Antagonistic regulation of salt and sugar chemotaxis plasticity by a single chemosensory neuron in Caenorhabditis elegans

Abstract

The nematode Caenorhabditis elegans memorizes various external chemicals, such as ions and odorants, during feeding. Here we find that C. elegans is attracted to the monosaccharides glucose and fructose after exposure to these monosaccharides in the presence of food; however, it avoids them without conditioning. The attraction to glucose requires a gustatory neuron called ASEL. ASEL activity increases when glucose concentration decreases. Optogenetic ASEL stimulation promotes forward movements; however, after glucose conditioning, it promotes turning, suggesting that after glucose conditioning, the behavioral output of ASEL activation switches toward glucose. We previously reported that chemotaxis toward sodium ion (Na⁺), which is sensed by ASEL, increases after Na⁺ conditioning in the presence of food. Interestingly, glucose conditioning decreases Na⁺ chemotaxis, and conversely, Na⁺ conditioning decreases glucose chemotaxis, suggesting the reciprocal inhibition of learned chemotaxis to distinct chemicals. The activation of PKC-1, an nPKC ε/η ortholog, in ASEL promotes glucose chemotaxis and decreases Na⁺ chemotaxis after glucose conditioning. Furthermore, genetic screening identified ENSA-1, an ortholog of the protein phosphatase inhibitor ARPP-16/19, which functions in parallel with PKC-1 in glucose-induced chemotactic learning toward distinct chemicals. These findings suggest that kinase–phosphatase signaling regulates the balance between learned behaviors based on glucose conditioning in ASEL, which might contribute to migration toward chemical compositions where the animals were previously fed.

Author summary

Caenorhabditis elegans responds to compounds that taste salty, bitter, sour, etc. However, its response to sweet compounds is unclear. Here, we show that C. elegans responds to glucose through a chemosensory neuron called ASEL. C. elegans avoids high concentrations of glucose and learns to approach glucose after feeding in the presence of high glucose, dependent on the action of ASEL. The ASEL neuron has been reported to respond to salt and promotes salt attraction after feeding in the presence of high salt. We find that the feeding-associated attractive responses to glucose and salt are antagonistic. When encountered with a mixture of salt and glucose during feeding, C. elegans changes its chemotactic response toward those chemicals according to the balance of each chemical in the mixture. C. elegans may memorize the concentrations of the chemical mixture during feeding and migrates to the chemical composition previously fed, which may promote opportunities for obtaining food. Furthermore, we find that kinase–phosphatase signaling, which modulates neurotransmission, in ASEL is required for chemotaxis based on information processing of salt and glucose.

Introduction

Animals sense information from complex environments and exhibit behavioral responses according to current and past situations. They receive multiple sensory cues, which are integrated in the nervous system to generate a repertoire of behaviors [1]. A model animal Caenorhabditis elegans integrates multiple external and internal cues and shows behavior appropriate to the situation using a simple nervous system and endocrine signals from peripheral organs. For example, when encountered with a repellent substance placed before food odor, C. elegans chooses to either cross the barrier of the repellent to approach food sources or retract to avoid the danger [2,3]. This decision-making behavior is dramatically altered by food deprivation. The hunger state enhances the ability to cross the barrier by insulin-like endocrine signaling from the intestine, by which the probability of obtaining food is increased [4]. As another example, C elegans distinguishes different kinds of bacterial food based on their chemical compounds, such as the odors of the bacteria. It prefers the smell of the pathogenic bacteria, Pseudomonas aeruginosa PA14, over that of nonpathogenic Escherichia coli OP50; however, it leans to avoid the smell of PA14 after training with the pathogen [5]. Pheromones can further modulate this learned pathogen avoidance through neuropeptidergic signaling that acts in the chemosensory neural circuit [6]. As demonstrated by the examples above, signal transduction pathways process multiple pieces of information and reflect them in appropriate behaviors; however, our understanding of their detailed mechanisms remains still limited.

C. elegans memorizes environmental signals such as chemical compounds and temperatures along with feeding experiences and shows behavioral responses based on the memory. Molecular mechanisms underlying these learned behavioral responses have been well studied. For example, they memorize concentrations of sodium chloride (NaCl) during feeding and are attracted toward those NaCl concentrations after feeding conditioning [7]. A pair of ASE gustatory neurons plays pivotal roles in NaCl chemotaxis plasticity. A right-sided ASE neuron, ASER, senses NaCl and promotes attraction toward NaCl concentrations presented during feeding. Levels of diacylglycerol (DAG) are altered in the ASER axon according to changes in the ambient NaCl concentration, which in turn modulate the activity of PKC-1, an nPKC ε/η ortholog. Worms migrate to the NaCl concentration at which they were fed through DAG–PKC-1 signaling dependent changes in neurotransmission from ASER [8–10]. A left-sided ASEL neuron also contributes to NaCl chemotaxis. The ASEL neuron senses Na⁺ and promotes chemotaxis toward Na⁺ only after feeding in the presence of high NaCl concentrations; however, the mechanism of Na⁺ chemotaxis plasticity is largely unknown [11]. Similarly, C. elegans memorizes the presence of odors and temperatures and increases attractive responses toward the sensory cues experienced during feeding [12–15]. It is proposed that C. elegans memorizes multiple sensory cues during feeding and approaches the environment where it was previously fed to increase the probability of obtaining food. However, it remains unclear how the multiple sensory cues that are memorized during feeding are integrated and processed to generate behaviors appropriate to the situation.

It has been unveiled how various taste substances are received in sensory organs in several organisms including mammals and C. elegans. In mammalian taste buds, Receptor (type II) taste bud cells respond to a single taste stimulus from sweet, bitter, and umami tastes, whereas Presynaptic (type III) cells respond to all tastes, including sour and salty [16]. In C. elegans, taste compounds are mainly received by the sensory cilia of chemosensory neurons in the amphid sensory organ [17]. ASE, ASH, and ASK sensory neurons have been reported to respond to salty taste NaCl, bitter, such as that from plant alkaloids, sour (acidic pH) and amino acids [18–21]. So far, it is unclear whether and how C. elegans responds to sweet substances, such as mono- and disaccharides, the sensing of which is beneficial for survival and has been extensively studied in mammals and insects [22].

Here, we show that C. elegans responds to monosaccharides, glucose and fructose. Although naive worms avoid these monosaccharides, they are attracted toward the monosaccharides after feeding conditioning in the presence of high concentrations of monosaccharides. The ASEL neuron responds to glucose and is required for glucose attraction after glucose conditioning. Similar to the action of ASER in NaCl chemotaxis plasticity, ASEL is activated when glucose concentration decreases and the behavioral output from ASEL is dramatically changed by glucose conditioning to promote high-glucose migration. We also examined the interaction between information of sugar (glucose) and salt (NaCl), which are both processed by the same sensory neuron ASEL, in chemotaxis plasticity. Interestingly, glucose conditioning reduced Na⁺ attraction. Conversely, NaCl conditioning reduced glucose attraction, suggesting that exposures to glucose and Na⁺ during conditioning have opposite effects on chemotactic responses toward those chemicals. Furthermore, we investigated the mechanism of chemotaxis regulation for glucose and Na⁺ in different directions after glucose conditioning. This mechanism might underlie a survival strategy by which animals migrate toward precise locations at which they were previously fed using memories of chemical compositions.

