Structure-activity relationships of benzylidene anabaseines in nicotinic acetylcholine receptors of cockroach nerve cords

Article abstract of DOI:10.1016/S0968-0896(02)00139-6

Ten analogues of 6′-chloro-3-benzylideneanabaseine (CBA) bearing substituents at the ortho- and the para-positions of the phenyl group were synthesized, together with two related compounds. The affinity of the synthesized compounds for nicotinic acetylcholine receptors (nAChRs) in the nerve cord of the American cockroach (Periplaneta americana L.) was examined by the radioligand binding assay using [³H]epibatidine (EPI), a nAChR agonist. All 12 tested compounds inhibited [³H]EPI binding, showing K_i values ranging from 14.6 to 6830 nM. The potency variation of para-substituted CBA analogues was explained by the steric (ΔB₁) and electronic (σ_p) parameters of the para-substituents, or by the steric parameter and the charge of the N1 nitrogen atom (qN₁). Among the CBA analogues, only two compounds containing a dimethylamino group and a methoxy group at the para-position showed high insecticidal activity against the German cockroach (Blattella germanica) when injected after pretreatment with metabolic inhibitors. High-affinity analogues of CBA might be suitable probes for use in classifying and characterizing insect nAChR subtypes.

Full text of DOI:10.1016/S0968-0896(02)00139-6

Bioorganic & Medicinal Chemistry 10 (2002) 2963–2971
Structure–Activity Relationships of Benzylidene Anabaseines in
Nicotinic Acetylcholine Receptors of Cockroach Nerve Cords
Israt Sultana, Izumi Ikeda and Yoshihisa Ozoe*
Department of Life Science and Biotechnology, Faculty of Life and Environmental Science, Shimane University,
Matsue, Shimane 690-8504, Japan
Received 19 February 2002;accepted 10 April 2002
Abstract—Ten analogues of 60-chloro-3-benzylideneanabaseine (CBA) bearing substituents at the ortho- and the para-positions of
the phenyl group were synthesized, together with two related compounds. The affinity of the synthesized compounds for nicotinic
acetylcholine receptors (nAChRs) in the nerve cord of the American cockroach (Periplaneta americana L.) was examined by the
radioligand binding assay using [3H]epibatidine (EPI), a nAChR agonist. All 12 tested compounds inhibited [3H]EPI binding,
showing Ki values ranging from 14.6 to 6830 nM. The potency variation of para-substituted CBA analogues was explained by the
steric (ꢀB1) and electronic (sp) parameters of the para-substituents, or by the steric parameter and the charge of the N1 nitrogen
atom (qN1). Among the CBA analogues, only two compounds containing a dimethylamino group and a methoxy group at the para-
position showed high insecticidal activity against the German cockroach (Blattella germanica) when injected after pretreatment with
metabolic inhibitors. High-affinity analogues of CBA might be suitable probes for use in classifying and characterizing insect
nAChR subtypes. # 2002 Elsevier Science Ltd. All rights reserved.
Introduction
cloned to date do not form a robust functional ligand-
gated channel unless coexpressed with a vertebrate b
subunit.8
The neurotransmitter acetylcholine (ACh) is released
from the terminal of the presynaptic neuron, and then
binds to the nicotinic acetylcholine receptor (nAChR)
on the postsynaptic neuron to mediate fast excitatory
neurotransmission between neurons. The nicotinic acet-
ylcholine receptor is a ligand-gated ion channel that is
composed of five subunits.1 ACh and other agonists
open the channel and enhance cation permeability,
which subsequently depolarizes the membrane. In
mammals and other vertebrates, nAChRs are expressed
both at the neuromuscular junction (NMJ) and in the
nervous system (NS).2,3 The nAChR at the NMJ con-
tains a1, b1, g (or e), and d subunits, whereas the
nAChR in the NS is composed of a2–a10 and b2–b4
subunits.4,5 On the other hand, insect nAChRs, which
are found only in the NS, apparently contain both a-
type and non-a-type subunits;for example, in the fruit
fly (Drosophila melanogaster Meigen), cDNAs encoding
four a subunits [ALS, SAD (or Da2), Da3, and Da4]
and two non-a subunits (ARD and SBD) have been
cloned.6,7 However, the subunit compositions and stoi-
chiometries of native nAChRs in all insect species
remain unknown, and the insect receptor subunits
Benzylidene anabaseines, which were initially derived
from a marine worm toxin anabaseine, have been
reported to be selective agonists for a7-nAChR in
mammals, and to exhibit antagonist activity at a4b2-
nAChR.9À12 One of the analogues (GTS-21, Fig. 1) is
the first drug candidate for the Alzheimer’s disease.13
Our previous electrophysiological study showed that
benzylidene anabaseines act as agonists in the nAChRs
of the American cockroach (Periplaneta americana
L.).14 3-Benzylideneanabaseine (BA) and 60-chloro-3-
benzylideneanabaseine (CBA) displayed high affinity for
P. americana nAChRs, while the parent compound
anabaseine showed little affinity.14 The introduction of a
nitro group at the para-position of the phenyl group led
to a decrease in the affinity of both BA and CBA,
though the presence of a chlorine atom at the 60-posi-
tion of the pyridine ring had little effect on affinity.
These findings prompted us to investigate the structure–
activity relationships of this class of agonists in P.
americana nAChRs in more detail, by synthesizing
additional analogues of CBA. In the present study,
using the nAChR agonist [3H]epibatidine (EPI) as a
radioligand, we examined the affinity of CBA analogues
for P. americana nAChRs and their insecticidal activity
against the German cockroach (Blattella germanica L.).
