Experience-dependent reconfiguration of thermoreceptors regulates neuronal response plasticity

SUMMARY

Neurons continuously adjust their properties as a function of experience. Precise modulation of neuronal responses is achieved by multiple cellular mechanisms that operate over a range of timescales. Primary sensory neurons rapidly adapt their sensitivities via posttranslational mechanisms including regulated trafficking of sensory molecules^1–4 but also alter their transcriptional profiles on longer timescales to adapt to persistent sensory stimuli^5–8. How diverse transcriptional and posttranscriptional pathways are coordinated in individual sensory neurons to accurately adjust their functions and drive behavioral plasticity is unclear. Here we show that temperature experience modulates both transcription and trafficking of thermoreceptors on different timescales in the C. elegans AFD thermosensory neurons to regulate response plasticity. Expression of the PY motif-containing adaptor protein (PYT-1) as well as the GCY-18 warm temperature-responsive guanylyl cyclase thermoreceptor⁹ is transcriptionally upregulated in AFD upon a temperature upshift^5,10. We find that as GCY-18 begins to accumulate at the AFD sensory endings, the GCY-23 cooler temperature-responsive thermoreceptor⁹ exhibits altered subcellular localization and increased retrograde trafficking, thereby increasing the functional GCY-18 to GCY-23 ratio in the AFD sensory compartment. Altered GCY-23 localization and trafficking require PYT-1-dependent endocytosis, and we show that PYT-1-mediated modulation of the GCY-18 to GCY-23 protein ratio at the AFD sensory endings is necessary to shift the AFD response threshold towards warmer values following the temperature upshift. Our results describe a mechanism by which transcriptional and posttranscriptional mechanisms are temporally coordinated across sensory receptors to fine tune experience-dependent plasticity in the response of a single sensory neuron type.

eTOC

Harris, Dutta et al. find that experience-dependent plasticity in the activation threshold of the AFD thermosensory neurons is mediated by modulating warm and cold thermoreceptor levels at the sensory ending. This is achieved by temporally coordinating transcriptional and posttranscriptional mechanisms in response to temperature experience, thereby optimizing cellular responses.

RESULTS and DISCUSSION

PYT-1 mediates endosomal localization of a thermoreceptor at the AFD sensory compartment upon a temperature upshift

C. elegans exhibits a preference for its recently experienced temperature when placed on a thermal gradient^11,12. This behavioral plasticity is mediated via adaptation of the sensory response (T*_AFD) and synaptic output threshold of the AFD thermosensory neuron pair^10,13–17. Thus, moving C. elegans from 15°C to 25°C for 4 hours increases both T*_AFD and the synaptic output threshold of AFD, thereby promoting navigation toward warmer temperatures^{10,11,13–17}. The GCY-8, GCY-18, and GCY-23 thermoreceptor guanylyl cyclases (rGCs: receptor guanylyl cyclases) are expressed specifically in AFD and localize to their sensory endings^18,19 (Figure 1A). While loss of all three rGCs abolishes AFD temperature responses^9,17, analyses of phenotypes upon loss or misexpression of single rGCs suggest that GCY-18 and GCY-23 confer responses to warmer and cooler temperatures, respectively^1,9. The role of GCY-8 in thermosensation is unclear and is not addressed here. These observations suggest the hypothesis that increasing the ratio of functional warm- to cold-responsive rGCs at the AFD sensory compartment may drive T*_AFD to warmer values following a temperature upshift.

Figure 1. A temperature upshift selectively relocalizes GCY-23 to an endosomal compartment in the AFD sensory ending in a PYT-1-dependent manner.

A) Cartoon of the AFD sensory ending showing localization of thermosensory rGCs in the microvilli membrane^5,18,19. Anterior is at left.

B) Representative images of endogenously tagged GCY-18::GFP and GCY-23::GFP localization within the AFD sensory endings under the indicated temperature shift conditions (top). Arrow indicates relocalized GCY-23::GFP to a dot at the center of the sensory compartment. Anterior is at left. Scale bar: 5 μm.

C) (Left) Representative images of GCY-23::GFP localization in the AFD sensory endings in the indicated temperature conditions. Scale bar: 2 μm. (Right) Average intensity distributions of GCY-23::GFP in the AFD sensory compartment from line scan analyses (red horizontal line in images at left). The percentage of sensory endings exhibiting GCY-23::GFP relocalization to a dot at the center of the AFD sensory ending (arrow) is indicated (see Star Methods and Figure S1A). n=50–66.

D) Representative images showing colocalization (arrow) of PYT-1::GFP and RFP::RAB-5 at the AFD sensory endings following a shift to 25°C for 4 hours. The percentage of sensory endings that exhibit colocalization is indicated. Scale bar: 2 μm. n=21.

E) (Left) Representative images of GCY-23::GFP colocalization with TagRFP::RAB-5 after a temperature upshift for 4 hours. The percentage of sensory endings that exhibit colocalization is indicated. Scale bar: 2 μm. n=25. (Right) Quantification of GCY-23::GFP fluorescence levels in the AFD sensory endings of animals grown at the indicated conditions in the shown genotypes. Each dot is the measurement from a single AFD sensory ending that excludes the GCY-23::GFP fluorescent signal from the RFP::RAB-5-marked region (dotted ellipse in images at left). n=22–26.

F) Representative images (left) and quantification (right) of GCY-18::GFP fluorescence in AFD sensory endings of animals grown at the indicated conditions in the shown genotypes. Each dot is the measurement from a single AFD sensory ending. Scale bar: 10 μm. n=19–22.

Horizontal and vertical lines indicate mean and SD, respectively. *** - different at p<0.001 (t test); ns – not significant. All data are from at least 2 independent experiments.

Also see Figures S1 and S2.

