Nature Neuroscience
7, 1096 - 1103 (2004)
Published online: 19 September 2004; | doi:10.1038/nn1316
Mechanism of extrasynaptic dopamine signaling in Caenorhabditis elegansDaniel L Chase, Judy S Pepper
& Michael R KoelleDepartment of Molecular Biophysics and Biochemistry, Yale University, New Haven, Connecticut 06520, USA.
Correspondence should be addressed to Daniel L Chase daniel.chase@yale.eduD1-like and D2-like dopamine receptors have synergistic and antagonistic effects on behavior. To understand the mechanisms underlying these effects, we studied dopamine signaling genetically in Caenorhabditis elegans. Knocking out a D2-like receptor, DOP-3, caused locomotion defects similar to those observed in animals lacking dopamine. Knocking out a D1-like receptor, DOP-1, reversed the defects of the DOP-3 knockout. DOP-3 and DOP-1 have their antagonistic effects on locomotion by acting in the same motor neurons, which coexpress the receptors and which are not postsynaptic to dopaminergic neurons. In a screen for mutants unable to respond to dopamine, we identified four genes that encode components of the antagonistic G o and Gq signaling pathways, including Go itself and two subunits of the regulator of G protein signaling (RGS) complex that inhibits Gq. Our results indicate that extrasynaptic dopamine regulates C. elegans locomotion through D1- and D2-like receptors that activate the antagonistic Gq and Go signaling pathways, respectively.Defects in dopamine signaling underlie schizophrenia1, drug addiction2 and Parkinson disease3. The effects of dopamine are mediated by heterotrimeric G proteins, as all dopamine receptors identified are seven-transmembrane proteins of the G protein−coupled receptor family4. However, the identity of the physiologically relevant signaling molecules and pathways activated by these receptors is less clear.
In mammals, dopamine acts through five receptors that, based upon structural, biochemical and pharmacological criteria, are grouped into two classes: D1-like and D2-like. Selective pharmacological agents and receptor knockouts have been used to show that signaling by D1- and D2-like receptors have opposite effects on behavior5,
6,
7. D1-like and D2-like signaling are not purely antagonistic, however, as simultaneous stimulation of D1-like and D2-like receptors reveals that they also have synergistic effects on certain behaviors5,
8. The physiological signaling mechanisms responsible for the antagonistic and synergistic effects of dopamine receptors have not been clearly established. D1- and D2-like receptors were initially thought to be expressed in distinct neurons. More recent work suggests that neurons with one receptor class at a high level also have lower levels of the other receptor class, so that both could signal within the same cells9. Biochemical experiments have shown that D1- and D2-like receptors can have opposite effects on adenylyl cyclase activity by signaling through G s/olf and Gi, respectively10,
11,
12,
13. It remains unclear, however, to what extent the opposing biochemical effects of Gs/olf and Gi signaling can account for the antagonistic effects of D1- and D2-like signaling on behavior.
In C. elegans, dopamine controls locomotion behavior14. The mammalian enzymes involved in dopamine synthesis, vesicle loading and reuptake by neurons all have C. elegans homologs, and mutants for each of these homologs have previously been analyzed15,
16. We have performed a genetic analysis of dopamine signaling in C. elegans and found that dopamine functions extrasynaptically to modulate locomotion rate by activating both D1- and D2-like receptors coexpressed in the cholinergic motor neurons. D1- and D2-like receptor signaling oppose each other in these cells by activating the antagonistic Gq and Go signaling pathways, respectively.
Results
Identification of the D2-like dopamine receptor DOP-3
DOP-1 and DOP-2 are C. elegans receptors with sequence similarity to mammalian D1-like and D2-like receptors, respectively. They also have been shown to bind dopamine with high affinity17,
18. To identify additional dopamine receptors, we did a blast search of the C. elegans genome using the human D2 receptor sequence as the search query. Eighteen putative receptors were identified, including DOP-1 and DOP-2. The predicted membrane-spanning regions of these receptors were used in a phylogenetic comparison with each other and with those of human dopamine, serotonin and muscarinic acetylcholine receptors (Fig. 1a). Besides DOP-1 and DOP-2, three additional C. elegans receptors (T14E8.3, C24A8.1 and C02D4.2) clustered with mammalian dopamine receptors. T14E8.3, hereafter referred to as DOP-3, showed very high similarity to the known D2-like dopamine receptor DOP-2, and was therefore selected for further analysis.
 | | |
Using RT-PCR, we isolated a cDNA for dop-3 and used it to deduce the DOP-3 amino acid sequence. DOP-3 is similar to the human D1 and D2 receptors throughout its length, particularly within the seven membrane-spanning regions (Fig. 1b). DOP-3 is more similar to the human D2 receptor than to the human D1 receptor (51% vs. 40% identity within the membrane-spanning regions and 38% vs. 24% identity throughout the entire protein). DOP-3 also contains a long third intracellular loop and a short carboxy (C)-terminal tail following the seventh membrane-spanning region, features typical of D2-like receptors4. Thus DOP-3, like DOP-2, appears to be a C. elegans D2-like dopamine receptor.
