NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Riddle DL, Blumenthal T, Meyer BJ, et al., editors. C. elegans II. 2nd edition. Cold Spring Harbor (NY): Cold Spring Harbor Laboratory Press; 1997.

Cover of C. elegans II

C. elegans II. 2nd edition.

Show details

Section IINeurotransmitter Metabolism and Function

The synthesis, packaging, release, and re-uptake of neurotransmitter molecules by neurons are highly regulated (Fig. 3). Most neurotransmitters are synthesized from common cellular metabolites by enzymes expressed specifically in neurons using the transmitter. In addition to classical transmitters such as acetylcholine, serotonin, and GABA, a few amino acids are neurotransmitters. Neurotransmitters are loaded into synaptic vesicles by vesicular transporters using a proton gradient as the source of energy. Distinct transporters are responsible for loading different transmitters. After the neurotransmitter is released by vesicular fusion, it diffuses across the synaptic cleft separating the pre- and postsynaptic cells and binds to receptors on the postsynaptic cell. Such receptors are highly specific for transmitters, and their expression in specific cells determines the response of the cell to the chemical signal. Receptors influence electrical activity in the postsynaptic cell either directly by permitting selective entry of ions or indirectly by activating second-messenger pathways. The signaling event is usually terminated by re-uptake of the transmitter from the synaptic cleft by plasma membrane transporters.

Figure 3. (A) Steps of the synaptic release cycle.

Figure 3

(A) Steps of the synaptic release cycle. Synaptic vesicles (or vesicle precursors) are synthesized in the soma and transported to synaptic terminals by axonal transport. Vesicles are loaded with neurotransmitter (more...)

Despite its simple anatomy, the C. elegans nervous system uses an array of classical neurotransmitters which approaches the complexity of vertebrate nervous systems. In this section, we review the molecules regulating the synthesis, packaging, and re-uptake of the major neurotransmitters in C. elegans, as well as the genetic and molecular analyses of transmitter receptors. We also briefly discuss neuropeptides in C. elegans, although we do not discuss the distinct mechanisms used in the biosynthesis, packaging, and release of these molecules. In many sections, we discuss results obtained using the large parasitic nematode Ascaris, which has the same number and organization of motor neurons in the ventral nerve cord as C. elegans (Stretton et al. 1978). These Ascaris studies provide electrophysiological data not yet available in C. elegans.

A. Acetylcholine

Acetylcholine appears to be the primary excitatory neurotransmitter controlling motor functions in C. elegans. Although physiological data are lacking, evidence for acetylcholine function in C. elegans is quite compelling and comes from pharmacological studies using acetylcholine agonists and antagonists (Lewis et al. 1980b), direct measurement of the presence of the transmitter in extracts (Hosono et al. 1987; Hosono and Kamiya 1991; Nguyen et al. 1995), measurement and characterization of the enzymes of acetylcholine synthesis and degradation (Johnson and Russell 1983; Kolson and Russell 1985a; Rand and Russell 1985a), and genetic analysis of mutants defective in acetylcholine synthesis (Rand and Russell 1984; Hosono et al. 1985).

There are several noteworthy aspects of acetylcholine metabolism in C. elegans. First, acetylcholine is the only neurotransmitter so far identified in C. elegans that is essential for viability. Animals totally deficient for acetylcholine synthesis ( cha-1 mutants, see below) are inviable, whereas animals deficient in GABA, serotonin, or dopamine are viable. In addition, the synthesis of the transmitter and its loading into synaptic vesicles are controlled by a novel type of eukaryotic operon with a novel nested structure (discussed below). This type of gene structure appears to be a general feature of cholinergic regulation in mammals as well (Bejanin et al. 1994; Erickson et al. 1994).

