Nudicauline and elatine as potent norditerpenoid ligands at rat neuronal α-bungarotoxin binding sites: Importance of the 2-(methylsuccinimido)benzoyl moiety for neuronal nicotinic acetylcholine receptor binding

Article abstract of DOI:10.1021/jm9604991

Methyllycaconitine (MLA, 1) is a novel, potent probe for mammalian and insect nicotinic acetylcholine receptors (nAChR) and displays remarkable selectivity toward neuronal [¹²⁵I]-α-bungarotoxin (αBgTX) binding sites that correspond to α7-type n

Full text of DOI:10.1021/jm9604991

4860
J . Med. Chem. 1996, 39, 4860-4866
Notes
Nu d ica u lin e a n d Ela tin e a s P oten t Nor d iter p en oid Liga n d s a t Ra t Neu r on a l
r-Bu n ga r otoxin Bin d in g Sites: Im p or ta n ce of th e 2-(Meth ylsu ccin im id o)ben zoyl
Moiety for Neu r on a l Nicotin ic Acetylch olin e Recep tor Bin d in g
David J . Hardick, Ian S. Blagbrough,* Gary Cooper, Barry V. L. Potter, Trevor Critchley,† and
Susan Wonnacott†
Department of Medicinal Chemistry, School of Pharmacy and Pharmacology, and School of Biology and Biochemistry,
University of Bath, Bath BA2 7AY, U.K.
Received J uly 5, 1996X
Methyllycaconitine (MLA, 1) is a novel, potent probe for mammalian and insect nicotinic
acetylcholine receptors (nAChR) and displays remarkable selectivity toward neuronal [125I]-
R-bungarotoxin (RBgTX) binding sites that correspond to R7-type nAChR in mammalian brain.
We have shown that, among a number of selected norditerpenoid alkaloids, elatine (2) and
nudicauline (3) are equipotent with, or better than, MLA (1) in binding to brain [125I]-RBgTX
binding sites, with IC50 values of 6.1, 1.7, and 7.6 nM, respectively. The 2-((S)-methylsuccin-
imido)benzoyl moiety of these ligands is crucial for high-affinity binding, whereas structural
modifications to the norditerpenoid core of the ligand can be tolerated without loss of activity
or selectivity. In addition to MLA (1), elatine (2), and nudicauline (3), we have examined
lycoctonine (4), inuline (6), lappaconitine (7), N-desacetyllappaconitine (8), delsoline (10),
delcorine (11), deltaline (12), condelphine (13), and karacoline (14). This study therefore extends
the range of norditerpenoids, other than MLA, which can be used to probe this important class
of nAChR. All 12 alkaloids were assessed for activity at [3H]nicotine binding sites which are
1
considered to represent R4â2 nAChR. Furthermore, the H and 13C NMR spectroscopic data
of MLA and elatine have been critically compared.
In tr od u ction
loids nicotine and muscarine were used to distinguish
nicotinic and muscarinic acetylcholine responses.10 More
recently, the snake polypeptide toxin RBgTX has been
crucial in the characterization of muscle nAChR11 and
certain neuronal nAChR that correlate with functional
receptors reconstituted from R7, R8, and R9 subunits
expressed in Xenopus oocytes.7-9,12
Methyllycaconitine (MLA, 1), isolated and purified
from various species of Delphinium and Consolida, is a
hexacyclic norditerpenoid alkaloid which is esterified at
the C18 oxygen atom with the 2-((S)-methylsuccinimi-
do)benzoyl moiety.1,2 MLA (1) is the most potent,
nonproteinaceous competitive antagonist at neuronal
Several hundred lycoctonine-type norditerpenoid al-
kaloids have now been isolated from Delphinium and
Consolida,13-15 and some of these may have similar
potential to MLA (1) in terms of potency and selectiv-
ity.16,17 A few of these natural products (<5%) possess
the 2-((S)-methylsuccinimido)benzoyl moiety of MLA (1).
Elatine (2) and nudicauline (3) (Figure 1) are two such
ligands which differ from MLA (1) within the norditer-
penoid core but contain this anthranilate ester moiety.
These probes are candidates for exploring the differ-
ences in the agonist/competitive antagonist recognition
site between various members of the nAChR family and
are lead compounds for the design and synthesis of novel
high-affinity ligands for neuronal nAChR subtypes.
Benn and J acyno have reviewed the chemical and
biological perspectives of these norditerpenoid alka-
loids.16 Indeed, J acyno and colleagues have recently
published that lycaconitine (similar to MLA (1), but
lacking the angular methyl group on the succinimide)
is 5-fold less active than MLA.17
vertebrate and invertebrate
[
125I]-R-bungarotoxin
(RBgTX) binding sites,3,4 which, in mammalian brain,
correspond to nicotinic acetylcholine receptors (nAChR)
containing the R7 subunit.4,5 Unlike RBgTX, MLA (1)
displays a 1000-fold preference for the neuronal RBgTX-
sensitive nAChR subtypes relative to muscle nAChR3,6
and blocks R7-like nAChR in hippocampal neurons and
Xenopus oocytes with picomolar affinity.5,7
nAChR are pentameric ligand-gated cation channels.
Whereas the well-characterized nAChR from vertebrate
muscle and Torpedo electroplaques are comprised of two
R and one each of â, γ (or ꢀ), and δ subunits, a plethora
of distinct but homologous subunits have been identified
in mammalian and avian nervous systems.8,9 Thus,
there is potential for enormous heterogeneity of neu-
ronal nAChR, and the identification and characteriza-
tion of these distinct nAChR would be greatly facilitated
by a range of selective probes. Historically, the alka-
* Author for correspondence: tel, (44-1225) 826795; fax, (44-1225)
826114; e-mail, i.s.blagbrough@bath.ac.uk.
We have recently established18 that hydrolysis of the
C18 ester bond in MLA (1) to yield lycoctonine (4)
† School of Biology and Biochemistry.
X Abstract published in Advance ACS Abstracts, October 15, 1996.