Identification and analysis of dopamine receptor knockouts
To determine the physiological roles of DOP-1, DOP-2 and DOP-3, we screened a C. elegans mutant library for deletion alleles of the genes encoding these receptors. We recovered two mutants (vs100 and vs101) for dop-1, one (vs105) for dop-2 and one (vs106) for dop-3 (Fig. 2). The deletion mutations are all predicted to severely compromise or eliminate receptor function by interrupting the open reading frames and/or deleting sequences encoding the conserved transmembrane domains.
| | |
We examined the receptor mutants for defects in dopamine signaling. Well-fed wild-type animals slow their locomotion when they encounter a bacterial lawn, and this 'basal slowing response' is controlled by dopamine signaling14. C. elegans moves by propagating sinusoidal bends along its body, and the frequency of these bends can be measured to assess the rate of locomotion. Using this measure, we found that wild-type animals slowed by 42% in response to bacteria, whereas cat-2 mutants, which are unable to synthesize dopamine19, showed no detectable slowing (Fig. 3a). The dop-1 and dop-2 mutants showed normal basal slowing responses (Fig. 3a). Another dop-1 deletion strain has recently been analyzed with similar results20. In contrast, the dop-3 mutant was completely defective in basal slowing (Fig. 3a). This defect could be rescued by transforming the dop-3 mutant with a transgene carrying the wild-type dop-3 gene, demonstrating that the basal slowing defect in the dop-3 mutant is due to loss of dop-3 function (Fig. 3a).
| | |
In food-deprived animals, serotonin induces an 'enhanced slowing response' beyond the dopamine-mediated basal slowing response seen in well-fed animals14. We found that the dopamine receptor mutants dop-1, dop-2 and dop-3, when food-deprived, all showed enhanced slowing responses indistinguishable from that of wild-type animals (16−22 body bends every 20 s in the absence of food and slowing to 2−3 body bends every 20 s in the presence of food). Thus DOP-3 acts specifically as a dopamine receptor to inhibit locomotion behavior in the basal slowing response.
We examined basal slowing in animals lacking all combinations of the DOP-1, DOP-2 and DOP-3 receptors (Fig. 3a). The dop-2 mutation had little or no effect on slowing in any genetic background, and we thus found no evidence for any role of dop-2 in basal slowing. Our double-mutant analysis did reveal a function for DOP-1: it antagonizes DOP-3 to control basal slowing. Whereas the dop-3 single mutant showed a complete basal slowing defect, we found that the dop-1 dop-3 double mutants behaved more like wild-type animals, showing significant slowing upon entering the bacterial lawn. Our second dop-1 allele gave similar results (data not shown).
Exposure of C. elegans to high concentrations of exogenous dopamine causes paralysis21. Wild-type animals showed a dosage-dependent increase in paralysis upon exposure to dopamine, and were completely paralyzed by 30 mM dopamine (Fig. 3b). Such high concentrations of dopamine may be required because the C. elegans cuticle acts as a permeability barrier22. The response of dop-2 mutants to exogenous dopamine was similar to that of wild-type animals, and dop-1 mutants were, if anything, slightly more sensitive to exogenous dopamine than wild-type animals (Fig. 3b). dop-3 mutants, however, were resistant to the paralytic effects of exogenous dopamine. At 30 mM dopamine nearly 65% of dop-3 mutants were capable of movement, whereas none of the wild-type animals moved. This suggests that exogenous dopamine exerts its paralytic effects by hyperactivating the same signaling pathways used by endogenous dopamine to slow locomotion, that is, through activation of DOP-3.
We tested mutants lacking all combinations of the receptors for sensitivity to exogenous dopamine (Fig. 3c). The dop-2 mutation had little effect when crossed into the other mutants. Interestingly, the dop-1 dop-3 double mutants, rather than showing the dopamine resistance seen in the dop-3 single mutant, were paralyzed by exogenous dopamine with sensitivity similar to that seen in the wild type. This indicates that exogenous dopamine activates both DOP-1 and DOP-3 to exert antagonistic effects on locomotion. This result is analogous to the antagonistic roles of DOP-1 and DOP-3 in mediating the effects of endogenous dopamine on the basal slowing response.
