SRD‐1 in AWA neurons is the receptor for female volatile sex pheromones in C. elegans males

Abstract

Pheromones are critical cues for attracting mating partners for successful reproduction. Sexually mature Caenorhabditis remanei virgin females and self‐sperm‐depleted Caenorhabditis elegans hermaphrodites produce volatile sex pheromones to attract adult males of both species from afar. The chemoresponsive receptor in males has remained unknown. Here, we show that the male chemotactic behavior requires amphid sensory neurons (AWA neurons) and the G‐protein‐coupled receptor SRD‐1. SRD‐1 expression in AWA neurons is sexually dimorphic, with the levels being high in males but undetectable in hermaphrodites. Notably, srd‐1 mutant males lack the chemotactic response and pheromone‐induced excitation of AWA neurons, both of which can be restored in males and hermaphrodites by AWA‐specific srd‐1 expression, and ectopic expression of srd‐1 in AWB neurons in srd‐1 mutants results in a repulsive behavioral response in both sexes. Furthermore, we show that the C‐terminal region of SRD‐1 confers species‐specific differences in the ability to perceive sex pheromones between C. elegans and C. remanei. These findings offer an excellent model for dissecting how a single G‐protein‐coupled receptor expressed in a dimorphic neural system contributes to sex‐specific behaviors in animals.

Introduction

Successful reproduction and maintenance of genetic diversity are vital to the survival of all species 1, 2. Numerous animals rely on chemical communication to locate members of the opposite sex and con‐specific partners, and in these cases, the release of a chemical signal, frequently a sex pheromone, is an indication of sexual maturity. This signal conveys differential information for both sexes and the fertility status of an individual. Previous studies on this topic have engendered the hypothesis that communication by chemical signals and production of a potent and specific sex pheromone would enhance species viability, and thus, selection pressure would drive evolutionary convergence to this communication mode 2, 3, 4, 5, 6, 7, 8, 9, 10.

Nematodes employ a similar chemoattraction scheme 4, 7, 11, and we found that nematode hermaphrodites and females use a chemical pheromone to attract males. In Caenorhabditis elegans, the chemosensory mechanism has been comprehensively elucidated at the molecular level from studies on hermaphrodites 12, and this provides a strong foundation for studying the poorly understood sex pheromone perception in males. Nematode sex pheromones can be classified broadly into volatile and non‐volatile pheromones. In C. elegans, non‐volatile sex pheromones are represented by the well‐studied water‐soluble ascarosides, which perform diverse biological functions, from short‐range male attraction, hermaphrodite repulsion, olfactory plasticity, and aggregation to dauer formation 11, 13, 14, 15, 16, 17. Conversely, volatile sex pheromones in nematodes are defined as long‐range chemical cues that help individuals locate mating partners from afar 4, 7. Little is currently known about the identity of these pheromones. Studies on better‐characterized sex pheromones from other animals, including insects and rodents, frequently provide information regarding three crucial characteristics: (i) the existence of the pheromones as a mixture of chemical components; (ii) the key enzymes required for their biosynthetic process; and (iii) the unique receptors required for their perception 2, 3, 4, 5, 6, 8, 14, 18. Because these pheromones are volatile, they are commonly low‐molecular‐weight chemicals.

In nematodes, the molecular mechanisms underlying the actions of volatile sex pheromones have been dissected to a limited extent. However, the findings of recent studies have offered an opportunity for further investigation. Con‐specific nematode adult males were reported to be attracted by volatile sex pheromones produced by (i) sexually mature virgin Caenorhabditis remanei females, (ii) self‐sperm‐depleted virgin C. elegans hermaphrodites, or (iii) sexually mature virgin C. elegans females but with defective self‐sperm 4, 7. Based on these findings, we hypothesized that closely related C. elegans and C. remanei might share evolutionarily conserved male‐specific receptors that detect at least one shared component in the volatile pheromone mixtures from the two species. Most chemosensory odorant receptors in C. elegans are G‐protein‐coupled receptors (GPCRs), and all six non‐volatile sex pheromone receptors reported thus far are also GPCRs 11, 19, 20, 21, 22; consequently, we speculated that this volatile sex pheromone receptor in the male is a GPCR, which might share downstream components with other GPCRs in eliciting chemoattraction behaviors 12. This hypothesis helped us formulate our search for this candidate receptor and facilitated parallel interrogation of its downstream components by using both molecular genetics and cell biology approaches, which led to the identification of critical sensory neurons and a putative receptor required for sex pheromone perception in the male nematode.

Results

Amphid neurons are required for pheromone perception

Given that C. elegans males differ from hermaphrodites by harboring several extra male‐specific cells whose sex‐specific functions remain poorly defined, the pheromone receptor was hypothesized to reside in one of these cells. The male‐specific cephalic neuron (CEM) was suggested to function in volatile sex pheromone perception 4, but no chemoreceptor of volatile sex pheromone has been identified to date in CEMs. Our previous study revealed that males from neither CEM‐defective mutant strains, ceh‐30(sm130)X and unc‐86(sm117)III, nor CEM‐ablated strains, completely lose their chemoattraction response toward volatile sex pheromones. The findings suggest that other chemosensory neurons could be critical for perceiving the chemical cues that elicit this male‐specific behavioral response. The functional components of sex pheromones are volatile and act in a sex‐ and stage‐specific manner 4, and to study their action here, we used the chemoattraction behavioral assay described in Materials and Methods (Fig 1A). The pheromone solution deposited on the lid of a Petri dish evoked the same normal chemoattractive behavior in wild‐type C. elegans males on the agar surface (Fig 1B), but after vaporization, the C. remanei female extract was no longer attractive to the males (Fig 1B). Therefore, the chemical likely acts through one or a subset of chemosensory amphid neurons required for detecting volatile attractants. Each amphid contains 12 sensory neurons whose neuronal endings are embedded in the sheath and socket cells. These neurons are responsible for diverse sensations, including physical and chemical sensation and nociception, as well as for navigation, lifespan regulation, and dauer formation. Mutants harboring defective amphids typically display aberrant chemosensation 12. We used the chemoattraction behavioral assay to test whether a mutant loses its chemoattraction to the applied sex pheromone; we tested mutants with known defects in specific amphid neurons that change their cell fate or specific cellular functions. Here, the attractiveness of animals toward a particular chemical is represented by a chemoattraction index (C.I.) (Fig 1A), the reduction in which implies the requirement of a cell or a molecular component functioning therein.

Figure 1. Amphid neurons are required for pheromone perception.

1. In our chemoattraction‐assay setup, nematode chemoattraction behavior is quantified using a measurable chemoattraction index (C.I.); the higher the C.I. measured in the assay, the stronger the attraction of the males to the sex pheromone.
2. After vaporization (VAC) at 4°C for 24 h, the Caenorhabditis remanei female extract failed to attract wild‐type Caenorhabditis elegans males. Conversely, a drop of the sex pheromone extract placed on the lid of a 60‐mm Petri dish evoked chemoattractive behavior in wild‐type C. elegans males (volatility test). Thus, the functional components in C. remanei female sex pheromone are volatile substances.
3. Mutant males with established defects in amphid neurons responded to C. remanei sex pheromone to a lesser degree relative to wild‐type, demonstrating that some or all of the amphid neurons are required for sex pheromone perception.

Data information: We assayed 400 males from each strain by using the sex pheromone chemoattraction assay. Two biological replicates were combined into a single value. Significance was determined using one‐way ANOVA with Bonferroni correction: ***P < 0.001. Means ± SEM (error bars) are shown.

We determined whether defects in specific amphid neurons lead to a loss of ability in sex pheromone perception. In daf‐6, daf‐10, che‐3, osm‐6, and osm‐3 mutant males exhibiting defective amphid‐neuron functions, diverse chemotactic processes are affected, including chemotaxis, osmotaxis, navigation, lifespan regulation, nociception, nose touch, male mating, and dauer formation 12, 23, 24, 25, 26, 27, 28 (Appendix Table S1). Accordingly, our results showed that sex pheromone attraction was reduced in the males of all these mutants (Fig 1C) and confirmed that amphid sensory neurons are required for sex pheromone perception. Notably, in osm‐3 mutant animals, the cilia of amphid neurons are defective, which suggests that the hypothesized sex pheromone receptors might reside on the cilia of one or more of the amphid neurons in the males.