S0022-2623(96)00499-2 CCC: $12.00 © 1996 American Chemical Society

Notes
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 24 4861
F igu r e 1. Structures of norditerpenoid alkaloids.
greatly diminishes nicotinic potency. In contrast, ad-
dition of the 2-((S)-methylsuccinimido)benzoyl moiety
to the norditerpenoid alkaloid aconitine (5) converts it
from a sodium channel activator with only weak affinity
for rat brain [125I]-RBgTX binding sites into a nicotinic
ligand with potency comparable to that of MLA (1).18
We now demonstrate the importance of this moiety for
nicotinic potency, and we report that a number of
natural and semisynthetic alkaloids which contain this
ester group are potent neuronal nAChR probes, as a
result of our biological analysis of 12 norditerpenoid
alkaloids at R7 and R4â2 nAChR.

4862 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 24
Notes
Ch em istr y
order at 48.42 and 39.86 ppm, respectively.21 Further-
more, we have assigned the elatine resonances at 36.99
and 35.39/35.19 ppm as the C2′′ and C3′′ succinimido
carbons, respectively, based on Pelletier’s assignments
of 37.0 (C2′′ succinimide) and 35.3 (C3′′ succinimide)
ppm for elatine. However, in our DEPT and correlation
spectroscopic studies with MLA, the C3′′ methylene
succinimido carbon resonates at 37.00 ppm, while the
C2′′ methine succinimido signal is unambiguously as-
signed to the split peak at 35.45 and 35.21 ppm. The
inconsistency in these literature assignments for elatine
is due to incorrect numbering of the methylsuccinimido
ring rather than an error in the interpretation of the
13C NMR spectrum. The C5′′ succinimido methyl
carbon resonance in MLA also appears as either a
broadened signal or a split peak (vide infra) at ap-
proximately 16.4 ppm, and the same phenomenon is also
observed in elatine. If the signal at 36.99 ppm, in our
spectrum, can be attributed to the succinimido meth-
ylene carbon, then it is possible that the split peaks
observed for the methine and methyl carbons are a
result of racemization at the chiral center under the
experimental conditions employed. However, this is
unlikely as this split peak phenomenon is also observed
with MLA which has been unambiguously synthesized
from inuline and homochiral (S)-methylsuccinic anhy-
dride where racemization of the carbon bearing the
methyl group is unlikely to occur.23 In the published
13C NMR spectral data for elatine (2), the signals at 28.4
and 27.9 ppm have been assigned to C2 and C12,
respectively.21 In the spectral data obtained with our
samples of elatine, these signals resonated at 27.80 and
26.36 ppm. However, in the 13C NMR spectra of MLA
(1), C12 resonates at lower field (28.70 ppm) compared
to C2 (26.07 ppm) unambiguously identified by DEPT
and 13C/1H heteronuclear correlation spectroscopy. Fol-
lowing comparison with our MLA 13C NMR data, it is
possible that the published assignments for C2/C12, C9/
C10, and the succinimido methine/methylene in elatine
(2) may have to be revised (interchanged).
In order to obtain alkaloids to test the structure-
activity relationships among these substituted nor-
diterpenoids quickly, we used readily available MLA (1)
as the template for functional group transformations.
This alkaloid was isolated and purified from Garden
Hybrid Delphinium according to literature procedures.1,2
Saponification of ester 1, with sodium hydroxide, af-
forded neopentyl-like alcohol 4 which could be readily
purified from even trace amounts of MLA (1).19 The
high affinity (nanomolar) of this nAChR antagonist
means that essentially all traces of this starting mate-
rial must be removed prior to assaying the resulting
semisynthetic ligands.19 Lycoctonine (4) was prepared
free from detectable MLA (1) and then converted into
inuline (6) by treatment with isatoic anhydride cata-
lyzed by 4-(N,N-dimethylamino)pyridine. Inuline (6)
was also synthesized by the controlled hydrolysis of
succinimide 1, but we found that inuline (6) and any
small quantities of unreacted MLA (1) were difficult to
separate by silica gel column chromatography, making
the anthranoylation of lycoctonine (4) a more expedient
route to the free aniline 6 than the partial hydrolysis
of MLA (1). Elatine (2) was prepared by the reaction
of MLA (1) with a large excess of formaldehyde diethy-
lacetal (as the solvent), in order to introduce the C7/C8
methylenedioxy bridge.20,21 We found this procedure to
be more practical than azeotropic distillation (Dean-
Stark) from benzene-formalin.21
NMR Da ta
The 1H NMR data for lycoctonine (4) are in good
agreement with the published data in the recent com-
1
prehensive review of the H NMR spectra of norditer-
penoid alkaloids by Hanuman and Katz.22 For lycoc-
tonine (4), the C8 hydroxyl functionality was assigned
at δ 4.15 by comparison with the H NMR spectrum of
MLA (1). The C18 hydroxyl functionality was assigned
within δ 3.58-3.62 following the conversion of lycocto-
nine (4) into inuline (6). The C7 hydroxyl functionality
1
1
Resu lts a n d Discu ssion
was not observed for 4 (in CDCl3). In the 400 MHz H
NMR spectrum of semisynthetic inuline (6) (from ly-
coctonine), the C18 methylene protons resonated as two
doublets at δ 4.12 and 4.13 (J AB), and the signal at δ
1.66 was assigned to the H7 hydroxyl proton. These
spectral data are in good agreement with those obtained
with inuline (6) obtained by the partial hydrolysis of
MLA (1). The 1H NMR spectrum of elatine (2) displayed
the characteristic methylenedioxy functional group
resonance at δ 5.07 and broadened succinimido reso-
nances in the 13C NMR spectrum similar to those
observed for MLA (1).
The 13C NMR data for elatine (2) are all in good
agreement with those previously assigned by Pelletier
and co-workers, and the assignments presented here are
based upon those published data.21 There is consider-
able similarity between the structures of the norditer-
penoid alkaloids MLA (1) and elatine (2), the only
difference being the introduction of a C7/C8 methylene-
dioxy bridge in 2 compared to the two tertiary alcohols
in 1 (see Figure 1). Interestingly, the C9 and C10
carbon resonances appear at significantly different
frequencies to those found in MLA (1). Hence, our
results indicate that C9 and C10 resonate at δ 43.22
and 46.07, respectively, in MLA (1), compared to elatine
(2) where C9 and C10 have been assigned in the reverse
The binding constants (IC50) of 12 lycoctonine- and
aconitine-type norditerpenoid alkaloids for inhibition of
[
125I]-RBgTX binding to rat brain membranes are given
in Table 1; inhibition curves for six of these compounds,
representing the range of potencies observed, are il-
lustrated in Figure 2. Three of these compounds, MLA
(1), elatine (2), and nudicauline (3), are potent inhibitors
of binding with IC50 values of 7.6, 6.1, and 1.7 nM,
respectively, showing that nudicauline (3) is even more
potent than MLA (1). These high-affinity alkaloids are
distinguished by the presence of the 2-((S)-methylsuc-
cinimido)benzoyl moiety, but they differ in substituents
around the norditerpenoid core. In comparison, the
greatly reduced potency of lycoctonine (4) emphasizes
the importance of the 2-((S)-methylsuccinimido)benzoyl
moiety for activity.