Results

C. elegans is attracted toward glucose and fructose after feeding conditioning in the presence of the monosaccharides

A study reported that C. elegans is attracted to a region containing 75 mM glucose on an agar plate [23]. To test the response of C. elegans to glucose, we prepared an agar plate with a glucose gradient (S1A Fig) and examined chemotaxis. Because the worms learn to be attracted to NaCl after feeding conditioning in the presence of high concentrations of NaCl [7] (Fig 1A first block), we examined glucose chemotaxis after feeding conditioning in the presence of glucose. Worms were attracted to glucose after feeding conditioning with 50 or 100 mM glucose for 5 h (Fig 1A second block, Fig 1B first block). By contrast, naive worms showed significant avoidance of glucose (Fig 1A second block). The significant attraction to glucose was not observed after starvation conditioning in the presence of glucose for 5 h, suggesting that the learned attraction to glucose requires both exposure to glucose and feeding experience during conditioning (Fig 1B second block). Similarly, the worms showed significant attraction to fructose after fructose conditioning in the presence of food and significant avoidance of fructose without fructose conditioning (Fig 1A third block). By contrast, the worms showed no substantial response to the disaccharide sucrose after sucrose conditioning in the presence of food (Fig 1A fourth block). They showed significant attraction toward glucose and fructose after fructose and glucose conditioning, respectively, in the presence of food (Fig 1C). These data suggest that C. elegans is attracted toward the monosaccharides glucose and fructose after feeding conditioning in the presence of the monosaccharides. Hereafter, glucose was mainly used as a chemical cue to study the worms’ response to monosaccharides.

Fig 1. C. elegans memorizes the presence of monosaccharides during feeding.

(A) After feeding conditioning with (“conditioned”) or without (“naïve”) 100 mM of chemical indicated on the x-axis, chemotaxis to the chemical used for conditioning was tested. A chemotaxis index was determined according to the following equation: chemotaxis index = (N_A − N_B) / (N_all − N_C), where N_A and N_B are the number of worms in areas of high and low concentrations of chemical cues, respectively, N_all is the total number of worms on a test plate, and N_C is the number of worms in the area around the starting position. Positive and negative chemotaxis indices mean migration toward high and low concentrations of chemicals, respectively (see also S1A Fig). Each blue or black dot represents a chemotaxis index calculated in each chemotaxis assay after conditioning with or without the chemicals indicated on the x-axis, respectively. n = 6–8 assays. (B) After conditioning with several concentrations of glucose in the presence (“Food (+)”) or absence (“Food (−)”) of food, chemotaxis to glucose was tested. n = 4 assays. (C) After conditioning with fructose (Fru) or glucose (Glc), chemotaxis to glucose (Glc) or fructose (Fru) was tested, respectively. n = 6 assays. Bars represent mean values; error bars represent SEM. One-sample, two-tailed t-test against zero value with Bonferroni correction: *P < 0.05, **P < 0.01.

A left-sided ASE gustatory neuron, ASEL, responds to glucose and is required for glucose attraction after glucose conditioning

We explored chemosensory neurons required for glucose chemotaxis in C. elegans. We used the dyf-11 mutant, in which sensation to water-soluble chemicals is largely disrupted due to sensory cilium defects. These defects can be recovered in a cell-type-specific manner by the expression of dyf-11(+) cDNA in restricted cell types [7,24]. The dyf-11 mutants showed no significant chemotactic response to glucose even after glucose conditioning. We performed cell-specific rescue experiments in the glucose chemotaxis defect of the dyf-11 mutant to explore chemosensory neurons that function in glucose chemotaxis. C. elegans has 11 types of chemosensory neurons, the ciliated endings of which are exposed to the outside and can directly sense external chemicals in the amphid, inner labial and phasmid sensory organs [17]. Among those chemosensory neurons, we focused on the ASE classes of chemosensory neuron in glucose chemotaxis, because ASE neurons play major roles in chemotaxis to various water-soluble chemicals, such as NaCl, cAMP and biotin [18,25], and we intended to examine a possible interaction between multiple chemical cues received in a particular class of chemosensory neurons. dyf-11 cDNA was expressed in either ASEL or ASER neurons, or 9 classes of amphid and phasmid chemosensory neurons other than the ASE neuron. The glucose chemotaxis defects were significantly recovered by dyf-11 expression in chemosensory neurons other than ASE or only in ASEL but not in ASER (Fig 2A). These data suggest that multiple chemosensory neurons, including ASEL, could regulate glucose chemotaxis. To test the requirement of ASEL in glucose chemotaxis, we examined the effect of ASEL ablation on glucose chemotaxis. Genetic ablation of ASEL using caspase [26] caused a significant defect in glucose attraction after glucose conditioning (Fig 2B), suggesting that the ASEL neuron is required for glucose attraction after glucose conditioning. On the other hand, glucose avoidance seems to be mainly regulated by sensory neurons other than ASEL, because ASEL ablation had no substantial effect on glucose avoidance without conditioning (Fig 2B).

Fig 2. The ASEL neuron functions in glucose attraction after glucose conditioning.

(A, B) After feeding conditioning with (conditioned) or without (naïve) glucose, chemotaxis to glucose was tested. Wild type without transgene and dyf-11(pe554) mutants expressing dyf-11(+) cDNA under the dyf-11 promoter, the odr-4 promoter, the gcy-5 promoter or the gcy-7 promoter, which drive expression in ciliated sensory neurons, amphid chemosensory neurons except ASE, ASER or ASEL, respectively, or without the transgene were tested (A). Wild type (“wt”) and strain OH8585 (“ASEL (−)”), in which ASEL is genetically ablated, were tested (B). One-sample, two-tailed t-test against zero value with Bonferroni correction: *P < 0.05, **P < 0.01. n = 7–8 (A) or 9 (B) assays. (C, D) Calcium responses of ASEL when glucose concentration changes after conditioning with (“conditioned”) or without (“naïve”) glucose in the presence of food. Time course of the average fluorescence intensity ratio (YFP/CFP) of YC2.60 relative to the basal ratio (R/R₀) in ASEL (C). The glucose concentration was switched from 15 to 0 mM at 50 s and then returned to 15 mM at 100 s. Quantification of calcium responses of ASEL upon decrease in glucose concentration (D). Peak fluorescence intensity ratios (R/R₀) of YC2.60 in ASEL during the 10-s period after the decrease in glucose concentration were plotted (left). Differences between the average fluorescence intensity ratios (R/R0) within 0–50 s and 50–100 s were plotted (right). Calcium responses in wild type (naïve, n = 14; conditioned, n = 13) and unc-13(e51) (conditioned, n = 12). Two-tailed Welch’s t-test: *P < 0.05. Box plots represent the median (central line), 25^th and 75^th percentiles (the box), and the smallest and largest values that are not outliers (the whiskers). Individual data points are superimposed on the box plots.