We note that all receptor mutants, including the triple mutant, can be paralyzed by sufficiently high concentrations of exogenous dopamine (Fig. 3c). The triple mutant also shows significant basal slowing (Fig. 3a). Thus dopamine exerts some of its effects through receptors other than DOP-1, DOP-2 and DOP-3.
Analysis of DOP-1 and DOP-3 expression patterns
To begin to address whether DOP-1 and DOP-3 might antagonize each other within the same cells, we examined the expression patterns of DOP-1 and DOP-3. We generated chromosomally integrated transgenes in which the dop-1 and dop-3 promoters directed expression of the green or red fluorescent proteins (GFP or RFP). We found that dop-1 reporters were expressed in support cells in the head, as well as in neurons in the head, ventral cord and tail (Fig. 4a). dop-3 reporters were also expressed in neurons of the head, ventral cord and tail. However, at a gross level, expression of the dop-1 and dop-3 reporters did not appear to overlap. dop-3 reporter expression was also seen in some body-wall muscles, although this was weak and variable, and in two mechanosensory neurons known as PVD neurons. Although we did not identify every neuron expressing the dop-1 or dop-3 reporters, they were not expressed in the command interneurons that affect locomotion nor in any dopaminergic neurons, as expression of the RFP reporters for dop-1 and dop-3 did not overlap expression of integrated GFP reporters for glr-1 or dat-1, which label these two classes of cells, respectively23,
24 (data not shown). We note that recently described dop-1:GFP reporters are expressed in fewer cells than we observed25,
20. The dop-1 promoter used in our studies contained additional sequences that extended to the gene adjacent to dop-1, and thus likely contained dop-1 enhancer regions that were lacking in the previously studied reporters.
| | |
Ventral cord motor neurons innervate the body wall muscles to directly control locomotion26, and we carried out a detailed analysis of dop-1 and dop-3 reporter expression in these cells (Fig. 4b−d). The ventral cord contains two types of motor neurons: cholinergic neurons stimulate muscle contraction and GABAergic neurons inhibit contraction. We crossed integrated GFP reporters that specifically label the cholinergic or GABAergic neurons into strains expressing integrated dop-1:rfp or dop-3:rfp transgenes to identify the dop-1 and dop-3 expressing cells (data not shown). We found that dop-1 reporters were expressed specifically in the cholinergic motor neurons, whereas dop-3 reporters were expressed strongly in the GABAergic neurons. Upon close inspection, dop-3 also showed weaker expression in most cholinergic motor neurons, thus overlapping dop-1 expression in these cells (Fig. 4b−d). dop-1 also showed expression in the PVD neurons, which is thus a second site of coexpression with dop-3.
DOP-1 antagonizes DOP-3 in cholinergic motor neurons
Given that the ventral cord motor neurons express DOP-1 and DOP-3 and directly control locomotion, we tested the hypothesis that DOP-1 and DOP-3 act in these cells to mediate the effects of dopamine on locomotion. We used promoters active in the cholinergic or GABAergic neurons of the ventral cord to express DOP-1 or DOP-3 in these cells and tested the ability of such transgenes to rescue the defects of dop-1 or dop-3 mutants.
For rescue of dop-3 function, we tested for the ability of dop-3 transgenes to reduce the dopamine resistance of the dop-3 mutant. Control and transgenic strains were tested for sensitivity to 30 mM dopamine (Fig. 5). Wild-type animals were dopamine sensitive (4% moving under these conditions), whereas dop-3 mutants were resistant (64% moving) (Fig. 5a). We expressed DOP-3 from the acr-2 promoter, which is active in cholinergic neurons of the ventral cord27. This strongly rescued the dop-3 mutant defect, such that only 8% of the transgenic animals were able to move (Fig. 5b). We also expressed DOP-3 in the GABAergic neurons using the unc-47 promoter28. This resulted in partial rescue (32% of transgenic animals moving). Thus DOP-3 mediates the effects of dopamine on locomotion in both cholinergic and GABAergic motor neurons of the ventral cord, although it appears that the cholinergic neurons may be the primary site of DOP-3 function.
| | |
Because the only defects caused by the dop-1 mutation were observed in the dop-3 mutant background, for rescue of dop-1 function we tested the ability of dop-1 transgenes to increase dopamine resistance in the dop-1 dop-3 double mutant toward levels seen in the dop-3 single mutant. Control dop-1 dop-3 animals were sensitive to 30 mM dopamine, whereas dop-3 single mutants were resistant (Fig. 5a). Because dop-1 reporters were expressed specifically in the cholinergic motor neurons, we expressed DOP-1 from the acr-2 promoter and saw significant rescue of DOP-1 function (22% moving, compared to 2% moving in control dop-1 dop-3 animals not expressing DOP-1) (Fig. 5b). Thus DOP-1 and DOP-3 are coexpressed in cholinergic motor neurons, and both function in these cells to mediate antagonistic effects of dopamine on locomotion.
