Skip to main content
NCBI home page
Chemical Senses logoLink to Chemical Senses
. 2018 Jan 24;43(3):169–180. doi: 10.1093/chemse/bjy001

Identification of Odor Blend Used by Caenorhabditis elegans for Pathogen Recognition

Soleil E Worthy 1, German L Rojas 2, Charles J Taylor 1, Elizabeth E Glater 2,✉
PMCID: PMC6018680  PMID: 29373666

Abstract

Animals have evolved specialized pathways to detect appropriate food sources and avoid harmful ones. Caenorhabditis elegans can distinguish among the odors of various species of bacteria, its major food source, but little is known about what specific chemical cue or combination of chemical cues C. elegans uses to detect and recognize different microbes. Here, we examine the strong innate attraction of C. elegans for the odor of the pathogenic bacterium, Serratia marcescens. This initial attraction likely facilitates ingestion and infection of the C. elegans host. Using solid-phase microextraction and gas chromatography coupled with mass spectrometry, we identify 5 volatile odors released by S. marcescens and identify those that are attractive to C. elegans. We use genetic methods to show that the amphid chemosensory neuron, AWCON, senses both S. marcescens-released 2-butanone and acetone and drives attraction to S. marcescens. In C. elegans, pairing a single odorant with food deprivation results in a reduced attractive response for that specific odor. We find that pairing the natural odor of S. marcescens with food deprivation results in a reduced attraction for the natural odor of S. marcescens and a similar reduced attraction for the synthetic blend of acetone and 2-butanone. This result indicates that only 2 odorants represent the more complex odor bouquet of S. marcescens. Although bacterial-released volatiles have long been known to be attractive to C. elegans, this study defines for the first time specific volatile cues that represent a particular bacterium to C. elegans.

Keywords: food, odor mixture, olfaction, Serratia marcescens, solid-phase microextraction

Introduction

The nematode Caenorhabditis elegans lives in rotting fruit and plant matter, where it must navigate a complex environment containing some microbes which are nutritious food and some which are pathogenic (Félix and Duveau 2012). Caenorhabditis elegans can distinguish among the odors of various species of bacteria and shows strong innate preferences for particular bacterial species (Shtonda and Avery 2006; Ha et al. 2010; Harris et al. 2014). Although a few attractants released by specific microbes have been identified, it is unknown what specific chemical cue or combination of chemical cues C. elegans uses to recognize and detect different microbes (Beale et al. 2006; Niu et al. 2010; Werner et al. 2014; Choi et al. 2016; Hsueh et al. 2017). Is recognition of a natural bacterial odor blend based on several components, a few key components or a single unique component? Identifying these ethologically relevant chemical cues is essential to understanding the molecular basis of interactions between C. elegans and microbes.

Here, we examine how C. elegans detects the pathogenic bacterium, Serratia marcescens, because S. marcescens has been shown to influence the behavior of C. elegans in several ways. First, the odor of S. marcescens is highly attractive to C. elegans and this attraction likely facilitates consumption and subsequent intestinal infection that kills the worm after 2–3 days (Pujol et al. 2001; Mallo et al. 2002; Kurz et al. 2003; Zhang et al. 2005; Glater et al. 2014). Thus, S. marcescens likely releases volatiles that are attractive to C. elegans. Second, although C. elegans is initially attracted to S. marcescens, after several hours the worms leave the lawn of pathogenic bacteria. A secreted, non-volatile, natural bacterial product mediating this repulsion from specific strains of S. marcescens has been identified and is serrawettin W2, a cyclic lipodepsipentapeptide (Pradel et al. 2007). Third, the behavioral response of C. elegans to the odor of S. marcescens exhibits plasticity and is modified after the worm has encountered S. marcescens. After C. elegans is infected by S. marcescens, C. elegans avoids the odor of this specific pathogenic bacterium (Zhang et al. 2005). Fourth, attraction to S. marcescens is also modified on an evolutionary timescale. It is likely that attraction for S. marcescens is a fast evolving trait because different strains of C. elegans isolated from different regions around the world show natural variation in preference for the odor of S. marcescens (Glater et al. 2014). In addition, in an experimental co-evolution study, C. elegans exhibited avoidance of S. marcescens after being passaged for 30 generations with the pathogen (Penley and Morran 2017). Therefore, we propose to examine 2 questions. First, what volatiles initially attract C. elegans to S. marcescens? Second, after the odor of S. marcescens is paired with an adverse experience, what volatiles are later used to recognize and avoid S. marcescens?

In this study, we define the attractive volatiles necessary for recognition of S. marcescens by C. elegans. We first identify the major volatiles released by S. marcescens using solid-phase microextraction and gas chromatography coupled with mass spectrometry and then determine which of these volatiles are attractive. We use genetic methods to determine the sensory neurons necessary for detection of S. marcescens and S. marcescens-released volatiles. Last, we use aversive conditioning assays in which the odor of S. marcescens is paired with food deprivation and then animals are tested for chemotaxis to individual odorants released by S. marcescens, synthetic mixtures of these odorants, or the natural odor of S. marcescens, to determine what component or components best represent S. marcescens. Taken together, we determine that 2 odorants, acetone and 2-butanone, are sufficient to represent S. marcescens to C. elegans and are detected by the AWCON sensory neuron.

Materials and methods

Nematode growth and strains

Strains were grown and maintained under standard conditions at 20°C on nematode growth media (NGM) (Brenner 1974). Animals were grown on plates seeded with Escherichia coli HB101 ATCC 33694. Some strains were provided by the CGC, which is funded by NIH Office of Research Infrastructure Programs (P40 OD010440), Dr. Cornelia Bargmann, Dr. Sreekanth Chalasani, Dr. Elissa Hallem and Dr. Paul Sternberg. CX4 odr-7(ky4) X (Sengupta et al. 1994), CX3938 lim-4(ky403) X (Sagasti et al. 1999), PR680 che-1(p680) I (Dusenbery et al. 1975), CX4651, osm-9(ky10) ocr-2(ak47) V (Tobin et al. 2002), CX6339 ceh-36(ky640) X (Lanjuin et al. 2003), CX6161 nsy-5(ky634) I (Chuang et al. 2007), CX10232 nsy-7(tm3080) IV (Lesch et al. 2009), CX9190 nsy-1(ky542) II (Wes and Bargmann 2001), EAH2 gcy-9 (tm2816) X (Carrillo et al. 2013), CX13078 tax-4(p678) III (Dusenbery et al. 1975), CX13790 tax-4(p678); kyEx4233 (ceh-36*::tax-4 sl2::GFP and myo3::mCherry), CX13359 tax-4(p678); kyEx3945 (sra-9::tax-4 sl2::GFP and elt-2::mCherry) (Macosko et al. 2009), CX11576 kyEx3097 (sra-9::TeTx::mcherry and elt-2::GFP) (Macosko et al. 2009), CX5868 kyIs140 I (str-2::GFP); kyEx613 (odr-3::TeTx; odr-1::RFP and elt-2::GFP) generated by Alvaro Sagasti, CX13458 nsy-7(tm3080); kyEx2125 (odr-3p::nsy-7 and ofm-1::dsRed), PS5716 tax-4(p678); syEx1045 (gcy-33::tax-4 and myo-2::gfp) (Hallem and Sternberg 2008).

