Ligands for the common allosteric site of acetylcholine M2-receptors: Development and application

Article abstract of DOI:10.1016/S0031-6865(99)00028-X

Ligands for the allosteric site of acetylcholine M₂ receptors are able to retard the dissociation of simultaneously bound ligands for the orthosteric site. This effect promotes receptor occupation by the orthosteric ligand. The allosteric effec

Full text of DOI:10.1016/S0031-6865(99)00028-X

Pharmaceutica Acta Helvetiae 74 Ž2000. 149–155
www.elsevier.comrlocaterpharmactahelv
Ligands for the common allosteric site of acetylcholine M -receptors:
2
development and application
U. Holzgrabe a,), W. Bender , H.M. Botero Cid , M. Staudt , R. Pick , C. Pfletschinger ,
a
a
a
a
a
b
b
b
E. Balatkov ´a , C. Tr a¨ nkle , K. Mohr
a
Department of Pharmaceutical Chemistry, Institute of Pharmacy and Food Chemistry, UniÕersity of W u¨ rzburg, Am Hubland, 97074 W u¨ rzburg, Germany
b
Department of Pharmacology and Toxicology, Institute of Pharmacy, UniÕersity of Bonn, An der Immenburg 4, 53121 Bonn, Germany
Abstract
Ligands for the allosteric site of acetylcholine M2 receptors are able to retard the dissociation of simultaneously bound ligands for the
orthosteric site. This effect promotes receptor occupation by the orthosteric ligand. The allosteric effect opens various therapeutic
perspectives, e.g., in organophosphorus poisoning. The aim of our studies was to optimize the affinity of the modulators for the common
allosteric binding site of muscarinic M2 receptors, the orthosteric site of which was liganded with the N-methylscolopamine. The
phthalimido substituted hexane-bisammonium compound W84 served as a starting point. Previous molecular modelling studies revealed
two positive charges and two aromatic imides in a sandwich-like arrangement to be essential for a high allosteric potency. A
three-dimensional quantitative structure activity relationship Ž3D QSAR. analysis predicted compounds with substituents of increasing
size on the lateral imide moieties to enhance the affinity for the allosteric binding site. Thus, we synthesized and pharmacologically
evaluated compounds bearing ‘‘saturated’’ phthalimide moieties as well as phthalimidines with substituents of systematically increasing
size in position 3 or on the aromatic ring at one or both ends of the molecule. Within each series, QSAR could be derived: 1.
‘‘Saturation’’ of the aromatic ring of the phthalimide moiety results in less potent compounds. 2. Increasing the size of the substituents in
position 3 of the phthalimide enhances the potency. 3. Putting substituents on the aromatic part of the phthalimide increases the potency
more effectively: the introduction of a methyl group in position 5 gave a compound with a potency in the nanomolar concentration range
which was subsequently developed as the first radioligand for the allosteric binding site. q 2000 Elsevier Science B.V. All rights
reserved.
Keywords: Allosteric site; Acetylcholine M2 receptor; 3D QSAR analysis; N-methylscopolamine; W84; Organophosphorus poisoning; Allosteric
modulation; Pain; Dementia
1
. Introduction
paradigm we consider the interaction at acetylcholine M2
receptors between structurally heterogeneous allosteric
modulators and orthosteric antagonists, such as N-methyl-
scopolamine ŽNMS., ŽChristopoulos et al., 1998; Holz-
grabe and Mohr, 1998. or agonists ŽGharagozloo et al.,
1999.. The allosteric modulators can influence both the
Whereas the concept of allosteric modulation for en-
zymes ŽKoolman and R o¨ hm, 1996. and for ion channel
receptors is well-established, this phenomenon is hardly
recognized in the case of G-protein coupled receptors. As a
ligand association and dissociation resulting either in a
reduction or in an elevation of ligand equilibrium binding.