We have previously postulated6 that the ester within
MLA (1) provides a carbonyl oxygen that, together with
the tertiary nitrogen atom, conforms to the Beers and
Reich pharmacophore24 for nicotinic binding. Thus, the
ester carbonyl oxygen in MLA (1) could donate a lone
pair of electrons to form a hydrogen bond at the ligand
binding site. However, inuline (6) is the anthranoyl
ester of lycoctonine (4); therefore inuline incorporates
the carbonyl ester functional group of MLA (1), yet it is

Notes
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 24 4863
Ta ble 1. Affinity of a Series of Norditerpenoid Alkaloids at Rat Neuronal R7-Type nAChR
compound
R1
R2
R3
R4
R5
R6
R7
R8
IC50 (M)
methyllycaconitine (1)
nudicauline (3)
lycoctonine (4)
inuline (6)
lappaconitine (7)
N-desacetyllappaconitine (8)
delsoline (10)
elatine (2)
delcorine (11)
deltaline (12)
condelphine (13)
karacoline (14)
MeO
MeO
MeO
MeO
MeO
MeO
HO
MeO
MeO
MeO
HO
2-(methylsuccinimido)benzoate
2-(methylsuccinimido)benzoate
HO
2-aminobenzoate
(see Figure 1)
(see Figure 1)
MeO
2-(methylsuccinimido)benzoate
MeO
H
MeO
H
MeO
MeO
MeO
MeO
H
OH
OH
OH
OH
H
H
OH
OCH2O
OCH2O
OCH2O
OH
OH
OH
OH
OH
OH
OH
H
H
H
H
OH
OH
H
H
H
H
H
H
H
H
H
H
H
OMe
OAc
OMe
OMe
OMe
OMe
OMe
OMe
OMe
OMe
OAc
OH
7.6 ( 3.4 × 10-9
1.7 ( 0.7 × 10-9
1.0 ( 0.1 × 10-5
1.6 ( 0.6 × 10-6
9.6 ( 5.5 × 10-5
6.8 ( 0.9 × 10-6
1.9 ( 0.5 × 10-5
6.1 ( 1.5 × 10-9
5.3 ( 1.4 × 10-5
1.1 ( 0.4 × 10-4
1.6 ( 0.3 × 10-6
1.8 ( 0.5 × 10-6
H
MeO
MeO
HO
AcO
H
H
HO
H
H
H
H
H
H
OH
OH
HO
H
H
ester carbonyl oxygen atom of MLA (1) is crucial in the
Beers and Reich pharmacophore,24 and this distance will
be similar in lappaconitine (7) and N-desacetyllappa-
conitine (8) but different from that found in MLA (1).
Both 7 and 8 are weak inhibitors of [125I]-RBgTX binding
(IC50 ) 96 and 6.8 µM, respectively). This may reflect
alterations in the distance geometry (vide supra), the
absence of the 2-(S)-methylsuccinimido ring, or other
structural changes (see Figure 1) within these ligands.
Delsemine (9) is similar in structure to MLA (1)
except that the 2-(S)-methylsuccinimido ring has been
opened with ammonia to produce the corresponding bis-
amides. Ammonia can react with either carbonyl of the
2-(S)-methylsuccinimido ring to give bis-amides with the
methyl group either R or â to the primary amide.26
Delsemine (9), as a mixture of these regioisomers, has
been reported to be over 2 orders of magnitude less
active (IC50 ) 0.36 µM) than MLA (1) at rat brain [125I]-
RBgTX binding sites,27 a result which provides further
evidence that this succinimido ring is significant for high
potency.
F igu r e 2. Binding curves for inhibition of [125I]-RBgTX
binding to rat brain membranes by representative norditer-
penoids. Brain membranes were incubated with [125I]-RBgTX
in the presence and absence of excess unlabeled RBgTX (to
define nonspecific binding) or serial dilutions of norditerpe-
noids, to determine their ability to compete for specific ligand
binding sites: (4) nudicauline (3), (9) MLA (1), (O) elatine (2),
(b) condelphine (13), (0) N-desacetyllappaconitine (8), (2)
delcorine (11). Values are the means from at least three
independent assays, with SEM indicated by the vertical bars.
Data points have been fitted to the nonlinear Hill equation to
generate the curves shown; IC50 values derived from the curves
are given in Table 1.
The methylsuccinimido ring, which is found in MLA
(1) and other related potent ligands, may help to
maintain the appropriate distance geometry between
the tertiary nitrogen atom of the norditerpenoid and the
carbonyl oxygen of the ester bond.24,28 1H and 13C NMR
spectroscopic studies with MLA (1), elatine (2), and
related small molecule analogues containing the 2-
(methylsuccinimido)benzoyl moiety (on a neopentyl
alcohol) display broad lines for the methyl resonances,
indicating that the 5-membered heterocycle may un-
dergo slow rotation on the NMR time scale.23 Inuline
(6) (anthranoyllycoctonine) satisfies the pharmacophore
requirements of the electronegative oxygen atom and
the tertiary nitrogen atom,24,28 but it is possible that
the anthranoyl aromatic ring is no longer in the correct
orientation for high-affinity binding. N-Acylated ana-
logues of inuline (6), e.g., delsemine (9), should possess
similar electron bond distributions to those found in
MLA (1), but the observed potency of this alkaloid 9
(vide supra) is significantly reduced when compared to
that of MLA (1). Furthermore, the modest activity of
inuline (6) (IC50 ) 1.6 µM, Table 1) cannot be attributed
even to trace impurities of MLA (1), as it was synthe-
sized from a sample of lycoctonine (4) itself prepared
at least 2 orders of magnitude less active than MLA at
inhibiting [125I]-RBgTX binding. This result indicates
that the anthranoyl ester alone is insufficient for
optimum potency, and we propose that the 2-(S)-
methylsuccinimido moiety of MLA is significant for its
activity. Recent toxicity studies (LD50 values) in mice
have revealed that the removal of this moiety from MLA
(1) reduces the LD50 by more than 2 orders of magni-
tude.25 This work highlighted the importance of an
amide or succinimide functionality at the anthranilic
acid aniline functional group to enhance mammalian
toxicity and concluded that anthranoyl lycoctonine
(inuline, 6) was a pivotal toxic alkaloid in the transition
between low and high toxicity.25
Lappaconitine (7) and N-desacetyllappaconitine (8)
are both norditerpenoid alkaloids which possess an
anthranoyl moiety directly esterified, via a tertiary
alcohol, to the A ring of the norditerpenoid core struc-
ture (at C4). MLA (1) and other potent norditerpenoid
alkaloids studied in this series differ at this point by
having an extra carbon atom (C18) between the A ring
and the alkyl oxygen of the anthranoyl ester bond. The
distance between the tertiary nitrogen atom and the

4864 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 24
Notes
and efficiently purified from MLA (1). The same source
of lycoctonine (4) had little activity (IC50 ) 10 µM, Table
1).
