Introduction

Nematodes form a vast array of ecological relationships, from specialized parasite-host dependencies to nematode-microbial interactions, each one demanding exquisitely fine-tuned sets of sensory palates that span multiple modalities (1–8). Yet, the number of sensory neurons across diverse nematode species seems to be constrained (9–11). Several free-living and parasitic nematode species examined by serial section electron microscopy have nearly identical numbers of 12-13 pairs of head sensory neurons, known as the amphid neurons (11–15). How does sensory diversity arise within this neuronal constraint? When coupled with well-described neuronal anatomy, this conserved neuron count allows for detailed comparisons at the single-cell resolution and represents an opportunity to interrogate how sensory cues are processed by anatomically similar nervous systems to produce species-specific or developmental stage-dependent behavioral outputs. To identify the genetic basis for patterning different nervous systems and to understand the processes that underlie evolutionary changes in adapting to different environments, several comparative model systems have been developed to promote comparisons to the well-studied nematode Caenorhabditis elegans at the genetic and cellular levels, including the predatory entomophilic nematode, Pristionchus pacificus (11, 16–19).

As expected from their association with insects in the wild, the olfactory preferences of P. pacificus are distinct from those of C. elegans and the human parasite Strongyloides stercoralis (2, 3, 20), but little is known about P. pacificus responses to water-soluble compounds. In C. elegans, the main salt receptor neuron class comprises of a bilateral pair of left and right ASE neurons (ASEL and ASER), which serve to induce an attractive locomotory response toward an increase in salt concentration (21). The gene che-1 (chemosensory defective) encodes a transcription factor that is exclusively expressed in the ASE neurons and is required for their proper differentiation, such that a che-1 mutant results in defective salt attraction (22–25).

One major role of C. elegans CHE-1 is to promote lateral asymmetry in the ASE neurons. The left and right ASE neurons asymmetrically express receptor-type guanylyl cylases (rGCs, encoded by gcy genes) (26). This finding led to the realization that the ASE neurons are lateralized, such that the left and the right ASE neurons differentially respond to distinct salt ions (27–30). This observation led in turn to the identification of a complex gene regulatory network that genetically programs the distinct sensory potentials of the left and right ASE neurons (31), which includes a miRNA, lsy-6, at the top of this gene regulatory network (31–33). However, the lsy-6 miRNA evolved selectively in the Caenorhabditis genus (34) and is absent in P. pacificus. Moreover, P. pacificus does not show an expansion of the ASEL-type and ASER-type rGCs, as C. elegans does (Hong et al., 2019). With these two genomic observations in mind, we had previously proposed that the ASE neurons are unlikely to be lateralized in P. pacificus (11).

However, we now revise this view in light of our work on mapping chemosensory responses in P. pacificus on the level of behavior and neuronal activity. We demonstrate that P. pacificus does in fact show lateralized chemosensory profiles, indicating that P. pacificus must have evolved independent means to establish ASE laterality. We also show that the tastant palate of P. pacificus is distinct from that of C. elegans, and that its dependence on ASE, as well as its terminal selector transcription factor che-1, have also diverged.

Results

P. pacificus and C. elegans differ in their behavioral responses to salts

Previous cross-species comparisons between P. pacificus and C. elegans indicated strong differences in olfaction preferences that reflect their divergent evolutionary histories and ecology (2). To identify the neurons that mediate gustation, we first compared the chemosensory profiles of these two species toward water-soluble ions. In this survey, we found ammonium salts to be the strongest attractants to wildtype P. pacificus adult hermaphrodites (NH₄Br, NH₄Cl, NH₄I), with NH₄I significantly more attractive to P. pacificus compared to C. elegans (Fig. 1). Notably, in contrast to C. elegans (35), we find that P. pacificus is only moderately and weakly attracted to NaCl and LiCl, respectively (Fig. 1). Also, P. pacificus is repulsed by acetate salts (NaAc, NH₄Ac), which induce attractive responses in C. elegans (36). We conclude that P. pacificus and C. elegans display differences in their salt preferences.

Figure 1.

Ppa che-1 shows similarities and difference to Cel che-1 in both expression and function

The C. elegans che-1 mutant (chemotaxis-defective) was originally isolated for its inability to respond to a broad panel of salt tastants (22, 35), including those described above (29, 30, 36). Che-1 was found to encode for a Zn finger transcription factor that is exclusively expressed in the ASE neuron pair (25), which through laser ablations had been found to be the main salt receptor neurons (37). che-1 was found to control the entire differentiation of the ASE neurons, including the expression of putative receptors of the GCY receptor guanylyl cyclase family (23–25, 38).

