![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
Vol. 62, Issue 4, 778-787, October 2002
Department of Neurochemistry, Faculty of Medicine (H.F., M.K.H., T.Hag.), Department of Biophysics and Biochemistry, Faculty of Science (S.Y.), and Institute of Molecular and Cellular Biosciences, the University of Tokyo, Tokyo, Japan (K.N.); Core Research for Evolutional Science and Technology (H.F., H.H., T.Ham.); Institute for Biomolecular Science, Gakushuin University, Tokyo, Japan (T.Hag.); Genomic Sciences Center, RIKEN Yokohama Institute, Yokohama, Japan (T. Ham., Y.M., H.H., S.Y.); and Suntory Institute for Bioorganic Research, Osaka, Japan (M.I.)
 |
Abstract |
|---|
 Top
 Abstract
 Introduction
Materials and Methods
Results
Discussion
References
|
|---|
Many biogenic amines evoke a variety of physiological responses by acting on G protein-coupled receptors. We have determined the conformation of two acetylcholine analogs, (S)-methacholine and (2S,4R,5S)-muscarine, bound to the M2 muscarinic acetylcholine receptor (M2 mAChR) by NMR spectroscopy. The analysis of the transferred nuclear Overhauser effect indicated that the receptor selectively recognized the conformers of (S)-methacholine and (2S,4R,5S)-muscarine with the gauche O-C2-C1-N dihedral angle at +60°. This is distinct from the predominant conformations of these ligands in solution with O-C2-C1-N dihedral angle (+80~85°) in the absence of the M2 mAChR, as assessed by analyses of the coupling constants and nuclear Overhauser effect spectroscopy. We have also built a molecular model of the M2 mAChR-(S)-methacholine complex, based on the X-ray crystallographic structure of rhodopsin. This model indicated that the conformation with the gauche O-C2-C1-N dihedral angle at +55.5°, which is similar to the one determined by NMR measurement, is energetically favored in the binding of (S)-methacholine to the receptor. We suggest that this conformation represents the binding of the agonist to the M2 mAChR in the absence of G protein.
| |
Introduction |
|---|
|
Top
Abstract
Introduction
Materials and Methods
Results
Discussion
References
|
|---|
The
muscarinic acetylcholine receptors (mAChRs) belong to the family of G
protein-coupled receptors (GPCRs). Upon binding to acetylcholine, mAChR
transmit signals via the activation of G proteins. Five subtypes of
mAChRs (M1-M5) have been
identified and found to be broadly expressed in the central nervous
system and in the peripheral tissues (Kubo et al., 1986
; Bonner et al., 1987, 1988; Peralta et al., 1987). In the central nervous system, the
mAChRs are known to modulate learning and memory and to regulate the
sensory, motor, and autonomic systems. In peripheral tissues, the
mAChRs mediate parasympathetic activities, such as ileum muscle contraction.
GPCRs comprise one of the largest superfamilies. Because many of them
have important pharmacological roles, their characteristics, such as
the recognition of ligands, are now the targets of pharmaceutical interests (Flower, 1999). No detailed structural information has been
available for the GPCRs, except for rhodopsin (Palczewski et al., 2000)
and the ligand binding domain of metabotropic glutamate receptor
(Kunishima et al., 2000). The receptor-bound conformations of pituitary
adenylate cyclase activating polypeptide (Inooka et al., 2001) and
glutamate (Kunishima et al., 2000) have recently become available but
there is no direct information on how biogenic amines, such as
acetylcholine, are recognized by GPCRs, such as mAChRs. To determine
the nature of the ligand-receptor interaction, an understanding of the
conformations of the interacting ligands as well as the architecture of
their binding pocket would be necessary.
The mutagenesis studies of the mAChRs have revealed some key residues
for acetylcholine binding, including Asp 103, and Tyr 104 and Tyr 403, located in the third and sixth transmembrane regions, respectively
(Wess et al., 1991; Page et al., 1995; Vogel et al., 1997; Ward et al.,
1999). The Asp 103 residue is likely to interact with the quaternary
amine group of acetylcholine through charges and Tyr 104 and Tyr 403 may interact with the ester group of acetylcholine through hydrogen
bonds or 
interaction involving the benzene group. Thus, the binding
of acetylcholine and other agonists seems to be governed by the
orientation of their quaternary amine and ester groups, and this
orientation is defined mainly by the dihedral angles of O-C2-C1-N (Fig.
