Abbreviations used : ACh, acetylcholine ; CFTR, cystic fibrosis transmembrane conductance regulator ; GIRK, G protein-gated inwardly rectifying K+ channel ; hK, high K+ ; 5-HT, 5-hydroxytryptamine ; i3 loop, third intracellular loop ; mAChR, muscarinic acetylcholine receptor ; m1-m5 and hm1-hm5, five mAChR subtypes and five human mAChR subtypes, respectively ; NMS, N-methylscopolamine.
Address correspondence and reprint requests to Dr. N. J. Cho at Department of Biochemistry, College of Natural Sciences, Chungbuk National University, Cheongju, 361-763, Korea.
Abstract : We have isolated a cDNA clone from the nematode Caenorhabditis elegans that encodes a protein of greatest sequence similarity to muscarinic acetylcholine receptors. This gene codes for a polypeptide of 682 amino acids containing seven putative transmembrane domains. The amino acid identities, excluding a highly variable middle portion of the third intracellular loop, to the human m1-m5 receptors are 28-34%. When this cloned receptor was coexpressed with a G protein-gated inwardly rectifying K+ channel (GIRK1) in Xenopus oocyte, acetylcholine was able to elicit the GIRK current. This acetylcholine-induced current was substantially inhibited by the muscarinic antagonist atropine in a reversible manner. However, another muscarinic agonist oxotremorine and antagonists scopolamine and pirenzepine had little or negligible effects on this receptor. Taken together, these results suggest that the cloned gene encodes a G protein-linked acetylcholine receptor that is most similar to but pharmacologically distinct from muscarinic acetylcholine receptors.
Acetylcholine (ACh) is believed to play a key role in various nervous functions such as contraction of skeletal muscles, emotion, perception, cognition, learning, and memory. ACh has been known to exert its effects by binding two kinds of postsynaptic integral membrane proteins, termed nicotinic and muscarinic receptors. Nicotinic receptors are cation ion channels, whereas muscarinic receptors mediate the regulation of ion channels by activating signal-transducing G proteins and intracellular effector systems.
Muscarinic acetylcholine receptors (mAChRs), like other members of the G protein-linked receptor superfamily, have a predicted structure of seven transmembrane domains. Since Kubo et al. (1986) reported the cloning of a cDNA for pig mAChR, mAChR genes have been isolated from a variety of organisms including humans. In mammals, five mAChR subtypes (m1-m5) have been identified (Bonner et al., 1987, 1988). The biochemical responses mediated by the five mAChR subtypes are known to be different. The m1, m3, and m5 subtypes are shown to preferentially stimulate phosphatidylinositol metabolism by pertussis toxin-insensitive mechanisms, whereas the m2 and m4 subtypes preferentially inhibit adenylyl cyclase by pertussis toxin-sensitive mechanisms (Peralta et al., 1988 ; Jones et al., 1992). The amino acid sequence comparison between the receptor subtypes reveals a substantial variation in the third intracellular loop (i3 loop) region, which is considered to be the site of G protein recognition. Thus, it is believed that the receptor subtypes mediate different biochemical responses via coupling to different G proteins. Electrophysiological studies also showed that mAChR subtypes modulate different kinds of ion channels. For instance, the m1 subtype, when expressed in Xenopus oocyte, activates a Ca2+-dependent Cl- channel, whereas the m2 subtype regulates a K+ channel that is directly stimulated by a G protein (Kubo et al., 1986, 1993 ; Fukuda et al., 1987).
We have begun to study the cholinergic system, using the nematode Caenorhabditis elegans as an experimental animal. C. elegans is suitable for this study because its nervous system is very simple and because a variety of genetic and molecular biological techniques are readily applicable. Previous studies have shown that C. elegans mAChRs bind N-[3H]methylscopolamine ([3H]NMS) with high affinity and that ligand binding properties of C. elegans mAChRs are generally similar to those of mammalian counterparts (Culotti and Klein, 1983 ; You et al., 1996). In this study, we identified a novel G proteinlinked ACh receptor in C. elegans, which is different from the known mAChRs in ligand binding properties.
MATERIALS AND METHODS
ACh chloride, carbamylcholine chloride (carbachol), atropine sulfate, (-)-scopolamine hydrochloride, oxotremorine free base, pirenzepine dihydrochloride, 5-hydroxytryptamine (5-HT) creatinine sulfate, and (±)-isoproterenol hydrochloride were purchased from Sigma.