To assess whether che-1 performs a similar function in salt perception for P. pacificus as for C. elegans, we analyzed the expression and function of the P. pacificus che-1 ortholog. In a previous paper, we reported that the 5’ region of the sole Pristionchus che-1 ortholog directs reporter expression to the Ppa ASE and Ppa ASG neuron classes (11). Re-examination of our provisional cell identifications using a newly-generated che-1p::GFP strain with stronger neurite expression revealed highly elaborated finger-like dendritic endings in the more anterior amphid neuron that could unambiguously be assigned to the AFD neurons (Fig. 2A-C), prompting us to reassign expression of che-1 to ASE and AFD.

Figure 2.

We confirmed that the che-1 reporter transgene indicates the full expression of the endogenous che-1 locus by tagging the endogenous che-1 locus with an ALFA-tag (39). The che-1::ALFA animals showed staining in 2 pairs of head neurons whose position is consistent with being the ASE and AFD neurons (Fig. 2D-E). By crossing the che-1 reporter transgene into a che-1 mutant background (see below), we also found that che-1 autoregulates its own expression (Fig. 2F), as it does in C. elegans (25).

Next, we examined whether P. pacificus salt responsiveness shows similar che-1 dependence as in C. elegans. We generated two putative null alleles in the P. pacificus che-1 homolog using CRISPR/Cas9 genome engineering, through introduction of small deletions in the first exon of the gene, thereby resulting in frameshift and premature stops (Fig. 3A). Both Ppa che-1 alleles exhibited defects in attraction towards NH₄Cl and LiCl compared to wildtype (Fig. 3B). However, unlike in C. elegans, Ppa-che-1 mutants showed no detectable difference in responses to NH₄l, NaCl, and NaAc.

Figure 3.

We considered two different possibilities for the behavioral differences of Ppa che-1 and Cel che-1 mutants. P. pacificus may use sensory neurons other than ASE to sense these cues, or alternatively, Ppa-che-1 may not have the same fundamental impact on ASE function in P. pacificus as it does in C. elegans. To explore these different possibilities, we silenced che-1 expressing neurons by expressing codon-optimized HisCl channel under control of the che-1 promoter. Histamine-treated che-1p::HisCl animals showed complete loss of attraction to NH₄Br, NH₄Cl, and NH₄I. Histamine treatment did not significantly alter their repulsive response to NaAc (Fig. 3C). These findings corroborate that NaAc is sensed by neurons other than ASE (or AFD, in which the che-1 promoter also drives HisCl). Since NH₄I sensation is affected by silencing of che-1(+) neurons but is unaffected in che-1 mutants, ASE differentiation may be more greatly impacted by the silencing of ASE than by the loss of che-1.

The ASE neurons show left/right asymmetric responses to salt

To assess whether Ppa ASE neurons show the same lateralized response to salt ions as Cel ASE neurons, we generated transgenic P. pacificus lines that express RCaMP in ASE neurons and assessed calcium responses to attractive salts (Fig. 4, S1-S2). Specifically, we looked for changes in calcium levels immediately after the addition and removal of specific salts. When 250 mM NH₄Cl is administered, an ‘ON’ response is observed as calcium transiently increases in the left ASE neuron (ASEL). In contrast, an ‘OFF’ response was observed as calcium sharply dips in the right ASE neuron (ASER) before quickly returning to baseline when the salt was removed, which suggests hyperpolarization of this neuron (Fig. 4A-B)(29). However, when presented with a 10-fold lower concentration of 25 mM NH₄Cl, the ‘OFF’ response completely disappeared in ASER while the ‘ON’ response became more pronounced in ASEL. The ASER responses to 250 mM and 25 mM NaCl (Fig. 4D) were very similar to the ‘OFF’ response (including hyperpolarization) observed for NH₄Cl, but the ASEL responses differ between the two salts: instead of the ‘ON’ response expected of ASEL, we observed a relatively weak ‘OFF’ response without the characteristic hyperpolarization but accompanied by an attenuated ‘bump’, a profile that we classify as an ‘OFF-2’ response (Fig. 4C). Finally, we examined the response to NH₄l and found that it also elicited laterally asymmetric responses, but yet again distinct from the responses to both NH₄Cl and NaCl, with ASEL showing a strong ‘ON’ response, and ASER showing an ‘ON-OFF’ biphasic response to 250 mM NH₄l (Fig. 4E-F)(40, 41). Interestingly, the ASER exhibited an ‘OFF’-only response to 25 mM NH₄l, which was not observed for the same concentration of NH₄Cl and NaCl, and thus may reflect the higher response to NH₄l than to NH₄Cl and NaCl in the behavior assays. Altogether, P. pacificus ASE neurons clearly show left/right asymmetric responses to salt attractants and these asymmetric responses show similarities and differences to the C. elegans ASE taste neurons (see Discussion).

