(Received for publication, February 27, 1997, and in revised form, June 18, 1997)
From the Departments of Physiology and Medicine, Medical College of Virginia, Richmond, Virginia 23298-0711
Muscarinic m2 and
m4 receptors couple preferentially to inhibition of
adenylyl cyclase, whereas m1, m3, and
m5 receptors couple preferentially to activation of
phospholipase C- and in some cells to stimulation of cAMP. Smooth
muscle cells were shown to express adenylyl cyclases types V and/or VI.
Acetylcholine (ACh) stimulated the binding of
[35S]GTP
S·G
complexes in smooth muscle membranes
to G
q/11 and G
i3 antibody. Binding to
G
q/11 antibody was inhibited by the m3
receptor antagonist, 4-DAMP, and binding to G
i3 antibody
was inhibited by the m2 receptor antagonist,
N,N
-bis[6[[(2-methoxyphenyl)methyl]amino]hexyl]-1,8-octanediamine tetrahydrochloride (methoctramine). The decrease in basal cAMP (35 ± 5%) induced by ACh in dispersed muscle cells was
accentuated by 4-DAMP or G
antibody (55 ± 8 to
63 ± 6%). In contrast, methoctramine, pertussis toxin (PTx), or
G
i3 antibody converted the decrease in cAMP to increase
above basal level (+28 ± 5 to +32 ± 6%); the increase in
cAMP was abolished by 4-DAMP or G
antibody. In muscle
cells where only m3 receptors were preserved by selective
receptor protection, ACh caused only an increase in cAMP that was
abolished by 4-DAMP. Conversely, in muscle cells where only
m2 receptors were preserved, ACh caused an accentuated decrease in cAMP that was abolished by methoctramine or PTx. In conclusion, m2 receptors in smooth muscle couple to
inhibition of adenylyl cyclases V/VI via G
i3, and
m3 receptors couple to activation of the enzymes via
G
q/11.
Complementary DNA clones encoding the full sequences of eight isoforms of mammalian adenylyl cyclase (types I-VIII) and the partial sequences of two additional isoforms (types IX and X) have been isolated (1-10). The amino acid sequences (range 1064-1248 residues) are arranged in two cassettes of six transmembrane-spanning domains (9-11). Overall homology is 60% with some cytoplasmic regions exhibiting up to 93% amino acid identity. The presumed catalytic domains (C1a and/or C2a) are homologous to the corresponding domains of membrane-bound homodimeric and soluble heterodimeric guanylyl cyclases (11, 12). Structural homology among the various isoforms is most evident between types II and IV and types V and VI. All the isoforms are expressed in the brain, and types V and VI are the predominant isoforms in the periphery (8-11).
Although functional regulation of adenylyl cyclases is diverse, three
broad categories can be distinguished comprising types I, III, and
VIII, types II, IV, and probably VII, and types V and VI (10, 11). All
adenylyl cyclases are activated by the diterpene, forskolin, and the
subunit of Gs. Types I and VIII, which are expressed
exclusively in neurons, are stimulated by submicromolar concentrations
of Ca2+ and calmodulin, whereas type III, which is more
widely expressed, is stimulated by low micromolar concentrations of
Ca2+ (13-16); type I is effectively inhibited by G
but only moderately inhibited by Gi and Go
(17-21). Types II and IV are not stimulated by
Ca2+/calmodulin or inhibited by Gi but are
stimulated by G
(19, 21-24). Stimulation by G
, initially
thought to be conditional on concurrent stimulation by
G
s, is now viewed as highly synergistic, with only
modest stimulation by G
alone (21). Types V and VI are inhibited
by Gi and by submicromolar concentrations of cytosolic
Ca2+ elicited by capacitative Ca2+ influx but
not by Ca2+ release from sarcoplasmic stores (24, 25-28);
inhibition by Go or stimulation by G
remains
uncertain. Both types V and VI contain consensus sequences for
phosphorylation and exhibit feedback inhibition, by
cAMP-dependent protein kinase (29, 30).