AWA neurons are required for pheromone perception

We next tested mutant C. elegans males exhibiting defects in specific amphid neurons. In these mutants, either terminal cell fates are altered or neuronal identities are changed in one or more amphid neurons, and in certain cases, atypical morphological features are present, and thus, the amphid‐neuron functions are impaired. The tested mutants were lin‐11, odr‐7, ocr‐2, mps‐1, unc‐3, ceh‐37, ceh‐36, ttx‐1, che‐1, eat‐4, and ttx‐3 (Appendix Table S1). Because ASJ‐ and ADF‐defective mutants were not viable, we used caspase‐1‐based genetic ablation to eliminate ASJs and ADFs in worms (Fig EV1A). Males of all these strains were used in the same chemoattraction assay, and distinct mutants in which the same neuron was affected were then cross‐checked for consistency of results and used to generate a holistic representation of the role played by each amphid neuron.

Figure EV1. Males in which ASJ or ADF neurons are genetically ablated respond normally to sex pheromone.

Human caspase 1 (ICE, or CASP1) was transcriptionally driven by cell‐specific promoters in Ptrx‐1p::ICE (ASJ) and Psrh‐142::ICE (ADF) in pPD95.77 vector. Pvha‐6p::rfp was used as the injection marker and was expressed in the hypodermis and intestine.

1. Ptrx‐1p::gfp is specifically expressed in ASJ neurons in the head region of animals of both sexes (top figures). The transgenic animals displaying the injection‐marker signal show no GFP signal at the ASJ position (bottom figures), which implies that ASJs were ablated. Arrowheads: ASJ neurons. Scale bars = 50 μm.
2. Psrh‐142::gfp is specifically expressed in ADF neurons in the head region of animals of both sexes (top figures). The transgenic animals displaying the injection‐marker signal show no GFP signal at the ADF position (bottom figures), which indicates the removal of ADF neurons. Arrowheads: ADF neurons. Scale bars = 50 μm.
3. Males in which ASJ and ADF neurons were genetically ablated responded normally to sex pheromone in the chemoattraction assays. ADF‐specific rescue of srd‐1 failed to restore wild‐type sex pheromone response in males. We assayed 400 males from each transgenic line and tested three independent transgenic lines. Two biological replicates were combined into a single value. Significance was determined using one‐way ANOVA: ***P < 0.001. Error bars: SEM.

Most of the tested mutant strains responded well to the applied sex pheromone, with their C.I. values being close to that of wild‐type C. elegans males (Fig 2A); these mutants were defective in AWB, AWC, ADF, ADL, AFD, ASE, ASG, ASH, ASI, ASK, and ASJ neurons. The only two exceptions were the lin‐11 and odr‐7 mutant males, which showed markedly attenuated attractive responses to the pheromone (Fig 2A). lin‐11 plays an essential role in AWA cell‐fate differentiation by activating the expression of the odr‐7‐encoded Zn‐finger transcription factor. Subsequently, odr‐7 expression is autoregulated in AWAs, where the respective differentiation program occurs 29, 30, 31. Therefore, AWA cell‐specific differentiation and morphological features are abolished in odr‐7(ky4)X mutants 29, 30, 32. Concurrently, odr‐7 expression in AWAs inhibits AWC‐specific olfactory receptor gene expression in AWA. Thus, our results identified AWA as a critical amphid neuron required for sex pheromone perception, and this was further corroborated by the finding that laser ablation of AWAs in wild‐type males caused a loss of the sex pheromone perception ability in a single‐worm chemoattraction assay (Fig 2B).

Figure 2. AWA chemosensory neurons are responsible for pheromone detection.

1. Inspection of specific amphid‐neuron‐defective mutants revealed that AWA neurons are most likely responsible for pheromone perception. Here, an edge connects neurons to a genetic mutant if the neurons are documented to be affected in the mutant. The color code of edges represents the average C.I. of multiple genetic mutants with impaired neuronal functions or specified abnormal neuron identities. The region colored green shows the range of wild‐type sex pheromone response in the male. We assayed 400 males from each strain in the sex pheromone chemoattraction assay. Two biological replicates were combined into a single value. Significance was determined using one‐way ANOVA with Bonferroni correction: ***P < 0.001. Means ± SEM (error bars) are shown.
2. AWA‐laser‐ablated males were not attracted by the sex pheromone in single‐worm chemoattraction assays. Both AWA‐defective mutant males (lin‐11 and odr‐7) and AWA‐laser‐ablated males failed to reach the sex pheromone within 30 min. Wild‐type Caenorhabditis elegans males responded normally to the sex pheromone, and wild‐type C. elegans hermaphrodites were not attracted by same‐sex pheromone. The single‐worm chemoattraction assay was performed as described in Materials and Methods. Sample size of males in AWA‐laser‐ablation experiment: 12; all other experiments: 100 worms.

AWA role in pheromone perception depends on GPCR pathways

The chemosensory function of AWA neurons is well documented 12, 33, 34, 35. Because numerous volatile odorants are sensed by GPCRs, we hypothesized that the receptors for sex pheromones are also GPCRs. In this scenario, the downstream signal transduction components of GPCRs might be required for the pheromone perception, and elimination of these components would impair or abolish sex pheromone perception ability in males. We selected four mutants, daf‐11, odr‐3, osm‐9, and grk‐2, which harbor defective GPCR pathway components, for examination in our chemoattraction assay (Fig 3A). The four genes encode these molecules: daf‐11, a guanylate cyclase 36, 37; odr‐3, a G‐protein α‐subunit. 38; grk‐2, a G‐protein‐coupled receptor kinase 39, 40; and osm‐9, a Ca²⁺‐permeable transient receptor potential vanilloid (TRPV) channel protein 39, 40 (Appendix Table S1). daf‐11 is required for cGMP‐gated channel function, osm‐9 and grk‐2 affect TRPV channel function, and odr‐3 is required for both functions. Males of all four mutant strains showed strong deficits in the perception of sex pheromones from C. elegans and C. remanei (Fig 3B), which implies that the sex pheromone signal transduction occurs through the cGMP‐gated channel and the TRPV channel. Notably, AWA‐specific rescue of odr‐3 function in mutant males partially restored their sex pheromone response (Fig 3C), which strongly suggests that a GPCR‐type receptor for sex pheromones is likely located in AWA neurons.

Figure 3. GPCR signaling cascade components are required for pheromone perception.

1. If the sex pheromone receptor is a GPCR as hypothesized, it could elicit chemoattraction through the signal transduction pathways that control cGMP‐gated channels and TRPV channels. Knocking out any key component in the two pathways should reduce or abolish chemoattraction behavior.
2. Mutant males of four GPCR signaling cascade components failed to respond to Caenorhabditis remanei sex pheromone, suggesting that the sex pheromone receptor is a GPCR and that both aforementioned signaling pathways are involved in pheromone perception. We assayed 400 males from each strain in the sex pheromone chemoattraction assay. Two biological replicates were combined into a single value. Significance was determined using one‐way ANOVA with Bonferroni correction: ***P < 0.001. Means ± SEM (error bars) are shown.
3. In males from both AWA‐specific‐rescue odr‐3 mutant strain and odr‐3‐rescued mutant strain, the sex pheromone response was significantly restored in single male worm arrival assays. However, males from the CEM‐specific‐rescue odr‐3 mutant strain failed to reach the sex pheromone within 30 min. In odr‐3 and him‐5 strain experiments, the sample sizes were 12 and 20 worms, respectively; in all other experiments, 40 worms were used.

Identification of the GPCR SRD‐1 as the potential sex pheromone chemoreceptor that functions in AWA neurons

We speculated that a GPCR expressed in AWA neurons would likely be used for pheromone perception. Thus, we profiled genes that are specifically expressed in male AWA neurons by using a poly‐A‐binding protein (PABP) mRNA pull‐down approach 41, 42. For this analysis, we ectopically expressed the gene fem‐3 in all neurons of wild‐type hermaphrodites to masculinize their nervous system, including AWA neurons 43, 44. The successful pull‐down of AWA‐specific genes was confirmed by testing for the presence of dpy‐13 (hypodermal gene), myo‐3 (muscle gene), and odr‐10 (gene specifically overexpressed in AWA) in total RNA samples: Whereas odr‐10 was found to be highly enriched in the AWA‐enriched samples, dpy‐13 and myo‐3 were depleted in these samples. The pull‐down experiments were repeated independently four times for each experimental group. Three out of the four datasets were found to be highly consistent, and the mean ratio of each gene from the three consistent replicas was compiled. Comparison between transcriptome profiles revealed critical differences between the whole‐worm and male AWA neurons, which allowed identification of male AWA‐enriched transcripts, including those encoding GPCRs.

