Developmental history modulates adult olfactory behavioral preferences via regulation of chemoreceptor expression in Caenorhabditiselegans

Abstract

Developmental experiences play critical roles in shaping adult physiology and behavior. We and others previously showed that adult Caenorhabditiselegans which transiently experienced dauer arrest during development (postdauer) exhibit distinct gene expression profiles as compared to control adults which bypassed the dauer stage. In particular, the expression patterns of subsets of chemoreceptor genes are markedly altered in postdauer adults. Whether altered chemoreceptor levels drive behavioral plasticity in postdauer adults is unknown. Here, we show that postdauer adults exhibit enhanced attraction to a panel of food-related attractive volatile odorants including the bacterially produced chemical diacetyl. Diacetyl-evoked responses in the AWA olfactory neuron pair are increased in both dauer larvae and postdauer adults, and we find that these increased responses are correlated with upregulation of the diacetyl receptor ODR-10 in AWA likely via both transcriptional and posttranscriptional mechanisms. We show that transcriptional upregulation of odr-10 expression in dauer larvae is in part mediated by the DAF-16 FOXO transcription factor. Via transcriptional profiling of sorted populations of AWA neurons from control and postdauer animals, we further show that the expression of a subset of additional chemoreceptor genes in AWA is regulated similarly to odr-10 in postdauer animals. Our results suggest that developmental experiences may be encoded at the level of olfactory receptor regulation, and provide a simple mechanism by which C. elegans is able to precisely modulate its behavioral preferences as a function of its current and past experiences.

Introduction

Conditions experienced during early development have profound effects on adult phenotypes. Fetal malnutrition is a major risk factor for metabolic disorders in human adults, and adverse experiences in early life influence adult stress responses in many animal species (de Gusmao Correia et al. 2012; Hanson and Gluckman 2014; Smith and Ryckman 2015; Alyamani and Murgatroyd 2018; Cater and Majdic 2021). In addition to modulating general life history traits, early experiences can also affect specific adult behavioral phenotypes. For instance, exposure to an odorant during a critical developmental period has been shown to subsequently modulate the responses of adults to that odorant (Nevitt et al. 1994; Remy and Hobert 2005; Jin et al. 2016; Hong et al. 2017). Adult phenotypic plasticity as a consequence of differential developmental experiences may allow adaptation to variable environments to optimize fitness (Sommer 2020). Despite the prevalence and critical role of early experiences in shaping adult phenotypes, the underlying mechanisms are not fully understood.

Caenorhabditis elegans adults develop via one of 2 alternative developmental trajectories based on environmental conditions experienced during their first and second larval stages (Cassada and Russell 1975). While larvae continue in the reproductive cycle through 4 larval stages (L1–L4) under favorable conditions, adverse conditions experienced during L1 instead drive larvae into the dauer diapause stage. Dauer larvae undergo extensive morphological, neuroanatomical, and behavioral remodeling that maximizes their ability to survive harsh conditions, distinguishing them from their L3 larval counterparts (Popham and Webster 1979; Albert and Riddle 1983; Riddle 1988; Britz et al. 2021). When growth conditions improve, they exit the dauer stage and resume development into reproductive adults. We and others previously showed that adult C. elegans that transiently passed through the dauer stage [henceforth referred to as postdauer (PD) adults] differ markedly from adult animals that bypassed the dauer stage (referred to as control adults) in their life history traits including longevity, stress resistance, and fecundity (Hall et al. 2010, 2013; Ow et al. 2018, 2021). This phenotypic plasticity is correlated with extensive changes in transcriptional profiles, as well as modification of the chromatin landscape (Hall et al. 2010, 2013; Ow et al. 2018; Vidal et al. 2018; Bhattacharya et al. 2019). Thus, isogenic populations of adult C. elegans hermaphrodites retain a molecular and phenotypic memory of their developmental history.

Food seeking is a critical behavioral drive and is subject to extensive modulation as a function of an animal’s internal state and external conditions (Pool and Scott 2014; Heisler and Lam 2017; Kim et al. 2017; Flavell et al. 2020). Dauer larvae do not feed and the morphologies of their sensory neurons are extensively remodeled (Popham and Webster 1979; Albert and Riddle 1983; Riddle 1988; Britz et al. 2021). Dauer larvae retain the ability to detect and respond to environmental cues including food-related chemical cues to assess whether conditions have improved sufficiently to trigger exit from the dauer stage and reentry into the reproductive cycle. Chemosensory responses of C. elegans dauer larvae have been assessed in a limited set of studies and appear to be partly distinct from those of adults or L3 larvae (Albert and Riddle 1983; Hallem et al. 2011; White et al. 2019; Vertiz et al. 2021). Whether PD adults retain these behavioral differences or exhibit further plasticity in food-seeking behaviors as a consequence of their distinctive developmental trajectory is largely unknown. The expression of many chemoreceptor genes in sensory neurons is altered in dauer larvae (Peckol et al. 2001; Nolan et al. 2002; Vidal et al. 2018), and the spatial patterns and/or levels of a subset of these genes are subsequently maintained or further altered in PD adults (Peckol et al. 2001; Hall et al. 2010; Vidal et al. 2018). Altered chemoreceptor expression profiles provide a plausible mechanism for chemosensory behavioral plasticity, but a correlation between changes in chemoreceptor gene expression and altered food-seeking behaviors in PD animals remains to be established.

Here, we show that PD adults exhibit increased sensitivity to a panel of bacterially-produced attractive odorants including the chemical diacetyl, low concentrations of which are sensed by the AWA olfactory neuron pair. Consistently, both dauer larvae and PD adults also exhibit increased diacetyl-evoked responses in the AWA olfactory neurons. We find that the AWA-expressed odr-10 diacetyl receptor gene is upregulated in dauer larvae; this upregulation is mediated in part via the DAF-16 FOXO transcription factor. While odr-10 expression is subsequently downregulated in PD adults relative to levels in dauer larvae, levels of the ODR-10 receptor protein in the AWA cilium, the site of primary odorant transduction, are retained at higher levels in PD as compared to control adults. Via transcriptional profiling of sorted populations of AWA neurons from control and PD animals together with examination of expression from endogenous reporter-tagged alleles, we further find that the expression of a subset of additional AWA-expressed chemoreceptor genes is also upregulated in dauers and is maintained at higher levels in PD adults. Together, our data demonstrate that altered chemoreceptor levels can underlie developmental stage- and history-dependent olfactory behavioral plasticity in C. elegans and highlight the complexity of mechanisms regulating expression of individual chemoreceptor expression genes in this organism.

Materials and methods

Strains and growth conditions

All C. elegans strains were maintained on nematode growth medium (NGM) seeded with Escherichia coli OP50 at 20°C unless stated otherwise. Dauer larvae were generated by picking 8–10 L4s from continuously growing animal populations onto 10-cm NGM plates and allowing growth and reproduction until food exhaustion (typically 7–9 days) at 25°C. Starved larvae were washed off plates with S-Basal buffer, and dauer larvae were selected by treatment with 1% SDS for 30 min. To recover PD adults, a droplet containing 100–700 dauer larvae was pipetted onto a 10-cm NGM plate seeded with OP50 and allowed to grow for 48 h at 20°C. L4 larvae were picked onto fresh, seeded NGM plates 1 day prior to imaging.