In combination with antagonists such as atropine, the
therapy of organophosphorus poisoning can take advantage
)
Corresponding author. Tel.: q49-931-888-5460; fax: q49-931-888-
5
0
494; e-mail: holzgrab@pharmazie.uni-wuerzburg.de
031-6865r00r$ - see front matter q 2000 Elsevier Science B.V. All rights reserved.
PII: S0031-6865Ž99.00028-X

1
50
U. Holzgrabe et al.rPharmaceutica Acta HelÕetiae 74 (2000) 149–155
of the retardation of the dissociation. The allosteric eleva-
tion of endogenous acetylcholine binding ŽBirdsall et al.,
1
999. might be beneficial in the treatment of pain and
dementia.
The goal of the investigations, presented here, was to
optimize the affinity of the modulators for the common
allosteric binding site of muscarinic M2 receptors, the
orthosteric site of which was liganded with the antagonist
N-methylscolopamine. The phthalimido substituted
alkane-bisammonium compound W84 ŽFig. 1. was taken
Fig. 2. Pharmacophore model.
as a starting point.
Previous molecular modelling studies ŽHolzgrabe et al.,
1
996. revealed two positively charged nitrogens and two
aromatic imides in a sandwich-like arrangement to be
essential for a high allosteric potency Žsee Fig. 2.. An
In each project it was tried to built series of systemati-
cally varied compounds which can be used to establish
QSAR analyses.
additionally performed three-dimensional quantitative
structure activity relationship Ž3D QSAR. analysis
ŽHolzgrabe and Hopfinger, 1996. predicted compounds
with substituents of increasing size on the lateral imide
moieties to enhance the affinity for the allosteric binding
site. This pharmacophore hypothesis was supported by the
result of a study in which the length of the spacers
between the pharmacophoric elements was varied ŽNassif-
Makki et al., 1999.. The distance between the positive
2. Common allosteric site
QSAR analyses of series of compounds are only valid
when all compounds bind in the same mode to the same
binding site. Generally, the allosteric action is evaluated as
the quantification of an effect; the EC50,diss value is the
concentration of the modulator at which the rate of radioli-
gand dissociation is reduced by 50% ŽJepsen et al., 1988..
However, experiments to antagonize allosteric effects re-
vealed that structurally closely related allosteric agents
may nevertheless reveal a divergent mode of allosteric
action ŽTr a¨ nkle and Mohr, 1997.. In order to gain a more
˚
charges was determined to be at least 10 A, which is
equivalent to about seven methylene groups; the spacer
between the positive charge and the lateral imide moiety
has to be long enough to allow the phthalimides to prop-
erly reach the aromatic positions. In the case of the W84
compounds a length of three methylene groups is neces-
sary.
Taking the findings of the theoretical studies into ac-
count, three projects were initiated:
Ž1. In order to check whether aromatic imides in lateral
positions are necessary for a high allosteric potency the
direct insight into the binding events, a high affinity
radioligand for the allosteric binding site was developed.
Introduction of a methyl group in position 5 of the phthal-
imide ring ŽTr a¨ nkle et al., 1998. yielded a compound,
designated as dimethyl-W84 Žsee Fig. 1., with particularly
high allosteric potency Žfor comparison with other com-
pounds see Tr a¨ nkle et al., 1996.. Tritiation of dimethyl-
phthalimide moieties were replaced with corresponding
‘‘saturated’’ imides and, additionally, the hexahydrophthal-
imide was stepwise reduced to a succimide ring.
W84 gave the first radioalloster suitable for direct binding
Ž2. Considering the QSAR analysis the allosteric po-
measurements at the allosteric site ŽTr a¨ nkle et al., 1998..
tency should increase with increasing steric size of the
substituent on the phthalimide. Thus, various substituents
were introduced in position 3 and 5 of the phthalimide.
Ž3. Since it is unlikely that the allosteric binding site is
symmetrical the structural variations mentioned above were
performed at one end and at both ends of the symmetrical
parent compound W84.