The expression of adenylyl cyclase isoforms in smooth muscle has not
been determined. The regulatory pattern suggested by our previous
studies in gastrointestinal smooth muscle is consistent with the
presence of types V and/or VI. Agonist-induced cAMP formation is
mediated by Gs (31, 32), and forskolin-stimulated cAMP formation is inhibited, depending on the agonist, by Gi1,
Gi2, Gi3, and Go (33-35). Thus,
inhibition induced by somatostatin (acting via sstr3) is mediated by
Gi1 and Go (33), and inhibition induced by
opioid agonists (acting via µ,
, and
receptors) is mediated by
Gi2 and Go (34); inhibition induced by
adenosine (acting via A1 receptors) is mediated by
Gi3 but not by Go (35). Forskolin-stimulated cAMP formation is inhibited in feedback fashion by
cAMP-dependent protein kinase (36). Phorbol esters have no
effect on basal or forskolin-stimulated cAMP formation (37). However,
inhibition of cAMP formation induced by agonists acting via
Gi1 or Gi2 (but not Gi3) is partly
reversed by concomitant activation of
PKC1; the effect reflects
selective PKC-dependent phosphorylation of
G
i1 and G
i2 but not G
i3 or
G
o (37).
It is well established that the muscarinic receptors, m2
and m4, are preferentially coupled to inhibition of
adenylyl cyclase; the odd-numbered receptors, m1,
m3, and m5, are preferentially coupled to
phosphoinositide hydrolysis but, in some cells, can also increase the
levels of cAMP (38-44). The mechanisms responsible for the increase in
cAMP are likely to reflect the type of adenylyl cyclase expressed in
various cells. In cells expressing predominantly Ca2+/calmodulin-sensitive adenylyl cyclases types I, III,
and VIII, activation could result from Ca2+ mobilization
and activation of PKC (43). This, however, is unlikely in cells
expressing predominantly Ca2+/calmodulin-insensitive types
II and IV or in cells expressing types V and VI that are inhibited by
physiological levels of Ca2+ (10, 28). In the present
study, we show that types V and/or VI, the isoforms of adenylyl cyclase
expressed in smooth muscle cells, are inhibited by m2
receptors via the subunit of Gi3 and concurrently
activated by m3 receptors via the
subunits of
Gq/11.
Muscle cells were isolated from the circular muscle layer of the rabbit stomach by successive enzymatic digestion, filtration, and centrifugation as described previously (31, 33). Briefly, muscle strips were incubated for 30 min at 31 °C in 15 ml of HEPES medium containing 0.1% collagenase (type II) and 0.1% soybean trypsin inhibitor. The composition of the medium was 120 mM NaCl, 4 mM KCl, 2.6 mM KH2PO4, 2 mM CaCl2, 0.6 mM MgCl2, 25 mM HEPES, 14 mM glucose, and 2.1% Eagle's essential amino acid mixture. After washing, the tissues were re-incubated in the same medium for 30 min. The digested tissue was washed with enzyme-free medium, and the cells were allowed to disperse spontaneously for 30 min. Suspensions of single muscle cells were harvested by filtration through 500-µm Nitex mesh. The suspensions were centrifuged twice for 10 min at 350 × g.
In experiments with G protein antibodies, the cells were permeabilized as described previously (31, 33) by incubation for 10 min with 35 µg/ml saponin in a medium containing 20 mM NaCl, 100 mM KCl, 5 mM MgSO4, 1 mM NaH2PO4, 25 mM NaHCO3, 0.34 mM CaCl2, 1 mM EGTA, and 1% bovine serum albumin. The cells were centrifuged at 350 × g for 5 min, washed free of saponin, and resuspended in the same medium.