For all experiments involving growth on plates containing auxin, NGM media was supplemented with the synthetic auxin 1 mM 1-naphthaleneacetic acid from a 500 mM stock solution dissolved in 95% ethanol. Corresponding control NGM media was supplemented with an equal volume of 95% ethanol.

To collect adult animals that underwent L1 larval arrest, gravid adult hermaphrodites were bleached, and eggs were allowed to hatch overnight in M9 buffer containing 0.1% Triton X-100. A total of 500–1,000 L1 larvae were subsequently placed onto 10-cm NGM plates seeded with OP50 and grown at 20°C until adulthood.

Genetics

All strains were constructed using standard genetic methods. Crosses were validated for the presence of the desired mutations using PCR-based amplification and/or sequencing. To generate strains for auxin-induced degradation specifically in AWA, animals were injected with a plasmid driving TIR1 under the AWA-specific gpa-4Δ6 promoter (PSAB1283: gpa-4Δ6p::TIR1::SL2::mScarlet; Supplementary Table 1) at 5 ng/μl together with the unc-122p::dsRed coinjection marker at 50 ng/μl. A complete list of strains used in this work is available in Supplementary Table 2.

Molecular biology

Gene editing was performed using CRISPR/Cas9 and repair templates. Plasmids PSAB1279 (odr-10::splitGFP₁₁) and PSAB1280 (gpa-4Δ6p::splitGFP_1-10) (Supplementary Table 1) were generated using traditional cloning from splitGFP constructs (KP#3315 and KP#3316, gift from J. Kaplan lab). An asymmetric donor template (Dokshin et al. 2018) was amplified from plasmid PSAB1279 to create long and short PCR fragments of 2,308 and 600 bp, respectively, containing approximately 1 kb homology arms for CRISPR-mediated insertion. The asymmetric hybrid template was injected (each fragment at 250 ng/μl) together with crRNA (20 ng/μl; IDT Integrated DNA Technologies), tracrRNA (20 ng/μl: IDT), Cas9 protein (25 ng/μl; IDT), and unc-122p::gfp (40 ng/μl) as the coinjection marker. F1 animals expressing the injection marker were isolated. F2 progeny were screened by PCR for the insertion and confirmed by sequencing to obtain odr-10(oy158) (Supplementary Table 2).

To mutate the predicted DAF-16 binding site upstream of odr-10 in odr-10(oy158), a conserved GTAAACA binding site 815 bp 5′ of the odr-10 start codon (Wexler et al. 2020) was mutated to GTCCCCA to generate odr-10(oy170). The injection mix contained an odr-10 promoter repair template with the mutated site along with 32 bp 5′ and 3′ homology arms (590 ng/µl; IDT), dpy-10 repair template (100 ng/µl, IDT), Cas9 protein (25 ng/µl; IDT), crRNA (20 ng/μl; IDT), and tracrRNA mix (20 ng/μl: IDT). F1 animals with Dpy phenotypes were isolated and F1 and F2 progeny were screened by PCR and sequencing. Confirmed mutants were backcrossed to remove the dpy-10 allele.

Chemotaxis behavioral assays

Chemotaxis assays were performed essentially as described previously (Bargmann et al. 1993; Troemel et al. 1997). Behavioral attraction and avoidance assays were performed on 10-cm round or square plates, respectively. Each assay was performed in duplicate each day, and data are reported from biologically independent assays performed over at least 3 days. Behaviors of control and experimental animals were examined in parallel each day.

Calcium imaging

Calcium imaging was performed essentially as previously described, using custom microfluidics devices (Chronis et al. 2007; Neal et al. 2015; Khan et al. 2022). Imaging was performed on an Olympus BX52WI microscope with a 40X oil objective and Hamamatsu Orca CCD camera. Video recordings were performed at 4 Hz. All odorants were diluted in filtered S-Basal buffer. 20 μM fluorescein was added to 1 buffer channel to confirm correct fluid flow in microfluidics devices. 1 mM (−)-tetramisole hydrochloride (Sigma L9756) was used to immobilize animals during imaging. To prevent animals from clogging the microfluidics loading arena and chip, 1 μl of poloxamer surfactant (Sigma P5556) was added to the S-Basal loading buffer. AWA neurons were imaged for 1 cycle of 30 s buffer/30 s odor/30 s buffer, or for 1 cycle of 30 s buffer/10 s odor/20 s buffer stimuli.

Recorded images were aligned with the template Matching plugin in Fiji (NIH) and cropped to include the AWA neuron soma and surrounding background fluorescence. The region of interest (ROI) was defined by outlining the AWA cell bodies, and an area of background fluorescence was chosen for background subtraction. To correct for photobleaching, an exponential decay was fit to the fluorescence intensity values for the first 30 s and the last 20 s of imaging. The resulting value was subtracted from original intensity values. Peak amplitude was calculated as the maximum change in fluorescence (F − F₀) in the 10 s following odor addition; F₀ was set to the average ΔF/F₀ value for 5 s before odor onset. Data visualization was performed using RStudio (Version 1.4.1717). Photomask designs for customized adult and dauer microfluidic imaging chips were adapted from (Chronis et al. 2007) and are available at https://github.com/SenguptaLab/PDplasticity (also see below). AWA neuron mean baseline fluorescence (Supplementary Fig. 1c) was calculated by taking the average ΔF/F₀ during the first 25 s of imaging (0–25 s) in each animal. Reported data were collected from biologically independent experiments over at least 2 days.

Dauer microfluidics imaging device

Designs were based on olfactory imaging chips previously described (Chronis et al. 2007) with modifications to accommodate dauer larvae. To constrict the thinner dauer larvae in a similar manner to adults, it was necessary to reduce the cross-sectional area of the worm trap. This was accomplished by placing 10-μm-wide posts of decreasing width culminating in a 5-μm gap into which the worm nose was constricted (https://github.com/SenguptaLab/PDplasticity). Posts were used instead of fully narrowing the channel to preclude limiting fluid velocity while loading animals into the channel. AutoCAD drawings were used to generate an ink photomask (outputcity.com) which was subsequently used to generate master molds with 10 μm feature depth using negative photoresist SU-8 3005 (Supplementary Files 3 and 4). To prevent delamination due to small features, a uniform 10-μm layer of SU-8 3005 photoresist was applied first to silicon wafers.