Competition binding experiments with the prototype mod-
ulators alcuronium, W84 and gallamine went parallel with
3
the inhibition of the w HxNMS dissociation which indicated
3
that the high affinity binding of w Hxdimethyl-W84 reflects
the occupation of the common allosteric site of the M2
receptor.
3
. Replacement of the phthalimide with corresponding
‘‘saturated’’ moieties
In previous studies, the unilateral replacement of the
phthalimide with a morpholino or hydroxy substituents
ŽHolzgrabe et al., 1997. or shortening of the lead com-
pound W84 down to monoquaternary compounds ŽKostenis
Fig. 1. Structural formula of the parent compound W84 and dimethyl-W84.

U. Holzgrabe et al.rPharmaceutica Acta HelÕetiae 74 (2000) 149–155
151
et al., 1994. resulted in a considerable reduction of the
going from the saturated cyclohexane-1,2-dicarbonic acid
imide through the dimethyl- and monomethyl- to the non-
substituted succimide.
A positive correlation between the allosteric potency
and the lipophilicity of the lateral moiety was found
suggesting that with increasing size of the hydrophobic
imides the contact area with the receptor protein is en-
larged.
allosteric potency indicating the importance of the lateral
aromatic rings. However, the question remains whether an
aromatic annelation of the imide is necessary or whether a
hydrophobic moiety is sufficient for the interaction with
the receptor protein. Thus, a cis and trans-annelated cyclo-
hexane and a trans-cyclohexene imide compound 1–3 in
addition to a dimethyl- Ž4., a monomethyl- Ž5. and an
unsubstituted Ž6. succimide derivative were synthesized by
conversion of the corresponding 1,2-dicarbonic acids to the
log
Ž
1rEC50
.
s1.412
Ž
"0.61
.
log Pq4.195
Ž
"0.57
.
X
X
anhydrides, condensation of the N , N -dimethyl-
aminopropanamine with the anhydrides giving the imides
and connecting two imide units with dibromoheptane Žsee
Fig. 3.. The allosteric potencies of these new compounds
2
2
ns6; r s0.91; ss0.33; Fs40.1; Q s0.83;
sPRESS s0.452
1
Ž .
X
were compared to the C r3 -phth compound ŽChoo and
7
2
Mitchelson, 1989. which was previously reported to have a
where n is the number of compounds, r is the square of
the correlation coefficient, s the standard deviation, F the
ratio of explained to unexplained variance, Q is the cross
validated r by using the leave-one-out procedure ŽQ can
adopt values between 1 and zero. and sPRESS is the
standard deviation from the predictive residual sum of
squares. Since the cyclohexene derivative shows a steeper
concentration–effect-curve, the data of this compound were
omitted.
higher potency than the corresponding hexane compound
2
ŽTrankle et al., 1996. ŽEC value of 390 nM vs. 1300
¨
5
0
X
2
2
nM.. In comparison to C r3 -phth all concentration–ef-
7
fect-curves except with the cyclohexene derivative com-
pound 3 were shifted in a parallel fashion to the right
indicating the same mode of action but a lower potency
Žsee Fig. 3.. Interestingly, the decrease in potency resem-
bles the diminution of the size of the lateral moiety in
3
Fig. 3. Structural formula of the ‘‘saturated’’ imides and concentration–effect-curves for reducing the apparent rate constant of w HxNMS dissociation from
M2 receptors in porcine heart membranes.

1
52
U. Holzgrabe et al.rPharmaceutica Acta HelÕetiae 74 (2000) 149–155
within this series of compounds, this stereochemical find-
ing is a hint that the phthalimide binds into a pocket at the
entrance of the receptor which has steric restrictions. Thus,
in the next step, it was aimed to probe the size of the
hypothesized binding pocket.