We previously showed that gcy-18 mRNA levels are upregulated within 4 hours after a temperature upshift²⁰. To investigate how transcriptional and posttranscriptional mechanisms are coordinated to regulate rGC protein concentrations, we imaged endogenous fluorophore-tagged GCY-18 and GCY-23 in the AFD sensory compartment following a temperature change. Both GCY-18::GFP and GCY-23::GFP localized specifically to AFD microvilli in animals cultivated overnight at 15°C^5,18,19 (Figure 1B, 1C). After a 4 hour shift to 25°C, the distribution and localization pattern of GCY-18::GFP at the sensory compartment was unaltered (Figure 1B, also see Figure S2B). In contrast, although GCY-23::GFP remained in the microvilli, a subset redistributed to the center of the sensory endings in a fraction of animals (Figure 1B, 1C, Figure S1A).

The pattern of GCY-23 enrichment at the center of the AFD sensory compartment resembled that of the PY motif transmembrane 1 (PYT-1) adaptor protein 4 hours after a temperature upshift⁵ (Figure S1B, S1C). We previously showed that AFD-specific pyt-1 expression is rapidly induced via an activity-dependent transcriptional pathway 1–4 hours after a shift to 25°C (summarized in Figure S1B), and that pyt-1 mutants are defective in resetting T*_AFD to warmer values⁵. Endogenously tagged GCY-23::TagRFP and PYT-1::GFP colocalized at the center of the AFD sensory ending after a shift to 25°C for 4 hours (Figure S1C). PY motif containing adaptor proteins localize to endosomes where they interact with WW domain containing E3 ubiquitin ligases via their PY motifs, and recruit transmembrane receptors for ubiquitination and degradation^21–29. Endosomes are present at the periciliary membrane compartment (PCMC) at the base of cilia in C. elegans sensory neurons^30,31. While endosomes have not been reported in AFD sensory endings, we previously observed vesicles in this region using electron microscopy³². We examined whether endosomes are present in the AFD microvilli, and whether PYT-1 localizes to these organelles to direct GCY-23 to endocytic pathways.

We found that fluorescently tagged RAB-5, an early endosome marker^33,34, was concentrated in the center of the AFD sensory ending in a temperature- and PYT-1-independent manner (Figure 1D, 1E, Figure S1D). PYT-1::GFP co-localized with TagRFP::RAB-5 after a 4 hour shift to 25°C (Figure 1D), indicating that PYT-1 is present in endosomes. A subset of GCY-23::GFP also colocalized with TagRFP::RAB-5 at the center of the sensory ending after a shift to 25°C (Figure 1E), suggesting that PYT-1 may recruit GCY-23 to endosomes in the AFD sensory compartment. Consistently, GCY-23 relocalization to the center of the AFD sensory endings after a temperature upshift was reduced in pyt-1 mutants (Figure 1C, Figure S2A). No effects were observed on the overall distribution of GCY-18::GFP at the AFD sensory compartment in pyt-1 mutants (Figure S2B). We infer that PYT-1 selectively modulates the redistribution of GCY-23 at the AFD sensory endings following a temperature upshift.

We next examined whether temperature-dependent GCY-23 relocalization to endosomes was accompanied by reduced abundance in the AFD sensory ending. Overall levels of GCY-23::GFP in the sensory ending were grossly unaltered by temperature or in pyt-1 mutants (Figure S2C). Reasoning that the subset of endosome-associated GCY-23 protein may be non-functional, we excluded the GCY-23::GFP signal from the RAB-5-marked region when quantifying GCY-23 levels. This analysis showed that GCY-23 abundance in the AFD microvilli decreased following a temperature upshift in a pyt-1-dependent manner (Figure 1E). In contrast, the sensory ending abundance of GCY-18 increased after a shift to 25°C independently of PYT-1, consistent with transcriptional upregulation^5,20 (Figure 1F). We suggest that following a temperature upshift, PYT-1 promotes selective relocalization of GCY-23 to endosomes thereby decreasing functional levels of this receptor in the microvilli, without affecting GCY-18 abundance or distribution.

Trafficking of GCY-23 and GCY-18 is regulated differentially by temperature and PYT-1

Sensory signaling proteins are typically synthesized in the soma and trafficked anterogradely via the dendrite to the receptive endings of sensory neurons. Conversely, proteins removed from the sensory compartment are endocytosed and targeted for degradation or recycling either locally or following retrograde trafficking to the soma^30,35–39. We tested whether endosomal localization of GCY-23::GFP in the AFD sensory endings is accompanied by increased retrograde dendritic trafficking. We examined trafficking of GCY-23::GFP and GCY-18::GFP in the distal AFD dendrite directly adjacent to the sensory compartment as a function of temperature (Figure 2A). Although few mobile GFP-containing particles were detected at 15°C, we observed a significant increase in mobile GCY-23 and GCY-18 puncta after a 4 hour shift to 25°C (Figure 2B, 2C, 2E, 2F, Figure S3A, S3B, Videos S1–S4). Mobile GCY-23::GFP-containing puncta predominantly trafficked retrogradely from the AFD sensory ending to the soma (Figure 2D, Figure S3A, Videos S2), whereas GCY-18::GFP containing puncta showed a bias toward anterograde movement to the sensory ending (Figure 2G, Figure S3B, Video S4). These observations suggest that the sensory ending distribution and trafficking of GCY-23 and GCY-18 are differentially regulated following warming. Specifically, functional GCY-18 levels at the AFD sensory endings are upregulated via increased transcription and anterograde flux, whereas functional GCY-23 levels are likely reduced by endosomal localization and enhanced retrograde trafficking.

Figure 2. PYT-1 promotes retrograde trafficking of GCY-23 upon a temperature upshift.