We next performed calcium imaging of the ASE neurons in response to changes in glucose concentration. ASEL and ASER are reported to receive different ions. ASEL responds to increases and ASER responds to decreases in concentrations of ions to promote attraction to those ions [26]. We found that ASEL, but not ASER, responded to a decrease in glucose concentration, consistent with the notion that ASEL plays important roles in glucose chemotaxis (Figs 2C top and S2). Calcium levels of ASEL remained high while glucose concentration was low and rapidly dropped when glucose concentration was increased. After glucose conditioning, ASEL was initially activated at the same level as that without conditioning when glucose concentration decreased (Fig 2C top and 2D left). By contrast, the average calcium level during low glucose period was significantly higher after glucose conditioning than that without conditioning (Fig 2C top and 2D right), implying differences in ASEL properties between worms with and without glucose conditioning. The ASEL neuron of an unc-13 mutant, in which synaptic transmission is largely disrupted due to exocytosis defects, responded to the decrease in glucose concentration at the same level as that in the wild type (Fig 2C bottom and 2D). The result suggests that synaptic transmission is not required for the glucose response in ASEL, although we cannot rule out the possibility that unc-13-independent communication, such as neuropeptidergic signaling, modulates the ASEL response.

Behavioral output of ASEL activation is changed after glucose conditioning

The calcium imaging experiments suggested that ASEL is activated when the glucose concentration is decreased with or without glucose conditioning (Fig 2C and 2D). We next monitored worm movements following the activation of ASEL using a multiworm-tracking system combined with an optogenetic tool [11]. Channelrhodopsin2 (ChR2) was expressed in ASEL and blue light was illuminated for 10 s to depolarize ASEL. When ASEL was activated by ChR2, the probability of forward locomotion significantly increased and reached a peak within 10 s in worms without conditioning (Figs 3A, 3B and S3, gray). The forward locomotion probability tended to decrease after switching off the blue light possibly because of hyperpolarization effect as a rebound reaction and returned to the baseline within 30 s. By contrast, a positive effect on forward locomotion probability was not observed by optogenetic ASEL activation in worms after glucose conditioning. Rather, ASEL activation caused a decreasing tendency in forward locomotion probability after glucose conditioning (Figs 3A, 3B and S3, blue). These data suggest that the behavioral output of ASEL activation is altered by glucose conditioning. We propose a model in which ASEL is activated upon decreased glucose concentration and promotes or suppresses forward movement in worms without conditioning or after glucose conditioning, respectively, thereby switching the response from negative to positive chemotaxis through glucose conditioning.

Fig 3. Behavioral outputs after glucose concentration changes are switched by glucose conditioning.

Probability changes in forward movement by optogenetic activation of ASEL (A, B). (A) After feeding conditioning with (blue traces) or without (gray traces) glucose, probabilities of forward movement were monitored in worms expressing ChR2 in ASEL on agar plates, containing 5 mM glucose. The ASEL neuron was activated by blue light illumination for 10 s (shaded in blue). To activate ChR2, all-trans retinal (ATR) was applied during conditioning. The same procedure was used without ATR as a control (S3 Fig). (B) Changes in forward probability were quantified by subtracting average forward probability before photostimulation (-5 to -1 s) from that during photostimulation (3 to 7 s). Bars represent mean values; error bars represent SEM. n = 11–15. One-sample, two-tailed t-test against zero value with Bonferroni correction: *P < 0.05, **P < 0.01. Two-tailed Welch’s t-test: ^##P < 0.01. Changes in pirouette frequency according to glucose concentration changes (C-G). After feeding conditioning with (blue traces) or without (gray traces) glucose, pirouette frequencies were monitored in the wild-type (C), the OH9019 ASEL-ablated strain (D), dyf-11(pe554) (E), dyf-11(pe554) strains expressing dyf-11(+) in ASEL (F), or ASER (G). Migration bias was quantified by using the pirouette index, defined as the difference of mean probability of pirouette between negative dC/dT domain and positive dC/dT domain. Positive and negative pirouette indices mean migration bias toward high and low glucose concentrations, respectively (H). Bars represent mean values; error bars represent SEM. n = 11–16. One-sample, two-tailed t-test against zero value with Bonferroni correction: **P < 0.01.

A previous study reported that in C. elegans, chemotaxis is achieved by changing the turning rate as a function of concentration change [27]. We monitored tracks of worms during glucose chemotaxis using a multitracking system [28]. Each track was divided into periods of smooth swimming (runs) and frequent turning (pirouettes), based on previously defined criteria, and the frequency of pirouettes during glucose chemotaxis was determined [7,27]. Pirouette frequency increased during increasing glucose concentration (dC/dT > 0) compared to decreasing glucose concentration (dC/dT < 0) in wild type worms without conditioning (Fig 3C and 3H first block, naïve). By contrast, those responses were reversed after glucose conditioning (Fig 3C and 3H first block, conditioned). These changes in pirouette responses to glucose led to avoidance or attraction of glucose in worms without conditioning or those after glucose conditioning, respectively. Both ASEL-ablation and dyf-11 mutant worms did not show significant changes in pirouette frequencies during glucose chemotaxis (Fig 3D, 3E and 3H second and third blocks). The expression of dyf-11 in ASEL, but not ASER, recovered increase in pirouette frequency in response to decreased glucose concentration after glucose conditioning (Fig 3F, 3G and 3H fourth and fifth blocks, conditioned). These data are consistent with the notion that ASEL action is required and sufficient for glucose attraction after glucose conditioning.

NaCl and glucose sensation by ASEL during feeding reduce chemotaxis to glucose and Na⁺, respectively

We previously reported that C. elegans is attracted toward Na⁺ after NaCl conditioning in the presence of food [11]. ASEL responds and drives attraction to Na⁺ in a way different from glucose: ASEL is activated by increased Na⁺ concentration and promotes forward locomotion after NaCl conditioning, thereby driving attraction to Na⁺. We examined both glucose and Na⁺ chemotaxis after exposure to a mixture of glucose and NaCl in the presence of food. Glucose conditioning in the absence of NaCl promoted glucose attraction, and interestingly, caused strong Na⁺ avoidance (Fig 4A left, first block and 4B, third bar). Glucose attraction switched to glucose avoidance in the presence of NaCl along with glucose during conditioning (Fig 4A left, blue bars). Conversely, the Na⁺ avoidance decreased and disappeared by addition of NaCl along with glucose during conditioning (Fig 4A left, red bars). After NaCl conditioning in the absence of glucose, the worms showed glucose avoidance and Na⁺ attraction; however, these responses gradually decreased by glucose addition along with NaCl during conditioning (Fig 4A right). These data suggest that sensing of NaCl and glucose during feeding, reduced glucose and Na⁺ chemotaxis, respectively, and the ratio of these chemicals during conditioning determines the direction and extent of chemotaxis to Na⁺ and glucose. To rule out that high osmolality and not the chemical properties of glucose affect Na⁺ avoidance after glucose conditioning, worms were exposed to sucrose in the presence of food, at the same concentration as glucose. Sucrose conditioning did not lead to significant Na⁺ avoidance, suggesting that high osmolality during conditioning is not sufficient for the induction of Na⁺ avoidance (Fig 4B, fifth bar). On the other hand, fructose conditioning led to significant Na⁺ avoidance, suggesting that Na⁺ avoidance is induced by sensation of either monosaccharide, glucose or fructose, in the presence of food (Fig 4B, fourth bar).