Genetic screen for dopamine-resistant mutants
To identify additional molecules involved in dopamine signaling, we isolated mutants with a phenotype similar to that of dop-3 mutants. Resistance to exogenous dopamine is much more easily observed than is a defect in basal slowing, so we screened for the dopamine resistance phenotype. We placed second-generation progeny of 15,000 mutagenized hermaphrodites on agar plates containing 40 mM dopamine and, after 10−30 min of exposure, selected rare mutants capable of sustained body movement. In this way, we isolated nine independent dopamine-resistant strains. Wild-type animals placed on 40 mM dopamine plates became paralyzed within 1 min and often assumed a distended body posture (Fig. 6a, left panel), while mutants isolated from the screen, such as eat-16(vs72), continued to move freely for at least 10 min (Fig. 6a, right panel).
| | |
Genetic complementation tests and DNA sequence analysis showed that all nine mutations isolated in the screen were alleles of four genes (goa-1, dgk-1, eat-16 and gpb-2) that have been previously analyzed genetically. Each of these genes encodes a neural G protein signaling molecule (discussed in detail below) conserved in mammals (Table 1)29,
30,
31,
32,
33,
34,
35,
36,
37. Our failure to isolate dop-3 mutants in the screen indicates that the screen was not large enough to identify every gene that can mutate to give dopamine resistance.
Mutants from the screen have dopamine signaling defects
We selected one mutant allele to represent each of the four genes identified in the screen and used a dopamine dose-response analysis to quantify dopamine resistance. Previously characterized null alleles for goa-1, eat-16 and dgk-1 are available33,
34,
38 and were tested in this assay. For gpb-2 we used the allele isolated in the screen.
The mutants tested were resistant to dopamine when compared to the wild type. goa-1 null mutants were completely resistant to all concentrations of dopamine tested (Fig. 6b). All other mutants, including the null mutants for dgk-1 and eat-16, were more like the dop-3 mutant in showing partial resistance to dopamine. These results show that GOA-1, the C. elegans ortholog of the neural G protein Go29, is absolutely required for the paralysis induced by exogenous dopamine, whereas DGK-1, GPB-2 and EAT-16 may only modify this dopamine response.
To determine whether the genes identified in our screen are involved in endogenous dopamine signaling, we tested the mutants for basal slowing. As each strain tested had a slightly different overall locomotion rate, we calculated the percent slowing induced by bacteria for each strain and compared these values to the 42% slowing seen in the wild type (Fig. 6c). All mutants resistant to exogenous dopamine showed significantly decreased slowing. The slowing defects ranged from severe (i.e., no detectable slowing in eat-16) to mild (32% slowing in dgk-1), but all mutants isolated in the screen for resistance to exogenous dopamine were also significantly defective for endogenous dopamine signaling.
Go antagonizes Gq to mediate dopamine signaling
GOA-1, DGK-1, EAT-16 and GPB-2 are expressed throughout the C. elegans nervous system and are components of two opposing G protein signaling pathways that have been characterized through genetic and molecular studies (Supplementary Fig. 1 online). Signaling by GOA-1, the C. elegans ortholog of human Go, antagonizes signaling by EGL-30, the C. elegans ortholog of human Gq, to set the levels of egg laying, locomotion and other behaviors30,
31,
33,
39. The RGS proteins EGL-10 and EAT-16 are GTPase activating proteins that inhibit GOA-1 Go and EGL-30 Gq, respectively33,
40. GPB-2, the C. elegans ortholog of human G 5, is an obligate subunit of both EGL-10 RGS and EAT-16 RGS35,
36,
37. DGK-1, the C. elegans ortholog of human DGK , encodes a diacylglycerol kinase32 and is a putative downstream effector of GOA-1 Go, whereas EGL-8, the C. elegans ortholog of human PLC, is a downstream effector of EGL-30 Gq41.