Radioligand Binding to Muscarinic Receptors in Dispersed Muscle CellsRadioligand binding to dispersed muscle cells was done as described previously (33). Muscle cells were suspended in HEPES medium containing 1% bovine serum albumin. Triplicate aliquots (0.5 ml) of cell suspension (106 cells/ml) were incubated for 15 min with 1 nM [3H]scopolamine alone or in the presence of acetylcholine, methoctramine, or 4-DAMP. Bound and free radioligand were separated by rapid filtration under reduced pressure through 5-µm polycarbonate Nucleopore filters followed by repeated washing (4 times) with 3 ml of ice-cold HEPES medium containing 0.2% bovine serum albumin. Nonspecific binding was measured as the amount of radioactivity associated with the muscle cells in the presence of 10 µM unlabeled ligand. Specific binding was calculated as the difference between total and nonspecific binding (mean ± S.E. 33 ± 6%). IC50 values were calculated from competition curves using the P.fit program (Biosoft; Elsevier Publishing, Cambridge, UK).
Measurement of cAMP in Dispersed Muscle Cells by RadioimmunoassaycAMP was measured in dispersed cells by radioimmunoassay as described previously (33, 34). Aliquots (0.5 ml) containing 106 cells/ml were incubated with 0.1 µM acetylcholine, and the reaction was terminated after 60 s with 6% cold trichloroacetic acid (v/v). The mixture was centrifuged at 2,000 × g for 15 min at 4 °C. The supernatant was extracted three times with 2 ml of diethyl ether and lyophilized. The samples were reconstituted for radioimmunoassay in 500 µl of 50 mM sodium acetate (pH 6.2) and acetylated with triethylamine/acetic anhydride (3:1 v/v) for 30 min. cAMP was measured in duplicate using 100-µl aliquots and expressed as pmol/106 cells.
Selective Protection of Muscarinic ReceptorsA technique of
selective receptor protection was used to determine the presence and
function of m2 and m3 receptors. The technique was previously used to determine the co-existence and function of
opioid µ, , and
receptors (45), 5-hydroxytryptamine
5-HT2 and 5-HT4 receptors (32), histamine
H1 and H2 receptors (46), and tachykinin
NK1, NK2, and NK3 receptors (47).
The technique involves protection of one receptor type with a selective
agonist or antagonist followed by inactivation of all unprotected
receptors by brief treatment with a low concentration of
N-ethylmaleimide. In the present study, the selective
m2 receptor antagonist, methoctramine (10 nM),
and m3 receptor antagonist, 4-DAMP (10 nM),
were used in separate experiments to protect m2 and
m3 receptors, respectively. Freshly dispersed muscle cells
were incubated with one antagonist at 31 °C for 2 min followed by
addition of 5 µM N-ethylmaleimide for 20 min.
The cells were centrifuged twice at 150 × g for 10 min
to eliminate the protective antagonist and N-ethylmaleimide and resuspended in fresh HEPES medium. The cAMP response of cells treated in this fashion was compared with the response of untreated (naive) cells. Previous studies have shown that the coupling of protected receptors to signaling pathways remains intact (32, 46, 47).
Smooth muscle cells incubated with N-ethylmaleimide without
protective ligand lost their ability to respond to receptor-linked agonists but retained their ability to respond to agents that bypass
receptors (e.g. ionomycin, KCl, forskolin), implying that post-receptor mechanisms were intact (46, 47).
Cell homogenates were prepared from dispersed gastric muscle cells and solubilized on ice for 1 h in 20 mM Tris (pH 8.0), 1 mM EDTA, 1 mM dithiothreitol, 100 mM NaCl, and 0.5% sodium cholate. For control studies, homogenates were prepared from rat brain. The suspension was centrifuged at 13,000 × g for 5 min. Solubilized membrane proteins (60-70 µg) were resolved by 7.5% SDS-polyacrylamide gel electrophoresis and electrophoretically transferred to nitrocellulose membranes. After incubation in 5% non-fat dry milk to block nonspecific antibody binding, the blots were incubated for 12 h at 4 °C with antibodies to types II, III, and IV, and a common antibody to types V and VI, and then for 1 h with anti-rabbit IgG conjugated with horseradish peroxidase. The bands were identified by enhanced chemiluminescence.
