The Signaling Pathway of Caenorhabditis elegans Mediates Chemotaxis Response to the Attractant 2-Heptanone in a Trojan Horse-like Pathogenesis*

The nematode Caenorhabditis elegans exhibits behavioral responses to a wide range of odorants associated with food and pathogens. A previous study described a Trojan Horse-like strategy of pathogenesis whereby the bacterium Bacillus nematocida B16 emits the volatile organic compound 2-heptanone to trap C. elegans for successful infection. Here, we further explored the receptor for 2-heptanone as well as the pathway involved in signal transduction in C. elegans. Our experiments showed that 2-heptanone sensing depended on the function of AWC neurons and a GPCR encoded by str-2. Consistent with the above observation, the HEK293 cells expressing STR-2 on their surfaces showed a transient elevation in intracellular Ca2+ levels after 2-heptanone applications. After combining the assays of RNA interference and gene mutants, we also identified the Gα subunits and their downstream components in the olfactory signal cascade that are necessary for responding to 2-heptanone, including Gα subunits of egl-30 and gpa-3, phospholipase C of plc-1and egl-8, and the calcium channel of cmk-1 and cal-1. Our work demonstrates for the first time that an integrated signaling pathway for 2-heptanone response in C. elegans involves recognition by GPCR STR-2, activation by Gα subunits of egl-30/gpa-3 and transfer to the PLC pathway, indicating that a potentially novel olfactory pathway exists in AWC neurons. Meanwhile, since 2-heptanone, a metabolite from the pathogenic bacterium B. nematocida B16, can be sensed by C. elegans and thus strongly attract its host, our current work also suggested coevolution between the pathogenic microorganism and the chemosensory system in C. elegans.

Because Caenorhabditis elegans lives in soil inhabited by various microorganisms, the capability to sense environmental stimuli is essential for the survival of the animal, including differentiating pathogens from food bacteria (1,2). Generally, the discrimination of pathogens mainly depends on the 12 pairs of amphid chemosensory neurons with cilia structures (3,4). Among the C. elegans chemosensory systems, the olfactory system is presumed to be one of the most sensitive and effective tools because of its specificity in detecting volatile chemicals at a distance. Furthermore, it has also been shown that subtle alterations in chemical structures among different odorants can result in substantial differences in perceived olfactory quality, which helps animals navigate in their environment and allows them to generate diverse behavioral responses to odorants. To achieve this remarkable discriminatory power of the olfactory system, C. elegans employs at least three pairs of olfactory neurons, including the AWA, AWB, and AWC neurons, to detect olfactory stimuli (5). In most cases, the worms detect attractive compounds emitted by food mainly using the olfactory neurons AWA and AWC (6,7); while AWB neurons are responsible for sensing repulsive volatile chemicals that may be produced by pathogens (8,9). Though part of the odorant specificity might be due to the odorant differences detected by AWA and AWC, or the functional asymmetry between the left and the right AWC, these two factors cannot account for all the odorant specificity, suggesting that additional intracellular mechanisms exist to establish odorant specificity (10,11).
Intracellular components are thought to account for the majority of discriminatory power, and some of them have been identified through the genetic analysis of mutations that affect olfactory sensing of specific chemicals. However, until now, ODR-10 and the attractant diacetyl are the only known G protein-coupled receptor (GPCR) 2 and its specific ligand in C. elegans. Worms with the odr-10 mutation have a 100-foldreduced sensitivity to the volatile attractant diacetyl, but retain a normal chemotaxis to other odorants detected by the AWA or AWC neuron (7,12). Despite the involvement of different molecules, a common pathway for olfaction transduction in either AWA or AWC has been proposed, which consists of recognition by specialized GPCRs, activation of the G protein, and transduction of the odorant signal to channel proteins (5). However, the integration of the specific signaling molecules, the individual sensory receptor that responds to its ligands and the signal cascade still require further elucidation.