A) Cartoon of an AFD neuron. Dashed rectangle indicates the dendritic region imaged.

B, E) Representative images showing GCY-23::GFP (B) or GCY-18::GFP (E) puncta in an AFD dendrite in the indicated conditions and genotypes. Orange arrowheads mark individual puncta. Scale bar: 10 μm.

C, F) Quantification of mobile GCY-23::GFP (C) or GCY-18::GFP (F) puncta number observed during a 60s video in AFD dendrites from the indicated conditions and genotypes. n=21–30.

D, G, H) Quantification of the bias towards anterograde versus retrograde movement of GCY-23::GFP (D, H) or GCY-18::GFP (G) particles in AFD dendrites from the indicated conditions and genotypes. The average flux of retrograde moving particles was subtracted from the average flux of anterograde moving particles such that positive and negative values indicate a bias towards anterograde or retrograde flux, respectively (see Star Methods). Vertical and horizontal lines indicate mean and SD, respectively. n=21–32. The rab-5 DN allele used in H was rab-5(S33N).

For all panels, each dot is the measurement from one AFD dendrite. Horizontal and vertical lines in C, F indicate mean and SD, respectively. ** and *** - different at p<30.01 and p<0.001, respectively (t test); ns – not significant. All data are from at least 2 independent experiments.

Also see Figure S3 and Videos S1–S4.

To determine whether thermoreceptor trafficking is PYT-1-dependent, we examined dendritic movement of GCY-18- and GCY-23-containing particles in pyt-1 null mutants. Both the number of mobile GCY-23::GFP-containing particles and their net retrograde flux were decreased in pyt-1 mutants (Figure 2B–D, Figure S3A, Video S1, Video S2). However, mutations in pyt-1 did not alter the number or net anterograde trafficking of GCY-18::GFP-containing puncta (Figure 2E–G, Figure S3B, Video S3, Video S4). To test whether endocytosis is required for PYT-1-dependent GCY-23 trafficking, we overexpressed a dominant negative allele of RAB-5 in AFD^40–42, and measured dendritic flux of GCY-23::GFP after a temperature upshift. Inhibiting endocytosis significantly decreased retrograde flux of GCY-23::GFP, phenocopying pyt-1 mutants (Figure 2H, Figure S3C). Taken together, these results indicate that after a temperature upshift, PYT-1 is transcriptionally upregulated and localizes to endosomes at the AFD sensory ending, where it targets GCY-23 for endosomal sorting and retrograde trafficking. PYT-1 may remove a subset of GCY-23 from the microvilli membrane and/or directly sort GCY-23 into endosomes^21,22. Thus, in the absence of PYT-1, functional GCY-23 levels in the AFD microvilli are expected to increase, thereby decreasing the GCY-18:GCY-23 ratio and disrupting T*_AFD plasticity.

The PY motifs of PYT-1 are necessary for regulation of GCY-23 trafficking

The PY motifs of PY motif containing adaptor proteins are required for their endosomal localization and recruitment of target proteins^{22–24,26,29}. We noted two canonical (LPSY and PPEY) and a non-canonical (VPYY) PY motif in the predicted intracellular C-terminal domain of PYT-1 (Figure 3A, Figure S4A). A subset of PYT-1 orthologs in related nematodes contains three canonical PY motifs (Figure S4A), suggesting that the non-canonical motif in C. elegans PYT-1 may be functional.

Figure 3. PYT-1 regulates retrograde trafficking of GCY-23 via its PY motifs.

A) Cartoon of domain organization of C. elegans PYT-1 and related PY motif-containing adaptor proteins.

B) (Left) Representative images of wild-type PYT-1::GFP and PYT-1(3XPY)::GFP localization in the AFD sensory endings marked by the expression of myrTagRFP. The percentage of sensory endings showing PYT-1 localization to the center of the sensory ending (arrow) is indicated. Scale bar: 5 μm. (Right) Quantification of overall levels of PYT-1::GFP in the AFD sensory endings from the indicated conditions and genotypes. Quantification was performed on images without coexpression of myrTagRFP. Each dot is the measurement from a single AFD sensory ending. n=9–76. ***: different at p<0.001 (one-way ANOVA with Tukey’s correction).

C) Quantification of the bias towards anterograde versus retrograde movement of particles in AFD dendrites of animals from the indicated conditions and genotypes. Vertical and horizontal lines indicate mean and SD, respectively. n=11–19. ***: different at p<0.001 (t test); ns – not significant.

D) (Left) GCaMP6s traces from AFD in animals of the indicated genotypes before and after a temperature upshift in response to a temperature ramp (green line). Thick lines and shading: average ΔG/R change (see Star Methods) and SEM, respectively. Dashed vertical lines: T*_AFD for the indicated genotypes. (Right) Quantification of T*_AFD in animals of the indicated genotypes calculated from traces at left. n=45–158. ***: different at p<0.001 (one-way ANOVA with Tukey’s correction); ns – not significant.

E) (Left) Cartoon depicting a typical thermotaxis assay after worms are shifted to 25°C for 4 hours. The trajectory of one worm is shown, and the x displacement is indicated for this animal. (Right) Quantification of x displacement from thermotaxis assays. Each dot is the average x displacement of all animals in a single assay with 15–25 animals. n = 7–9 assays. ** and ***: different between indicated at p<0.01 and 0.001, respectively (one-way ANOVA with Tukey’s correction); ns – not significant.

Each dot in B, C, D is a measurement from a single AFD neuron. Horizontal and vertical lines in B, D indicate mean and SD, respectively. All data are from at least 2 independent experiments.

Also see Figure S4.