*Corresponding author. Tel.: +81-852-32-6573;fax: +81-852-32-
6902;e-mail: ozoe-y@life.shimane-u.ac.jp
0968-0896/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(02)00139-6

2964
I. Sultana et al. / Bioorg. Med. Chem. 10 (2002) 2963–2971
We report here our results, and discuss the interaction
of agonists with insect nAChRs.
lower than that of 3 or 4. The introduction of a methoxy
group at the ortho-position decreased the potency of 5,
indicating that the sterical bulkiness of this substituent
might be unfavorable. Compound 2, with a methyl
group at the para-position, displayed moderate potency,
with a Ki of 259 nM. Compound 7, with a dimethyl-
amino group at the para-position, showed lower potency
(Ki=56.1 nM) than that of 3 or 4, and higher potency
than that of 2 or 6. Figure 3C shows the dose-response
curves of analogues containing an electron-withdrawing
group at the para-position of the phenyl group. These
compounds, excepting a fluoro analogue (8) with a
moderate potency (Ki=306 nM), demonstrated micro-
molar potency to inhibit [3H]EPI binding to P. amer-
icana nAChRs (Table 1).
Results
Chemistry
In the present study, we synthesized 10 analogues (2–11)
of CBA (1) and 13 by the reaction of an appropriate
aldehyde with 60-chloroanabaseine, which was obtained
from ethyl 6-chloronicotinate and 2-piperidone (Fig. 2,
Scheme 1). 5-Benzylidene-6-phenyl-2,3,4-trihydropyr-
idine (14) was similarly prepared by the reaction of
benzaldehyde with 2-phenyl-2,3,4,5-tetrahydropyridine,
which was obtained from ethyl benzoate and 2-piper-
idone.
To investigate the effect of replacement of the benzyli-
dene moiety with a different group, we performed a
binding assay of the cinnamylidene congener (13) of
CBA (1), which retained potency but was approximately
14-fold less potent (Ki=292 nM) than CBA (Fig. 3D,
Table 2). To determine the importance of the nitrogen
atom on the pyridine ring, we tested a compound (14)
that had a benzene ring instead of a pyridine ring.
Compound 14 (Ki=371 nM) was approximately 18-fold
less potent than CBA (1), indicating that the presence of
the nitrogen atom on the pyridine ring is important,
possibly as a hydrogen bond acceptor, but not crucial,
which finding was in accord with observations on other
nicotinic ligands.15
Potency of benzylidene anabaseines and related
compounds in P. americana nAChRs
All tested compounds inhibited specific [3H]EPI binding
to P. americana nerve cord membranes in the nanomo-
lar to micromolar range (Table 1). Figure 3A and 3B
show the dose-dependence of the inhibition of [3H]EPI
binding by CBA analogues bearing an electron-donat-
ing group(s) at the para- and the ortho-positions of the
phenyl group. Based on the IC50 values estimated from
the dose-response data, we calculated the Cheng–Pru-
soff’s inhibition constant Ki for each compound. Of the
six compounds tested,
3
(Ki=18.2 nM) and
4
(Ki=14.6 nM), with a hydroxy group(s), demonstrated
the highest potency to inhibit [3H]EPI binding. The
affinity of 5 (Ki=36.4 nM) and 6 (Ki=192 nM), bearing
a methoxy group(s), for P. americana nAChRs was
Quantitative structure–activity relationship (QSAR)
analysis
To determine which factors govern the variation in
potency, regression analysis was performed for the pKis
of CBA (1) and nine mono-substituted CBA analogues
Figure 2. Structures of synthesized benzylidene anabaseines and rela-
ted compounds.
Figure 1. Structures of nAChR agonists.

I. Sultana et al. / Bioorg. Med. Chem. 10 (2002) 2963–2971
2965
Scheme 1. Reagents: (a) LDA, Me3SiCl;(b) (1) ethyl 6-hydroxynicotinate;(2) HCl;(c) POCl 3;(d) RCHO, AcOH, AcONa.
(2, 3, 5, 7–12), using various parameters.16 The follow-
ing equation was derived using the electronic parameter
sp and the steric parameter ꢀB1 of the para-sub-
stituents:
from the axis connecting the a-atom of the substituent
with the rest of the molecule. The ꢀB1 scale is based on
H=0. Table 3 lists the pKi values and the physico-
chemical parameters used to derive the equation. Figure
4 displays a good correlation between the observed and
the calculated pKi values. Eq (1) indicates that the
potency of the compounds decreases with an increase in
the electron-withdrawing ability of the substituents, and
that the minimum width of the para-substituent of the
phenyl group is a crucial factor for the binding of the
compounds to P. americana nAChRs.
pKi ¼ À0:96ðÆ0:28Þꢀp À 2:60ðÆ0:49ÞꢀB1 þ 7:75ðÆ0:26Þ
n ¼ 10; s ¼ 0:35; r ¼ 0:96; F2;7 ¼ 39:51
ð1Þ
In this and the following equation, n is the number of
compounds included in each equation; s, the standard
deviation; r, the correlation coefficient;and F, the F
value of the correlation. B1 is a STERIMOL para-
meter17 defined as the minimum width of the substituent
Eq (2), shown below, was formulated on the basis of the
charge (qN1) of the imine nitrogen atom, which was
calculated by the semi-empirical molecular orbital
method AM1. As shown in Figure 5, the aromatic rings
of the CBA molecule thus constructed were twisted out
of the approximate plane of the tetrahydropyridine ring,
which finding is in accord with the results of the earlier
calculation for para-chloro BA.18
Table 1. Potency of CBA analogues to inhibit specific [3H]EPI bind-
ing to nerve cord membranes of American cockroaches and their
insecticidal activity against German cockroaches
pKi ¼ À88:44ðÆ37:41ÞqN1 À 2:66ðÆ0:61ÞꢀB1
À 6:01ðÆ5:99Þ
n ¼ 10; s ¼ 0:42; r ¼ 0:94; F2;7 ¼ 25:42
ð2Þ
Eq (2) indicates that the activity of the compounds
decreases as the charge of the nitrogen atom becomes
less negative. Thus, the results prove that the electronic
effects of the para-substituents are reflected in the char-
acteristics of the N1 nitrogen atom.
