluations are as follows.
1)
TFA, CH2Cl2
0 °C rt
O
2) ClCOOMe
Et3N, THF
D
NPhth
O
3) NaBH4, MeOH
–78 °C
22D
18D
39% (4 steps from
)
4) phthalimide
DEAD, PPh3
benzene
>99.5% de after
recrystallization from hexane.
Structure confirmed by X–ray
(see Figure 4).
1) hydrazine
EtOH
(2S)
(1S)
2) Boc2O, Et3N
MeOH
COOH
Cytotoxicity was first evaluated on eight cell lines
(HEK293T, MRC5, MRC5 high-density, MDA–MB231,
LNCaP, A549, PC3, and Jurkat cells) by CellTiter–Glo®
Luminescent Cell Viability Assay (Promega). Selected results
are shown in Figure S1 (see the Supporting Information). It was
found some compounds are cytotoxic at 0.3 μM with low
potency. For example, (+)–CAMP (2) inhibited 16% of
proliferation of the human embryonic kidney (HEK293T) cells,
while other isomers were totally inactive. Both (–)–TAMP (3)
and (+)–TAMP (4) inhibited MRC5 (high density) (the normal
human lung fibroblast cell line) proliferation with potency of
7% and 4%, respectively, while (–)–CAMP (1) and (+)–CAMP
(2) made no effect. To A549 cells (the human epithelial of lung
carcinoma), only (+)–TAMP (4, 0.3 μM) showed 13% of
inhibition on the proliferation. Only a weak cytotoxicity was
thus observed with the GABA analogs, however, the effects
were apparently specific to the cell lines.
NH2•HCl
3) 6 M hydrochloric
acid, 60 °C
(+)–TAMP•HCl
4
( •HCl)
22D
68% (3 steps from
)
Scheme 6. Synthesis of (+)–TAMP•HCl (4•HCl)
The enantiomeric (+)–TAMP was also synthesized in this
study by employing, A) cyclopropane 19L of 86% de obtained
in Table 3 (entries 5, 6), or B) a combination of d–menthol
chiral auxiliary and quinine–derived organocatalyst 16 for
construction of the (S,S)–trans–cyclopropane core. The latter
procedure B is shown in Scheme 6. Cyclopropanation of
d–menthyl acrylate (17D), prepared from d–menthol and
acryloyl chloride in 86% yield, by using quinine-derived
catalyst 16 proceeded in 46% yield to give (S,S)–isomer 18D as
a major diastereomer. The diastereomeric ratio of 18D was
determined to be 77:23, which, without separation, was
subjected to the same reaction sequence as for the synthesis of
(–)–TAMP (Scheme 5) to give rise to diastereomerically pure
Cytotoxicity was also assayed on 53 human cancer cell
lines panel. Selected results are shown in Figure S2. It was
found (+)–CAMP (2) and (–)–TAMP (3) showed

antiproliferation activity on Caov–4 (17% inhibition at 0.3 μM)
and DMS–53 (12% inhibition at 3 μM), respectively, while
other analogs were inactive. It is again interesting that the
effects of the GABA analog diastereomers are specific to the
cancer cells.
our work is in progress to improve the activity by synthetically
modifying the structures of (+)–CAMP (2) and (–)–TAMP (3).
Inhibition of ATF6 was evaluated by the reporter gene
assay and only (+)–CAMP (2) was found to be active among
four analogs 1–4. Since some anticancer drugs inactivate ATF6,
the observed inhibition by structurally simple GABA analog,
(+)–CAMP (2), would be of interest. It should be noted here,
however, that (+)–CAMP (2) was shown to inhibit also the
tumor suppressor protein p53.
This paper describes synthesis and oncological activity of
conformationally restricted GABA analogs. (+)–CAMP (2) was
found to have weak but broad activities. The SAR data shown
here would provide a molecular basis for the oncological
activities of GABA and analogs, which are left unexplored.