Isolation of the C. elegans G protein-linked ACh receptor cDNA clone
C. elegans (N2, wild-type strain) was grown either on nematode growth medium plates (Brenner, 1974) or in liquid culture (Sulston and Brenner, 1974) using E. coli strain OP50 as a food source.
The cDNA clone was obtained by PCR, using single-stranded DNA, prepared from a C. elegans mixed-stage cDNA library (Stratagene), as a template. The primers were designed based on the genomic sequence of a C. elegans cosmid clone in the databank (C15B12 ; accession no. U23529) (Wilson et al., 1994). PCR was performed in a total volume of 50 μl containing 1.5 mM MgCl2, 1 × reaction buffer (Amersham), 200 μM of each dNTP, 2.5 U Taq polymerase (Amersham), 1 μM forward primer (MF no. 1), 1 μM reverse primer (MR no. 1), and 20 ng of template DNA. The PCR condition was as follows : an initial denaturation at 95°C for 5 min, 30 cycles of amplification at 95°C for 1 min, 55°C for 2 min, and 72°C for 3 min, and a final extension at 72°C for 7 min. The PCR products were ligated into pCRII vector (Invitrogen) and the nucleotide sequences of the cDNA clones were determined by the dideoxynucleotide chain termination method, using the Cy5 AutoRead sequencing kit and ALFexpress DNA sequencer (Pharmacia). We identified one clone (pCMR10), which contains the identical nucleotide sequence to the corresponding genomic sequence (Wilson et al., 1994). This gene was named gar-1 (for G protein-linked ACh receptor). Primer sequences are as follows : MF no. 1, 5′ -ATCGATTGCTTCTCATCTTG-3′ ; MR no. 1, 5′-ATTTTGGATCCAGATCTATGCAACT-3′ (the shaded sequence G was changed from the sequence A to create a BamHI site).
Genomic Southern blot analysis
Genomic DNA was isolated from C. elegans as described by Emmons et al., (1979). The genomic DNA (5 μg) was digested with restriction enzymes, electrophoresed through a 0.8% (wt/vol) agarose gel, and transferred to a nitrocellulose filter by using 20 × saline-sodium citrate as described by Sambrook et al. (1989). The blot was hybridized to a 32P-labeled C. elegans G protein-linked ACh receptor probe. The incubation was performed at 37°C for 16 h in a hybridization buffer [6×saline-sodium citrate, 30% (vol/vol) formamide, 0.5% (wt/vol) sodium dodecyl sulfate, 5× Denhardt's solution, and 0.1 mg/ml salmon sperm DNA]. The blot was washed twice (10 min each) at 37°C in 2× saline-sodium citrate/0.5% (wt/vol) sodium dodecyl sulfate and twice (20 min each) at 50°C in 0.2× saline-sodium citrate/0.1% (wt/vol) sodium dodecyl sulfate. The probe DNA was labeled with [32P]dCTP (3,000 Ci/mmol, Amersham), using a random primed DNA labeling kit (Boehringer Mannheim).
Preparation of stage-specific RNA and RT-PCR
Gravid worms were treated with alkaline hypochlorite solution (Emmons et al., 1979) and hypochlorite-resistant embryos were then hatched in M9 buffer overnight at 20°C. Hatched L1 larvae were placed on nematode growth medium plates and allowed to develop to various development stages before harvest for RNA preparation. Harvested worms were frozen in liquid nitrogen and ground to a powder using a mortar and pestle. RNA was prepared from the frozen worm powder by use of the Tri reagent (Sigma) following the procedure recommended by the manufacturer.