The chemotaxis of C. elegans toward food bacteria has been described by detecting both water-soluble and volatile clues. Consistently, the volatile compounds that have been found to be attractive to worms are mostly natural products of bacterial metabolism, including alcohols, ketones, aldehydes, esters, amines, sulfhydryls, organic acids, and aromatic and heterocyclic compounds (6,13). Our previous work reported a bacterial strain, B. nematocida B16, which can infect and kill C. elegans using the strategy of a "Trojan horse" (14). Potent volatile organic compounds produced by this bacterium are more attractive to worms than those from ordinary dietary bacteria, luring the nematode and permitting successful infection. Among volatile organic compounds, 2-heptanone belonging to methyl ketone, a compound found widely in nature, has been reported to serve as a biological pheromone in some species (15). Our previous work has revealed an interesting phenomenon: the fatal pathogenic bacterium B. nematocida 16 can imitate this beneficial signal to trap animals. The interaction that the bacterial pathogens produce the potent signal to lure its hosts and the worms sense this attractant to be trapped suggests an interesting coevolution between the pathogenic microorganism and the chemosensory system in C. elegans.
The volatile compounds of 2-heptanone has also been shown as one of the most important attractants since the absence of 2-heptanone would significantly weaken the capability to attract nematodes (35). In our current investigation, we explored the receptor and transductory pathway in C. elegans using 2-heptanone as a ligand. Our experiments showed that the detection of 2-heptanone depended on the function of a GPCR encoded by str-2 that expressed on the AWC on neuron. We also identified the G␣ subunits and their downstream components in the olfactory signal cascade that are necessary for responding to 2-heptanone. The olfactory pathway in the chemosensory neuron AWC includes G␣ subunits of egl-30 and gpa-3, phospholipase C of plc-1 and egl-8, and the calcium channel of cmk-1 and cal-1, which differs from the previously reported transductory pathway in AWC mediated by the cyclic nucleotide-gated channel (CNG) consisting of TAX-4 and TAX-2 subunits. Therefore, our current investigation proposes a potentially novel olfactory signal pathway for 2-heptanone sensing in AWC olfactory neurons.

C. elegans Detects 2-Heptanone Depending on Functions of AWC Olfactory
Neurons-Chemotaxis of C. elegans was measured by establishing a gradient of attractant from a point source and observing the accumulation of animals at the attractant source. Because the volatile organic chemical 2-heptanone diluted by ethanol was spotted on the surface of the agar, we spotted ethanol on the other side of the agar as a control for the effect of ethanol. Since C. elegans adapts to an attractant after prolonged exposure, the steady-state number of animals at an attractant reflected the result of both attraction and adaptation processes. To minimize the effects of adaptation, we did pre-experiments to calculate the time of the process. It was found that the worms immediately oriented their movements and then they congregated at the 2-heptanone, and the accumulation of worms at 2-heptanone would be complete within 1h (Fig. 1A). The rapid response of C. elegans to 2-heptanone suggested that 2-heptanone diffused into the air where it could be detected by the nematodes. To verify this possibility, 2-heptanone was spotted on the lid of the Petri plate so that the attractant was not in direct contact with the agar. The animals in the assays moved toward the source of 2-heptanone immediately, confirming that 2-heptanone was sensed by olfactory chemosensory neurons as a volatile signal. To further characterize the responses to volatile attractants, dose assays for 2-heptanone were performed. It was indicated that the animals responded well to 1 l of attractants at either 10 Ϫ1 dilutions or 10 Ϫ2 dilutions (7.2 ϫ 10 Ϫ1 mol/liter or 7.2 ϫ 10 Ϫ2 mol/liter).