Figure 4.

The AFD neurons also respond to salts in P. pacificus

The RCaMP line that we used to assess calcium responses in ASE is also expressed in AFD, allowing us to simultaneously examine the calcium responses in the AFD (AM12) neurons (Fig. 5, S1, S3-S5). Surprisingly, we detected a distinctly ‘ON-OFF’ biphasic response to all 3 salt types at both concentrations. Specifically, although we observed weaker or comparable responses in AFD neurons when compared to either ASE neuron’s response towards 250 mM NH₄Cl, NaCl and NH₄l (Fig. 5A-B, 5E-F, 5I-J), the AFD responses were more robust than ASER toward 25 mM NH₄Cl, NaCl, and NH₄I (Fig. 5D, 5H, 5L). Specifically, AFD neurons responded strongly to 25 mM NaCl when neither one of the ASE neurons showed a positive response (Fig. 5G-H). Averaging the calcium transients separately by AFD left versus right did not result in significant differences in the shape of the neuronal calcium responses, with the exception of the AFDR responses to higher versus lower concentrations of NH₄I (Fig. S4-S5). We have not further pursued whether these AFD responses are a reflection of a direct perception of salt or a secondary consequence of communication of salt-perceptive neurons (like ASE) to AFD.

Figure 5.

A target of che-1, the guanylyl cyclase gcy-22.3 is required for chemotaxis

We further explored the asymmetric salt perception by the ASE neurons, which in C. elegans is largely mediated through distinct receptor-type guanylyl cyclases (rGC proteins, encoded by gcy genes), which confer salt specificity via their extracellular domains (Ortiz 2009, Smith 2013). Our previous genome survey of Ppa homologs of gcy genes has revealed patterns that made us question whether Ppa gcy genes are convincing candidates for lateralized chemotactic responses. Specifically, we noted that ASER-expressed C. elegans gcy genes and ASEL-expressed C. elegans gcy genes have only expanded in the Caenorhabditis genus (11). One outlier to this pattern is the Cel gcy-22 gene, which is expressed in ASER, but has not expanded in C. elegans (11, 42). However, this gene has duplicated several times in P. pacificus, resulting in 5 putative Ppa-gcy-22 paralogs (11). We fused the promoter of one of these paralogs, Ppa-gcy-22.3, to gfp and found that transgenic animals express GFP exclusively in ASER (Fig. 6A), identical to the C. elegans gcy-22 ortholog (42). We confirmed its expression in ASER by analyzing animals that carry both the Ppa gcy-22p::gfp reporter and the Ppa che-1p::rfp reporter, showing a unilateral overlap of these reporters in ASER (Fig. 6B).

Figure 6.

To assess whether Ppa-gcy-22.3 is a potential effector of Ppa che-1 function, we crossed the gcy-22.3 reporter into che-1(ot5012) mutants. We found expression of gcy-22.3 was eliminated (Fig. 6C), leading us to conclude that gcy-22.3 is a potential effector of che-1 function, identical to its homolog in C. elegans.

To determine whether and which of the observed salt responses is mediated by the ASER-expressing gcy-22.3, we generated a putative gcy-22.3 null mutant through CRISPR/Cas9 genome editing (2 bp complex deletion that introduces frameshift)(Fig. 7A) and examined its response to the higher salt concentration. The ‘OFF’ response to 250 mM NH₄Cl was notably abolished in the ASER neuron in the loss-of-function gcy-22.3 mutant, while the ‘ON’ response in the ASEL remained intact (Fig. 7B-C). However, the responses to 250 mM NaCl were not significantly reduced in the either the ASEL or ASER neuron in the gcy-22.3 mutant (Fig. 7D-E). Furthermore, the gcy-22.3 mutation also reduced the ‘ON’ portion of the AFD biphasic response following presentation of 250 mM NH₄Cl and NaCl (Fig. 7F-G). Our findings suggest GCY-22.3 is possibly a P. pacificus salt receptor that can affect the physiological responses to salt in the ASER gustatory neuron, as well as other gustatory neurons that do not express this receptor (AFDs).