Imaging and image analysis

Animals were mounted on 10% agarose pads set on microscope slides and immobilized using 10 mM (−)-tetramisole hydrochloride. Imaging for expression level quantification was performed on an inverted spinning disk confocal microscope (Zeiss Axiovert with a Yokogawa CSU22 spinning disk confocal head and a Photometerics Quantum SC 512 camera). Optical sections were acquired at 0.27 μM sections using either a 63× oil immersion objective when imaging neuron cell bodies, or a 100× oil immersion objective when imaging the cilia using SlideBook 6.0 software (Intelligent Imaging Innovations, 3i). z-projections of all optical sections at maximum intensity were generated using SlideBook 6.0 or FIJI/ImageJ (NIH).

Quantification of GFP or mNeonGreen levels in AWA soma was performed by tracing the ROI of each neuron cell body, and subtracting background fluorescence. The mean fluorescence of each cell body was calculated in Fiji. For the quantification of ciliary ODR-10 levels in adult animals, an ROI of the periciliary membrane compartment (PCMC) and primary stalk was obtained and the integrated density of total fluorescence in the ROI was calculated using the corrected total cell fluorescence to take into account differences in the sizes of different ROIs and the highly branched structures of AWA cilia. Since the AWA cilia branches are collapsed in dauer larvae and are relatively smaller in L3 larvae, the ROI was traced around the entire PCMC and cilia branches. Three background ROIs were obtained and averaged to calculate the mean background fluorescence.

Collection of AWA neurons

To obtain control animals, adult hermaphrodites expressing the stably integrated transgene gpa-4Δ6p::myrGFP (PY10421, Supplementary Table 2) were bleached, and eggs allowed to hatch in M9 with 0.1% Triton-100. 50,000–100,000 growth-synchronized L1 larvae were plated onto 15-cm 8P growth plates seeded with 1 ml of an overnight culture of E. coli NA22 grown in 2XYT media. L1 larvae were allowed to grow for 40–48 h at 20°C to obtain populations of L4 larvae. To obtain PD animals, SDS-selected dauer larvae were plated onto 15 cm 8P plates seeded with 1 ml of an overnight culture of E. coli NA22 grown in 2XYT media, and allowed to recover to the L4 stage for 20–24 h at 20°C.

Animals were collected and cells dissociated essentially as described previously (Taylor et al. 2021). In brief, animals were washed off plates with M9 buffer and incubated in lysis buffer [200 mM dithiothreitol, 0.25% sodium dodecyl sulfate (SDS), 20 mM HEPES buffer (4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid), 3% sucrose] for 5 min. Lysed animals were washed 5× with egg buffer (118 mM NaCl, 48 mM KCl, 2 mM CaCl₂, 2 mM MgCl₂, 25 mM HEPES, pH 7.3) following which the worm pellet was resuspended in 750 μl of 15 mg/ml Pronase (Sigma P8811) in egg buffer. To promote dissociation, the mixture was pipetted frequently for 20 mins and progress of dissociation was monitored under a light microscope. Pronase digestion was terminated by adding L-15-10 media (L-15 media + 10% volume FBS, HI) at 4°C. Dissociated worms were centrifuged at 4°C and the worm pellet was resuspended in egg buffer and recentrifuged at 4°C. The supernatant was filtered through a 35-µm filter in preparation for FACS sorting. A 25-μl aliquot of the dissociated worm pellet was pipetted directly into TRIzol for analysis as the whole worm sample.

Prior to performing FACS, 0.5 μl of 1 mg/ml DAPI was added to 1 ml of the cell suspension to exclude dead cells. AWA neurons were captured by isolating the high GFP+ and low DAPI compartment and cells were sorted directly into TRIzol. Cells from this compartment were also sorted onto glass slides and imaged under a fluorescence light microscope to confirm the presence of intact GFP+ cells. Following cell collection, collection tubes were briefly spun down and frozen at −80°C. 1,000–10,000 cells were isolated per sorting run.

RNA sequencing

RNA was extracted from collection tubes using standard chloroform and isopropanol precipitation. The RNA pellet was resuspended in RNAase-free water and any DNA digested using the RNase-Free DNase kit (Qiagen). RNA was eluted using the RNeasy MinElute Cleanup Kit (Qiagen). mRNA was isolated and libraries amplified using Smart-Seq v4 and library quality checked with Bioanalyzer using Agilent High Sensitivity Gel. Libraries were sequenced on Illumina NextSeq500. Library preparation and sequencing was performed at the MIT BioMicro Center (https://biology.mit.edu/tile/biomicro-center/).

Raw RNA-Seq reads were quality checked, adapter trimmed, and aligned to the C. elegans genome using STAR ALIGNER (https://github.com/alexdobin/STAR). Principal component analyses were performed on the normalized read counts of the 10,000 most variable genes across all samples. Differential expression analysis was performed using DESeq2 with lfc shrinkage on. Differential expression was determined with a significance cut-off of 0.05 (adjusted P-value; Wald test with Benjamini–Hochberg corrections) and the indicated log₂ fold change cut-off of either −1.5 and 1.5 or −2 and 2. AWA enrichment analysis was performed using the web-based Tissue Enrichment Analysis tool with default parameters (https://www.wormbase.org/tools/enrichment/tea/tea.cgi) (Angeles-Albores et al. 2016).

Statistical analyses

Normal distribution of the data was assessed using the Shapiro–Wilks test (GraphPad Prism v9.0.2). Normally distributed data were analyzed using a 2-tailed Welch’s t-test or ANOVA. Nonnormally distributed data were analyzed using a Mann–Whitney t-test or Kruskal–Wallis test across multiple conditions. Post hoc corrections for multiple comparisons were applied to data in which more than 2 groups were analyzed. The specific tests used and corrections applied are indicated in each figure legend.

Results

PD adult animals exhibit enhanced attraction to a subset of volatile odorants

C. elegans adult animals are strongly attracted to a subset of chemicals that indicates the presence of nutritious bacteria, the major food source for these nematodes (Ferkey et al. 2021). Attractive volatile odorants are primarily sensed by the AWA and AWC olfactory neurons pairs in the bilateral head amphid organs of C. elegans (Bargmann et al. 1993). Consistently, adult hermaphrodite animals grown continuously under favorable environmental conditions (control adults; Fig. 1a) exhibited robust attraction to a range of concentrations of odorants sensed by the AWA and/or AWC neurons (Fig. 1b).

Fig. 1.

Olfactory responses to a panel of volatile attractants is increased in PD adults. a) Cartoon of growth conditions for the generation of control and PD adults. b) Behavioral responses of wild-type control and PD adults to a panel of volatile odorants at the indicated dilutions. Chemotaxis index = (number of animals at the odorant − number of animals at the diluent ethanol)/(number of animals at the odorant + number of animals at ethanol). Each dot is the chemotaxis index of a single assay plate containing ∼50–300 adult hermaphrodites. Bars represent the mean; error bars are SEM. The behaviors of control and PD animals were assayed in parallel in duplicate; ≥3 independent experiments. *, **, and *** indicate different at P < 0.05, 0.01, and 0.001, respectively (2-tailed Welch’s t-test). c) Average changes in GCaMP2.2b fluorescence in AWA soma to a 30-s pulse of 10⁻⁵ diacetyl in control and PD adults. Shaded regions indicate SEM. n ≥ 16 animals (1 neuron per animal) each. d) Quantification of peak fluorescence intensity changes in AWA soma expressing GCaMP2.2b to a 30-s pulse of diacetyl at the indicated concentrations. Bars represent mean, error bars are SEM. n ≥ 16 animals (1 neuron per animal) each. Control and PD adults were examined in parallel over at least 3 days. ** and *** indicate different at each concentration at P < 0.01 and 0.001, respectively (2-tailed Welch’s t-test); ns, not significant.