From the obtained data, we identified 357 GPCRs that are enriched in male AWAs. We considered these data in combination with the data reported by another group regarding genes that are preferentially expressed in males 45, and by using 2× enrichment as the cutoff, we shortlisted 50 GPCR candidates potentially expressed in male AWAs. The results of transcriptional‐reporter‐expression assays confirmed that 9 of the 50 genes were highly expressed in male AWA neurons (Fig EV2A). Moreover, mutant and RNAi analyses for these GPCR genes revealed that one of these male AWA‐specific GPCRs, SRD‐1, is likely required for pheromone perception (Figs EV2B and 4A): The srd‐1(eh1)II (hereafter “srd‐1”) mutant males lacked the ability to sense the sex pheromone in our chemoattraction assay (Fig 4A) but responded normally to diacetyl and pyrazine; this implies that the mobility and the chemosensory system of the srd‐1 mutants were functional and that the impairment of chemoattraction behavior was solely due to the loss of function of a single GPCR encoded by srd‐1.

Figure EV2. Expression patterns of GPCR candidates and their functional verification.

1. Expression pattern of Psrx‐76::gfp, Pstr‐260::gfp, Psrt‐12::gfp, Pstr‐44::gfp, Psrt‐7::gfp, Psrab‐13::gfp, Pstr‐116::gfp, and Pstr‐164::gfp in adult males: (a–h) GFP images; (a′–h′) RFP images; (a″–h″) merged images. Podr‐7::rfp served as a reporter of AWA neurons (arrowheads). GFP expression was observed in AWA neurons in all these reporter lines. Scale bar = 50 μm.
2. Summary of all GPCR genes examined using transcriptional reporters, where mutant strains or RNAi‐treated animals were analyzed.
3. Psrd‐1::gfp was expressed in ASI neurons in both males and hermaphrodites. Expression pattern of Psrd‐1::gfp: (i–l) GFP images; (i′–l′) RFP images; (i″–l″) merged images. Pdaf‐7::rfp and Pstr‐3::rfp served as reporters of ASI neurons (arrowheads). Scale bar = 50 μm.

Figure 4. Characteristics of SRD‐1 as the receptor required for sex pheromone perception and expression pattern of srd‐1 in Caenorhabditis elegans .

1. srd‐1 mutant males were not attracted to the sex pheromone but were attracted to the positive‐control chemoattractants, 1:1,000 diacetyl and 10 mg/ml pyrazine. This demonstrated that SRD‐1 is required and specific for sex pheromone perception.
2. Summary of srd‐1 expression pattern. The transcriptional reporter Psrd‐1::gfp was detected in both male and hermaphrodite ASI neurons in the head region and only in male AWA and ADF neurons.
3. SRD‐1::GFP fusion construct revealed predominant SRD‐1 localization at the cilia of AWAs in C. elegans males. Scale bar = 5 μm.
4. Examination of srd‐1 transcriptional reporter revealed that srd‐1 is expressed in AWA neurons in males but not hermaphrodites. A transcriptional reporter for odr‐7 indicated the locations of AWAs (arrowheads). Scale bar = 50 μm.
5. Schematic diagram of reprogramming of chemotaxis in C. elegans. Rescue‐experiment results demonstrated the necessity of SRD‐1 in sex pheromone perception; ectopic expression of SRD‐1 in the chemorepulsion‐responsible neuron, AWB, revealed that this SRD‐1 expression is sufficient for altering sex pheromone preference in C. elegans.
6. AWA‐specific rescue of srd‐1 restored sex pheromone perception ability in both males and hermaphrodites.
7. Ectopic expression of srd‐1 in AWB neurons elicited a distinct repulsive behavior toward the sex pheromone in both males and hermaphrodites.

Data information: For each transgenic strain, three independent lines were examined. We assayed 400 males from each strain in the sex pheromone chemoattraction assay. Two biological replicates were combined into a single value. Significance was determined using one‐way ANOVA with Bonferroni correction: ***P < 0.001. Means ± SEM (error bars) are shown.

Sexually dimorphic expression pattern of srd‐1

SRD‐1 is a predicted chemosensory receptor, but its biological function remains undocumented 46. srd‐1::GFP expression is not detected in che‐3 mutants, which harbor defective sensory cilia, or in tax‐2 mutants, in which sensory signal transduction is defective; srd‐1 expression is repressed by developmental entry into the dauer stage 47. Troemel et al 46 also reported that SRD‐1::GFP is expressed in both ADF and ASI neurons in males but only in ASI neurons in hermaphrodites (Fig 4B).

After identifying SRD‐1, we characterized its expression profile by using transcriptional reporters and the translational fusion‐reporter GFP‐tagged SRD‐1, both of which were transcriptionally driven by the gene's own promoter (3‐kb sequence upstream of the predicted translational start site), and SRD‐1::GFP was found to be localized subcellularly at the cilia of male AWA neurons (Fig 4C). The cilia localization of SRD‐1 also agreed with the morphological features of AWA neurons, which reconfirmed the previous finding of its AWA specificity 48. In the head region, the Psrd‐1::gfp signal overlapped with that of the AWA‐specific reporter Podr‐7::rfp (Fig 4D) in all tested males and with that of the ASI reporters Pdaf‐7::rfp and Pstr‐3::rfp (Fig EV2C) in both sexes. No Psrd‐1::gfp signal was detected in any tested hermaphrodite AWA neurons. Therefore, srd‐1 displays a sexually dimorphic expression profile in C. elegans, being highly expressed in ASI, ADF, and AWA neurons in males and only expressed in ASI neurons in hermaphrodites (Fig 4B).

unc‐3, which encodes an Olf‐1/EBF‐family transcription factor, is required for promoting ASI‐specific gene expression, and it represses alternative neuronal programs 49, 50. Because loss of unc‐3 function did not result in defective sex pheromone perception (Fig 2), ASI neurons are unlikely to be necessary for sex pheromone sensing. By contrast, srd‐1 exhibits a sexually dimorphic expression profile in ADFs and thus might be required in sex pheromone perception. Because ADF function was not entirely abolished in the mps‐1 mutant strain, we performed genetic ablation with caspase‐1 to eliminate ADFs. The ADF‐ablated males responded well to the sex pheromone (Fig EV1C), and to further confirm that ADF neurons are not required for sex pheromone chemosensation, we performed ADF‐specific rescue in srd‐1 mutant animals. Restoring srd‐1 only in ADF did not rescue defective sex pheromone chemotaxis (Fig EV1C). Collectively, these results suggest that srd‐1 expressed in ADF neurons is insufficient for restoring sex pheromone perception, and no available evidence indicates that ASI and ADF neurons are required for volatile sex pheromone perception.

AWA‐specific expression of SRD‐1 confers responsiveness to sex pheromone

To ascertain whether the loss of srd‐1 function in AWA alone can adequately explain the failure of srd‐1 mutant males to respond to sex pheromone, we performed AWA‐specific rescue experiments in srd‐1 animals. SRD‐1 expression driven by its endogenous promoter or an AWA‐specific promoter (odr‐7 promoter) in the extrachromosomal array partially rescued the mutant phenotype in chemoattraction assays (Fig 4E and F). In both transgenic strains, the sex pheromone perception ability was restored to a similar level: 62 and 65% responsiveness, using, respectively, srd‐1 endogenous promoter and odr‐7 promoter (Fig 4F). These results and the srd‐1 expression pattern together suggest that SRD‐1 functions in AWA neurons to confer pheromone responsiveness in males.

We next sought to understand the sufficiency of SRD‐1 in conferring the sex pheromone response. AWB neurons are normally hardwired to respond to repulsive volatile odorants, with the response leading to a withdrawal behavior, and a previous study indicated that ectopic expression of an AWA‐specific receptor in AWB neurons alone was adequate for reversing the transgenic animal's behavior toward a chemical attractant from attraction to repulsion 51. Thus, for our assay here, srd‐1 cDNA under the control of an AWB‐specific promotor (from str‐1) was successfully introduced into the srd‐1 mutant strain (Fig 4E and G); when these transgenic males were used in the chemotaxis assay, the animals that expressed SRD‐1 only in AWB neurons displayed a clear repulsive behavior toward the sex pheromone (Fig 4G) at a level similar to wild‐type males responding to a well‐characterized repellant chemical, 2‐nonanone. Therefore, we conclude that mis‐expression of srd‐1 alone is sufficient for altering the sex pheromone preference of C. elegans males.