1. Synthesis and Vesicular Transport

Acetylcholine is synthesized by choline acetyltransferase (ChAT), which is encoded by the cha-1 gene. The C. elegans protein is 36% identical to pig ChAT and 34% identical to Drosophila ChAT (Alfonso et al. 1994a). In mammals, the enzymes that synthesize neurotransmitters are often associated with synaptic vesicles (Kuhn et al. 1990; Bon et al. 1991; Carroll 1994), and biochemical and immunochemical studies in C. elegans suggest that at least part of the ChAT activity is membrane-associated (Rand and Russell 1985a; J. Duerr, pers. comm.). Viable cha-1 mutants are small, slow growing, and uncoordinated (coily and jerky going backward; Rand and Russell 1984; Hosono et al. 1985). They also have slow pharyngeal pumping and a slow irregular defecation cycle (Thomas 1990; Avery 1993a) and are resistant to inhibitors of acetylcholinesterase (Rand and Russell 1984). Rare lethal alleles of cha-1 (e.g., m324) represent the null phenotype; animals homozygous for such alleles are able to hatch, but they can barely move or feed and die as small, shrunken L1 larvae (Rand 1989; Avery and Horvitz 1990; Alfonso et al. 1994a). The temperature effects of some cha-1 alleles make them particularly useful: cn101 homozygotes are temperature-sensitive ts-Unc and p1182 is ts-lethal (Hosono et al. 1985; Rand 1989).

Neurotransmitters are transported into synaptic vesicles by specific transporter molecules. The C. elegans unc-17 gene was shown to encode a synaptic vesicle acetylcholine transporter (Alfonso et al. 1993). unc-17 encodes a predicted 58-kD hydrophobic protein with 12 transmembrane domains and homology with two rat synaptic vesicle dopamine transporters (Erickson et al. 1992). Using the unc-17 cDNA, Varoqui et al. (1994) were able to identify a homolog from Torpedo, and a Torpedo clone was then used to identify rat and human homologs (Erickson et al. 1994). Transfected mammalian cells expressing either the C. elegans unc-17 gene or the Torpedo, rat, or human VAChT genes make a protein that binds a specific inhibitor of VAChT (vesamicol). The transfected rat gene was able to mediate specific vesamicol-inhibitable accumulation of acetylcholine. These results confirm that these homologs all encode acetylcholine transporters.

Both viable and lethal alleles of unc-17 have been identified, and they lead to phenotypes similar to those of the cha-1 mutants described above (Brenner 1974; Rand and Russell 1984; Alfonso et al. 1993). Thus, a total deficiency for the acetylcholine transporter seems to have the same effect as a total deficiency for acetylcholine synthesis, demonstrating that the transporter is essential for neural function and for survival. In addition, antibody staining of the developmentally arrested cha-1 and unc-17 hatchees revealed a superficially normal nervous system (J. Duerr, pers. comm.). It therefore appears that most (or perhaps all) of the processes required for proper neural development can occur in the complete absence of acetylcholine or cholinergic function.

2. A Cholinergic Operon with Conserved Structure

Molecular analysis indicated that the cha-1 and unc-17 mRNAs were derived by alternative splicing of a common precursor (Fig. 4). The two genes use a common 5′-untranslated exon; the remainder of the unc-17 gene is nested within the long first intron of the cha-1 gene (Alfonso et al. 1994b). Thus, the sequential steps of acetylcholine synthesis and vesicle loading are encoded by different genes within a single, complex transcription unit. This same genomic organization has also been shown in mammals: The human (and rat) VAChT gene is nested within the long first exon of the ChAT structural gene (Fig. 4) (Bejanin et al. 1994; Erickson et al. 1994). Although there are some clear differences between the nematode and mammalian genes (e.g., the number of introns in ChAT and in VAChT differ from nematodes to mammals, and the mammalian gene may have additional promoters not present or not yet identified in C. elegans), the overall similarity is striking and suggests that the gene structure is critical for function and/or regulation.

Figure 4. The structure of the cholinergic gene locus is conserved from C.

Figure 4

The structure of the cholinergic gene locus is conserved from C. elegans to mammals. In C. elegans, the unc-17 (more...)