Transgenes

Extra chromosomal transgenes were made by injection of DNA plasmids and a fluorescent coinjection marker into the gonads of young adult hermaphrodites. Two to 3 independent lines from each injection were assayed. DNA was cloned into psm vector. AWC ceh-36p* ends: ctcacatccatctttctggcgactgtttcatt…gcctgcccccgcatgcacaa; ASK sra-9p ends: gcatgctatattccaccaaa....tgtgcatcaatcatagaaca; tax-4 cDNA ends: atgtcaacggcggaacctgc.…ctgaatccttgctcaaatag; odr-3p::nsy-7 cDNA as described (Lesch and Bargmann 2010); sra-9p::TeTx::mCherry as described (Macosko et al. 2009); and odr-3p::TeTx generated by Alvaro Sagasti.

Bacterial strains and reagents

Bacterial strains S. marcescens ATCC 274 and E. coli HB101 ATCC 33694 were obtained from the American Type Culture Collection.

Bacterial choice assay

The 2-choice bacterial choice assay was modified from Zhang et al. (2005). Briefly, bacteria were grown overnight in Luria Broth (LB) at 26°C, centrifuged, and then resuspended in LB at an OD600 of 1.0 for S. marcescens or an OD600 of 10 for E. coli HB101, and 25 μL of each bacterial suspension was spotted onto an NGM plate and incubated for 5 h at 20°C. At these OD600 values, both bacteria had approximately the same cellular density: at OD600 of 1.0, S. marcescens yields 2.1 × 109 ± 1.5 × 109 colony-forming units (cfu) per ml; at OD600 of 10, E. coli HB101 yields 3.2 × 109 ± 1.7 × 109 cfu per mL. Adult animals were washed 3 times in S-basal buffer and 50–200 animals were placed near the center of an NGM plate, equidistant from the 2 bacterial patches. Animals were allowed to move freely for 1 h, then 5 μL of 1 M sodium azide was added to each bacterial patch to immobilize worms on bacterial patches, and then worms were counted. The bacterial choice index is the number of worms on S. marcescens minus the number of worms on E. coli divided by the number of worms on both bacterial patches. For E. coli with acetone and 2-butanone assays, E. coli was prepared as above at OD600 of 10. Immediately before worms were placed on the assay plates, 2 μL of 1 mM 2-butanone diluted in ethanol and 2 μL of 10 mM acetone diluted in ethanol was pipetted on one E. coli patch and 4 μL of ethanol was pipetted on the other E. coli patch.

Chemotaxis assays

Chemotaxis assays were performed using 10 cm square chemotaxis plates as described (Tsunozaki et al. 2008). In brief, assay agar was 2% agar, 1 mM MgSO4, 1 mM CaCl2 and 5 mM phosphate buffer (pH 6.0). Chemical dilutions were in ethanol at the concentrations indicated in figure legends. 2 μL of diluted chemical were pipetted on one side of the plate, 2 μL of ethanol on the other side, and 2 μL of 1 M sodium azide on both sides to anaesthetize animals that reached odor or ethanol sources. For the chemical bouquet, diluted chemicals were combined and then 2 μL were pipetted on one side of the plate. Animals were washed 2 times in S-basal, once in water and then 50–200 placed in the center of the plate, and the distribution of animals counted after 1 h. The chemotaxis index is the number of worms on odor side minus the number of worms on diluent side divided by the total number of worms that have left the origin.

Aversive olfactory conditioning assays

The odor conditioning assays were modified from Bargmann and Colbert (1995). Adult worms were washed 3 times in S-basal and then transferred to an NGM plate without bacterial food (approximately 500 worms were transferred to each plate). For naïve animals, another NGM plate was used for the lid and inverted above the plate of worms. For the odor-conditioned animals, 4 spots of 3 μL neat chemical was pipetted on an NGM plate, the plate was used for a lid and inverted above the plate with worms. The 2 plates were sealed with parafilm and incubated for 90 min. Then the worms were assayed in chemotaxis assays described above to pure odorants or mixtures of odorants at concentrations indicated in figure legends. For bacterial-conditioned assays, naïve animals were prepared in the same way as for odor conditioned assays. For bacterially conditioned animals, S. marcescens bacteria was grown overnight in LB at 26°C and then resuspended (OD600 = 10). Five spots of 100 μL of the bacterial suspension was spotted onto an NGM plate and incubated for 5 h at 20°C. Then the NGM plate with S. marcescens bacteria was used for a lid and inverted above plate of worms on NGM plate without food, sealed with parafilm and incubated for 90 min. Then the worms were assayed in chemotaxis assays described above to pure odorants or mixtures of odorants at concentrations indicated in figure legends. For the odor of S. marcescens, 5 μL of S. marcescens (OD600 = 1.0) or 5 μL of LB media was pipetted on NGM agar plate and incubated at 20°C for 6.5 h. Then the agar plug with bacteria or media was cut out with a sterile blade and placed on the lid of a chemotaxis plate directly above 2 μL of 1 M sodium azide on the chemotaxis plate.

For all behavioral assays, data points are means of n ≥ 5 assays with between 50–200 animals each, repeated on at least 2 different days.

Solid-phase microextraction gas chromatography mass spectrometry

Bacteria were prepared as for bacterial choice assay. Bacteria were grown overnight in LB at 26°C, centrifuged, and then resuspended at an OD600 = 10. Two NGM plates were prepared, each with 9 spots of 25 μL of bacterial suspension. For the controls, 25 μL of LB media without bacteria was spotted on NGM plate. Plates were incubated for 5 h at 20°C. Then 18 squares (each approximately 8 × 8 mm) of the NGM agar with 25 μL bacteria suspension or LB media were placed in a GC-MS glass vial at 20°C for 1 h (Supplementary Figure 1). Each agar square was extracted using a metal spatula that was sterilized by placing in ethanol, flaming, and allowing to cool at room temperature briefly before use. Samples were analyzed by SPME-GC-MS using the Agilent 6890 GC System equipped with a Restek, Rtx-5 column, and Agilent 5973 Mass Selective Detector. The temperature program was: hold 2 min at 35°C, increased to 140°C at a rate of 10°C/min, and then increased to 250°C at a rate of 100°C/min and hold at 250°C for 3 min. MS ranged from 30 to 550 in full scan mode. VOCs were identified with the NIST 11 (National Institute of Standards and Technology) mass spectral library and pure chemical standards run following the same parameters as for bacterial samples. Bacterial samples of S. marcescens were prepared 3 times on different days and then analyzed, bacterial samples of E. coli of were prepared twice on different days and then analyzed, and LB control samples were run immediately before or after each bacterial sample. To determine approximate absolute quantities of volatiles, pure chemicals of known quantity were run and the quantities of compounds in S. marcescens headspace were calculated relative to the total peak areas of the standards. All standard chemicals were at least 98% purity and were purchased from Sigma-Aldrich (USA).

Results

Identification of S. marcescens attractive volatiles

Although S. marcescens is a bacterial pathogen that can kill C. elegans, the odor of S. marcescens is more attractive to the wild-type C. elegans strain N2 than its standard laboratory food source, E. coli (Zhang et al. 2005; Glater et al. 2014). We evaluated olfactory preferences of C. elegans using a bacterial choice assay in which worms migrate to one of 2 patches of bacteria on opposite sides of an agar plate (Figure 1A) (Zhang et al. 2005). In this assay, the animals’ first approach over 1–2 h is dominated by their olfactory preferences for volatile odors released by the bacteria and C. elegans shows a strong preference for S. marcescens over E. coli (Figure 1B).

