G protein-coupled receptors receive many of the neural, hormonal, and paracrine inputs to gastrointestinal (GI) smooth muscle cells. This article examines the major G protein-coupled receptors, G proteins, and effectors that mediate responses to enteric neuromuscular transmitters. Excitatory transmitters primarily couple through Gq/11 and Gi/Goproteins and elicit responses via formation of inositol trisphosphate and diacylglycerol and inhibition of adenylyl cyclase. Several inhibitory transmitters couple through Gs and activation of adenylyl cyclase. There are interesting examples, however, of inhibitory transmitters apparently using pathways regulated by Gq/11 to elicit responses by localized Ca2+ release and activation of Ca2+-dependent ion channels. G protein-coupled receptors may also be differentially expressed by smooth muscle cells and interstitial cells of Cajal, which may increase the diversity of responses and allow specialized innervation of GI muscle tissues.
- gastrointestinal motility
- interstitial cells of Cajal
- enteric nervous system
a considerable volume of work has led to a reasonable understanding of the major neuromuscular transmitters and postjunctional receptors and effectors that regulate gastrointestinal (GI) motility. Excitatory motoneurons release ACh and neurokinins, such as substance P and neurokinin A (NKA). Inhibitory motoneurons release nitric oxide (NO), vasoactive intestinal polypeptide (VIP), pituitary adenylate cyclase-activating peptide (PACAP), and possibly ATP. Most of the enteric neurotransmitters are coupled to postjunctional responses via G protein-coupled receptor pathways. NO is the black sheep in the family, because its effects are mediated by cGMP-dependent mechanisms. This article describes the G protein-coupled receptors, the G proteins linking the receptors to function, and some of the cellular effectors activated by enteric neuromuscular transmitters.
STRUCTURAL FEATURES OF G PROTEIN-COUPLED RECEPTORS
G protein-coupled receptors constitute a very large family of cell-surface receptors (for review, see Ref. 11). Well over 100 members of this family have been identified. These receptors mediate responses to a very wide variety of transcellular signaling molecules, including peptides, amino acids, small amines, and derivatives of fatty acids. In many cases, multiple subtypes of receptors exist for the same signaling molecule.
G protein-coupled receptors share many common structural features. The similarities in structure through this large family of receptors suggest a common origin in evolution. They are composed of a single amino acid strand that traverses the membrane seven times. The extracellular amino terminus and specific amino acids in the extracellular loops connecting the transmembrane domains convey specific receptor-binding characteristics (ligand-binding domain). The intracellular carboxy-terminal end and the intracellular loops contain amino acid sequences that allow the receptors to couple with specific G proteins.
G proteins are heterotrimeric (α-, β-, and γ-subunits), membrane-associated, GTP-binding proteins with GTPase activity. A large variety of G proteins have been isolated. These signaling molecules are categorized by their α-subunits, of which there are 16 distinct forms in four major families (i.e., αs, αi/αo, αq, and α12). Five isoforms of β- and at least 12 γ-subunits have also been characterized. G protein-coupled receptors typically associate with the G proteins of specific families due to sequence-specific binding sites. Although this is the rule, there are examples of receptor promiscuity in which receptors can associate with G proteins of several families. Binding of a ligand to a G protein-coupled receptor causes a conformation change in the receptor and association of the receptor to specific types of G proteins. This triggers the intracellular signaling cascade illustrated in Fig.1.
It was recently postulated that G protein-coupled receptor function is regulated by GTPase-activating proteins or GAPs (also referred to as RGS proteins, see Ref. 1). These are small proteins, first cloned from yeast, that may provide negative feedback on G protein signaling or cross talk with other receptor systems. The rate at which G protein activation of an effector is terminated depends on the intrinsic GTPase activity of the subunit. Normally slow, RGS proteins bind to subunits and enhance the rate of GTP hydrolysis, thus decreasing the effective period of G protein signaling. This is a fascinating new area of investigation, because defects in RGS expression or function could markedly increase the duration of signaling. Alternatively, overexpression of RGS proteins could effectively block signaling via G protein-coupled receptors.