To test whether these PY motifs are required for PYT-1 localization and/or function, we mutated all three PY motifs (pyt-1(3XPY)) at both the GFP-tagged and untagged endogenous loci of pyt-1. A GFP-tagged PYT-1(3XPY) protein was no longer enriched in endosomes but was instead upregulated and present throughout the AFD sensory endings (Figure 3B). A similar role for PY motifs on the stability of the Drosophila Commisureless (Comm) adaptor protein and localization to endosomes has been reported previously^24,26. Mutating the PY motifs also significantly reduced net retrograde flux of GCY-23 in AFD dendrites upon a temperature upshift similar to the phenotype of pyt-1(null) mutants (Figure 3C, Figure S4B). These observations indicate that the PY motifs of PYT-1 are necessary for regulating the temperature-dependent removal of GCY-23 from AFD sensory endings via retrograde dendritic trafficking.

The defective retrograde trafficking of GCY-23 in pyt-1(3XPY) mutants is expected to lower the GCY-18:GCY-23 ratio in the AFD sensory endings, resulting in T*_AFD plasticity defects. After a temperature upshift, the T*_AFD plasticity defect of pyt-1(3XPY) mutants phenocopied that of the null mutant (Figure 3D). At 15°C, T*_AFD was minimally affected in pyt-1 mutants, consistent with low pyt-1 expression at this temperature⁵ (Figure 3D). We infer that as in pyt-1(null) mutants, the failure to route GCY-23 to the retrograde trafficking pathway in pyt-1(3XPY) mutants upon a temperature upshift results in defects in T*_AFD resetting to a warmer value. We noted that the pyt-1(null) but not pyt-1(3XPY) mutants exhibited decreased response amplitude particularly when shifted to 25°C (Figure 3D). Although the reason for this discrepancy is currently unclear, it is possible that the mislocalized PYT-1(3XPY) protein retains partial ability to interact with sensory compartment-localized proteins including GCY-23 to modulate response amplitude. We previously showed that pyt-1(null) mutants also exhibit defects in their ability to shift their behavioral preference to warmer temperatures following a temperature upshift⁵. However, pyt-1(3XPY) mutants did not exhibit a thermotaxis defect under these conditions (Figure 3E). The thermotaxis defect observed in pyt-1(null) mutants may result from the combined effect of reducing both the temperature response amplitude and T*_AFD response threshold.

PY motif adaptor proteins recruit WW domain containing E3 ubiquitin ligases via their PY motifs to ubiquitinate target proteins²⁴. The WWP-1 and HECW-1 WW domain containing E3 ubiquitin ligases are strongly expressed in AFD⁴³. However, T*_AFD was unaffected in animals doubly mutant for wwp-1 and hecw-1 (Figure S4C), suggesting that PYT-1 may modulate GCY-23 trafficking via recruitment of either an alternative ubiquitin ligase or a distinct WW domain-containing protein.

PYT-1-mediated regulation of GCY-23 trafficking is necessary for temperature experience-dependent plasticity in T*_AFD

We next explored the functional implications of temperature experience- and PYT-1-dependent trafficking of GCY-23. Our results suggest that pyt-1 mutants exhibit a lower T*_AFD upon a temperature upshift due to a failure to reduce functional cold temperature-responsive GCY-23 levels, thereby lowering the GCY-18:GCY-23 ratio in the AFD sensory compartment. This model predicts that loss of gcy-23 or gcy-18 would suppress or enhance the T*_AFD plasticity defect of pyt-1 mutants, respectively.

As expected, T*_AFD in wild-type animals changed from a cooler to a warmer temperature upon a 4 hour shift from 15°C to 25°C⁵; this change was significantly decreased in pyt-1 mutants (Figure 4A–C). Consistent with the hypothesis that an increased GCY-18:GCY-23 ratio is necessary for increasing T*_AFD, in a genetic background in which GCY-18 was the only remaining thermoreceptor, T*_AFD plasticity resembled that of wild-type animals upon a shift to 25°C (Figure 4A, 4C). Moreover, no reduction in T*_AFD was observed upon introduction of the pyt-1 null mutation into this genetic background (Figure 4A, 4C). This result supports the hypothesis that loss of gcy-23 is sufficient to suppress the T*_AFD defect of pyt-1 mutants, and that PYT-1 does not target GCY-18 to modulate T*_AFD plasticity. In contrast, in a genetic background in which GCY-23 was the only remaining thermoreceptor, T*_AFD was significantly lower than in wild-type animals upon a temperature upshift (Figure 4B, 4C). The T*_AFD defect of these animals was enhanced upon additional loss of pyt-1 in agreement with functional GCY-23 levels being increased in pyt-1 mutants (Figure 4B, 4C). We noted that the loss of pyt-1 also increased the calcium response amplitude in animals expressing GCY-18 alone (Figure 4A), suggesting that in the absence of GCY-23, PYT-1 may target other proteins such as channels for endocytosis and degradation resulting in a lower response amplitude that is then alleviated upon loss of pyt-1.

Figure 4. PYT-1-dependent modulation of GCY-23 at the AFD sensory ending is necessary for T*_AFD plasticity.

A, B) (Left) Cartoon showing expression of GCY-18 (A) or GCY-23 (B) alone in gcy-23 gcy-8 or gcy-8 gcy-18 double mutants, respectively. (Middle and Right) GCaMP6s traces and quantification of T*_AFD in animals of the indicated genotypes before (0 hours) a temperature upshift (middle), and after a 4 hour temperature upshift (right). Green line: temperature ramp. Thick lines and shading: average ΔG/R change and SEM, respectively. Each dot is a measurement from a single animal. n=29–38; 2 independent experiments. Horizontal and vertical lines indicate mean and SD, respectively. ***: different at p<0.001 (one-way ANOVA with Tukey’s correction); ns – not significant.