Compound
R1
R2
[3H]EPI binding Insecticidal activity
Ki (nM)
Mortality (%)
at 10 mg
1 (CBA)
2
3
4
5
6
7
8
9
H
CH3
OH
OH
OCH3
H
H
H
21.2 (14.6–30.7)a,b
259 (198–340)a
33.3Æ15.3c
10.0Æ10.0c
20.0Æ10.0c
10.0Æ0.0c
66.7Æ5.8c,d
10.0Æ0.0c
96.7Æ5.8c,e
13.3Æ5.8c
23.3Æ15.3c
10.0Æ0.0c
3.3Æ5.8c
Insecticidal activity
18.2 (15.7–21.1)a
On the basis of the fact that nAChR agonists such as
nicotine and imidacloprid (Fig. 1) exhibit insecticidal
activity,19 we investigated the insecticidal activity of all
synthesized compounds by injecting them into adult
male German cockroaches. Compounds containing a
methoxy group (5) and a dimethylamino group (7) at
the para-position showed the highest insecticidal activity
OH 14.6 (10.8–19.5)a
H
36.4 (26.6–49.7)a
OCH3 OCH3 192 (171–216)a
N(CH3)2
F
Cl
Br
CN
NO2
H
H
H
H
H
H
56.1 (44.1–71.1)a
306 (226–414)a
6500 (4640–9100)a
6830 (3910—11,900)a
4750 (3240–6950)a
3730 (2950–4720)a,b
10
11
12
13.3Æ5.8c
among the CBA analogues;their LD values were esti-
50
mated to be 8.43 and 2.68 mg/cockroach (Table 1),
respectively, based on the dose–mortality relationships
shown in Figure 6. CBA was almost equipotent to
(À)nicotine, which gave about 37% mortality at a dose
of 10 mg. Compounds with an electron-withdrawing
a95% confidence limits are given in parentheses.
bTaken from ref 14.
cMeanÆstandard deviation (n=3).
dLD50=8.43 (7.46–9.52) mg/roach.
eLD50=2.68 (2.21–3.24) mg/roach.

2966
I. Sultana et al. / Bioorg. Med. Chem. 10 (2002) 2963–2971
Figure 3. Dose–response curves of benzylidene anabaseines and related compounds in the inhibition of specific [3H]EPI binding to P. americana
nerve cord membranes. Membrane-bound radioactivity was measured after the incubation of membranes with [3H]EPI at 24 ꢀC for 2 h in the pre-
sence and absence of a test compound. Nonspecific binding was determined in the presence of 5 mM unlabeled (Æ)-EPI. Specific binding was cal-
culated as the difference between total and nonspecific binding. Data are the meanÆstandard deviation of three experiments, each done in duplicate,
and fitted to the logistic equation.
Discussion
group showed low insecticidal activities;these com-
pounds exhibited comparatively lower affinity for P.
americana nAChRs. However, other compounds, such
as 3 and 4, showed unexpectedly low insecticidal activ-
ities, although these compounds with an electron-
donating group(s) exhibited high affinity.
We demonstrated that benzylidene anabaseines act as
agonists in cockroach (P. americana) nAChRs. In
mammalian nAChRs, benzylidene anabaseines are ago-
nists for homooligomeric a7-nAChR but behave as
Table 2. Potency of cinnamylidene anabaseine and benzylidene phenyl trihydropyrididne to inhibit specific [3H]EPI binding to nerve cord mem-
branes of American cockroaches and their insecticidal activity against German cockroaches
Compound
R3
R4
[3H]EPI
binding
Insecticidal activity
Ki (nM)
Mortality (%)
at 10 mg
13
14
C6H5CH¼CH
60-Cl–30-C6H3N
C6H5
292 (214–398)a
371 (293–469)a
10.0Æ0.0
C6H5
30.0Æ10.0
a95% confidence limits are given in parentheses.

I. Sultana et al. / Bioorg. Med. Chem. 10 (2002) 2963–2971
2967
Table 3. Regression analysis of structure–activity relationships of 60-
chloro-3-(4-substituted benzylidene)anabaseines
methyl GTS-21 (4OH-GTS-21, Fig. 1), the major meta-
bolite of GTS-21 in humans, showed Kis of 0.17 and
0.45 mM in PC12 and SK-N-SH cell membranes,
respectively. The observation that the potency of the 4-
hydroxy analogue is higher than that of the 4-methoxy
analogue in human SK-N-SH cell membranes is the
same as that in the potencies of 4 and 6 in P. americana
nerve cord membranes.
Nicotine is an agonist in both vertebrate and inverte-
brate nAChRs.19 However, neonicotinoids such as imi-
dacloprid have been reported to act as selective agonists
for insect versus vertebrate nAChRs. At the pH of the
insect body fluid, the basic nitrogen atom on the pyrro-
lidine ring of nicotine is protonated as a one-unit posi-
tively charged center, which is one of the sites of
interaction with the receptor, as is the case in the qua-
ternary nitrogen atom of ACh. In contrast, the corre-
sponding nitrogen atom of neonicotinoids bears a
partial positive charge, which is conferred by a neigh-
boring electron-withdrawing group such as a nitro or a
cyano group.21 According to Yamamoto,19 neonico-
tinoids are supposed to interact with insect nAChRs,
but not with their vertebrate counterparts, due to this
partial positive charge of the amine nitrogen atom on
the imidazolidine ring. On the other hand, it was pro-
posed that the negatively charged tip of the nitro group
of imidacloprid is important for selective interaction
with insect nAChRs.22À24 Matsuda et al.7 suggested that
the interaction of amino acid residues of insect nAChRs
with the nitro group enhances the partial positive charge
of the tertiary amine nitrogen atom to a level compar-
able to that of the quaternary nitrogen atom of ACh.
The selectivity of neonicotinoid insecticides for insect
versus vertebrate nAChRs implies that there might be
structural differences in the agonist binding site between
animal species. The electron-withdrawing para-sub-
stituents of CBA analogues could be expected, then, to
exert effects similar to those of the nitro group of
imidacloprid on the imine nitrogen atom;that is, they
might make the imine nitrogen atom partially positive.