Interestingly, AWA‐specific expression of srd‐1 in hermaphrodites also evoked a clear attractive behavior and ectopic srd‐1 expression in AWB neurons in hermaphrodites also elicited a distinct repulsive behavior toward the C. remanei female sex pheromone (Fig 4F and G). Therefore, the downstream signal transduction process for the srd‐1 product appeared to be present in both sexes.

Sex pheromone can excite AWA neurons of wild‐type males through srd‐1

The chemoattraction experiments discussed thus far were conducted using single animals or bulk populations to obtain end‐point readouts. The information is obtained after multiple steps in the animals, from perception of the chemical signal to subcellular signal processing and cell–cell communication relay. To visualize the effect of the chemical signal detection and immediate cellular response triggered by SRD‐1, we used the calcium indicator GCaMP5 for real‐time visualization sensory‐neuron excitation in the presence of SRD‐1 by monitoring the influx of calcium ions upon exposure to the chemical signal. We have generated a transgenic worm strain carrying an AWA‐specific promoter (Podr‐7 promoter) that drives the expression of the GCaMP reporter gene. The transgene was crossed into srd‐1 mutant C. elegans (Fig 5A). Here, if AWA neurons are activated upon pheromone stimulation, the fluorescence intensity in AWA would be increased (Fig 5B). Upon exposure to the volatile sex pheromone, the GCaMP signal intensity started to increase in the AWA‐neuron cell body, dendrites, and axon, and the intensity peaked at roughly 30 s after pheromone stimulation and then gradually decreased within the observation window of ~ 3 min.

Figure 5. SRD‐1 is necessary for sex pheromone to activate a calcium response in AWAs.

1. The GCaMP5 reporter was used to visualize activation of the SRD‐1 receptor by the sex pheromone in AWA neurons. To facilitate real‐time imaging of the process, the heads of the transgenic worms were immobilized on soft‐agar pads permeable to sex pheromone.
2. Excitation of the GCaMP5 reporter in AWA neurons could be observed through both lateral and top views, but the top view was clearer and more suitable for quantification. Arrowheads: AWA neurons. Scale bar = 50 μm.
3. Relative fluorescence‐intensity change was quantified through calcium imaging. Wild‐type males showed receptor excitation by sex pheromones from Caenorhabditis elegans hermaphrodites or Caenorhabditis remanei females. This response was lost in srd‐1‐mutant‐strain males but was restored following AWA‐specific rescue of srd‐1. Red triangles: time points of sex pheromone addition. Three males were assayed from each strain.

By comparing the calcium‐imaging data generated from non‐AWA‐specific stimuli (H₂O and M9 buffer) with those from AWA‐specific odorants (diacetyl and pyrazine), we demonstrated the specificity of this assay: AWAs were activated only by the AWA‐specific odorants (Fig EV3A). Moreover, we demonstrated that the sex pheromone but not the male extractant evoked a robust change in AWA fluorescence intensity in wild‐type males (Figs 5C and EV3A), which indicates that this excitation is sex pheromone‐specific. The activation signal, calculated as the mean of the maximal value of the GFP signal‐intensity changes, represents the attractive response toward the sex pheromone. Intriguingly, the C. elegans hermaphrodite sex pheromone also elicited a scintillation excitation of the GCaMP signal, whereas the C. remanei female sex pheromone evoked a more stable GCaMP signal of higher intensity (Figs 5C and EV3A). The chemoattraction‐assay results further showed that the number of C. elegans hermaphrodites necessary for preparing an extract that evokes a prominent male attractive response was 20‐fold higher than the number required in the case of C. remanei virgin females. Therefore, we speculate that the difference in the signal pattern generated by hermaphrodite and female extracts might be due to the distinct concentrations of the active ingredient(s). To verify this, we compared the excitation pattern of male AWA neurons upon exposure to different concentrations of C. remanei female sex pheromone: As the concentration of the female extract was lowered, the GCaMP signal gradually decreased (Fig EV3B), and at 1,000,000‐fold dilution of the extract, the excitation was completely lost (was at the background level; Fig EV3B). These results suggest that the excitation pattern of male AWA neurons depends on the concentration of the sex pheromone active ingredient(s).

Figure EV3. AWA‐specific odorants can excite AWA neurons but this response depends on chemoreceptors.

1. AWAs of wild‐type Caenorhabditis elegans males can be activated by AWA‐specific odorants, diacetyl and pyrazine, and C. elegans or Caenorhabditis remanei female/hermaphrodite extractants, but not M9 buffer or ddH₂O. AWA neurons of srd‐1 mutant males can respond to distinct AWA‐specific odorants. Red triangles: time points of sex pheromone addition.
2. After 1,000‐fold dilution, C. remanei sex pheromone elicits a C. elegans sex pheromone‐like response pattern in calcium imaging, but does not evoke any calcium response after 1,000,000‐fold dilution. Red triangles: time points of sex pheromone addition.
3. Female extract of C. remanei excites AWA neurons in wild‐type C. elegans but not srd‐1 mutant hermaphrodites. The excitation was restored following AWA‐specific expression of srd‐1 in the mutant hermaphrodites. Red triangles: time points of sex pheromone addition.

We next used males and hermaphrodites of wild‐type, srd‐1, and transgenic srd‐1‐rescued strains in the calcium‐imaging reporter assay and recorded their AWA activation by the sex pheromone. AWA neurons were activated by the sex pheromone in wild‐type and srd‐1‐rescued males but not mutated‐strain males (Fig 5C). The responses closely matched the chemoattraction‐assay data implicating SRD‐1 function in AWA and revealed that the presence of SRD‐1 is tightly correlated with the ability of AWA neurons to respond to pheromone stimulation (cellular excitation) and the associated behavioral response. As a control, we showed that in the srd‐1 strains, a response could still be evoked by other AWA‐specific odorants (e.g., diacetyl) in the absence of SRD‐1 (Fig EV3A), which indicates that other cellular chemoreceptor relay functions were intact. Unexpectedly, AWA neurons in wild‐type and srd‐1‐rescued hermaphrodites were excited upon sex pheromone treatment, but the excitation of these neurons was lost in srd‐1 hermaphrodites (Fig EV3C). Because AWA‐neuron excitation is insufficient for evoking an attractive response in hermaphrodites, either additional differences must exist between male and hermaphrodite AWA neurons or additional sex‐specific cellular requirements must govern the translation of neuronal excitation into behavioral responses. Moreover, the ability of this sex pheromone to excite wild‐type hermaphrodite AWA neurons, but not srd‐1‐mutant AWA neurons, indicates that hermaphrodite AWA neurons express SRD‐1 at a very low level (below the detection limit of a transcriptional reporter).

C‐terminal (CT) region of SRD‐1 defines the interspecies variations of sex pheromone perception strength

Androdioecious C. elegans and dioecious C. remanei are closely related within the ELEGANS group of Caenorhabditis. Intriguingly, despite their distinct mating systems and the evolutionary divergence in these androdioecious and dioecious species, the males of both species are attracted to the female/hermaphrodite nematode volatile sex pheromones. We reported that C. elegans males were less attracted to the attractant as compared with males of three closely related Caenorhabditis species 4. We hypothesized that these perception responses could differ due to (i) a change in the expression pattern or level of the receptors, (ii) species‐specific change in the SRD‐1‐coding sequence, or (iii) both.

First, we examined whether native expression levels of endogenous srd‐1 in different species vary, which could account for the interspecies difference in pheromone responses. Quantitative RT–PCR results showed that srd‐1 expression level was lower in C. elegans males than in C. remanei males (Fig EV5B). Unexpectedly, the gene copy numbers/expression levels were not correlated with the sex pheromone perception ability: Higher expression levels of srd‐1 in C. elegans did not correlate with stronger pheromone perception ability (Fig EV5C and D). Therefore, the expression level is not likely to be either a key contributing factor or a limiting factor that can account for the perception strength difference.

Figure EV5. SRD‐1 gene‐expression level does not produce notable differences in sex pheromone chemosensitivity.

