Introduction

Ethanol is a widely used beverage. Its abuse causes severe damage to human health and creates serious social problems, including an increased risk of cardiomyopathy, liver complaints, traffic accidents, loss of work and learning problems1,2,3. Alcohol alters the behaviour of both invertebrates and vertebrates. Upon exposure to ethanol, humans, rodents and Drosophila exhibit responses to ethanol ranging from disinhibition and euphoria at low doses to incoordination, lethargy and even death at higher doses4,5. In Caenorhabditis elegans (C. elegans), ethanol intoxication suppresses locomotion, swimming and egg laying6. A number of identified genes contribute to alcoholism7,8. Ethanol acts on a wide spectrum of target sites. Ethanol inhibits NMDA9, AMPA10 and kainate11 receptors. It enhances GABAA12 and glycine receptors13,14, potentiates ion currents through the ligand-gated ion channel of the 5-hydroxytryptamine-3 (5-HT3)-type serotonin receptor15,16 and modulates the neuronal acetylcholine receptor17. Ethanol acts on a calcium-activated large conductance BK potassium channel, SLO-1, to inhibit neuronal activity and plays a key role in alcohol intoxication in the nematode C. elegans6,18.

Animals develop resistance to toxins. There are two general routes for evolving resistance to a toxin. One is to evolve a reduced sensitivity to the toxin’s effects. The second route is to minimize the amount of toxin that reaches the target organ or tissue. The underlying mechanisms include detoxification, excretion, sequestration and reduced absorption19. C. elegans develops tolerance to ethanol. Multiple genes that control acute sensitivity to ethanol have been identified in C. elegans7,20. The cAMP-PKA signal transduction pathway functions in ethanol tolerance in Drosophila5. nNOS/NO/cGMP/cGKII signalling positively regulates sensitivity to the sedative hypnotic effects of a high ethanol dose in mice21,22. However, it is still important to identify novel genes and signalling pathways that are involved in acute ethanol tolerance.

Due to feasible genetic and neuronal manipulations, rich genetic tools and vertebrate neuronal protein homologues, the nematode C. elegans is a useful model animal to identify novel ethanol target sites and molecular bases for ethanol tolerance. Ethanol affects C. elegans physiological activities and behaviours. The worms exhibit uncoordinated behaviour and intoxication when exposed to 100 mM (~0.6%) and 500 mM (~3%) ethanol, respectively6. Several molecules, including SLO-1, NPR-1 and the switching defective/sucrose nonfermenting (SWI/SNF) chromatin-remodelling complex, are involved in modifying acute tolerance to ethanol in C. elegans6,23,24. NPR-1 negatively modulates acute ethanol tolerance. The C. elegans strains with a lower function or loss-of-function alleles recover more quickly from exposure to ethanol than N2 animals23. N2 animals express a high-activity form of the NPR-1 receptor that suppresses output of a circuit detecting hyperoxia25 mainly by blocking output of RMG interneurons26. Here, using in vivo Ca2+ imaging, behavioural tests and chemogenetic experiments, we present evidence that suggests that the GCY-35/GCY-36—TAX-2/TAX4 signalling pathway in O2 sensory neurons URX, AQR and PQR, mediates the ethanol excitatory effect on these neurons and functions in acute functional ethanol tolerance in npr-1 animals.

Results

GCY-35/GCY-36—TAX-2/TAX-4 signalling pathway is involved in acute functional ethanol tolerance in npr-1 worms