However, the situation differed between the two types of
compounds;the electron-withdrawing para-substituents
of CBA analogues slightly decreased the negative charge
of the imine nitrogen atom, and thereby resulted in a
reduction in affinity for P. americana nAChRs (Table 3).
Compound
R1
sp
qN1
ꢀB1
pKi
Calcd
Eq 1 Eq 2
Obsd
1
2
3
5
7
8
9
10
11
12
H
CH3
OH
0.00 À0.158 0.00 7.67 7.75 7.97
À0.17 À0.159 0.52 6.59 6.57 6.67
À0.37 À0.158 0.35 7.74 7.20 7.04
À0.27 À0.159 0.35 7.44 7.10 7.12
OCH3
N(CH3)2 À0.83 À0.161 0.35 7.25 7.64 7.30
F
Cl
Br
CN
NO2
0.06 À0.155 0.35 6.51 6.79 6.77
0.23 À0.155 0.80 5.19 5.46 5.57
0.23 À0.154 0.95 5.17 5.07 5.08
0.66 À0.151 0.60 5.32 5.56 5.75
0.78 À0.146 0.70 5.43 5.19 5.04
antagonists for heterooligomeric a4b2-nAChR.10À13 In
this respect, P. americana nAChRs that responded to
benzylidene anabaseines resemble a7-nAChR. The CBA
analogues (3, 4) with a hydroxy group(s) at the para-
position of the phenyl group displayed the highest affi-
nity for P. americana nAChRs, showing Kis of ca. 15–
18 nM. These compounds are 16- to 20-fold more
potent than (À)nicotine and only 2- to 3-fold less potent
than the nicotinic agonist insecticide imidacloprid;in
the same assay system, the Kis of (À)nicotine and imi-
dacloprid were reported to be 496 and 5.95 nM, respec-
tively.20 GTS-21, an a7-nAChR agonist drug candidate,
inhibited the specific binding of the a7-nAChR ligand
[
125I]a-bungarotoxin to rat PC12 cell membranes
(Ki=0.31 mM) and to human SK-N-SH cell membranes
(Ki=23 mM).11 The affinity of the 60-chloro analogue (6)
of GTS-21 for P. americana nAChRs is apparently
comparable to that of GTS-21 in rat a7-nAChR,
although the radioligands used are different. para-Des-
Figure 4. Relation between observed and calculated pKi values: (A) eq 1;(B) eq 2.

2968
I. Sultana et al. / Bioorg. Med. Chem. 10 (2002) 2963–2971
The results of QSAR analysis revealed that two factors
affect the binding of para-substituted CBA analogues to
P. americana nAChRs;the electronic and steric proper-
ties of the substituents. One of the two factors is the
inductive and resonance effect of the substituents. The
electron-withdrawing para-substituents reduce the elec-
tron density of the imine nitrogen atom, which is
important in interaction with nAChRs, through the
Finally, we measured the insecticidal activity of the
synthesized compounds by injecting them into German
cockroaches. Of the 14 analogues tested, only two com-
pounds with high activity in vitro showed high insecti-
cidal activity, with LD50s of <10 mg/cockroach.
However, these compounds were estimated to be
approximately 8- and 25-fold less active than imidaclo-
prid (LD50=21.7 nmol/g20). Other 12 analogues tested
showed only low insecticidal activity. Appropriate
hydrophobicity of compounds is an important factor in
their penetration into the central NS of insects. The
hydrophilic, readily ionizable properties of benzylidene
anabaseines might reduce their penetrability to the
binding site. Alternatively, they might be transformed
into amine or other inactive forms in the body of the
cockroach.
conjugated double bonds.
A similar relationship
between the pKi values and the imine nitrogen basicity
(pKa) of benzylidene anabaseines has been reported,
evidenced by their inhibition of [3H]methyl carba-
mylcholine binding to rat brain membranes.25 The pre-
sent study revealed the minimum width of the
substituents to be another important factor affecting
affinity of the analogues for P. americana nAChRs. A
portion of the space around the substituent, though not
the whole space, could be restricted in the binding site.
Conclusion
In the current study, we showed that CBA analogues
such as 1, 3, and 4 have a high affinity for cockroach (P.
americana) nAChRs. Both the electronic and the steric
properties of the para-substituent of the phenyl group
affect the potency of CBA analogues;the greater the
electron-donating ability and the smaller the minimum
width of the substituents, the higher the potency of the
analogues. A higher negative charge of the imine nitro-
gen atom results in a high affinity. The high-affinity
agonists found here might prove to be ligands useful in
the classification and characterization of insect nAChR
subtypes.
Experimental
Chemistry
General. Melting points were determined with a Yanako
MP-500D apparatus and are uncorrected. 1H NMR
spectra were measured in DMSO-d6, using tetra-
methylsilane as an internal standard on a JEOL JNM-
A400 (400 MHz) instrument. Mass spectra were
obtained at 70 eV on a Hitachi M-80B spectrometer.
General chemicals were purchased from Wako Pure
Chemical Industries, Ltd., except (Æ)-EPI and piper-
onyl butoxide (PB), which were purchased from Sigma-
Aldrich Co. and Tokyo Chemical Industry Co., Ltd.,
respectively. [3H](Æ)-EPI (1250.6 GBq/mmol) was
obtained from NEN Life Science Products, Inc. Propyl
2-propynyl phenylphosphonate (NIA16388) was a gift
from Prof. K. Nishimura, Osaka Prefecture University,
Osaka, Japan.
Figure 5. 60-Chloro-3-benzylideneanabaseine modeled by the AM1
calculation.
60-Chloro-3-(4-methylbenzylidene)anabaseine (2). 4-Methyl
benzaldehyde (280 mg, 2.34 mmol) and 60-chloro-
anabaseine (CANA)14 (227 mg, 1.17 mmol) were added
to an aqueous solution (20 mL) containing 0.6 M acetic
acid and 0.2 M sodium acetate with stirring, and the
mixture was heated at 60 ꢀC for 48 h. After cooling, the
solution was made acidic with 5% HCl and washed with
EtOAc to remove residual benzaldehyde. The solution
was then made alkaline with a saturated NaHCO3
Figure 6. Dose–mortality curves of 5 and 7 in insecticidal assays using
German cockroaches.