C) Data from 4 hour temperature upshifts replotted as the effect of the pyt-1(oy160) allele relative to each control genotype. The mean of all wild-type animals in the 4 hour upshift condition is set as baseline (horizontal dotted line). For each data point in each condition, the wild-type mean was subtracted to calculate the difference between that data point and the wild-type mean. Symbols for statistical comparisons are from the comparisons made on the raw T*_AFD values in A and B.

D) Model for the temperature-dependent reorganization of AFD sensory ending rGC content. See text for details.

In summary, our results describe a mechanism that coordinates the levels of functional thermosensory receptors in AFD sensory endings through activity-dependent transcriptional and posttranscriptional pathways to fine-tune neuronal responses. Upon a 4 hour temperature upshift from 15°C to 25°C, levels of the warm-responsive GCY-18 thermoreceptor increase via both upregulated expression and anterograde trafficking. A temperature upshift also transcriptionally upregulates the PY motif containing PYT-1 adaptor protein which localizes to endosomes in the AFD sensory compartment. PYT-1 then recruits the cool-responsive GCY-23 thermoreceptor to endosomes for endocytosis and retrograde trafficking to the soma likely for degradation. The resulting increase of the GCY-18:GCY-23 ratio at the sensory endings drives T*_AFD towards warmer values (summarized in Figure 4D).

Modulation of receptor levels at the membrane provides a simple mechanism by which to alter cellular responses. Consequently, receptors are subject to multiple modes of regulation. Expression levels of receptors from different subfamilies are regulated by signaling and cellular activity^5,44–48, receptor and channel functions are modified via posttranslational modifications such as phosphorylation^38,49–51, and protein levels are altered by regulated internalization and recycling or degradation^52–54. While gcy-18 is under stimulus-dependent transcriptional control^5,20, GCY-23 appears to be regulated via selective endocytosis directed by PYT-1. Despite 68% amino acid identity in their intracellular domains, PYT-1 specifically targets GCY-23 via mechanisms that remain to be determined. Conceptually and mechanistically, stimulus-dependent modulation of GCY-23 via PYT-1 shares similarities with regulation of the axon guidance receptor Robo by the PY motif-containing adaptors Commissureless (Comm) in Drosophila, and Ndfip1 and PRRG4 in vertebrates, as well as with modulation of amino acid transporters by PY motif-containing arrestin-like proteins in yeast^22–27,55. In each case, surface receptor levels are regulated by adaptor proteins (e.g. also see^56,57) whose expression is under tight spatiotemporal control, and which share little sequence homology beyond PY motifs. Tuning of receptor availability via adaptors which are themselves subject to regulation may provide an additional layer of control to fine-tune cellular properties in response to changing ligand concentrations or environmental stimuli over different timescales.

Resource Availability

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Piali Sengupta (sengupta@brandeis.edu).

Materials Availability

All C. elegans strains and plasmids generated in this study are available on request to the lead contact.

Data and Code Availability

• Data for all Figures are uploaded to https://doi.org/10.6084/m9.figshare.30822701.

• Code for analysis of calcium imaging and fluorescence measurement data DOIs are listed in the Key Resources Table.

• Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.

KEY RESOURCES TABLE.