Identification of Muscarinic Receptor-activated G ProteinsG proteins selectively activated by acetylcholine were
identified by the method of Okamoto et al. (48). Ten ml of
muscle cell suspension (2 × 106 cells/ml) were
homogenized in 20 mM HEPES medium (pH 7.4) containing 2 mM MgCl2, 1 mM EDTA, and 2 mM dithiothreitol. After centrifugation at 27,000 × g for 15 min, the crude membranes were solubilized for 60 min at 4 °C in 20 mM HEPES medium (pH 7.4) containing 2 mM EDTA, 240 mM NaCl, and 1% CHAPS. The
membranes were incubated for various periods at 37 °C with 60 nM [35S]GTPS in a solution containing 10 mM HEPES (pH 7.4), 100 µM EDTA, and 10 mM MgCl2. The reaction was stopped with 10 volumes of 100 mM Tris-HCl medium (pH 8.0) containing 10 mM MgCl2, 100 mM NaCl, and 20 µM GTP, and the mixture was placed in wells pre-coated with specific G protein antibodies. After incubation for 2 h on ice, the wells were washed three times with phosphate buffer solution containing 0.05% Tween 20, and the radioactivity from each well was
counted. Coating with G protein antibodies (1:1000) was done after the
wells were first coated with anti-rabbit IgG (1:1000) for 2 h on
ice. The selective m2 receptor antagonist, methoctramine, and m3 receptor antagonist, 4-DAMP, were used to identify
the receptor subtype coupled to a given G protein.
Results were expressed as means ± S.E. of n separate experiments and evaluated statistically using Student's t test for paired or unpaired values.
Materials125I-cAMP,
[3H]scopolamine, and [35S]GTPS were
obtained from NEN Life Science Products; HEPES was from Research
Organics, Cleveland, OH; soybean trypsin inhibitor and collagenase
(type II) were from Worthington; 4-DAMP, methoctramine,
p-fluorohexahydro-siladifenidol (p-F-HHSiD) were
from Research Biochemicals International, Natick, MA;
N-ethylmaleimide and all other chemicals were from Sigma. Antibodies to adenylyl cyclase types II, III, and IV, a common antibody
to types V and VI, and antibody to G
were obtained from
Santa Cruz Biotechnology, Santa Cruz, CA. Pertussis toxin and
antibodies to G
q/11, G
i1-2,
G
i3, G
o, and G
s and peptide fragments against which antibodies to G
q/11
(QLNLKEYNLV) and G
i3 (KNNKECGLY) were raised were
obtained from Calbiochem. The ability of these antibodies to block
activation or inhibition of specific effector enzymes
(phospholipase-
1, phospholipase-
3, nitric oxide synthase,
adenylyl cyclase) has been demonstrated in recent studies (31-35), and
a concentration of 10 µg/ml was found to be maximally effective.
Studies of cloned
muscarinic m3 and m2 receptors have established
4-DAMP (Kd 0.5 nM) and methoctramine
(Kd 12 nM) as selective antagonists of
m3 and m2 receptors, respectively (39, 49-52).
Competition binding studies in dispersed gastric smooth muscle cells
using [3H]scopolamine as radioligand confirmed the
selectivity of the two antagonists. In muscle cells where only
m3 receptors were preserved, [3H]scopolamine
binding was preferentially inhibited by 4-DAMP (IC50 0.4 ± 0.1 nM) with a potency 104 greater
than that of methoctramine (IC50 4 ± 1 µM) (Fig. 1). Conversely,
in muscle cells where only m2 receptors were preserved, [3H]scopolamine binding was preferentially inhibited by
methoctramine (IC50 8 ± 3 nM) with a
potency 2 × 103 greater than that of 4-DAMP
(IC50 15 ± 4 µM). The affinities of
4-DAMP and methoctramine in cells where only one native receptor type
was preserved matched closely those obtained with cloned m2
and m3 receptors.
