A Gain-of-function Mutation in Adenylate Cyclase Confers Isoflurane Resistance in Caenorhabditis elegans 

• Although volatile general anesthetics inhibit neurotransmitter release by a mechanism not fully understood, genetic evidence in Caenorhabditis elegans  has shown that a major mechanism of action of volatile anesthetics acting at clinical concentrations in this animal is presynaptic inhibition of neurotransmission

• An additional component of the presynaptic volatile anesthetic mechanism was demonstrated by showing that activation of neuronal adenylate cyclase antagonizes isoflurane inhibition of locomotion in C. elegans 

VOLATILE general anesthetics (VAs) have a complex set of actions on neurotransmission that summate to produce general anesthesia.1VAs promote inhibitory synaptic transmission by potentiation of γ-aminobutyric acid type A and glycine receptors and decrease excitatory transmission by multiple potential mechanisms. Transmitter release is reduced at both inhibitory and excitatory synapses by VAs, but VA potency and efficacy are greater at excitatory synapses.1Thus, the net presynaptic VA effect should be a decrease in central nervous system excitability. Biochemical and electrophysiologic evidence have implicated inhibition of sodium channels as one molecular mechanism whereby VAs inhibit neurotransmitter release.1However, blockade of sodium channels does not account for the entire inhibition of transmitter release, at least at some synapses2; rather, the transmitter release machinery that lies mechanistically downstream of the sodium channel is a good candidate as the residual VA target.

In the nematode Caenorhabditis elegans , we found that an unusual mutation in the unc-64  gene, which encodes C. elegans  neuronal syntaxin, fully blocked the behavioral effects of clinical concentrations of isoflurane (by this we mean concentrations that fall within the range used for human anesthesia, as much as two times the minimum alveolar concentrations of isoflurane, which produces 0.62 mM aqueous concentration; isoflurane EC⁵⁰against coordinated locomotion in C. elegans  at 22°C = 0.7–1 vol%; 1 vol% isoflurane at 22°C = 0.58 mM).3,–,5Syntaxin is one of three essential presynaptic SNARE proteins that acts in concert with other proteins to mediate fusion of synaptic vesicles with the presynaptic membrane.6The unusual mutation, designated unc-64  (md130 ), produces a truncated syntaxin that dominantly antagonizes VA sensitivity along with expressing reduced wild-type syntaxin, resulting in other phenotypes consistent with decreased excitatory transmitter release. Importantly, other unc-64  alleles with similarly decreased transmitter release phenotypes were hypersensitive to VAs; for example, unc-64(md130 ) has a 20–30-fold higher isoflurane EC⁵⁰than the otherwise phenotypically similar VA hypersensitive unc-64(md1259 ) and unc-64(js21 ) mutants.5Thus, the truncated syntaxin is not indirectly antagonizing VA sensitivity by reducing transmitter release; rather, the data are most consistent with interaction of the truncated syntaxin with another protein essential for VA sensitivity.

To identify the relevant syntaxin-interacting protein(s) essential for VA sensitivity, we tested C. elegans  mutants isolated in a screen for suppressors of syntaxin reduction-of-function phenotypes.7,8The logic of testing these mutants is that one or more of the suppressor mutations might lie in the putative syntaxin-interacting VA target and might be VA resistant. Indeed, we reported previously that some of these suppressors were VA resistant; however, the level of resistance was not as great as that in unc-64(md130 ) and was most likely attributable to an indirect effect on elevation of transmitter release.7Here we report on an additional syntaxin suppressor mutant whose level of resistance is similar to that of md130  and define in part the mechanism whereby it regulates VA sensitivity in C. elegans .