1).
|
Attempts have been made to determine the conformation of the muscarinic
agonists producing the physiological effects by synthesizing and
examining the diastereomers of conformationally rigid acetylcholine analogs (Portoghese, 1970; Casy, 1975). In the majority of the cases,
the trans isomers of the rigid analogs were more potent than
the cis isomers in inducing the muscarinic activities.
Therefore, the pharmacologically active conformation of the muscarinic
agonists has been assumed to have the trans O-C2-C1-N
dihedral angle (~180°) (Taylor and Insel, 1990).
Because of the recent development of the Sf9/baculovirus system and its
culturing techniques, sufficient quantities of many GPCRs (milligram
order) have become available for biophysical studies, such as NMR. The
conformation of small molecules interacting with large molecules such
as proteins can be determined using the TRNOE method (Clore and
Gronenborn, 1982). The intensities of negative NOE signals
(TRNOE), observed because of protein-ligand interaction provide the
spatial information between protons of the bound state of the ligands.
By using this NMR method and an energy calculation, the conformation of
acetylcholine bound to the nicotinic acetylcholine receptor has been
suggested (Behling et al., 1988). However, the receptor-bound
conformation of acetylcholine cannot be determined only by NMR data
because the four methylene protons cannot be distinguished on an NMR
spectrum, due to the chemical equivalency of the two sets of methylene
protons (Fig. 1). The acetylcholine analogs methacholine and muscarine
show an individual signal for each proton due to the difference in the
chemical shifts. The full assignment of the protons enables us to
determine the receptor-bound conformation based solely on the TRNOE intensities.
In this article, we report the conformations of
(S)-methacholine and
(2S,4R,5S)-muscarine bound to the
mAChR and discuss the physiological significance. The preliminary
results have been reported elsewhere (Furukawa et al., 2001).
| |
Materials and Methods |
|---|
|
Top
Abstract
Introduction
Materials and Methods
Results
Discussion
References
|
|---|
Expression and Purification of the M2 mAChR Mutant
(M2 Mutant).
A mutant of mAChR
M2 subtype (M2 mutant) was
constructed from the M2 mAChR cDNA, as in the
previous report (Hayashi and Haga, 1996). The mutation included the
deletion of the central part of the protease-susceptible third
intracellular loop (233-380), the replacement of the putative
glycosylation residues Asn 2, 3, 6, and 9 with Asp to prevent molecular
heterogeneity, and the addition of a hexa-histidine tag and a thrombin
cleavage site at the C terminus for purification. The Sf9 cells, grown
to a density of 3.0 × 106 cells/ml at 28°
in Cell Master (Wakenyaku, Kyoto, Japan), were infected with the
recombinant virus for the M2 mutant at an m.o.i. of 5 to 10 in the presence of 1 µM atropine sulfate. The cells were
harvested after 48 h of infection. The membrane fraction, prepared
as described previously (Hayashi and Haga, 1996), was solubilized with
1% digitonin (Wako Pure Chemicals, Osaka, Japan) and 0.3% sodium
cholate and was loaded onto 3-(2'-aminobenzhydryloxy)tropane (ABT)-agarose affinity chromatography gel (Haga and Haga, 1985). The
eluted fraction was loaded onto a hydroxyl apatite column and was
washed with 12 volumes of a buffer containing 10 mM potassium phosphate
buffer, pH 7.4, 0.1 mM atropine sulfate, and 0.3% sodium cholate
(Dojindo, Kumamoto, Japan). The M2 mutant was
eluted with a buffer containing 1 M potassium phosphate buffer, pH 7.4, 0.1 mM atropine sulfate, and 0.3% sodium cholate. The eluate was
loaded onto a PD10 column (Sigma) equilibrated with a buffer containing 10 mM
Tris(hydroxymethyl-d3)amino-d2-methane-deuterium
chloride (Sigma-Aldrich, St. Louis, MO), pH 7.0, and 0.2% sodium
cholate in D2O. After the addition of 1.5 mM
(S)-methacholine or
(2S,4R,5S)-muscarine to the void
volume fraction, the receptor was concentrated to 50 µM. Crude
soybean phosphatidylcholine (Avanti Polar Lipids, Birmingham, AL) was
added to a concentration of 1.5 mg/ml before NMR measurements.