RT-PCR was performed by using the RNA PCR kit (Takara). The first strand of cDNA was synthesized from 0.9 μg of each stage-specific total RNA in a 5-μl solution containing 5 mM MgCl2, 10 mM Tris-HCl (pH 8.3), 50 mM KC1, 1 mM of each dNTP, 5 U RNase inhibitor, 0.125 μM oligo(dT) primer, and 1.25 U of avian myeloblastosis virus (AMV) reverse transcriptase XL. The incubation was performed at 30°C for 10 min, followed by 42°C for 1 h. The reaction was terminated by heating at 99°C for 5 min. For PCR, 3 μl of the RT reaction mixture was added to a 12-μl solution containing 1 × Ex Taq buffer, 0.25 μM forward primer (MF no. 1), 0.25 μM reverse primer (MR no. 1), and 0.075 U of Takara Ex Taq polymerase. The PCR condition was as follows : an initial denaturation at 94°C for 2 min, 41 cycles of amplification at 94°C for 30 s, 57°C for 30 s, and 72°C for 2 min, and a final extension at 72°C for 8 min. The PCR product (2 μl) was analyzed by gel electrophoresis.
cRNA preparation and expression in Xenopus oocytes
The cRNAs for the C. elegans G protein-linked ACh receptor, pig cardiac mAChR (Fukuda et al., 1987), human β2-adrenergic receptor (Kobilka et al., 1987), mouse 5-HT2c (previously termed 5-HT1c) receptor (Lubbert et al., 1987), human cystic fibrosis transmembrane conductance regulator (CFTR) (Bear et al., 1991), and rat G protein-gated inwardly rectifying K+ channel (GIRK1) (Dascal et al., 1993) were prepared by in vitro transcription, using the cRNA Megascript kit (Ambion) (Uezono et al., 1993). The resulting cRNAs were identified on a gel and quantified spectrophotometrically.
Stages IV and V oocytes were removed from female Xenopus laevis (Xenopus I Inc.) and then defolliculated by incubation for 1.5-2 h with 2 mg/ml collagenase (type IA, Sigma) in OR-Mg solution (82 mM NaCl, 20 mM MgCl2, 2 mM KCl, 5 mM HEPES, pH 7.5). Oocytes were microinjected with cRNAs and incubated for 2-3 days at 18°C in ND96 solution (96 mM NaCl, 2 mM KCl, 1.8 mM CaCl2, 1 mM MgCl2, 5 mM HEPES, pH 7.5) supplemented with 2.5 mM sodium pyruvate and 50 μg/ml gentamicin.
Two-electrode voltage clamp was performed with a GeneClamp 500 amplifier (Axon Instruments). The voltage electrode and the current electrode were pulled, using a horizontal puller (Sutter Instruments), to have resistances of 1-5 MΩ. The electrodes were filled with 3 M KCl. The oocytes were placed in a chamber perfused with Ca2+-free ND96. The holding potential was set at -80 mV. In the case of recording GIRK current, the solution was changed to high-K+ (hK) solution (96 mM KCl, 2 mM NaCl, 1.8 mM CaCl2, 1 mM MgCl2, 5 mM HEPES, pH 7.5). Drugs dissolved in Ca2+-free ND96 or hK solution were applied to the recording chamber via bath perfusion lines. The data were collected on a chart recorder and stored on video cassette recorder tape through Digidata (Instrutech) for later analysis.
In an attempt to isolate a cDNA clone encoding C. elegans mAChR or its homologue, we performed a PCR, using C. elegans cDNA library as a template. The primers were designed based on the sequence information of a possible mAChR gene reported by the C. elegans genome sequencing consortium (Wilson et al., 1994) (see Materials and Methods). A cDNA clone, which contains an open reading frame of 682 amino acids, was identified by restriction enzyme mapping and sequence analysis (Fig. 1). Comparison of this cDNA sequence with the corresponding genomic sequence indicated that the C. elegans gene contains 16 introns in the coding region. Seven putative transmembrane domains, which is a characteristic feature of G protein-linked receptors, could be identified by hydropathy profile (Fig. 1). Other features of G protein-linked receptors were also found in this protein, for example, a potential N-glycosylation site (Asn3) and two cysteine residues engaged in a disulfide linkage in the second and third extracellular segments (Fig. 2).
Database searches indicated that the cloned receptor is most similar to mAChRs. We compared the amino acid sequence of this receptor with those of five human mAChR subtypes (hm1-hm5 ; Fig. 2). Excluding the highly variable i3 loop except 20 amino acids on each side, the overall amino acid sequence identities to hm1, hm2, hm3, hm4, and hm5 are estimated to be 32, 34, 28, 32, and 32%, respectively. The amino acid sequence similarities of the same region to hm1, hm2, hm3, hm4, and hm5 are about 52, 52, 46, 51, and 50%, respectively. Site-directed mutagenesis analyses of mAChRs have identified several conserved amino acid residues that are important for muscarinic ligand binding (Wess et al., 1995). The sequence comparison revealed that the C. elegans receptor lacks some of these residues (Fig. 2). This sequence difference may partly explain the observation that the C. elegans receptor and mAChRs displayed distinct ligand binding properties (see Fig. 6).