N2 and npr-1 animals show different rates of recovery from ethanol exposure23. A previous work has shown that in N2 animals, the high-activity form of the NPR-1 receptor blocks output from the AQR, PQR, and URX O2 sensing neurons25. We speculated that the different rates at which N2 and npr-1 animals recovered from ethanol exposure reflect differences in the functions of this O2 sensing circuit. To test this possibility, we examined the behaviour of mutants defective in the function of this circuit, including gcy-35 and gcy-36. We recorded the locomotion of acute ethanol-treated worms on agar plates as described6,23,27. In brief, we transferred worms deprived of food for 30 min to the test NGM plates that were bacteria unseeded and ethanol added (final concentration, 500 mM) and then recorded the worms’ movement over 2 min beginning at 10 and 30 min using a CCD camera at 1 frame/s. The relative locomotion speeds (treated/untreated) at 10 min and 30 min of exposure to an exogenous dose of 500 mM ethanol were used as indicators of sensitivity to and acute adaptation or tolerance to ethanol as previous reports suggested6,23,27. Our results showed that almost all single gene mutants of gyc-35, gcy-36, tax-2 and tax-4 exhibited ethanol sensitivity and acute functional ethanol tolerance similar to N2 worms (Supplementary Fig. 1). npr-1(ad609) showed increased acute functional tolerance (Fig. 1A, Supplementary Movies 1–4) as reported previously23. Interestingly, npr-1(ad609); gcy-35(ok769) double mutants moved slower than npr-1(ad609), similar to N2 worms at 30 min. Sensitivities to ethanol (indicated by the relative locomotion speed at 10 min) in all strains were similar (Fig. 1A, Supplementary Movies 5 and 6). The gene gcy-35 is expressed in URX, AQR and PQR sensory neurons28. We used the gcy-32 promoter29 to drive gcy-35 cDNA expression specifically in these three types of neurons in the double mutant. The extrachromosomal expression of gcy-35 cDNA in these neurons fully rescued the locomotion phenotype in the double mutants (Fig. 1A, Supplementary Movies 7 and 8). The gene gcy-36 co-expresses with gcy-35 in URXs, AQR and PQR28. GCY-35 and GCY-36 form a soluble guanylyl cyclase (sGC) heterodimer in C. elegans30. Similar to npr-1(ad609); gcy-35(ok769) double mutant, npr-1(ad609); gcy-36(db42) double mutant showed a reduced relative speed compared with that in npr-1(ad609) and similar to that in N2 worms (Fig. 1B). These results suggest that gcy-35/gcy-36 is involved in acute ethanol tolerance in C. elegans. The sGC GCY-35/GCY-36 complex catalyses the biosynthesis of cGMP, and the latter acts on the downstream effector, the TAX-2/TAX-4 channel, to play a role in chemosensations in C. elegans25,31,32,33. We then examined the role of tax-2/tax-4 in acute functional ethanol tolerance. Our data showed that npr-1(ad609); tax-4(p678) and npr-1(ad609); tax-2(p671) double mutants manifested a slower speed compared with that of the npr-1 mutant, similar to that in N2 worms (Fig. 1C,D). The extrachromosomal expression of tax-4 cDNA driven by the gcy-32 promoter markedly restored the npr-1 locomotion phenotype in the npr-1(ad609); tax-4(p678) double mutant (Fig. 1C). Collectively, these results suggest that the GCY-35/GCY-36—TAX-2/TAX-4 signalling pathway likely function in URX, AQR and PQR neurons and is required for the acute tolerance in npr-1 animals.