In C. elegans, two pairs of olfactory neurons, AWA and AWC, were in charge of detecting the attractive odorants (6,7,10). Therefore, to identify which olfactory neurons were essential for chemotaxis to 2-heptanone, we first tested the chemotaxis to 2-heptanone using the RNAi method of knocking down the genes odr-7 and osm-9 required for the function of AWA, and the genes odr-1, tax-2, and tax-4 required for AWC. Our results showed that genes responsible for AWC olfactory neurons, including odr-1, tax-2, and tax-4, decreased the chemotaxis to 2-heptanone (data not shown). Then we further validated the chemotaxis to 2-heptanone using the mutants of odr-1 (n1936) and odr-7 (ky4). Our results showed that the odr-1 mutation animals were defective in sensing 2-heptanone at dilutions of 10 Ϫ1 and 10 Ϫ2 , while the odr-7 mutant retained the capability of sensing 2-heptanone (Fig. 1B). Furthermore, the mutant of ceh-36 (ks86), which lacked AWC chemosensory neurons, was also assayed for the chemotaxis to 2-heptanone. The result from the ceh-36 (ks86) mutation was similar to that from odr-1 (n1936), suggesting that the defect in AWC neurons would impede the sensing of 2-heptanone (Fig. 1E).
GPCR STR-2 of AWC on Neuron Is Required for Sensing 2-Heptanone-GPCRs have been reported as the most common olfactory receptors in C. elegans and mammalians. The genome of C. elegans encodes more than 500 putative GPCRs (2,8,16). Because some GPCRs that express asymmetrically on the chemosensory neurons are usually localize in the cilia, possess variable expressional levels, and are highly divergent between C. elegans and other species, it is reasonable to hypothesize that GPCRs enable the different chemotaxis responses for the different chemical molecules in C. elegans (2,5).
Because two known GPCRs, including STR-2 and SRA-13, have been demonstrated to be expressed in AWC olfactory neurons, the worms with the mutation of str-2 (ok3148) and sra-13 (zh13) were tested for chemotaxis to 2-heptanone. It was shown that the animals with str-2 (ok3148) mutation were defective in sensing 2-heptanone; but the animals with sra-13 (zh13) mutation retained a normal response to 2-heptanone (Fig. 1C). To further confirm the specificity between the receptor STR-2 and its signal molecule 2-heptanone, the mutant of str-2 (ok3148) was tested for chemotaxis to benzaldehyde, the other attractant sensed by AWC olfactory neurons, and to 2-ethyl alcohol, an attractant sensed by AWA. Consistently, our data illustrated that mutation of str-2 had no effects on worms sensing benzaldehyde and 2-ethyl alcohol, suggesting that the response to 2-heptanone mediated by the receptor STR-2 is specific (Fig. 1D).
STR-2 has been reported to be asymmetrically expressed in one of AWC neurons (AWC on ) (17,18). Thus, the mutants of nsy-1 (ok593) and inx-19 (ky634), which had two AWC on or two AWC off , respectively, were employed to clarify the roles of STR-2 involved in 2-heptanone sensing. Compared with the wild-type worms, the animals with nsy-1 (ok593) mutation, which possessed two AWC on neurons, represented comparable sensitivity to 2-heptanone (p Ն 0.05) (Fig. 1E). Contrarily, the worms with inx-19 (ky634) mutation lost the ability to respond to the odor of 2-heptanone because of the lack of expression of STR-2 (p Ͻ 0.01) (Fig. 1E). On the other hand, to determine whether the expression of STR-2 in AWC neurons could be induced by its signal 2-heptanone, we used the worms of kyIs140 [str-2::GFP ϩ lin-15(ϩ)] to observe the changes in STR-2 expression. It was found that the STR-2 indeed was asymmetrically strongly expressed in one AWC neuron and weakly expressed in ASI neurons that sense the situations such as crowding and starvation during dauer formation, which is consistent with the previous report ( Fig. 2B) (19). Furthermore, the fluorescence of the STR-2::GFP in AWC on neuron was enhanced after adding the signal molecule 2-hepatonane (Fig. 2C).