Figure 1.

Figure 1.

Open in a new tab

Caenorhabditis elegans prefers Serratia marcescens over E. coli. (A) Cartoon of the bacterial choice assay. Approximately 50–200 worms are placed on an agar plate between 2 patches of bacteria, which they can approach by olfactory chemotaxis. Animals were allowed to move freely for 1 h before plates were scored. Cartoon adapted from Glater et al. 2014. (B) Bacterial choice index of wildtype animals. WT show a significant preference for S. marcescens (OD600 = 1.0) over E. coli (OD600 = 10), *** indicates P < 0.001, one sample Student’s t-test, compared to 0, n = 25 assays. Error bars represent SEM.

We used solid-phase microextraction (SPME) gas chromatography-mass spectrometry (GC-MS) to identify the volatiles present in the headspace of S. marcescens. The headspace refers to the volume of air above the bacterial sample within the GC-MS glass vial. The bacteria were grown on agar plates in the same manner as grown for bacterial choice assays (Supplementary Figure 1). We found consistently in 3 replicate samples made on different days that S. marcescens headspace contained 5 major volatile organic compounds (VOC) that were not present in the headspace of E. coli or the control sample, LB media spotted on NGM agar. We used the NIST 11 (National Institute of Standards and Technology) mass spectral library to identify tentatively the VOCs and confirmed the identity of the VOCs using standards made from pure chemicals. The unique VOCs identified in S. marcescens cultures were: acetone, dimethyl sulfide, 2-butanone, ethyl acetate, and dimethyl disulfide (Figure 2). Retention times of identified VOCs are in Supplementary Table 1.

Figure 2.

Figure 2.

Open in a new tab

Gas chromatography-mass spectrometry of headspace of Serratia marcescens, E. coli and LB media control. (A) Overnight liquid cultures of bacteria were spotted on NGM agar plates and incubated for 5 h, NGM agar squares were placed inside a GC-MS glass vial for 1 h. Then a SPME fiber was inserted into the sample vial containing the bacteria to sample the volatiles in the headspace. Representative total ion chromatograms for S. marcescens (OD600 = 10) (top), E. coli (OD600=10) (middle), and LB media control (bottom). Peaks unique to S. marcescens are numbered and labeled: 1 = acetone, 2 = dimethyl sulfide, 3 = 2-butanone, 4 = ethyl acetate, 5 = dimethyl disulfide. Prominent other peaks not unique to S. marcescens: a = air, b = ethanol, c = isopropyl alcohol, d = 1-butanol, e = indole. Asterisks indicate “fiber peaks,” volatile organic compounds released by the SPME fiber. Peaks were identified with NIST 11 (National Institute of Standards and Technology) mass spectral library and confirmed with known standards. See sample retention times in Supplementary Table 1. (B) Inset shows magnified chromatogram for approximate retention times 2–7 min. Peaks are labeled in the same manner as in (A).

We tested each of the candidate odorants in S. marcescens headspace at a range of concentrations in chemotaxis assays. Of the 5 VOCs, C. elegans showed a strong preference for 2-butanone and acetone at a range of concentrations, but weak or no preference for the other 3 odorants at most concentrations (Figure 3A and C). Ethyl acetate was weakly attractive at high concentrations of 1 M and undiluted stock (Figure 3D). Dimethyl sulfide and dimethyl disulfide were not significantly attractive at all concentrations tested (Figure 3B and E). These results are generally consistent with previous studies that have shown that 2-butanone is attractive at a range of concentrations, and that acetone is attractive when tested at 1:100 dilution (approximately 100 mM) (Bargmann et al. 1993). However, previous work has shown that 1:100 dilution of ethyl acetate (approximately 100 mM) and 1:100 dilution of dimethyl disulfide (approximately 100 mM) to be weakly attractive, but we see no attraction in our chemotaxis assays at these concentrations (Bargmann et al. 1993; Hsueh et al. 2017). This difference in chemotaxis indexes may be because we used square chemotaxis plates while these previous studies used round plates and the concentration gradient encountered by the worms may differ.

Figure 3.

Figure 3.

Open in a new tab

Major attractive volatiles released by Serratia marcescens are acetone and 2-butanone. Chemotaxis to different concentrations of identified volatiles in S. marcescens headspace. All volatile organic compounds (VOCs) were diluted in ethanol. (A) Acetone, (B) dimethyl sulfide, (C) 2-butanone, (D) ethyl acetate, and (E) dimethyl disulfide. Gray bars indicate concentration of VOCs used in odor bouquet at the approximate concentrations in S. marcescens headspace (5 mM acetone, 0.5 mM dimethyl sulfide, 0.5 mM 2-butanone, 0.1 mM ethyl acetate and 1 mM dimethyl disulfide). a indicates P < 0.01, one sample t-test compared to 0; n ≥ 6 assays. Error bars represent SEM. (F) Chemotaxis to odor mixtures as labeled; concentrations are those found in S. marcescens headspace (5 mM acetone, 0.5 mM dimethyl sulfide, 0.5 mM 2-butanone, 0.1 mM ethyl acetate, and 1 mM dimethyl disulfide). ***P < 0.001, *P < 0.05, ANOVA with Dunnett compared to bouquet with all 5 odorants, n ≥ 6 assays. Error bars represent SEM.

Next, we tested an odorant mixture of all 5 volatiles found in S. marcescens headspace at the approximate concentrations detected in the headspace (5 mM acetone, 0.5 mM dimethyl sulfide, 0.5 mM 2-butanone, 0.1 mM ethyl acetate, and 1 mM dimethyl disulfide, Figure 3F). As expected from the results of chemotaxis with single odorants, the bouquet of all 5 odorants was very attractive. Next, we tested different subsets of the odorants (Figure 3F). Acetone is the most attractive compound because the bouquet lacking acetone is significantly less attractive than the bouquet of all 5 volatiles, although this bouquet is still somewhat attractive. This remaining attraction is likely due to 2-butanone, the second most attractive compound in the mixture. The 3 remaining odorants do not contribute to the attractiveness of the bouquet because the bouquet of all 5 odorants is as attractive as the blend of only acetone and 2-butanone. Taken together, this indicates that 2-butanone and acetone are likely the major attractive volatiles in S. marcescens headspace.

Identification of sensory neurons that detect S. marcescens volatiles

In order to evaluate whether the attractive volatiles we identified are what C. elegans uses to recognize S. marcescens, we next determined which sensory neuron(s) detect S. marcescens bacterial odor and then whether these sensory neurons also detect the identified attractive volatiles, acetone, and 2-butanone. Caenorhabditis elegans has 2 pairs of olfactory neurons that sense attractive volatile molecules, AWC and AWA, and 3 pairs of chemosensory neurons that sense repulsive odors, AWB, ASH, and ADL. To determine which neurons are important in preference for S. marcescens, we tested well-characterized C. elegans mutants with defects in the cell fates of specific chemosensory neurons in bacterial choice assays (Figure 4A). Mutations that affected the AWC neuron exhibited defective choice behavior while other mutants exhibited wild-type choice behavior. The odr-7 mutants and lim-4 mutants, which have defects in AWA or AWB cell fates, respectively (Sengupta et al. 1994; Sagasti et al. 1999) resembled wild-type animals in their preference for S. marcescens in the choice assay. The double mutant osm-9 ocr-2, which lacks sensory function of ASH, ADL and AWA neurons (Tobin et al. 2002) also showed a strong preference for S. marcescens. However, the mutant ceh-36, which has a defect in AWC cell fate (Lanjuin et al. 2003; Koga and Ohshima 2004) had a reversed olfactory preference, preferring E. coli to S. marcescens. The mutant ceh-36 also affects the differentiation of ASE taste neurons (Chang et al. 2003; Koga and Ohshima 2004), but the che-1 mutant, which also lacks ASE neurons, retained the normal preference for S. marcescens. These results suggest that AWC neurons are important for S. marcescens preference.