To test whether passage through the dauer stage influences the olfactory preference behaviors of adult animals, we grew wild-type animals under food-restricted conditions to promote entry into the dauer stage as previously described (Fig. 1a) (Ow et al. 2018). Dauer larvae were collected via SDS-mediated selection and subsequently allowed to resume reproductive growth on bacterial food (Fig. 1a). Similar to control adults, young PD adult animals were also robustly attracted to both AWA- and AWC-sensed volatile odorants (Fig. 1b). We found that PD adults typically exhibited enhanced attraction to lower odorant concentrations relative to control adults (Fig. 1b). Animals that experienced starvation-induced developmental arrest at the L1 larval stage did not exhibit similar enhanced attraction responses as adults (Supplementary Fig. 1a). PD animals have previously been shown to exhibit decreased avoidance of the aversive pheromone ascr#3 (Sims et al. 2016). Control and PD animals avoided high concentrations of benzaldehyde to a similar extent (Supplementary Fig. 1b) although the aversion responses of PD adults to high concentrations of octanol were decreased (Supplementary Fig. 1b). We conclude that the behavioral responses of adult animals to multiple attractive volatile chemicals are sensitized upon passage through the dauer diapause stage.

We next tested whether increased behavioral attraction correlates with enhanced odorant responses in olfactory neurons by examining odorant-evoked changes in intraneuronal calcium dynamics using a genetically encoded calcium indicator. We focused our attention on the odorant diacetyl, low concentrations of which are sensed by the AWA olfactory neuron pair (Bargmann et al. 1993), and for which the cognate receptor and signal transduction mechanisms have been described (see below) (Sengupta et al. 1996; Colbert et al. 1997; Roayaie et al. 1998). Low concentrations of diacetyl consistently evoked responses of larger amplitude in the AWA neurons of PD as compared to control adults (Fig. 1, c and d) consistent with the observed increased behavioral responses to this chemical. This amplitude difference is unlikely to arise simply due to differences in the expression levels of the AWA-expressed calcium sensor (Supplementary Fig. 1c). Together, these data indicate that neuronal responses of adult animals to the attractant diacetyl are enhanced upon passage through the dauer stage.

Ciliary levels of the diacetyl receptor ODR-10 are increased as a function of developmental history

We and others previously reported that control and PD animals exhibit significant differences in gene expression profiles including in individual sensory neurons (Hall et al. 2010; Sims et al. 2016; Vidal et al. 2018; Bhattacharya et al. 2019). Each chemosensory neuron type in C. elegans expresses multiple G protein-coupled receptors (GPCRs), expression of a subset of which has previously been shown to be regulated by external and internal state (Troemel et al. 1995; Peckol et al. 2001; Lanjuin and Sengupta 2002; Nolan et al. 2002; Gruner et al. 2014; Ryan et al. 2014; Vidal et al. 2018). Low concentrations of diacetyl are sensed by the ODR-10 olfactory receptor which is expressed specifically in the AWA neurons and localizes to their sensory cilia, the primary site of olfactory signal transduction (Sengupta et al. 1996). Behavioral and neuronal responses to diacetyl have previously been shown to be regulated by feeding state, developmental stage and somatic sex via expression changes in odr-10 (Ryan et al. 2014; Wexler et al. 2020), but whether odr-10 expression is also modulated via dauer passage is unclear. Since state-dependent regulation of individual olfactory receptor genes provides a simple mechanism for mediating odorant-selective behavioral plasticity, we tested the hypothesis that altered expression of odr-10 underlies the observed plasticity in diacetyl responses in PD adults.

An endogenous odr-10 allele tagged with t2A::mNeonGreen (odr-10::t2A::mNG) was specifically expressed in both AWA neurons (Fig. 2a) (McLachlan et al. 2022). Since the t2A peptide is self-cleaving, mNeonGreen levels allow assessment of transcriptional and translational, but not posttranslational, regulation of odr-10 (Ahier and Jarriault 2014; Stewart-Ornstein and Lahav 2016). However, expression levels of the fluorescent reporter protein were not significantly altered in PD as compared to control adults (Fig. 2b). Since this reporter does not allow assessment of ciliary ODR-10 protein levels, we next quantified ODR-10 protein levels in AWA cilia from the endogenous odr-10 allele tagged with the split-GFP reporter GFP₁₁ [odr-10(oy158)], and reconstitution of GFP via expression of the GFP_1-10 fragment under a constitutive AWA-specific promoter (Kamiyama et al. 2016). We confirmed that tagging odr-10 with GFP₁₁ had no effect on behavioral responses to diacetyl, and that these animals continued to exhibit enhanced attraction to diacetyl upon passage through the dauer stage (Supplementary Fig. 2a). The reconstituted ODR-10 fusion protein was localized to the extensively branched AWA cilia in both control and PD animals (Fig. 2c). Since the complex architecture of the AWA cilium precluded precise quantification of overall ciliary protein levels in adult animals, we restricted our analysis to assessing ODR-10 protein levels at the ciliary base (henceforth referred to as the periciliary membrane compartment or PCMC) and in the primary ciliary stalk. We found that reconstituted ODR-10::GFP levels in AWA cilia were significantly higher in PD than in control animals (Fig. 2, c and d). These observations raise the possibility that increased ciliary levels of ODR-10 protein in PD animals may contribute to the increased diacetyl responses of PD animals.

Fig. 2.