MAJOR RECEPTORS AND G PROTEIN SIGNALING PATHWAYS INVOLVED IN ENTERIC NEUROMUSCULAR TRANSMISSION
Muscarinic Receptors and Coupling to Cellular Effectors
Evidence from molecular cloning studies indicates that there are separate genes encoding five muscarinic receptor subtypes. Because of the cross-reactivities of most ligands for muscarinic receptors and the expression of multiple types of receptors in single cells, it is difficult to unravel the complicated signaling pathways in native myocytes. The expression of cloned receptors and signaling elements in cell lines has greatly increased our understanding of the specific pathways activated by G protein-coupled receptors and the specific receptor-G protein interactions that mediate these effects. A caveat to this approach is that receptors have been known to exhibit promiscuity in recombinant systems and to couple with atypical G proteins.
GI muscles express the m2 and m3 subtypes of muscarinic receptors. These receptors are present in smooth muscles in a ratio of ∼80% m2 to 20% m3 (7). Receptors of the m2 type are coupled to the Gi/Gofamily of G proteins, which at a biochemical level causes inhibition of adenylyl cyclase. Reduction in cAMP production is generally excitatory in GI muscles, since many downstream effects of cAMP-dependent protein kinase (A kinase) are inhibitory (see discussion below regarding cAMP-dependent mechanisms). The specific molecular interactions that convey specificity between m2 receptors and specific subunits have been deduced through the use of mutant m2 receptors and G protein α-subunits. Normally, m2 receptors have very low affinities for Gqα. Hybrid Giα/Goα subunits in which the last five amino acids of Gqα were replaced with corresponding sequences of Giα or Goα readily associate with m2 receptors. Thus m2 receptors have a sequence that recognizes the COOH-terminal amino acids of Giα or Goα. The element responsible is a short sequence containing Val385, Thr386, Ile389, and Leu390 (“VTIL motif”) located between the sixth transmembrane domain and the i3 intracellular loop of the m2 receptor (39).
Recent evidence suggests that m2 receptors may also signal through the G protein βγ-subunit to specific members of the MAP kinase superfamily (MAP kinase and junkinase, JNK) via a phosphorylation cascade beginning with activation of the guanine nucleotide exchange protein, p21Ras, and the GTP-binding protein, Rac1 (10). Such a pathway may serve a proliferative role and stimulate growth, but the significance, if any, of this pathway in GI smooth muscles is not yet known.
Receptors of the m3 type are more clearly understood in terms of their role in smooth muscles. These receptors couple to membrane-bound phospholipase C-β (PLC-β) via pertussis toxin-sensitive Gq/11 proteins (8). Mutagenesis studies suggest that several residues at the amino and carboxy terminals of the i3 intracellular loop, together with several hydrophilic residues in the i2 loop (i.e., Ala488, Ala489, Leu492, and Ser493), provide the specificity of the m3 receptor for Gq/11(39). Activation of PLC-β resulting from ligand occupation of m3 receptors leads to increased hydrolysis of phosphatidylinositol 4,5-diphosphate (PIP2) and formation of inositol 1,4,5-trisphosphate (IP3) and diacylglycerol (DAG). These molecules each activate specific pathways that provide excitatory signals in smooth muscles; IP3 causes mobilization of Ca2+ from the sarcoplasmic reticulum. Ca2+ from stores sums with Ca2+ provided by influx mechanisms and enhances the force of contraction. DAG activates protein kinase C (PKC), which affects ion channels and may enhance the sensitivity of the contractile apparatus to Ca2+. The function served by PKC is complicated, and considerable controversy still exists regarding the precise role of this pathway in the regulation of contractility of smooth muscles (e.g., Ref. 12).
Most receptors that couple to Gq/11 display rapid, but partial, desensitization, which is evidenced by the transient nature of IP3 accumulation and the observation that preexposure to agonist causes reduction in responses when the same agonist is reapplied. Mechanisms of desensitization affecting G protein-coupled receptors include phosphorylation of receptors and receptor internalization (e.g., for review, see Ref. 2). Gq/11-coupled signaling could also be desensitized by depletion of PIP2, the substrate for IP3 synthesis. Evidence from model cell systems suggests that there is significant depletion of PIP2 within minutes of receptor activation (27).