Finally, to confirm STR-2 protein is an odorant receptor for 2-heptanone, it was necessary to express this GPCR on the membrane of mammalian cell line HEK293T and then observe the cell response. Therefore, we transiently transfected mammalian cells with a plasmid encoding STR-2 protein with a mycepitope tag at its N terminus (str-2-pCMV-Tag 3C), which was FIGURE 1. Chemotaxisto 2-heptanone depends on the chemosensory neuron AWC on and the GPCR STR-2 expressed on it. A, schematic diagram of chemotaxis assay. A thin layer of agar in a Petri dish was used as a substrate for chemotaxis, since C. elegans moves efficiently over the agar surface. About 100 washed adult animals were placed near the center of a 10-cm assay plate with the attractant that was generally diluted by ethanol at one side of the plate and the control (ethanol) at the opposite side. The numbers of animals within the attractant area and the control area were counted at 1 h. A chemotaxis index was calculated as (animal numbers at attractant-animal numbers at counter-attractant)/total animal number in assay. B, chemotaxis assay of 2-heptanone on the wild-type, odr-1(n1936), and odr-7(ky4) animals at 1:10 or 1:100 dilutions. C, chemotaxis assay of 2-heptanone on the worms with the str-2 (ok3148) and sra-13 (zh13) mutations. D, response of str-2 (ok3148) animals to benzaldehyde and 2-ethyl alcohol. E, chemotaxis assay to 2-heptanone using the worm lines of ceh-36 (ks86), nsy-1 (ok593), and inx-19 (ky634). The worms with ceh-36 and inx-19 mutations represented the obviously decreased sensing of 2-heptanone. The statistical differences were analyzed using one-way ANOVA, n.s., p Ն 0.05; ***, p Ͻ 0.01. expected to be exposed to the external face of the plasma membrane. To evaluate STR-2 expression, immunofluorescent staining was performed, and the result demonstrated that myc-STR-2 was heterologously expressed and was mainly located on the cell membrane (Fig. 3A). To assay STR-2 function, we measured intracellular Ca 2ϩ levels using fura-2-based Ca 2ϩ imaging. Upon perfusion of 2-heptanone into HEK293T cells that had been transfected with the plasmids encoding myc-STR-2, a transient increase in intracellular Ca 2ϩ was observed in cells (Fig. 3B). The graphs of time-dependent changes in intracellular Ca 2ϩ suggested that the Ca 2ϩ elevation response was maximal about 3.5min after the onset of 2-heptanone application (at 90s) and then decayed within 1.5 min; but such responses were never seen in the negative control transfected with the blank vector (Fig. 3C). Quantitative analysis of 800 cells having myc-STR-2 expression or the blank vector revealed statistically significant changes in fluorescence when the cells with myc-STR-2 sensed 2-heptanone (Fig. 3D). Similarly, to further confirm the specificity between the receptor STR-2 and its signal molecule 2-heptanone, we replaced 2-heptanone with benzaldelhyde. As expected, the cells expressing myc-STR-2 exhibited no changes of intracellular Ca 2ϩ under the stimulation of benzaldehyde (Fig. 3E).
The G Protein ␣ Subunits of GPA-3 and EGL-30 Mediate the Signal Pathway Sensing 2-Heptanone-The olfaction in C. elegans is transduced from environmental stimuli to intercellular by the heterotrimeric G protein-coupled signal pathway. The genome of C. elegans encodes 21 G␣ subunits that play an activate role in the G protein signaling cascades, and therefore we tested the changes of chemotaxis to 2-heptanone with the 21 G␣ gene mutations. Our results showed that three G␣ mutations have effects on the chemotaxis to 2-heptanone, including odr-3 (n2130), gpa-3 (pk35), and egl-30 (ad805) (Fig. 4A).