I. Sultana et al. / Bioorg. Med. Chem. 10 (2002) 2963–2971
2969
solution and extracted with EtOAc. The EtOAc layer
was dried over anhydrous Na2SO4 and concentrated to
give a solid mass. The crude solid was purified by silica-
gel column chromatography (n-hexane/EtOAc=2:1)
followed by recrystallization from n-hexane/EtOAc to
afford 2 (128 mg) as a white solid: yield 37%;mp 107.6–
60-Chloro-3-(4-dimethylaminobenzylidene)anabaseine (7).
The title compound was obtained from 4-dimethylami-
nobenzaldehyde and CANA, as described for 2. The
product was treated with HCl and 2-propanol, and the
resulting solid mass was recrystallized from methanol/2-
propanol to afford the dihꢀydrochloride as a red solid:
yield 40%;mp 121.2–123.7 C (dec); 1H NMR d 2.03 (q,
J=5.8 Hz, 5-CH2, 2H), 2.95 (t, J=5.8 Hz, 4-CH2, 2H),
3.06 (s, N(CH3)2, 6H), 3.75 (t, J=5.8 Hz, 6-CH2, 2H),
6.79 (d, J=9.1 Hz, Ph, 2H), 7.12 (s,=CH, 1H), 7.59 (d,
J=9.1 Hz, Ph, 2H), 7.79 (d, J=8.3 Hz, 50-CH, 1H), 8.07
(dd, J=2.4, 8.3 Hz, 40-CH, 1H), 8.62 (d, J=2.4 Hz, 20-
CH, 1H);HRMS m/z (M+) calcd for C19H20ClN3
325.1346, found 325.1370.
108.6 ꢀC; H NMR d 1.70 (q, J=6.1 Hz, 5-CH2, 2H),
1
2.30 (s, CH3, 3H), 2.76 (t, J=6.1 Hz, 4-CH2, 2H), 3.74
(t, J=6.1 Hz, 6-CH2, 2H), 6.54 (s,=CH, 1H), 7.19 (d,
J=8.1 Hz, Ph, 2H), 7.29 (d, J=8.1 Hz, Ph, 2H), 7.55 (d,
J=8.3 Hz, 50-CH, 1H), 7.91 (dd, J=2.4, 8.3 Hz, 40-CH,
1H), 8.46 (d, J=2.4 Hz, 20-CH, 1H);HRMS m/z (M+)
calcd for C18H17ClN2 (M+) 296.1080, found 296.1077.
60-Chloro-3-(4-hydroxybenzylidene)anabaseine (3). The
title compound was obtained from 4-hydroxy-
benzaldehyde and CANA, as described for 2. 3: a light
yellow solid;yield 36%;mp 211.8–212.0 ꢀC; 1H NMR d
1.71 (q, J=5.7 Hz, 5-CH2, 2H), 2.76 (t, J=5.7 Hz,
4-CH2, 2H), 3.71 (t, J=5.7 Hz, 6-CH2, 2H), 6.47
(s,=CH, 1H), 6.77 (d, J=8.7 Hz, Ph, 2H), 7.25 (d,
J=8.7 Hz, Ph, 2H), 7.54 (d, J=8.3 Hz, 50-CH, 1H), 7.89
(dd, J=2.6, 8.3 Hz, 40-CH, 1H), 8.44 (d, J=2.6 Hz, 20-
CH, 1H), 9.76 (s, OH, 1H);HRMS m/z (M+) calcd for
C17H15ClN2O 298.0873, found 298.0865.
60-Chloro-3-(4-fluorobenzylidene)anabaseine (8). The title
compound was obtained from 4-fluorobenzaldehyde
and CANA, as described for 2. 8: a white solid;yield
56%;mp 87.4–87.8 C; H NMR d 1.72 (q, J=6.1 Hz,
5-CH2, 2H), 2.75 (dt, J=2.0, 6.1 Hz, 4-CH2, 2H), 3.75
(t, J=6.1 Hz, 6-CH2, 2H), 6.60 (s,=CH, 1H), 7.22 (d,
J=7.8 Hz, Ph, 1H), 7.24 (d, J=7.8 Hz, Ph, 1H), 7.47 (d,
J=7.8 Hz, Ph, 1H), 7.50 (d, J=7.8 Hz, Ph, 1H), 7.56 (d,
J=8.3 Hz, 50-CH, 1H), 7.93 (dd, J=2.4, 8.3 Hz, 40-CH,
1H), 8.48 (d, J=2.4 Hz, 20-CH, 1H);HRMS m/z (M+)
calcd for C17H14ClFN2 300.0830, found 300.0830.
ꢀ
1
60-Chloro-3-(2,4-dihydroxybenzylidene)anabaseine
The title compound was obtained from 2,4-dihydroxy-
(4).
60-Chloro-3-(4-chlorobenzylidene)anabaseine (9). The
title compound was obtained from 4-chlorobenzald-
ehyde and CANA, as described for 2. 9: a white solid;
yield 69%;mp 109.4–109.6 ꢀC; 1H NMR d 1.72 (q,
J=6.2 Hz, 5-CH2, 2H), 2.75 (dt, J=1.9, 6.2 Hz, 4-CH2,
2H), 3.75 (t, J=6.2 Hz, 6-CH2, 2H), 6.59 (s,=CH, 1H),
7.44 (2s, Ph, 4H), 7.56 (d, J=8.3 Hz, 50-CH, 1H), 7.93
(dd, J=2.3, 8.3 Hz, 40-CH, 1H), 8.48 (d, J=2.3 Hz, 20-
CH, 1H);HRMS m/z (M+) calcd for C17H14Cl2N2
316.0534, found 316.0486.
benzaldehyde and CANA, as described for 2. 4: an
orange solid;yield 46%;mp 163.1–166.4 ꢀC (dec); H
1
NMR d 1.70 (q, J=5.4 Hz, 5-CH2, 2H), 2.70 (t,
J=5.4 Hz, 4-CH2, 2H), 3.73 (br, 6-CH2, 2H), 6.26–6.30
(m, Ph, 1H), 6.67 (s,=CH, 1H), 7.20 (d, J=8.3 Hz, Ph,
1H), 7.55 (d, J=8.1 Hz, 50-CH, 1H), 7.88 (dd, J=2.2,
8.1 Hz, 40-CH, 1H), 8.43 (d, J=2.2 Hz, 20-CH, 1H), 9.56
(s, OH, 1H), 9.59 (s, OH, 1H);HRMS m/z (M+) calcd
for C17H15ClN2O2 314.0822, found 314.0846.