Neurokinin Receptors and G Protein Coupling
Three neurokinin receptors (NK1, NK2, and NK3) mediate the effects of tachykinins (substance P, NKA, and NKB) in the GI tract. Postjunctional cells within the tunica muscularis express NK1 and NK2 receptors. Recent studies (32,35) with specific neurokinin receptor antibodies show expression of NK1 by interstitial cells of Cajal (ICC) and of NK2 by smooth muscle cells. If immunohistochemistry reports the expression pattern of neurokinin receptors with fidelity, then the differences in expression suggest possible specialization of innervation by tachykinin-containing excitatory motoneurons.
There are significant regions of homology between the neurokinin receptors (28). However, each shows a different order of specificity toward the tachykinins. NK1 is most potently stimulated by substance P, NK2 shows highest affinity for NKA, and NK3 has highest affinity for NK3. All three neurokinin receptors couple to pertussis toxin-insensitive G proteins, and the major signaling pathway is by Gq/11 coupling to PLC and IP3 production (23). Neurokinin receptors desensitize on repeated application of agonist, but NK1 receptors are the most strongly desensitized. The mechanisms of desensitization and receptor regulation of neurokinin receptors have been studied in considerable detail, and a recent review covers many of the important findings (2). There are several consensus sequences in the NK1 receptor for G protein receptor kinases (GRK) and PKC, particularly in the i3 intracellular loop and the carboxy terminus of the receptor. Phosphorylation of NK1 receptors leads to uncoupling from G protein subunits. Similar phosphorylation sites are not available for the NK2 receptor. The difference in regulation by phosphorylation may account for the differences in desensitization between these receptors (9). Differences in the endocytosis of receptors are unlikely to account for the different rates of desensitization. Resensitization of NK1 receptors occurs when phosphorylation reverses, and evidence suggests this occurs primarily after endocytosis and during recycling. Dephosphorylation of NK2 receptors does not appear to be required for resensitization; however, the processes regulating NK2 receptors are less well understood. The interesting process of G protein-coupled receptor regulation is now being investigated in a number of systems, and this area of investigation should yield important information about agonist responses in the GI tract.
VIP and PACAP Receptors and G Protein Coupling
Receptors for VIP and PACAP are part of a diverse family of G protein-coupled receptors that are activated by an array of large peptides. Because VIP and PACAP are coexpressed by enteric inhibitory neurons, they may both contribute to inhibitory responses (e.g., Ref.24). Contributions of the two peptides and the functional receptors expressed appear to vary between preparations, but this variability may be due to the difficulties in characterizing G protein-coupled receptors with pharmacological techniques. VIP-2 receptors are the predominant form in the GI tracts of a variety of species (38). When expressed in null cells, VIP-2 receptors couple through Gs to activate adenylyl cyclase and have equal affinity for VIP, PACAP-27, and PACAP-38. There has also been a report of a second VIP/PACAP receptor in GI muscles that couples to effectors via Giα1–2(e.g., see Ref. 25). However, the structure of this receptor has not been published. Murthy et al. (25) have suggested activation of parallel pathways by VIP and PACAP involving cAMP- and cGMP-mediated components in smooth muscle relaxation.
The observations above suggest that VIP and PACAP would generally have similar effects in GI muscles (25). Several studies, though, have described distinct effects for VIP and PACAP that appear to be mediated by different signaling pathways and different effectors. For example, Parkman and colleagues (31) reported dual effects of PACAP in guinea pig gallbladder. These appeared to be mediated by PACAP-preferring and PACAP/VIP receptors, suggesting that PACAP has effects separate from VIP. In the rat distal colon antagonists for VIP receptors, VIP-(10—28), and PACAP receptors, PACAP-(6—38), blocked different aspects of inhibitory response elicited by electrical field stimulation of neurons; VIP-(10—28) blocked responses mediated by activation of large-conductance Ca2+-activated K+ channels (BK channels) and PACAP-(6—38) blocked parallel responses due to small-conductance, apamin-sensitive Ca2+-activated K+ channels (SK channels; Ref.16). These data suggest that VIP and PACAP, which appear to be coreleased in response to enteric inhibitory nerve stimulation, couple to distinct receptors and have distinct coupling mechanisms and ionic conductances in this preparation. As described below, for SK channels to be activated by PACAP, coupling of PACAP-selective receptors through Gq/11 is suggested. PACAP I receptors have these characteristics.