Figure 1
figure 1

The genes encoding GCY-35/GCY-36 and TAX-2/TAX-4 are required for  acute functional ethanol tolerance in npr-1 worms. (A–D) The relative locomotion speed (ethanol treated/untreated) at 10 and 30 min after ethanol exposure in (A) wild type N2, npr-1(ad609), npr-1(ad609); gcy-35(ok769) mutant and the gcy-32p::gcy-35 transgene, the same worm strains are not listed hereinafter, (B) npr-1(ad609); gcy-36(db42), etc., (C) npr-1(ad609); tax-4(p678), gcy-32p::tax-4 transgene, etc., (D) npr-1(ad609); tax-4(p678), etc. The data are analyzed by two-way ANOVA. In (A), F Time (1, 75) = 75.63, P < 0.0001; F genotype (3, 75) = 7.667, P < 0.0001; F genotype * Time (3, 75) = 11.86, P = 0.0002. In (B), F Time (1, 49) = 50.35, P < 0.0001; F genotype (2, 49) = 12.53, P < 0.0001; F genotype * Time (2, 49) = 8.229, P = 0.0008. In (C), F Time (1, 58) = 61.91, P < 0.0001; F genotype (3, 58) = 16.08, P < 0.0001; F genotype * Time (3, 58) = 8.584, P < 0.0001. In (D), F Time (1, 43) = 55.93, P < 0.0001; F genotype (2, 43) = 18.66, P < 0.0001; F genotype * Time (2, 43) = 16.26, P < 0.0001. In (A–D), significant posttest comparisons with npr-1(ad609) at any given time point are indicated with ***p ≤ 0.001. In (A,C), significant posttest comparisons with double mutant strain at any given time point are indicated with ###p ≤ 0.001.

Full size image

NPR-1 and the oxygen-sensitive sGC GCY-35/GCY-36 complex modulate worm locomotion speed in different oxygen levels. Hawaiian wild strain CB4856 and npr-1(ad609) mutants move faster at high ambient oxygen levels (21%) than the N2 worm does25,34. To test the possible impact of high ambient oxygen levels on the locomotion of ethanol-treated worms, we developed a simple device to make a low ambient oxygen level and examined worm locomotion under this condition (Supplementary Fig. 2A, see methods for details). The O2 levels were assayed by measuring O2-induced quenching of the fluorescent dye Ru(phen)3Cl235. After infusing a mixed gas of 7% [O2] for nearly 12 min, the O2-level reduced to 7% in the device (Supplementary Fig. 2B). The npr-1(ad609) mutant moved slower under low ambient O2 levels than in air (Supplementary Fig. 2C) as expected, in agreement with a previous report34. This result verifies that this simple device worked well for the tests under a low oxygen level. We recorded the worm locomotion speed under low and high ambient O2 levels. Our results showed that the O2 levels had no significant impact on relative (Supplementary Fig. 2D) and absolute speeds in all ethanol (500 mM)-treated (Supplementary Fig. 2F) and untreated (Supplementary Fig. 2E) worms at 30 min, although ethanol treatment slowed the worms. These results suggest that the O2 levels have no effect on the movement in short-term food-deprived worms, regardless of whether the worms were ethanol treated or untreated.

Ethanol activates URX and AQR sensory neurons

Our data indicate that the GCY-35/GCY-36—TAX-2/TAX-4 signalling pathway in URXs, AQR and PQR plays a role in acute functional ethanol tolerance in C. elegans. A previous study reported that optogenetically activating these three types of neurons induces worm to move faster34. We hypothesized that these neurons sense ethanol and mediate the acute functional ethanol tolerance effect. To test this possibility, we used the genetically encoded calcium sensor R-GECO136 combined with a microfluidic device to record in vivo somatic calcium responses to ethanol in these neurons. In brief, a transparent polydimethylsiloxane (PDMS) microfluidic device was used to immobilize and expose worms to ethanol stimulation31,37,38 (see methods for details). We recorded Ca2+ transients of neuronal soma in response to an ethanol concentration of 500 mM, also used in the behavioural assay. As expected, URXs in wild-type N2 worms and npr-1(ad609) mutants displayed robust Ca2+ responses to 500 mM ethanol stimulation, and the responses in npr-1(ad609) mutants were stronger than those in the N2 worms (Fig. 2A). This may explain why npr-1 animals show acute functional ethanol tolerance compared with N2 worms. The AQR neuron also displayed robust Ca2+ responses to the 500 mM ethanol stimulation (Supplementary Fig. 3). PQR neuron localizes in the worm’s tiny tail. It is not easy to immobilize the worm tail. We did not record the Ca2+ response in this neuron. As URXs exhibited more robust Ca2+ signals and had almost no delay, we recorded URXs only thereafter.