Figure 7.

Discussion

Our study has revealed several insights into the substrates of evolutionary changes between two distantly related nematode species, P. pacificus and C. elegans. We used intracellular calcium levels as our readout for neuronal activity with a genetically encoded calcium sensor in two pairs of che-1-expressing amphid sensory neurons– the first calcium imaging study in P. pacificus. We show that three neuron types (ASE left, ASE right and AFD neurons) each have distinct calcium responses to specific ion concentrations, revealing diversity at the single neuron level. We have identified the first laterally asymmetric marker in P. pacificus, gcy-22.3p::GFP, with its expression limited to ASER homolog (AM7). Our unexpected discovery that neuronal asymmetry is present in the ASE homologs between two distantly related nematode species, despite the lack of a lsy-6 homolog, suggests that functional lateralization in P. pacificus may be mediated by a different genetic pathway compared to C. elegans.

P. pacificus and C. elegans have diverged taste palates

We have shown that while C. elegans is attracted to acetate salts, P. pacificus avoids these acetates. Previous studies have shown that ammonium acetate (NH₄Ac) is sensed both as a water-soluble compound as well as a volatile odorant and is mediated by different signaling pathways (36). Like C. elegans, it is therefore likely that ammonium and acetate ions involve a different set of neurons in P. pacificus (i.e. non-che-1 expressing neurons), based on the finding that histamine-treated che-1p::HisCl animals did not significantly attenuate their repulsive response to NaAc. We find that sodium chloride is a common attractant for both P. pacificus and C. elegans, although the magnitude of response is lower than previously published results (29). This result is likely due to differences in generating the salt gradients. Nevertheless, we find that P. pacificus neurons have a distinct response to this salt when compared to those observed in C. elegans neurons. The P. pacificus ASEL neuron responds to a decrease in NaCl concentration (likely sodium), as evidenced by the ‘OFF-2’ response profile, whereas the C. elegans ASEL neuron responds to an increase in sodium concentration. However, in both nematodes, the ASER neuron responds to a decrease in chloride concentration. Additionally, the P. pacificus ASER neuron exhibits a unique ON-OFF response to ammonium iodide, whereas in C. elegans, no ON-OFF type of response is seen in the ASE neurons– the C. elegans ASER neuron responds only to a decrease in iodide concentration. Collectively, our findings highlight the divergence of the P. pacificus salt-sensory neurons compared to those observed in C. elegans.

P. pacificus ASE neurons have narrower sensitivity range

The sensitivity range of the P. pacificus ASER responses is significantly less compared to the ASEL responses, as well as to C. elegans ASER responses. Whereas the C. elegans ASER has a 40-fold sensitivity range in the ‘OFF’ response to the removal of various concentrations of NaCl (1-40 mM)(29, 43–45), the P. pacificus ASER showed the ‘OFF’ response only to 250 mM NH₄Cl but not to a 10-fold reduction concentration of NH₄Cl (25 mM). For the ASEL in contrast, the response to 25 mM was just as strong as to 250 mM NH₄Cl (10-fold) and comparable to the 8-fold concentration range observed for C. elegans ASEL towards NaCl. Alternatively, the magnitude of these sensitivity differences may be also partially due to differences among calcium indicators (i.e. GCaMP and Cameleon), but multiple P. pacificus che-1p::GCaMP strains did not exhibit sufficient basal fluorescence to allow for image tracking and direct comparison. The narrower sensitivity range in P. pacificus chemosensation was also observed for attraction to volatile odors (2), which span only 10-fold, versus up to 10,000-fold in attractive odors for C. elegans (21).