Except where noted, C. elegans  mutant strains were obtained from the Caenorhabditis  Genetics Center, which is funded by the National Institutes of Health–National Center for Research Resources (Bethesda, Maryland). Strains were grown as described previously on nematode growth media agar.9N2 var  Bristol was the wild-type strain and the genetic background for all mutants. Mutant strains used for this work are as follows: LGI:  unc-13(e376), unc-13(e376); acy-1(js127)III, unc-13(s69); acy-1(js127)III, unc-13(s69), gsa-1(ce81 ); LGIII:  unc-64(e246), acy-1(js127), acy-1(js127)/+, acy-1(js127) unc-64(e246), acy-1(js127)/+ unc-64(e246), acy-1(nu329), acy-1(js127); snb-1(md247)V, acy-1(js127); unc-10(md1117)X, crh-1(n3315), acy-1(js127) crh-1(n3315), unc-64(e246); jsEx558[acy-1(P260S ) Prol-6::GFP], unc-64(e246); jsEx570[acy-1(+ ) Prol-6::GFP], unc-64(e246 ); jsEx571 [acy-1(+ ) Prol-6::GFP], unc-64(e246 ); jsEx575 [acy-1(  L2444S) Prol-6::GFP], unc-64(e246); jsEx579 [Psnb-1::acy-1(P260S ) Prol- 6::GFP], unc-64(e246);  jsEx580[Pmyo-3::acy-1  (P260S) Prol-6:: GFP]; LGV:  snb-1(md247 ); LGX:  unc-10(md1117), kin-2(ce179 ); Extrachromosomal only:  jsEx558 [acy-1(P260S ) Prol-6::GFP], jsEx570 [acy-1(+ ) Prol-6::GFP], jsEx571 [acy-1(+ ) Prol-6::GFP], jsEx575 [acy-1(  L2444S) Prol-6::GFP], jsEx579 [Psnb-1::acy-1(P260S ) Prol-6::GFP], jsEx580 [Pmyo-3::acy-1(P260S ) Prol-6::GFP], jsEx676 [gsa-1  (+) Prol-6::GFP]. js127  double mutants with snb-1(md247 ), unc-10(md1117), unc-13(e376)  and unc-13(s69 ) were constructed using rbf-1(js232 )-marked Unc strains. rbf-1  is located at the center of chromosome III and was used to mark the non-js127  chromosome III. js127;him-8  males were mated into rbf-1 ;Unc hermaphrodites and hyperactive (js127  homozygous) progeny selected from the broods of heterozygous animals. The next generation was examined for nonhyperactive progeny (js127 ;Unc). Homozygosity of md247  and md1117  was verified molecularly, whereas js127  was deduced by the absence of js232  scored molecularly.

Germline transformation was accomplished by coinjecting the plasmid of interest, pBluescript carrier DNA (200 μg/ml), and the dominant transformation marker ^prol-6 ::GFP (pPHGFP1 at 20–35 μg/ml).10Transformants were selected by scoring green fluorescent protein (GFP) expression on an epifluorescence dissecting microscope, and stably transformed lines were isolated. jsEx570  and jsEx571  were obtained from injecting pPR1522 into unc-64(e246 ) at 15 μg/ml. A pAC2 (10 μg/ml) injection into unc-64(e246 ) yielded jsEx558 , and a pAC3 (15 μg/ml) injection into unc-64(e246 ) produced jsEx575 . pAC2 and pAC3 were also introduced independently into the balanced null allele acy-1(pk1279)/dpy-17(e164 ) to determine whether these altered forms were capable of rescuing null mutant lethality. pAC5 and pAC6 were each transformed into unc-64(e246 ) and the lines jsEx579  and jsEx580  isolated, respectively. Once stable lines were established, individual arrays were outcrossed to remove the unc-64(e246 ) mutation. jsEx676  was created by the coinjection of pPD118.33 and pRP1505.