Fig 4. Conditioning with glucose and sodium ions decreases chemotaxis to sodium ions and glucose, respectively.

After feeding conditioning with the indicated chemicals, chemotaxis to glucose (A, blue) or Na⁺ (A, red, B–D) was tested. 0, 25, or 50 mM NaCl and 50 mM glucose were used for conditioning (A, left); 0, 25, or 50 mM glucose and 50 mM NaCl were used for conditioning (A, right). The osmolarity was adjusted with sucrose to be the same in the conditioning plates in each experiment (A). Conditioning was performed with or without (“control” or “cont”) 100 mM of NaCl, glucose, fructose or sucrose (B–D). The wild type was used (A and B). The wild type and dyf-11(pe554) mutants with or without transgenes expressing dyf-11(+) under the dyf-11 promoter or the ASEL-selective gcy-7 promoter were used (C). The 2-ASEL strain, OH7621, was used (D). Bars represent mean values; error bars represent SEM. n = 8–9 (A), 8 (B, D), or 8–15 (C). One-sample, two-tailed t-test against zero value with Bonferroni correction: *P < 0.05, **P < 0.01 (A-D). One-way ANOVA followed by Dunnett’s post hoc test (A): ^#P < 0.05, ^##P < 0.01.

We next examined glucose-induced Na⁺ avoidance using the sensory cilia-defective dyf-11 mutant and the dyf-11 mutant, in which dyf-11 expression, and likely cilia function, is rescued only in ASEL. The dyf-11 mutant did not show prominent Na⁺ chemotaxis even after NaCl or glucose conditioning (Fig 4C, the second block). By contrast, recovery of the sensory cilium of ASEL was sufficient for attraction and avoidance of Na⁺ after NaCl and glucose conditioning, respectively, suggesting that sensory perception by ASEL is required for two directions of Na⁺ chemotaxis after NaCl and glucose conditioning (Fig 4C, the fourth block). Unlike wild type worms and the dyf-11 mutant whose sensory cilium was recovered by the dyf-11 promoter, the dyf-11 mutant whose sensory cilium of ASEL was recovered showed strong Na⁺ attraction even without NaCl conditioning (Fig 4C), which implies that sensory perception from ciliated sensory neurons other than ASEL represses Na⁺ attraction in naive wild type worms. We noticed that a similar phenotype was observed in worms whose neuronal identity of ASER is converted to ASEL identity (2-ASEL strain) [26,29]. The 2-ASEL strain showed strong attraction to Na⁺ with or without NaCl conditioning, and glucose conditioning caused Na⁺ avoidance (Fig 4D). These data imply that ASER may function to reduce Na⁺ attraction, except after conditioning with high-NaCl in the presence of food.

PKC-1 promotes glucose attraction after glucose conditioning

DAG/PKC-1 signaling plays significant roles in chemosensory neurons, including ASER gustatory and AWC olfactory neurons, to regulate chemotaxis plasticity [7,30]. One of the functions of this signaling pathway in ASER is the promotion of chemotaxis toward high salt by increased neurotransmission from ASER to downstream interneurons [9,10]. We examined glucose chemotaxis in a loss-of-function (lf) mutant of pkc-1, pkc-1(nj3), which showed a strong defect in high-salt migration [7]. The pkc-1(nj3) mutant showed a strong defect in glucose attraction after glucose conditioning (Fig 5A, Conditioned). In the pkc-1 mutant, no significant effect was observed in glucose chemotaxis by glucose conditioning (Fig 5A, naive vs conditioned in pkc-1(lf)). These results suggest that PKC-1 is required for glucose attraction induced by glucose conditioning. On the other hand, the pkc-1 mutation promoted Na⁺ attraction without conditioning and the Na⁺ chemotaxis was not further increased by NaCl conditioning, suggesting that PKC-1 may repress Na⁺ attraction in naïve worms and reduction in PKC-1 activity may underlie Na⁺ attraction after NaCl conditioning (Fig 5B). The pkc-1 mutant showed Na⁺ attraction instead of avoidance after glucose conditioning (Fig 5B, Glucose). Thus, PKC-1 is required for Na⁺ avoidance after glucose conditioning. Next, we examined the effects of overexpression of PKC-1 in either ASEL or ASER. Expression of a gf form of PKC-1 in ASEL, but not in ASER, promoted glucose chemotaxis bias toward attraction in worms without conditioning (Fig 5C, Naive). On the other hand, the gf form of PKC-1 in ASEL or ASER promoted Na⁺ chemotaxis bias toward avoidance or attraction, respectively (Fig 5D), suggesting that PKC-1 functions in opposite directions in ASEL and ASER: increased PKC-1 action in ASEL or ASER reduces or increases Na⁺ attraction, respectively. These behavioral phenotypes caused by pkc-1(gf) expression are consistent with the hypothesis that PKC-1 activity in ASEL promotes both glucose attraction and Na⁺ avoidance after high-glucose conditioning and the previous report suggesting that PKC-1 activity in ASER promotes NaCl attraction after high-NaCl conditioning [10].

Fig 5. PKC-1 in ASEL promotes glucose attraction and Na⁺ avoidance after glucose conditioning.

(A, C) After feeding conditioning with (“conditioned”) or without (“naïve”) glucose, chemotaxis to glucose was tested. (B, D) After feeding conditioning with or without (“control”) glucose or NaCl, chemotaxis to Na⁺ was tested. The wild type and pkc-1(nj3lf) (A, B) and wild type worms with or without (“control”) transgenes expressing PKC-1(A160E, gf) in ASEL or ASER using the gcy-7 promoter or the gcy-5 promoter, respectively (C, D). The data using the wild type worms without a transgene in panel C are the same as those in Fig 2A, because these experiments were simultaneously conducted and the experimental data were split to the two panels. Bars represent mean values; error bars represent SEM. n = 6–7 (A), 6 (B), 8 (C), or 9–20 (D). One-sample, two-tailed t-test against zero value with Bonferroni correction: **P < 0.01 (A, B). two-tailed Welch’s t-test with Bonferroni correction: ^## P < 0.01 (A). One-way ANOVA followed by Dunnett’s post hoc test: ^##P < 0.01 (B), **P < 0.01 (C, D).