Figure 2
figure 2

O2 Sensory neuron URXs are activated by ethanol. (A) Ethanol (500 mM) evoked somal calcium signals of URXs in wild type N2 worm and npr-1(ad609) mutant. The left panels show mean intensities of Ca2+ signals (in solid traces) with S.E.M. (in grey), the right panels display summaries of the peak (12 s–22 s) signals, and the number of independent tests for each genotype or each concentration of ethanol is indicated, similarly hereinafter. ***p ≤ 0.01 compared with wild type N2 worm. (B–D) Somal Ca2+ levels of URXs in responses to 500 mM ethanol in animals (B) unc-31(e928), (C) unc-13(e1091) and (D) unc-9(fc16) mutants. (E) Somal Ca2+ levels of URX in responses to ethanol of different concentrations in npr-1(ad609) mutant.

Full size image

It is possible that signals transmitted from other neurons result in Ca2+ transients in the tested neurons. We thus employed an unc-13 and unc-31 lof mutation to validate whether the Ca2+ transients in URX are induced by presynaptic neurons. UNC-13 is essential for synaptic vesicle exocytosis and neurotransmitter release39, and UNC-31 plays a major role in Ca2+-dependent regulative exocytosis of dense-core vesicles and release of neuropeptides40,41. The lof mutants were used to analyse the response in sensory neurons under neurotransmission isolation38,42,43. In unc-31(e928) and unc-13(e1091) mutants, URXs also displayed robust ethanol-elicited Ca2+ signals similar to those in N2 worms (Fig. 2B,C). These data indicate that the Ca2+ response to ethanol in URXs is not likely a result of chemical neurotransmission from its presynaptic neurons. However, URX neurons connect with RMG and IL2 neurons by gap junctions44, and it possible that these electrical synaptic connections cause Ca2+ transients in URXs. The innexins (invertebrate connexin analogue) that compose gap-junction channels in URX are not yet known. In virtue of the fact that RMGs and IL2s express unc-7 and unc-9, respectively45, and that UNC-7 and UNC-9 form heterotypic channels to mediate electrical synapses46, we thus employed unc-9(fc16) mutants to examine ethanol-elicited Ca2+ signals in URXs. Our results showed that URXs in unc-9(fc16) displayed even stronger Ca2+ transients in response to 500 mM ethanol (Fig. 2D). In summary, these data support that URXs respond to ethanol in a cell-autonomous manner.

Next, we assayed URX Ca2+ responses to different concentrations of ethanol in the npr-1(ad609) mutant. Our data indicated that the Ca2+ responses in URXs were well related with the ethanol concentrations; however, they were not well fitted by the Hill function (Fig. 2E). Does ethanol alter oxygen levels in the M13 buffer and then affect URX neurons? We then assayed dissolved oxygen levels in the ethanol-M13 solutions with various ethanol concentrations. Our results showed that ethanol levels did not significantly influence dissolved oxygen levels in the solutions (Supplementary Fig. 4). The above results suggest that ethanol directly activates gcy-35/gcy-36-expressing neurons.

GCY-35/GCY-36—TAX-2/TAX-4 signalling pathway mediates ethanol activating O2 sensory neurons

To examine the role of the GCY-35/GCY-36—TAX-2/TAX-4 signalling pathway in ethanol activation of O2 sensory neurons, we first examined ethanol-evoked URX Ca2+ signals in npr-1(ad609); gcy-35(0k769) double mutants and the gcy-35 neuron-specifically reconstituted transgene. URXs exhibited robust Ca2+ signals in response to 500 mM ethanol in the npr-1(ad609) worms and almost no response in the npr-1(ad609); gcy-35(ok769) worms (Fig. 3A). Notably, transgenic rescue of gcy-35 in three O2-sensitive neurons by extrachromosomal expression of the cDNA driven by the gcy-32 promoter markedly restored the Ca2+ signals in URXs (Fig. 3A). We next examined URX Ca2+ signals in acute ethanol exposed npr-1; gcy-36 double mutants. As expected, URXs did not show obvious ethanol-evoked Ca2+ transients in npr-1(ad609); gcy-36(db42) (Fig. 3B) and npr-1(ad609); gcy-36(db66) (Supplementary Fig. 5A), similar to the response in the npr-1; gcy-35 double mutant (Fig. 3A). Collectively, these data support that the GCY-35/GCY-36 complex is necessary for neuronal excitation by ethanol in URXs.