The results of our screen showed that mutations that reduce GOA-1 Go signaling (goa-1 and dgk-1 mutations) or increase EGL-30 Gq signaling (eat-16 RGS mutations) confer resistance to exogenous dopamine. A gpb-2 G5 mutation, vs74, was also isolated in the screen. This missense mutation (Ala117Thr) caused defects such as hyperactive egg laying (data not shown) that were not seen in the gpb-2 null mutant35, but that were seen in two previously isolated special mutants of gpb-2 (ref. 36). Like vs74, these special gpb-2 mutants were more strongly dopamine resistant than the gpb-2 null mutant (data not shown), and also had single amino acid changes in the beta-propeller region of GPB-2 G5. It has been suggested that the special gpb-2 mutations specifically disrupt EAT-16 RGS function while leaving EGL-10 RGS function relatively unaffected36. Thus the vs74 mutation is predicted to specifically increase EGL-30 Gq signaling. This is consistent with our hypothesis that mutations that cause dopamine resistance either reduce GOA-1 Go signaling or increase EGL-30 Gq signaling.
We tested our hypothesis by altering GOA-1 Go or EGL-30 Gq signaling in four different ways. We began by examining egl-10 RGS, egl-30 Gq and egl-8 PLC mutants for dopamine sensitivity. Our hypothesis predicts that these mutations should cause hypersensitivity to dopamine by either increasing GOA-1 Go signaling (egl-10 RGS) or decreasing EGL-30 Gq signaling (egl-30 Gq and egl-8 PLC). Indeed, we found that egl-30 Gq mutants were more sensitive to exogenous dopamine than were wild-type animals (Fig. 7). One hundred percent of egl-30 Gq mutants were paralyzed after exposure to 15 mM dopamine, whereas 55% of wild-type animals continued to move. We also found that egl-8 PLC and egl-10 RGS mutants were more sensitive to dopamine than were wild-type animals (Fig. 7). As a fourth test of our hypothesis, we examined the dopamine resistance of animals that overexpressed EGL-10 RGS from a chromosomally integrated transgene. Overexpressing EGL-10 RGS has previously been shown to strongly reduce signaling through GOA-1 Go40. As predicted, animals overexpressing EGL-10 RGS were dopamine resistant (Fig. 7).
| | |
Discussion
Extrasynaptic dopamine acts through coexpressed receptors
Our results in C. elegans show striking parallels to observations that D1-like and D2-like receptors have antagonistic effects in mammals, and establish C. elegans as a model in which to analyze the mechanism of this antagonism. We found that a C. elegans D2-like receptor, DOP-3, acts to slow locomotion, whereas a D1-like receptor, DOP-1, antagonizes this effect.
DOP-1 and DOP-3 reporters are highly expressed in the cholinergic and GABAergic ventral cord motor neurons, respectively, both of which directly control locomotion. In mammals, D1- and D2-like receptors are highly expressed in different neurons of the neostriatum, but confocal microscopy shows that neurons expressing one receptor type at a high level also contain the other receptor type at a low level9. This parallels our observations that cholinergic motor neurons, in addition to high levels of DOP-1 expression, also show low levels of DOP-3 expression. To determine the sites of receptor function, we transgenically expressed the receptors in specific neurons of the ventral cord. We found that DOP-1 and DOP-3 are not only coexpressed in cholinergic neurons, they both actually function in these cells to control locomotion. DOP-3 also has some function in GABAergic neurons, which do not express DOP-1. Thus DOP-1 and DOP-3 apparently directly antagonize each other in the same cells, and may also indirectly antagonize each other's effects by acting in different cells.
DOP-1 and DOP-3 mediate signaling by extrasynaptic dopamine. The eight dopaminergic neurons in C. elegans make synapses onto 41 other neuron types in the animal, and only one of the 58 ventral cord neurons is in this set26. This indicates that dopamine must diffuse from its site of release and function as a neurohormone in C. elegans. Our results are consistent with a previous study in C. elegans in which cells directly postsynaptic to dopaminergic neurons were ablated without eliminating the basal slowing response14. In mammals, dopamine also functions as a neurohormone: dopamine receptors are extrasynaptic42 and are expressed in regions of the brain that do not receive dopaminergic innervation43. In addition, functional studies suggest dopamine has effects remote from its sites of release44.