Ciliary ODR-10 levels are increased in PD adults. a) Representative images of AWA neurons expressing odr-10::t2A::mNeonGreen from the endogenous odr-10 locus in control and PD adults. Arrows indicate AWA cell bodies. Anterior is at left. Scale bars: 5 µm. b) Quantification of ODR-10::t2A::mNeonGreen fluorescence in AWA soma in control and PD adults. Each dot is a measurement from an individual neuron. Bars represent the mean; error bars are SEM. n ≥ 39 neurons (≥21 animals). Control and PD adults were imaged in parallel on the same day; ≥3 independent experiments. ns, not significant. c) Representative images of reconstituted ODR-10::GFP expression in AWA cilia of control and PD adults. MKS-5::tagRFP marks the ciliary transition zones. Arrows indicate the PCMC. Anterior is at top. Scale bars: 5 µm. d) Quantification of total ciliary reconstituted ODR-10::GFP fluorescence in AWA PCMC and primary stalk in control and PD adults. Each dot is a measurement from an individual animal. Bars represent the mean, and error bars are SEM. n ≥ 29 animals. Control and PD adults were imaged in parallel; ≥3 independent experiments. *** indicates different at P<0.001 (2-tailed Welch’s t-test). e) Behavioral responses of wild-type control and wild-type control animals overexpressing ODR-10::tagRFP under the gpa-4Δ6 promoter (OX ODR-10; 5 ng/µl) to the indicated dilutions of diacetyl. Each dot represents the chemotaxis index of a single assay plate containing ∼50–300 adult hermaphrodites. Bars represent the mean; error bars are SEM. A subset of control and experimental chemotaxis assays were performed in parallel over at least 3 days. * and ** indicate different at each concentration at P < 0.05 and <0.01, respectively (2-tailed Welch’s t-test). F) Quantification of peak fluorescence intensity changes in AWA to a 10-s pulse of 10⁻⁷ diacetyl. Responses of wild-type control animals injected with gpa-4Δ6p::odr-10::tagRFP at 5 and 30 ng/µl are shown. Bars represent mean, error bars are SEM. n ≥ 16 animals (1 neuron per animal) each. Animals were examined over 2 days with the exception of OX ODR-10 (30 ng/µl). * and ** indicate different at P < 0.05 and <0.01, respectively (Kruskal–Wallis with Dunn’s multiple comparisons test).

To establish whether increasing ODR-10 levels alone in AWA is sufficient to enhance diacetyl response sensitivity even in control animals, we overexpressed odr-10 from the constitutive gpa-4Δ6 promoter and examined diacetyl responses via both behavioral assays and imaging of diacetyl-evoked intracellular calcium dynamics. As shown in Fig. 2e, the behavioral responses of control animals overexpressing odr-10 were higher than those of wild-type control animals across multiple concentrations of diacetyl. Overexpression of odr-10 was also sufficient to increase diacetyl-evoked calcium responses in AWA in control adults (Fig. 2f and Supplementary Fig. 2b). We also addressed the possibility that upregulation of a diacetyl receptor other than ODR-10 confers enhanced diacetyl responses in PD animals. However, neither control nor PD odr-10(ky32) animals (Sengupta et al. 1996) responded to diacetyl (Supplementary Fig. 2, c and d), indicating that this receptor is essential for diacetyl responses in both conditions. We conclude that increased ODR-10 levels may be sufficient to account for the enhanced diacetyl responses of PD animals.

odr-10 expression and diacetyl behavioral responses are increased in dauer larvae

C. elegans dauer larvae as well as the analogous infective juvenile larvae of parasitic nematodes have been shown to exhibit distinct olfactory behaviors (Vertiz et al. 2021). The expression of many neuronal genes including chemoreceptor genes is also markedly altered in dauer larvae (Peckol et al. 2001; Nolan et al. 2002; Hall et al. 2010; Vidal et al. 2018; Bhattacharya et al. 2019). The dauer-specific expression patterns of a subset of these genes is maintained in PD adults, while the expression of other genes is further altered to a PD-specific pattern or restored to the pattern observed in control adults (Peckol et al. 2001; Hall et al. 2010; Vidal et al. 2018). We asked whether odr-10 expression is modulated in dauers, following which ciliary protein levels may be subsequently maintained at higher levels in PD animals.

Expression of the odr-10::t2A::mNG reporter remained restricted to the AWA neurons in dauer larvae (Fig. 3a). In contrast to our observations in PD adults, odr-10::t2A::mNG expression was strongly upregulated in dauer animals as compared to levels in L3 larvae (Fig. 3, a and b). Consistently, reconstituted ODR-10::GFP protein levels were also increased in AWA cilia in dauers (Fig. 3, c and d).

Fig. 3.

Dauer larvae exhibit upregulated odr-10 expression and enhanced diacetyl responses. a) Representative images of AWA neurons in L3 and dauer larvae expressing ODR-10::t2A::mNeonGreen from the endogenous odr-10 locus. Arrows indicate AWA cell bodies. Anterior is at left. Scale bar: 5 µm. b) Quantification of ODR-10::t2A::mNeonGreen fluorescence in AWA neurons of the indicated animals. Each dot is a measurement from a single neuron. Bars represent the mean; error bars are SEM. n ≥ 31 neurons (≥18 animals). L3 and dauer larvae were imaged on the same day, ≥3 independent experiments. *** indicates different between indicated at P < 0.001 (2-tailed Welch’s t-test). c) Representative images of reconstituted ODR-10::GFP expression in AWA cilia of L3 and dauer larvae. MKS-5::tagRFP marks the ciliary transition zones. Arrows indicate the AWA PCMC. Anterior at top. Scale bars: 5 µm. d) Quantification of total reconstituted ODR-10::GFP fluorescence in the AWA PCMC and primary stalk in L3 and dauer animals. Each dot is a measurement from an individual animal. Bars represent the mean total fluorescence, error bars are SEM. n ≥ 23 animals. *** indicates different between indicated at P<0.001 (2-tailed Welch’s t-test). e) Average changes in GCaMP2.2b fluorescence in AWA to a 30-s pulse of 10⁻⁵ dilution of diacetyl in control and PD adults, and dauer larvae. Shaded regions indicate SEM. Control and PD adult data are repeated from Fig. 1d. n ≥ 16 neurons (1 neuron per animal). e) Quantification of peak fluorescence intensity changes in AWA expressing GCaMP2.2b to a 30-s pulse of 10⁻⁵ dilution of diacetyl. Each dot is a measurement from a single neuron. Bars represent mean fluorescence, error bars are SEM. n ≥ 16 neurons (1 neuron per animal). Control and PD adult data are repeated from Fig. 1d. * and *** indicate different at P < 0.05 and <0.001, respectively (1-way ANOVA with Tukey’s multiple comparisons test).

We tested whether upregulation of odr-10 expression in dauers correlates with increased diacetyl responses in these animals. To examine diacetyl-evoked calcium responses in AWA, we modified the microfluidics imaging device typically used for imaging C. elegans adults (Chronis et al. 2007) to accommodate the thinner and smaller dauer larvae (see Materials and Methods), although we were unable to use these or the adult imaging devices to examine L3 larvae. In response to a pulse of 10⁻⁵ dilution of diacetyl, dauer animals exhibited markedly increased responses as compared to control or PD adults (Fig. 3, e and f). Altered expression levels of the calcium sensor in AWA are unlikely to account for the observed increase in diacetyl responses in dauer larvae (Supplementary Fig. 1c). We could not reliably assess the behavioral responses of dauers to diacetyl in our behavioral assays in part due to their unique locomotory patterns (Gaglia and Kenyon 2009; Bhattacharya et al. 2019), although we note that a previous report indicated that dauers exhibit decreased behavioral responses to diacetyl (Vertiz et al. 2021). We conclude that odr-10 may be upregulated upon entry into the dauer stage possibly via transcriptional mechanisms and correlates with increased diacetyl responses. While this transcriptional upregulation is not maintained in PD adults, increased ODR-10 protein levels in AWA cilia correlates with enhanced diacetyl responses in PD adults.