ODR-3 has been thought of as the main G␣ mediating the olfactory signal pathway, and it has been described to disturb the development of cilia in chemosensory neurons. The proper structure of cilia is required for chemosensory neurons to detect most odors, hence we assayed the chemotaxis of the mutant odr-3 (n2130) to some other attractive molecules of benzaldehyde and 2-ethyl alcohol besides 2-heptanone. The mutation of odr-3 indeed disturbed the detection of the other two odors (Fig. 4B), consistent with the results previously described. However, it was noticed that odr-3 RNAi had a normal chemotaxis to 2-heptanone as well as to the other two tested odors, benzaldehyde and 2-ethyl alcohol (Fig. 4B). Since the different results we had obtained from the mutant odr-3 (n2130) and odr-3 RNAi, we employed qPCR experiment to validate the effect of odr-3 knock-down. Our data of qPCR obviously demonstrated that, compared with the negative control, the expressional level of odr-3 was retained less than half in the animals of odr-3 RNAi (Fig. 4C). Considering that odr-3 RNAi was performed after L1 stage, and it should have little effect on cilia development, it is reasonable to speculate that, besides the potential role in transducing the signal for sensing 2-heptanone, the disturbance of cilia development may be responsible for the defect in chemotaxis for all the tested odors.
To further confirm the involvement of egl-30, we rescued the expression of egl-30 under the control of the srsx-3 promoter, which initiated gene expression specifically in AWC neurons. Most of the chemotaxis to 2-heptanone was rescued in the transgenic animals (Fig. 4D), suggesting that the G␣ subunits EGL-30 participated in mediating the olfactory signal of 2-heptanone.
We tested the chemotaxis for the gpa-3 (pk35) mutation together with egl-30 RNAi or egl-30 (ad805) mutation together with gpa-3RNAi, and the change in the chemotaxis was cumulative (Fig. 4E), suggesting that GPA-3 and EGL-30 can activate the downstream molecules independently.
PLC Pathway Functions in the Olfactory Signal Pathway-Stimulation of a GPCR and its downstream G protein can activate some classical signaling cascades, including PKA, PLC, and cGMP pathways. Next, we tested the effects on chemotaxis to 2-heptanone of the genes located in the three pathways.
Downstream of the PLC pathway, the Ca 2ϩ influx can be observed since the endoplasmic reticulum releases Ca 2ϩ to the cytoplasm. To further confirm that PLC pathway mediated the olfactory signal transduction for sensing 2-heptanone, we determined Ca 2ϩ release using the strain carrying Ca 2ϩ reporter GCaMP3 under the control of the AWC-specific promoter of str-2 gene. After the addition of 2-heptanone, our results showed that the AWC neurons responded to 2-heptanone by a change in intracellular Ca 2ϩ concentration, suggesting the molecule of 2-heptanone could be transduced by the olfactory signal pathway of PLC as well as the intracellular Ca 2ϩ (Fig. 5B).
However, to our surprise, the animals with odr-1, daf-11, tax-2, and tax-4 mutations, which belong to the genes in cGMP pathway, were also defective in chemotaxis to 2-heptanone (Fig. 6A). The cyclic nucleotide-gated channel encoded by the tax-4 and tax-2 genes has been reported to be essential for the chemotaxis of AWC to a variety of volatile odors (20 -22). The other two genes, daf-11 and odr-1, encoding the receptor guanylate cyclase in AWC olfactory neurons, likely function as heterodimers to synthesize cGMP for the opening of the TAX-4/TAX-2 channel (23-25). Hence, another possibility was raised because it had been reported that the genes odr-1/daf-11 and tax-2/tax-4 were necessary for maintaining the expression of the gene str-2 in C. elegans (26,27). To test the hypothesis, we first determined if the expression level of odr-1 could be induced by the signal molecule of 2-heptanone. Compared with the negative control without signal molecule, it was found that the addition of 2-heptanone could not enhance the changes of odr-1 expression in AWC neurons (Fig. 6B). Then we further assayed the changes of STR-2 in the worms with odr-1 RNAi. It was observed that str-2::GFP expressed weakly both with and without the addition of 2-heptanone (Fig. 6C). However, when animals of odr-1 RNAi were cultured on media with extra cGMP addition, the str-2::GFP expression was rescued and the animals also presented a relatively normal chemotaxis to 2-heptanone (Fig. 6D). Finally, we expressed str-2 using the srsx-3 promoter in odr-1 (n1936), and the chemotaxis to 2-heptanone was rescued in the transgenic animals (Fig. 6E). Collectively, our current data demonstrated that those genes, including guanylyl cyclases odr-1 and daf-11 as well as cGMP gated cation subunits tax-2 and tax-4, seemed to play more important roles in maintaining str-2 expression to influence the olfactory signal transduction of sensing 2-heptanone.