1. Ce‐srd‐1 cDNA in which 10 Cr‐srd‐1 polymorphisms were substituted did not elicit higher responsiveness to the sex pheromone as compared with wild‐type Ce‐srd‐1 cDNA. Thus, these 10 polymorphisms do not confer differential sensitivity of the receptor. subECL: srd‐1 transgenic line carrying the Ce‐srd‐1 cDNA in which the 10 indicated Cr‐srd‐1 polymorphisms in the ECL region were substituted.
2. Endogenous expression level of Ce‐srd‐1 in Caenorhabditis elegans is lower than that of Cre‐srd‐1 in Caenorhabditis remanei.
3. Left: Comparison of the expression level of endogenous Ce‐srd‐1 and of srd‐1 in three transgenic lines carrying Ce‐srd‐1 cDNA in high copy number. Right: Comparison of C.I. of wild‐type C. elegans, srd‐1, and three srd‐1 transgenic lines carrying Ce‐srd‐1 cDNA. High expression did not confer increased responsiveness to the sex pheromone relative to that in wild‐type C. elegans, which implies that expression level is not correlated with response strength in the males.

Data information: We assayed 400 males from each wild‐type and mutant strain in the sex pheromone chemoattraction assay. Two biological replicates were combined into a single value. For transgenic‐line studies, each bar represents a total of 1,200 males from four independent transgenic lines: 20 assays were performed for two independent transgenic lines and 10 assays were performed for the other two independent transgenic lines, and the tests were conducted over 2–4 separate days. Expression levels and copy numbers were measured using qPCR and are normalized relative to the housekeeping gene rpl‐32. Significance was determined using one‐way ANOVA: ***P < 0.001. Error bars: SEM.

By analyzing the protein sequence of SRD‐1 from C. elegans, Caenorhabditis briggsae, Caenorhabditis remanei, and Caenorhabditis brenneri, we found that the orthologs share a typical GPCR structure harboring seven transmembrane helices (Fig 6A). Amino acid sequence alignment showed that the most distinct interspecies difference was within the CT region (Figs 6A and EV4). Moreover, 10 amino acid variations were identified in the extracellular loop (ECL) regions of SRD‐1 (Fig EV4A). To test whether the protein sequence variation is responsible for the altered perception ability in C. elegans and C. remanei, srd‐1 cDNAs from these two species were expressed in the C. elegans srd‐1 mutant under the control of C. elegans srd‐1 promoter (Fig 6D). The srd‐1 mutant males rescued by expressing C. remanei srd‐1 cDNA were more responsive to sex pheromones than those rescued by expressing C. elegans srd‐1 cDNA even when the transcripts were detected at a similar level (Fig 6C and D). Thus, the protein‐coding sequence of C. remanei srd‐1 appears to confer comparatively stronger pheromone perception ability. Next, we evaluated whether this disparity was due to the difference in the CT region or polymorphisms in the ECL region. The CT tail of a GPCR is typically involved in the pre‐assembly of signaling complexes and relay of the signal to downstream components, which, in turn, determine the activity of the GPCR. Therefore, we hypothesized that the distinct CT domains of SRD‐1 orthologs could account for the differential responses exhibited by males of different species toward female sex pheromone. To confirm the necessity of the CT region in signal transduction, we attempted rescue of srd‐1 mutants by using C. elegans and C. remanei srd‐1 cDNAs lacking the CT‐coding sequence (Fig 6D). No improvement was detected when any CT‐deleted construct was used for the rescue, which suggests that the CT region is critical for SRD‐1 activities. To test whether the CT region alone could sufficiently account for the difference in the perception activity, the CT sequences of C. elegans and C. remanei SRD‐1 were swapped and the transgenic animals expressing the resulting hybrid cDNAs were used in the chemoattraction rescue assay. Notably, replacement of the C. elegans SRD‐1 CT with the C. remanei CT region resulted in stronger chemoattraction in the transgenic strain, resembling what is observed when C. remanei SRD‐1 is expressed (Fig 6E). The reverse was also true: Replacement of the C. remanei SRD‐1 CT with the C. elegans CT region (Fig 6E) led to a weaker chemoattractive response. Most importantly, the respective percentage increase and decrease were similar in magnitude, mirroring the difference found in assays conducted on animals carrying wild‐type C. elegans and C. remanei SRD‐1 orthologs (Fig 6B).

Figure 6. SRD‐1 C‐terminal (CT) region contributes to species‐specific activities of SRD‐1 homologs in sex pheromone perception.

1. Predicted topological analysis revealing that SRD‐1 from four closely related Caenorhabditis species feature similar structures, which contain seven transmembrane helices and show considerable species‐to‐species variation in the CT region. Sequences of the four SRD‐1 proteins are highly conserved. For the four sequences, the alignment identity score was analyzed by using Benchling with Clustal Omega.
2. Wild‐type Caenorhabditis elegans males are less attracted to sex pheromone than wild‐type Caenorhabditis remanei males.
3. Expression level of srd‐1 cDNA in srd‐1 transgenic lines. srd‐1 mutant males were rescued using, separately, wild‐type C. elegans and wild‐type C. remanei srd‐1 cDNAs. Transgenic lines expressing the molecule at a similar level were selected for use in chemoattraction assays. Four independent transgenic lines were selected, and the expression‐level data were combined into a single bar. The expression levels were measured using qPCR.
4. srd‐1 mutant males rescued with Ce‐srd‐1 cDNA showed lower responsiveness than those rescued with Cre‐srd‐1 cDNA, which indicates that sex pheromone perception ability depends primarily on the protein‐coding region and not the regulatory sequences in the promoter. The sex pheromone perception defect of srd‐1 mutants was not rescued in males expressing CT‐truncated SRD‐1 proteins. The findings show that the CT region is essential for sex pheromone perception.
5. srd‐1 mutant males rescued by using Ce‐srd‐1 cDNA together with Cre‐srd‐1 CT cDNA were more responsive to sex pheromone than those rescued using Cre‐srd‐1 cDNA with Ce‐srd‐1 CT cDNA. The results of the CT‐swapping experiment demonstrated that sex pheromone perception ability is determined by the CT region and not the rest of the SRD‐1 protein.

Data information: We assayed 400 males from each wild‐type strain in the sex pheromone chemoattraction assay. Two biological replicates were combined into a single value. For transgenic‐line studies, each bar represents a total of 1,200 males from four independent transgenic lines: 20 assays were performed for two independent transgenic lines and 10 assays were performed for the other two independent transgenic lines, and the tests were conducted over 2–4 separate days. Significance was determined using one‐way ANOVA with Bonferroni correction: ***P < 0.001. Mean ± SEM (error bars) are shown.

Figure EV4. SRD‐1 protein sequence alignment.

1. Four SRD‐1 sequences were aligned using Clustal Omega (http://www.ebi.ac.uk/Tools/msa/clustalo/). In the extracellular loop (ECL) regions, 10 polymorphisms have been identified, and the CT region shows more diverse variation than the rest of the protein.
2. Schematic diagram of protein sequence alignment performed using the alignment function Benchling (https://benchling.com/).

Lastly, we tested whether the 10 amino acid polymorphisms in the SRD‐1 ECL regions contribute to the differences in SRD‐1 activity. These 10 amino acids in C. elegans SRD‐1 were substituted with corresponding residues from other species through site‐directed mutation in the C. elegans gene. Introduction of the resulting cDNA did not elicit any improvement of responsiveness to the sex pheromone in the rescue experiment, which eliminates the possibility that these polymorphisms contribute substantially to the level of biological activities in C. remanei SRD‐1 (Fig EV5A). Considering these data, we conclude that the variation in sex pheromone perception ability is primarily contributed by the CT region of different SRD‐1 orthologs. The change in male responsiveness to the sex pheromone is probably a result of sequence changes introduced at the CT region when responsiveness to female pheromone no longer imposes a critical selection pressure on survival in hermaphrodites.

Discussion

Other sex pheromone receptors might be present in nematodes

Our study uncovered a critical component in the volatile sex pheromone perception pathway in C. elegans. The results suggest that dimorphic expression of SRD‐1 dictates the sex‐specific response to the same chemical cue. This study identified SRD‐1 as being crucial for sex pheromone perception in male nematodes, but our data do not rule out the involvement of other receptors.