To examine whether C. elegans contains multiple genes for the cloned receptor, genomic Southern blot analysis was performed using the cloned cDNA as a probe at moderate stringency. All the strongly hybridizing bands were of expected size (Fig. 3), suggesting that this receptor may be encoded by a single gene. A few weakly hybridizing bands (an ~6-kb EcoRI fragment, an ~3.5-kb HindIII fragment, and an ~3-kb EcoRV fragment) were also observed that may represent either other subtypes of the receptor or other closely related receptors.
We performed RT-PCR experiments to examine at which stage the cloned receptor gene transcript is present. A band of expected size (~2 kb) was observed at all developmental stages (Fig. 4), including the dauer stage (data not shown), indicating that the gene transcript is present throughout the developmental stages.
To determine whether the cloned cDNA encodes a functional receptor, we expressed the cDNA in Xenopus oocytes. The receptor cRNA was injected into oocytes either alone or together with GIRK1 cRNA. Replacement of Ca2+-free ND96 solution by hK solution elicited an inward current (IhK). Amplitude of IhK in water-injected oocytes (57.0 ± 4.6 nA, n = 5) (Fig. 5A) was not significantly different from that in oocytes injected with the receptor cRNA alone (49.0 ± 4.8 nA, n = 5) (Fig. 5B). The current amplitude was larger in oocytes injected with GIRK1 alone (120.8 ± 9.4 nA, n = 15) (Fig. 5C). When oocytes injected with both the receptor and GIRK1 cRNAs were treated with 1 μM ACh, an additional increase (295.4 ± 56.3 nA, n = 21) in the inward current was observed (Fig. 5D). In contrast, water-injected oocytes and oocytes injected with the receptor cRNA alone or GIRK1 cRNA alone were not affected by treatment with 1 μM ACh. These results demonstrate that the cloned cDNA encodes a functional ACh receptor. The results also suggest that the cloned receptor couples to a G protein, possibly of the Gi/Go class, as the GIRK1 channel is known to be regulated by many Gi/Go-linked receptors.
We compared the pharmacological properties of the C. elegans G protein-linked ACh receptor with those of pig cardiac mAChR (m2), which is also known to couple to the activation of the GIRK1 channel. Oocytes expressing pig cardiac mAChR and GIRK1 produced the inward current (415.0 ± 79.3 nA, n = 9) by treatment with 1 μM ACh. The muscarinic antagonist atropine (1 μM) decreased the ACh-induced GIRK current mediated by the C. elegans receptor by 76.3 ± 6.2% (n = 13) in a reversible manner (Fig. 5D), whereas atropine blocked most of the current (90.2 ± 7.5%, n = 3) mediated by the pig cardiac mAChR (data not shown). Other potent antagonists, scopolamine (1 μM) and pirenzepine (1 μM), reduced the current mediated by the C. elegans receptor by only 15.0 ± 7.0% (n = 5) and 30.5 ± 16.0% (n = 3), respectively (Fig. 6A2 and B2). By contrast, they were able to reduce the current mediated by the pig cardiac mAChR by 97.6 ± 2.4% (n = 3) and 83.8 ± 6.0% (n = 3), respectively (Fig. 6A1 and B1). Although carbachol (1 μM), an ACh analogue, could induce the GIRK current in the oocytes expressing the C. elegans receptor (76.0 ± 9.9 nA, n = 5) (data not shown), the potent muscarinic agonist oxotremorine (1 μM) failed to elicit any significant response (n = 4) (Fig. 6C2). These data show that the cloned receptor is pharmacologically different from the pig cardiac mAChR and other previously known mAChRs.