4
. Symmetrical substituent variations at the phthalim-
ide skeleton
X
Fig. 4. Correlation between the allosteric potency of C r7 -phth and 1, 2,
3
Since the preceding study had revealed a cycloalkane
moiety to be disadvantageous as a lateral system, in the
following the aromatic phthalimide skeleton was kept the
same whilst one carbonyl functional was replaced with
hydrogens Ž7., as well as hydroxy Ž8. and alkoxy groups
Ž9. and with phenyl Ž10. and benzyl Ž11. and benzylidene
rings Ž12. Žsee Fig. 5.. Whereas the hydroxy and alkoxy
derivatives can be easily obtained by reduction of the
carbonyl group with NaBH4 and conversion with the
corresponding alcohol in acidic media, other synthesis
pathways had to be developed for the other compounds.
The synthesis of the phenyl substituted compound 10
started off with the reduction of phenylphthalazone with
ZnrHCl to give the phthalimidine which was alkylated
4
–6 and the lipophilicity.
Even though the correlation is found to be highly
significant, the difference in potency between the trans-
and cis-diastereomers of the cyclohexane compound can-
not be explained by the calculated lipophilicity Žsee Fig.
4
.. Since the trans isomer of the cyclohexane-imide steri-
cally resembles the phthalimide more than the cis isomer,
the higher potency of the trans-isomer can be easily
understood. However, even though no significant correla-
tion of the potency with steric parameters was found
Fig. 5. Structural formula of the symmetrical phthalimido substituted allosteric modulators 7–12.

U. Holzgrabe et al.rPharmaceutica Acta HelÕetiae 74 (2000) 149–155
153
with dibromoalkane in presence of NaH. Two of these
molecules were connected with 1,6-bisŽdimethyl-
amino.hexane to obtain the bisammonium salt. The benzyl
and benzylidene compounds 11 and 12, resp., were
achieved via the benzalphthalide which was converted to
the imidine; this imidine can be either reduced to give the
benzyl product 11 or the obtained ErZ isomers separated
for the preparation of the benzylidene bisammonium salts
1
2 ŽBotero Cid et al., 1999..
The parallel concentration–effect-curves for these com-
pounds exhibit an allosteric potency of the same order of
magnitude as the parent compound W84. Whereas most of
the alkoxy substituted compounds 9 are less potent than
W84, the benzyl and phenyl substituted ones 10 and 11
show a slightly higher potency; interestingly, all benzyli-
dene isomers 12 are more than one order of magnitude
more potent than W84. Again, the correlation between the
potency and the lipophilicity Žp . was checked, but the
regression analysis results in a poor correlation. The ben-
zylidene derivatives turned out to be outliers. Omitting
these compounds gave a significant, but still poor positive
correlation:
Fig. 6. Correlation between the allosteric potency and the volume of the
lateral substituents of the compounds 7–10 and 12.
benzyl substituted compound is one order of magnitude
less potent than the benzylidene compounds although the
volume of both substituents is almost the same. The strik-
ing difference between these compounds is the difference
in rigidity caused by the double bond. In the case of the
benzylidene compounds the smaller loss of entropy on
binding to the receptor protein ŽAndrews, 1993. might be
log
Ž
1
Ž
EC50
.
s0.304
Ž
"0.15
.
pq5.464
Ž
"0.19
.
the reason for the high potency. Remarkably, the stereo-
chemistry of the benzylidene groups attached to the phthal-
imide skeleton, EE, ZZ and EZ, does not influence the
potency. Since the alkane bisammonium compounds were
found to be highly flexible ŽHolzgrabe and Hopfinger,
2
2
ns8; r s0.80; ss0.167; Fs23.4; Q s0.67;
sPRESS s0.21
Ž .
2
In addition, steric parameters, such as volume, surface
and refractivity calculated by means of Hyperchemr
Chemplus, were checked. Using a parabolic model a poor
correlation with the volume was found wherein the benzyl
compound was an outlier. Omitting this compound resulted
in a striking correlation ŽFig. 6.:
1
996., it is tempting to speculate that either the aromatic
ring of the phthalimide or of the benzylidene ring can
serve as the partner for interaction with the binding site.