NMR Measurement.
The NMR measurements were performed on a
Bruker ARX-600 spectrometer (Bruker, Osaka, Japan) at 296°K for the
NOESY and TRNOESY experiments, and on a Bruker ARX-800 spectrometer at
296°K for the one-dimensional spectroscopy experiments. The
assignment of the proton signals for (S)-methacholine was
based on the previous reports (Casy et al., 1971), and that for
(2S,4R,5S)-muscarine was based on
correlation spectroscopy, total correlation spectroscopy, and NOESY spectra.
/2-shifted, squared, sine bell window function in the F1 and F2
dimensions; Fourier transformed; and zero-filled to confer the final
data matrices.
For the NOESY experiments with the free (S)-methacholine and
the (2S,4R,5S)-muscarine, a mixing
time of 1.2 s was used. For the two-dimensional TRNOESY
experiments of the ligands in the presence of the purified
M2 mutant, the mixing time was randomized at ± 10% from 50 and 150 ms for (S)-methacholine and for
(2S,4R,5S)-muscarine, respectively, to
cancel the correlation spectroscopy peaks. Two TRNOESY spectra were
acquired in each experiment: one with the sample in the presence of the
purified M2 mutant (50 µM) and
(S)-methacholine or
(2S,4R,5S)-muscarine (1.5 mM) and the
other with the same sample except for the addition of 1.5 mM atropine.
The latter spectrum was subtracted from the former to acquire the
spectrum representing the specific binding of
(S)-methacholine or
(2S,4R,5S)-muscarine to the
M2 mutant. The intensities of the TRNOE
cross-peaks were roughly linear up to 100 ms, suggesting that the spin
diffusion effect is negligible in this range.
The coupling constants of the ligands in solution were determined using
high-resolution one-dimensional 1H-NMR
measurements with a Bruker ARX-800 spectrometer. The free induction
decay was multiplied by the exponential or Gaussian function, and then
was Fourier-transformed.
[3H]NMS and [35S]GTP
S Binding
Assays.
The binding assays were carried out on virus-infected Sf9
cell membrane fractions expressing the M2 mutant
or the M2 mutant fused to the 
subunit of the
G protein Gi1 at the C terminus (M2-Gi1), as described
in the previous studies (Furukawa and Haga, 2000; Guo et al., 2001).
The [3H]N-methylscopolamine
([3H]NMS) binding (1.5 nM) was displaced by
acetylcholine, carbamylcholine, (S)-methacholine,
(R)-methacholine, or atropine. The agonist-stimulated binding of guanosine
5'-O-(3-[35S]thio)triphosphate
([35S]GTPS) (50 nM) to the
M2-Gi1 was measured in
the presence of 1 µM GDP and various concentrations of the ligands.
The volume of the above reactions was 200 µl. After incubating the
mixtures at 30°C for 1 h, the membrane fraction was trapped on a
GF/B glass fiber filter and the radioactivity bound to the membrane was measured.
The M2 mAChR Modeling.
The model of
photoactivated rhodopsin (metarhodopsin II) was constructed by adopting
the rigid body movement of transmembrane segments 3 through 6 to the
crystal structure of nonactivated rhodopsin (Palczewski et al., 2000),
in accordance with Borhan et al. (1996) and Farrens et al. (2000).