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Bacterial and virus strains
E. coli	Caenorhabditis	WormBase: OP50;
	Genetics Center (CGC)	WormBase: WBStrain00041969
Chemicals, peptides, and recombinant proteins
Tetramisole Hydrochloride	Sigma-Aldrich	Cat# L9756
Alt-R S.p. Cas9 Nuclease V3	IDT	Cat# 1081059
Alt-R CRISPR-Cas9 tracrRNA	IDT	Cat# 1072532
Deposited data
Raw data	This study	https://doi.org/10.6084/m9.figshare.30822701
Experimental models: Organisms/strains
C. elegans: Strain N2	Caenorhabditis	WormBase: N2;
	Genetics Center (CGC)	WormBase: WBStrain00000001
gcy-18(oy165[gcy-18::gfp])	Harris et al.5	PY12303
gcy-18(oy165[gcy-18::gfp]); pyt-1(oy160)	This study	PY12355
gcy-23(oy200[gcy-23::gfp])	This study	PY12361
gcy-23(oy200[gcy-23::gfp]); pyt-1(oy160)	This study	PY12363
gcy-23(oy186[gcy-23::tagRfp]);pyt-1(oy169[pyt-1::gfp])	This study	NHS22
pyt-1(oy169[pyt-1::gfp]);oyEx756[ttx-1p::tagRfp::rab-5 + unc-122p::tagRfp]	This study	NHS23
gcy-23(oy200[gcy-23::gfp]);oyEx756[ttx-1p::tagRfp::rab-5 + unc-122p::tagRfp]	This study	PY12364
gcy-23(oy200[gcy-23::gfp]); pyt-1(oy160);oyEx756[ttx-1p::tagRfp::rab-5 + unc-122p::tagRfp]	This study	NHS24
gcy-23(oy186[gcy-23::tagRfp])	This study	PY12353
gcy-23(oy186[gcy-23::tagRfp]); pyt-1(oy160)	This study	PY13003
gcy-23(oy200[gcy-23::gfp]); oyEx784[gcy-8p::rab-5(S33N) + unc-122p::tagRfp]	This study	PY13001
pyt-1(oy169[pyt-1::gfp])	Harris et al.5	PY12306
pyt-1(syb9663 3XPY)	This study	PHX9663, SunyBiotech
pyt-1(oy169 oy217[3XPY])	This study	PY13002
pyt-1(oy169[pyt-1::gfp]); Ex[gcy-8p::myrtagRfp + unc-122p::tagRfp]	This study	NHS33
pyt-1(syb9663 3XPY); Ex[gcy-8p::myrtagRfp + unc-122p::tagRfp]	This study	NHS34
wyIs629[gcy-8p::GCaMP6s + gcy-8p::mCherry + unc-122p::gfp]	Hawk et al.16	DCR3055
pyt-1(oy160); wyIs629[gcy-8p::GCaMP6s + gcy-8p::mCherry + unc-122p::gfp]	Harris et al.5	PY12332
pyt-1(syb9663); wyIs629[gcy-8p::GCaMP6s + gcy-8p::mCherry + unc-122p::gfp]	This study	PY13000
wwp-1(oy187); hecw-1(gk3102); wyIs629[gcy-8p::GCaMP6s + gcy-8p::mCherry + unc-122p::gfp]	This study	NHS29
gcy-23(oy200[gcy-23::gfp]); pyt-1(syb9663)	This study	NHS25
gcy-23(oy150) gcy-8(oy44); wyIs629[gcy-8p::GCaMP6s + gcy-8p::mCherry + unc-122p::gfp]	Hill et al.1	PY12104
gcy-23(oy150) gcy-8(oy44); pyt-1(oy160);wyIs629[gcy-8p::GCaMP6s + gcy-8p::mCherry + unc-122p::gfp]	This study	PY12359
gcy-8(oy44) gcy-18(nj38); wyIs629[gcy-8p::GCaMP6s + gcy-8p::mCherry + unc-122p::gfp]	Hill et al.1	PY12102
gcy-8(oy44) gcy-18(nj38); pyt-1(oy160); wyIs629[gcy-8p::GCaMP6s + gcy-8p::mCherry + unc-122p::gfp]	This study	PY12360
Recombinant DNA
ttx-1p::tagRfp::rab-5	This study	PSAB1393
gcy-8p::rab-5(S33N)	This study	PSAB1394
	Gift from Ashish K.	PSAB1315
gcy-8p::myrtagRfp	Maurya
Software and algorithms
Matlab 2024b	Mathworks	https://www.mathworks.com/products/matlab.html
MetaMorph	Molecular Devices	https://www.moleculardevices.com/products/cellular-imagingsystems/acquisitionand-analysissoftware/metamorph-microscopy
FIJI/Imagej	https://imagej.net/people/	https://imagej.net/software/fiji/downloads
Slidebook 6	Intelligent-Imaging	https://www.intelligentimaging.com/slidebook
ZEN	Zeiss	https://www.zeiss.com/microscopy/en/products/software/zeiss-zen.html
Prism 6	Graphpad	https://www.graphpad.com/scientificsoftware/prism/
NIH COBALT	Papadopoulos and	https://www.ncbi.nl
	Agarwala58	https://www.ncbi.nlm.nih.gov/tools/cobalt/re_cobalt.cgi
Jalview 2.11.4.1	Waterhouse et al.59	https://www.jalview.org/
Worm tracking and analysis software	Luo et al62	https://github.com/samuellab/MAGATAnalyzer
Software for analysis of calcium imaging	This study	https://github.com/SenguptaLab/Harris_Dutta_PYT

STAR METHODS

EXPERIMENTAL MODEL AND STUDY PARTICIPANT DETAILS

C. elegans strains and genetics

All strains used in this work are listed in the Key Resources Table. Worms were grown at 20°C on nematode growth media (NGM) plates seeded with E. coli OP50. The wild-type strain used was C. elegans variety Bristol strain N2. Strains containing multiple mutations or gene edits were generated using standard genetic manipulations, and verified by PCR-based genotyping and/or sequencing. All experiments were performed on day one adult hermaphrodite worms.

METHOD DETAILS

Generation of transgenic strains

Transgenesis was performed using experimental plasmids at 2–10 ng/μl and coinjection marker plasmids at 50 ng/μl.

Plasmid construction

Promoter sequences and cDNAs were amplified from plasmids or a C. elegans cDNA library generated from a population of mixed stage animals, respectively⁵⁸. Plasmids were constructed using standard restriction enzyme cloning or Gibson assembly (New England BioLabs). The dominant negative RAB-5(S33N) mutation⁴⁰ was introduced into the overhangs during Gibson assembly.

CRISPR/Cas9-based genome engineering

All crRNAs, tracrRNAs and Cas9 protein were obtained from Integrated DNA technologies (IDT). Injection mixes were prepared generally according to published protocols⁵⁹. All gfp and tagRfp insertions were made immediately before the stop codon of each relevant gene in the N2 genetic background.

Reporter-tagged alleles: gcy-23(oy186[gcy-23::tagRfp]:

A donor plasmid was created by Gibson assembly of the tagRfp sequence flanked by ~1 kb homology arms 5’ and 3’ of the gcy-23 stop codon, respectively, and insertion into pMC10 (gift of M. Colosimo). The tagRfp sequence was inserted using a crRNA (5’-CAACAAAATTTCTCACAGCT-3’) and a tagRfp donor with ~1 kb homology arms amplified from the donor plasmid.

gcy-23(oy200[gcy-23::gfp]:

A donor plasmid was created by Gibson assembly of the gfp sequence flanked by ~1 kb homology arms 5’ and 3’ of the gcy-23 stop codon, respectively, and insertion into pMC10 (gift of M. Colosimo). The gfp sequence was inserted using a crRNA (5’-CAACAAAATTTCTCACAGCT-3’) and a gfp donor with ~1 kb homology arms amplified from the donor plasmid.

wwp-1 deletion allele:

The wwp-1(oy187) allele was generated according to published protocols⁵⁹. 2 crRNAs (5’ – CACAAATGACAGCGAAACGG - 3’ and 5’-ATGAGGGTTATACAATAATT-3’) targeting sequences upstream and downstream of the wwp-1 coding region and an ssODN donor (5’-actgactagtagtacttaacatcttcattcccacctattgtataaccctcatatttcttctcacccacac-3’) containing 35 bp of homology 5’ and 3’ to the cut sites were injected along with Cas9 protein. The injection mix contained: ssODN donor (1000 ng/μl), crRNAs (200 ng/μl each), tracrRNA (100 ng/μl), Cas9 (250 ng/μl), and co-injection marker (unc-122p::gfp (50 ng/μl)). The progeny of transgenic animals was subsequently examined for the presence of the desired deletion via sequencing.

pyt-1::gfp(oy217 3XPY):

Cas9 protein, guide RNAs and a gBlock were ordered from IDT. To make the repair template, a gBlock was subcloned into pBluescript, PCR amplified and melted before adding to the injection mix. The injection mix included the pRF4 rol-6(gf) plasmid and was injected into pyt-1(oy169) animals and animals were singled. 96 F1 progeny were singled from plates that contained rollers, plates were allowed to starve, and were then screened by PCR for the expected change.

pyt-1::gfp(oy217 3XPY) guide RNA (5’–3’):

TACATGAAACCTGAAGAAGTTGG

pyt-1::gfp(oy217 3XPY) gBlock (5’–3’):

AAAGAATAAAATTAGAACAATGTACTATATAtATGAAgCCgGAgGAgGTgGGAAAAGCTATAGGAAAACGTTTGAATCAATGTGAAAAAGGAGAGTGTTATTCAGATGTGGAGATAGAAGCcGCgGgtaagtttgaaattattttagattttcttaaagttttaatatcctcataggttaagtgaaatacgaacatttctagatcgcgtatttacaaatagttttttgtgaggcagatattatattttacagCCGCTTCAAGTGTAAATATGCCTCAAATCCTATTATCATCAGAAGAGCATGCgGCtGCtGCTTATGAACTTGAATCAGCACGTGCATCCCCAGCTGCTGCAGCTGATGACGTCATGTACTGCGATCAACTGAATCGATCATTTCAAAACTTACTATCAGCAAGAgGCGGCCGCAAAATGAGTAAAGGAGAAGAACTTTTCACTGGAGTTGTCCCAATTCTTGTTGAATTAGA

Calcium imaging

Temperature-evoked calcium responses in AFD were measured as described previously⁵. Animals were cultivated at 20°C then shifted to 15°C at the L4 stage. The next day, well-fed adults were either imaged directly following removal from the 15°C incubator (0 hours condition) or moved to 25°C for 4 hours and then imaged (4 hours condition). Animals were immobilized with 10mM tetramisole hydrochloride on 5% agarose pads on cover glass, and a second cover glass was placed on top of the specimen for imaging. The specimen was imaged under a microscope equipped with a custom Peltier temperature control system on the stage. The specimen was subjected to a linear temperature ramp at 0.05°C/s using a temperature controller (Accuthermo FTC200), an H-bridge amplifier (Accuthermo FTX700D), and a thermistor (McShane TR91-170). Videos of GCaMP6s fluorescence at the AFD sensory endings were acquired using a Zeiss 10X air objective (NA 0.3) on a Zeiss Axioskop2 Plus microscope with a Hamamatsu Orca digital camera (Hamamatsu). The transgene expressing GCaMP6s also contains gcy-8p::mCherry, and mCherry fluorescent signal was acquired in parallel with GCaMP6s signal. Metamorph software (Molecular Devices) was used to operate the microscope. Data were analyzed using custom scripts in MATLAB (Mathworks) (https://github.com/SenguptaLab/Harris_Dutta_PYT). T*_AFD was calculated as previously described⁹. In brief, ΔF/F traces were visualized in MATLAB, and the frame at which fluorescence began to steeply increase was selected. The temperature corresponding to this frame is quantified as T*_AFD. GCaMP6s ΔF traces were normalized to mCherry fluorescence to account for any differences in transgene expression between animals or genotypes, and the resulting Δ green/red (ΔG/R) traces are shown in the figures.

Microscopy

In all microscopy experiments, well-fed one day-old adult animals grown under indicated temperature conditions were immobilized with 20 mM tetramisole and mounted on 10% agarose pads on slides.

High-resolution AFD sensory ending images:

High resolution images of the localization of GCY-18, GCY-23, PYT-1, and RAB-5 at the AFD sensory endings were acquired using a Zeiss LSM 880 AiryScan or a Zeiss LSM 980 AiryScan2 Confocal system in the AiryScan configuration with a 63x oil objective (NA 1.4). For representative images shown, image resolution was enhanced via post-processing with AiryScan joint deconvolution in the Zeiss ZEN software package.

Analyses of AFD sensory ending protein abundance:

Images were acquired on a Zeiss Axio Observer with a Yokogawa CSU-X1 spinning disk confocal head (3i Marianas system) with a 100x oil objective (NA 1.4). Images were processed in ImageJ, and expression was quantified from a sum slices projected z-stack as corrected total cell fluorescence (CTCF) using the equation CTCF = Integrated Density – (Area of selected cell ROI × Mean fluorescence of a nearby background ROI).

Analyses of GCY-23::GFP protein abundance at the AFD sensory ending excluding the endosomal compartment:

Images were acquired using a Zeiss LSM 980 AiryScan2 Confocal system. Images were processed in ImageJ, and GCY-23::GFP abundance was quantified from a sum slices projected z-stack as corrected total cell fluorescence (CTCF) using the equation CTCF = Integrated Density – (Area of selected cell ROI × Mean fluorescence of a nearby background ROI). The endosomal compartment as marked by TagRFP::RAB-5 was then manually selected as an ROI; this ROI was used in the GFP channel to measure the GCY-23::GFP CTCF in the endosomal compartment. This CTCF was subtracted from the total CTCF, yielding the values quantified in Figure 1E.