pRP1522, a 14-kb genomic clone containing the acy-1  locus,11was obtained from Celine Moorman, Ph.D. (Hubrecht Laboratory, Centre for Biomedical Genetics, Utrecht, The Netherlands) and Ronald Plasterk, Ph.D. (Professor, Hubrecht Laboratory, Centre for Biomedical Genetics). The single base pair acy-1(js127 ) lesion was introduced into this clone using the DpnI-mediated site-directed mutagenesis protocol12to create a genomic clone (pAC2) encoding the P260S mutant form of ACY-1. Specifically, pRP1522 was mutagenized using OL#897 (ATTCAGTCTGTGATGTCT-AAAAAGGTACGCA) and OL#898 (TGCGTACCTTTTTAGACATCACAGACTGAAT). Similarly, pRP1522 was mutagenized using OL#955 (TTGGCAAGAAAGGATTCTGAGTTGGAGACACAG) and OL#956 (CTGTGTCTC-CAACTCAGAATCCTTTCTTGCCAA) to create pAC3, an acy-1  genomic clone harboring an L244S lesion modeled after the constitutively active adenylate cyclase mutation isolated in Dictyostelium .13Both constructs were sequenced to verify proper introduction of the desired lesion. Using this same mutagenesis method, a BamHI restriction site was engineered before the acy-1  initiation site by mutagenizing pAC2 using OL#1026 (TCTT-GTCTTCTGGATCCATGGACGACGATGT) and OL#1027 (ACATCGTCGTCCATGGATCCAGAAGACAAGA) to produce pAC4. The acy-1  promoter was then replaced with the snb-1  synaptobrevin promoter to yield a construct, pAC5, for selective expression of the P260S form in neurons. This was accomplished by swapping the SacII-BamHI fragment of pAC4 with the SacII-BamHI fragment of the polymerase chain reaction (PCR) product resulting from OL#1028 (GTTAGTATCATTC-GAAACATACC) and OL#1029 (AGCTTTCCGCGGAAATC-TAGG) amplification of the snb-1  promoter from pRM248. pAC6 was created in the same manner with OL#1028 and OL#1030 (GCTGCGGCCGCGGGTCGGC) used to amplify the myo-3  promoter from pPD96.52 to study muscle selective expression. pPD118.33 is a Pmyo-2 :GFP, and pRP1505 is a wild-type gsa-1  construct and was obtained from Ronald Plasterk, Ph.D.14 

A nonclonal genetic screen was performed using conventional mutagenesis.15,unc-64  (e246 ) L4-staged hermaphrodites were mutagenized for 4 h in 50 mM ethyl methanesulfonate. To recover recessive mutants, second generation self-progeny were examined for mobile animals. From a given plate of approximately 50–150 F1 progeny, at most one F2 candidate suppressor was selected and clonally passaged. Suppression was verified in the next generation. and subsequent backcrossing was initiated. This was generally performed by mating with wild-type males; F1 double heterozygous males were then crossed into unc-64(e246 ) and several Unc progeny hermaphrodites clonally passaged. The next generation was screened for moving animals to reticulate the unc-64  Sup double mutants. After at least two rounds of backcrossing, suppressors were outcrossed from unc-64  (e246 ) to obtain the single mutant suppressor. Presence of the suppressor was verified in the strain by reintroducing the unc-64  (e246 ) allele and showing retention of suppression of paralysis. In seven rounds of screening a total of 24,000 haploid genomes, 14 suppressors were recovered. Other suppressors isolated in the screen have been described elsewhere.7,8 

All mapping and complementation tests were performed using standard genetic methods.16Suppressors were grouped initially based on their behavioral phenotype and subsequently placed into complementation groups by complementation assays. js127  was a single allele isolate and was placed on chromosome III based on linkage to lon-1 . Specifically, js127 e246  males were mated into lon-1 e246  hermaphrodites, resulting in js127 e246/lon-1 e246  non-Lon cross progeny. These cross progeny were e246  homozygous but were phenotypically non-Unc because of single copy js127  (i.e., js127  acted dominantly to suppress the e246  Unc phenotype). From this transheterozygote, 7 of 31 Lon progeny also were phenotypically non-Unc through acquisition of a single copy of js127  by recombination. This localization was refined using three-factor mapping with lon-1  and dpy-18. lon-1 dpy-18 unc-64  was placed over js127 e246 , and all 18 Lon non-Dpy recombinants failed to segregate js127  placing js127  close by or to the left of lon-1 . Fine structure mapping was performed using the single nucleotide polymorphisms of the Hawaiian strain, CB4856.17Single nucleotide polymorphisms that altered restriction enzyme recognition sites were chosen for analysis because they could be scored simply by PCR amplification of that genomic region and subsequent restriction enzyme digestion. First, a dpy-1 daf-2 js127 lon-1 unc-64  strain was constructed by placing dpy-1(e1) daf-2(e1370ts ) in trans to js127 lon-1 unc-64 . 122 Sup Lon Unc progeny were passaged clonally, and nine plates contained Dpy Daf progeny. Dpy Daf animals were then raised at the permissive temperature, 20°C for daf-2 , to assess locomotion to verify the presence of js127 . Then, CB4856 males were mated into this strain, and the resulting cross progeny males were mated back into dpy-1 daf-2 js127 lon-1 unc-64  hermaphrodites, to obtain at most one recombinant chromosome per progeny. Progeny of this cross were screened for Lon non-Daf animals, which would only result from a single recombination event between daf-2  and lon-1 . Two hundred thirty-nine recombinants were isolated, and then each recombinant chromosome was homozygosed and scored for js127  by its Sup phenotype. The location of each recombination event was determined by scoring the genotype of each single nucleotide polymorphisms. PCR primers and location of single nucleotide polymorphisms are available upon request. Examination of the Sup Lon Unc class of recombinants placed js127  to the right of cosmid F10F2, whereas information from the Lon Unc class of recombinants positioned js127  to the left of cosmid C35D10. This region of approximately 273 Kb contained an estimated 82 genes, one of which was an adenylate cyclase gene, acy-1 . Sequencing of the acy-1  open reading frame from js127  revealed a C > T single nucleotide change resulting in a P260S lesion in the protein. A PCR-digestion assay was developed to score the presence of this lesion molecularly. PCR amplification of a 210-bp product using OL#872 (TCTTGAAGAGG-CCGGATACATT) and OL#896 (AAAATGCATGCGTAGCCTTTTAG) followed by digestion with Bgl I resulted in a restriction pattern of 192 bp and 18 bp from wild type and a 210-bp undigested fragment from js127 .