Upregulation of odr-10 expression in dauer larvae is mediated in part via the DAF-16 FOXO transcription factor

The TGF-β and insulin signaling pathways act in parallel to regulate dauer formation in response to adverse environmental conditions (Riddle and Albert 1997; Fielenbach and Antebi 2008). Insulin signaling inhibits nuclear translocation of the DAF-16 FOXO transcription factor, and DAF-16 is nuclear-localized in dauer larvae (Lin et al. 1997; Ogg et al. 1997; Aghayeva et al. 2021). Since this molecule has been implicated in the altered regulation of sensory gene expression in dauer larvae (Bhattacharya et al. 2019; Wexler et al. 2020; Aghayeva et al. 2021), we tested whether the observed upregulation of odr-10 expression in dauers is mediated in part via DAF-16-dependent transcriptional regulation.

Since daf-16 mutants cannot enter the dauer stage, we assessed the effects of daf-16 depletion via auxin-induced degradation specifically in AWA (Nishimura et al. 2009; Zhang et al. 2015). We expressed the auxin receptor TIR1 in AWA in a strain in which the endogenous daf-16 locus has been edited to include a degron tag (Aghayeva et al. 2020, 2021). To obtain larger numbers of dauer larvae, these experiments were performed in temperature-sensitive daf-2 insulin receptor mutants that constitutively enter the dauer stage even under favorable environmental conditions (Fig. 4a) (Riddle et al. 1981; Gems et al. 1998). As in wild-type animals, odr-10::t2A::mNG expression levels were upregulated in daf-2 dauer larvae as compared to expression levels in control or PD daf-2 adults in the absence of auxin treatment (Fig. 4b). We found that auxin-mediated depletion of DAF-16 in AWA decreased, although did not fully abolish, the upregulated odr-10 expression observed in dauer animals (Fig. 4, a-i and b). Growth on auxin had little effect on odr-10 expression levels in adult animals that bypassed the dauer stage (Fig. 4, a-ii and b), although we note that odr-10 was previously identified as a putative DAF-16-regulated gene in a comparison of the neuronal transcriptomes of daf-2 and daf-16; daf-2 adult animals (Kaletsky et al. 2016). Addition of auxin only during dauer recovery also did not affect odr-10 expression in adult animals (Fig. 4, a-iii and b). We infer that DAF-16 function in AWA is partly necessary during dauer entry to transcriptionally upregulate odr-10 expression, although we are unable to exclude the possibility that DAF-16 is not fully depleted in AWA under these conditions.

Fig. 4.

DAF-16 FOXO may be partly necessary for upregulation of odr-10 expression in dauer larvae. a) Cartoons of different growth conditions of daf-2(e1368) animals expressing daf-16::mNeptune::AID, odr-10::t2A::mNeonGreen and gpa-4Δ6p::TIR1 with or without 1 mM auxin. b) Quantification of mean ODR-10::t2A::mNeonGreen fluorescence in AWA in animals grown with or without 1 mM auxin in conditions indicated in Ai-Aiii. Each point is a measurement from an individual AWA neuron. Bars represent mean fluorescence, error bars are SEM. n ≥ 46 neurons (≥23 animals) for dauers; n ≥ 40 neurons (≥22 animals) for adults. Dauer larvae and control and PD adults grown with or without 1 mM auxin were imaged in parallel; 3 independent experiments. *** indicates different between indicated at P < 0.001 (2-tailed Welch’s t-test); ns, not significant. c) (Left) Diagram of genomic odr-10 locus with arrowhead indicating location of predicted DAF-16 binding motif (Wexler et al. 2020). (Right) Quantification of total reconstituted ODR-10::GFP fluorescence in AWA PCMC and stalk. mut indicates animals in which the conserved DAF-16 binding sequence within the endogenous odr-10 promoter has been mutated. Each dot is a measurement from an individual animal. Bars represent the mean total fluorescence, error bars are SEM. n ≥ 29 animals. Adult control and PD wild-type data are repeated from Figure 2D. *** indicates different at P < 0.001 (1-way ANOVA with Tukey’s multiple comparison test); ns, not significant.

A putative DAF-16 binding site in the proximal regulatory sequences of odr-10 was previously shown to be necessary for starvation-dependent upregulation of odr-10 expression driven from a transcriptional reporter in C. elegans males (Wexler et al. 2020). However, mutating this site in the endogenous odr-10 locus via gene editing had no effect on the upregulated levels of ciliary ODR-10 protein in either dauer or PD animals (Fig. 4c). Multiple DAF-16 binding sites in odr-10 regulatory sequences may contribute redundantly in the context of the endogenous promoter to the upregulation of odr-10 expression in dauer larvae. Alternatively, DAF-16 may act indirectly to regulate endogenous odr-10 expression.

AWA neurons exhibit distinct gene expression profiles in control and PD adults

In addition to diacetyl, PD adults also exhibit increased responses to the AWA-sensed odorants pyrazine and 2,4,5-trimethylthiazole (Fig. 1b). Although the receptors for these chemicals are as yet unidentified, this observation suggests that the expression of chemoreceptors in addition to odr-10 in AWA may also be altered as a function of dauer passage. To test this notion, we dissociated control and PD L4 larvae expressing GFP specifically in AWA, collected GFP-labeled populations of AWA neurons via fluorescence-activated cell sorting (FACS), and performed transcriptional profiling (Taylor et al. 2021). To obtain large populations of growth-synchronized L4 animals for cell sorting, the control population was grown from L1-arrested larvae (see Materials and Methods). In parallel, we also transcriptionally profiled populations of dissociated but unsorted cells from control and PD L4 animals.

Principal component analyses indicated that with the exception of 1 sample, the RNA-Seq profiles of all biologically independent replicates of sorted control and PD AWA neurons were present in a cluster distinct from that obtained from cells collected from whole animals (Supplementary Fig. 3a). AWA-expressed genes were the most enriched in the dataset collected from sorted AWA neurons (Supplementary Fig. 3b). Moreover, the 20 most highly expressed genes in AWA as predicted by the CeNGEN neuronal profiling project (Taylor et al. 2021) were robustly represented as upregulated in the AWA RNA-Seq data as compared to data from whole animals (Supplementary Fig. 3c). These observations indicate that we successfully enriched and profiled populations of AWA neurons.

Comparison of the gene expression changes in the datasets from control and PD whole animals indicated that the expression of multiple GPCR genes was altered as a function of dauer passage (Fig. 5, a and b and Supplementary File 1). There was minimal overlap with previously published datasets of genes differentially regulated in PD vs. control whole animals (Hall et al. 2010; Ow et al. 2018), likely due to differences in conditions of animal growth and sample preparation although we note that genes predicted to be involved in GPCR signaling were also previously identified as a differentially expressed category (Hall et al. 2010). Up- or downregulated chemoreceptor genes are predicted to be expressed in multiple chemosensory neuron types, indicating that modulation of chemoreceptor gene expression as a function of developmental history is mediated at the level of individual chemoreceptor genes and not sensory neuron types. However, despite the established importance of neuropeptide and hormonal signaling in regulating dauer entry (Riddle and Albert 1997; Fielenbach and Antebi 2008), the expression of only a few predicted neuropeptide genes appeared to be altered in PD as compared to control animals (Fig. 5b and Supplementary File 1).