ATP Receptors and G Protein Coupling
ATP has long been suggested to act as the “fast” transmitter in enteric inhibitory neurotransmission, and the hyperpolarization effects of ATP are thought to be mediated by P2Y receptors. This hypothesis is difficult to test definitively, because specific pharmacological agents to block P2Y receptors are not available. P2Y receptors form a distinct subset of G protein-coupled receptors. At least seven entities have been described. These receptors are typically coupled to phospholipase C via Gq/11, but inhibition of adenylyl cyclase has also been reported (30). Expressed P2Y receptors have been categorized pharmacologically by the rank order of potency of agonists; some P2Y receptors prefer pyrimidines to purines. Lacking good pharmacological agents to distinguish between receptor subtypes makes it very difficult to determine the active form of P2Y receptor(s) present in GI muscles. Most studies have suggested the presence of P2Y1 receptors, because of the typically higher potency of 2-methylthio-ATP over ATP, but a recent report suggests that P2Y2 receptors are dominant in gastric smooth muscle cells (26). This class of P2Y receptors has no apparent preference for ATP over UTP, raising the interesting possibility that UTP could also be a candidate for the fast inhibitory transmitter.
COUPLING OF G PROTEINS TO RESPONSES
Much has been learned about the mechanisms that mediate neural responses in postjunctional cells. There are two general levels of regulation that lead to modulation of contractile responses:1) mechanisms directed at regulating cytoplasmic Ca2+ concentration, and 2) mechanisms that regulate the sensitivity of the contractile apparatus to Ca2+. In phasic GI muscles modulation of cytoplasmic Ca2+concentration comes primarily from electrical events that regulate Ca2+ entry through L-type Ca2+ channels. Further control of cytoplasmic Ca2+ is accomplished by Ca2+ release from, and uptake into, stores within cells. Entry, release, and uptake of Ca2+ are each controlled by downstream effects of activation of G protein-coupled receptors. Mechanisms regulating the Ca2+sensitivity of the contractile apparatus also appear to be controlled by G protein-coupled receptors, but the mechanisms for this phenomenon are less fully understood than the mechanisms that regulate cytoplasmic Ca2+ concentration.
Ca2+ entry into most GI muscles is controlled by rather subtle changes in the open probability of L-type Ca2+ channels (e.g., see Ref. 33for review). The more frequently these channels open and the longer they stay open directly correlate with changes in cytoplasmic Ca2+. This mechanism is the major determinant of contractile force in phasic muscles. L-type Ca2+ channels are voltage dependent, and membrane potential appears to be the main factor regulating Ca2+ channel open probability. Excitatory transmitters enhance inward current via activation of nonselective cation channels (although examples of inhibition of outward currents are also in the literature; reviewed in Ref. 33). The increase in inward current yields depolarization: enhanced membrane potential depolarization during electrical slow waves and increased probability for firing of action potentials. These events all lead to enhanced Ca2+ channel open probability and greater Ca2+influx. Inhibitory transmitters primarily act to increase outward current via activation of several types of K+ channels. The net increase in outward current tends to hyperpolarize membrane potential, reduce the duration of electrical slow waves, and decrease the probability of action potential generation. In both cases, transmembrane potential is the main factor regulating Ca2+current (33).
Stimulation of muscarinic receptors causes activation of nonselective cation channels in GI muscles from a variety of species. This conductance has been observed and characterized by several laboratories, and several generalizations are possible. Little or none of the nonselective cation current is available in unstimulated GI muscles. The conductance is activated by receptor occupation and a pertussis toxin-sensitive G protein, suggesting coupling through Gi/Go(13). The single-channel conductance appears to be ∼20–30 pS, and the channels are voltage dependent over the physiological range of potentials. A number of ions and drugs block this conductance (e.g., Gd3+, Cd2+, Ni2+, quinine, and Fenamate); however, specific blockers have not been identified. Although the nonselective channels are permeable to Ca2+, they conduct very little Ca2+ in physiological ionic gradients. The channels are nearly equally permeable to Na+ and K+, and the current produced by the conductance reverses at ∼0 mV. Because GI muscle cells operate at potentials negative to 0 mV, the current through nonselective cation channels is inward, carried predominantly by Na+. Entry of Na+ has no direct effect on the contractile process, but the depolarization caused by entry of positive charge increases the open probability of Ca2+ channels.