Molecular modelling investigations supported this hypoth-
esis ŽBotero Cid et al., 1999..
Summarizing this part of the study: since no steric
restriction and no stereoselectivity concerning the sub-
stituent in position 3 could be found the binding pocket is
supposed to be rather large.
2
log
Ž
1rEC50
.
sq0.0000867
Ž
"0.000026 vol
.
y0.0977 "0.031
Ž
.
volq32.76
Ž
"9.26
.
2
2
ns11; r s0.95; ss0.18; Fs80.4; Q s0.91;
sPRESS s0.24
3
Ž .
5
. Asymmetrical substituent variations at the phthalim-
ide skeleton
From this analysis the minimum potency was calculated
to be related with a volumes in a range between 540 and
3
˚
5
75 A , which corresponds to the methoxy group. Groups
Due to the simplicity of the synthesis most of the newly
synthesized allosteric modulators were symmetrical. How-
ever, the binding site of the modulators is unlikely to be
symmetrical. In order to check whether both ends of the
modulators have a different contribution to the potency we
aimed to synthesize asymmetrical compounds. Due to the
results of the 3D QSAR analysis is was predicted that
enlargement of one phthalimide moiety should increase the
potency even with the contralateral phthalimide kept un-
changed. On the one hand the synthesis started off with the
alkylation of dimethylaminopropylphthalimide with a huge
excess of dibromohexane without any solvent in order to
of smaller and bigger size induced a higher potency. The
increase in potency going along with the diminution of the
substituent size might be caused by a different sort of
interaction between this part of the phthalimide Žcarbonyl
and hydroxy groups versus alkoxy groups. and the recep-
tor protein. The increase in potency connected with the
expansion of the alkane ether chain goes along with the
elevation of the intermolecular hydrophobic contact sur-
face with the receptor protein ŽB o¨ hm and Klebe, 1996.
and, therefore, with the enhancement of the affinity of the
compounds to the receptor binding site. Interestingly, the

1
54
U. Holzgrabe et al.rPharmaceutica Acta HelÕetiae 74 (2000) 149–155
avoid the bisamination of dibromohexane. On the other
hand, a corresponding 1,2-dicarbonic acid was converted
via the anhydride to the dimethylaminopropylimide which
can be connected to the above obtained bromo-compound
to give the asymmetrical compounds 13 ŽFig. 7..
With exception of the ‘‘pyridophthalimide’’ all aro-
matic compounds were found to have a higher potency
than the parent compound W84. Especially the naphthalim-
ide compound shows a very high potency. In order to
quantify the potency, it was checked whether a correlation
between the potency and lipophilic or steric parameters
could be found. Again, the correlation with the lipophilic-
ity was poor in comparison with the volume of the lateral
substitutent ŽBender et al., 1999. ŽFig. 8..
2
log
Ž
1rEC50
.
sy0.000089
Ž
"0.000036 vol
.
Fig. 8. Correlation between the allosteric potency and the volume of the
lateral substituents of the compounds 13.
q0.108 "0.041
Ž
.
voly25.85
Ž
"12.2
.
2
2
ns8; r s0.93; ss0.151; Fs41.6; Q s0.898;
sPRESS s0.199
of corresponding symmetrical and asymmetrical phthalim-
ides will be compared in Section 6.
4
Ž .
In contrast to Eq. Ž3., the allosteric potency is running
through an maximum volume which amounts to 600 to
3
˚
6. Symmetrical versus asymmetrical substituent varia-
tions
6
50 A . These findings clearly indicate that the hypothe-
sised binding area has a defined shape: On the one hand,
there is obviously much space for hydrophobic substituents
at position 3 whereas on the other hand there is pocket
which can perfectly take in and interact with the naphthal-
imide Žor phenothiazine moiety ŽHolzgrabe et al., 1997...