In C. elegans, selective isolation of mRNA is not straightforward; therefore, we adopted the approach of pan‐neural expression of fem‐3 to achieve masculinization of the hermaphrodite nervous system and then used a single‐cell mRNA pull‐down method to obtain male AWA‐enriched mRNA species. Furthermore, we analyzed our data in combination with the microarray data of Jiang et al 45 characterizing differential gene expression between males and hermaphrodites. Our mRNA pull‐down assay and the subsequent microarray experiment were designed based on the expectation that the sex pheromone receptor would stand out among all GPCRs by virtue of its mRNA being highly enriched in masculinized AWA neurons or in the male population. Among all the male AWA‐enriched transcripts, we tested only the top 50 GPCR‐encoding candidates from the ranked list generated from both datasets. Thus, it is likely that we missed other sex pheromone receptors whose transcript enrichment was below the cutoff level (2×); such receptors could potentially act in conjunction with SRD‐1 but are probably less differentially expressed than SRD‐1. The likelihood of other primary sex pheromone receptors being present is high, although these receptors potentially make minor activity contributions because srd‐1 mutation alone can almost entirely eliminate the chemoattraction ability in our assay (C.I. close to zero), and re‐introduction of this single gene in an alternative neuron can elicit a strong alteration of pheromone response, from attraction to repulsion. Therefore, if other receptors exist corresponding to additional active ingredients in the sex pheromone extract, these receptors likely play an augmentative role. Moreover, we are aware of the low resolution at which our chemoattraction assay reveals subtle differences in receptor activity based on the C.I., and thus, additional tracking protocols for real‐time monitoring of individual worm movements might allow superior detection of features evoked by distinct biologically active ingredients if they are present in the pheromone extract.

We have demonstrated that the extracts obtained from sexually mature C. remanei females and C. elegans self‐sperm‐depleted hermaphrodites can attract con‐specific adult males and that extracts prepared from larval stages, adult males, or mated females/hermaphrodites elicit no such behavior 4. Thus, GC‐MS‐profile comparison between these extracts from worms of divergent genetic backgrounds and developmental stages should facilitate the identification of candidate chemicals possessing sex pheromone activities. The availability of these chemical components would provide new reagents for detecting receptor molecules exhibiting distinct biological activities.

Possible sex pheromone signal transduction pathway

As reported in the results section, activation of the volatile sex pheromone signaling pathway involves several components in the cGMP‐gated and TRPV channel signal transduction pathways, including the G‐protein ODR‐3, the G‐protein kinase GRK‐2, the channel‐forming protein OSM‐9, and the guanylate cyclase DAF‐11. In C. elegans, GPCR signal transduction mainly relies on cGMP‐gated and TRPV channels 12. The OSM‐9 and OCR‐2 TRPV channel subunits participate in primary signal transduction in AWA and ASH neurons. ocr‐2 encodes a TRPV ion channel, and its activity is essential for several chemotactic responses 52, 53, 54. Although OSM‐9 is expressed in 10 pairs of head neurons, OSM‐9 and OCR‐2 are co‐expressed only in four pairs of head sensory neurons: AWA, ADL, ADF, and ASH. Mutations of osm‐9 or ocr‐2 produce variable impact, ranging from negligible to severe defects in AWA‐specific odorant responses 52, 54. Thus, the effect of each mutation on sex pheromone response might not be identical. For example, although C. elegans hermaphrodites rely on AWAs to detect diacetyl, ocr‐2(ak47) hermaphrodites retain partial response to diacetyl 32, 54.

Similarly, AWA neurons depend on TRPV function contributed by the activities of osm‐9, ocr‐1, and ocr‐2 54, and ocr‐2 mutant strain exhibits impaired chemotaxis in response to several AWA‐specific odorants; however, AWA cell‐specific signal transduction‐related gene expression and morphological features are not entirely abolished in ocr‐2(ak47)IV mutant. Therefore, although we expected mutations of osm‐9 or ocr‐2 to cause similar chemosensation defects, this does not eliminate the possibility of the osm‐9 mutant but not ocr‐2 mutant showing defective sex pheromone perception, as was observed here. In the absence of a complete characterization of the sex pheromone signal transduction process, our chemoattraction‐assay data suggest only that these two genes play unique biological roles in this pheromone perception context. Further investigation is required to elucidate how this TRPV channel function is executed.

When AWA cell‐specific signal transduction genes and morphological features are eliminated in odr‐7(ky4)X mutant, the mutant animals fail to respond to all odorants detected by AWA 29, 30, 31. Similarly, AWA‐specific genes are not expressed in lin‐11 mutants. These results together with those obtained using AWA‐laser‐ablated males suggest the involvement of AWA among all the amphid neurons in the sex pheromone perception process. Regardless of what the downstream activities are, be they associated with the cGMP‐gated channel or the TRPV channel, investigation should be focused on their interplay in AWA neurons.

Role of CEM in volatile sex pheromone perception

Our search for the sex pheromone receptor began based on the speculation that the receptor would be expressed in CEM, but our results revealed that SRD‐1 functions in the AWA neuron. Considering our previous demonstration that CEM ablation only leads to partial loss of sex pheromone perception ability, one straightforward interpretation of the findings is that CEM provides only an augmenting chemosensory‐input neuron together with other unidentified chemoreceptors. This possibility cannot be eliminated because we searched for sex pheromone receptors only in AWA neurons in this study. CEM could still detect compounds within the sex pheromone mixture that are not ligands of SRD‐1. This speculation is based on observations from other species: Removal of minor components from sex pheromone preparations led to a partial diminution of perception strength, and several active components were detected by distinct receptors 2, 6, 11, 19, 55, 56. Direct testing of the aforementioned possibility would be facilitated by comprehensive identification of active compounds in the volatile nematode sex pheromone and the use of single‐cell mRNA pull‐downs for studying CEM neurons.

A second scenario that is relatively more likely is that CEM is involved in a modulatory role that might support the function of AWA neurons in sex pheromone perception. White et al 57 previously concluded that the functions of CEM, AWA, and AWC in chemoattraction could compensate for one another if one of the neurons is impaired during the L3 developmental stage. Thus, CEM might modulate the function of the sex pheromone chemoreceptor expressed on AWA neurons and help establish the sexual identity of AWA and thereby ensure that SRD‐1 expression and function can be sustained. Alternatively, CEM might act as part of a dual‐sensory input in which the CEM signal is integrated with the input of AWA by an interneuron. Measurement of SRD‐1 expression levels in AWA neurons in CEM‐ablated animals or in a strain harboring feminized CEM neurons would be necessary for testing this possibility.

Sex‐seeking versus food‐seeking

The discovery of SRD‐1 as a receptor required for sex pheromone perception and its function in AWA neurons suggests potential interactions between the molecular mechanisms that govern sex‐seeking and food‐seeking behaviors. The ODR‐10 receptor in AWA is essential for food‐seeking 58: When odr‐10 is knocked out, food‐leaving and exploratory behaviors are enhanced in males; ODR‐10 is weakly and strongly expressed in males when food is abundant and scarce, respectively; and ODR‐10 expression decreases when males reach sexual maturity. Consequently, modulation of the expression level of the food‐seeking receptor promotes exploratory behavior in adult males and facilitates the location of mates in a food‐abundant environment.

Intriguingly, SRD‐1 expression was shown to be strong when worms reached adulthood and weak in larvae, and SRD‐1 expression level was low in dauers or dauer pheromone‐treated animals 47. Therefore, a simple scenario could be one where an apparent negative correlation exists between the expression levels of ODR‐10 and SRD‐1 in sexually mature males and this represents an alternative drive for sex versus food at the reproductively active stage. It is tempting to suggest this control mechanism in the case of worms in which food‐seeking and sex‐seeking behaviors might be prioritized depending on food availability, with mate searching being promoted when the food supply is adequate. This priority might be governed by two separately responsible transcriptional programs, with one dominating over the other at specific times: At larval stages, the food‐searching mode is prioritized, whereas sexually mature males put mate‐searching above food‐searching behavior by showing reduced attraction to the food source and exhibiting a preference for females. Thus, being well‐fed and sexually mature are coupled with the high‐level expression of sex pheromone‐related receptors and a reduced urge for locating food sources.

We suggest that chemosensation is modulated by developmental stages, food supply, and other social conditions, and this could potentially be explained by the existence of genetic interactions governing all necessary transcriptional programs acting within AWA neurons. The documentation of srd‐1 expression in males and the sex pheromone responsiveness of males at various developmental stages and under distinct food supply and social conditions offer a new experimental paradigm for examining signal integration at both single‐animal and single‐cell levels.