Fig. 5.

Gene expression is altered in AWA following passage through the dauer stage. a) MA plot showing enrichment and log₂ fold changes of differentially expressed genes in PD vs. control whole worm RNA-Seq libraries. The names of a subset of genes from different gene families are indicated. Up- and downregulated genes were determined by differential expression analysis with a log₂ fold change cut-off >1.5, Padj < 0.05; NS, not significant. b) Waterfall plot of differentially expressed genes from indicated gene families in control and PD whole animal RNA-Seq data. Up- and downregulated genes were determined by differential expression analysis with a log₂ fold change cut-off >1.5, Padj < 0.05. c) MA plot showing enrichment and log₂ fold changes of differentially expressed genes in PD vs. control AWA RNA-Seq libraries. The names of a subset of genes from different gene families are indicated. Up- and downregulated genes were determined by differential expression analysis with a log₂ fold change cut-off >1.5, Padj < 0.05; NS, not significant. d) Waterfall plot of differentially expressed genes from indicated gene families in PD vs. control AWA RNA-Seq libraries. Up- and downregulated genes were determined by differential expression analysis with a log₂ fold change cut-off >1.5, Padj < 0.05.

In contrast to the gene expression changes in the whole animal dataset, the expression of only a small number of GPCRs appeared to be altered in PD vs. control AWA neurons (Fig. 5, c and d and Supplementary File 2). As expected, odr-10 transcript levels were not significantly changed in PD adults. However, we noted that the expression of several neuropeptide genes and in particular, multiple transcription factors was significantly altered in PD vs. control adults (Fig. 5d and Supplementary File 2). Affected transcription factors belong to multiple subfamilies including the greatly expanded nuclear hormone receptor family, members of which are predicted to be coexpressed in sensory neurons along with putative chemoreceptor GPCRs, and which have been suggested to act as receptors for external and internal cues (Sural and Hobert 2021; Taylor et al. 2021). Together, these results indicate that passage through the dauer stage alters the expression of genes from multiple families, and that the expression of genes in individual sensory neurons such as AWA is also affected.

The expression of multiple AWA-expressed chemoreceptor genes is upregulated in dauers and is maintained at higher levels in PD adults

As described above, while odr-10 expression is not altered in PD adults, expression of this gene is transcriptionally upregulated in dauer larvae following which ciliary ODR-10::GFP protein levels are maintained at higher levels in PD adults. Since our RNA-Seq experiments indicated that the mRNA levels of only a small number of GPCR genes are altered in PD as compared to control AWA neurons, we tested whether additional AWA-expressed chemoreceptor genes are regulated similarly to odr-10.

Since we were unable to efficiently dissociate dauer larvae likely due to their modified cuticle (Cassada and Russell 1975) and thus could not profile populations of sorted AWA neurons from dauer larvae, we instead examined the expression of a subset of endogenously tagged chemoreceptor genes. Similar to odr-10, mRNA levels of the putative AWA-expressed chemoreceptors srd-27, srd-28, and str-44 were also not significantly upregulated in AWA in PD adults (Fig. 5, c and d and Supplementary File 2) although srd-28 was identified as a gene significantly upregulated in PD adults in the whole animal dataset (Fig. 5b). srd-27, srd-28 and str-44 endogenously tagged with t2A::mNG were specifically expressed bilaterally in the AWA olfactory neurons in L3 larvae and control adults (Fig. 6a) (McLachlan et al. 2022). Expression of all 3 genes was upregulated in AWA in dauer larvae as compared to L3 larvae with srd-28 and str-44 showing stronger changes (Fig. 6, a–d). However, unlike odr-10 whose expression remained restricted to AWA in all examined stages, all 3 chemoreceptor genes were expressed in additional neurons in the head in dauer larvae (Fig. 6a). Expression in all neurons including in AWA was subsequently downregulated in PD adults relative to dauer larvae, although srd-28, str-44, and srd-27 retained higher expression levels in PD as compared to control adults (Fig. 6, a–d). As in the case of odr-10, depletion of DAF-16 during dauer entry via auxin-induced degradation (Fig. 4a-i) also decreased but did not abolish upregulation of srd-28::t2A::mNG expression in dauer larvae (Fig. 6e). We conclude that the expression of multiple AWA-expressed chemoreceptor genes is likely transcriptionally upregulated in dauer larvae in part via DAF-16-dependent mechanisms, and that increased levels of a subset of these receptors may be maintained in PD adults possibly in part via posttranscriptional mechanisms.

Fig. 6.

The expression of a subset of AWA-expressed chemoreceptor genes may be regulated similarly to odr-10 upon dauer passage. a) Representative images of receptor::t2A::mNeonGreen expression from the corresponding endogenous loci in L3, dauer, control adult, and PD adult stage animals. AWA neurons are marked via expression of gpa-4Δ6p::tagRFP. Arrows indicate AWA neurons; asterisks indicate ectopic expression in other cell types. Anterior is at left. Scale bars: 5 µm. b–d) Quantification of mean receptor::t2A::mNeonGreen fluorescence in AWA neurons in animals of the indicated developmental stages. Each dot is a measurement from an individual neuron. Bars represent mean; error bars are SEM. n ≥ 28 neurons (≥18 animals). L3 and dauer larvae were imaged in parallel; 3 independent experiments. Control and PD adults were imaged in parallel; 3 independent experiments. *, **, and *** indicates different at P < 0.05, <0.01, and <0.001 (Kruskal–Wallis with Dunn’s multiple comparisons test); ns, not significant. e) Quantification of mean SRD-28::t2A::mNeonGreen fluorescence in AWA in daf-2(e1368) animals expressing daf-16::mNeptune::AID and gpa-4Δ6p::TIR1 grown with or without 1 mM auxin. Animals were grown as indicated in the cartoon in Fig. 4a-i. Each dot is a measurement from an individual neuron. Bars represent mean fluorescence, error bars are SEM. n ≥ 37 neurons (≥22 animals). Control and dauers grown on plates with and without 1 mM auxin plates were imaged in parallel; 3 independent experiments. *** indicates different at P < 0.001 (2-tailed Welch’s t-test); ns, not significant.

Discussion

We show here that adult C. elegans hermaphrodites that transiently passed through the dauer developmental stage exhibit enhanced olfactory responses associated with food-seeking as compared to adult animals that bypassed this stage. Increased sensitivity to the odorant diacetyl is correlated with upregulated expression of the diacetyl receptor ODR-10 in dauer larvae, and higher levels of the ODR-10 protein in AWA olfactory cilia in PD adults. Via transcriptional profiling, we further show that levels of a subset of additional AWA-expressed chemoreceptors are also modulated by developmental stage and trajectory. Our results suggest that state- and experience-dependent cues are integrated to differentially modulate individual chemoreceptor levels in a single sensory neuron type likely via both transcriptional and posttranscriptional mechanisms, thereby providing a possible mechanism underlying developmental history-dependent olfactory behavioral plasticity.