3. Cholinergic Neurons

The primary tools for identifying cholinergic neurons in C. elegans have been antibodies to ChAT and UNC–17. The cellular expression pattern of these two proteins is virtually identical (J. Duerr, pers. comm.). Immunoreactivity to both proteins is observed primarily in a punctate staining pattern in synaptic regions (Alfonso et al. 1993; J. Duerr, pers. comm.), as expected for synaptic or synaptic-vesicle-associated proteins. In addition, the anti-ChAT stain also appears weakly in nonsynaptic regions of cells (J. Duerr, pers. comm.). Transgenic animals that overexpress cha-1 and unc-17 have increased synaptic staining and also have enough staining in cell bodies to confirm the identification of cholinergic neurons.

Almost all of the ChAT-positive cells appear to be motor neurons. Strongly staining cells include six of the eight classes of ventral cord motor neurons (VA, VB, VC, DA, DB, AS), three types of pharyngeal motor neurons (M1, M2, M5), and putative sublateral motor neurons (SAA, SAB, SIA, SIB, SMB, SMD) (J. Duerr, pers. comm.). A few other neurons also appear to be cholinergic; these include ALN, PLN, and SDQ, as well as some cells yet to be identified (see Appendix 2) (J. Duerr, pers. comm.). These results are in agreement with published Ascaris enzyme and physiology data for ventral nerve cord motor neurons. The Ascaris counterparts of the DA, DB, and AS cells contain ChAT, and their excitatory output is blocked by cholinergic blockers, whereas the DD and VD counterparts do not contain ChAT and their output is blocked by GABAergic, but not cholinergic, blockers (Johnson and Stretton 1985; Segerberg and Stretton 1993).

4. Acetylcholinesterase

Unlike other neurotransmitters, whose synaptic action is terminated by rapid re-uptake mechanisms, acetylcholine is hydrolyzed in the synaptic cleft by the enzyme acetylcholinesterase (AChE). Three classes of AChE activity (designated A, B, and C) have been identified, partially purified, and characterized from C. elegans. The classes have distinct kinetic properties, substrate specificities, and inhibitor sensitivities (Johnson and Russell 1983; Kolson and Russell 1985b) and are controlled by three unlinked genes, ace-1 X, ace-2 I, and ace-3 II (Culotti et al. 1981; Johnson et al. 1981, 1988; Kolson and Russell 1985a). ace-1 encodes the class-A enzyme which is 42% identical to AChE from Torpedo and humans, 41% identical to human butyryl~cholinesterase, and 35% identical to Drosophila AChE (Johnson et al. 1981; Arpagaus et al. 1994). ace-2 mutants lack class-B activity and ace-3 mutants lack class-C activity; these genes are presumed to encode the respective AChE classes (Culotti et al. 1981; Johnson et al. 1988). The existence of three C. elegans AChE genes is in contrast to Drosophila, which has only one AChE gene, Ace (Hall and Kankel 1976), and vertebrates, which have one AChE gene as well as a second gene encoding a closely related butyrylcholinesterase activity (Taylor and Radic 1994). Catalytically, classes A and B are both somewhat similar to vertebrate AChE (Johnson and Russell 1983), but the class-C enzyme is quite unusual: It has an extremely low Km for acetylcholine and has so far been found only in nematodes (Kolson and Russell 1985b). Vertebrate AChE molecules are often covalently attached to collagen-like tails to help anchor them in the basement membrane (Johnson et al. 1977). In contrast, none of the C. elegans AChE forms appears to be covalently attached to a collagen-like tail. However, within each enzyme class are multiple size forms, some of which appear to be membrane-bound (Johnson and Russell 1983), so that much of the AChE activity is likely to be on the external surface of cell membranes.