Furthermore, when semisynthetic 3-deoxyaconitine (15)
was 18-O-demethylated and regioselectively esterified
at the unmasked C18 neopentyl alcohol with the 2-((S)-
methylsuccinimido)benzoyl moiety, the binding affinity
at rat [125I]-RBgTX binding sites was equal to that
recorded for MLA (1).18 This is despite the fact that
3-deoxyaconitine (15) is significantly different from
lycoctonine (4) within the norditerpenoid core, e.g., the
aconitine-type alkaloid 15 contains 6R-methoxy, 7â-
hydrido, 8â-acetate, 13â-hydroxy, 14R-benzoate, and
15R-hydroxy substitutents (see Figure 1) compared to
a lycoctonine-type alkaloid. Furthermore, aconitine (5)
and MLA (1) display markedly different pharmacology.
However, without the 2-(S)-methylsuccinimido group
and with the anthranoyl ester functional group intact
(cf. inuline, 6), a decrease in the binding affinity of 2
orders of magnitude was observed.18 This is analogous
to the difference in activity we have observed between
MLA (1) (which contains the succinimido group) and
inuline (6).
From our detailed examination of the relative poten-
cies of these alkaloids, the importance of substituents
on the norditerpenoid core can also be appraised. In
particular, elatine (2) was equipotent (Table 1) with
MLA (1), despite the replacement of C7 and C8 hydroxyl
groups with a methylenedioxy bridge (see Figure 1). We
therefore conclude that the C7 and C8 hydroxyl groups
are not essential for the donation of a hydrogen atom
to form a hydrogen bond at the receptor binding site.
They are the only two (tertiary) hydroxyl groups present
in MLA (1), although they cannot be discounted as the
source of lone pairs of electrons to form a hydrogen bond.
Nudicauline (3), in which the C14 O-methyl ether of
MLA (1) has been replaced with an acetate functional
group, was marginally more potent than MLA at [125I]-
RBgTX binding sites (Table 1, Figure 2). Kukel and
J ennings have recently shown that within a series of
five 2-(methylsuccinimido)benzoyl norditerpenoid alka-
loids tested at rat neuronal [125I]-RBgTX binding sites,
all were within 1 order of magnitude in binding affinity
when compared to MLA (1).27 Hence, glaudelsine,
which has the lowest binding affinity in their series and
differs from MLA (1) in having a â-hydroxy functional
group at C6 (i.e., not a â-methoxy functional group),
inhibited [125I]-RBgTX binding in rat brain membranes
with an IC50 of 16 nM compared to MLA (1) with an
IC50 of 1.7 nM. Unfortunately, the key chemical struc-
tures of these norditerpenoid alkaloids are incorrect in
their paper and also in the corresponding erratum,
where active compounds are shown with incorrect
stereochemistry.27
The R3- and R4-containing nAChR are several orders
of magnitude less sensitive to MLA than R7-containing
nAChR.29 All compounds were assessed for activity at
[3H]nicotine binding sites which are considered to
represent nAChR comprised of R4 and â2 subunits.30
Only those compounds with nanomolar affinities at
[
125I]-RBgTX binding sites showed any appreciable
activity at [3H]nicotine binding sites. IC50 values for
MLA (1), nudicauline (3), and elatine (2) were 44, 9.4,
and 8.3 µM, respectively. Thus, these norditerpenoids
are at least 1000-fold more active at R7-containing
nAChR. A preference for [125I]-RBgTX binding sites is
retained by all the other compounds which displayed
micromolar affinities for this (R7) nAChR but were
unable to displace [3H]nicotine binding even at the
highest concentration tested (100 µM).
Other norditerpenoid alkaloids tested in our study,
none of which contains the 2-((S)-methylsuccinimido)-
benzoyl functionality, were only weakly active at rat
neuronal [125I]-RBgTX binding sites as shown in Table
1. Hence, we propose that structural modifications to
selective functional groups within the norditerpenoid
alkaloid can be tolerated without loss of significant
activity or selectivity compared to altering the 2-((S)-
methylsuccinimido)benzoyl moiety. The C18 O-methyl
ethers delsoline (10) and delcorine (11) were of low
activity. The least potent compound in our study was
deltaline (12) (Table 1) with an IC50 of 110 µM. Delta-
line lacks an oxygen atom at C18 (Figure 1), and it is
also the only compound examined which contains an
acetate group at C6 on the â-face. Deltaline (12)
contains an additional tertiary alcohol at C10 (R7 ) HO,
Table 1). However, in other respects, e.g., the presence
of the C7/C8 methylenedioxy bridge, deltaline is similar
in its norditerpenoid core substitution pattern to elatine
(2), one of the most potent ligands in our study.
Delcorine (11) also contains this methylenedioxy bridge,
yet it is 4 orders of magnitude less active than elatine
(2) (see Table 1). This difference in activity is analogous
to that observed between lycoctonine (4) and MLA (1).
Con clu sion s
These results demonstrate unequivocally the impor-
tance of the methylsuccinimido ring for potency of
lycoctonine-type norditerpenoid alkaloids at mammalian
neuronal nAChR labeled with [125I]-RBgTX. There is
now considerable scope for the synthesis of alternative
substituted succinimide or maleimide analogues of MLA
(1) which can be used to explore the domain surrounding
this region of the ligand at the receptor (agonist) binding
site. These results also demonstrate that hydrogen
atom donation from C7 and/or C8 hydroxyl functional
groups, in forming a putative hydrogen bond, is not an
absolute requirement for the binding of MLA (1) to R7
nAChR. There appears to be a tolerance of certain
functional group alterations/deletions within the MLA
norditerpenoid core which do not affect binding affinity.