Dopamine signaling mechanisms in C. elegans and mammals
Our results indicate that the D2-like receptor DOP-3 signals through Go to inhibit locomotion, and that this effect is antagonized by signaling of the D1-like receptor DOP-1 through Gq. Our screen for dopamine-resistant mutants showed that mutations that decrease Go signaling, like DOP-3 mutations, lead to dopamine resistance. Analysis of a panel of preexisting signaling mutants confirmed this result, and showed that mutations that decrease Gq signaling, like DOP-1 mutations, have the opposite effects. We found that expression of DOP-1 or DOP-3 in ventral cord cholinergic motor neurons using the acr-2 promoter was sufficient to mediate effects of dopamine on locomotion, and the Go and Gq signaling pathways have been shown to control locomotion by acting in these same cells. Indeed, GOA-1 Go and EGL-30 Gq are both expressed in cholinergic motor neurons, and the acr-2 promoter has previously been used to transgenically express activated forms of these G proteins32,
45. It was found that GOA-1 Go inhibits acetylcholine release from these cells, whereas EGL-30 Gq stimulates acetylcholine release, thus controlling locomotion. EGL-8 PLC and DGK-1 were similarly shown to act in the ventral cord cholinergic motor neurons to oppositely control acetylcholine release and affect locomotion32,
45. Taken together, these results support our model (shown in Supplementary Fig. 1 online) in which DOP-3 and DOP-1 have opposing effects on locomotion by acting in the same cells to directly activate the antagonistic Go and Gq signaling pathways, respectively.
We propose that much of the observed antagonism between D1-like and D2-like receptor signaling in mammals can be explained by their signaling through Gq and Go, respectively. Four lines of evidence support this proposal. First, our work demonstrates that dopamine acts through a D2-like receptor (DOP-3) and Go in C. elegans, and dopamine in the mammalian brain also acts through Go using D2-like receptors. In fact, most D2-like receptors in the brain are coupled to Go, as D2-like receptors largely lost their coupling to G proteins in brains from Go knockout mice46. Second, our work suggests that a D1-like receptor (DOP-1) signals through Gq in C. elegans, and D1-like dopamine receptors can also activate Gq in mammalian brain47. Third, we found that the Go and Gq signaling pathways oppose each other to mediate the physiological effects of dopamine in C. elegans. Whereas the relationship between Go and Gq signaling in mammals is unclear, the antagonistic character of Go and Gq signaling in C. elegans is well-established33. The high degree of conservation, between C. elegans and mammals, of all the proteins in these pathways suggests that Go and Gq may also function antagonistically in the mammalian brain. Finally, the physiological consequences of antagonistic signaling by Gq and Go in C. elegans parallel the antagonistic effects of mammalian D1-like and D2-like receptors. This is most evident in studies of the control of neurotransmitter release. In mammals, D1-like receptors activate release of neurotransmitters including GABA and acetylcholine in certain brain regions, whereas D2-like receptors inhibit such neurotransmitter release48,
49. In C. elegans, Gq signaling seems to increase acetylcholine release at ventral cord motor neuron synapses, whereas Go signaling decreases release45,
32. If the mechanism of Gq and Go signaling is conserved from C. elegans to mammals, signaling by D1-like receptors through Gq, and D2-like receptors through Go, could explain the observed antagonistic effects of these receptors on neurotransmitter release in mammals.
Methods
Phylogenetic analysis of dop-3.
Transmembrane regions for each putative C. elegans receptor were identified by manual sequence comparison to the transmembrane regions of mammalian G protein−coupled receptors. The amino acid sequences of the transmembrane domains were catenated and aligned by the program MegAlign (DNASTAR Inc.), and the phylogenetic tree was plotted.
Behavioral assays.
To quantify basal slowing, nematode growth media (NGM) plates with bacteria were prepared by spreading 35  l of HB101 (absorbance at 600 nm, A600 = 0.70−0.75) bacteria across each plate and incubating overnight at 37 °C. Plates without bacteria were also incubated overnight and all plates were cooled to room temperature (20−23 °C) before use. Assays were performed on staged adults selected as late L4 larvae and aged 15−19 h at 20 °C. Animals were washed free of bacteria in 1 ml of S basal for 3 min in a watch glass and then placed on assay plates. For assays in the presence of bacteria, animals were placed on the bacterial lawn. Then, 90 s later, the number of body bends made by each animal was counted for five consecutive 20-s intervals. At least six animals were examined for each genotype for a total of 30 separate measurements. Assays of the enhanced slowing response were as described14.