Locomotion assays were performed at room temperature (20–22°C) on at least 20 young adult hermaphrodites by collecting serial charge-coupled device camera images with an LG3 frame grabber (Scion Corporation, Frederick, MD) every 2.5–5 s at a magnification between 0.5× and 0.8×. Plates were undisturbed on the microscope for 5–10 min before imaging was initiated. A series of images of basal locomotion were collected before dropping a metal rod from a constant height onto the plates to serve as a mechanical stimulus to excite the animals.18A similar series was collected after this mechanical stimulus. Locomotory velocity was calculated between successive images by measuring the linear displacement in the position of the tail of each animal. Velocities over four consecutive images were calculated and averaged to assess both basal and stimulated locomotion. Assays were performed in triplicate.

Acute sensitivity to the acetylcholinesterase inhibitor aldicarb was assayed by transferring 20–25 animals to plates containing aldicarb and monitoring the time course of animal paralysis.19Animals were counted as paralyzed if they appeared hypercontracted and failed to move even if prodded with a platinum wire. Aldicarb, 2-methyl-2-[methylthio] proprionaldehyde O-[methylcarbamoyl]oxime, was obtained from Chem Services, Inc. (West Chester, PA) and prepared as a 100-mM stock solution in 70% ethanol. Aldicarb was added to the nutrient growth medium agar after autoclaving.

Isoflurane dispersal assays were performed at 22–24°C using young adult hermaphrodites, as described previously.20Worms were transferred in S-basal buffer to 9.5-cm agar dispersal plates seeded at their edge with Escherichia coli  bacteria. Dispersal plates were then placed in various atmospheric concentrations of isoflurane (measured subsequently by gas chromatography) and the animals allowed to disperse. The fraction of adults (approximately 50 per assay plate) present in the bacterial ring divided by the total number of adults after 40 min was scored as the dispersal index.

Endogenous cAMP concentrations were measured from young adult animals using a competitive binding enzyme-linked immunosorbent assay (Amersham Biosciences, Piscataway, NJ).21Wild-type, acy-1(nu329 ), and acy-1(js127 ) animals were grown on the adenylate-cyclase–deficient E. coli  strain DHP1F-glnV44(AS) recA1 endA1 gyrA96(NalR) thi1 hsdR17 spoT1 rfbD1 cyaA.22Synchronized cultures of adults were obtained using standard methods. Approximately 200 μl packed adults were resuspended in 1 ml lysis buffer and sonicated with three 20-s pulses using a microtip. Protein concentrations were measured using a Bradford assay (Biorad, Hercules, CA). Assays were performed in triplicate.