The nonselective cation conductance activated by muscarinic stimulation is facilitated (as distinct from activated) by intracellular Ca2+, thus either Ca2+ entry or Ca2+ release from stores in the vicinity of nonselective cation channels augments the conductance. This property of the nonselective cation conductance might lead to positive feedback: inward current via nonselective cation channels produces depolarization, depolarization increases Ca2+ influx, and Ca2+ influx potentiates nonselective cation current. We (20) have used the term “Ca2+-induced Ca2+ entry” to describe this phenomenon. A recent report has suggested that the conductance is gated by m2 receptors (Gi/Gocoupled) and modulated by m3 receptors (3). The latter may have been due to Gq/11-coupled increase in Ca2+ release from stores; however, these experiments were performed in cells dialyzed with Ca2+ buffer to clamp intracellular Ca2+, suggesting that m3 stimulation may, in some way, cross-facilitate m2 function.
Stimulation of tachykinin receptors with NKA and substance P also activates a nonselective cation conductance in GI muscles (22). A major difference, though, was noted between the conductances activated by muscarinic and tachykinin receptor occupation, in that the latter was insensitive to intracellular Ca2+. This suggests that different types of nonselective cation channels are linked to the different G protein-coupled receptor populations. There may be specialized assembly of specific types of nonselective cation channels with certain classes of receptors. Therefore, activation of a given receptor may only activate the nonselective cation channel(s) within its unit or domain. This hypothesis is supported by the observation that the maximal currents activated by ACh and substance P are additive (22).
cAMP- and cGMP-dependent pathways provide the major inhibitory mechanisms that regulate GI muscles. A variety of inhibitory transmitters and hormones elicit responses via enhanced production of cAMP. These include norepinephrine, VIP, PACAP, adenosine (P1 receptors), and secretin. cAMP produces paradoxical effects on L-type Ca2+ currents. Generally speaking, inhibitory agents should cause less Ca2+ influx, but cAMP-dependent mechanisms enhance Ca2+ current in smooth muscle myocytes (e.g., see Ref. 18). This effect is likely to be due to phosphorylation by cAMP-dependent protein kinase A, which is known to increase Ca2+ current in the heart and other tissues (see Ref. 37). This trend, which would tend to make agonists that utilize cAMP-dependent mechanisms excitatory, is counteracted in GI muscles by concomitant activation of a variety of K+ channels. Activation of K+ channels tends to limit excitability and decrease the open probability of Ca2+ channels. The types of K+ channels that are expressed and utilized in cAMP-dependent responses vary in different species and in different parts of the GI tract. At a minimum, it is known that ATP-dependent K+ channels (40), BK channels (5), 15-pS delayed rectifier K+ channels, and 90-pS voltage-dependent K+ channels (19) can be activated in GI muscle cells by cAMP-dependent mechanisms. Other channels, such as an inwardly rectifying K+ channel, may also be regulated by cAMP-dependent mechanisms and contribute to the changes in resting potential noted in response to agonists that enhance cAMP. The properties of these K+ channels (i.e., dependence on voltage, metabolism, and intracellular Ca2+) vary considerably, so each class of channel may contribute to different parts of the response to an agonist. For example, the voltage dependence of some channels may cause them to be activated at the resting potential. A small increase in the open probability of these channels can dramatically affect (either hyperpolarize or stabilize) resting potential. The voltage dependence of activation of other channels occurs over a more positive range of potentials. These channels are activated during slow waves and action potentials, and increases in their open probabilities can reduce the amplitude of slow waves and block action potential generation. BK channels are both voltage and Ca2+dependent, so these channels may exert effects primarily when muscles have been prestimulated with excitatory agonists.
Electrical mechanisms are supplemented and further regulated by release and uptake of Ca2+ from stores. Many of the conductances discussed above are Ca2+ dependent (for review, see Ref. 6), so subtle changes in Ca2+near the membrane can significantly influence open probability. Recent studies (29) have demonstrated fundamental Ca2+ release events (Ca2+ sparks) that can regulate the open probabilities of Ca2+-dependent channels. Altering uptake and release mechanisms offers an important regulatory control on spark activity and the open probabilities of these channels. Besides regulating membrane proteins, release of Ca2+ can supplement cytoplasmic Ca2+ levels. This results from G protein-coupled receptors that couple through Gq/11 and the production of IP3.