ENSA-1 regulates chemotaxis after glucose conditioning in parallel with PKC-1

Although the pkc-1(lf) mutation promoted Na⁺ attraction in worms without conditioning, glucose conditioning significantly decreased Na⁺ chemotaxis (Fig 5B). To further examine molecules required for Na⁺ chemotaxis plasticity after glucose conditioning, we performed forward genetic screening (see detail in the Methods section). We mutagenized the 2-ASEL strain, OH7621, and screened for mutants that showed Na⁺ attraction even after glucose conditioning. The isolated mutant JN4784 showed strong Na⁺ attraction after glucose conditioning (S4B Fig). We identified a candidate region in which the causative mutation is located by a method based on whole-genome sequencing of the backcrossed progeny of JN4784 with either mutant or wild type phenotypes (see Methods). Within the candidate region, we found a nonsense mutation (Gln25*), pe4796, in ensa-1, which encodes an ortholog of the protein phosphatase inhibitor ARPP-16/19 (S5B Fig). Introduction of the fosmid WRM065cF04, which contains ensa-1, reduced increased Na⁺ attraction of the ensa-1(pe4796) mutant (S4C Fig). Two insertion/deletion mutants of ensa-1, pe4791 and pe4792, which were obtained by using a CRISPR/Cas9-based method, showed significant Na⁺ attraction in worms without conditioning, similar to ensa-1(pe4796) (Fig 6A). Furthermore, they showed defects in glucose attraction and Na⁺ avoidance after glucose conditioning (Fig 6A and 6B, blue bars). Both ensa-1 and pkc-1 mutants showed significant attraction toward Na⁺ without conditioning (Fig 6A and 6C, black bars) and glucose conditioning significantly decreased Na⁺ chemotaxis of these mutants (Fig 6A and 6C, compare black and blue bars). These results suggest that ensa-1 is required for chemotaxis plasticity toward glucose and Na⁺, similar to pkc-1. Expression of ensa-1 cDNA only in ASEL significantly rescued both glucose attraction and Na⁺ avoidance after glucose conditioning, suggesting that ensa-1 functions in ASEL to promote glucose attraction and Na⁺ avoidance after glucose conditioning (Fig 6D).

Fig 6. ENSA-1 in ASEL promotes glucose attraction and Na⁺ avoidance after glucose conditioning.

(A-D) After feeding conditioning with or without (“control”) the indicated chemicals, chemotaxis to glucose or Na⁺ was tested. Bars represent mean values; error bars represent SEM. n = 6 (A), 8 (B), 9–13 (C), or 6–12 (D). One-sample, two-tailed t-test against zero value with Bonferroni correction: *P < 0.05, **P < 0.01. One-way ANOVA followed by Dunnett’s post hoc test: ^#P < 0.05, ^##P < 0.01 (A, C). (E) Expression patterns of ENSA-1::Venus driven by the pan-neuronal rgef-1 promoter (top) or the ASEL-selective gcy-7 promoter (bottom). Head regions of adult worms are shown. An enlarged image of the cell body of ASEL is shown on the upper left.

To examine the genetic interaction between ensa-1 and pkc-1, we tested Na⁺ chemotaxis of the pkc-1; ensa-1 double mutant under naive conditions and after conditioning with NaCl or glucose (Fig 6C). The pkc-1; ensa-1 double mutant showed no chemotactic bias after NaCl conditioning (Fig 6C right, a red bar). Together with the finding that the single mutants of pkc-1 and ensa-1 showed significant attraction to Na⁺ after NaCl conditioning (Fig 6A and 6C, red bars), these data suggest that redundant functions of PKC-1 and ENSA-1 are required for Na⁺ attraction after NaCl conditioning. On the other hand, the pkc-1; ensa-1 double mutant showed significant Na⁺ attraction both in the naive condition and after glucose conditioning, similar to the single mutants of pkc-1 and ensa-1 (Fig 6A and 6C, black and blue bars). However, unlike the single mutants, glucose conditioning had no significant effect on Na⁺ chemotaxis in the pkc-1; ensa-1 double mutant (Fig 6A and 6C, compared black and blue bars). In addition, the double mutant showed decreased Na⁺ chemotaxis in the naïve condition compared to the single mutants of pkc-1 and ensa-1 (Fig 6A and 6C, black bars, compared pkc-1(nj3); ensa-1(pe4791) with pkc-1(nj3) or ensa-1(pe4791) by one-way ANOVA followed by Dunnett’s post hoc test: P = 0.000476 (vs pkc-1), 0.00443 (vs ensa-1)). These results suggest that ENSA-1 functions in parallel with PKC-1 in the regulation of Na⁺ chemotaxis under naïve conditions and Na⁺ chemotaxis plasticity after glucose conditioning.

PKC-1 is strongly localized to the presynaptic region in the ASER neuron [8]. Unlike PKC-1, an ENSA-1::Venus fusion protein was distributed throughout the neuron, including the cell body, axon, and dendrite, implying that ENSA-1 could function with PKC-1 in the presynaptic region and other subcellular sites in ASEL after glucose conditioning (Fig 6E).

Discussion

In this study, we report that worms avoid glucose and fructose; however, they learn to be attracted to monosaccharides after cultivation with high concentrations of the monosaccharides in the presence of food (Fig 7). Although these monosaccharides are important for energy production, high concentrations of glucose are known to be harmful to animals, including C. elegans. High-glucose conditions can cause mitochondrial damage, generate oxidative stress and reduce the life span in C. elegans [31,32]. Therefore, the avoidance of high concentrations of glucose would be beneficial for survival. On the other hand, in the natural environment, C. elegans is abundant in microbe-rich rotting plant matter, such as deposing fruits and stems [33], where high concentrations of sugar could exist. Monosaccharides may be used as chemical cues to memorize previous feeding locations, and worms are attracted to and remember these locations in the natural environment. C. elegans may flexibly alter its behavioral responses to various environmental signals, including hazardous substances, to obtain food in harsh natural environments.

Fig 7. A summary of chemotaxis plasticity induced by sugar and salt conditioning.

The ASEL neuron responds to a decrease in glucose concentration and increases the forward locomotion probability in naive animals grown without glucose, thereby promoting glucose avoidance (left, Naive). It has been reported that ASEL shows little or no response to Na⁺ in animals grown in NaCl-free conditions by calcium imaging experiments ([11]; left). After high-glucose conditioning in the presence of food, the ASEL neuron increases the turning frequency in response to decreased glucose concentration, thereby promoting glucose attraction (top right, black lines). On the other hand, although experimentally not confirmed, increased Na⁺ could activate ASEL [11], which increases the turning frequency and promotes Na⁺ avoidance after high-glucose conditioning (top right, gray lines). A phosphatase inhibitor ARPP-16/19 ortholog, ENSA-1, and an nPKCε/η ortholog, PKC-1, are required for both glucose attraction and Na⁺ avoidance in ASEL. After high-NaCl conditioning, ASEL is strongly activated by increased Na⁺ concentration and increases forward locomotion probability, thereby promoting Na⁺ attraction ([11]; bottom right, black lines). Conversely, although experimentally not confirmed, decreased glucose can activate ASEL and increase forward locomotion probability, which promotes glucose avoidance after high-NaCl conditioning (bottom right, gray lines). Therefore, the differences in direction of sensory responses upon concentration changes between glucose and Na⁺ can explain the differences in direction of chemotactic responses toward glucose and Na⁺ after conditioning. It should be noted that ENSA-1 and PKC-1 are required also for repression of Na⁺ attraction in naive conditions and promotion of Na⁺ attraction after NaCl conditioning, although their functional sites are obscure.