Concentration/response curves were fit by nonlinear regression using the equation: y = min + (max − min)/(1 + ([Iso]/EC⁵⁰)^−k). The minimum was constrained to 0. The EC⁵⁰s were used as the measure of the isoflurane sensitivity of the strains. EC⁵⁰s were compared for statistical differences by simultaneous curve fitting, as described by Waud23using GraphPad Prism 5 Software (GraphPad, Software, Inc., San Diego, CA). The error values following the EC⁵⁰values are the error of the fit. Error values for cAMP concentrations were SD of triplicate assays. Error values for aldicarb assays were SEM of triplicate assays. Locomotion rates and cAMP concentrations were compared by two-sided t  test. The time at which half of the animals were paralyzed in the aldicarb paralysis assays was compared for statistical differences by simultaneous curve fitting using GraphPad Prism 5 Software. Statistically significant differences were at the P < 0.05 level. For multiple comparisons, the significance threshold was less than 0.05 per number of comparisons.

The js127  mutation was isolated in a screen for mutations that improve the locomotion of the unc-64  syntaxin reduction-of-function allele, e246. js127  strongly suppressed the slow uncoordinated locomotion of unc-64(e246 ) (fig. 1A, B). Indeed, after stimulation, the js127 e246  double-mutant strain moved at speeds indistinguishable from wild-type animals and was the strongest suppressor mutant isolated in the screen (fig. 1B). To test whether this suppression of locomotion was associated with an increase in acetylcholine release, aldicarb sensitivities were measured. Aldicarb is an acetylcholinesterase inhibitor that is widely used to measure the levels of cholinergic transmission in C. elegans  mutants with the caveat that aldicarb sensitivity is an indirect measure of transmitter release and only assays acetylcholine release.19Mutants with a decrease in acetylcholine release are more resistant to paralysis by aldicarb, and this can be conveniently measured by kinetic assays. js127 e246  was significantly less resistant to aldicarb than was unc-64  (e246 ) (fig. 1C), consistent with an enhancement of syntaxin's function to mediate synaptic vesicle fusion and transmitter release. The js127  mutation was outcrossed from unc-64(e246 ), and its isoflurane sensitivity was measured. js127  was strongly resistant to isoflurane with an EC⁵⁰more than three times greater than that of the wild-type strain and fully resistant to concentrations of isoflurane in the clinical range (fig. 1D).

To identify the genetic lesion responsible for the phenotypes in js127 , the suppression of the sluggish locomotion phenotype of e246  was mapped genetically. The suppression phenotype mapped to a 273-Kb interval on the left arm of chromosome III (fig. 2A). The C. elegans  genome sequence predicts 82 genes in this interval, one of which is acy-1 , which encodes an adenylate cyclase previously shown to regulate neurotransmitter release in C. elegans  and, by its closest homologs, in mammals.24,–,26The acy-1 gene was sequenced in the js127  mutant and a C > T transition mutation was found in codon 260, resulting in proline to serine missense lesion (fig. 2B). Proline 260 lies within a highly conserved region at the N-terminal end of the C1a domain, one of the cytoplasmic catalytic domains. To confirm that the P260S mutation was indeed responsible for the js127  phenotypes, an acy-1  genomic plasmid was constructed with the P260S mutation, and transgenic animals expressing the plasmid were generated. The P260S transgene strongly suppressed the locomotion defects of unc-64(e246 ) to levels similar to those of the js127e246  (fig. 2C). Likewise the transgene in the absence of unc-64(e246 ) conferred high-level isoflurane resistance, actually greater than that in js127 [P < 0.007 vs. acy-1  (js127 )] (fig. 2D). Based on the mapping, identification of a lesion, and phenocopy by transformation, we conclude that the C > T transition resulting in a P260S change in acy-1  is the js127  mutation. Confirming our assignment of js127  to acy-1 , this identical ACY-1(P260S) lesion was independently isolated in another laboratory in a similar screen for suppressors of a different mutation that reduces neurotransmitter release in C. elegans .25This suggests that relatively few mutations in acy-1  result in this phenotype.