Schild analysis of the relative potencies of the two antagonists in inhibiting acetylcholine-induced contraction confirmed that 4-DAMP was about 6 × 103-fold more potent than methoctramine as an antagonist of m3 receptors that mediate contraction. Contraction in dispersed muscle cells was measured by scanning micrometry as described previously (33, 34). Muscle cells in which only m2 receptors were preserved lost the ability to contract in response to acetylcholine, whereas muscle cells in which only m3 receptors were preserved retained fully their ability to contract (EC50 0.4 nM in muscle cells expressing one or both receptor types). The concentration of acetylcholine that elicited maximal contraction (0.1 µM) abolished [3H]scopolamine binding. The radioligand binding and pharmacological studies in dispersed gastric smooth muscle cells confirmed the selectivity of the antagonists and the validity of the receptor protection technique.
Expression of Adenylyl Cyclase Isoforms in Gastric Smooth MuscleWestern blot analysis of homogenates derived from
dispersed gastric smooth muscle cells using antibodies to adenylyl
cyclase type II, type III, type IV, and a common antibody to types V
and VI disclosed the presence of type V and/or type VI only (Fig. 2). In contrast, homogenates from rat
brain disclosed the presence of types II, III, IV, and V/VI (Fig. 2).
The selective expression of types V and/or type VI in smooth muscle
conforms to the predominant expression of these two types in peripheral
tissues (8-11). The properties of adenylyl cyclase in gastric and
intestinal smooth muscle are consistent with the properties of types V
and VI; both isozymes are inhibited by Gi and
G
o (unlike types II and IV) (33, 34) and are not
stimulated by Ca2+ or calmodulin (unlike types I, III, and
VIII) (10, 28).
Identification of G Proteins Coupled to Muscarinic m2 and m3 Receptors
Incubation of solubilized muscle
cell membranes with acetylcholine (0.1 µM) and
[35S]GTPS (60 nM) for 20 min caused a
significant, time-dependent increase in the binding of
[35S]GTP
S·G
complexes to wells pre-coated with specific antibody to
G
q/11 and G
i3 but not to wells pre-coated
with antibodies to G
s, G
i1-2, or
G
o (Fig. 3 and Table I).
The increase in bound radioactivity reflected
acetylcholine-dependent activation of the dissociated
subunits of Gq/11 and Gi3 by
[35S]GTP
S. The increase in the binding of
[35S]GTP
S·G
complexes to wells pre-coated with
G
q/11 antibody was abolished by the m3
receptor antagonists, 4-DAMP (0.1 µM), and
p-fluorohexahydro-siladifenidol (0.1 µM) but
not by the m2 receptor antagonist, methoctramine (0.1 µM) (Fig. 3 and Table I); control binding in the absence
of acetylcholine was not affected by 4-DAMP or methoctramine (control
binding at 20 min 3414 ± 377 versus 3620 ± 467 and 3547 ± 587 with 4-DAMP and methoctramine, respectively).
Conversely, the increase in the binding of
[35S]GTP
S·G
complexes to wells pre-coated with
G
i3 antibody was abolished by methoctramine but not by
4-DAMP or p-fluorohexahydro-siladifenidol (Fig. 3 and Table
I); control binding in the absence of acetylcholine was not affected by
4-DAMP or methoctramine (control binding at 20 min 3287 ± 419 versus 3711 ± 653 and 3428 ± 592 with 4-DAMP and
methoctramine, respectively). The pattern implied that m3 receptors were coupled to Gq/11 while m2
receptors were coupled to Gi3.
|
Peptides I and II, comprising the G protein sequences against which the
Gi3 and G
q/11 antibodies, respectively,
were raised, were used to block the binding of GTP
S·G
complexes
to the corresponding antibody. Peptide I inhibited the binding of
GTP
S·G
complexes to G
i3 antibody in a
concentration-dependent fashion, whereas peptide II had no
effect (Fig. 3 and Table I). Conversely, peptide II inhibited the
binding of GTP
S·G
complexes to G
q/11 antibody in
a concentration-dependent fashion, whereas peptide I had no effect (Fig. 3 and Table I). It is noteworthy that the peptides inhibited the control binding of GTP
S·G
complexes to the
corresponding antibody as well as the increase in binding induced by
acetylcholine.