60-Chloro-3-(4-methoxybenzylidene)anabaseine (5). The
title compound was obtained from 4-methoxy-
benzaldehyde and CANA, as described for 2. 5: a
cream-colored solid;yield 43%;mp 78.4–78.6 ꢀC; 1H
NMR d 1.71 (q, J=6.1 Hz, 5-CH2, 2H), 2.77 (dt,
J=2.0, 6.1 Hz, 4-CH2, 2H), 3.73 (t, J=6.1 Hz, 6-CH2,
2H), 3.77 (s, OCH3, 3H), 6.54 (s,=CH, 1H), 6.96 (d,
J=8.8 Hz, Ph, 2H), 7.36 (d, J=8.8 Hz, Ph, 2H), 7.55
(d, J=8.3 Hz, 50-CH, 1H), 7.91 (dd, J=2.4, 8.3 Hz, 40-
CH, 1H), 8.45 (d, J=2.4 Hz, 20-CH, 1H);HRMS m/z
(M+) calcd for C18H17ClN2O 312.1029, found
312.1032.
60-Chloro-3-(4-bromobenzylidene)anabaseine (10). The
title compound was obtained from 4-bromobenzalde-
hyde and CANA, as described for 2. 10: a white solid;
yield 30%;mp 119.5–119.8 ꢀC; 1H NMR d 1.71 (q,
J=6.2 Hz, 5-CH2, 2H), 2.74 (dt, J=2.0, 6.2 Hz, 4-CH2,
2H), 3.75 (t, J=6.2 Hz, 6-CH2, 2H), 6.57 (s,=CH, 1H),
7.35 (d, J=8.5 Hz, Ph, 2H), 7.56 (d, J=8.2 Hz, 50-H, 1H),
7.58 (d, J=8.5 Hz, Ph, 2H), 7.93 (dd, J=2.5, 8.2 Hz, 40-
CH, 1H), 8.48 (d, J=2.5 Hz, 20-CH, 1H);HRMS m/z
(M+) calcd for C17H14BrClN2 360.0029, found 360.0135.
60-Chloro-3-(4-cyanobenzylidene)anabaseine (11). The
title compound was obtained from 4-cyanobenzalde-
hyde and CANA, as described for 2. 11: a white solid;
yield 53%;mp 140.1–142.6 ꢀC; 1H NMR d 1.71 (q,
J=5.9 Hz, 5-CH2, 2H), 2.75 (t, J=5.9 Hz, 4-CH2, 2H),
3.78 (t, J=5.9 Hz, 6-CH2, 2H), 6.66 (s,=CH, 1H), 7.55
(d, J=8.4 Hz, 50-H, 1H), 7.58 (d, J=8.3 Hz, Ph, 2H),
7.83 (d, J=8.3 Hz, Ph, 2H), 7.94 (d, J=8.4 Hz, 40-CH,
1H), 8.49 (s, 20-CH, 1H);HRMS calcd for C 18H14ClN3
m/z (M+) 307.0876, found 307.0875.
60-Chloro-3-(2,4-dimethoxybenzylidene)anabaseine (6).
The title compound was obtained from 2,4-dimethoxy-
benzaldehyde and CANA, as described for 2. 6: a yellow
solid;yield 35%;mp 118.5–119.3 ꢀC; H NMR d 1.70
1
(q, J=5.9 Hz, 5-CH2, 2H), 2.68 (t, J=5.9 Hz, 4-CH2,
2H), 3.70 (s, OCH3, 3H), 3.76 (t, J=5.9 Hz, 6-CH2, 2H),
3.79 (s, OCH3, 3H), 6.55–6.60 (m, Ph, 2H), 6.61
(s,=CH, 1H), 7.32 (d, J=8.3 Hz, Ph, 1H), 7.56 (d,
J=8.1 Hz, 50-CH, 1H), 7.90 (dd, J=2.4, 8.1 Hz, 40-
CH, 1H), 8.45 (d, J=2.4 Hz, 20-CH, 1H);HRMS m/z
(M+) calcd for C19H19ClN2O2 342.1135, found
342.1150.
60-Chloro-3-(cinnamylidene)anabaseine (13). The title
compound was obtained from cinnamaldehyde and
CANA, as described for 2. The product was treated

2970
I. Sultana et al. / Bioorg. Med. Chem. 10 (2002) 2963–2971
with HCl and 2-propanol, and the resulting solid mass
was recrystallized from methanol/2-propanol to afford
the dihydrochloride as a light green solid: yield 27%;
mp 150.2–157.4 ꢀC (dec); 1H NMR d 2.05 (q, J=5.7 Hz,
5-CH2, 2H), 2.93 (t, J=5.7 Hz, 4-CH2, 2H), 3.79 (t,
J=5.7 Hz, 6-CH2, 2H), 7.05 (d, J=11.2 Hz,=CH, 1H),
7.34–7.72 (m,=CH, Ph, 7H), 7.84 (d, J=8.3 Hz, 50-CH,
1H), 8.11 (dd, J=2.4, 8.3 Hz, 40-CH, 1H), 8.65 (d,
J=2.4 Hz, 20-CH, 1H);HRMS m/z (M+) calcd for
C19H17ClN2 308.1080, found 308.1074.