The ability to reproduce the phenotypes of js127  by transformation suggests that the mutation confers a gain of function to ACY-1. Consistent with this hypothesis, js127  heterozygotes significantly suppress the sluggish locomotion of unc-64  (e246 ) (fig. 3A). js127  also dominantly confers an aldicarb hypersensitivity phenotype, whereas the reduction-of-function allele acy-1  (nu329 )27is aldicarb resistant (fig. 3B). Likewise for isoflurane sensitivity, the acy-1  (nu329 ) allele has an isoflurane hypersensitive phenotype, opposite that of acy-1  (js127 ) (fig. 3, C and D). However, unlike transgenic expression of ACY-1(P260S), transgenes expressing additional wild-type ACY-1 or ACY-1 with a mutation previously shown to confer constitutive activity on Dictyostelium  adenylate cyclase13were not isoflurane resistant, nor did these transgenes suppress slow locomotion of unc-64  (e246 ) phenotypes (fig. 3, D and E). Thus, P260S appears to be a particularly strong gain-of-function mutation. To test directly the hypothesis that js127  was a gain-of-function mutation, we compared whole animal cAMP concentrations in acy-1  (js127 ), acy-1  (nu329 ), and wild-type animals. Consistent with the genetic data, cAMP concentrations were significantly higher in js127  than in wild-type animals and significantly lower in nu329  (fig. 3F). Thus, we conclude that js127  confers an increase in ACY-1 adenylate cyclase activity.

ACY-1 is expressed throughout the C. elegans  nervous system and in body wall muscles.14,27Thus, it is possible that enhanced ACY-1 activity in muscle cells, rather than neurons, is responsible for the js127  phenotypes. To test this hypothesis, acy-1  (P260S) was expressed selectivity in neurons or muscle using cell-type–specific promoters. acy-1  (P260S) driven by the pan-neuronal promoter Psnb-1  strongly suppressed the slow locomotion of unc-64  (e246 ), whereas expression in muscle with the Pmyo-3  promoter produced no discernible suppression (fig. 4A). Similarly, panneuronal acy-1  (P260S) produced high-level resistance to isoflurane, whereas the isoflurane sensitivity of the muscle acy-1  (P260S) was similar to that of wild type (fig. 4, B and C). We conclude that ACY-1 adenylate cyclase acts in neurons to suppress the syntaxin mutant phenotype and regulate isoflurane sensitivity.

To define the pathway whereby ACY-1 regulates transmitter release and isoflurane sensitivity, we tested the phenotypes of mutations in genes that might lie in the pathway. Adenylate cyclase normally is stimulated by Gαs. C. elegans  has one Gαs gene, gsa-1 , which has been shown to promote cholinergic transmitter release and neurodegeneration.14,24,25,27,28We found that similar to acy-1(js127 ), an activating mutation, ce81 ,24in gsa-1  was strongly resistant to isoflurane (fig. 5A). Likewise, animals transformed with additional copies of wild-type gsa-1  were also isoflurane resistant (fig. 5A). Protein kinase A (PKA) is a classic downstream target of adenylate cyclases and has been implicated in ACY-1 signaling.25,29We tested a loss-of-function allele of kin-2 , which encodes a negative regulatory subunit of PKA. Consistent with ACY-1 signaling through PKA to regulate isoflurane sensitivity, the kin-2  loss-of-function mutant was strongly isoflurane resistant (fig. 5A). The cAMP response element binding protein (CREB) is a transcription factor that can be activated by PKA phosphorylation and regulates the expression of numerous genes.30CREB is most clearly implicated in synaptic plasticity and neural development but also has been show to promote the expression of presynaptic syntaxin.31Thus, we considered the hypothesis that ACY-1 might promote synaptic transmission and reduce isoflurane sensitivity by activating CREB. However, a null mutation in the only C. elegans  homolog of CREB,32 crh-1 , had normal sensitivity to isoflurane and did not suppress the isoflurane resistance of js127  in the acy-1  (js127 ) crh-1  (null) double mutant (fig. 5A). The crh-1  (null) mutant was resistant to aldicarb, consistent with the hypothesis that crh-1  does promote cholinergic neurotransmission (fig. 5B); however, as for isoflurane resistance, the crh-1  (null) mutant did not suppress the aldicarb hypersensitivity phenotype of acy-1  (js127 ). A notable caveat to attributing the aldicarb-resistant phenotype to the crh-  1(null) mutant is that only one mutant was tested and the phenotype was not rescued by transformation. Thus, the aldicarb resistance could be attributable to an unknown background mutation. However, the data definitively show that C. elegans  CREB does not act downstream of ACY-1 to control neurotransmission and isoflurane sensitivity.