The interactions with smaller and bigger substituents are
less strong causing a lower potency. At this point, the
questions still remains, whether the allosteric modulators
have to be symmetrical or asymmetrical. Therefore, pairs
6
.1. Comparison of the differences in potency upon sym-
metrical and asymmetrical Õariations of lateral moieties
The diminution of one phthalimide to a succimide
resulted in the series of W84 compounds in a loss of
potency of DlogŽ1rEC .s0.88 ŽBender et al., 1999..
50
Replacement of both phthalimides with the succimide in
the corresponding heptane series induced about twice the
loss of potency: DlogŽ1rEC .s2.54. This clearly indi-
5
0
cates that the phthalimide moiety is important in both
places.
6
.2. Stepwise replacement of the phthalimide with ana-
logue imides
Preliminary investigations showed that the replacement
of one phthalimide in W84 with the benzylidene phthalimi-
dine moiety results in an increase in potency in compari-
son to W84 which could be further enhanced by the
replacement of the second phthalimide with benzylidene
phthalimidine moiety ŽHolzgrabe et al., 1997.. Similar
observations were made for the stepwise replacement of
the phthalimide with the above discussed optimal naphthal-
imide substituent. The symmetrical compound appears to
be about 100 times more potent than W84 and seems to be
the most active compound reported till now. It is interest-
ing to find out in the next step whether an asymmetrical
compound composed of an hexane-bisammonium-chain
connecting a benzylidene phthalimidine and a naphthalim-
Fig. 7. Structural formulae of the asymmetrically substituted phthalimide
compounds 13.

U. Holzgrabe et al.rPharmaceutica Acta HelÕetiae 74 (2000) 149–155
155
ide will show a higher or lower potency. Investigations in
this direction are in progress.
Botero Cid, H.M., Pick, R., Holzgrabe, U., Mies-Klomfaß, E., Kostenis,
E., Mohr, K., 1999. Structure-activity relationships in a series of
bisquaternary phthalimidine derivatives modulating the muscarinic
M -receptor allosterically. J. Med. Chem., submitted.
2
Choo, L.K., Mitchelson, F., 1989. Characterization of the antimuscarinic
X
effect of heptane-1,7-bisŽdimethyl-3 -phthalimidopropyl ammonium
bromide. Eur. J. Pharmacol. 162, 429–435.
7
. Conclusions
Christopoulos, A., Lanzafame, A., Mitchelson, F., 1998. Allosteric inter-
actions of muscarinic cholinoceptos. Clin. Exp. Pharmacol. Physiol.
25, 185–194.
Gharagozloo, P., Lazareno, S., Popham, A., Birdsall, N.J., 1999. Al-
losteric interactions of quaternary strychnine and brucine derivatives
with muscarinic acetylcholine receptors. J. Med. Chem. 42, 438–445.
Holzgrabe, U., Hopfinger, A.J., 1996. Conformational analysis, molecular
shape comparison and pharmacophore identification of different al-
losteric modulators of muscarinic receptors. J. Chem. Inf. Comp. Sci.
Taken together, it can be concluded that the 3D QSAR
analysis, which predicted higher affinity to the allosteric
binding site by enlargement of the non-overlap volume of
the modulators, has led to highly potent compounds char-
acterized by a rigid hydrophobic moiety in position 3 of
the phthalimide and a large aromatic area annellated to the
imide. These findings initiate the synthesis of new modula-
tors of higher allosteric potency and may open the perspec-
tive to find compounds suitable for therapeutic purposes.
3
6, 1018–1024.
Holzgrabe, U., Mohr, K., 1998. Allosteric modulation of antagonist
binding to muscarinic receptors. Drug Discovery Today 3, 214–222.
Holzgrabe, U., Wagener, M., Gasteiger, J., 1996. Variation of the oxime
function in bispyridinium-type allosteric modulators of M -cholino-
2
ceptors. J. Mol. Graph. 14, 185–193.