Activating transcription factor (ATF6) is a type II
transmembrane
protein
containing
a
cytosolic
cAMP-responsive element-binding protein and ATF basic
leucine zipper domain, and observed as one of the unfolded
protein response (UPR) components in carcinoma of liver and
uterus.29 It has been suggested ATF6α, one of the two
homologues of ATF6 in the mammalian genome, promotes
hepatocarcinogenesis by regulation of target genes. UPR
components are one of the targets for cancer treatment, and
some drugs including retaspimycin hydrochloride (IPI–504)
inactivates the transcription factors such as ATF6.30 Here, we
have evaluated inhibition of ATF6 signaling by the four GABA
analogs 1–4 by a reporter gene assay in HEK293 cells, and
found that 10 μM of (+)–CAMP (2) weakly inhibits 17% of the
signal (Figure S3). Other analogs made no effect.
p53 is the tumor suppressor protein.31 Upon
carcinogenesis, p53 leads the cell to apoptotic cell death. By a
reporter gene assay in HEK293 cells, here, (+)–CAMP (2) was
found to weakly inhibit 23% and 28% of the p53 signaling at
0.3 and 10 μM, respectively (Figure S3). Other analogs were
totally inactive.32
Experimental
Procedures for all chemical syntheses are described in the
Supporting Information.
Single-crystal X–ray diffraction experiment of
(phthalimidomethyl)cyclopropane 22D
Single-crystal X–ray analysis of 22D was performed on a
Bruker SMART APEX CCD area (graphite–monochromated
Mo–Ka radiation ( = 0.71073 Å)) with a nitrogen flow
temperature controller. Data collection was performed at 183 K.
Empirical absorption corrections were applied using the
SADABS program. The structures were solved by direct
methods (SHELXS–97) and refined by full-matrix least squares
calculations on F2 (SHELXL–97) using the SHELX–TL
program package. Non-hydrogen atoms were refined
anisotropically; hydrogen atoms were fixed at calculated
positions and refined using a riding model. Crystallographic
data of the structure is summarized in Tables S2–S6.
CCDC–1936107 contains the supplementary crystallographic
data for this paper. The data can be obtained free of charge
from The Cambridge Crystallographic Data Centre via
http://www.ccdc.cam.ac.uk/data_request/cif.
From these cytotoxic assays and reporter gene assays,
specific actions of GABA analog stereoisomers (1–4) in
oncological aspects were shown as a preliminary SAR data.
Among them, (+)–CAMP (2) was found to show weak, but
definitive and diverse activities.
Conclusion
In this work, we have established a new synthetic entry to
the four stereoisomers of 2–aminomethyl–1–cyclopropanecar
boxylic acid. The cis analogs (CAMP) were synthesized over
10 steps starting from 2–furaldehyde (5), through
diastereoselective cyclopropane formation via diazene
intermediate 7D/7L employing d– or l–menthol as a chiral
auxiliary. On the other hand, the trans analogs (TAMP) were
Cytotoxicity assay
Antiproliferative activities of the GABA analogs 1–4
were determined upon treatment for 20 h (HEK293T, Caov–4,
DMS–53) or 72 h (MRC5–high density, A549) using the
CellTiter–Glo® luminescent Cell Viability Assay (Promega,
Madison, WI), according to the manufacture's protocol.
synthesized
by
enantioselective
organocatalytic
cyclopropanation on d– or l–menthyl acrylate as a key step,
where we observed double asymmetric induction effect. The
enantioselective synthesis of TAMP was thus successfully
achieved over 8 steps starting from d– or l–menthol.
During the synthetic studies, configurational analyses
based on experimental and theoretical 13C NMR chemical shifts
were successfully demonstrated on the intermediary diazenes
7D and 7D’. From RMS an DP4 analyses, 7D and 7D’ were
reasonably assigned as trans– and cis–isomers, respectively,
which were later confirmed by leading to (–)–CAMP (1). It was
thus shown here that the 13C analysis can be applied to bicyclic
diazenes, which cannot be readily assigned by NMR analysis
based on JH,H coupling constants.
Reporter gene assay (for inhibitory activity)
HEK293 cells seeded in 48-well plates were transfected
with pGL4.28 (Promega). Twenty-four hours after transfection,
the cells were exposed to the GABA analogs 1–4 and
thapsigargin (for ATF6,33 5 h) or doxorubicin (for p53,34 20 h).
Luciferase activities were measured using Steady–Glo®
Luciferase Assay System (Promega) and a luminescencer.