Figure 3
figure 3

The genes encoding GCY-35/GCY-36 and TAX-2/TAX-4 mediate ethanol-evoked Ca2+ responses in O2 sensitive neuron URXs. The somal Ca2+ transients in URXs in response to ethanol (500 mM) in (A) npr-1(ad609); gcy-35(ok769), gcy-32p::gcy-35 transgene, etc., (B) npr-1(ad609); gcy-36(db42), etc., (C) npr-1(ad609); tax-4(p678), etc., and (D) npr-1(ad609); tax-2(p671), etc. One-way ANOVA. In (A), F (2, 46) = 14.17, P < 0.0001; in (C), F (2, 48) = 16.77, P < 0.0001. ***p ≤ 0.001 compared as indicated.

Full size image

The cGMP-gated TAX-2/TAX-4 channel acts downstream of the sGC GCY-35/GCY-36 complex. For testing the role of the channel in URX activation by ethanol, we used npr-1; tax-2 and npr-1; tax-4 mutants to examine Ca2+ responses to ethanol in the neurons. URXs displayed no visible increase in Ca2+ transients in npr-1(ad609); tax-4(p678) (Fig. 3C), npr-1(ad609); tax-2(p671) (Fig. 3D), and npr-1(ad609); tax-4(ks28) (Supplementary Fig. 5B), similar to URX responses in npr-1; gcy-35 and npr-1; gcy-36 double mutants (Fig. 3A,B, Supplementary Fig. 5A). The extrachromosomal expression of tax-4 cDNA driven by the gcy-32 promoter obviously restored URX robust Ca2+ responses to ethanol (Fig. 3C).

The Ca2+ responses of URXs in npr-1(ad609) were well related with ethanol concentrations from 50 mM to 1.5 M (Fig. 2E). To determine whether the GCY-35/36—TAX-2/TAX-4 signalling pathway is also indispensable for URX responses to a high concentration of ethanol in npr-1 animals, we recorded the Ca2+ transients in URXs in response to 1.5 M ethanol in the double mutants npr-1(ad609); gcy-35(0k769), npr-1(ad609); tax-4(p678) and npr-1(ad609); tax-2(p671). Our result showed that all the double mutants displayed no visible ethanol-evoked URX Ca2+ signals (Supplementary Fig. 6A–C). The extrachromosomal expression of gcy-35 and tax-4 cDNA driven by the gcy-32 promoter markedly restored URX robust Ca2+ responses to ethanol in npr-1(ad609); gcy-35(0k769) (Supplementary Fig. 6A) and npr-1(ad609); tax-4(p678) (Supplementary Fig. 6B), respectively. These data demonstrate that the GCY-35/GCY-36—TAX-2/TAX-4 signalling pathway mediates ethanol activation of O2 sensory neurons over a wide range of concentrations.

To test whether the GCY-35/GCY-36 complex is an ethanol receptor, we ectopically co-expressed gcy-35 and gcy-36 cDNA in the neurons that express tax-2/tax-4 and no gcy-35/gcy-36. We first tested whether the candidate neurons respond to ethanol. We found that ASIs and ASGs displayed no ethanol-evoked Ca2+ signals; however, ASKs and BAGs exhibited ethanol-elicited Ca2+ responses (Supplementary Fig. 7). Thus, ASIs and ASGs were chosen for our test. Unexpectedly, ectopic co-expression of gcy-35 and gcy-36 did not confer ethanol sensitivity on ASIs or ASGs (Supplementary Fig. 7). Currently, we cannot exclude that this was due to a technical failure. Whether this complex is an ethanol receptor still needs further study.