The genetic analysis of C. elegans AChE was one of the early triumphs of C. elegans biochemical genetics. All of the mutants were identified by “brute-force” screening, using class-specific single-worm enzyme assays. Animals homozygous for any one of the ace genes have no discernible behavioral or developmental phenotype (although ace-2 animals are hypersensitive to AChE inhibitors). Two of the double-mutant combinations ( ace-1 ; ace-3 and ace-2 ; ace-3 ) are also behaviorally and developmentally essentially normal (Johnson et al. 1988). However, the ace-1 ; ace-2 double mutant is uncoordinated (Culotti et al. 1981), and the ace-1 ; ace-2 ; ace-3 triple mutant is paralyzed and developmentally arrested. Embryonic development is relatively unimpaired, but the animals are unable to grow beyond the hatching stage (Johnson et al. 1988). Thus, most of the functions of the AChE isoforms are individually dispensable but collectively essential. Histochemical assays indicate that classes A and B are present in the nerve ring and the ventral ganglion, with somewhat less staining in the nerve cords and the preanal ganglion (Culotti et al. 1981). Mosaic analysis suggests that ace-1 expression is required in muscle cells (Herman and Kari 1985; Johnson et al. 1988). However, this analysis does not exclude the possibility that ace-1 is also expressed in some neurons.

5. Acetylcholine Receptors and Other Postsynaptic Components

Pharmacological studies suggest that cholinergic transmission at C. elegans neuromuscular junctions is mediated postsynaptically by ligand– gated receptors of the nicotinic acetylcholine receptor (nAChR) family (Lewis et al. 1980b; Avery and Horvitz 1990). Application of nicotine leads to hypercontraction of body wall muscle and modulates pharyngeal muscle action. Although the neuromuscular nAChRs from C. elegans and vertebrates have pharmacological similarities, there are some important differences (Lewis et al. 1980b; Fleming et al. 1993). In particular, toxins such as α–bungarotoxin, which bind very tightly to vertebrate nAChRs, are not effective against the C. elegans receptor, whereas the anti–helminthic levamisole is a potent agonist of C. elegans neuromuscular nAChR receptors. The sensitivity of nematodes to levamisole has provided an effective method of identifying mutants affecting nAChR subunits and associated proteins (Lewis et al. 1980a).

At least 16 nAChR subunit genes are thought to be expressed in C. elegans (Fleming et al. 1993; T. Barnes, pers. comm.), and the role of five of these subunits in C. elegans neurotransmission has been studied in some detail. The unc-29 , unc-38 , and lev-1 genes encode three nAChR subunits that are components of the levamisole–sensitive receptor found in body wall muscle. unc-38 encodes an α-subunit, whereas lev-1 and unc-29 encode non–α subunits (J.T. Fleming et al., in prep.). Coexpression of UNC–38, UNC–29, and LEV–1 in Xenopus oocytes results in levamisole-induced currents (J.T. Fleming et al., in prep.). Genetic and biochemical studies of unc-38 and unc-29 indicate that both of the subunits are essential for levamisole–sensitive nAChR function, but lev-1 is not essential (Lewis et al. 1987b; J.T. Fleming et al., in prep.). Direct evidence that unc-29 , unc-38 , and lev-1 are present in muscle is lacking, although unc-29 is required in the MS, C, and D lineages, which give rise to most of the musculature, suggesting a requirement in muscle (L. Miller and S. Kim, pers. comm.). Because null mutants of all three genes retain some coordinated locomotion, it is likely that additional non-levamisole-sensitive nAChRs are expressed in muscle. Furthermore, cholinergic transmission in the pharynx is probably mediated by a distinct nicotinic receptor complex because unc-29 mutants have normal pharyngeal pumping (Avery 1990). A candidate to encode such a receptor is eat-18 , a gene necessary for pharyngeal muscle to respond to nicotine (Raizen et al. 1995).