Acetylcholine (0.1 µM) caused a
significant 35 ± 5% decrease in basal cAMP levels of dispersed
gastric muscle cells (basal level, 3.9 ± 0.6 pmol/106
cells) (Fig. 4). Neither 4-DAMP nor
methoctramine alone had any effect on basal cAMP (4.5 ± 0.4 pmol/106 cells). Pretreatment of the cells with 0.1 µM methoctramine converted the decrease induced by
acetylcholine to a significant 32 ± 6% increase above basal
level (Fig. 4); the increase in the presence of methoctramine was
inhibited in a concentration-dependent fashion by 4-DAMP
(EC50 0.5 nM) and abolished by 0.1 µM 4-DAMP (Figs. 4 and 5).
Pretreatment of the cells for 1 h with 200 ng/ml PTx so as to
uncouple m2 receptors from Gi3 also converted
the acetylcholine-induced decrease in cAMP to a significant 36 ± 5% increase above basal level; the increase in cAMP was abolished by
0.1 µM 4-DAMP (Fig. 4).
Conversely, pretreatment of the cells with 0.1 µM 4-DAMP
accentuated the decrease in cAMP induced by acetylcholine (63 ± 6% versus control response
35 ± 5% with
acetylcholine alone; p < 0.01); the decrease was
reversed in a concentration-dependent fashion by methoctramine
(EC50 1 nM) and abolished by 0.1 µM methoctramine (Figs. 5 and
6). The accentuated decrease in cAMP in
the presence of 0.1 µM 4-DAMP was abolished by
pretreatment of the cells with PTx (Fig. 6). The pattern of response to
acetylcholine reflected concurrent inhibition of cAMP mediated by
m2 receptors and stimulation mediated by m3
receptors.
The protein kinase C inhibitor, calphostin C (1 µM), had
no effect on the increase in cAMP induced by acetylcholine in the presence of methoctramine or PTx (+37 ± 4 and +39 ± 5%,
respectively, versus +35 ± 5%) or the accentuated
decrease in cAMP induced by acetylcholine in the presence of 4-DAMP
(59 ± 5% versus
63 ± 6%).
Permeabilized muscle cells were used to identify the G
proteins coupled to m2 and m3 receptors by
functional blockade with G protein antibodies. Basal cAMP and the
decrease in cAMP induced by acetylcholine were not affected by
permeabilization (basal level, 4.2 ± 0.5 pmol/106
cells; acetylcholine-induced decrease of cAMP, 33 ± 3%).
Pretreatment of permeabilized muscle cells with methoctramine, PTx, or
G
i3 antibody (10 µg/ml) converted the decrease in cAMP
to an increase above basal level (+27 ± 5, +35 ± 5, and
+28 ± 5%, respectively) (Fig. 7).
Preincubation of the cells for 1 h with G
antibody (10 µg/ml) abolished the increase in cAMP induced by all three agents
(Fig. 7), and preincubation with antibodies to G
q/11 or G
i1-2 (each 10 µg/ml) had no effect (range of
response +27 ± 7 to +30 ± 6%). The increase in cAMP
induced by G
i3 antibody was also abolished by 4-DAMP
(2 ± 5%).
Preincubation of permeabilized muscle cells for 1 h with
G antibody (10 µg/ml) accentuated the decrease in cAMP
induced by acetylcholine (
51 ± 6% versus control
response
35 ± 5%; p < 0.05) (Fig.
8); preincubation with antibodies to
G
q/11 or G
i1-2 had no effect (
33 ± 6% and
29 ± 3%). The accentuated decrease in cAMP induced
by G
antibody was abolished by methoctramine, PTx, or
G
i3 antibody (Fig. 8). 4-DAMP also accentuated the
decrease in cAMP induced by acetylcholine (
56 ± 4%
versus control response
33 ± 5%; p < 0.01) (Fig. 8). The accentuated decrease induced by 4-DAMP was
abolished by preincubation of the cells with G
i3 antibody (10 µg/ml) (Fig. 8) but not with antibodies to
G
q/11 or G
i1-2 (
53 ± 6% and
57 ± 3%).
The pattern of response elicited by treatment of permeabilized muscle
cells with specific G protein antibodies implied that the
acetylcholine-induced decrease in cAMP was mediated by m2 receptors via Gi3, whereas the increase in cAMP was
mediated by m3 receptors via G
q/11.