We found that the ASEL neuron is activated by decreased glucose concentrations and is required for glucose attraction. ASEL activation was reported to promote forward movements [11,18]. We confirmed that optogenetic stimulation of ASEL promotes forward moving probability, thereby promoting glucose avoidance in response to decreased glucose concentration. By contrast, after glucose conditioning, ASEL activation decreased forward movements, suggesting that glucose conditioning alters the action of neural circuits downstream of ASEL and switches chemotaxis from avoidance to attraction of glucose (Fig 7). This phenomenon is reminiscent of plastic changes in neural circuits downstream of ASER in chemotaxis plasticity after high-NaCl conditioning in the presence of food—high-NaCl conditioning alters excitatory synaptic transmission from ASER to AIB interneurons, thereby promoting attraction to high-NaCl [9,10]. ASEL may promote glucose attraction after high-glucose conditioning in the presence of food through a mechanism similar to that used in the ASER neural circuit. A similar mechanism acting in different neural circuits could contribute to robustness of food-seeking behavior based on food-associated learning to multiple chemicals, such as salts and monosaccharides.

In mammals, sugars, including glucose and fructose, are received in the taste bud as sweet tastes. In addition, glucose is received by sweet taste receptors expressed in the intestine and pancreatic β-cells to control plasma glucose levels. Glucose sensing in the intestine promotes glucose absorption into the intestine dependent on glucagon-like peptide secretion and subsequent intestinal neuronal activation, which prevents hyperglycemia when a glucose-rich meal is ingested [34]. Glucose sensing in pancreatic β-cells promotes calcium signals followed by insulin secretion to reduce blood glucose levels [35]. Biochemical and fluorescent imaging experiments have revealed that ambient glucose is incorporated into the body of C. elegans [32,36]. Thus, ambient glucose may be received in peripheral organs also in C. elegans. In this paper, we have shown that the ASEL neuron responds to glucose and plays important roles in glucose chemotaxis. However, we cannot rule out the possibility that other chemosensory neurons and/or peripheral organs sense glucose and are involved in the regulation of glucose conditioning-induced chemotaxis. As seen in the examples of behavioral control based on multiple information processing introduced at the beginning of the paper [4,6], neuropeptidergic signaling in the nervous system or inter-tissue interaction may be involved in feeding experience-dependent glucose chemotaxis.

The ASEL neuron has been shown to respond to several ions, including sodium, lithium, and magnesium [26]. ASEL responds to increase in concentrations of these ions. We have reported that ASEL promotes attraction to Na⁺ after conditioning with Na⁺ in the presence of food [11]. In this study, we observed the negative effects of Na⁺ and glucose conditioning on chemotaxis to glucose and Na⁺, respectively. In addition to attractive responses to chemicals exposed during feeding conditioning, aversion of chemicals different from those exposed to during conditioning may promote food-seeking behavior by increasing accuracy to arrive at the chemical composition at conditioning. After glucose conditioning, optogenetic ASEL activation caused reduced forward moving probability. Under the circumstances, decrease in glucose concentrations, which activates ASEL, will be followed by reduced forward movement, thereby glucose attraction can be promoted after glucose conditioning. Given that ASEL is activated by an increase in Na⁺ concentration [11], it is expected that ASEL activation after increasing Na⁺ concentration would reduce forward movement, thereby promoting aversion of Na⁺ after glucose conditioning. Therefore, activation of ASEL to concentration changes of glucose and Na⁺ in an opposite direction may underlie behavioral responses to those chemicals in the opposite direction after glucose conditioning (Fig 7). C. elegans AWC olfactory neurons were reported to change the response patterns to bacterially produced medium-chain alcohols in the context presenting more attractive odorants sensed by AWC, thereby switching from attraction to avoidance of medium-chain alcohols [37]. The chemosensory neurons of C. elegans can drive multiple sensory outputs according to the situation to possibly increase the chance of survival.

DAG-PKC signaling has been identified as an important signaling pathway that regulates the direction and strength of taxis responses to temperature, salt, and odor in C. elegans [7,30,38]. PKC-1 promotes excitatory synaptic transmission from ASER to AIB interneurons to promote high-NaCl attraction after high-NaCl conditioning [9,10]. Moreover, it promotes glutamatergic neurotransmission from AWC olfactory neurons to promote odorant attraction [39,40]. Here, we report that chemotaxis plasticity after glucose conditioning depends on PKC-1 in ASEL. Although the activities of interneurons downstream of ASEL warrant further investigation, glucose conditioning may alter properties of neurotransmission from ASEL to downstream interneurons through DAG-PKC signaling in ASEL similar to that in ASER after high-NaCl conditioning. Our forward genetic screening approach identified ENSA-1 (ARPP-16/19 ortholog), which regulates chemotaxis after glucose conditioning in parallel with PKC-1. ARPP-16/19 acts as an inhibitor of protein phosphatase 2A, PP2A, and is regulated by phosphorylation through MAST3 and PKA kinases [41]. Interestingly, KIN-4, the MAST3 kinase ortholog, regulates thermotaxis plasticity by regulating synaptic transmission from AFD thermosensory neurons to AIY interneurons [42]. ENSA-1, which is distributed throughout the ASEL neuron, could interact with a variety of components of signaling pathways. It will be interesting to study how molecular signaling through PKC-1 and ENSA-1 regulates chemotaxis after glucose conditioning in ASEL (Fig 7).

Materials and methods

Maintenance of C. elegans strains

C. elegans Bristol strain N2 was used as the wild type. The C. elegans strains were grown and maintained on NGM plates [43] seeded with Escherichia coli strain NA22 at 20°C, except for in calcium imaging and optogenetics experiments, in which E. coli strain OP50 was used as a food source. When C. elegans is grown with E. coli NA22 as a food source, the intestine has a tendency to enlarge, pushing the intestine forward and bringing it closer to the neuronal population. In this case, autofluorescence of the intestine can affect the fluorescence imaging. To avoid this, we used E. coli OP50 for calcium imaging. We used the NA22 strain for the behavioral experiments because NA22 has a higher growth rate than OP50, and more worms can be obtained by feeding with the NA22 food source. C. elegans strains used in this study are listed in S1 Table.

DNA constructs and transgenesis

We used GATEWAY cloning (ThermoFisher) to generate plasmids for YC2.60, ensa-1, or ensa-1::venus expression. We generated the destination vector carrying the sequence of YC2.60, ensa-1, or ensa-1::venus, namely pDEST-YC2.60, pDEST-ensa-1::sl2::cfp, or pDEST-ensa-1::Venus::sl2::cfp, respectively, and then inserted the rgef-1, gcy-5, or gcy-7 promoter sequence upstream of each gene through the LR reaction with the entry vector carrying each promoter. Venus was fused to the C-terminal region of ensa-1 cDNA just before the stop codon by a PCR-based method. DNA constructs were injected at concentrations of 10–30 ng/μL with a co-injection marker, i.e., lin-44p::mCherry (20 ng/μL), unc-122p::mCherry (20 ng/μL), or myo-3p::Venus (10 or 20 ng/μL), and a carrier DNA, namely pPD49.26. Injection mixtures were prepared to a final concentration of 100 ng/μL in total.