Figure 4.

Figure 4.

Open in a new tab

AWCON mediates Caenorhabditis elegans preference for Serratia marcescens. (A) Bacterial choice (S. marcescens OD600 = 10; E. coli OD600 = 1.0) index of mutants affecting olfactory neuron cell fates or function. (B) Rescue of bacterial choice behavior in tax-4(p678) mutants expressing tax-4 cDNA in AWC (ceh-36* promoter) but not BAG (gcy-33 promoter), ASI (str-3 promoter), or ASK (sra-9 promoter). (C) Bacterial choice behavior of N2 worms expressing tetanus toxin in AWC and AWB weakly (odr-3 promoter) or ASK (sra-9 promoter). (D) nsy-7 rescued in AWC and AWB (odr-3 promoter). ***P < 0.001, compared to WT by ANOVA with Dunnett, n ≥ 6 assays. (E) Effect of different concentrations of S. marcescens (OD600 = 1.0 or OD600 = 0.1) on bacterial choice index of WT and nsy-1 mutant (2 AWCON neurons); the concentration of E. coli (OD600 = 10) was held constant. P value was generated by 2-way ANOVA for interaction of genotype and S. marcescens concentration (OD600 = 1.0 or 0.1); *P <0.05, n ≥ 5 assays.

We also tested the role of BAG neurons because these neurons, which respond to carbon dioxide, have been shown to be important for C. elegans to leave the lawn of S. marcescens after several hours (Brandt and Ringstad 2015). To ask whether the BAG neurons are important for the initial attraction of C. elegans to S. marcescens, we tested gcy-9 mutants which have defective BAG neurons (Hallem et al. 2011). We found that gcy-9 mutants showed wild-type preference for S. marcescens in the choice assay (Figure 4A). Thus BAG neurons are important for leaving S. marcescens, but not for preference for S. marcescens over E. coli.

AWC sensory transduction requires a cyclic nucleotide-gated channel (TAX-4) that is localized to the sensory cilia of multiple chemosensory neurons including AWC (Komatsu et al. 1996; Chang et al. 2003). The tax-4 mutants were defective in S. marcescens preference, consistent with a role for AWC chemosensation in the choice (Figure 4A). Serratia marcescens preference was restored by tax-4 expression under the ceh-36* promoter, which is selective for AWC neurons (Chang et al. 2003; Koga and Ohshima 2004), but was not rescued by tax-4 expression gcy-33 (BAG) promoter, consistent with the gcy-9 mutant data, or under a control sra-9 (ASK) promoter (Figure 4B). Serratia marcescens preference was also disrupted by a tetanus toxin transgene that blocked synaptic release from AWC and AWB neurons, but not by a control ASK tetanus toxin transgene (Figure 4C). Thus the strong preference for S. marcescens likely requires the sensory function and synaptic output of AWC.

The AWC neuron pair consists of 2 cells, AWCON and AWCOFF, that express different chemoreceptor genes and detect partly overlapping sets of odors (Troemel et al. 1999; Wes and Bargmann 2001). Animals mutant for the genes nsy-5 and nsy-7 have cell fate transformations that eliminate the AWCON neuron, resulting in 2 AWCOFF neurons (Chuang et al. 2007; Lesch et al. 2009). Like ceh-36 mutants, nsy-5 and nsy-7 were reversed in choice behaviors, preferring E. coli to S. marcescens (Figure 4A). Expression of nsy-7 in AWC under the odr-3 promoter fully rescued S. marcescens preference in a nsy-7 background (Figure 4D), as well as rescuing the AWCON cell fate (Lesch et al. 2009). By contrast, the nsy-1 mutant, which lacks AWCOFF neurons (Sagasti et al. 2001), retained a strong preference for S. marcescens (Figure 4A). The effects of nsy-5, nsy-7, and nsy-1 indicate that AWCON drives S. marcescens preference.

The cell fate transformation in nsy-1 results in 2 AWCON neurons, and in a standard choice assay nsy-1 appeared to have an even stronger preference for S. marcescens than wild type (Figure 4A). To confirm this result, we presented a 10-fold lower amount of S. marcescens bacteria in the choice with E. coli, a configuration that greatly diminished the wild-type preference for S. marcescens over E. coli (Figure 4E). In this sensitized choice assay, nsy-1 mutants had a strongly enhanced preference for S. marcescens (Figure 4E). These results indicate that the number of AWCON neurons can determine S. marcescens preference: animals with 2 AWCON neurons have an even stronger preference for S. marcescens than wild-type animals with one AWCON neuron.

Taken together, C. elegans had a strong preference for the odor of S. marcescens over the odor of E. coli in a bacterial choice assay. Serratia marcescens preference is highly dependent on a specific class of olfactory neuron, AWCON, which is known to sense a variety of attractive volatile odors. Mutants selectively lacking AWCON or both AWC neurons had a reversed preference favoring E. coli, suggesting that in the absence of AWC neurons, non-AWC neurons can drive either attraction toward E. coli or repulsion from S. marcescens odors.

AWCON detects S. marcescens volatiles: 2-butanone and acetone

Does the AWCON neuron which mediates the olfactory preference for S. marcescens over E. coli also mediate the attraction to acetone and 2-butanone? The AWCON neuron can respond to more than one odorant because a single sensory neuron in C. elegans expresses many different kinds of olfactory G-protein coupled receptors (Bargmann 2006). The volatile odorant 2-butanone has previously been shown to be sensed by the AWCON neuron, but the neuron responding to acetone has not been previously determined (Bargmann 2006; Tsunozaki et al. 2008). We found that acetone was likely also sensed by the AWCON neuron because genetic mutants that affect AWCON function (ceh-36 and nsy-7) fail to chemotax towards acetone (Figure 5A). In addition, tax-4 mutants were defective in acetone chemotaxis, but wildtype acetone chemotaxis was restored by tax-4 expression under the ceh-36* promoter, which is selective for AWC neurons (Figure 5B).

Figure 5.

Figure 5.

Open in a new tab

AWCON mediates chemotaxis to acetone and Escherichia coli with added acetone and 2-butanone. (A) Acetone chemotaxis behavior of genetic mutants affecting olfactory neuron cell fates. (B) Rescue of acetone chemotaxis in tax-4(p678) mutants expressing tax-4 cDNA in AWC (ceh-36* promoter). (C) Bacterial choice between patch of E. coli (OD600 = 10) with 10 mM acetone and 1 mM 2-butanone diluted in ethanol and patch of E. coli (OD600 = 10) with diluent ethanol. N2 shows preference for E. coli with acetone and 2-butanone and nsy-7 mutant shows no preference. (D) Rescue of E. coli (OD600 = 10) with 2-butanone and acetone versus E. coli (OD600=10) alone in tax-4(p678) mutants expressing tax-4 cDNA in AWC (ceh-36* promoter) ***P < 0.001, **P < 0.01 compared to WT by ANOVA with Dunnett, n ≥ 6 assays. Error bars represent SEM.