In conclusion, we discovered SRD‐1 as a key receptor molecule functioning in AWA neurons in C. elegans for sex pheromone perception. Expression of this receptor in male AWA confers sex pheromone‐induced neuronal excitation and correlates closely with the positive response of the males toward females. This finding provides an entry point for dissecting the molecular mechanism governing the sex‐specific behavior that is manifested through the dimorphic nervous system of C. elegans.

Materials and Methods

Worm strains

N2 CB4088:him‐5(e1490)V was used as wild‐type C. elegans and featured a high incidence of the male population. EM464 was used for C. remanei, CB5161 for C. brenneri, and AF16 for C. briggsae. CB4018 fog‐2(q71), which lacks self‐sperm, was used to produce C. elegans female sex pheromone. In the test of amphid‐neuron‐defective mutants, we used KC459:daf‐6(e1377)X, him‐5(e1490)V; KC678:daf‐10(e1387)IV, him‐5(e1490)V; DR292:che‐3(e1379)I, him‐8(e1489)IV; KC677:osm‐6(p811)V, him‐8(e1489)IV; KC649:osm‐3(p802)IV, him‐5(1490)V; MT633:lin‐11(n389)I, him‐5(e1467)V; KC632:odr‐7(ky4)X, him‐5(e1490)V; KC706:ocr‐2(ak47)IV, him‐5(e1490)V; KC633:ceh‐36(ks86)X, him‐5(e1490)V; KC634:ceh‐37(ok642)X, him‐5(e1490)V; KC652:mps‐1(ok376)II, him‐5(e1490)V; KC631:ttx‐1(p767)V, him‐8(e1489)IV; KC710:che‐1(p674)I, him‐5(e1490)V; KC694:eat‐4(ky5)III, him‐5(e1490)V; and KC534:unc‐3(e151)X, him‐5(e1490)V. To ablate ASJ and ADF neurons, we used KC1455:him‐5(e1490)V, wxEX[Ptrx‐1p::gfp unc‐54 + Pvha‐6p::rfp]; KC1456:him‐5(e1490)V, wxEX[Ptrx‐1p::ICE::unc‐54 3′UTR + Ptrx‐1p::gfp unc‐54 + Pvha‐6p::rfp]; KC1457:him‐5(e1490)V, wxEX[Psrh‐142::gfp unc‐54 + Pvha‐6p::rfp]; and KC1458:him‐5(e1490)V, wxEX[Psrh‐142::ICE::unc‐54 3′UTR + Psrh‐142::gfp unc‐54 + Pvha‐6p::rfp]. The effects of mutations on the GPCR signaling transduction pathway were examined using KC463:daf‐11(m47)V, him‐8(e1489)IV; KC643:odr‐3(n2150)V, him‐8(e1489)IV; KC639:osm‐9(ky10)IV, him‐5(e1490)V; and KC705:grk‐2(rt97)III. For the SRD‐1 mutant strain, we tested KC301:srd‐1(eh1)II, him‐5(e1490)V. For the AWA cell‐specific mRNA pull‐downs, we used KC1099:unc‐119(e2498)III, him‐5(e1490), wxIs63[M142 + Prab‐3::fem3 + Podr‐7::flag‐tag::pab‐1::unc‐54 3′UTR]. To identify the SRD‐1 expression pattern, we used KC1438:srd‐1(eh1)II, him‐5(e1490)V, wxEX[Psrd‐1::srd‐1::gfp unc‐54 3′UTR]; and KC1439:srd‐1(eh1)II, him‐5(e1490)V, wxEX[Psrd‐1::gfp unc‐54 3′UTR + Podr‐10::rfp unc‐54 3′UTR]. The srd‐1 rescue experiment was performed using KC1440:srd‐1(eh1)II, him‐5(e1490)V, wxEX[Psrd‐1::srd‐1::gfp unc‐54 3′UTR + Pgfi::gfp unc‐54 3′UTR]; KC1441:srd‐1(eh1)II, him‐5(e1490)V, wxEX[Podr‐7::srd‐1::gfp unc‐54 3′UTR + Pgfi::gfp unc‐54 3′UTR]; KC1459:srd‐1(eh1)II, him‐5(e1490)V, wxEX[Psrh‐142::srd‐1::gfp unc‐54 3′UTR + Pcrm1b::rfp unc‐54 3′UTR]; and KC1442:srd‐1(eh1)II, him‐5(e1490)V, wxEX[Pstr‐1::srd‐1::gfp unc‐54 3′UTR + Pcrm1b::rfp unc‐54 3′UTR]. For the calcium‐imaging experiment, a calcium indicator was expressed in AWAs of both wild‐type and srd‐1‐mutated animals: KC1443:him‐5(e1490)V, wxEX[Podr‐7::GCaMP5 unc‐54 3′UTR + Pgfi::gfp unc‐54 3′UTR]; and KC1444:srd‐1(eh1)II, him‐5(e1490)V, wxEX[Podr‐7::GCaMP5 unc‐54 3′UTR + Pgfi::gfp unc‐54 3′UTR]. For SRD‐1 CT‐region studies, we used KC1445:srd‐1(eh1)II, him‐5(e1490)V, wxEX[Psrd‐1::srd‐1:gfp unc‐54 3′UTR + Pcrm1b::rfp unc‐54 3′UTR]; KC1446:srd‐1(eh1)II, him‐5(e1490)V, wxEX[Psrd‐1::srd‐1 ▵CT::gfp unc‐54 3′UTR + Pcrm1b::rfp unc‐54 3′UTR]; KC1447:srd‐1(eh1)II, him‐5(e1490)V, wxEX[Psrd‐1::Cre‐srd‐1::gfp unc‐54 3′UTR + Pcrm1b::rfp unc‐54 3′UTR]; KC1448:srd‐1(eh1)II, him‐5(e1490)V, wxEX[Psrd‐1::Cre‐srd‐1▵CT::gfp unc‐54 3′UTR + Pcrm1b::rfp unc‐54 3′UTR]; KC1449:srd‐1(eh1)II, him‐5(e1490)V, wxEX[Psrd‐1::Ce‐srd‐1▵CT::Cre‐srd‐1 CT::gfp unc‐54 3′UTR + Pcrm1b::rfp unc‐54 3′UTR]; KC1450:srd‐1(eh1)II, him‐5(e1490)V, wxEX[Psrd‐1::Cre‐srd‐1▵CT::Ce‐srd‐1 CT::gfp unc‐54 3′UTR + Pcrm1b::rfp unc‐54 3′UTR]; KC1451:srd‐1(eh1)II, him‐5(e1490)V, wxEX[Psrd‐1::Ce‐srd‐1▵CT::Cbr‐srd‐1 CT::gfp unc‐54 3′UTR + Pcrm1b::rfp unc‐54 3′UTR]; and KC1452:srd‐1(eh1)II, him‐5(e1490)V, wxEX[Psrd‐1::Ce‐srd‐1▵CT::Cbre‐srd‐1 CT::gfp unc‐54 3′UTR + Pcrm1b::rfp unc‐54 3′UTR].

Chemoattraction assay and single‐worm arrival assay

To obtain the sex pheromone, synchronized C. remanei females and fog‐2 mutant C. elegans females were separated from the mixed population at the L4 stage, and 24 h later, five virgin adult female C. remanei (EM464) or 100 virgin adult female fog‐2 C. elegans were placed in 100 μl of M9 buffer for 6 h at 25°C. The supernatant was first tested on wild‐type C. elegans or C. remanei males for quality control. The assay was performed on a microscope slide coated with a 2‐ml layer of chemotaxis agar (1.5% agar, 25 mM NaCl, 1.5 mM Tris‐base, and 3.5 mM Tris–Cl); the slide was placed in an empty 5.5‐cm Petri dish. Two 2 μl drops of 1 M sodium azide solution were pipetted 3 cm apart and allowed to dry, after which one 2 μl drop of the supernatant and one drop of the control buffer were separately added on top of the sodium azide spots. Next, 20 synchronized young adult males were placed in the middle of the slide, with the starting point equidistant from the test and control spots (1.5 cm), and 30 min later, paralyzed worms were scored based on their location. The C.I. was calculated by determining the difference between the numbers of worms at the test and control spots and then dividing this by the total number of tested worms; in this assay, C.I. = 1.0 indicates that all worms are attracted to the test spot, whereas a C.I. value near zero implies that the test material is not attractive to the worms. The C.I. values from multiple trials are summarized and shown with the calculated SEM (error bars). All chemoattraction assays were performed on two separate days with controls and freshly prepared sex pheromone samples. The sample size was at least 20 assays (400 animals) for each test. Significance was determined using one‐way ANOVA with Bonferroni correction: ***P < 0.001. The experimental setting for the single‐worm arrival assay was the same as that used in the chemoattraction assay, but only individual worms were examined in each assay and the location of the worms was recorded every 30 s. To test the volatility of the pheromone, we performed two sets of experiments: One, 100 μl of the sex pheromone extract was vaporized at 4°C for 24 h, then added back to 100 μl, and the resulting solutions were tested with wild‐type C. elegans males; and two, the attraction assay was repeated by placing the 2 μl pheromone‐solution drop on a Petri dish cover directly above the sodium azide spot, which prevented direct contact of males with the pheromone solution.