(7 mL), and the suspension was centrifuged in the same
manner as above. The supernatant was decanted, and
the surface of the resulting pellet was washed with buf-
fer A (2 mL) again. The pellet was then suspended in
ice-cold 50 mM Tris–HCl containing 120 mM NaCl (pH
7.4) (buffer B) and stored at À80 ꢀC until use. The pro-
tein-dye binding method27 was used to determine pro-
tein concentration, with bovine serum albumin (BSA) as
a standard.
Binding assays with [3H]EPI. The procedure of this
assay is based generally on the method of Orr et al.26
Either a 20-mL aliquot of buffer B for the determination
of total binding, a 20-mL aliquot of buffer B containing
unlabeled (Æ)-EPI (final concn 5 mM) for the determi-
nation of nonspecific binding, or a 20-mL aliquot of
buffer B containing a test compound for the determina-
tion of inhibitory activity was added to test tubes. Both
buffer B (100 mL) containing [3H]EPI (final concn 2 nM)
and buffer B (80 mL) containing nerve cord membranes
(40 mg as protein) were added to all test tubes, which
were then incubated at 24 ꢀC for 2 h. Whatman GF/C
filters were presoaked in buffer B containing 10 mg of
BSA/mL for 2 h. The reaction was terminated by rapid
filtration through the GF/C filters under reduced pres-
sure, using a Brandel M-24 cell harvester. The filters
were rapidly washed twice with 2 mL of 10 ꢀC-cold buf-
fer B containing 2 mg of BSA/mL. The filter disks were
removed and placed in 10 mL of toluene–Methyl Cello-
solve-based scintillation fluid, and the bound radio-
activity was determined using a Beckman LS 6000SE
liquid scintillation counter. Test compounds were first
dissolved in DMSO and diluted with buffer B. The final
concentrations of DMSO in assay solutions were lower
than 0.1% (v/v), which did not affect the binding activ-
ity of membranes. Each experiment was performed in
duplicate and repeated at least three times. IC50 values
were estimated from the mean values of inhibition per-
centage at three to five compound concentrations, using
the Probit method. The inhibition constant Ki was cal-
culated according to the method of Cheng and Prus-
off.28
6-Phenyl-2,3,4,5-tetrahydropyridine (PTP). Diisopropy-
lamine (6.0 mL, 45 mmol) and 1.6 M n-butyllithium/
hexane (30 mL, 48 mmol) were added to dry THF
(21 mL) with stirring at À70 ꢀC under a nitrogen atmo-
sphere. 2-Piperidone (4.2 g, 14 mmol) in dry THF
(20 mL) and trimethylsilyl chloride (6.0 mL, 42 mmol)
were added to the solution, stirred at À70 ꢀC for 15 min,
and warmed to room temperature (solution A). Diiso-
propylamine (6.0 mL, 45 mmol) and 1.6 M n-butyl-
lithium/hexane (30 mL, 48 mmol) were added to dry
THF (21 mL) with stirring at À70 ꢀC under a nitrogen
atmosphere (solution B). Solution A was added to
solution B and stirred at À70 ꢀC after 15 min. Ethyl
benzoate (1.75 g, 11.7 mmol) was added to the mixture
and stirred for 15 min. The reaction mixture was
warmed to room temperature and stirred for 18 h. The
resulting precipitate was collected, dissolved in HCl
(20 mL), and heated under reflux at 120 ꢀC for 24 h.
After cooling, the HCl was removed by an evaporator.
The brown residue was recrystallized from methanol/
acetone to afford the hydrochloride salt of PTP (611 mg,
33%) as a gray solid: mp 73.8–74.5 ꢀC (dec); 1H NMR d
1.89 (m, 4-CH2, 5-CH2, 4H), 3.28 (br s, 3-CH2, 2H),
3.75 (br s, 6-CH2, 2H), 7.65 (t, J=7.6 Hz, Ph, 2H), 7.77
(t, J=7.6 Hz, Ph, 1H), 8.00 (d, J=7.6 Hz, Ph, 2H);
HRMS m/z (M+) calcd for C11H13N 159.1048, found
159.1050.
5-Benzylidene-6-phenyl-2,3,4-trihydropyridine (14). The
title compound was obtained from benzaldehyde and
PTP, as described for 2. 14: a light yellow solid;yield
9.0%;mp 139.1–141.8 ꢀC (dec); 1H NMR d 1.71 (q,
J=6.1 Hz, 5-CH2, 2H), 2.77 (dt, J=1.6, 6.1 Hz, 4-CH2,
2H), 3.71 (t, J=6.1 Hz, 6-CH2, 2H), 6.58 (s,=CH, 1H),
7.28–7.96 (m, Ph, 10H);HRMS m/z (M+) calcd for
C18H17N 247.1361, found 247.1386.
Insecticidal assays. Adult male German cockroaches
(Blattella germanica L.) were immobilized in an ice-cold
jar, and an acetone solution (1 mL) containing PB
(30 mg) and NIA16388 (10 mg) was topically applied to
each roach abdomen. After being kept at 25 ꢀC for 1 h,
the cockroaches were again immobilized in an ice-cold
jar. A DMSO solution (0.25 mL) containing a test com-
pound was injected into the ventral side of each roach
abdomen. The cockroaches were kept with water at
25 ꢀC, and after 24 h, the mortality rate was recorded.
Each dose was given to 10 cockroaches, and each assay
was performed in triplicate. LD50 values were calculated
from the mean values of mortality at five concentra-
tions, using the Probit method.
Biology
Preparation of membranes from cockroach nerve cords.