These results have implications for the design and
synthesis of less structurally complex, subtype-selective
nAChR ligands.31,32
Exp er im en ta l Section
NMR spectra were recorded using J eol GX 270 (operating
at 270 MHz for 1H) or J eol EX 400 (operating at 400 MHz for
1H or 100 MHz for 13C) spectrometers. Chemical shift values
are on the δ scale (parts per million) and referenced to either
The results presented here for lycoctonine-type alka-
loids are also in agreement with our studies with
aconitine-type norditerpenoids which we have synthe-
sized substituted with anthranoyl and 2-((S)-methyl-
succinimido)benzoyl moieties respectively.18 Thus, con-
delphine (13) and karacoline (14), which are substituted
with hydrogen at C7 (R4 in Table 1) and are therefore
typical Aconitum alkaloids, displayed low activity.
1
tetramethylsilane, δ 0.00 in CDCl3 for H NMR, or deuteriated
chloroform, at δ 77.00 for 13C NMR. NMR data are reported
sequentially as integral, multiplicity, assignment, and coupling
constant (J ). The 1H and 13C NMR assignments for all
norditerpenoid alkaloids were based on detailed comparisons
1
with H/1H homonuclear, 1H/13C heteronuclear, and 13C DEPT

Notes
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 24 4865
spectroscopic studies of MLA and elatine and are in agreement
were concentrated to afford pure inuline (6) (5 mg, 7%), while
a further quantity of impure inuline (6) was also recovered
(15 mg): 1H NMR (CDCl3) δ 7.81 (1H, dd, aromatic H6, 8.5,
1.5 Hz), 7.29 (1H, td, aromatic H4, 7.0, 1.8 Hz), 6.68-6.64 (2H,
m, aromatic H3, H5), 5.78 (2H, br s, NH2), 4.13 (1H, d, H18a,
11.6 Hz (J AB)), 4.12 (1H, d, H18b, 11.6 Hz (J AB)), 4.01 (1H, s,
H8 (OH)), 3.91 (1H, s, H6), 3.61 (1H, t, H14, 4.6 Hz), 3.42 (3H,
s, OMe), 3.37 (3H, s, OMe), 3.35 (3H, s, OMe), 3.26 (3H, s,
OMe), 3.23 (1H, dd, H16, 8.8, 7.0 Hz), 3.08 (1H, dd, H9, 6.7,
4.6 Hz), 3.00 (1H, dd, H1, 10.1, 7.3 Hz), 2.98-2.88 (2H, m,
H17, Ha of NCH2), 2.86-2.78 (1H, m, Hb of NCH2), 2.73 (1H,
d, H19a, 11.6 Hz), 2.61 (1H, dd, H15, 15.3, 8.8 Hz), 2.50-2.42
(2H, m, H12a, H19b), 2.34 (1H, dd, H13, 7.0, 4.6 Hz), 2.25-
2.12 (1H, m, H2), 2.12-2.04 (1H, m, H2), 2.00-1.94 (2H, m,
H10), 1.89-1.80 (1H, m, H12b), 1.80-1.72 (2H, m, H5, H3a),
1.67 (1H, dd, H15, 15.3, 7.0 Hz), 1.66 (1H, s, H7 (OH)), 1.63-
1.53 (1H, m, H3b), 1.07 (3H, t, NCH2CH3, 7.0 Hz); 13C NMR
(CDCl3) δ 167.72 (CO, ester), 150.73 (aromatic C2), 134.23 and
130.63 (aromatic C4, C6), 116.77 and 116.15 (aromatic C3, C5),
110.20 (aromatic C1), 90.82 (C6), 88.42 (C7), 83.96 and 83.86
(C1, C14), 82.53 (C16), 77.45 (C8), 68.53 (C18), 64.45 (C17),
57.90, 57.72, 56.18, and 55.69 (C1, C6, C14, C16, 4 × OMe),
52.40 (C19), 50.92 (N-CH2), 50.27 (C5), 48.99 (C11), 46.04
(C10), 43.17 (C9), 38.13 (C13), 37.49 (C4), 33.52 (C15), 32.16
(C3), 28.64 (C12), 26.08 (C2), 14.09 (NCH2CH3).
Ela tin e (2). Purified MLA (1) (270 mg, 0.39 mmol) was
suspended in formaldehyde diethylacetal (20 mL) containing
p-toluenesulfonic acid monohydrate (0.5 g, 2.6 mmol). The
resulting white suspension was then heated to 80 °C where-
upon it cleared. A gummy precipitate appeared at this stage,
so DMSO (3 mL) was added to aid solubilization. The reaction
mixture was heated at 80 °C for 44 h and then cooled to 25
°C, and benzene (8 mL) was added. The orange-colored
solution was heated for a further 2 h, with azeotropic distil-
lation using a Dean-Stark trap, before being cooled to 25 °C.