An interesting new twist on Gq/11coupling is highlighted by the actions of ATP (via P2Y receptors) and possibly those of PACAP. ATP, a putative inhibitory neurotransmitter, is thought to mediate the fast hyperpolarization response characteristic of enteric inhibitory neural responses in most species. These responses appear to be mediated through P2Y receptors that are linked to PLC-β. As described above, this pathway leads to Ca2+ release, which is difficult to imagine as an inhibitory signal. However, it should be noted that the ion channels activated by P2Y agonists in GI muscles are small-conductance Ca2+-activated K+ channels (SK channels) (see Ref. 17). These channels are activated on release of Ca2+ from intracellular stores. This has prompted the hypothesis that P2Y receptors are coupled to highly localized Ca2+ release (a sparklike event mediated or initiated by IP3) that influences membrane conductances close to the site of release without raising cytoplasmic Ca2+.
Questions remain about the mechanisms of PACAP responses in GI muscles. How does this peptide activate the same apamin-sensitive conductance as ATP (24)? Activation of SK channels by ATP is linked to Ca2+ release (see above), but the standard view of PACAP receptors in GI muscles (i.e., PACAP receptors coupled to activation of adenylyl cyclase via Gs) does not explain activation of apamin-sensitive channels. Other agonists that raise cAMP levels do not activate SK channels (unpublished observations). Most of the splice variants for PACAP I receptors have been shown to activate both adenylyl cyclase and PLC (34). Expression of certain splice variants of PACAP I receptors that permit coupling to PLC via Gq in parallel with activation of adenylyl cyclase might produce Ca2+ release events that could activate SK channels in GI muscles. Another explanation might come from recent studies (36) on adrenal chromaffin cells in which PACAP was shown to release Ca2+ via a caffeine/ryanodine-sensitive mechanism. These data suggest that multiple mechanisms may link PACAP to localized Ca2+ release and activation of SK channels.
A final level of regulation comes from dual modulation of the Ca2+ sensitivity of the contractile apparatus. The sensitivity is reduced by cAMP- and cGMP-dependent mechanisms and raised by excitatory agonists, such as ACh and the neurokinins (reviewed in Ref. 33). The mechanism by which Ca2+ sensitivity is altered is not entirely clear, but it appears likely to involve regulation of the thin filament elements in smooth muscles (12). Thus stimulation of G protein-coupled receptors that converge on cAMP or IP3 and DAG results in highly integrated responses composed of the actions of several effectors that regulate the electrophysiology, Ca2+ metabolism, and contractile apparatus of GI muscle cells.
ICC AS PRIMARY RECEPTOR/TRANSDUCER CELLS
Recent evidence suggests that ICC mediate inhibitory neurotransmission in GI muscles (4). ICC are tightly associated with enteric motoneurons, and synaptic-like structures can be readily observed between varicose nerve terminals and ICC in areas of dense innervation, such as the deep muscular plexus in the small intestine (S. M. Ward and X.-Y. Wang, unpublished observations). ICC may express specialized receptor mechanisms and/or effector proteins that transduce inhibitory neural inputs. These cells are connected to smooth muscle cells via gap junctions, so electrical responses of ICC must be considered when discussing the responsiveness of smooth muscle tissues to agonists. Studies of isolated and cultured ICC have shown that these cells are responsive to a number of neurotransmitters that utilize G protein-coupled receptors. At the present time, it is unclear what role ICC have in excitatory neurotransmission or responses to hormones and paracrine substances, but several studies have demonstrated specialized receptor populations (such as NK1and somatostatin 2A; Refs. 32, 35) and ionic conductances (effector proteins) in ICC (21). Specialized mechanisms in ICC to receive and transduce chemical signals may be a general phenomenon in the GI tract, and numerous tissue responses could be mediated through this important class of cells. Specialized distribution of G protein-coupled receptors in smooth muscles and ICC may help explain some of the complications that have arisen from pharmacological studies on GI tissues and heterogeneous cell populations.
I am grateful to Dr. C. William Shuttleworth for reading and commenting on the manuscript.
Address reprint requests to K. M. Sanders.
↵* Fourth in a series of invited articles on G Protein-Coupled Receptors in Gastrointestinal Physiology.
This study was supported by National Institute of Diabetes and Digestive and Kidney Diseases Grants PO1 DK-41315 and RO1 DK-40569.
- Copyright © 1998 the American Physiological Society