AWA cell‐specific mRNA pull‐down experiment

To establish the male AWA single‐cell transcriptional profile, mRNAs were pooled from AWA neurons ectopically expressing PABP under control of an AWA‐specific promoter (odr‐7 promoter). PABP was fused with FLAG‐tag and was extracted using a FLAG antibody, and the AWA‐enriched transcripts harboring poly‐A tails that were thus pulled down together were subject to microarray analysis. Masculinization plasmid Prab‐3::fem3 and AWA‐specific PABP‐expression plasmid Podr‐7::flag‐tag::pab‐1::unc‐54 3′UTR were injected together with the unc‐119 transformation marker M142ΔXbaI 59, with unc‐119 mutant used as the transformation background. After injection, the stabilized transgenic line was integrated by performing UV treatment. After integration, animals were outcrossed for four rounds to obtain a clean background. The integration‐line‐packed worms were fixed in 1% paraformaldehyde in M9 buffer, and the paraformaldehyde was inactivated with 0.133 V of glycine (125 M, pH 3.5). Animals were lysed using a cell disrupter (at 25 kpsi), and after sedimenting the debris through centrifugation, the RNA bound to FLAG::PAB‐1 was enriched using ANTI‐FLAG® M2 Affinity Gel. The RNA extraction steps were implemented as described in TRIzol manufacturer's manual. Quality assessments of mRNA samples and oligonucleotide microarrays were conducted by Welgene Biotech. Co., Ltd. An Agilent 2100 bioanalyzer was used in combination with the RNA 6000 LabChip® kit to obtain detailed information regarding the condition of the mRNA samples in the form of highly sensitive electropherograms. The mRNA samples were processed using an Agilent Quick Amp Labeling Kit (Agilent, USA) to generate fluorescent complementary RNA (cRNA) for use with Agilent oligonucleotide microarrays. Oligonucleotide microarrays were hybridized using an Agilent C. elegans Oligo 4 × 44 K Microarray (Agilent, USA) in an Agilent microarray hybridization chamber, following the protocol of Agilent Two‐color Microarray‐Based Gene Expression Analysis 60. Genes whose expression differed between the samples could be identified by scanning the microarrays with a laser‐based detection system, Agilent Microarray Scanner. The obtained images were analyzed using the software Agilent Feature Extraction 10.5.1.1. In each independent microarray hybridization, the total cy3 and cy5 signals were normalized to each other by using the method of rank consistency linear LOWESS normalization. The pull‐down RNA sample was labeled with cyanine 5 (excited by 532‐nm laser), and the total RNA sample was labeled with cyanine 3 (excited by 633‐nm laser). For each hybridization, the ratio of the signals from the fem‐3‐masculinized‐animal pull‐down RNA (Cy‐5) to the signal from the total RNA (Cy‐3) was calculated, which represented the fold enrichment of the genes.

ASJ and ADF genetic ablation

To genetically ablate ASJ or ADF neurons, human caspase 1 (ICE or CASP1) transcription was driven using an ASJ‐specific promotor, thioredoxin (trx‐1) promoter (1 kb; a gift from the Joy Alcedo laboratory), or an ADF‐specific promotor, srh‐142 promoter (3 kb). Successful ablation of ASJs or ADFs was demonstrated using an ASJ reporter (Ptrx‐1::gfp) or ADF reporter (Psrh‐142::gfp).

Calcium imaging

Intracellular calcium signals were measured by monitoring Ca²⁺ binding to a calcium indicator, GCaMP5; the binding increases green‐fluorescence intensity in cells. A disposable worm‐mounting and chemical‐loading device, the thin‐layer micro‐diffusion (TLMD) unit, was used for detecting neuronal excitation by measuring the changes in fluorescence intensity. A 4% agarose pad (0.2 mm in thickness and 1.3 mm in radius) was loaded between two autoclave tapes. Well‐fed transgenic worms were transferred to an empty worm plate for 5 min to clean off the attached bacterial residues before mounting, and 0.2 μl of 1 mM levamisole was applied onto the agarose‐gel pad; the worms transferred to the gel pads were partially immobilized by levamisole. A clean 22 × 40 mm coverslip was laid lightly over the TLMD unit, and 8 μl of the test chemical solutions was loaded at the 30‐s time point. The loaded solution was rapidly sucked into the space between the slide and coverslip through capillary action, and the diffusion of chemicals started when the micro‐gel pad was surrounded by the loaded solution.

Time‐lapse imaging parameters and post‐experiment video conversion

For neural‐excitation assays, we used a spinning‐disk microscopy system, which comprised a Carl Zeiss Axio Observer Z1 Inverted Microscope equipped with a Piezo Z stage insert for glass slides, a spinning‐disk unit Yokogawa CSU‐X1, and an Andor iXon897BV EMCCD camera. A Nikon 60× oil‐immersion objective lens was used. Time‐lapse images were acquired at 1 frame/s for 3 min. The test samples were added manually at ~ 30 s after filming was started. Data analysis and video stacking were performed in Imaris image‐analysis program. AWA‐neuron cell‐body regions were selected and analyzed. The normalized GFP intensity was calculated by normalizing the recorded average intensity at each time point against the average intensity at t = 0. The relative intensity values are presented here.

Chimeric SRD‐1 construction

For expression of C. elegans srd‐1 cDNA, C. remanei srd‐1 cDNA, and the cDNAs encoding their CT‐region‐truncated forms in srd‐1 mutant strain, we constructed the plasmids Psrd‐1::srd‐1::gfp unc‐54 3′UTR unc‐54 3′UTR, Psrd‐1::cre‐srd‐1::gfp unc‐54 3′UTR, Psrd‐1::srd‐1▵CT::gfp unc‐54 3′UTR, and Psrd‐1:: Ce‐srd‐1▵CT::gfp unc‐54 3′UTR; in both sets of plasmids, srd‐1 was under the control of the C. elegans srd‐1 promoter and fused to the unc‐54 3′UTR. Caenorhabditis elegans srd‐1 cDNA, C. remanei srd‐1 cDNA, and the cDNAs for their CT‐truncated forms were PCR‐amplified using cDNAs from C. elegans N2 strain and C. remanei EM464, and the C. elegans srd‐1 promoter was PCR‐amplified from genomic DNA extracted from C. elegans N2 strain and then cloned into pPD95.75 by Gibson assembly. The CT‐swapped srd‐1 cDNA plasmids, Psrd‐1:Ce‐srd‐1▵CT::Cre‐srd‐1 CT::gfp unc‐54 3′UTR and Psrd‐1::Cre‐srd‐1▵CT::Ce‐srd‐1 CT::gfp unc‐54 3′UTR, were constructed through PCR by using Psrd‐1::srd‐1::gfp unc‐54 3′UTR unc‐54 3′UTR and Psrd‐1::cre‐srd‐1::gfp unc‐54 3′UTR plasmids as template and fused to the CT region of their orthologs by Gibson assembly. All constructions were confirmed by sequencing.

Author contributions

KLC supervised the entire project; XW, CMC, and CY contributed to the amphid‐neuron and GPCR signal cascade studies; YZ conducted the microarray and expression‐pattern analysis; YZ, XW, and HNY performed the SRD‐1 identification and characterization; XW and HNY conducted the GCaMP5 experiments; XW performed the CT study; and XW contributed to the analysis of the results and wrote the manuscript in consultation with KLC.

Conflict of interest

The authors declare that they have no conflict of interest.

Supporting information

Appendix

Expanded View Figures PDF

Review Process File

EMBO Reports (2019) 20: e46288

Data availability

The microarray data from this publication have been deposited to the GEO database (https://www.ncbi.nlm.nih.gov/geo) and assigned the identifier (accession number: GSE112610).