Then, the model of the activated form of the M2
mAChR was built using the model of metarhodopsin II, employing the
"Homology" module. The transmembrane bundles were built by
replacing the amino acid residues of the helices of rhodopsin with the
sequence of the M2 mAChR and the loop structures were constructed by using the fragment library in the "Biopolymer" module of Insight II (Accelrys, Princeton, NJ). The structure was
energy-minimized by the use of the DISCOVER 3 force field. The model of
(S)-methacholine was docked to the putative binding site of
the M2 mAChR model, guided by the ligand-receptor
interactions between the quaternary amine-Asp 103 and the carbonyl
oxygen-Tyr 403. The molecular dynamics with simulated annealing,
employing DISCOVER 3 force field, was carried out with the
(S)-methacholine bound model involving the amino acid
residues within 9 Å from the ligands.
| |
Results |
|---|
|
Top
Abstract
Introduction
Materials and Methods
Results
Discussion
References
|
|---|
Preparation of the M2 Mutant.
Throughout this
work, we have used the M2 mutant without the
putative glycosylation site and the central part of the third intracellular loop, and with the hexa-histidine tag at the C terminus, which has been previously shown to possess the intrinsic activities of
the M2 mAChR (ligand binding and G protein
activation) (Hayashi and Haga, 1996; Furukawa and Haga, 2000).
). The membrane fraction
was solubilized by the mixture of digitonin and sodium cholate, and was
purified to homogeneity using ABT-agarose and hydroxyl apatite columns (Haga and Haga, 1985) (Fig. 2). The upper
band in the SDS-PAGE was thought to represent the dimeric form of the
M2 mutant because it was stained with the
anti-hexa-histidine tag polyclonal antibodies (data not shown). The
washing of the hydroxyl apatite column with the sodium cholate-based
buffer partially removed the digitonin bound to the receptor. This step
was effective for avoiding nonspecific binding of
(S)-methacholine or
(2S,4R,5S)-muscarine to the digitonin micelles, which produced nonspecific TRNOE signals. Such signals were
observed in the presence of more than 10 mg/ml digitonin. The
concentrations of digitonin and sodium cholate in the sample containing
2 mg/ml M2 mutant were approximately 5 mg/ml and
2 mg/ml, respectively, as determined by the band intensities in thin-layer chromatography (data not shown). The addition of soybean phosphatidylcholine at 1.5 mg/ml further diminished the nonspecific TRNOE and increased the stability at higher temperature (23°C). The
M2 mutant thus prepared retained approximately
>95% and 90% of the
L-[3H]quinuclidinyl
benzilate or [3H]NMS binding activity in the
presence of 1.5 mM of (S)-methacholine or
(2S,4R,5S)-muscarine after three days
at 4 and 23°C, respectively: the Ki
values estimated under NMR conditions were essentially the same as the
Ki and EC50
values estimated for M2 mutants and M2-Gi1 fusion proteins
in the membranes (Table 1). The purified M2 mutant did not show any sign of degradation
after the NMR measurements, as assessed by SDS-PAGE (Fig. 2).
|
|
Characteristics of the Ligand Binding.
The data points in the
[3H]NMS displacement curves were well fitted to
the equation for a one-site model (data not shown). The affinities (in
Ki) for acetylcholine,
(S)-methacholine, and (2S,4R,5S)-muscarine were similar to
each other (20-80 µM), whereas that for (R)-methacholine
was approximately 20-fold lower than that of acetylcholine (Table 1),
and are consistent with the previous reports for the
M1 mAChR (Page et al., 1995).
showed
agonist-dependent [35S]GTPS binding in the
presence of 1 µM GDP. The EC50 values as well
as the [35S]GTPS binding maxima were similar
between acetylcholine and (S)-methacholine (Table 1).
Consistent with the previous report on the fusion protein of the
2-adrenergic receptor and the subunit of G
protein Gs (Seifert et al., 1998), the addition
of the G protein subunit was not necessary to observe
agonist-dependent [35S]GTPS binding to the
M2-Gi1. The
S-isomer of methacholine was chosen for the NMR analysis
because its characteristics resembled those of the physiological
ligand, acetylcholine. The EC50 and [35S]GTPS binding maximum values for
(R)-methacholine were 100-fold higher and 2-fold lower than
those for acetylcholine, respectively. Therefore, the
R-isomer is a partial agonist of the
M2 mutant.