The results obtained in naive muscle cells expressing both
m2 and m3 receptors were corroborated in cells
where only one receptor type was preserved. In cells where only
m3 receptors were preserved, acetylcholine caused only an
increase in cAMP (+27 ± 3%), similar to that elicited in naive
cells when m2 receptors were blocked with methoctramine or
uncoupled with PTx, or when Gi3 was blocked with
G
i3 antibody (Fig. 9). The
increase in cAMP induced by acetylcholine was abolished by 4-DAMP but
was not affected by methoctramine or PTx.
In cells where only m2 receptors were preserved,
acetylcholine elicited an accentuated decrease in cAMP (52 ± 4%) (Fig. 9), similar to that elicited in naive cells in the presence
of 4-DAMP or after treatment with G
antibody (Figs. 6
and 9). The accentuated decrease in cAMP was abolished by methoctramine
or PTx but was not affected by 4-DAMP (Fig. 9).
The present study confirmed the co-existence of muscarinic
m2 and m3 receptors on gastric smooth muscle
cells and demonstrated the differential coupling of the two receptor
types to adenylyl cyclase. Muscarinic m2 receptors, the
predominant receptor type expressed in gastrointestinal smooth muscle
(40, 53), are known to be coupled to inhibition of adenylyl cyclase via
a PTx-sensitive G protein (40, 54-56). The present study identified
this G protein as Gi3. In addition, the present study
demonstrated a direct coupling of m3 receptors to
activation of adenylyl cyclase via the subunits of
Gq/11. Although coupling of m3 receptors to
activation of adenylyl cyclase was known to occur in some cell types
(38, 41), the mechanism(s) underlying this effect had not been
determined.
The evidence for the differential coupling of m2 and
m3 receptors was based on a combination of experimental
strategies. (a) GTPS·G
complexes activated by
m3 receptors bound selectively to G
q/11
antibodies, whereas GTP
S·G
complexes activated by m2 receptors bound selectively to G
i3
antibodies. No acetylcholine-induced increase in binding to
G
i1-2 or G
o antibodies could be detected. The binding of GTP
S·G
complexes to G
i3
or G
q/11 antibodies was selectively blocked by peptide
fragments against which these antibodies were raised. (b)
Blockade of m2 receptors or their uncoupling from G
proteins by PTx converted the decrease in cAMP induced by acetylcholine
to increase above basal level; the increase was blocked by a selective
m3 receptor antagonist and by a common antibody to
G
implying that activation of adenylyl cyclase was
mediated by G
derived from m3-dependent activation of Gq/11. (c) Concurrent activation
of adenylyl cyclase mediated by m3 receptors attenuated the
predominant inhibition mediated by m2 receptors; blockade
of the stimulatory effect with an m3 receptor antagonist or
G
antibody accentuated the decrease in cAMP; the
accentuated decrease was abolished by methoctramine, PTx, and
G
i3 antibody. (d) The results obtained in
naive muscle cells expressing both receptor types were corroborated
using muscle cells in which only one receptor type was preserved. Thus,
cells where only m2 receptors were preserved responded to
acetylcholine by an accentuated decrease in cAMP which was abolished by
an m2 receptor antagonist or PTx; conversely, cells where
only m3 receptors were preserved responded to acetylcholine
with only an increase in cAMP that was abolished by an m3
receptor antagonist.
The selectivity of the antagonists used for receptor protection was
demonstrated by radioligand binding, and the measured IC50
values closely matched those derived from measurements in cells
expressing cloned m2 or m3 receptors (39,
49-52). Pharmacological analysis confirmed the validity of the
receptor protection technique that had previously been used to
characterize a variety of receptors co-expressed on smooth muscle cells
and coupled to the same or distinct signaling pathways (e.g.
histamine H1 and H2 receptors (46),
5-HT2 and 5-HT4 receptors (32), adenosine
A2b and A1 receptors (35), tachykinin
NK1, NK2, and NK3 receptors (47), and opioid µ, , and
receptors (34)). Only receptors were inactivated while post-receptor mechanisms were spared; in particular, neither basal nor forskolin-stimulated cAMP formation was affected (32,
35).