P. pacificus taste neurons exhibit a unique biphasic response

Although the left-ON and right-OFF responses are conserved in the ASE neurons in both species, the biphasic response of ASER to 250 mM NH₄I has not been observed in either of the ASE neurons towards salts. In C. elegans, hyperosmotic stimulus such as 1M glycerol, or high concentrations or long duration of CuSO₄ exposure both result in biphasic responses by the ASH neurons that sense noxious chemicals (41, 46). Olfactory neurons that mediate avoidance behavior such as the AWB neurons can also exhibit a biphasic response to the presence and removal of a high concentration isoamyl alcohol that normally elicits avoidance behavior (47). The neuron-specific response can also be dependent on the concentration of the chemical compound, since the ASER to 25 mM NH₄I was a weak ‘OFF’ rather than a biphasic one. In contrast to the salt-dependent response types by P. pacificus ASE neurons, the P. pacificus AFD neurons also show exclusively biphasic responses with various amplitudes. Further characterization will help determine if biphasic responses can also be found in other sensory neuron types in P. pacificus, specifically those neurons mediating avoidance behavior.

AFD are potentially polymodal neurons

Broadly speaking, C. elegans chemosensory neurons have been classically characterized as specialized neurons for dedicated modalities such as water-soluble chemicals (ASE), volatile odorants (AWA, AWB, AWC), noxious chemicals (ASH), pheromones (ADL, ADF, ASK), as well as light (ASJ) and temperature (AFD)(1, 37, 48– 51). Advances in multi-neuron calcium recordings have since shown that a given odor within a certain concentration range is detected by different ensembles of the 12 amphid neuron classes, including the AFD neurons (52–54). Unexpectedly, we found that the P. pacificus AFD neurons exhibit a distinctive biphasic response to all three salts tested (NH₄Cl, NaCl, and NH₄I), which differ from the C. elegans AFD calcium responses (54, 55). Moreover, the loss of the receptor gcy-22.3 reduced the AFD ‘ON’ response to NH₄Cl and NaCl, indicating ASER has only a limited contribution to the AFD response. The strong positive AFD response to 25 mM NaCl that is absent in both ASE neurons further supports the likelihood that AFD receives inputs, possibly neuropeptidergic, from other amphid neurons. The integration of thermosensation and chemosensation is important for memory-regulated behavior. In C. elegans, maximum chemotaxis indices toward NH₄Cl occurs when there is concordance between cultivation temperature and assay temperature (56). The C. elegans AFD neurons are also important for gustatory aversive learning in NaCl avoidance (45). Given the influence of environmental temperature on the P. pacificus mouth-form plasticity and the wide range of micro-climates that wild strains P. pacificus have been isolated from (57–59), temperature and taste preferences could be regulated at multiple genetic levels during crucial developmental decisions.

Changes in the gene regulatory architecture of sensory neuron specification

We found that the key regulator of Cel ASE identity, che-1, is also expressed in Ppa ASE and may play a similar role as a terminal selector in this neuron, based on its effect on ASE-mediated behavior and regulation of the Ppa-gcy-22.3 gene. However, the stronger behavioral effect of silencing of che-1 expressing neurons compared to a che-1 mutant background could either indicate that che-1 does not have as broad a role in controlling ASE differentiation in P. pacificus versus C. elegans. It is also possible that a developmental loss of ASE differentiation may result in compensatory changes in the chemosensory system during early development, as was observed in the C. elegans mating pheromone response by males (60).

Unexpectedly, we found that unlike in C. elegans, the Ppa che-1 gene is also expressed in the AFD neurons. Since we cannot record neural activity in AFD in a che-1 mutant (the che-1p::RCaMP driver fails to be expressed properly in che-1 mutants due to autoregulation), and do not yet have molecular markers for Ppa AFD neurons, we cannot assess whether che-1 affects AFD neuron differentiation.

Perhaps the most striking difference in the gene regulatory architecture of ASE neuron specification is the apparent lack of the key regulator of ASE asymmetry in P. pacificus, the miRNA lsy-6. In C. elegans, the expression of lsy-6 exclusively in ASEL is prepatterned via an early embryonic Notch signal (33) and serves to downregulate the homeodomain transcription factor cog-1 in the ASEL neuron (61). Through a network of downstream regulatory events, asymmetry of rGCs eventually becomes established (31). cog-1 and several asymmetrically expressed downstream effectors of cog-1, such as the die-1, lim-6 and fozi-1 transcription factors are conserved in P. pacificus, but whether the function of these factors in controlling P. pacificus ASE laterality is conserved remains to be determined. In this context it is intriguing to note that another prominent gene regulatory pathway that is controlled by miRNAs in C. elegans, the heterochronic pathway (let-7s, lin-4), appears to have diverged in P. pacificus as well, despite the conservation of the overall physiological readouts of this pathway (temporal patterning of cell lineage divisions)(62). It is tempting to speculate that miRNA-meditated regulatory process is particularly labile.