The Conformations of (S)-Methacholine and
(2S,4R, 5S)-Muscarine in
Solution.
The coupling constants between HS1
and H2
(3JHS1-H2) and
between HR1 and H2
(3JHR1-H2) were
1.3 and 9.4 Hz for (S)-methacholine and 1.4 and 9.4 Hz for
(2S,4R,5S)-muscarine, respectively, in
the 1H NMR spectra at 800 MHz (Fig.
3). By using the modified Karplus equation (Haasnoot et al., 1980) for
3JHS1-H2, the
O-C-C-N dihedral angles of (S)-methacholine and
(2S,4R,5S)-muscarine were estimated to
be +83° and +82°, respectively. The estimated angle for
(S)-methacholine is essentially the same as those reported by Casey et al. (1971): approximately +90° for the O-C-C-N
dihedral angle determined from the coupling constants of
3JHS1-H2 (1.5 Hz) and 3JHR1-H2
(8.8 Hz). On the other hand, Partington et al. (1972) proposed
that -methacholine exists as a 3:1 mixture of the two gauche conformers with O-C-C-N dihedral angles of +60° and
60°. Although it is theoretically possible to assume that
-methacholine and muscarine take two or more conformations that are
in equilibrium, we consider it more probable that they exist
predominantly in a single conformation rather than in the two
gauche conformations with a ratio of 3 to 1 for the
following reasons: 1) Molecular mechanics calculation (Discover III)
indicated that the conformation with the O-C-C-N dihedral angles at
+81° is energetically the most stable, and that the trans
and the gauche conformations with +180° and 60°
O-C-C-N dihedral angles is not energetically preferable because of the
steric hindrance between the C2-methyl and N-methyl groups
(data not shown). 2) The present results for O-C2-C1-N dihedral angles
of (S)-methacholine and
(2S,4R,5S)-muscarine (+83° and
+82°) are similar to those estimated form their crystal structures (+85° and +73°), which is consistent with the idea that they are stable in these conformations rather than in the trans or
gauche conformations with +180° or 60° O-C-C-N
dihedral angles. 3) The NOE signal between HR1
and H2 was too weak to consider the presence of the gauche
conformer with O-C-C-N dihedral angle at 60°.
|
The Conformations of the Receptor-Bound (S)-Methacholine and (2S,4R,5S)-Muscarine. The negative NOE signals for both (S)-methacholine and (2S,4R,5S)-muscarine were best observed at the ligand-to-receptor ratio of 30. No signal for either ligand was detected in the control buffer containing detergents (0.5% digitonin and 0.2% sodium cholate) and lipids but not the M2 mutant. Furthermore, the negative NOE signals disappeared when atropine, an antagonist of the mAChRs, was added to the sample solution containing the receptor protein. These control experiments indicated that the negative intramolecular NOE signals observed were produced by the specific ligand-receptor interaction. The TRNOESY spectra were obtained in the presence and absence of atropine. Then, the former spectrum was subtracted from the latter to confer the final spectrum representing the specific binding of the above agonists to the M2 mutant. No proton peak, besides the one for the acetyl protons in (S)-methacholine, overlapped with those for atropine in the 1H NMR spectrum (data not shown). In the NOESY or TRNOESY experiment, there was no overlap. Therefore, there was no interference caused by the presence of the atropine signal in this structural analysis.
For (S)-methacholine, the TRNOE signals were observed starting from the mixing time
m = 25 ms. We
chose to use m = 50 ms for the measurement
because it minimized the appearance of indirect NOEs via spin diffusion
and still conferred significant NOE intensities for the distance
calculations. The pattern of the cross-peaks in the TRNOESY spectrum
was consistent with the conformation with the gauche
O-C2-C1-N dihedral angle (Fig. 4). The
differences in the cross-peak patterns between the NOESY (free) and
TRNOESY (receptor-bound) spectra were the ratios of the peak
intensities, H2-+NMe3/H2-HS1,
and
Me2-+NMe3/Me2-HS1.