Two nAChR subunits expressed in neurons have also been characterized in detail. deg-3 encodes an α–7–like nAChR subunit expressed in touch receptor neurons and some other neurons (Treinin and Chalfie 1995). Dominant deg-3 mutations result in neuronal degeneration of cells expressing this gene (presumably due to increased channel activity), whereas loss of deg-3 activity confers no mutant phenotype. Another nAChR receptor subunit gene, now called acr-2 , is expressed in multiple classes of motor neurons (Squire et al. 1995; Y. Jin and H.R. Horvitz, pers. comm.). An apparent dominant mutation in this gene, n2420, exhibits a shrinking uncoordinated phenotype (Y. Jin and H.R. Horvitz, pers. comm.). Coexpression of acr-2 and unc-38 in Xenopus oocytes is associated with levamisole-inducible currents (Squire et al. 1995). Mutations in unc-38 are capable of suppressing the n2420 phenotype (Y. Jin and H.R. Horvitz, pers. comm.); thus, UNC–38 may be expressed in neurons as well as muscle and may form a functional receptor complex with ACR–2 in vivo.

In addition to the 5 nAChR subunits already characterized, 11 apparent nicotinic receptor subunits are present in the genomic sequence data already released (T. Barnes, pers. comm.), suggesting that as many as 50 nAChR subunits may be expressed in C. elegans. Evidence also exists for muscarinic acetylcholine receptors in C. elegans: Culotti and Klein (1983) described a membrane-bound, high-affinity, saturable binding activity for N-methylscopolamine and quinuclidinyl benzilate, two potent muscarinic receptor blockers. In addition, Avery and Horvitz (1990) reported that muscarinic agonists and antagonists can modulate pharyngeal pumping. However, there are no known mutants defective in the putative muscarinic receptor(s).

Genetic and biochemical approaches have also identified several postsynaptic nonreceptor components required for cholinergic transmission. The levamisole resistance locus unc-50 encodes a nonreceptor molecule of novel structure; homologs of unc-50 exist in eukaryotes from yeast to vertebrates (M. Hengartner, pers. comm.). Because levamisole-binding activity appears to be decreased in unc-50 mutants (Lewis et al. 1987b), the protein seems most likely to play a part in receptor gene transcription or receptor assembly. The levamisole resistance genes unc-63 and unc-74 are also candidates to encode nonreceptor components, since AChR homologs have not been found in the genomic intervals where these genes reside (T. Barnes, pers. comm.).


The vast majority of GABAergic cells in C. elegans are inhibitory motor neurons. However, there are also a few excitatory GABAergic motor neurons. The evidence for GABA function in C. elegans includes pharmacological studies using GABA-related compounds, immunohistochemical demonstration of the presence of GABA in specific cells, and analysis of mutants defective in GABA synthesis and function (McIntire et al. 1993a,b). Mutants with GABAergic transmission defects are viable, although they have several motor defects. The most obvious phenotype is a tendency to contract dorsal and ventral body wall muscle simultaneously in response to touch ("shrinker” phenotype), which appears to result from lack of function of the GABA-containing DD and VD inhibitory motor neurons (McIntire et al. 1993a,b).

Three genes have been identified in C. elegans that are required in the presynaptic neuron for GABAergic transmission. unc-25 encodes a protein that is approximately 45% identical to mammalian glutamic acid decarboxylase (GAD) (Y. Jin and H.R. Horvitz, pers. comm.), the enzyme that synthesizes GABA from glutamic acid. unc-25 mutants have no apparent GABA (McIntire 1993a), and extracts prepared from these animals lack GAD activity (C. Johnson and A. Stretton, pers. comm.). After synthesis, GABA is transported into vesicles by a vesicular transporter, a molecule that has not been molecularly identified in vertebrates. unc-47 mutants accumulate elevated levels of GABA immunoreactivity, a phenotype consistent with a defect in GABA transport. unc-47 probably encodes the GABA vesicular transporter as UNC-47 contains a series of membrane-spanning domains similar to those of many other transporters (K. Schuske and E. Jorgensen, pers. comm.). Finally, there is evidence that the unc-46 gene product is required presynaptically in neurons for GABAergic transmission (McIntire et al. 1993a); however, evidence also exists for postsynaptic unc-46 function (Reiner and Thomas 1995).