For isolating ensa-1(pe4791) and ensa-1(pe4792) insertion/deletion mutants using a CRISPR/Cas9 system [44], we injected two plasmid DNAs, including Cas9 cDNA under the rgef-1 promoter or ensa-1 sgRNA under the U6 promoter, at 30 or 50 ng/μL, respectively, with a co-injection marker, namely myo-3p::Venus (10 ng/μL), into wild type N2. We note that it has been found that Cas9 expression by the rgef-1 promoter, which is known to drive expression in the nervous system, could yield deletion mutants in the process of generating neuron-specific gene knockdown strains based on the somatic CRISPR/Cas9 system. The target sequence (AGAGCTTATGGGCAAATTGG) in ensa-1 sgRNA includes the first exon of ensa-1. We screened F1 animals harboring deletions or insertions around the ensa-1 target sequence by PCR-based genotyping, using the MultiNA microtip electrophoresis system (Shimadzu). F2 animals harboring frameshift mutations in the first exon of ensa-1 on two homologous chromosomes were isolated and mutation sites were determined.

Behavioral assays

In all behavioral assays, we handled with chemotaxis buffer, which contains 25 mM potassium phosphate (pH 6.0), 1 mM CaCl₂, and 1 mM MgSO₄. For chemotaxis assays, except for optogenetics experiments (Fig 3A and 3B) and worm-tracking experiments (Fig 3C–3H), we used circular agar plates (approx. 85 mm in diameter) with chemical gradients (S1A Fig, shown as chemotaxis-test plates). Chemical gradients on the test plates were produced as described [7,11]. For chemotaxis to sugars (d-glucose, d-fructose, and sucrose), two cylindrical 2% agar blocks, including 0 mM or 50 mM sugar in chemotaxis buffer, were placed 3 cm from the center of a 2% agar plate, including the chemotaxis buffer, in opposite directions for 23–25 h at 20°C and removed just before the assay. For chemotaxis to Na⁺, two cylindrical 2% agar blocks, including 50 mM NH₄Cl or NaCl in chemotaxis buffer, were placed 3 cm from the center of a 2% agar plate, including 50 mM NH₄Cl in chemotaxis buffer, in the opposite directions for 18–24 h at 20°C and removed just before the assay. For conditioning, adult worms were transferred to modified NGM plates, including 100 mM d-glucose, d-fructose, sucrose, or NaCl instead of ~51 mM NaCl in standard NGM, unless otherwise noted, for 4–5 h at 20°C. NaCl-free NGM plates were used for negative control (shown as “naïve” or “control”). E. coli NA22 cultured in NaCl-free LB liquid was seeded on the conditioning plates as a food source. After conditioning, worms were placed at the center of the chemotaxis-test plate and allowed to crawl for 45 min. Fifteen to two hundred worms were used in each assay. The chemotaxis index was determined according to the equation shown in S1A Fig.

For worm-tracking experiments, we used a rectangular agar plate (diameter approx. 78.5 mm × 121 mm) as a locomotion-test plate (S1B Fig). To form a linear glucose concentration gradient, a rectangular 2% agar block, including 50 mM glucose, was placed on a quartered area of a 2% agar plate for 23–25 h at 20°C and removed just before the assay. For glucose conditioning, adult worms were transferred to a modified NGM plate, including 100 mM d-glucose instead of ~51 mM NaCl in standard NGM, for 4–6 h at 20°C. E. coli NA22 cultured in NaCl-free LB liquid was seeded on the conditioning plate as a food source. After conditioning, 30–50 worms were placed at the center of the test plate and recorded at one frame per second for 10 min using a multiworm-tracking system [28]. The analysis of tracking data was performed as reported [45]. The binarized images were used to extract the center-of-gravity coordinates of the worms. The trajectories of the extracted center-of-gravity coordinates with a length of 1 mm or longer were used as the worm’s trajectories in subsequent data analysis. To determine the concentrations of glucose on the test plate, 15 cylindrical agar blocks were excised from the central region of the test plate (S1B Fig) and the glucose concentrations were measured using Amplex Red Glucose/Glucose Oxidase Assay Kit (ThermoFisher). The cubic spline curve was created based on the average of three trials and used for subsequent locomotion analyses as the glucose concentration on the test plate (S1C Fig). Based on this glucose concentration gradient, concentration changes per second (dC/dt) during locomotion were determined (S1B Fig). According to the definition in [45], periods of pirouette, consecutive sharp turns separated by less than 3.18 s, at all time points were determined and the frequencies of pirouette were calculated according to the time derivative of glucose concentration by 0.1 mM s^-1 bins. Bins with <25 data points were omitted from further analysis. Pirouette index was defined as the difference of probability of pirouette between negative dC/dT rank and positive dC/dT rank.

Optogenetics experiments were performed as described [11]. We used a circular agar plate (approx. 85 mm in diameter), including 2% agar and 5 mM d-glucose in chemotaxis buffer, as a locomotion-test plate. We used transgenic worms expressing Channelrhodopsin2 in ASEL in the mutant background lite-1, which encodes a blue light photoreceptor. For glucose conditioning, adult worms were transferred to a modified NGM plate, including 10 μM of all-trans retinal (ATR) and 0 mM or 100 mM d-glucose instead of ~51 mM NaCl in standard NGM, overnight at 20°C. E. coli OP50 cultured in NaCl-free LB liquid was seeded on the conditioning plate as a food source. The modified NGM plate without ATR was used as negative control. After conditioning was complete, ~50 worms were placed at the center of the test plate and recorded at one frame per second for 10 min using a multiworm-tracking system [28]. Blue light (peak wavelength = 470 mm; 0.2 mW/mm²) was delivered by a ring-shaped light emitting diode (CCS Inc, LDR2-90BL) after 100 s of recording without light. For each experiment, light pulses of 10 s illumination were applied five times, with 80 s intervals between each pulse. Forward movement probability was calculated as the ratio of worms during forward locomotion, i.e., except during pirouettes, sharp turns, or pauses, at each time point, and values of five trials were averaged in each experiment. At least 11 experiments were performed for each condition.

Calcium imaging

Experiments were conducted as described [7]. We used YC2.60 as a calcium indicator. For conditioning, adult worms were transferred to a modified NGM plate, including 0 mM or 100 mM d-glucose instead of ~51 mM NaCl in standard NGM, overnight at 20°C. E. coli OP50 cultured in NaCl-free LB liquid was seeded on the conditioning plate as a food source. After conditioning was complete, worms were physically immobilized in a microfluidic device and an imaging buffer (25 mM potassium phosphate [pH 6.0], 1 mM CaCl₂, 1 mM MgSO₄, and 0.02% gelatin; osmolarity was adjusted to 350 mOsm with glycerol) containing 15 mM d-glucose was delivered to the tip of the nose. The glucose concentration contained in the imaging buffer was changed from 15 to 0 mM and then recovered to 15 mM after 50 s. Fluorescence intensities of CFP and YFP were simultaneously monitored at a rate of two frames per second. Fluorescence intensities in the cell body of ASEL or ASER were analyzed with custom-made scripts using the ImageJ software. The average fluorescence intensity of the ratio of YFP to CFP between 5 and 15 s from the start of recording was set as R₀, and the fluorescence intensity ratio of YFP to CFP relative to R₀ (R/R₀) was calculated for a time series of images.