In conclusion, our work illustrates how comparative behavioral and genetic analyses in nematodes is a powerful strategy to uncover substrates of evolutionary change in simple nervous systems.

Materials and Methods

Nematode Strains

P. pacificus and other nematode strains were maintained at ∼20°C on NGM plates seeded with E. coli OP50 for food as described previously (63); these are derived from standard C. elegans culture methods (64). P. pacificus and other nematode strains used are listed in Table S1.

Chemotaxis assays

The assay for assessing response to salt gradients was adapted from C. elegans and P. pacificus chemotaxis assays (2, 21, 42). Overnight salt gradients were established on 10 cm chemotaxis plates containing 20 ml agar (5 mM KPO4, 1 M CaCl₂, 3% Bacto-agar, 1 mM MgSO₄) by adding 10 µl of 2.5 M salt solutions for 16 hours. Then, another 4 µl of the same salt solution or water control was added to reinforce the gradient 4 hours before the assay. Just prior to the assay, 1.5 µl of 1 M sodium azide was added to both the attractive salt (A) and the control (C) spots. Young P. pacificus adult hermaphrodite from near-saturated cultures were washed 3x with distilled water and collected by centrifuging at 2000 rpm for 2 minutes. Approximately 200 worms were loaded onto the edge of each assay plate between the gradient sources, and at least 10 combined worms have to reach the scoring arenas to be considered. At least 12 assays constituted each experimental trial, and multiple trials were conducted and averaged for each condition. The Chemotaxis Index (CI) for each end-point assay plate is defined as (A - C)/(A + C). To conduct conditional knockdowns of neurons, 5 M histamine dihydrochloride (Sigma Aldrich H7250) stock solution in sterilized deionized water (Arrowhead CA) was filter-sterilized and top plated onto the agar plates to a final histamine concentration of 25 mM histamine approximately 10 minutes before loading the worms to commence the assay. Most assays lasted ∼3.5 hours at room temperature in line with the speed of P. pacificus without food, with ∼40% of the animals participating. When using the Ppa-che-1p::optHisCl(csuEx83) strain, only animals expressing Ppa-egl-20p::RFP tail marker from the extrachromosomal array were scored. Because of P. pacificus’ strong aversion to acetate, we could not easily assess the individual contributions of salt ions in a saturated background of ammonium acetate as conventionally practiced in C. elegans studies (30).

Ppa-che-1p::HisCl strain

To make the Ppa-che-1p::optHisCl, the P. pacificus codon-optimized histamine-gated chloride channel sequence used in C. elegans (HisCl)(65, 66) was designed using (https://hallemlab.shinyapps.io/ Wild_Worm_Codon_Adapter/) and custom synthesized (Twist Bioscience), and subsequently inserted behind the che-1 promoter (3.1 kb containing the first exon and intron)(11) to create the pHC30 plasmid construct. This Ppa-che-1p::optHisCl plasmid (2 ng/µl) along with PS312 genomic DNA (80 ng/µl) and Ppa-egl-20p::RFP co-injection marker (2 ng/µl) were individually digested with HindIII and assembled as the injection mix to create csuEx83.

Ppa-che-1p::RCaMP reporter strain

To make the Ppa-che-1p::optRCaMP, we generated a transgenic worm strain expressing the codon-optimized genetically-encoded calcium indicator (GECI), jRCaMP1a, in the neurons of interest (67, 68). jRCaMP1a is an improved red GECI based on mRuby with comparable sensitivity to GCaMP6 (69). The codon-optimized RCaMP sequence was custom synthesized (Twist Bioscience), and then subcloned using the plasmid pJET1.2/blunt (Thermo Fisher Scientific) to create the pMM2 plasmid construct. Using Gibson Assembly (E2611, New England Biolab), the codon-optimized RCaMP sequence was introduced downstream of the Ppa-che-1 promoter sequence to create the pMM5 plasmid construct. Separately, we also generated a Ppa-che-1p::optGFP transcriptional reporter with codon-optimized GFP (pMM3) to enable the localization of the che-1-expressing neurons during video acquisition (70). The Ppa-che-1p::optRCaMP (2 ng/µl) and Ppa-che-1p::optGFP constructs (1 ng/µl), along with PS312 genomic DNA (80 ng/µl) and Ppa-egl-20p::RFP (1.5 ng/µl) were individually digested with HindIII and assembled as the injection mix to create csuEx93. Despite multiple attempts, we were unable to generate an equivalent che-1p::GCaMP transgenic line with sufficient basal level of GCaMP expression for a comparison to the RCaMP calcium dynamics.