Both of the ratios were lower in the TRNOESY spectrum than in the NOESY
spectrum (a/b > a*/b* and c/d > c*/d* in Fig. 4). This
indicated the shift in the degree of the O-C2-C1-N dihedral angle of
(S)-methacholine as it became bound to the
M2 mutant. In addition, the observation that the
intensities of the TRNOE cross-peaks, Me2-HS1
(d*) and 2-methyl-HR1 (e*), were consistently equal at various lengths of mixing time (m = 25, 50, and 100 ms) (data not shown), indicated that these distances
were approximately equal for the bound ligand. The equal distance
between Me2-HS1 and Me2-HR1
indicated that the O-C2-C1-N dihedral angle was approximately +60°
and that the binding of the ligand to the receptor led to a 20~30°
rotation from the predominant conformation of the free form.
|
m = 150 ms,
was required to detect a sufficiently intense TRNOE. The patterns of
the H2-H1 and H3-H1 cross-peaks in the TRNOESY spectrum (receptor-bound form) and the NOESY spectrum (free form) differed from each other (Fig.
5). As found with
(S)-methacholine, the ratio of the peak intensities,
H2-+NMe3/H2-HS1,
decreased as (2S,4R,5S)-muscarine
became bound to the receptor (a/b > a*/b*), indicating the
decrease in the degree of the O-C2-C1-N dihedral angle. In the NOESY
spectrum, the order of intensities was
HR1-HR3 > HS1-HR3 > HS1-HS3 > HR1-HS3 (d > f > e > c), whereas that in the TRNOESY spectrum was
HR1-HR3 > HS1-HR3 > HR1-HS3 > HS1-HS3 (d* > f* > c* > e*). This indicated that HS3 moved away from
HS1 whereas HR1 moved
closer to Hs3 as a result of binding to the
receptor (Fig. 5B). This pattern of TRNOE signals, in comparison with
that of the NOE signals, indicated that the conformation of the
receptor-bound (2S,4R,5S)-muscarine takes the gauche O-C2-C1-N dihedral angle at less than
+80~90°, most probably around +60°.
|
Ligand Docking to the M2 mAChR Binding Site.
The
amino acid residues of the M2 mAChR were
successfully fitted into a model derived from the crystal structure of
rhodopsin (Palczewski et al., 2000) to form the seven bundles of helices characteristic of the GPCRs. We took the results obtained by
Farrens et al. (1996) and Borhan et al. (2000) into account by moving the transmembrane segments 3 and 6 outward, and rotating the
transmembrane segment 6 in a counter-clockwise direction (facing the
extracellular surface) at the ligands at the binding site. The
S-isomer of methacholine was fitted into the putative
ligand-binding site of this molecular model, guided by the two
intermolecular interactions, between the quaternary amine and Asp 103 and between the acetyl group and Tyr 403. The most stable conformation
for (S)-methacholine had the gauche rather than
the trans conformation with the O-C-C-N dihedral angle at
+55.5°, as assessed by molecular dynamics with simulated annealing,
employing the DISCOVER 3 force field with the freedom to change its
conformation (Figs. 6 and 7). The
arbitrary fitting of (S)-methacholine
with the trans conformation,
followed by the energy optimization, resulted in the conversion to the gauche conformation. This confirmed that the
gauche conformation was preferred over the trans
for (S)-methacholine bound to the M2
receptor model.
|
|
; Lu and Hulme, 1999). Recently Lu et al.
(2001) proposed a model for the complex of acetylcholine and M1 mAChR based on the structure of rhodopsin and
their mutational studies, where they also assumed that acetylcholine
bound to the receptor takes the gauche form, adopting our
preliminary result (Furukawa et al., 2001). A particular difference
between our molecular model and the one by Lu et al. (2001) is that our
model includes the rotation of the transmembrane segment 6, and
thereby, the disposition of the Tyr 403 residue. This disposition of
Tyr 403 brings about a hydrogen bond network among
(S)-methacholine, Tyr 104, and Thr 190. Without rotation of
transmembrane segment 6 and disposition of Tyr 403, the formation of
this hydrogen bond network does not seem to be possible.