Finally, we tested for suppression of js127  phenotypes by reduction of function mutations in three transmitter-release machinery proteins: UNC-10, rab-3-interacting molecule (RIM); SNB-1, synaptobrevin; and UNC-13, mUNC13. For isoflurane resistance and aldicarb sensitivity, the mutations in all three genes strongly suppressed js127  (fig. 5, A and C, D, E). Thus, js127  adenylate cyclase activation does not bypass the core vesicular fusion machinery to produce VA resistance or enhance transmitter release. However, for locomotion rate, only the locomotion of a strong unc-13  allele (s69 ) was not improved by js127  (fig. 5F). unc-13(e376 ), a weaker allele, still moved significantly better in the background of js127  (fig. 5F). Similarly, the locomotion rates of both snb-1  partial loss of function and unc-10  null mutants were improved significantly in a js127  mutant background. Thus, UNC-10 and perhaps SNB-1 (the epistatic relationship of SNB-1 to ACY-1 is not definitive given the snb-1  allele is not null) are not required for the locomotion-promoting activity of ACY-1. By contrast, UNC-13, at least at the level of sensitivity of these assays, is epistatic to acy-1  (js127 ).

Through screening of mutations that suppress the phenotypes of a syntaxin reduction of function mutant, we have identified a gain-of-function mutation of C. elegans  (ACY-1 adenylate cyclase) that strongly antagonizes isoflurane sensitivity. Our data are consistent with an ACY-1 signaling pathway as shown in figure 6. With regard to anesthetic mechanisms, the most central question posed by this study is whether ACY-1 is an anesthetic target and the js127  mutation directly blocks volatile anesthetic inhibition of ACY-1. This hypothesis seems unlikely in light of our previous findings. Although not as VA resistant as acy-1(js127 ), other mutants that suppress unc-64(e246 ) are also VA resistant.7In general, we have found that environmental conditions or mutations such as acy-1(js127 ) that enhance neurotransmitter release confer VA resistance.7,20,33Likewise, mutants with reduced neurotransmission have been found to be hypersensitive to VAs.5,20,33Thus, the anesthetic phenotype of acy-1  (js127 ) is most easily explained as being attributable to indirect enhancement of the process that VAs block.

In C. elegans  only the truncated syntaxin and unc-13  mutants have been found to deviate from the correlation between the levels of neurotransmitter release and VA resistance.5,34The truncated syntaxin acts in a dominant fashion to block VA effects on transmitter release without otherwise detectably altering behavior or neurotransmission.5,34The VA resistance of the syntaxin mutant can be suppressed by overexpression of wild-type UNC-13, consistent with a model where the truncated syntaxin in a dose-dependent mechanism blocks VA inhibition of UNC-13 activity. The unc-13  mutants, despite having reduced transmitter release, were also VA resistant, and a strain with a membrane-targeted UNC-13 was VA resistant, suggesting the model that VAs block membrane association of UNC-13.34Thus, we have previously proposed that UNC-13 is a presynaptic target for clinical concentration of VAs in C. elegans .34 

Might UNC-13 be a direct target of PKA and thereby offer a testable hypothesis for the unusually strong VA resistance of acy-1  (js127 )? Indeed, among the release machinery mutants tested, only the strong unc-13  allele, s69 , was found to be incompetent for js127  suppression of its uncoordinated locomotion. However, little spontaneous or evoked exocytosis is detected from cholinergic unc-13(s69 ) motor neurons by electrophysiologic assays.35,–,37Thus, although the formal interpretation of our genetic epistasis experiments is that UNC-13 lies downstream of ACY-1, this result may derive from the fact that unc-13(s69 ) has essentially no transmitter release for ACY-1 to enhance, rather than UNC-13 being the direct target of the GSA-1–ACY-1–PKA pathway. In addition, UNC-13-mUNC13 has not been reported to be a PKA target. It is more likely that PKA phosphorylates some intermediate target whose activity requires UNC-13. UNC-13 is a diacyl glycerol-binding presynaptic protein that interacts with syntaxin, RIM, calmodulin, and other presynaptic proteins to promote neurotransmitter release.38A reasonable candidate PKA target is UNC-10 RIM. In mammals, PKA has been shown to phosphorylate RIM, and this phosphorylation is necessary for PKA-dependent presynaptic long-term potentiation in mouse cerebellar neurons.39,–,41RIM interaction with mUNC13 is necessary for normal synaptic vesicle priming in mouse hippocampal neurons,42and RIM binding to mUNC13 has been shown to reduce the concentrations of mUNC13 homodimers, which are autoinhibitory.43In addition to disinhibition of mUNC13, RIM has been shown to promote presynaptic localization of P- and Q-type calcium channels near the active zone and interact with other presynaptic proteins, including Rab3; it also may serve a scaffolding function.39,44,45However, in C. elegans , if UNC-10 RIM is the ACY-1/PKA target, it is not an essential target because acy-1  (js127 ) is capable of significantly improving the locomotion of an unc-10  null mutant. Likewise for the VA presynaptic mechanism, UNC-10 is nonessential because unc-10  null mutants are normally sensitive to isoflurane.34 