To test whether 2-butanone and acetone can drive food choice, we conducted a modified bacterial choice assay. Escherichia coli was spotted on both sides and the plate and then immediately before the assay acetone (10 mM) and 2-butanone (1 mM) were added to E. coli patch on one side and the diluent ethanol was added to the E. coli patch on other side. We found that wild-type N2 animals preferred the E. coli patch with acetone and 2-butanone at these concentrations, 2-fold higher than the approximate concentrations in the S. marcescens headspace, but not at lower concentrations (Figure 5C, data not shown). We then tested the role of the AWCON neuron in detecting 2-butanone and acetone in the presence of E. coli. First, we examined nsy-7 mutants in the modified 2-butanone and acetone E. coli choice assay. As expected, the nsy-7 mutants which lack the AWCON neuron showed no preference for the E. coli patch with 2-butanone and acetone over E. coli alone (Figure 5C). Second, we tested tax-4 mutants and found that they were defective in the E. coli plus 2-butanone and acetone assay (Figure 5D). Preference for E. coli with 2-butanone and acetone was restored by tax-4 expression in AWC neurons. These results indicate that the AWCON neuron drives preference for acetone and 2-butanone in the presence of background volatiles released by E. coli.

Conditioning with S. marcescens odor

Our results indicate that the detection of 2-butanone and acetone by the AWCON neuron can drive preference for S. marcescens. However, how does C. elegans recognize the odor of S. marcescens? In other words, does the 5 component odor blend, 2-component odor blend of acetone and 2-butanone, or a single odorant represent S. marcescens to C. elegans? To answer this question, we conducted a modified version of an aversive conditioning assay (Bargmann et al. 1993; Bargmann and Colbert 1995). In aversive conditioning assays, worms are exposed to an odorant in the absence of food for 90 min. These “conditioned” worms show a lower preference for that specific odorant than “naïve” worms that were starved in the absence of the odorant. We then assayed their chemotaxis to the natural odor of S. marcescens, single odorants present in the S. marcescens headspace, and blends of these odorants (Figure 6A). We hypothesized that C. elegans would reduce its preference for the volatile odorants that resemble its representation of the odor of S. marcescens. We tested conditioned worms for chemotaxis towards the natural odor of S. marcescens bacteria and found that C. elegans showed reduced chemotaxis towards the natural odor. We then tested pure odor components in the headspace of S. marcescens. We tested chemotaxis towards 1 mM acetone, a concentration approximately 5-fold more dilute than in the S. marcescens headspace because in previous pure odor conditioning experiments, worms showed a reduction in preference for odors more dilute than the conditioning odor (Bargmann and Colbert 1995). We found that C. elegans conditioned with S. marcescens reduced its chemotaxis towards acetone. This reduction in chemotaxis was similar to the reduction in chemotaxis towards the natural odor of S. marcescens (Figure 6B). For comparison, we tested the other components at 1 mM. We found a reduction in preference for the attractive odorant, 2-butanone, but not for neutral odorants, ethyl acetate and dimethyl disulfide (Figure 6B). As expected, the conditioned worms did not show a reduction in chemotaxis towards isoamyl alcohol, a known attractive odor detected by AWC, but not present in the headspace of S. marcescens (Bargmann et al. 1993).

Figure 6.

Figure 6.

Open in a new tab

Aversive olfactory conditioning with odor of Serratia marcescens reduces preference to 2-butanone and acetone. (A) Schematic of aversive bacterial conditioning assay. Adult worms are washed off food, starved for 90 min in the presence of odor of S. marcescens (OD600 = 10) growing on NGM agar lid (S. marcescens conditioned) or NGM agar lid alone (naïve) and then washed again and tested in chemotaxis assays to various single odorants and odor mixtures. (B) Naïve and conditioned worms were tested for chemotaxis to natural odor of S. marcescens (bacterial patch (OD600 = 1.0) versus LB media on NGM agar on lid of chemotaxis plate); single odorant components (1 mM) present in S. marcescens headspace; to a mixture of all 5 odorants at 5-fold lower concentrations than approximate concentrations in S. marcescens headspace (1 mM acetone, 0.1 mM dimethyl sulfide, 0.1 mM 2-butanone, 0.02 mM ethyl acetate, and 0.2 mM dimethyl disulfide); a mixture of 1 mM acetone and 0.1 mM 2-butanone; and 10 mM isoamyl alcohol, an odorant not found in S. marcescens headspace. P values were generated by 2-way ANOVA for interaction of tested chemotaxis odor and condition (naïve or S. marcescens conditioned); each tested chemical odor or mixture of odors was compared to natural S. marcescens odor; as well as additional comparison between acetone alone and mixture of acetone and 2-butanone (ns, P = 0.51). *P < 0.05, n ≥ 5 assays. Error bars represent SEM.

We next tried odor mixtures. We tested the conditioned worms for chemotaxis towards the bouquet of all 5 odorants at a 5-fold lower concentration for all odorants than the approximate concentration of odorants in S. marcescens headspace. The conditioned worms showed a similar reduction in chemotaxis to the bouquet of all 5 odorants, the mixture of acetone and 2-butanone and the natural odor of S. marcescens. In addition, we found that the reduction in preference for acetone alone and for the mixture of acetone and 2-butanone were not significantly different (2-way ANOVA, P = 0.51). This likely indicates that the change in response to these odors is non-additive. Thus taken together, it seems the 2 components, acetone and 2-butanone, either alone or together, represent the more complex mixture of at least 5 volatiles in S. marcescens headspace.

Next we did the reciprocal experiment. We odor conditioned worms with different single odorants found in S. marcescens headspace paired with food deprivation for 90 min and then tested the conditioned and naïve worms in a S. marcescens and E. coli bacterial choice assay (Figure 7A). We hypothesized that conditioning with attractive odorants in the S. marcescens headspace would reduce preference for S. marcescens, but conditioning with neutral odorants would not. As expected, conditioning with acetone alone reduced S. marcescens preference greatly and conditioning with 2-butanone also reduced S. marcescens greatly (Figure 7B). In addition, conditioning to neutral odorants in the headspace of S. marcescens, ethyl acetate and dimethyl disulfide, did not reduce S. marcescens preference.

Figure 7.

Figure 7.

Open in a new tab

Aversive olfactory conditioning with single odorants reduces preference for Serratia marcescens. (A) Schematic of aversive odor conditioning assay. Adult worms are washed off food, starved for 90 min in the presence of odor on NGM agar lid (odor conditioned) or NGM agar lid alone (naïve) and then washed again and tested in bacterial choice assays between S. marcescens (OD600 = 1.0) and E. coli (OD600 = 10). (B) Odor conditioned worms were exposed to single odor components present in S. marcescens headspace (acetone, 2-butanone, ethyl acetate, and dimethyl disulfide) as well as a single control odorant not found in S. marcescens headspace (isoamyl alcohol). In addition, worms were conditioned to a mixture of an odorant present in S. marcescens headspace, acetone, and absent, isoamyl alcohol (IAA). ***P < 0.001, odor conditioned worms compared to naïve worms by ANOVA with Dunnett, n ≥ 6 assays. Error bars represent SEM.