The solvents were then removed in vacuo to furnish an orange
oil which was dissolved in dichloromethane (20 mL) and then
washed with saturated aqueous sodium hydrogen carbonate
solution (15 mL). The dichloromethane solution was dried
(Na2SO4), filtered, and concentrated in vacuo to give an orange
oil (301 mg). Purification by flash chromatography (silica gel,
5% methanol-dichloromethane) gave pure MLA, 83 mg, and
a mixture of MLA and elatine (2), 106 mg. This mixture was
purified further by HPLC using the following conditions:
column, C8-IK5 Inertpak (Capital HPLC, 25 cm × 10 mm i.d.);
eluent, 64% aqueous ammonium acetate (0.05 M buffered to
pH 5.0 with glacial acetic acid), 36% acetonitrile; flow rate, 4
mL/min; detection at 270 nm. Combined collected fractions
were adjusted to pH 8 with a few drops of saturated aqueous
sodium hydrogen carbonate solution before removing the
acetonitrile in vacuo. The remaining aqueous component was
then lyophilized (16 h) to afford a white powder. This was
suspended in distilled water (25 mL) and washed with dichlo-
romethane (4 × 20 mL). The combined organic fractions were
dried (Na2SO4), filtered, and concentrated in vacuo to afford
the title compound (28 mg, 10%), homogeneous by TLC and
free of (<0.1%) detectable MLA by HPLC:19 1H NMR (CDCl3)
inter alia δ 8.05 (1H, d, aromatic H6, 7.3 Hz), 7.67 (1H, td,
aromatic H4, 7.7, 1.6 Hz), 7.53 (1H, td, aromatic H5, 7.9, 1.3
Hz), 7.27 (1H, dd, aromatic H3, 7.7, 1.1 Hz), 5.07 (2H, s,
OCH2O), 4.13-4.00 (2H, m, H18), 3.43 (3H, s, OMe), 3.35 (3H,
s, OMe), 3.33 (3H, s, OMe), 3.26 (3H, s, OMe), 3.30-3.22 (1H,
m, H16), 2.80-2.90 (1H, m, NCH2a), 2.76 (1H, d, H19a, 11.6
Hz), 2.73-2.63 (1H, m, NCH2b), 2.60 (1H, dd, H12a, 13.9, 3.8
Hz), 2.44 (1H, dd, H15a, 14.8, 8.8 Hz), 2.40-2.30 (2H, m, H19b,
H13), 1.86 (1H, dd, H15b, 14.8, 7.7 Hz), 1.75-1.68 (1H, m,
H12b), 1.06 (3H, t, NCH2CH3 7.1) (succinimido proton reso-
nances at δ 3.05 (CH and CH2), 2.55 (CH2), and 1.50 (CH3)
were assigned using 1H/1H correlation spectroscopy); 13C NMR
(CDCl3) δ 179.87 (CO N), 175.92 (CO N), 164.09 (CO O), 133.55
(C4 aromatic), 132.96 (C2 aromatic), 131.21 (C6 aromatic),
129.94 (C3 aromatic), 129.38 (C5 aromatic), 127.11 (C1 aro-
matic), 93.53 (OCH2O), 92.14 (C7), 89.27 (C6), 83.36 and 83.33
(C1, C8), 81.55 (C16)*, 81.24 (C14)*, 69.76 (C18), 64.14 (C17),
58.93 (OMe), 57.80 (OMe), 56.15 (OMe), 55.20 (OMe), 53.41
(C5), 52.79 (C19), 50.48 (NCH2CH3), 49.95 (C11), 48.42 (C9)*,
39.86 (C10)*, 38.59 (C13), 37.15 (C4), 36.99 (C2′′ succinimide)*,
1
with recently reported H and 13C NMR spectroscopic data for
these alkaloids.21,22 Thin layer chromatography (TLC) was
performed using aluminum-backed TLC plates coated with
Kieselgel 60 F254 purchased from Merck. Silica gel 60 (35-75
µm) was purchased from Prolabo. Solvent A is dichlo-
romethane-methanol-concentrated aqueous ammonia, 100:
10:1, v/v/v. High-resolution mass spectra (FAB, mNBA matrix)
were recorded with a VG Analytical Autospec mass spectrom-
eter (by the EPSRC MS service at Swansea), and HPLC was
performed with a J asco PU-980 pump equipped with a J asco
UV-975 detector. Melting points were determined using a
Kofler hot-stage apparatus (Cambridge Instruments) and are
uncorrected.
MLA was isolated and purified from Garden Hybrid Del-
phinium as previously described and was >99.9% pure by
HPLC (using UV detection at 270 nm, with lappaconitine as
the internal standard).1,2,19 Furthermore, our isolated MLA
cochromatographs with and displays identical biological activ-
ity with an authentic sample of MLA, kindly provided for us
by Prof. M. Benn (University of Calgary, Canada). Unless
otherwise stated, all solvents were purchased from Fisons.
Acetonitrile, chloroform, dichloromethane, and methanol were
all HPLC grade. Absolute ethanol was purchased from Hay-
mans, and anhydrous DMF and DMSO were purchased from
Aldrich. Isatoic anhydride was recrystallized (absolute etha-
nol) to give a brown powder, mp 210-213 °C dec (lit.33 mp 240
°C dec, lit.34 mp 243 °C dec).
Lycocton in e (4). MLA (870 mg, 1.27 mmol) was dissolved
in absolute ethanol (20 mL) and aqueous potassium hydroxide
solution (2 M, 2 mL). The reaction mixture was stirred at 20
°C for 13 h, after which time TLC (solvent A, detection by
dipping in ninhydrin dissolved in n-butyl alcohol followed by
heating) showed that MLA had been hydrolyzed to a more
polar alkaloid. The pH of the ethanolic solution was adjusted
to 7 (paper) with aqueous hydrochloric acid solution (1 M)
before concentration in vacuo and lyophilization. The residue
was then suspended in dichloromethane, inorganic salts were
removed by filtration, and the filtrate was concentrated in
vacuo. The residue was then purified by flash chromatography
(solvent A) to give pure lycoctonine 4 (458 mg, 77%), homo-
geneous by TLC and >99.9% free from MLA by HPLC:19 1H
NMR (CDCl3) δ 4.15 (1H, s, H8 (OH)), 3.85 (1H, s, H6), 3.62-
3.58 (2H, m, H14, H18 (OH)), 3.45 (3H, s, OMe), 3.44 (1H, d,
H18a, 14.3 Hz), 3.41 (3H, s, OMe), 3.37 (1H, d, H18b, 14.3 Hz),
3.33 (3H, s, OMe), 3.25 (3H, s, OMe), 3.21 (1H, t, H16, 7.6
Hz), 3.06 (1H, t, H9, 5.2 Hz), 3.00-2.72 (4H, m, H1, H17,
NCH2), 2.66-2.50 (2H, m, H19a, H15a ), 2.41 (1H, dd, H12a,
13.0, 2.2 Hz), 2.33 (1H, dd, H13, 6.0, 4.8 Hz), 2.27 (1H, d, H19b,
12.0 Hz), 2.20-2.00 (2H, m, H2), 1.96-1.76 (2H, m, H10,
H12b), 1.72-1.45 (4H, m, H5, H15b, H3), 1.04 (3H, t, NCH2CH3,
7.1 Hz) (for lycoctonine, H7 OH was not observed,22 in CDCl3);
13C NMR (CDCl3) δ 90.40 (C6), 88.21 (C7), 84.15 and 83.78
(C1, C14), 82.50 (C16), 77.37 (C8), 67.38 (C18), 64.74 (C17),
57.73, 57.62, 56.09, and 55.62 (C1, C6, C14, C16, 4 × OMe),
52.58 (C19), 50.99 (NCH2), 49.45 (C5), 48.70 (C11), 45.92 (C10),
43.09 (C9), 38.37 (C4), 37.82 (C13), 33.42 (C15), 31.44 (C3),
28.57 (C12), 25.94 (C2), 13.99 (NCH2CH3); HRMS (FAB,
mNBA matrix) lycoctonine (C25H41NO7) found 468.2956 (MH+
calcd 468.2961).