Forward genetic screening

To isolate mutants defective in chemotaxis plasticity after glucose conditioning, we used the 2-ASEL strain (OH7621) [26,46]. OH7621 showed strong attraction to Na⁺ without conditioning; however, the attraction to Na⁺ was strongly reduced after glucose conditioning (Fig 4D). OH7621 was mutagenized and the F1 offspring were divided into 21 pools of approximately 6,000 each. Approximately 6000 worms of the next generation in each pool were conditioned with 100 mM glucose for 4–6 h and subjected to a Na⁺ chemotaxis assay. Worms attracted to higher Na⁺ were collected and the next generation of worms was cultured until they reached adulthood. This process was repeated eight times (S4A Fig). Finally, we isolated 3 worms from each pool and tested Na⁺ chemotaxis of offspring of each worm after glucose conditioning. One of the isolated strains, JN4784, showed strong attraction to Na⁺ after glucose conditioning (S4B Fig).

Genetic mapping

To identify the region containing the gene involved in Na⁺ chemotaxis defect after glucose conditioning in JN4784, we performed variant discovery mapping based on whole-genome sequencing of the backcrossed offspring of JN4784 [47]. The JN4784 mutant was backcrossed with the original strain, OH7621, and then 99 lines of F1 generation progeny were isolated. Those cross progenies were cultivated across generations and their Na⁺ chemotaxis was tested after glucose conditioning two times in the F6 generation (S5A Fig). The top 20 lines of the average chemotaxis index were defined as the mutant-phenotype population and the bottom 10 lines were defined as the wt-phenotype population. The genome of each population was extracted and the genome fragment library was prepared using KAPA HyperPlus Library Preparation Kit for Illumina (NIPPON Genetics). Whole-genome sequencing was performed by multiplex paired-end Illumina Hiseq X Ten sequencing (Macrogen). Variants were extracted by mapping to the C. elegans reference genome at 104.07 average depth of coverage. After subtracting variants already present in OH7621, the variants in each mutant and wild type phenotype group were plotted according to their genomic positions. We found a candidate region with a high density of variants with a percentage close to 100% in chromosome I only in the mutant phenotype population (S5B Fig).

Confocal microscopy

Anesthetized worms at the adult stage were imaged on a 5% agarose pad. Z-series images (slice spacing of 1 μm) of the head regions were acquired with a Leica SP5 confocal microscope using a 63×/1.30 objective. Z-stack images were created with ImageJ software.

Statistical analysis

Statistical analyses were performed using R3.6.1 (http://www.R-project.org/). For multiple comparisons, ANOVA followed by Dunnett’s post hoc test or Bonferroni correction was performed as indicated in each figure caption. All behavioral and calcium imaging analyses were performed with at least four biological replicates. All results of statistical analyses are summarized in S2 Table. Raw data are available in S3 Table.

Supporting information

S1 Table. List of C. elegans strains.

(XLSX)

S2 Table. All results of statistical analyses.

(XLSX)

S3 Table. Raw data which were used for creating graphs.

(XLSX)

S1 Fig. Chemotaxis- and locomotion-test plates.

(A) Schematic of a chemotaxis-test plate used for chemotaxis assays, except for tracking analyses of worm locomotion. Fifteen to two hundred worms were used in each assay. (B) Schematic for a locomotion-test plate used for tracking analyses using a multiworm-tracking system. Circles represent areas excised for measurement of glucose concentrations. Thirty to fifty worms were used in each assay. (C) Glucose concentrations relative to positions on the y-axis in a test plate shown in B. Each curve from the three trials and the average curve smoothed by a cubic spline method are shown.

(PDF)

S2 Fig. Calcium imaging of ASER upon change in glucose concentration.

Calcium responses of ASER upon glucose concentration changes after conditioning with glucose in the presence of food. Time course of the average fluorescence intensity ratio (YFP/CFP) of YC2.60 relative to the basal ratio (R/R₀) in AESR. The glucose concentration was switched from 15 to 0 mM at 50 s and then returned to 15 mM at 100 s.

(PDF)

S3 Fig. Probability of forward movement in ChR2-expressing worms (ATR- control).

After feeding conditioning with (blue traces) or without (gray traces) glucose, probabilities of forward movement were monitored in worms expressing ChR2 in ASEL on agar plates, containing 5 mM glucose. Blue light was illuminated for 10 s (shaded in blue). As a control experiment in Fig 3A, all-trans retinal (ATR) was not applied during conditioning.

(PDF)

S4 Fig. Screening for mutants defective in Na⁺ avoidance after glucose conditioning.

(A) Schematic of screening for mutants defective in Na⁺ avoidance after glucose conditioning. See detail in the Methods section. (B, C) After feeding conditioning with (“glucose”) or without (“control”) glucose was complete, chemotaxis to Na⁺ was tested. Original (OH7621) and isolated mutant (JN4784) strains were used (B). The JN4796 strain, which was isolated by outcrossing JN4784 with the wild type N2, with (+) or without (−) the fosmid, WRM065cF04, including ensa-1 gene. See exact genotypes in S1 Table. Bars represent mean values; error bars represent SEM. n = 4–6 (B), 12 (C). One-sample, two-tailed t-test against zero value with Bonferroni correction: **P < 0.01 (B). Two-tailed Welch’s t-test: **P < 0.01 (C).

(PDF)

S5 Fig. Genetic mapping of the causative genes of the JN4784 mutant.

(A) After feeding conditioning with glucose, chemotaxis to Na⁺ was tested in JN4784, OH7621 and 99 of F6 cross progenies, which were isolated after crossing JN4784 with OH7621. Bars represent mean with SEM. n = 4 (JN4784, OH7621), 2 (cross progenies). (B) Variant frequencies in cross-progeny populations showing mutant (purple) or wild type (sky blue) phenotype. The horizontal axis represents physical positions on each chromosome.

(PDF)

Data Availability

All relevant data are within the paper and its Supporting Information files.

Funding Statement

M.T. was supported by Grants-in-Aid for Scientific Research (C) (18K06055) from the Ministry of Education, Culture, Sports, Science and Technology of Japan (MEXT). Y.I. was supported by Grants-in-Aid for Scientific Research (A) (22H00416), for Scientific Research (S) (17H06113), for Scientific Research on Innovative Area (25115010) and for Scientific Research on Innovative Area (20115001) from MEXT. The URL of Japan Society for the Promotion of Science: https://www.jsps.go.jp/english/ The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.