The procedure of Orr et al.26 was modified to prepare
the membrane fraction from the ventral nerve cords of
cockroaches. Adult male American cockroaches (P.
americana L.) were dissected in ice-cold 50 mM Tris–
HCl buffer containing 200 mM sucrose and 1 mM
EDTA (pH 7.4) (buffer A). The nerve cords isolated
from 10 cockroaches were homogenized with a glass–
Teflon homogenizer (25 strokes) in ice-cold buffer A
(2 mL). The homogenate was then centrifuged at
25,000g for 30 min. The supernatant was decanted, and
the surface of the resulting pellet was washed with buf-
fer A (2 mL). The pellet was then suspended in buffer A
Molecular modeling and QSAR analysis
The structure of CBA (1) was modeled using the mole-
cular modeling software SYBYL, version 6.5.29 The
coordinate was calculated using the SYBYL standard

I. Sultana et al. / Bioorg. Med. Chem. 10 (2002) 2963–2971
2971
values for bond lengths and angles. The initial coordi-
nate thus obtained was fully optimized by the semi-
empirical molecular orbital method AM1. The pKi
values [log(1/Ki (M))] of 10 para-substituted CBA ana-
logues in the inhibition of [3H]EPI binding to P. amer-
icana nerve cord membranes were analyzed using the
QSAR analysis software OREG 2.05.30
10. de Fiebre, C. M.;Meyer, E. M.;Henry, J. C.;Muraskin,
S. I.;Kem, W. R.;Papke, R. L. Mol. Pharmacol. 1995, 47, 164.
11. Meyer, E. M.;Kuryatov, A.;Gerzanich, V.;Lindstrom,
J.;Papke, R. L. J. Pharmacol. Exp. Ther. 1998, 287, 918.
12. Papke, R. L.;Meyer, E.;Nutter, T.;Uteshev, V. V. Eur. J.
Pharmacol. 2000, 393, 179.
13. Kem, W. R. Behav. Brain Res. 2000, 113, 169.
14. Sultana, I.;Hosokawa, C.;Nishimura, K.;Ikeda, I.;Ozoe,
Y. Insect Biochem. Mol. Biol. 2002, 32, 637.
15. Nishiwaki, H.;Nakagawa, Y.;Takeda, D. Y.;Okazawa,
A.;Akamatsu, M.;Miyagawa, H.;Ueno, T.;Nishimura, K.
Pest Manag. Sci. 2000, 56, 875.
Acknowledgements
16. Hansch, C.;Leo, A. (Eds.) Exploring QSAR: Funda-
mentals and Applications in Chemistry and Biology;American
Chemical Society: Washington, DC, 1995.
We thank T. Okumoto for his helpful advice.
17. Verloop, A.;Hoogenstraaten, W.;Tipker, J. In
Design;Arie ns, E. J., Ed.;Academic: New York, 1976;Vol. 7,
p 165.
Drug
References and Notes
1. Tierney, M. L.;Unwin, N. J. Mol. Biol. 2000, 303, 185.
2. McGehee, D. S.;Role, L. W. Annu. Rev. Physiol. 1995, 57,
521.
3. Lindstrom, J. In Ion Channels;Narahashi, T., Ed.;Plenum:
New York, 1996;Vol. 4, p 377.
18. Zoltewicz, J. A.;Prokai-Tatrai, K.;Bloom, L. B.;Kem,
W. R. Heterocycles 1993, 35, 171.
19. Yamamoto, I. In Nicotinoid Insecticides and the Nicotinic
Acetylcholine Receptor;Yamamoto, I.;Casida, J. E., Eds.;
Springer: Tokyo, 1999, p 3.
4. Lukas, R. J.;Changeux, J. P.;Le Nove re, N.;Albuquerque,
E. X.;Balfour, D. J. K.;Berg, D. K.;Bertrand, D.;Chiappi-
nelli, V. A.;Clarke, P. B. S.;Collins, A. C.;Dani, J. A.;
Grady, S. R.;Kellar, K. J.;Lindstrom, J. M.;Marks, M. J.;
Quik, M.;Taylor, P. W.;Wonnacott, S. Pharmacol. Rev. 1999,
51, 397.
5. Alexander, S. P. H.;Mathie, A.;Peters, J. A. (Compilers),
Trends in Pharmacological Science Nomenclature Supplement,
12th ed.;Elsevier: London, 2001;p 7.
20. Mori, K.;Okumoto, T.;Kawahara, N.;Ozoe, Y.
Manag. Sci. 2002, 58, 190.
21. Tomizawa, M.;Yamamoto, I. J. Pesticide Sci. 1993, 18, 91.
22. Kagabu, S.;Matsuno, H. J. Agric. Food Chem. 1997, 45,
276.
23. Nakayama, A.;Sukekawa, M. Pestic. Sci. 1998, 52, 104.
24. Tomizawa, M.;Lee, D. L.;Casida, J. E. J. Agric. Food
Chem. 2000, 48, 6016.
25. Kem, W. R. Invertebr. Neurosci. 1997, 3, 251.
26. Orr, N.;Shaffner, A. J.;Watson, G. B. Pestic. Biochem.
Physiol. 1997, 58, 183.
Pest
6. Tomizawa, M.;Casida, J. E. Pest Manag. Sci. 2001, 57,
914.
7. Matsuda, K.;Buckingham, S. D.;Kleier, D.;Rauh, J. J.;
Grauso, M.;Sattelle, D. B. Trends Pharmacol. Sci. 2001, 22,
573.
27. Bradford, M. M. Anal. Biochem. 1976, 72, 248.
28. Cheng, Y.;Prusoff, W. H. Biochem. Pharmacol. 1973, 22,
3099.
8. Schlutz, R.;Bertrand, S.;Chamaon, K.;Smalla, K.-H.;
Gundelfinger, E. D.;Bertrand, D. J. Neurochem. 2000, 74,
2537.
9. Hunter, B. E.;de Fiebre, C. M.;Papke, R. L.;Kem, W. R.;
Meyer, E. M. Neurosci. Lett. 1994, 168, 130.
29. SYBYL Molecular Modeling Software;Tripos Associates:
St. Louis, MO, USA.
30. Asao, M.;Shimizu, R.;Nakao, K.;Nishioka, T.;Fujita,
T. OREG 2.05;Japan Chemistry Program Exchange/Tanabe
Seiyaku Co.: Osaka, Japan.