In u lin e (An th r a n oyllycocton in e) (6). Lycoctonine (54
mg, 0.11 mmol) was dissolved in anhydrous DMF (1 mL), and
to this were added isatoic anhydride (38 mg, 0.23 mmol) and
4-(N,N-dimethylamino)pyridine as catalyst (6 mg, 0.4 equiv).
The reaction mixture was then heated to 70 °C and stirred at
this temperature for 17 h. TLC (solvent A) indicated that
starting material was still present, so a further aliquot of
isatoic anhydride (10 mg, 61 µmol) was added. After another
7 h, only a trace of starting material remained (TLC analysis,
solvent A), so the reaction mixture was cooled to 20 °C and
the DMF was removed in vacuo using a Kugelrohr distillation
apparatus. Dichloromethane (5 mL) was added to the residue,
and the resultant brown precipitate was filtered off. The
residual brown oil (54 mg) was purified by flash chromatog-
raphy on silica gel, eluting first with 2% methanol in dichlo-
romethane followed by gradually increasing methanol concen-
trations to 5% in dichloromethane. Homogeneous fractions

4866 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 24
Notes
(10) Dale, H. H. The action of certain esters and ethers of choline,
and their relation to muscarine. J . Physiol. (London) 1914, 6,
147-190.
(11) Heinemann, T.; Changeux, J . P. Structural and functional
properties of the acetylcholine receptor in its purified and
membrane-bound states. Annu. Rev. Biochem. 1978, 47, 317-
357.
(12) Clarke, P. B. S. The fall and rise of neuronal R-bungarotoxin
binding proteins. Trends Pharmacol. Sci. 1992, 13, 407-413.
(13) Pelletier, S. W.; J oshi, B. S. Carbon-13 and proton NMR shift
assignments and physical constants of norditerpenoid alkaloids.
In Alkaloids: Chemical and Biological Perspectives; Pelletier,
S. W., Ed.; Springer Verlag: New York, 1991; Vol. 7, pp 297-
564.
(14) Yunusov, M. S. Diterpenoid alkaloids. Nat. Prod. Rep. 1991, 8,
499-526.
(15) Yunusov, M. S. Diterpenoid alkaloids. Nat. Prod. Rep. 1993, 10,
471-486.
(16) Benn, M. H.; J acyno, J . M. The toxicology and pharmacology of
diterpenoid alkaloids. In Alkaloids: Chemical and Biological
Perspectives; Pelletier, S. W., Ed.; J ohn Wiley and Sons: New
York, 1983; Vol. 1, pp 153-210.
(17) J acyno, J . M.; Harwood, J . S.; Lin, N.-H.; Campbell, J . E.;
Sullivan, J . P.; Holladay, M. W. Lycaconitine revisited: partial
synthesis and neuronal nicotinic acetylcholine receptor affinities.
J . Nat. Prod. 1996, 59, 707-709.
35.39 and 35.19 (C3′′ succinimide)*, 34.83 (C15), 31.70 (C3),
27.80 (C2)*, 26.36 (C12)*, 16.56 and 16.28 (C5′′ succinimide),
13.91 (NCH2CH3) (* denotes that these assignments may have
to be interchanged). The HPLC sample (>99.9% pure) dis-
played appropriate HRMS data: 695.3603 (MH+, 100), also
694.3476 (7); calcd C38H51O10N2 requires 695.3544 and
C38H50O10N2 requires 694.3465.
All synthesized alkaloids gave satisfactory 1H and 13C NMR
spectral data consistent with the proposed structures.13,22 All
other alkaloids were from Latoxan (Rosans, France).
Nicotin ic Bin d in g Assa ys. Competition binding assays
for [125I]-RBgTX or [3H]nicotine binding sites in rat brain
membranes were carried out as previously described.35 In
brief, rat brain P2 membranes (1-2 mg of protein/mL) were
incubated with serial dilutions of competing ligand and 1 nM
[
125I]-RBgTX or 10 nM [3H]nicotine for 3 h at 37 °C or for 1 h
at 20 °C, respectively. Bound radioligand was separated by
centrifugation in the case of [125I]-RBgTX or filtration in the
case of [3H]nicotine. Total and nonspecific binding were
determined in the absence and presence, respectively, of 1 µM
RBgTX or 10-4 M nicotine. Potential ligands (as their free
bases, except for nudicauline perchlorate and lappaconitine
hydrobromide) were dissolved in ethanol to a stock concentra-
tion of 10-3 M and stored at 4 °C. Concentrations that inhibit
radioligand binding by 50% (IC50 value) were determined by
fitting the data to the Hill equation.
(18) Hardick, D. J .; Cooper, G.; Scott-Ward, T.; Blagbrough, I. S.;
Potter, B. V. L.; Wonnacott, S. Conversion of the sodium channel
activator aconitine into a potent R7-selective nicotinic ligand.
FEBS Lett. 1995, 365, 79-82.
Ack n ow led gm en t. This study was supported finan-
cially by The Wellcome Trust (Project Grant 036214 and
Grant 045023) and the EU Biotechnology Programme.
B.V.L.P. is a Lister Institute Research Professor. We
are grateful to Latoxan (Rosans, France) for providing
some of the norditerpenoid alkaloids, to Dr. J . A.
Ballantine and the National EPSRC Mass Spectrometry
Service (Swansea, Wales) for obtaining the HRMS data,
and to Professor M. H. Benn (University of Calgary,
Canada) for the generous provision of an authentic
sample of MLA citrate at the commencement of these
studies. We wish to thank Dr. M. G. Rowan (University
of Bath) for helpful discussions and for his interest in
this project.
(19) Coates, P. A.; Blagbrough, I. S.; Lewis, T.; Potter, B. V. L.;
Rowan, M. G. An HPLC assay for the norditerpenoid alkaloid
methyllycaconitine, a potent nicotinic acetylcholine receptor
antagonist. J . Pharm. Biomed. Anal. 1995, 13, 1541-1544.