Upon entry into the dauer stage, odr-10 expression is transcriptionally upregulated in part via DAF-16. However, activation of DAF-16 alone appears to be insufficient to upregulate odr-10 expression since odr-10 expression is not upregulated in daf-2 insulin receptor mutants that did not enter the dauer stage. Although we are unable to exclude the possibility that DAF-16 is only partially depleted in AWA upon auxin treatment, it is likely that additional factors including transcription factors such as the DAF-3 SMAD or the DAF-12 nuclear hormone receptor that act downstream of dauer-promoting hormonal signals also play a role in regulating odr-10 expression (Fielenbach and Antebi 2008; Sims et al. 2016; Aghayeva et al. 2021). Similarly, altered expression of innexins and other chemoreceptors in dauers were shown to be only partly DAF-16-dependent (Aghayeva et al. 2021). Upon exit from the dauer stage, odr-10 expression is downregulated possibly due to inactivation of DAF-16 and the altered chromatin profile of these animals (Hall et al. 2010; Sims et al. 2016), but levels of ciliary ODR-10 protein continue to be maintained at higher levels than in control animals. Ciliary GPCRs are trafficked to the PCMC and further trafficked into and out of the cilium via diffusion and motor-driven transport (Mukhopadhyay et al. 2017; Nachury and Mick 2019). We previously showed that ciliary trafficking of chemoreceptors in different C. elegans sensory neurons is regulated by multiple neuron- and receptor-specific mechanisms and that these trafficking mechanisms are further modulated by sensory signaling (Brear et al. 2014; DiTirro et al. 2019). Sensory neuron cilia including those of AWA are extensively remodeled in dauers but their morphologies appear to be restored to those resembling those in control animals in PD adults (Albert and Riddle 1983; Britz et al. 2021) (this work). We propose that 1 or more ciliary trafficking mechanisms are altered in PD animals resulting in increased ODR-10 protein trafficking into, or decreased removal from, the AWA cilia in PD animals. In addition to transcriptional changes in chemoreceptor expression, regulated trafficking of ciliary GPCRs provides an additional mechanism to fine tune sensory responses as a consequence of developmental experience. However, increased ciliary ODR-10 levels in PD adults may also arise due to perdurance of the protein following dauer exit.

Why do PD adults enhance their responses to food-related odors? Stress, including starvation, experienced during early larval stages may indicate that food availability is unreliable. Increased attraction to food in PD adults may be a bet-hedging strategy that optimizes growth and survival in a variable environment. However, although animals also retain a cellular memory of starvation-induced L1 arrest (Jobson et al. 2015; Webster et al. 2018), this arrest does not appear to alter examined olfactory behaviors in adults, suggesting that stress assessed during a distinct period in development and/or dauer entry is required to drive behavioral plasticity in ensuing adults. Interestingly, we and others previously showed that expression of the osm-9 TRPV channel gene is strongly downregulated in the ADL nociceptive but not AWA neurons in PD adults, resulting in decreased responses to the ADL-sensed aversive ascr#3 pheromone (Sims et al. 2016). Decreased aversion to this pheromone by PD adults has been suggested to be a mechanism that inhibits dispersion in a crowded environment and promotes outcrossing (Sims et al. 2016). Coordinated differential modulation of responses in defined subsets of chemosensory neurons likely allow PD adults to optimize survival and reproduction.

In contrast to the main olfactory systems of vertebrates in which each olfactory sensory neuron expresses 1 or very few olfactory receptors (Ressler et al. 1993; Vassar et al. 1993; Chess et al. 1994), the coexpression of as many as 100 chemoreceptors in each chemosensory neuron type in C. elegans (Troemel et al. 1995; Vidal et al. 2018; Taylor et al. 2021) raises a unique challenge for this organism. Since coexpressed chemoreceptors are likely tuned to distinct odorants, modulating synaptic transmission from a single chemosensory neuron in C. elegans as a function of internal state would be expected to coordinately alter responses to a broad range of chemicals sensed by that neuron type. Regulation of individual chemoreceptors instead enables the animal to precisely target and modulate defined chemosensory behaviors. Consistent with this notion, expression of chemoreceptor genes is subject to complex modes of regulation. odr-10 expression is higher in hermaphrodites and juvenile male larvae than in adult males, and expression is upregulated in adults of both sexes upon starvation (Ryan et al. 2014; Wexler et al. 2020). The expression of multiple chemoreceptors in many sensory neuron types is dramatically altered in dauer larvae and subsequently further modulated in PD animals, whereas the expression of other receptors is sexually dimorphic (Troemel et al. 1995; Peckol et al. 1999; Nolan et al. 2002; Vidal et al. 2018). In the ADL nociceptive neurons, expression of the srh-234 chemoreceptor has been shown to be regulated via cell-autonomous and non cell-autonomous mechanisms (Gruner et al. 2014, 2016). Thus, the expression of each chemoreceptor gene may be modulated by multiple regulatory modules that act combinatorially to integrate distinct inputs and appropriately calibrate the behavioral response.

Individual chemosensory neurons in Aedes aegypti also express multiple receptors from different receptor subfamilies (Herre et al. 2022). The expression of subsets of these receptors has been shown to be modulated by feeding and reproductive state although whether receptor expression is causative to altered olfactory behavioral profiles is unclear (Matthews et al. 2016). In rodents, a subset of specialized olfactory neurons in rodents express multiple receptors of the MS4A family, each of which responds to ethologically relevant chemical stimuli (Greer et al. 2016). It will be interesting to assess whether expression of these receptors is also subject to extensive state-dependent modulation, and whether this modulation drives sensory behavioral plasticity as a function of current and past external and internal conditions.

Data Availability

All plasmids used in this work are listed in Supplementary Table 1. All strains used in this work are listed in Supplementary Table 2. Strains and plasmids are available upon request. Data underlying this article are available at https://github.com/SenguptaLab/PDplasticity. Sequencing files of the RNA-Seq experiments have been deposited in the Array Express database at EMBL-EBI (www.ebi.ac.uk/arrayexpress) under accession number E-MTAB-11823.

Supplementary material is available at Figshare: https://doi.org/10.25386/genetics.20752723. Supplementary files include: Supplementary Table 1 (plasmids used in this work), Supplementary Table 2 (strains used in this work), Supplementary Figs. 1–3, supplementary figure legends, supplementary references, Supplementary File 1—RNA-Seq data of differentially expressed genes from control and PD whole animals, Supplementary File 2—RNA-Seq data of differentially expressed genes from sorted control and PD AWA neurons, Supplementary File 3—schematic of microfluidics device used to image dauer larvae, and Supplementary File 4—AutoCAD file used to generate an ink photomask (outputcity.com).