Discussion
It is well known that C. elegans has chemotaxis to a variety of volatile compounds, which can mainly be mediated by the olfactory neurons. The olfactory responses for the different odors are associated with the most powerful behaviors in worms, such as pursuing food and escaping pathogens. Among those olfactory responses, C. elegans mainly employs four cells (left and right AWA or AWC cells) to sense and discriminate among a variety of volatile attractants (11). Previous studies also have revealed that two distinct pathways of olfactory signal transduction are employed in these two types of neurons based Regardless of whether either AWC or AWA is involved, the binding of odorant molecules (ligands) to specific receptors on the external surface of cilia is generally the first step for initiation of olfactory transduction. The str-2 gene encodes a seventransmembrane domain protein, a member of the str family of chemosensory receptors, asymmetrically expressed at the top of cilia in AWC on neurons and weakly expressed in the ASI neurons. Mutations in the str-2 gene obviously affect chemotaxis toward 2-heptanone, but had little effect on sensing benzaldehyde whose recognition also requires AWC function. Certainly, our data have shown that the seven-transmembrane receptor STR-2 was responsible for sensing the volatile odorant 2-heptanone when heterologously expressed in human HEK293 cells, and the behavioral response of STR-2 to 2-heptanone should be highly selective in vivo.
Heterotrimeric G protein-coupled signaling cascades are used to transduce olfactory signals varying from intercellular mediators to environmental stimuli. It has been revealed that a variety of G␣ subunits, including odr-3, gpa-3, gpa-5, and gpa-6 in AWA neurons as well as odr-3, gpa-2, gpa-3, and gpa-13 in AWC neurons, play either positively or negatively in chemosensation (16, 28 -30). Among them, gpa-3 gene was expressed in cilia, cell bodies and axons of AWA and AWC neurons. Since loss of gpa-3 activity exacerbated the chemotaxis defect of odr-3(lf) mutants to all odorants, its function has been suggested to be redundant to ODR-3 (30,32). In our current study, genes gpa-3 and egl-30 influenced animals' response to 2-heptanone, and the double mutant worms almost lost all chemotaxis to 2-heptanone. Compared with gpa-3, egl-30 was more widely expressed. However, when we specifically rescued the expression of egl-30 in AWC neurons, the chemotaxis of transgenic animals was restored, confirming that egl-30 in AWC neurons was required for chemotaxis to 2-heptanone. Therefore, two G␣ subunits, gpa-3 and egl-30, were sufficient for the olfactory transduction pathway of detecting 2-heptanone. The gene odr-3 has commonly been considered the major G␣ gene for olfactory transduction in either AWC or AWA, but its role in detecting 2-heptanone seemed different. Though the method of RNAi in neurons was believed to have less obvious the effects than in other organs of C. elegans, we think, the intact response of animals treated with ord-3 RNAi was not due to the inadequately knocked-down because using the RNAi method of knocking down the genes odr-1, tax-2, and tax-4 decreased the chemotaxis to 2-heptanone. Additionally, the gene of odr-3 is involved in cilium morphogenesis in olfactory neurons, and its amount determined the extent of cilium outgrowth in AWA and AWC neurons. For example, it has been proposed that ODR-3 expression defined the unique fan-like extension of the AWC olfactory cilia (29). Thus it is reasonable for us to speculate that the ODR-3's function for maintaining the AWC cilium structure should be partially responsible for the deficiency to sense 2-heptanone besides its potential role in transducing the signal.