Last, we preexposed worms to isoamyl alcohol, which is not present in S. marcescens headspace, and tested S. marcescens preference in the bacterial choice assay. As expected, the worms did not show a reduced preference for S. marcescens. Next, we investigated the response of the worms after being conditioned with an odor blend. We conditioned food deprived worms with a mixture of acetone (present in S. marcescens) and isoamyl alcohol (absent in S. marcescens) and then assayed preference. We hypothesized that worms would not lower their preference for S. marcescens because this blend of acetone and isoamyl alcohol would be very different from the odor of S. marcescens. However, they did reduce their preference. Thus, the worms seem to be attending to the components of the conditioning odor blend and so now avoid an odor blend that contains acetone, but is otherwise very different from the conditioning odor blend.

Discussion

Caenorhabditis elegans is attracted to volatile chemicals released by bacteria and can distinguish among the blends of volatiles released by different species of bacteria, which is important for finding nutritious food and avoiding harmful microbes (Bargmann et al. 1993; Zhang et al. 2005; Shtonda and Avery 2006). However, the precise odors that C. elegans uses to detect and recognize a specific bacterial species have not been identified. In this study, we determined that the volatile chemical cues mediating recognition of pathogenic S. marcescens are 2-butanone and acetone. The amphid chemosensory neuron, AWCON, senses both S. marcescens-released 2-butanone and acetone and is necessary for attraction to the natural odor of S. marcescens. In an aversive learning assay where the odor of S. marcescens is paired with food deprivation, C. elegans reduces its chemotaxis similarly for the blend of acetone and 2-butanone and the odor of S. marcescens, indicating that the blend of acetone and 2-butanone represents S. marcescens to C. elegans.

The neuronal basis of preference for S. marcescens

The standard C. elegans laboratory strain N2 has a strong preference for the odor of S. marcescens over the odor of E. coli HB101 in an olfactory choice assay. S. marcescens preference is highly dependent on a specific class of olfactory neuron, AWCON, which is known to sense a variety of attractive volatile odors. Interestingly, AWCON has been shown to recognize several ketones, including 2-butanone, 2-heptanone and in this study, acetone (Bargmann et al. 1993; Zhang et al. 2016). Mutants selectively lacking AWCON or both AWC neurons had a reversed preference favoring E. coli, suggesting that non-AWC neurons can drive either attraction toward E. coli or repulsion from S. marcescens odors.

In a previous study, N2 preference for the bacterium Pseudomonas aeruginosa PA14 over E. coli OP50 was found to require both AWC and AWB olfactory neurons (Ha et al. 2010). Although both Serratia and Pseudomonas preference required AWC, a complete reversal of preference was observed only in the S. marcescens–E. coli choice, and not in the Pseudomonas–E. coli choice. Moreover, in our assay the AWB mutant lim-4 did not affect S. marcescens preference. These results suggest that the identity of the neurons that mediate olfactory choice behavior and the magnitude of their effects vary based on the specific bacteria that are tested.

Volatiles in S. marcescens headspace

We identified 5 major volatiles in the S. marcescens headspace, or volume of air above bacteria within the GC-MS vial. We would like to highlight that there may be additional low abundance volatiles in the S. marcescens headspace that were not detected by the methods used in this study. We intentionally grew the bacteria for GC-MS analysis in as similar a way as possible to the way the bacteria were grown in the bacterial choice assay. In particular, this meant that the bacteria were grown on NGM agar for 5 h and then placed in a GC-MS glass vial for only 1 h because that because this means that the bacteria is in the same bacterial growth state as during 1 h bacterial choice assays (Supplementary Figure 1). However, this short incubation time in the GC-MS glass vial means that high abundance volatiles are detected, but low abundance volatiles may be under the threshold for detection. Nevertheless, we found that the high abundance volatiles that we did detect, acetone and 2-butanone, were sufficient to represent S. marcescens in aversive olfactory conditioning assays and to drive behavior.

To determine, the similarity of odor profiles of different strains of S. marcescens, we examined 2 additional strains of S. marcescens, pathogenic Db11 and its non-pathogenic derivative Db1140 (Kurz et al. 2003). Db11 and Db1140 are non-pigmented strains of S. marcescens that do not produce the red pigment prodigiosin (Thomson et al. 2000). Non-pigmented S. marcescens strains are more commonly found in hospitals and involved in human infections while pigmented strains, such as ATCC 274 used in this study, are more commonly found in natural environments (Hejazi and Falkiner 1997). We found that Db11 released volatiles similar to S. marcescens ATCC 274: acetone, 2-butanone and dimethyl disulfide, but not ethyl acetate or dimethyl sulfide (Supplementary Table 1 and Figure 2). Db11 also released 2-pentatone, which was not detected in the headspace of ATCC 274. We found that the headspace of non-pathogenic derivative Db1140 differed somewhat from pathogenic Db11. Db1140-released acetone and pyrrole, which was not detected in Db11 or ATCC 274, but not 2-butanone and dimethyl disulfide (Supplementary Table 1 and Figure 2). Interestingly, if Db1140 was grown over a longer period of time (5 h instead of 1 h) in the GC-MS glass vial from which the headspace was sampled, 2-butanone was detected, but not dimethyl disulfide (data not shown). This result likely means that Db1140 does release 2-butanone, but at much lower levels than Db11. Thus, the 3 S. marcescens strains examined are similar in that they all release the attractive odorants acetone and 2-butanone, but there are differences among the odor profiles of the strains as well.

Recognition of odor blends by C. elegans

We examined how C. elegans recognizes an odor blend. We found that although we detected 5 major volatiles in the headspace of S. marcescens, C. elegans was only attracted to 2 out of these 5 volatiles, acetone and 2-butanone. We next asked what component or components of the odor blend of S. marcescens results in reduced chemotaxis after pairing the odor of S. marcescens with food deprivation. We found that acetone and 2-butanone represent S. marcescens to C. elegans. This result is similar to insects, which in the majority of studies, use a few key components to recognize an odor blend of volatiles from host plants (Pham-Delegue et al. 1993; Bruce et al. 2005). Although C. elegans is estimated to have approximately 2000 olfactory receptors and may have the olfactory repertoire to respond to many components in a mixture, at least in this case, it only responded behaviorally to 2 representative components in olfactory association assays.

It is noteworthy that 2-butanone is one of the odorants for recognition of S. marcescens because 2-butanone is a commonly used odorant to study olfactory plasticity in the C. elegans literature. For example, when 2-butanone is paired with food deprivation, C. elegans decreases its attraction to 2-butanone (Bargmann and Colbert 1995). In addition, when 2-butanone is paired with feeding, C. elegans increases its attraction to 2-butanone (Torayama et al. 2007). Thus, this study shows that 2-butanone is likely an ethologically relevant stimulus to C. elegans.

These odorants, 2-butanone and acetone, are not unique to S. marcescens. In fact, they are ubiquitous and released by many species of bacteria, both pathogenic and non-pathogenic (Lemfack et al. 2014). We propose that because many species of bacteria release similar volatiles, C. elegans must distinguish among them by recognizing different ratios of the attractive components of the odor mixture. Indeed, insects have been shown mostly to recognize the volatiles of host plants by different ratios of constitutive components as opposed to components that are unique to specific host plants (Bruce and Pickett 2011). In order to test this hypothesis, the components and ratio of components needed for recognition of bacteria by C. elegans will need to be determined for several species of bacteria. We are currently working on this for bacterial species isolated from the natural microbiome of C. elegans (Samuel et al. 2016).