Activation of the O2-sensory neurons URXs, AQR and PQR is indispensable for acute functional ethanol tolerance

To verify the role of the O2 sensory neurons in acute functional ethanol tolerance, we used chemogenetics to silence their activities. We employed specific expression of Drosophila HisCl1 encoding a histamine-gated chloride channel driven by the gcy-32 promoter in npr-1(ad609) (Supplementary Fig. 8) and administration of 10 mM histamine47 to chemogenetically inhibit the neurons. Our result showed that chemogenetically silencing O2 sensory neurons markedly suppressed npr-1 acute functional ethanol tolerance without impacting worm sensitivity to ethanol (Fig. 4A), although HisCl1 expression in O2-sensitive neurons alone (without histamine treatment) in N2 worms also suppressed acute functional ethanol tolerance (Supplementary Fig. 9). This is possibly a result of the low probability of opening the HisCl1 channel without the ligand histamine or being activated by unknown ligands in worms. Taken together, these data suggest that activation of URX, AQR and PQR neurons is required for acute functional ethanol tolerance in npr-1 worms.

Figure 4
figure 4

Activation of O2 sensory neurons URXs, AQR and PQR is required for acute functional ethanol tolerance in npr-1 worms. (A) Chemogenetic inhibition of URXs, AQR and PQR suppresses the acute functional ethanol tolerance in animal. Two-way ANOVA. F Time (1, 62) = 54.02, P < 0.0001; F genotype (3, 62) = 2.050, P = 0.1160; F genotype * Time (3, 62) = 4.180, P = 0.0093. Significant posttests difference at 30 min in comparisons with mutant with or without histamine treatment, or the transgene without histamine application is the same, and indicated as **p ≤ 0.01. (B) Working model for the molecular basis of acute functional ethanol tolerance.

Full size image

In summary, in this study, we presented data supporting that the sGC heterodimer GCY-35/GCY-36—cGMP-gated channel (TAX-2/TAX-4) signalling pathway in O2 sensory neurons mediates the acute functional ethanol tolerance in npr-1 worms and that ethanol directly activates these neurons. The hypothesized molecular basis is shown in Fig. 4B.

Discussion

C. elegans lives in rotting fruit and vegetation that contain lower oxygen levels than that in air. The low oxygen level may protect the worms from the damage of reactive oxygen species25,48,49. C. elegans develops O2 sensation and moves faster under high O2 levels25,30,34. Due to metabolic pathways of bacteria, bacterial flourishing produces alcohol and reduces ambient oxygen levels in microenvironments. The presence of alcohol and a low oxygen level means a sufficient supply of bacterial food for worms. However, high levels of alcohol make worms intoxicated and uncoordinated. The nematodes develop sensations and behavioural responses to oxygen and ethanol. This endows worms with advantages for survival and population propagation in the wild. Loss-of-function mutations and the lower function 215F natural allele of NPR-1 make worms ethanol resistant (Supplementary Fig. 2D), which is consistent with a previous report23. The high level of ethanol (500 mM)-treated npr-1 animals move as fast at 7% O2 as at 21% O2 (Supplementary Fig. 2). Possible explanations are as follows: URXs display an increase in Ca2+ signals in response to the switch of O2 from 7% to 21%, similar to those evoked by 500 mM ethanol, as shown in Supplementary Fig. 10, and the O2 sensory neurons respond to these two stimuli almost equally. However, the response is possibly not additive. The aversive responses to the harmful stimuli of high levels of oxygen and acute functional tolerance to high levels of ethanol endue worms to avoid harmful environments, and these two functions possess similar adaptive significance.