CRISPR mutagenesis generated mutants

CRISPR/Cas9 mutagenesis was used to generate mutations (70, 71). crRNA and primer sequences, and induced mutations, are included in Table S3 and Fig. S6.

che-1 alleles (PPA01143)

Target crRNA, tracrRNA, and Cas9 nuclease were purchased from IDT Technologies (San Diego, CA). crRNA and tracrRNA were hydrated to 100 µM with IDT Duplex Buffer, and equal volumes of each (0.61 µl) were combined and incubated at 95°C for 5 minutes, then 25°C for 5 minutes. Cas9 protein (0.5 µl of 10 µg/µl) was added, then the mix was incubated at 37°C for 10 minutes. Ppa-egl-20p::RFP was used as a co-injection marker. To reach a final total volume of 40 µl, the Cas9-crRNA-tracrRNA complex was combined with pZH009 (Ppa-egl-20p::RFP) DNA to reach 50 ng/µl final concentration using nuclease-free water. F₁ progeny were screened for the presence of Ppa-egl-20p::RFP expression in the tail and candidate F₁‘s were sequenced to identify heterozygotes (71). ot5012 has a 4 bp insertion while ot5013 has an 8 bp complex insertion/deletion, and both mutations cause frameshift mutations and premature stop codons. Each allele was outcrossed two times to wildtype before characterization.

gcy-22.3 alleles (PPA04454)

Target crRNA, tracrRNA, and Cas9 nuclease were purchased from IDT Technologies (San Diego, CA). crRNA (RHL1400) and tracrRNA were hydrated to 100 µM with IDT Duplex Buffer, and equal volumes of each (0.61 µl) were combined and incubated at 95°C for 5 minutes, then 25°C for 5 minutes. Cas9 protein (0.5 µl of 10 µg/ µl) was added, then the mix was incubated at 37°C for 10 minutes. Ppa-egl-20p::RFP was used as a co-injection marker. To reach a final total volume of 40 µl, the Cas9-crRNA-tracrRNA complex was combined with pZH009 (Ppa-egl-20p::RFP) DNA to reach 50 ng/µl final concentration using nuclease-free water. F₁ progeny were screened for the presence of Ppa-egl-20p::RFP expression in the tail and candidate F₁‘s were sequenced to identify heterozygotes (71). csu181 has a 2 bp complex deletion, while csu182 has a 22 bp complex deletion, both mutations cause frameshifts and premature stop codons. Each allele was outcrossed two times to wildtype before characterization.

CHE-1:ALFA C-terminal tagging and immunostaining

For induction of site-specific insertions via CRISPR/Cas-9 mediated mutagenesis, target crRNA, tracrRNA, and Cas9 nuclease were treated as described above. Single-stranded DNA repair template containing the ALFA nanobody tag (RHL1551) with 35 bp of homology arms on the 5’ and 3’ sides were purchased from IDT. To minimize sequence identity, the two copies of the ALFA sequence contain silent mutations. The crRNA (RHL1396) and tracrRNA were hydrated to 100 µM with IDT Duplex Buffer, and equal volumes of each (0.61 µl) were combined and incubated at 95°C for 5 minutes, then 25°C for 5 minutes. F₁ animals expressing the co-injection marker egl-20::optRFP were lysed and checked by PCR for insertions.

Non-starved healthy cultures of ALFA-tagged CHE-1 (csu226[Ppa-che-1::2xALFA]) (from six 6 cm plate cultures) were washed with M9 and filtered (Sartorius 84g/m2, Grade 392) and processed as previously described (39). In brief, mixed stage worms were fixed overnight at 4°C on a nutator with 500 µl fixation buffer (4% paraformaldehyde in PBS). The worms were then incubated overnight at 37°C with 500 µl 4% β-mercaptoethanol dissolved in 1% Triton-X100 in 0.1M Tris pH 7.4, and digested in 200 μl of collagenase buffer with 200 units of collagenase type IV (Gibco, Thermo Fisher Scientific, Waltham, MA, USA) at 37°C for ∼3.5 hours. The partially digested worms were subsequently washed three times with 500 µl PBST. After collagenase treatment the centrifugation between the washes was done at low speed, 1000 RCF. After washes, the worms were stained in 50 µl 1% BSA in PBST with the primary antibody (1:100 FluoTag-X2 anti-ALFA-AZdye568, N1502, NanoTag Biotechnologies, Göttigen, Germany), overnight at 4°C in a nutator. Following the washes, worms were resuspended in 50 µl VectaShield mounting medium (H-1000, Vector Laboratories, USA) containing DAPI (Molecular Probes, Thermo Fisher Scientific) and gently mounted on freshly prepared 3% Noble agar pads. Images were acquired on a Leica DM6000 microscope.