To quantify dopamine resistance, young adults were transferred to plates containing dopamine at the concentrations indicated and incubated undisturbed for 20 min at room temperature. Animals were then inspected for spontaneous body bends during a 15-s period. A body bend was defined as the movement of a point in the animal immediately posterior to the pharynx through a minimum or maximum amplitude. All strains were analyzed at each concentration of dopamine on the same day using plates that were made in a single batch. Dopamine plates (6 cm) were prepared by autoclaving 1.7% Difco Bacto agar (Becton Dickinson) in water, cooling to  50 °C, adding glacial acetic acid to a concentration of 2 mM and dopamine to various concentrations, pouring and drying for 2 d at room temperature. Plates were stored inverted at 4 °C and used within one week.
Mutants analyzed in behavioral assays shown were as follows: goa-1(sa734), dgk-1(sy428), eat-16(ad702), gpb-2(vs74), eat-16(vs72), nIs51 (egl-10 overexpressor transgene), egl-10(md176), egl-8(md1971), egl-30(n686), cat-2(e1112), dop-1(vs100), dop-2(vs105), dop-3(vs106), dop-2(vs105); dop-1(vs100), dop-2(vs105); dop-3(vs106), dop-1(vs100) dop-3(vs106) and dop-2(vs105); dop-1(vs100) dop-3(vs106).
Transgenes.
To examine expression of dop-1 and dop-3, we modified the GFP reporter vector pPD95.77 (kind gift of A. Fire, Stanford University School of Medicine) to replace GFP coding sequences with those of mRFP1 (ref. 50), and promoters for each gene were inserted into both the GFP and RFP versions. For dop-1, a genomic fragment including 9,180 base pairs of promoter sequence and 70 base pairs of coding sequence was used, and for dop-3, 13,035 base pairs of genomic DNA (9,955 base pairs of promoter and sequences extending into the fourth coding exon) was used. The precise limits and primers used for amplification of dop-1 and dop-3 promoters are described in Supplementary Methods online. The lin-15(n765) animals were microinjected with 50 ng/l of a lin-15-rescuing plasmid as marker (pL15EK) along with 100 ng/l of a GFP or RFP reporter to generate extrachromosomal transgenes. Three transgenes were chromosomally integrated by UV/psoralen mutagenesis and named as follows: (i) dop-1:gfp, vsIs28, (ii) dop-1:rfp, vsIs22 and (iii) dop-3:rfp, vsIs33. Animals shown in Figure 4 were generated by crossing vsIs28 and vsIs33 animals. We also crossed a glr-1:gfp transgene that labels command interneurons (nuIs1), an integrated dat-1:gfp transgene that labels dopaminergic neurons (a gift of R. Nass, Vanderbilt University) and an unc-17:gfp transgene that labels cholinergic neurons (vsIs48) into the vsIs22 and vsIs33 backgrounds to analyze dop-1 and dop-3 RFP reporter expression in these cells.
For the rescue experiment in Figure 3a, dop-3(vs106) mutants were microinjected with 15 ng/l of a plasmid encoding myo-2:GFP as marker along with 5 ng/l of a 17,163 bp dop-3 genomic fragment, which included the same promoter region used to determine the expression pattern of dop-3. For the rescue experiments in Figure 5, acr-2 and unc-47 promoter fragments were cloned in front of coding sequences for DOP-3 or GFP in the vector pPD49.26 (gift of A. Fire), or no coding sequences were used to generate 'empty vectors.' An analogous acr-2:dop-1 construct was also generated. The promoter fragments extended the following lengths upstream of the translation start codons: acr-2, 3,198 bp and unc-47, 257 bp. As controls, myo-2:gfp was injected as a marker at 10 ng/l along with the acr-2 or unc-47 GFP constructs at 50 ng/l, and an acr-2 or unc-47 empty vector at 25 ng/g. To test for rescue, the empty vector constructs were replaced by the analogous constructs containing DOP-1 or DOP-3 coding sequences. The transgenic animals assayed for dopamine responsiveness were selected for fluorescence in ventral cord neurons to obtain animals that had retained their extrachromosomal transgenes in those cells.
Isolation of dopamine-resistant mutants.
Wild-type (N2) animals were mutagenized with ethyl methanesulfonate and cultured on NGM plates. F2 progeny of mutagenized animals were allowed to grow to adulthood, washed twice in water, transferred by pipette (along with 50 l water) to 40 mM dopamine plates, and the animals were distributed across the plate surface using 3 mm glass beads. After 10−30 min, during which time the water absorbed into the plates, animals capable of sustained movement were picked individually. Primary isolates were retested by placing 10−15 of their progeny onto 40 mM dopamine plates. Strains in which all progeny continued to move after 5 min were deemed dopamine resistant.
Accession numbers.