Analyses of GCY-18 and GCY-23 localization within the AFD sensory ending:

Images were acquired on a Zeiss Axio Observer with a Yokogawa CSU-X1 spinning disk confocal head (3i Marianas system) with a 100x oil objective (NA 1.4). GCY protein distribution patterns in the AFD sensory endings were analyzed in ImageJ by an experimenter blinded to the strain genotype. The patterns in each sensory ending were qualitatively placed into one of three categories: 1) center empty (a semicircular region of reduced fluorescence was apparent in the center of the sensory ending), 2) center concentrated (a bright “dot” of fluorescence was apparent in the center of the sensory ending), 3) neither (the center of the sensory ending exhibited neither reduced fluorescence nor a bright concentration of fluorescence) (representative images are shown in Figure S1A). For simplicity, the percentage in category 2 is presented in the figures. For analysis of the intensity distribution across the sensory ending, a line of pixel width 3 was drawn across the sensory ending horizontally (medial to lateral or lateral to medial) at a single z slice in the center plane of the AFD sensory ending, and the plot profile function in ImageJ was used to measure the intensity distribution across this line. Intensity distributions were imported into MATLAB, and a custom script was used to generate the traces displayed Figure 1C and Figure S2B. Briefly, the intensity distribution for each sensory ending was interpolated to 30 × points, then the average and SEM of all intensity distributions was plotted.

Analyses of GCY-18 and GCY-23 dendritic puncta movements:

Videos of GCY-18::GFP and GCY-23::GFP movement in the AFD dendrite were acquired on Zeiss Axio Observer with a Yokogawa CSU-X1 spinning disk confocal head (3i Marianas system) with a 100x oil objective (NA 1.4) for 240 frames with 250 ms exposure. To generate the kymographs, a line segment of 20–25 μm in the distal region of the AFD dendrite was drawn, i.e., starting from the base of the AFD sensory ending, and extending 20–25 μm proximally towards the soma. Kymographs were generated using the Multi Kymograph Plugin (ImageJ). Anterograde, retrograde, and stationary particles were manually identified by drawing line segments over each track. The anterograde/retrograde flux was calculated as the number of anterograde/retrograde moving tracks per unit distance (length of the line segment) and per unit time.

Quantification of colocalization percentages:

For colocalization of GCY-23::GFP, GCY-23::TagRFP, TagRFP::RAB-5, GFP::RAB-5, and PYT-1::GFP, high resolution images of the sensory endings were acquired as described above. Red and green channels were merged in ImageJ and colocalization at the AFD sensory endings was quantified by an experimenter blinded to the strain genotypes.

Thermotaxis behavior

Thermotaxis assays were performed as described previously⁵. Well-fed animals were cultivated at 20°C until the L4 stage, then moved to 15°C overnight. The next day, the one day-old adult animals were moved to 25°C for four hours to induce a shift in temperature preference. The temperature gradient on the assay pad was 19–23°C with a steepness of 0.18°C/cm across a 22.5cm square NGM agar pad. The temperature of the center and sides of the pad was confirmed with a digital thermometer before each assay. The pad was cut horizontally across the middle, and two genotypes were run simultaneously with one on the top and one on the bottom half. 15–25 animals in M9 buffer were placed along the center line of each pad at the start of the assay. Animal movement was monitored at 2 frames per second for 60 minutes using a Megapixel camera. Worms were tracked using custom LabView scripts and tracks were analyzed using custom MATLAB scripts⁶⁰ (https://github.com/samuellab/MAGATAnalyzer).

QUANTIFICATION AND STATISTICAL ANALYSIS

Plots of fluorescence intensity, dendritic trafficking quantifications, and AFD thermosensory response thresholds were generated with Prism 6. Sensory ending intensity distribution traces and GCaMP fluorescence traces were generated with MATLAB (Mathworks). Example images of fluorescent tags were generated using ImageJ. Multiple sequence alignments were made using NIH COBALT⁶¹ and plots (Figure S4A) were generated with Jalview 2.11.4.0⁶². Statistical analyses were performed in Prism 6. All results shown are from at least two biologically independent experiments. Statistical test details and the number of analyzed samples are reported in each figure legend.

Supplementary Material

1. Video S1. Trafficking of GCY-23::GFP at 15°C, related to Figure 2.

Representative video showing movement of GCY-23::GFP-containing particles in the distal AFD dendrite in a wild-type and pyt-1(oy160) animal at 15°C.

2. Video S2. Trafficking of GCY-23::GFP after a temperature upshift, related to Figure 2.

Representative video showing movement of GCY-23::GFP-containing particles in the distal AFD dendrite in a wild-type and pyt-1(oy160) animal following a temperature upshift from 15°C to 25°C for 4 hours.

3. Video S3. Trafficking of GCY-18::GFP at 15°C, related to Figure 2.

Representative video showing movement of GCY-18::GFP-containing particles in the distal AFD dendrite in a wild-type and pyt-1(oy160) animal at 15°C.

4. Video S4. Trafficking of GCY-18::GFP after a temperature upshift, related to Figure 2.

Representative video showing movement of GCY-18::GFP-containing particles in the distal AFD dendrite in a wild-type and pyt-1(oy160) animal following a temperature upshift from 15°C to 25°C for 4 hours.

SE: sensory ending, CB: cell body.

Highlights.

• Warm to cold thermoreceptor ratio modulates AFD sensory neuron response plasticity

• The warm thermoreceptor is transcriptionally upregulated upon a temperature upshift

• Increased expression of an adaptor targets the cold thermoreceptor to endosomes

• Coordinated transcriptional and trafficking pathways modulate thermoreceptor levels