In the olfactory signal cascade in C. elegans, the ion channel is the key component in the last step to sense and transduce receptor activity into electrical activity in olfactory cells (20,33). Here, a series of signal molecules, such as egl-8, plc-1, itr-1, cmk-1, and cal-1 have been shown to influence the chemotaxis to 2-heptanone, and therefore the PLC pathway, instead of the known ion channel, is proposed for chemosensory neuron AWC on sensing 2-heptanone. So far, the integration of a specific signaling pathway, including the individual sensory receptor that responds to its ligands as well as the downstream signal cascade, was suggested for sensing 2-heptanone (Fig. 7), which might be a potentially novel olfactory signal pathway existing in AWC olfactory neurons. Additionally, since 2-heptanone, a metabolite from the pathogenic bacterium B. nematocida B16, can be sensed by C. elegans and thus strongly attract its host (5), our current work also suggested coevolution between the pathogenic microorganism and the chemosensory system in C. elegans.
The wild type N2 and mutant strains were grown to adulthood on NGM (50 mM NaCl, 20 g/liter of agar, 2.5 g/liter of peptone, 1.0 mM cholesterol, 1.0 mM CaCl 2 , 1.0 mM MgSO 4 , and 25 mM potassium phosphate at pH 6.0) seeded with Escherichia coli OP50 under well-nourished conditions at 20°C using standard methods.
Chemotaxis Assays-All of the attractants were obtained from Sigma Chemicals. Chemotaxis assays were modified based on the assays of Ward (34) and Bargmann and Horvitz (31). Assay plates were 10-cm tissue culture dishes containing   The pathway of sensing 2-heptanone was affected by genes located on the PLC and cGMP pathways. Between them, the genes in the PLC pathway, such as phospholipase C of plc-1and egl-8, calcium channel of cmk-1and cal-1, were expected to be responsible for transducing the specific olfactory signals of 2-heptatone. However, the genes of cGMP pathway seemed to play more important roles in maintaining str-2 expression to influence the olfactory signal transduction.
10 ml of 1.6% agar. Two marks were made on the back of the plate at opposite sides of the plate about 0.5 cm from the edge of the agar. About 1 l of attractant diluted in ethanol was placed on the agar over one mark, and 1 l of ethanol was placed as the counter-attractant over the opposite mark. One microliter sodium azide with the concentration of 1 M was also placed at both the attractant source and the counter-attractant source. This drug could anesthetize animals within about a 0.5-cm radius of the attractant. Since it was found that the chemotaxis index of worms would reach the maximum and stable after 45 min, the chemotaxis assays were generally finished within 1 h.
During the assay, the well-fed adult animals were washed three times with M9 Buffer and once with water to free them of bacteria. Then the worms were placed near the center of a plate equidistant from the attractant and the counter-attractant. At 1 h after the assay began, the numbers of animals at the attractant and the counter-attractant areas were determined. The chemotaxis index was calculated as (animal numbers at attractantanimal numbers at counter-attractant)/total animal number in assay (Fig. 1A). The chemotaxis index could vary from 1.0 to Ϫ1.0.
RNA-mediated Interference-RNAi bacterial strains containing targeted genes were obtained from the Ahringer RNAi library. RNAi feeding experiments were performed on synchronized L1 larvae at 20°C until L4 larvae or young adult worms that were used in the chemotaxis assay.
Mammalian Cell Culture and Immunofluorescence-HEK293T cell lines were maintained in DMEM with 10% fetal calf serum and lacking pyruvate. Cells were plated onto poly-D-lysinecoated 15-mm glass coverslips in 6-well cell culture plates for 24 h before transfection. Lipofectamine 2000 (ThermoFisher Scientific) was used as the DNA carrier. Two micrograms of plasmid DNA (str-2-pCMV-Tag 3C or the negative control of blank vectors) were added to each culture dish individually, and then were incubated with the cells for 10 h at 37°C. Two days after transfection, the cells were fixed and permeabilized with methanol at Ϫ20°C for 20 min. Anti-c-myc was used at 1:500 dilution. The secondary antibody (goat anti-mouse(HϩL),PE conjugate) was used at a 1:100 dilution. Staining and visualization were carried out according to standard procedures.