Using anti-GABA antibodies, McIntire et al. (1993b) identified 26 neurons that contain GABA immunoreactivity, most of which are motor neurons (including DD, VD, RME, AVL, and DVB; see Appendix 2). In all of these cells, the GABA immunoreactivity was uniformly distributed throughout the cytoplasm and was clearly not restricted to synaptic regions. Expression of the unc-25 GAD gene is restricted to the same 26 neurons (Y. Jin and H.R. Horvitz, pers. comm.), suggesting that all of the GABAergic neurons in C. elegans have been identified.

GABA uptake by a plasma membrane transporter (or transporters) was analyzed by exposing GABA-deficient unc-25 animals to GABA and then staining for GABA immunoreactivity (McIntire et al. 1993a). Surprisingly, only 7 of the 26 GABA-immunoreactive cells take up GABA under these conditions, and there are also additional cells capable of GABA uptake. Thus, not all of the GABA-synthesizing cells express the GABA uptake activity and not all of the transporter-expressing cells can synthesize detectable levels of GABA (McIntire et al. 1993a).

Studies on GABA function, localization, and uptake in Ascaris are in good agreement with the C. elegans results. In Ascaris, GABA acts as an inhibitory transmitter at neuromuscular junctions (del Castillo et al. 1964). The 26 cells with strong and consistent GABA immunoreactivity appear to be the homologs of the 26 C. elegans GABA-positive cells (Guastella et al. 1991). An additional 10 Ascaris cells stain weakly or inconsistently with antisera specific for GABA, and many of these cells also contain GABA transport activity (Guastella and Stretton 1991). It is therefore likely that many (or all) of these cells correspond to the C. elegans cells described above which take up GABA but cannot synthesize it endogenously.

unc-49 mutants share many phenotypic characteristics with unc-25 , unc-46 , and unc-47 mutants. However, unc-49 mutant animals are resistant to the GABA receptor agonist muscimol (McIntire et al. 1993a). Molecular analysis has revealed that the unc-49 gene has the potential to encode several proteins with similarity to GABAA receptors of the ligand-gated ion channel superfamily (B. Bamber and E. Jorgensen, pers. comm.). Additionally, the exp-1 gene may encode a postsynaptic component regulating excitatory GABAergic transmission in the enteric muscles (Avery and Thomas, this volume; see also McIntire et al. 1993a).

C. Dopamine

Dopamine (3,4-dihydroxyphenylethylamine) was originally identified in C. elegans using the technique of formaldehyde-induced fluorescence (FIF; Sulston et al. 1975). In most organisms, dopamine is produced by the hydroxylation of tyrosine by tyrosine hydroxylase to form 3,4-dihydroxy~phenylalanine (DOPA), and the subsequent decarboxylation of DOPA to dopamine by aromatic amino acid decarboxylase (AAAD). Evidence from several organisms suggests that AAAD also decarboxylates 5-hydroxy~tryptophan to produce serotonin. Exogenous dopamine inhibits locomotion and egg laying, and these behavioral responses habituate (Schafer and Kenyon 1995; Ségalat et al. 1995), but it has not yet been proven that these responses are due to dopamine (as opposed to some related transmitter) in vivo.

Five cat (catecholamine-deficient) genes were identified that affected dopamine (Sulston et al. 1975). cat-2 , cat-3 , and cat-5 seem to affect primarily the morphology of the neurons or the subcellular localization of the FIF, but in cat-2 and cat-4 mutants, the level of dopamine is greatly reduced or absent. Dopamine is normally present in eight sensory cells (two ADE, two PDE, and four CEP neurons; Sulston et al. 1975). Unpublished data suggest that dopamine is required for the function of these neurons: Sensory behaviors (including foraging behavior and sensation of bacterial lawns) mediated by these cells are absent in dopamine-deficient cat-2 , cat-4 , or bas-1 (see below) mutants (J. Kaplan, pers. comm.; B. Sawin and H.R. Horvitz, pers. comm.). Dopamine is also present in three pairs of ray neurons (R5A, R7A, R9A) in the male tail (Sulston and Horvitz 1977).