(20) Rabinovich, M. S. Aconitine alkaloids. VI. Elatine. J . Gen. Chem.
USSR 1954, 24, 2211-2213; Chem. Abstr. 1957, 51, 16503g.
(21) Pelletier, S. W.; Desai, H. K.; Kulanthaivel, P.; J oshi, B. S.
Methylenation of some lycoctonine-type C19-diterpenoid alka-
loids: partial synthesis of delbruline, and elatine. Heterocycles
1987, 26, 2835-2840.
(22) Hanuman, J . B.; Katz, A. 1H-NMR spectra of norditerpenoid
alkaloids a review. J . Nat. Prod. 1994, 57, 1473-1483.
(23) Blagbrough, I. S.; Coates, P. A.; Hardick, D. J .; Lewis, T.; Rowan,
M. G.; Wonnacott, S.; Potter, B. V. L. Acylation of lycoctonine:
semi-synthesis of inuline, delsemine analogues and methyl-
lycaconitine. Tetrahedron Lett. 1994, 35, 8705-8708.
(24) Beers, W. H.; Reich, E. Structure and activity of acetylcholine.
Nature 1970, 228, 917-922.
Refer en ces
(25) Manners, G. D.; Panter, K. E.; Pelletier, S. W. Structure-activity
relationships of norditerpenoid alkaloids occuring in toxic lark-
spur (Delphinium) species. J . Nat. Prod. 1995, 58, 863-869.
(26) Pelletier, S. W.; Bhattacharyya, J . The nature of “delsemine”
from Delphinium semibarbatum and D. tricorne Michaux (Dwarf
Larkspur). Tetrahedron Lett. 1977, 2735-2736.
(27) Kukel, C. F.; J ennings, K. R. Delphinium alkaloids as inhibitors
of R-bungarotoxin binding to rat and insect neural membranes.
Can. J . Physiol. Pharmacol. 1994, 72, 104-107; 1995, 73, 145.
(28) Sheridan, R. P.; Nilakantan, R.; Dixon, J . S.; Venkataraghavan,
R. The ensemble approach to distance geometry: application to
the nicotinic pharmacophore. J . Med. Chem. 1986, 29, 899-906.
(29) Drasdo, A.; Caulfield, M. P.; Bertrand, D.; Bertrand, S.; Won-
nacott, S. Methyllycaconitine: a novel nicotinic antagonist. Mol.
Cell. Neurosci. 1992, 3, 237-243.
(30) Flores, C. M.; Rogers, S. W.; Pabreza, L. A.; Wolfe, B. B.; Kellar,
K. J . A subtype of nicotinic cholinergic receptor in rat brain is
composed of R4 and â2 subunits and is upregulated by chronic
nicotine treatment. Mol. Pharmacol. 1992, 41, 31-37.
(31) Sargent, P. B. The diversity of neuronal nicotinic acetylcholine
receptors. Annu. Rev. Neurosci. 1993, 16, 403-443.
(32) McDonald, I. A.; Cosford, N.; Vernier, J .-M. Nicotinic acetylcho-
line receptors: molecular biology, chemistry and pharmacology.
Annu. Rep. Med. Chem. 1995, 30, 41-50.
(1) Goodson, J . A. The alkaloids of the seeds of Delphinium elatum.
J . Chem. Soc. 1943, 139-141.
(2) Coates, P. A.; Blagbrough, I. S.; Hardick, D. J .; Rowan, M. G.;
Wonnacott, S.; Potter, B. V. L. Rapid and efficient isolation of
the nicotinic receptor antagonist methyllycaconitine from
Delphinium: assignment of the methylsuccinimide absolute
stereochemistry as S. Tetrahedron Lett. 1994, 35, 8701-8704.
(3) Wonnacott, S.; Albuquerque, E. X.; Bertrand, D. Methylly-
caconitine: a new probe that discriminates between nicotinic
acetylcholine receptor subclasses. In Methods in Neurosciences;
Conn, M., Ed.; Academic Press: New York, 1993; Vol. 12, pp
263-275.
(4) Barrantes, G. E.; Rogers, A. T.; Lindstrom, J .; Wonnacott, S.
R-Bungarotoxin binding sites in rat hippocampal and cortical
cultures: initial characterisation, colocalisation with R7 subunits
and up-regulation by chronic nicotine treatment. Brain Res.
1995, 672, 228-236.
(5) Alkondon, M.; Pereira, E. F. R.; Wonnacott, S.; Albuquerque, E.
X. Blockade of nicotinic currents in hippocampal neurons defines
methyllycaconitine as a potent and specific receptor antagonist.
Mol. Pharmacol. 1992, 41, 802-808.
(6) Ward, J . M.; Cockcroft, V. B.; Lunt, G. G.; Smillie, F. S.;
Wonnacott, S. Methyllycaconitine: a selective probe for neuronal
R-bungarotoxin binding sites. FEBS Lett. 1990, 270, 45-48.
(7) Palma, E.; Bertrand, S.; Binzoni, T.; Bertrand, D. Neuronal
nicotinic R7 receptor expressed in Xenopus oocytes presents five
putative binding sites for methyllycaconitine. J . Physiol. 1996,
491.1, 151-161.
(8) McGehee, D. S.; Role, L. W. Physiological diversity of nicotinic
acetylcholine receptors expressed by vertebrate neurons. Annu.
Rev. Physiol. 1995, 57, 521-546.
(33) Dictionary of Organic Compounds, 5th ed.; Chapman and Hall:
New York, 1982; Vol. 3, p 3370.
(34) CRC Handbook of Chemistry and Physics, 71st ed.; Lide, D. R.,
Ed.; CRC Press: Boca Raton, FL, 1990; p 3-295.
(35) Garcha, H. S.; Thomas, P.; Spivak, C. E.; Wonnacott, S.;
Stolerman, I. P. Behavioural and ligand-binding studies in rats
(9) Elgoyhen, A. B.; J ohnson, D. S.; Boulter, J .; Vetter, D. E.;
Heinemann, S. R9: an acetylcholine receptor with novel phar-
macological properties expressed in rat cochlear hair cells. Cell
1994, 79, 705-715.
with 1-acetyl-4-methylpiperazine, a novel nicotinic agonist.
Psychopharmacology 1993, 110, 347-354.
J M9604991