Acknowledgement
Cytotoxic activities were evaluated on diverse cell lines,
and some compounds were found to be active with low potency
but with apparent specificity; (+)–CAMP (2) and (+)–TAMP
(4) inhibited proliferation of HEK293T and A549 cell lines,
respectively, and both (–)–TAMP (3) and (+)–TAMP (4)
inhibited MRC5 (high density) cells. (+)–CAMP (2) and
(–)–TAMP (3) were found to be effective on the inhibition of
the cancer cell proliferation of Caov–4 and DMS–53,
respectively, while other analogs made no effect. Because of
the low potencies of antiproliferation on cancer cell lines panel
(see the Supporting Information for details), we have not yet
determined the cytotoxicity mechanism. To address the issues,
Daicel Co. is gratefully acknowledged for HPLC analysis
shown in Table 3. This work was supported by the grant for
2010-2012 Strategic Research Promotion (Nos. T2202, T2309,
T2401) of Yokohama City University, Japan.
Supporting Information
Synthetic procedures and characterization data for new
intermediates, procedures for single-crystal X–ray diffraction
experiment of 22D, Results for oncological assays, HPLC
profiles for diastereomer separation, and 13C NMR calculations
are
included.
This
material
is
available
on
https://doi.org/10.1246/##.

M. Yoshikawa, S. Ide, A. Suemasa, S. Kawamura, T.
Kobayashi, E. Masuda, Y. Ito, W. Hayakawa, T.
Katayama, S. Yamada, M. Arisawa, M. Minami, S. Shuto,
Bioorg. Med. Chem. 2013, 21, 4938.
References
#
Dedicated to Professor Fuyuhiko Matsuda (Hokkaido
University) on the occasion of his 65th birthday.
23. C. D. Papageorgiou, S. V. Ley, M. J. Gaunt, Angew. Chem.
Int. Ed. 2003, 42, 828.
1.
a) G. A. R. Johnston, Curr. Top. Med. Chem. 2002, 2, 903.
b) M. Ordóñez, C. Cativiela, Tetrahedron: Asymmetry
2007, 18, 3. c) R. B. Silverman, Angew. Chem. Int. Ed.
2008, 47, 3500. d) I. A. Levandovskiy, D. I. Sharapa, T. V.
Shamota, V. N. Rodionov, T. E. Shubina, Future Med.
Chem. 2011, 3, 223.
a) R. D. Allan, D. R. Curtis, P. M. Headley, G. A. R.
Johnston, D. Lodge, B. Twitchin, J. Neurochem. 1980, 34,
652. b) R. K. Duke, M. Chebib, V. J. Balcar, R. D. Allan,
K. N. Mewett, G. A. R. Johnston, J. Neurochem. 2000, 75,
2602.
24. C. D. Papageorgiou, M. A. Cubillo de Dios, S. V. Ley, M.
J. Gaunt, Angew. Chem. Int. Ed. 2004, 43, 4641.
25. a) M. N. Masuno, D. M. Young, A. C. Hoepker, C. K.
Skepper, T. F. Molinski, J. Org. Chem. 2005, 70, 4162. b)
K. He, Z. Zhou, G. Zhao, C. Tang, Heteroat. Chem. 2006,
17, 317.
2.
26. X. Sui, R. Zhu, G. Li, X. Ma, Z. Gu, J. Am. Chem. Soc.
2013, 135, 9318.
27. O. Mitsunobu, Synthesis 1981, 1981, 1.
28. S. S. Roberts, M. Mori, P. Pattee, J. Lapidus, R. Mathews,
J. P. O'Malley, Y. C. Hsieh, M. A. Turner, Z. Wang, Q.
Tian, M. J. Rodland, C. P. Reynolds, R. C. Seeger, S. R.
Nagalla, J. Clin. Oncol. 2004, 22, 4127.
29. a) C. Hetz, Nat. Rev. Mol. Cell Biol. 2012, 13, 89. b) M.
Wang, R. J. Kaufman, Nat. Rev. Cancer 2014, 14, 581.
30. J. Patterson, V. J. Palombella, C. Fritz, E. Normant,
Cancer Chemother. Pharmacol. 2008, 61, 923.
31. L. Xu, J. Massague, Nat. Rev. Mol. Cell Biol. 2004, 5,
209.