| |
Discussion |
|---|
|
Top
Abstract
Introduction
Materials and Methods
Results
Discussion
References
|
|---|
The determination of the ligand conformations that bind mAChRs has
been an important issue in the field of pharmacology (Portoghese, 1970). In the present study, we identified the conformations of two
muscarinic ligands, (S)-methacholine and
(2S,4R,5S)-muscarine, bound to the
M2 mAChR, as having the gauche
O-C2-C1-N dihedral angle at approximately +60°. This result was
further supported by the modeling study of the
mAChR-(S)-methacholine complex, based on the crystal
structure of rhodopsin (Palczewski et al., 2000), which showed that the
most energetically preferred conformation of
(S)-methacholine had a +55.5° O-C2-C1-N dihedral angle.
This receptor-bound conformation differs from the one in solution, because (S)-methacholine is thought to exist predominantly
as a conformer with an +80-85° O-C2-C1-N dihedral angle.
Numbers of studies using rigid acetylcholine analogs have shown that
their trans isomers induce more muscarinic activities than
the gauche isomers (Portoghese, 1970; Casy, 1975).
Therefore, there has been a general assumption that acetylcholine acts
on the mAChRs in a conformation similar to that of the trans
isomers of the rigid acetylcholine analogs (Taylor and Insel, 1990).
However, this assumption had some ambiguity, because the degree of the O-C2-C1-N dihedral angles in the active isomers varies from compound to
compound. For example, the dihedral angle of ACTM (Chothia and Pauling,
1970) and 3-dimethylamino-2-hydroxy-bicyclo[2,2,2]octane (Nelson and
Wilson, 1971) were +137° and +120°, respectively, although both of
the compounds were originally designed to represent the
trans conformer of acetylcholine (180°). On the other
hand, their inactive cis isomers, which were designed to
represent the gauche conformer of acetylcholine (~60°),
were found to have the eclipsed O-C2-C1-N dihedral angle at 0°
(Chothia and Pauling, 1970; Nelson and Wilson, 1971). In addition,
there were such cases as cis- and
trans-dimethyldiacetoxypiperidium iodide (Lewis et al.,
1973) and d- and
l-1-methyl-3-acetoxy-trans-decahydroquinolin methiodide (Smissman and Chappell, 1969), where the isomers
representing the gauche O-C2-C1-N dihedral angle induced
more muscarinic activity than those representing the
trans-O-C2-C1-N. The O-C2-C1-N dihedral angles of the
cis- and trans-dimethyldiacetoxypiperidium iodide isomers were approximately 60° and 180° (Lewis et al., 1973), and
those isomers of the d- and
l-1-methyl-3-acetoxy-trans-decahydroquinolin methiodide isomers were 74° and 169° (Stephen et al., 1972),
respectively. Together, these studies indicate that the rigid
acetylcholine analogs with a broad O-C2-C1-N dihedral angles (60 to
137°), rather than exclusively 180°, are able to induce the
muscarinic activities (Baker et al., 1971). Thus, these data are not
sufficient to determine the precise O-C2-C1-N dihedral angle at which
acetylcholine binds to the receptor and induces the pharmacological
activity. It should be noted that Partington et al. (1972)
pointed out the absence of simple correlation between the predominant
conformations and the potency of the drugs and that Schulman et al.
(1983) proposed the significance of the distance between the
quaternary amine nitrogen and the ester oxygen of agonists rather than
the O-C-C-N dihedral angle for the pharmacological activity.
Within the simplest assumption that the receptor-bound conformation of
a ligand represents the pharmacologically active form, the present
results indicate that the conformation with O-C2-C1-N at +60° induces
pharmacological activity. This interpretation is compatible with the
studies of cis-dimethyldiacetoxypiperidinium iodide and
d-1-methyl-3-acetoxy-trans-decahydroquinolin
methiodide, in which the cis- or d-isomers
(O-C2-C1-N dihedral angle at 60° and 74°) are preferred to the
trans- or l-isomers (O-C2-C1-N dihedral angle at
180° and 169°) in inducing pharmacological activities (Smissman and
Chappell, 1969; Lewis et al., 1973). However, it cannot explain the
pharmacological activities induced by many other rigid analogs with an
O-C2-C1-N dihedral angle such as 120°.