An alternative or additional ACY-1-PKA mechanism consistent with UNC-13 as the VA target is regulation of proteasome-dependent degradation of UNC-13. In Drosophila  neurons, synaptic DUNC-13 (Drosophila  UNC-13) concentrations were found to be positively regulated by cAMP and PKA.46Inhibition of cAMP-PKA signaling resulted in a rapid and substantial decrease in DUNC-13 concentrations at the synapse, and this decrease could be blocked with proteasome inhibitors. However, the mechanism whereby the cAMP-PKA pathway regulates the apparent proteasomal degradation of DUNC-13 is obscure.

Is the C. elegans  presynaptic VA mechanism described here relevant to the mammalian anesthetic mechanism? As stated, presynaptic inhibition of excitatory neurotransmitter release has been demonstrated in a variety of mammalian models.1Thus, a contribution of presynaptic anesthetic effects to general anesthesia seems likely. The presynaptic machinery in C. elegans  is highly conserved in humans,6,47and the VA concentrations to which the mutants in the presynaptic machinery are conferring resistance are in the clinical range. Thus, the C. elegans  presynaptic VA mechanism further elaborated here might reasonably contribute to general anesthesia in mammals. Experimental support for this conjecture has been reported. A truncated syntaxin based on the C. elegans  VA-resistant mutant was expressed in a rat neuroendocrine cell line and in hippocampus and found to antagonize the effects of clinical concentrations of isoflurane on neurosecretion and transmitter release.48These results argue that at least some aspects of the C. elegans  presynaptic mechanism are conserved in higher organisms.

An important issue to consider when discussing the potential relevance of the proposed C. elegans  presynaptic anesthetic mechanism to mammalian anesthesia is how to reconcile the fundamental function of UNC-13 orthologs with the differential inhibition by volatile anesthetics of transmitter release from distinct neuronal subtypes. In other words, if UNC-13 orthologs function at all synapses and are important presynaptic anesthetic targets, how might volatile anesthetics more potently inhibit excitatory synapses compared with inhibitory ones, as previously shown?1The mammalian UNC-13 homologs mUNC13-1, 2, and 3 have distinct functional roles that could contribute to the synapse-selective effects of VAs. Release from the majority of glutamatergic terminals in mouse hippocampus requires the mUNC13-1 isoform, whereas mUNC13-1 and mUNC13-2 function redundantly in γ-aminobutyric acid–mediated release, at least in the cerebral cortex and hippocampus.49In rat brain, mUNC13-1 is expressed throughout the central nervous system, whereas mUNC13-2 expression is restricted to the cerebral cortex and hippocampus. mUNC13-3 appears to be expressed exclusively in the cerebellum.50Intriguingly, mUNC13-1– and mUNC13-2–mediated release in mouse differs in their potentiation by diacyl glycerol; mUNC13-1 is less efficaciously potentiated.49These observations suggest the hypothesis that mUNC13-1, the closest homolog to C. elegans  UNC-13, may be more sensitive to VAs because its weak DAG potentiation is more efficaciously blocked by VAs than is that of mUNC13-2. Testing of this hypothesis is experimentally feasible with the availability of mouse knockout strains for each of the mUNC13 isoforms.