Calcium Imaging

To conduct calcium imaging, worms were trapped and imaged within a microfluidic PDMS chip as previously described while delivering stimuli directly to the nose of an immobilized worm (46, 72). The microfluidic chip was connected to a programmable valve controller (ValveBank) that enables the user to toggle between “stimulant ON” and “stimulant OFF” states. Specifically, the ValveBank allows controlled switching of flow from the two outer buffer channels in the chip such that either the control solution (stimulant OFF) or the stimulant solution (stimulant ON) flows over the worm nose (Fig. S7). Buffer solution (for the outer two channels) consisted of M9 buffer with 0.1% Tween-20 and 1 µg/ml fluorescein (73, 74). The worm loading solution consisted of M9 buffer with 0.1% Tween-20 and 1.5 mM tetramisole hydrochloride to immobilize the animal. Water-soluble stimulant solutions consisted of the following water-soluble compounds dissolved in nanopure/milli-Q water: ammonium chloride (NH₄Cl), ammonium iodide (NH₄I), and sodium chloride (NaCl) at concentrations of 250 mM and 25 mM. 750 mM NaCl elicited very inconsistent responses. The water-soluble stimulant solution was made by diluting the water-soluble compound in nanopure/milli-Q water with 0.1% Tween-20. The control solution consisted of 0.1% Tween-20 nanopure/milli-Q water. Worms were exposed to the control solution, stimulant solution, and then control solution using a 60-second program: 10 seconds stimulant OFF, 20 seconds stimulant ON, and 30 seconds stimulant OFF. As a negative control, animals were exposed only to the control solution, without switching channels, for the duration of the recording. Images were captured using a Zeiss Axio Observer Z1 inverted fluorescence microscope and a pco.panda 4.2 SCMOS camera. Changes in fluorescence intensity were measured in the neurons of interest while the worm was exposed to green light. Images were processed using MetaMorph software version 7.10.5.476. Images were captured at 500 ms exposure time at 2 fps because the baseline fluorescence in Ppa-che-1p::optRCaMP worms was too dim to capture viable data using the standard 100 ms exposure time. Baseline F₀ was measured as the average background-subtracted fluorescence from the first 9 seconds of each recording and change in fluorescence intensity was calculated as dF/F = (F - F₀) / F₀, as described (74). Data were analyzed and plotted using custom scripts generated in MATLAB versions R2021a–R2024a. This code is available at https://github.com/honglabcsun/Calcium-Imaging.

For bar plot comparisons between wildtype and gcy-22.3 mutants, we calculated minimum pre-stimulus and maximum post-stimulus % dF/F values using custom scripts generated in MATLAB. Min-Max data were exported as text files, manually converted and organized into an XLSX file (Microsoft Excel Office16) and imported into Prism GraphPad software (version 10) for data visualization and statistical analysis.

Nomenclature

Throughout the results section, P. pacificus genes will be referred to without the Ppa- prefix; if necessary for comparison to another species such as C. elegans (Cel-) will the Ppa- prefix be used.

Table 1.

Supplementary figures and tables

Figure S1.

Figure S2.

Figure S3.

Figure S4.

Figure S5.

Figure S6:

Figure S7.

Table S1.

Table S2.

Table S3.

Acknowledgements

This research is funded by NIH SC1GM140970 to RLH, NIH R56MH096881 to SHC. OH is funded by the HHMI. MM and RLH contributed in conception, design and acquisition of work. VL, HRC, DLC, IMD, NRK, KTQ, and SJC contributed to data acquisition. RLH, OH, and SHC contributed to the analysis and writing of the work. The authors declare that they have no competing interests. We would to also thank I. Martinez and C. Igreja for technical assistance. All data needed to evaluate the conclusions in the paper are present in the paper and the Supplementary Materials.