Fura-2 Ca 2ϩ Imaging-After 48 h of cell transfection, the cells were harvested and incubated in HBSS buffer (137 mM NaCl, 5.4 mM KCl, 4.5 mM glucose, 0.4 mM KH 2 PO 4 , 0.3 mM Na 2 HPO 4 , and 0.1% PluronicF-127, pH 7.2) containing 0.005 mM Fura-2-AM (Molecular Probes Life Tech) for 45 min at 37°C. The mixture of HBSS buffer with Fura was then removed. Following three washes in HBSS buffer, the cells were incubated in HBSS buffer for another 4.5 h at room temperature. After the cells were washed again, the compounds for visualization were dissolved in dimethyl sulfoxide and finally mixed in HBSS buffer. Fluorescence imaging was carried out according to standard procedures.
Construction of Transgenic Worm Lines-The genes str-2 and egl-30 under the control of the srsx-3 promoter were expressed in animals odr-1 (n1936) and egl-30 (ad805) respectively to generate rescue lines. The full encoding frames along with their 3Ј-UTR were obtained by PCR using N2 genomic DNA as a template. The resulting PCR products were purified from aga-rose gel by using a QIAquick Gel Extraction Kit (Qiagen). Then the purified PCR product, 10 ng/ml, was coinjected with pPD95.79. Two days after DNA injection, four worms expressing GFP were allowed to self-fertilize. The genotype of the transgenic lines was confirmed by PCR amplification of a portion of the gene using oligonucleotide primers.
Imaging Sensory-evoked Ca 2ϩ Transients in C. elegans-To confirm that 2-heptanone was detected by the AWC olfactory neuron, we used the worm strain CX11935KyEx3252 (str-2::GCaMP3 10 ng/l, coel::GFP 10 ng/l) to observe transient Ca 2ϩ in the AWC neuron in C. elegans. Animals were placed on the agar plate for the assay of chemotaxis, and 1 l of 2-heptanone at 100ϫ dilution was placed on the center of the plate. After 5 min, we picked signal animals to capture the image using fluorescence microscopy (Olympus, Melville, NY).
Quantitative Real-time PCR (qPCR)-The total RNA samples of worms, including the odr-3 RNAi and the negative control HT115 that was fed on the E. coli with blank vector, were extracted using the Trizol reagent from Invitrogen (Carlsbad, CA) at young adult. Random primed cDNA were obtained by reverse transcription of the total RNA samples with TransScriptReverse Transcriptase from TransGen Biotech (Beijing, China). A real-time PCR analysis was performed with the ABI Prism 7000 Sequence Detection System (Applied Biosystems) using SYBR Premix Ex Taq™ II (Takara, Dalian, China). The primers for the target gene odr-3 used as follows: odr-3F: 5Ј-TATCTGAACCTGGGTATCGG-3Ј and odr-3R: 5Ј-GCTATGGAGACTCGGTGAA-3Ј. The partial sequence of ␤-actin was used for an internal control, whose primers were 5Ј-GGATTCTGCAGAAACTCGGA-3Ј and 5Ј-CATCCA-GATCCGCCATGT-3Ј. PCR amplification followed by 40 cycles of 94°C for 30 s, 60°C for 30 s, 72°C for 40 s on ABI PRISM 7000 Real-Time PCR. The real-time PCR experiments were repeated three times for each reaction using independent RNA sample.
Author Contributions-Chunmei Zhuang and Xiaowei Huang wrote this paper. Xiaowei Huang designed the experiments. Ninghui Zhao, Yao Chen, Donghua Zhang, Jinyuan Yan, and Wei Zou did the experiments. Xiaowei Huang and Keqin Zhang analyzed the data.