We next explored the possibility that the cloned C. elegans receptor couples to other G proteins. A transient endogenous Ca2+-activated C1- current (1.24 ± 0.23 μA, n = 3) was generated, presumably via Go/Gq, by activation of 5-HT2C receptor (Fig. 7A1). By contrast, the cloned C. elegans receptor did not produce the current (n = 6) (Fig. 7A2), suggesting that the cloned receptor does not couple to Go/Gq in Xenopus oocyte. In other experiments, the receptor cRNA was coinjected into oocytes with CFTR cRNA to see whether the receptor couples to Gs. These oocytes did not show any response to ACh (n = 10) (Fig. 7B2), whereas oocytes injected with both β2-adrenergic receptor cRNA and CFTR cRNA produced the CFTR current by treatment with isoproterenol (662.0 ± 86.0 nA, n = 5) (Fig. 7B1). These results imply that the cloned receptor does not couple to Gs.
We have isolated a cDNA clone from C. elegans that encodes a G protein-linked ACh receptor. The isolated cDNA is presumed to synthesize a polypeptide that contains seven putative transmembrane domains, which is a characteristic feature of the G protein-linked receptor superfamily. When expressed in Xenopus oocytes with GIRK1, this protein evoked the GIRK current in response to ACh. Thus, this receptor is functionally similar to mAChRs. Database searches also indicated that this C. elegans protein is more homologous to mAChRs than any other known proteins. However, this protein proved different in drug specificity from mAChRs. Scopolamine (1 μM) reduced the ACh-induced activation of the C. elegans receptor by only 15%, whereas it almost completely blocked that of pig cardiac mAChR. In addition, oxotremorine (1 μM) caused little, if any, activation of the C. elegans receptor. Conceivably, this C. elegans receptor may represent an ancient form of mAChR or a novel type of ACh receptor.
Nucleotide sequence analysis showed that the C. elegans G protein-linked ACh receptor gene contains 16 introns. No intron has been found in the coding region of mammalian mAChR genes, although Drosophila mAChR gene has been reported to have three introns in the i3 loop region (Shapiro et al., 1989). This observation is in contrast to the general tendency that the genes of lower organisms have shorter and fewer introns than those of higher organisms. In the evolution of the cholinergic nervous system of higher organisms, introns may have been removed from the ACh receptor genes.
We performed electrophysiological recording assays using the Xenopus oocyte expression system to examine the coupling specificity of the C. elegans receptor to various G proteins. In Xenopus oocytes, an endogenous Cl- channel is activated by Ca2+ released from the internal stores by inositol 1,4,5-trisphosphate when Go/Gq-mediated phosphatidylinositol metabolism is stimulated by receptors such as 5-HT2C receptor (Lubbert et al., 1987 ; Quick et al., 1994). CFTR is a Cl- channel that is regulated by cyclic AMP-dependent protein kinase, and activation of Gs-linked receptors induces the CFTR current (Uezono et al., 1993). GIRK1 channel is known to be activated by Gi/Go-linked receptors (Dascal et al., 1993). Therefore, if CFTR or GIRK1 cRNA were introduced with a receptor cRNA into the oocytes, the coupling specificity of the receptor to G proteins could be determined. Our electrophysiological data showed that the activation by ACh of the cloned receptor in the oocytes generates only GIRK current, not CFTR or endogenous Ca2+-activated Cl- current, suggesting that the cloned receptor couples to Gi, not to Gs or to Go/Gq.
Previous ligand binding studies indicated the presence of high-affinity [3H]NMS binding sites in C. elegans (Culotti and Klein, 1983 ; You et al., 1996). We asked whether the cloned receptor binds [3H]NMS with high affinity. However, we failed to detect significant levels of [3H]NMS binding sites when the cloned receptor gene was expressed in heterologous systems such as COS-1 cells, CHO-K1 cells, and Xenopus oocytes. These results are consistent with the electrophysiological data showing the negligible effect of scopolamine on the C. elegans receptor. Thus, it is likely that the gene responsible for the high-affinity [3H]NMS binding sites in C. elegans is distinct from the one we cloned in this study. Amino acid sequence comparison revealed that each of the five mAChR subtypes is <35% identical to the C. elegans receptor, further suggesting that the C. elegans receptor does not belong to any known subtype of mAChRs and, therefore, may constitute a new class of G protein-linked ACh receptors.
In summary, we have identified a novel G protein-linked ACh receptor that behaves differently from conventional mAChRs. The finding of the G protein-linked ACh receptor in this study may indicate that ACh receptor system is more diverse than we understand at present. Further study on this receptor is needed to establish its functional role in the cholinergic nervous system, and it will be of considerable interest to see whether this receptor homologue is also present in other organisms.