On the other hand, there is the possibility that the receptor-bound
ligand assumes multiple conformations, depending on the state of the
mAChR. It is known that the M2 mutant, or
M2 mAChR, as well as other GPCRs, binds agonists
with high and low affinity, depending on the presence and absence of G
proteins and guanine nucleotides (Hulme et al., 1983; Haga et al.,
1986). In this study, the NMR measurements were carried out for the
purified M2 mutant in the absence of G protein.
Under these conditions, the M2 mAChR and the
M2 mutant are known to bind to the agonist with
low affinity (Hulme et al., 1983; Haga et al., 1986). This low-affinity
state represents the initial binding of the agonists to the receptor, before the association with G proteins. The receptors show
high-affinity agonist binding when they associate with G proteins in
the absence of guanine nucleotides. This high-affinity state
corresponds to the transition state for the GTP/GDP exchange in the
receptor-G protein complex, which leads to the pharmacological activity
(Gilman, 1987; Haga and Haga, 1987). The high-affinity state has
1000-fold higher affinity for the agonist than the low-affinity state
(Shiozaki and Haga, 1992). It will be interesting to elucidate whether
such a change in the state of the receptor accompanies the
conformational shift of the ligand, from the initially bound
conformation to another one with distinct O-C2-C1-N dihedral angles.
In summary, we determined the receptor-bound conformation of (S)-methacholine and (2S,4R,5S)-muscarine, in the absence of G protein, to have the gauche O-C2-C1-N dihedral angle at approximately +60°. The mAChR recognizes the conformer with an O-C2-C1-N dihedral angle at +60°, which differs from the predominant conformer in solution with an O-C2-C1-N dihedral angle at +80-85°.
| |
Acknowledgments |
|---|
We thank T. Okada for technical assistance with the Sf9/baculovirus, Drs. K. Wakamatsu and M. Murata for valuable discussions, and Dr. K. Sugase for help in the energy calculations. We also thank Dr. T. Kigawa for help in NMR measurements.
| |
Footnotes |
|---|
Received December 17, 2001; Accepted June 25, 2002
This work was supported by Core Research for Evolutional Science and Technology (CREST), Japan Science and Technology Corporation and the Japan Society for the Promotion of Science-Research for Future Program.
H.F. and T.H. contributed equally to this work.
Address correspondence to: Tatsuya Haga, Institute for Biomolecular Science, Gakushuin University, 1-5-1 Mejiro, Toshima-Ku, Tokyo, Japan 171-8588. E-mail: tatsuya.haga{at}gakushuin.ac.jp
| |
Abbreviations |
|---|
mAChR, muscarinic acetylcholine receptor;
GPCR, G protein-coupled receptor;
TRNOE, transferred nuclear Overhauser
effect;
NOE, nuclear Overhauser effect;
NOESY, nuclear Overhauser
effect spectroscopy;
TRNOESY, transferred nuclear Overhauser effect
spectroscopy;
NMS, N-methylscopolamine;
GTPS, guanosine 5'-O-(3-thio)triphosphate;
Sf9, Sporodoptera frugiperda;
PAGE, polyacrylamide gel
electrophoresis.
| |
References |
|---|
|
Top
Abstract
Introduction
Materials and Methods
Results
Discussion
References
|
|---|
2-adrenoceptor-Gs
fusion protein.
Eur J Biochem
255:
369-82[Medline].This article has been cited by other articles:
|
|
 |
|
  S.-J. Han, F. F. Hamdan, S.-K. Kim, K. A. Jacobson, L. M. Bloodworth, B. Li, and J. Wess Identification of an Agonist-induced Conformational Change Occurring Adjacent to the Ligand-binding Pocket of the M3 Muscarinic Acetylcholine Receptor J. Biol. Chem., October 14, 2005; 280(41): 34849 - 34858. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
 |
 |
|
 |
 |
 |
+
I{1.02 
i*