Attraction to pathogenic S. marcescens

It is surprising that C. elegans is strongly attracted to the odor of S. marcescens, a pathogenic bacterium that can kill infected animals in a few days. Interestingly, C. elegans is also attracted to the odors of other harmful microbes. In 2 recent studies, pathogenic bacterium Bacillus nematocida and the nematophagous fungus Arthrobotrys oligospora were shown to be attractive to C. elegans and to release ubiquitous volatiles as well (Niu et al. 2010; Hsueh et al. 2017). Although there should be a strong selection for avoidance of odors of harmful microbes, if harmful and harmless microbes release similar volatiles, how can C. elegans avoid them? On a short timescale, C. elegans can learn to avoid microbes that cause an adverse experience (Zhang et al. 2005). Over multiple generations, natural populations of C. elegans may evolve to avoid the microbes in its local environment. Indeed, in our previous work we found that level of attraction for the odor of S. marcescens among wild strains of C. elegans does vary, which indicates that the level of attraction may indeed be a rapidly evolving trait (Glater et al. 2014). In addition in an experimental co-evolution study after being passaged with the S. marcescens for 30 generations, C. elegans was shown to increase its avoidance of the pathogen, based on both olfactory and non-olfactory cues (Penley and Morran 2017). Thus, the ability to modify preferences for specific microbes within an organism’s lifetime as well as on an evolutionary timescale may be an important feature of how C. elegans navigates the complex microbial environment.

In conclusion, we have determined that 2 key components represent the natural blend of S. marcescens to C. elegans. This finding is both an important step forward for understanding the chemical basis of the interaction between C. elegans and S. marcescens as well as the first study to examine how the worm recognizes odor blends. Knowledge about how C. elegans recognizes natural odor blends is likely to be useful for increased understanding of how evolutionarily related parasitic nematodes recognize their human, animal or plant hosts, and for development of reagents that could be useful for nematode control. We expect that this work will provide the foundation for future work to elucidate further the chemical, neuronal, and molecular basis of microbe recognition by C. elegans and their role in host-pathogen coevolution.

Supplementary material

Supplementary material are available at Chemical Senses online.

Supplementary Material
Click here for additional data file. (947.2KB, pdf)

Funding

This work was supported by Pomona College and a grant to Pomona College from the Howard Hughes Medical Institute through the Precollege and Undergraduate Science Education Program. Some strains were provided by the CGC, which is funded by NIH Office of Research Infrastructure Programs (P40 OD010440).

Acknowledgements

The authors would like to thank Julia Lee and Alexandra Antonoplis for developing SPME GC-MS protocols for analyzing bacterial headspace and Drs. Cornelia Bargmann, Sreekanth Chalasani, Sabine Rech, Miri VanHoven, and Paul Sternberg for helpful discussion.