The GenBank accession number for DOP-3 is AY485309.
Note: Supplementary information is available on the Nature Neuroscience website.
Received 11 June 2004; Accepted 28 July 2004; Published online: 19 September 2004.
REFERENCES
- Knable, M.B. & Weinberger, D.R. Dopamine, the prefrontal cortex and schizophrenia. J. Psychopharmacol. 11, 123−131 (1997). | PubMed | ISI | ChemPort |
- Koob, G.F., Sanna, P.P. & Bloom, F.E. Neuroscience of addiction. Neuron 21, 467−476 (1998). | Article | PubMed | ISI | ChemPort |
- Lang, A.E. & Lozano, A.M. Parkinson's disease. First of two parts. N. Engl. J. Med. 339, 1044−1053 (1998). | Article | PubMed | ISI | ChemPort |
- Missale, C., Nash, S.R., Robinson, S.W., Jaber, M. & Caron, M.G. Dopamine receptors: from structure to function. Physiol. Rev. 78, 189−225 (1998). | PubMed | ISI | ChemPort |
- Gong, W., Neill, D.B., Lynn, M. & Justice, J.B., Jr. Dopamine D1/D2 agonists injected into nucleus accumbens and ventral pallidum differentially affect locomotor activity depending on site. Neuroscience 93, 1349−1358 (1999). | Article | PubMed | ISI | ChemPort |
- McNamara, F.N. et al. Congenic D1A Dopamine receptor mutants: ethologically based resolution of behavioral topology indicates genetic background as a determinant of knockout phenotype. Neuropharmacology 28, 86−99 (2003). | Article | ChemPort |
- Kelly, M.A. et al. Locomotor activity in D2 dopamine receptor-deficient mice is determined by gene dosage, genetic background, and developmental adaptations. J. Neurosci. 18, 3470−3479 (1998). | PubMed | ISI | ChemPort |
- Plaznik, A., Stefanski, R. & Kostowski, W. Interaction between accumbens D1 and D2 receptors regulating rat locomotor activity. Psychopharmacology (Berl.) 99, 558−562 (1989). | PubMed | ChemPort |
- Aizman, O. et al. Anatomical and physiological evidence for D1 and D2 dopamine receptor colocalization in neostriatal neurons. Nat. Neurosci. 3, 226−230 (2000). | Article | PubMed | ISI | ChemPort |
- Kimura, K., White, B.H. & Sidhu, A. Coupling of human D-1 dopamine receptors to different guanine nucleotide binding proteins. Evidence that D-1 dopamine receptors can couple to both Gs and Go. J. Biol. Chem. 270, 14672−14678 (1995). | Article | PubMed | ISI | ChemPort |
- Corvol, J.C., Studler, J.M., Schonn, J.S., Girault, J.A. & Herve, D. Golf is necessary for coupling D1 and A2a receptors to adenylyl cyclase in the striatum. J. Neurochem. 76, 1585−1588 (2001). | Article | PubMed | ISI | ChemPort |
- Liu, Y.F., Jakobs, K.H., Rasenick, M.M. & Albert, P.R. G protein specificity in receptor-effector coupling. Analysis of the roles of Go and Gi2 in GH4C1 pituitary cells. J. Biol. Chem. 269, 13880−13886 (1994). | PubMed | ISI | ChemPort |
- Senogles, S.E. The D2 dopamine receptor isoforms signal through distinct Gi proteins to inhibit adenylyl cyclase. A study with site-directed mutant Gi proteins. J. Biol. Chem. 269, 23120−23127 (1994). | PubMed | ISI | ChemPort |
- Sawin, E.R., Ranganathan, R. & Horvitz, H.R. C. elegans locomotory rate is modulated by the environment through a dopaminergic pathway and by experience through a serotonergic pathway. Neuron 26, 619−631 (2000). | Article | PubMed | ISI | ChemPort |
- Wintle, R.F. & Van Tol, H.H. Dopamine signaling in Caenorhabditis elegans—potential for parkinsonism research. Parkinsonism Relat. Disord. 7, 177−183 (2001). | Article | PubMed | ISI |
- Nass, R. & Blakely, R.D. The Caenorhabditis elegans dopaminergic system: opportunities for insights into dopamine transport and neurodegeneration. Annu. Rev. Pharmacol. Toxicol. 43, 521−544 (2003). |
|
standard error of the mean (s.e.m.) for two trials totaling at least 50 animals. DOP-3 was required for normal response to exogenous dopamine, whereas DOP-1 antagonized the effects of DOP-3.