Increasing awareness of the diverse regulation of various isoforms of adenylyl cyclases requires that the proposed mechanisms for activation or inhibition of the enzymes be consistent with the properties of the adenylyl cyclase(s) expressed in a given cell type (8-11). In the present study, adenylyl cyclase types II, III, and IV could not be detected in dispersed gastric muscle cells. Type V and/or type VI was detected by Western blot analysis since the common antibody could not distinguish between the two types. The tissue expression and regulatory features of adenylyl cyclase in smooth muscle are consistent with the absence of types I and VIII which are confined to neurons (13-16), and with the absence of types II, IV, and possibly VII which are not susceptible to inhibition by Gi (11, 21-23).
The types of adenylyl cyclase (V and/or VI) expressed in smooth muscle
are regulated in similar fashion and known to be inhibited by various
isoforms of Gi and by feedback phosphorylation by
cAMP-dependent protein kinase (25, 26, 29, 30). Both these
properties are evident in smooth muscle. Forskolin-stimulated cAMP
formation in gastric smooth muscle, for example, is augmented by
selective inhibition of cAMP-dependent protein kinase
activity with myristoylated protein kinase A inhibitor (36). Inhibition
of adenylyl cyclase in smooth muscle by various isoforms of
Gi and Go appears to be receptor-specific.
Inhibition by somatostatin sstr3 is mediated additively by
Gi1 and Go (33), whereas inhibition by opioid µ, , and
receptors is mediated additively by Gi2
and Go (34). Go is not involved in inhibition
by adenosine A1 or muscarinic m2 receptors that
are coupled to Gi3 (35, 37).
Inhibition by submicromolar concentrations of Ca2+
resulting from capacitative entry of Ca2+ is a distinctive
regulatory feature of adenylyl cyclases types V and VI (10, 21); this
feature, however, could not be demonstrated under our experimental
conditions. The measurement of cAMP in the present study was made
during the first 60 s when the concomitant rise in cytosolic
Ca2+ induced by acetylcholine is determined by inositol
1,4,5-trisphosphate-dependent release of Ca2+
from sarcoplasmic stores and thus precedes store depletion and capacitative Ca2+ entry (57, 58). Furthermore, suppression
of phosphoinositide hydrolysis and inositol
1,4,5-trisphosphate-dependent Ca2+ release by
Gq/11 antibody had no effect on the increase in cAMP mediated by m3 receptors. The concurrent activation of PKC
also had no effect on adenylyl cyclase activity in smooth muscle since inhibition of PKC activity by calphostin C had no effect on cAMP levels. As previously shown (37), however, PKC can influence adenylyl
cyclase activity in gastrointestinal smooth muscle indirectly by
selective phosphorylation of G
i1 and G
i2;
concurrent activation of PKC attenuated the inhibition of
forskolin-stimulated cAMP mediated by somatostatin sstr3
receptors coupled to Gi1 and by opioid µ,
, and
receptors coupled to Gi2, but not by adenosine A1 or muscarinic m2 receptors coupled to
Gi3 (33-35).
A novel aspect of this study was the ability of G derived from
the dissociation of Gq/11 to activate smooth muscle
adenylyl cyclases type V and/or VI. Activation was not conditional on
concurrent activation of Gs, as is the case for activation
of types II and IV (11, 21). The low abundance of G
in peripheral
tissues, particularly when derived from the dissociation of
Gq/11, raises the possibility that, when expressed in
smooth muscle, types V and VI may be unusually sensitive to activation
by G
or by specific combinations of
subunits.
In summary, muscarinic m2 receptors are coupled to
inhibition of adenylyl cyclases V and/or VI in smooth muscle via the
subunit of Gi3, whereas m3 receptors are
coupled to activation of the enzymes via the
subunits of
Gq/11. The cAMP response to muscarinic agonists reflects
the predominant inhibitory influence of m2 receptors.