References

  1. Bargmann CI. 2006. Chemosensation in C. elegans. The C. elegans Research Community, Wormbook. http://www.wormbook.org(accessed 31 January 2018). [Google Scholar]
  2. Bargmann CI, Colbert HA. 1995. Odorant-specific adaptation pathways generate olfactory plasticity in C. elegans. Neuron. 14: 803–812. [DOI] [PubMed] [Google Scholar]
  3. Bargmann CI, Hartwieg E, Horvitz HR. 1993. Odorant-selective genes and neurons mediate olfaction in C. elegans. Cell. 74:515–527. [DOI] [PubMed] [Google Scholar]
  4. Beale E, Li G, Tan MW, Rumbaugh KP. 2006. Caenorhabditis elegans senses bacterial autoinducers. Appl Environ Microbiol. 72:5135–5137. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Brandt JP, Ringstad N. 2015. Toll-like receptor signaling promotes development and function of sensory neurons required for a C. elegans pathogen-avoidance behavior. Curr Biol. 25:2228–2237. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Brenner S. 1974. The genetics of Caenorhabditis elegans.Genetics. 77:71–94. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Bruce TJ, Pickett JA. 2011. Perception of plant volatile blends by herbivorous insects–finding the right mix. Phytochemistry. 72:1605–1611. [DOI] [PubMed] [Google Scholar]
  8. Bruce TJ, Wadhams LJ, Woodcock CM. 2005. Insect host location: a volatile situation. Trends Plant Sci. 10:269–274. [DOI] [PubMed] [Google Scholar]
  9. Carrillo MA, Guillermin ML, Rengarajan S, Okubo RP, Hallem EA. 2013. O2-sensing neurons control CO2 response in C. elegans. J Neurosci. 33:9675–9683. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Chang S, Johnston RJ Jr, Hobert O. 2003. A transcriptional regulatory cascade that controls left/right asymmetry in chemosensory neurons of C. elegans. Genes Dev. 17:2123–2137. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Choi JI, Yoon K, Subbammal Kalichamy S, Yoon S-S, Il Lee J. 2016. A natural odor attraction between lactic acid bacteria and the nematode Caenorhabditis elegans. ISME J. 10:558–567. [DOI] [PMC free article] [PubMed]
  12. Chuang CF, Vanhoven MK, Fetter RD, Verselis VK, Bargmann CI. 2007. An innexin-dependent cell network establishes left-right neuronal asymmetry in C. elegans.Cell. 129:787–799. [DOI] [PubMed] [Google Scholar]
  13. Dusenbery DB, Sheridan RE, Russell RL. 1975. Chemotaxis-defective mutants of the nematode Caenorhabditis elegans. Genetics. 80:297–309. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Félix MA, Duveau F. 2012. Population dynamics and habitat sharing of natural populations of Caenorhabditis elegans and C. briggsae. BMC Biol. 10:59. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Glater EE, Rockman MV, Bargmann CI. 2014. Multigenic natural variation underlies Caenorhabditis elegans olfactory preference for the bacterial pathogen Serratia marcescens. G3 (Bethesda). 4:265–276. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Ha HI, Hendricks M, Shen Y, Gabel CV, Fang-Yen C, Qin Y, Colón-Ramos D, Shen K, Samuel AD, Zhang Y. 2010. Functional organization of a neural network for aversive olfactory learning in Caenorhabditis elegans. Neuron. 68:1173–1186. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Hallem EA, Spencer WC, McWhirter RD, Zeller G, Henz SR, Rätsch G, Miller DM 3rd, Horvitz HR, Sternberg PW, Ringstad N. 2011. Receptor-type guanylate cyclase is required for carbon dioxide sensation by Caenorhabditis elegans. Proc Natl Acad Sci USA. 108:254–259. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Hallem EA, Sternberg PW. 2008. Acute carbon dioxide avoidance in Caenorhabditis elegans. Proc Natl Acad Sci USA. 105:8038–8043. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Harris G, Shen Y, Ha H, Donato A, Wallis S, Zhang X, Zhang Y. 2014. Dissecting the signaling mechanisms underlying recognition and preference of food odors. J Neurosci. 34:9389–9403. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Hejazi A, Falkiner FR. 1997. Serratia marcescens. J Med Microbiol. 46:903–912. [DOI] [PubMed] [Google Scholar]
  21. Hsueh Y-P, Gronquist MR, Schwarz EM, Nath RD, Lee C-H, Gharib S, Schroeder FC, Sternberg PW. 2017. Nematophagous fungus Arthrobotrys oligospora mimics olfactory cues of sex and food to lure its nematode prey. ELife. 6:e20023. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Koga M, Ohshima Y. 2004. The C. elegans ceh-36 gene encodes a putative homemodomain transcription factor involved in chemosensory functions of ASE and AWC neurons. J Mol Biol. 336:579–587. [DOI] [PubMed] [Google Scholar]
  23. Komatsu H, Mori I, Rhee JS, Akaike N, Ohshima Y. 1996. Mutations in a cyclic nucleotide-gated channel lead to abnormal thermosensation and chemosensation in C. elegans. Neuron. 17:707–718. [DOI] [PubMed] [Google Scholar]
  24. Kurz CL, Chauvet S, Andrès E, Aurouze M, Vallet I, Michel GP, Uh M, Celli J, Filloux A, De Bentzmann S et al. 2003. Virulence factors of the human opportunistic pathogen Serratia marcescens identified by in vivo screening. EMBO J. 22:1451–1460. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Lanjuin A, VanHoven MK, Bargmann CI, Thompson JK, Sengupta P. 2003. Otx/otd homeobox genes specify distinct sensory neuron identities in C. elegans. Dev Cell. 5:621–633. [DOI] [PubMed] [Google Scholar]
  26. Lemfack MC, Nickel J, Dunkel M, Preissner R, Piechulla B. 2014. mVOC: a database of microbial volatiles. Nucleic Acids Res. 42:D744–D748. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Lesch BJ, Bargmann CI. 2010. The homeodomain protein hmbx-1 maintains asymmetric gene expression in adult C. elegans olfactory neurons. Genes Dev. 24:1802–1815. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Lesch BJ, Gehrke AR, Bulyk ML, Bargmann CI. 2009. Transcriptional regulation and stabilization of left-right neuronal identity in C. elegans. Genes Dev. 23:345–358. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Macosko EZ, Pokala N, Feinberg EH, Chalasani SH, Butcher RA, Clardy J, Bargmann CI. 2009. A hub-and-spoke circuit drives pheromone attraction and social behaviour in C. elegans. Nature. 458:1171–1175. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Mallo GV, Kurz CL, Couillault C, Pujol N, Granjeaud S, Kohara Y, Ewbank JJ. 2002. Inducible antibacterial defense system in C. elegans. Curr Biol. 12:1209–1214. [DOI] [PubMed] [Google Scholar]
  31. Niu Q, Huang X, Zhang L, Xu J, Yang D, Wei K, Niu X, An Z, Bennett JW, Zou C et al. 2010. A Trojan horse mechanism of bacterial pathogenesis against nematodes. Proc Natl Acad Sci USA. 107:16631–16636. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Penley MJ, Morran LT. 2017. Host mating system and coevolutionary dynamics shape the evolution of parasite avoidance in Caenorhabditis elegans host populations. Parasitology. 144:1–8. [DOI] [PubMed] [Google Scholar]
  33. Pham-Delegue MH, Bailez O, Blight MM, Masson C, Picard-Nizou AL, Wadhams LJ. 1993. Behavioural discrimination of oilseed rape volatiles by the honeybee Apis mellifera L. Chem Senses. 18:483–494. [Google Scholar]
  34. Pradel E, Zhang Y, Pujol N, Matsuyama T, Bargmann CI, Ewbank JJ. 2007. Detection and avoidance of a natural product from the pathogenic bacterium Serratia marcescens by Caenorhabditis elegans. Proc Natl Acad Sci USA. 104:2295–2300. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Pujol N, Link EM, Liu LX, Kurz CL, Alloing G, Tan MW, Ray KP, Solari R, Johnson CD, Ewbank JJ. 2001. A reverse genetic analysis of components of the Toll signaling pathway in Caenorhabditis elegans. Curr Biol. 11:809–821. [DOI] [PubMed] [Google Scholar]
  36. Sagasti A, Hisamoto N, Hyodo J, Tanaka-Hino M, Matsumoto K, Bargmann CI. 2001. The CaMKII UNC-43 activates the MAPKKK NSY-1 to execute a lateral signaling decision required for asymmetric olfactory neuron fates. Cell. 105:221–232. [DOI] [PubMed] [Google Scholar]
  37. Sagasti A, Hobert O, Troemel ER, Ruvkun G, Bargmann CI. 1999. Alternative olfactory neuron fates are specified by the LIM homeobox gene lim-4. Genes Dev. 13:1794–1806. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Samuel BS, Rowedder H, Braendle C, Félix MA, Ruvkun G. 2016. Caenorhabditis elegans responses to bacteria from its natural habitats. Proc Natl Acad Sci USA. 113:E3941–E3949. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Sengupta P, Colbert HA, Bargmann CI. 1994. The C. elegans gene odr-7 encodes an olfactory-specific member of the nuclear receptor superfamily. Cell. 79:971–980. [DOI] [PubMed] [Google Scholar]
  40. Shtonda BB, Avery L. 2006. Dietary choice behavior in Caenorhabditis elegans. J Exp Biol. 209:89–102. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Thomson NR, Crow MA, McGowan SJ, Cox A, Salmond GP. 2000. Biosynthesis of carbapenem antibiotic and prodigiosin pigment in Serratia is under quorum sensing control. Mol Microbiol. 36:539–556. [DOI] [PubMed] [Google Scholar]
  42. Tobin DM, Madsen DM, Kahn-Kirby A, Peckol EL, Moulder G, Barstead R, Maricq AV, Bargmann CI. 2002. Combinatorial expression of TRPV channel proteins defines their sensory functions and subcellular localization in C. elegans neurons. Neuron. 35:307–318. [DOI] [PubMed] [Google Scholar]
  43. Torayama I, Ishihara T, Katsura I. 2007. Caenorhabditis elegans integrates the signals of butanone and food to enhance chemotaxis to butanone. J Neurosci. 27:741–750. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Troemel ER, Sagasti A, Bargmann CI. 1999. Lateral signaling mediated by axon contact and calcium entry regulates asymmetric odorant receptor expression in C. elegans. Cell. 99:387–398. [DOI] [PubMed] [Google Scholar]
  45. Tsunozaki M, Chalasani SH, Bargmann CI. 2008. A behavioral switch: cGMP and PKC signaling in olfactory neurons reverses odor preference in C. elegans. Neuron. 59:959–971. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Werner KM, Perez LJ, Ghosh R, Semmelhack MF, Bassler BL. 2014. Caenorhabditis elegans recognizes a bacterial quorum-sensing signal molecule through the AWCON neuron. J Biol Chem. 289: 26566–26573. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Wes PD, Bargmann CI. 2001. C. elegans odour discrimination requires asymmetric diversity in olfactory neurons. Nature. 410:698–701. [DOI] [PubMed] [Google Scholar]
  48. Zhang C, Zhao N, Chen Y, Zhang D, Yan J, Zou W, Zhang K, Huang X. 2016. The Signaling pathway of Caenorhabditis elegans mediates chemotaxis response to the attractant 2-heptanone in a trojan horse-like pathogenesis. J Biol Chem. 291:23618–23627. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Zhang Y, Lu H, Bargmann CI. 2005. Pathogenic bacteria induce aversive olfactory learning in Caenorhabditis elegans. Nature. 438:179–184. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Material
Click here for additional data file. (947.2KB, pdf)

Articles from Chemical Senses are provided here courtesy of Oxford University Press

RESOURCES