Acknowledgements
Holzgrabe, U., Staudt, M., Bender, W., Mesagna, P., Pfletschinger, C.,
Tr a¨ nkle, C., Mohr, K., 1997. Allosteric modulation of antagonist
binding to the acetylcholine M -receptor: unilateral structural varia-
Thanks are due to the Deutsche Forschungsgemein-
schaft DFG and to the Fonds der Chemischen Industrie,
Deutschland, for financial support, to the KAAD for the
grant given to HMBC, to the EU for Erasmus grants given
to EB, PM and CP, as well as to Irmela Schwesig, Ilona
Knoblauch, Frauke M o¨ rschel and Iris Witten for their
skillful technical assistance and to Hugo Kubinyi ŽBASF,
Ludwigshafen, FRG. for providing the BILIN program.
2
tions of alkane bisammonium-type modulators. Abstract DPhG and
SGPhW Symposium, Z u¨ rich.
Jepsen, K., L u¨ llmann, H., Mohr, K., Pfeffer, J., 1988. Allosteric stabiliza-
3
tion of H-N-methylscopolamine binding in guinea-pig myocardium
by an antidote against organophosphate intoxication. Pharmacol. Tox-
icol. 63, 163–168.
Koolman, J., R o¨ hm, K.-H., 1996. Color Atlas of Biochemistry. Thieme
Verlag, Stuttgart, pp. 107–108.
Kostenis, E., Holzgrabe, U., Mohr, K., 1994. Allosteric effect on M2-
cholinoceptors of derivatives of the alkane-bis-ammonium compound
W84 — comparison with bispyridinium-type allosteric modulators.
Eur. J. Med. Chem. 29, 947–953.
Nassif-Makki, T., Tr a¨ nkle, C., Bejeuhr, G., Cambareri, A., Pfletschinger,
C., Kostenis, E., Mohr, K., Holzgrabe, U., 1999. Bisquaternary
ligands of the common allosteric site of M2 acetylcholine receptors:
optimization of the distances between the pharmacophoric elements.
J. Med. Chem. 42, 849–858.
References
Andrews, P.R., 1993. Drug-receptor interactions. In: Kubinyi, H. ŽEd..,
3
D QSAR in Drug Design — Theory Methods and Applications.
ESCOM, Leiden, pp. 13–40.
Bender, W., Staudt, M., Tr a¨ nkle, C., Mohr, K., Holzgrabe, U., 1999.
Probing the size of a hydrophobic binding pocket within the allosteric
Tr a¨ nkle, C., Mohr, K., 1997. Divergent modes of action among cationic
allosteric modulators of muscarinic M receptors. Mol. Pharmacol.
2
site of muscarinic acetlylcholine M -receptors. Life Science, in press.
51, 674–682.
2
Birdsall, N.J.M., Farries, T., Gharagozloo, P., Kobayashi, S., Lazareno,
S., Sugimoto, M., 1999. Subtype-selective positive cooperative inter-
actions between brucine analogs and acetylcholine at muscarinic
receptors: functional studies. Mol. Pharmacol. 4, 778–787.
B o¨ hm, H.-J., Klebe, G., 1996. Was l a¨ ßt sich aus der molekularen
Erkennung in Protein–Ligand-Komplexen f u¨ r das Drug-Design neuer
Wirkstoffe lernen?. Angew. Chem. Int. Ed. Engl. 35, 2588–2614.
Tr a¨ nkle, C., Kostenis, E., Burgmer, U., Mohr, K., 1996. Search for lead
structures to develop new allosteric modulators of muscarinic recep-
tors. J. Pharmacol. Exp. Ther. 279, 926–933.
Tr a¨ nkle, C., Mies-Klomfaß, E., Botero Cid, M.H., Holzgrabe, U., Mohr,
3
K., 1998. Identification of a w Hxligand for the common allosteric site
of muscarinic acetylcholine M2 receptors. Mol. Pharmacol. 54, 139–
145.