D. Serotonin

Serotonin (5-hydroxytryptamine, or 5-HT) is a common neuro~transmitter in vertebrates and invertebrates, and there is persuasive evidence for its function in C. elegans. It has been identified in C. elegans neurons by formaldehyde-induced fluorescence (Horvitz et al. 1982) and by anti-serotonin immunostaining (Desai et al. 1988; McIntire et al. 1992). In vertebrates, serotonin is usually synthesized from tryptophan in two steps: hydroxylation by the enzyme tryptophan hydroxylase, and decarboxylation by AAAD. In C. elegans, exogenous serotonin stimulates egg laying and pharyngeal pumping and inhibits locomotion and defecation (Horvitz et al. 1982; Ségalat et al. 1995). Serotonin is also required for male mating behavior (Loer and Kenyon 1993).

A number of genes have been identified that affect serotonin metabolism or function. Mutants in bas-1 (biogenic amine synthesis-defective) are deficient in serotonin and dopamine (Loer and Kenyon 1993). However, these mutants can accumulate exogenous serotonin in the “normal” serotonin cells as well as the “dopamine” cells described above (Loer and Kenyon 1993). Wild-type animals are also able to accumulate exogenous 5-hydroxy~tryptophan (5-HTP, the precursor of serotonin) in the “serotonin” and “dopamine” cells and convert it to serotonin. Treatment of bas-1 mutants with exogenous 5-HTP does not lead to any serotonin immunofluorescence (Loer and Kenyon 1993), which is consistent with bas-1 mutants having a defect in AAAD. bas-1 maps near (and could be the same as) a putative AAAD gene identified by the Genome Sequencing Project (C. Loer, pers. comm.). Other putative AAAD homologs have been identified (see, e.g., Marra et al. 1993), but it is not known what part, if any, these other genes may play in neurotransmitter metabolism.

The cat-4 gene, originally identified because of its effect on dopamine cells (Sulston et al. 1975), is also deficient in serotonin (Desai et al. 1988), but it does not appear to have a defect in serotonin uptake or in the decarboxylation of 5-HTP to make serotonin (Loer and Kenyon 1993).

Several genes have also been identified that affect the response to exogenous serotonin (Ségalat et al. 1995). The best characterized such gene is goa-1 , which encodes a Go subunit apparently required for transduction of a signal from a metabotropic serotonin receptor (Mendel et al. 1995; Ségalat et al. 1995; Jorgensen and Rankin, this volume).

There are at least ten cells in C. elegans hermaphrodites (and a greater number in males) with significant anti-serotonin immunoreactivity (see Appendix 2) (Desai et al. 1988; Loer and Kenyon 1993; G. Garriga; B. Sawin and H.R. Horvitz; both pers. comm.). The NSM cells have the strongest and most consistent staining. These pharyngeal cells have varicosities, fine branches, and endings on the surface of the pharynx (Albertson and Thomson 1976), suggesting that serotonin might be released into the pseudocoelom and have a humoral function. The male-specific CP neurons also have strong serotonin immunoreactivity; they are involved in male mating behavior, and animals lacking serotonin ( bas-1 , cat-4 ) are defective in this behavior (Loer and Kenyon 1993). In addition, the HSN cells, which are required for egg laying (Trent et al. 1983), contain serotonin (Desai et al. 1988), but serotonin does not appear to be essential for HSN function (Weinshenker et al. 1995).

E. Glutamate and Other Transmitters

Glutamate acts as both an excitatory and inhibitory neurotransmitter in C. elegans. In the pharynx, the M3 motor neurons appear to be inhibitory and glutamatergic and act by opening a chloride channel (J. Dent et al., pers. comm.). avr-15 , a gene involved in resistance to avermectin (C. Johnson, pers. comm.), is a good candidate to encode a glutamate receptor (or a receptor subunit), because pharyngeal muscle of avr-15 mutants does not respond to pulses of glutamate (J. Dent et al., pers. comm.; Avery and Thomas, this volume). Additionally, two genes encoding subunits of a glutamate-gated chloride channel have been isolated by functional expression cloning in Xenopus oocytes (Cully et al. 1994).