32. GABA analogs 1–4 were found to be totally inactive in
other 15 inhibition assays (SMAD signaling, Wnt
signaling, and so on).
33. Y. Wang, J. Shen, N. Arenzana, W. Tirasophon, R. J.
Kaufman, R. Prywes, J. Biol. Chem. 2000, 275, 27013.
34. Y. Wang, K.-M. Debatin, H. Hug, BMC Mol. Biol. 2001,
2, 8.
3.
4.
R. Sakai, K. Suzuki, K. Shimamoto, H. Kamiya, J. Org.
Chem. 2004, 69, 1180.
a) M. Oikawa, S. Sasaki, M. Sakai, R. Sakai, Tetrahedron
Lett. 2011, 52, 4402. b) M. Oikawa, S. Sasaki, M. Sakai,
Y. Ishikawa, R. Sakai, Eur. J. Org. Chem. 2012, 2012,
5789. c) M. Sakai, Y. Ishikawa, S. Takamizawa, M.
Oikawa, Tetrahedron Lett. 2013, 54, 5911. d) M. Sakai, K.
Tanaka, S. Takamizawa, M. Oikawa, Tetrahedron 2014,
70, 4587. e) K. Tanaka, M. Sakai, S. Takamizawa, M.
Oikawa, Chem. Lett. 2015, 44, 253.
For preliminary studies on this work, see references 6a
and 6b.
a) M. Oikawa, Y. Sugeno, Y. Ishikawa, H. Tukada, Synlett
2013, 24, 886. b) Y. Sugeno, Y. Ishikawa, M. Oikawa,
Synlett 2014, 25, 987.
R. Galeazzi, G. Mobbili, M. Orena, Tetrahedron:
Asymmetry 1997, 8, 133.
R. K. Duke, R. D. Allan, M. Chebib, J. R. Greenwood, G.
A. R. Johnston, Tetrahedron: Asymmetry 1998, 9, 2533.
I. R. Baxendale, M. Ernst, W.-R. Krahnert, S. V. Ley,
Synlett 2002, 2002, 1641.
5.
6.
7.
8.
9.
10. V. Rodríguez-Soria, L. Quintero, F. Sartillo-Piscil,
Tetrahedron 2008, 64, 2750.
11. M. T. Rispens, E. Keller, B. de Lange, R. W. J. Zijlstra, B.
L. Feringa, Tetrahedron: Asymmetry 1994, 5, 607.
12. In this paper, compound number with L denotes that
bears l-menthyl ester, and D denotes d-menthyl ester.
13. O. M. Moradei, L. A. Paquette, Org. Synth. 2003, 80, 66.
14. a) N. Grimblat, A. M. Sarotti, Chem. Eur. J. 2016, 22,
12246. b) K. Tanaka, H. Manabe, R. Irie, M. Oikawa,
Bull. Chem. Soc. Jpn. 2019, 92, in press (DOI:
10.1246/bcsj.20190096).
15. Tutorial and Userʼs Guide, Spartan '18 for Windows,
Macintosh and Linux, Wavefunction Inc., 2018.
16. T. A. Halgren, J. Comp. Chem. 1996, 17, 490.
17. W. Hehre, P. Klunzinger, B. Deppmeier, A. Driessen, N.
Uchida, M. Hashimoto, E. Fukushi, Y. Takata, J. Nat.
Prod.
2019,
published
online
(DOI:
10.1021/acs.jnatprod.9b00603).
18. S. G. Smith, J. M. Goodman, J. Am. Chem. Soc. 2010,
132, 12946.
19. K. Morokuma, R. Irie, M. Oikawa, Tetrahedron Lett.
2019, 60, 2067.
20. A. Y. Sukhorukov, S. L. Ioffe, Chem. Rev. 2011, 111,
5004.
21. a) M. Ueda, H. Miyabe, H. Sugino, O. Miyata, T. Naito,
Angew. Chem. Int. Ed. 2005, 44, 6190. b) X. Lu, L. Deng,
Angew. Chem. Int. Ed. 2008, 47, 7710.
22. a) D. J. Aitken, S. D. Bull, I. R. Davies, L. Drouin, J.
Ollivier, J. Peed, Synlett 2010, 2010, 2729. b) K. Nakada,