The cGMP signalling pathway is involved in sensations and many physiological activities50. In C. elegans, it plays essential roles in the chemosensation of soluble and volatile compounds32, oxygen25,30 and CO251, as well as in photosensation52 and water sensation31. It is also involved in long-term physiological processes, including satiety, lifespan, lethargy, innate immunity and development53,54,55,56. GCY-35 and GCY-36, consisting of a sGC complex that is expressed in sensory neurons URX, AQR and PQR, are required in URX sensory neurons to sense increases in O2 levels and affect corresponding behaviours25,30,34. The heme-binding domain of GCY-35 binds molecular oxygen, NO and CO25,57. Is the GCY-35/GCY-36 complex an ethanol receptor? Unfortunately, calcium responses to various concentrations of ethanol could not be well fitted by the Hill function, and ectopic co-expression of both gcy-35 and gcy-36 cDNA in ASI or ASG neurons that expressed tax-2 and tax-4 was not able to induce a Ca2+ response to ethanol, albeit technical failure cannot be excluded so far. These results hint that the complex is unlikely to be an ethanol receptor. This open question needs further study. However, overall, the present study demonstrates that GCY-35/GCY-36—TAX-2/TAX-4 signalling mediates the ethanol excitatory effect on O2 sensory neurons and acute worm function in ethanol tolerance. As many molecular mechanisms in neuronal activities are conserved in the animal kingdom, our finding adds a new and important aspect to the cGMP signalling pathway and provides a new clue to understanding the complicated mechanisms of alcohol effects.

Methods

Assays of worm behavioral responses to acute ethanol treatment

The ethanol effect on worm locomotion was tested as described6,23,27, with minor modifications. In short, 10 ml of hot 2% nematode growth media (NGM) was poured into a 6 cm diameter Petri dish and dried for 2 hour at 37 °C. Anhydrous ethanol was added onto the test plate unseeded with bacterium to achieve the final concentration of 500 mM used in the behavioral tests. The dish was covered with lid, sealed with Parafilm and stored for more than 2 hour at room temperature to allow ethanol diffuse. 10 young adult worms were transferred into a plate unseeded with bacterium and allow them to move freely for 30 min, then were transferred onto an area that was circled by a filter paper ring (inner diameter 1.5 cm) soaked with 20 mM CuSO4 solution to limit worm exploration range on agar plates added with or without ethanol for ethanol treated or untreated locomotion test. Worm locomotion was recorded under a Zeiss Discovery V8 stereomicroscope (Carl Zeiss MicroImaging GmbH, Göttingen, Germany). The image sequences were captured with an Andor iXonEM + DV885K EMCCD camera (Andor Technology plc., Springvale Business Park, Belfast, UK) at 1 frame per second. The worm locomotion speed was analyzed by use of multi-worm Tracker software58. The relative speeds were calculated as: Speed ratio = speed ethanol treated/speed ethanol untreated. The relative locomotion speed during period of 2 min at 10 and 30 min after worm exposure to ethanol were used as indicators for worm sensitivity and functional tolerance to ethanol, respectively.

Chemogenetics

A chemogenetic method was used to inhibit neuron activity as described47. In short, a Drosophila chloride channel, HisCl1, was ecto-expressed specifically in the sensory neurons URXs, AQR and PQR driven by gcy-32 promoter, and 10 mM histamine was applied to activate the channel. For making the histamine-NGM agar plates, we added 1 M histamine dihydrochloride (Sigma-Aldrich, ST Louis, USA) stock solution (in ultra-pure water) into NGM agar at ∼65 °C immediately before pouring plates.

Data analysis

Data statistical analysis was conducted by use of GraphPad Prism 6 (GraphPad Software, Inc., CA, USA). We used one-way analysis of variance to test the mean values among three or more than three samples with single factor, and used two-way analysis of variance to determine the significant difference between groups for two factors (time and genotypes). Next, we used Bonferroni’s t-test correction for multiple comparisons. Results are presented as the mean values ± SEMs with number of experimental replications (n).