Analysis of glr-1 mutants, lacking a protein with 40% identity to the vertebrate AMPA glutamate receptor subunit, suggests that several classes of sensory neurons use glutamate as an excitatory transmitter (Hart et al. 1995; Maricq et al. 1995). The most compelling case exists for the sensory neuron ASH, but Hart et al. (1995) also proposed that certain mechanosensory neurons are glutamatergic (see Driscoll and Kaplan, this volume). However, additional neurons are also likely to be glutamatergic since the glr-1 receptor is expressed in 17 classes of neurons (Hart et al. 1995; Maricq et al. 1995). Several other genes encoding homologs of mammalian excitatory glutamate receptors have been identified by polymerase chain reaction (PCR) methods and by the Genome Sequencing Project (A. Maricq and C. Bargmann, pers. comm.).

Octopamine (p-hydroxyphenyl~ethanolamine) has been detected in C. elegans extracts, and exogenous octopamine stimulates movement and inhibits egg laying (Horvitz et al. 1982). Thus, its biological actions appear to antagonize those of serotonin, but it is not yet known which cells contain octopamine.

Several additional small molecules known to be neurotransmitters in other organisms (including adenosine, epinephrine, glycine, aspartate, histamine, and norepinephrine) have not yet been identified as transmitters in C. elegans. In most cases, the technology (e.g., immunohistochemistry and enzyme assays) available for analysis of these transmitters is not as good as that available for the transmitters described above. Therefore, as new and more sensitive methods are developed, it is possible that the existence and function of some of these transmitters will be identified in C. elegans.

F. Peptides

In general, peptide neurotransmitters/neuromodulators appear to be regulated, transported, and released by mechanisms different from those used by the classical small molecule neurotransmitters. In both invertebrates and vertebrates, families of related peptides are often processed from precursor proteins by specific peptidases, and this appears to be true in C. elegans as well. To date, only one class of peptide has been studied in C. elegans: the FMRFamide group, which includes FLRFamide peptides (the terminology is derived from the carboxy-terminal amino acid sequence of the peptides). The only gene to be analyzed thus far is flp-1 (FMRFamide like peptide; Rosoff et al. 1992), which encodes an alternatively spliced transcript. flp-1 encodes proteins that may be processed to give eight related FLRFamide peptides. Other possible peptide-precursor genes have been identified by the Genome Sequencing Project, but these genes and their putative products have not yet been studied.

flp-1 mutants have not yet been characterized and the precise function(s) of the encoded peptides is not known. However, the peptide FLRFamide was shown to potentiate the effects of serotonin on vulval muscles (Schinkmann and Li 1992). Peptide immunolocalization was performed using an antibody recognizing carboxy-terminal Arg-Phe-NH2 epitopes (Schinkmann and Li 1992). This antibody is expected to react with most or all FMRFamide and FLRFamide peptides and perhaps many others; it should thus identify all of the flp-1 -containing peptides and perhaps many more. These peptides have been immunolocalized to the VC motor neurons of the ventral nerve cord and approximately 25 additional cells throughout the body, including both interneurons and motor neurons (see Appendix 2) (Schinkmann and Li 1992). It is quite possible that the different flp-1 peptides are expressed in different sets of cells.

In Ascaris, where the identification and separation of neuropeptides have been more carefully analyzed, it is estimated that there may be as many as a dozen families of distinct bioactive peptides, many of which have highly specific cellular localizations and physiological effects (Sithigorngul et al. 1990; Stretton et al. 1991; Cowden et al. 1993; Cowden and Stretton 1995; see Jorgensen and Rankin, this volume). It is therefore likely that a more intensive biochemical and cellular analysis of C. elegans will reveal comparable complexity.

Copyright © 1997, Cold Spring Harbor Laboratory Press.
Bookshelf ID: NBK20199


  • PubReader
  • Print View
  • Cite this Page

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...