1. Developmental Biology
  2. Neuroscience
Download icon

Feeding state-dependent regulation of developmental plasticity via CaMKI and neuroendocrine signaling

  1. Scott J Neal
  2. Asuka Takeishi
  3. Michael P O'Donnell
  4. JiSoo Park
  5. Myeongjin Hong
  6. Rebecca A Butcher
  7. Kyuhyung Kim  Is a corresponding author
  8. Piali Sengupta  Is a corresponding author
  1. Brandeis University, United States
  2. Daegu Gyeongbuk Institute of Science and Technology (DGIST), Republic of Korea
  3. University of Florida, United States
Research Article
  • Cited 10
  • Views 2,188
  • Annotations
Cite this article as: eLife 2015;4:e10110 doi: 10.7554/eLife.10110

Abstract

Information about nutrient availability is assessed via largely unknown mechanisms to drive developmental decisions, including the choice of Caenorhabditis elegans larvae to enter into the reproductive cycle or the dauer stage. In this study, we show that CMK-1 CaMKI regulates the dauer decision as a function of feeding state. CMK-1 acts cell-autonomously in the ASI, and non cell-autonomously in the AWC, sensory neurons to regulate expression of the growth promoting daf-7 TGF-β and daf-28 insulin-like peptide (ILP) genes, respectively. Feeding state regulates dynamic subcellular localization of CMK-1, and CMK-1-dependent expression of anti-dauer ILP genes, in AWC. A food-regulated balance between anti-dauer ILP signals from AWC and pro-dauer signals regulates neuroendocrine signaling and dauer entry; disruption of this balance in cmk-1 mutants drives inappropriate dauer formation under well-fed conditions. These results identify mechanisms by which nutrient information is integrated in a small neuronal network to modulate neuroendocrine signaling and developmental plasticity.

https://doi.org/10.7554/eLife.10110.001

eLife digest

Living organisms have the remarkable ability to adapt to changes in their external environment. For example, when conditions are favorable, the larvae of the tiny roundworm C. elegans rapidly mature into adults and reproduce. However, when faced with starvation, over-crowding or other adverse conditions, they can stop growing and enter a type of stasis called the dauer stage, which enables them to survive in harsh conditions for extended periods of time. The worms enter the dauer stage if they detect high levels of a pheromone mixture that is produced by other worms—which indicates that the local population is over-crowded. However, temperature, food availability, and other environmental cues also influence this decision.

A protein called TGF-β and other proteins called insulin-like peptides are produced by a group of sensory neurons in the worm's head. These proteins usually promote the growth of the worms by increasing the production of particular steroid hormones. However, high levels of the pheromone mixture, an inadequate supply of food and other adverse conditions decrease the expression of the genes that encode these proteins, which allows the worm to enter the dauer state. It is not clear how the worm senses food, nor how this is integrated with the information provided by the pheromones to influence this decision.

To address these questions, Neal et al. studied a variety of mutant worms that lacked proteins involved in different aspects of food sensing. The experiments show that worms missing a protein called CaMKI enter the dauer state even under conditions in which food is plentiful and normal worms continue to grow. CaMKI inhibits entry into the dauer stage by increasing the expression of the genes that encode TGF-β and the insulin-like peptides in sensory neurons in response to food.

Neal et al.'s findings reveal how CaMKI enables information about food availability to be integrated with other environmental cues to influence whether young worms enter the dauer state. Understanding how food sensing is linked to changes in hormone levels will help us appreciate why and how the availability of food has complex effects on animal biology and behavior.

https://doi.org/10.7554/eLife.10110.002

Introduction

Discrete alternate phenotypes arising from a single genotype in response to varying environmental cues is referred to as polyphenism (Michener, 1961; Mayr, 1963; Stearns, 1989). Well-described polyphenic traits include the exhibition of wings on locusts, caste hierarchy in social insects, and environmental sex determination in reptiles (Nijhout, 2003; Beldade et al., 2011; Simpson et al., 2011). In well-studied cases as in insects, it has been shown that animals integrate sensory cues during specific developmental stages to promote the expression of alternate phenotypic traits via regulation of endocrine and neuromodulatory signaling (Simpson et al., 2011; Watanabe et al., 2014). Environmental cues that trigger developmental plasticity include pheromones, temperature, mechanical stimuli, and food (Nijhout, 2003; Simpson et al., 2011). In particular, nutrient availability and quality during development is a major regulator of polyphenism in many species (Wheeler, 1986; Greene, 1989; Pfennig, 1992; Bento et al., 2010). Although the extent and adaptive value of polyphenism has been extensively discussed (Pfennig et al., 2010; Moczek et al., 2011), the underlying molecular and neuronal mechanisms that allow animals to sense and integrate signals from food and feeding-state signals in the context of other cues to regulate phenotypic plasticity are not fully understood.

Caenorhabditis elegans exhibits polyphenism in response to environmental cues sensed during a critical period in development. Shortly following hatching, C. elegans larvae assess crowding in their environment via concentrations of a complex mixture of small molecules called ascarosides (collectively referred to as dauer pheromone) produced by conspecifics (Golden and Riddle, 1982, 1984c; Jeong et al., 2005; Butcher et al., 2007; Edison, 2009; Ludewig and Schroeder, 2013). High concentrations of one or more of these chemicals promote entry of larvae into the alternate stress-resistant and long-lived dauer developmental stage, whereas under uncrowded conditions, larvae continue in the reproductive cycle (Cassada and Russell, 1975) (Figure 1A). Although pheromone is the instructive cue for dauer entry, additional cues, such as temperature and food availability, also regulate this binary decision (Golden and Riddle, 1984a, 1984b, 1984c; Ailion and Thomas, 2000) (Figure 1A). Thus, high (low) concentrations of food or low (high) temperature can efficiently inhibit (promote) pheromone-induced dauer formation, allowing animals to assess and integrate diverse sensory cues in order to make a robust developmental choice.

Figure 1 with 2 supplements see all
CMK-1 acts in the AWC and ASI/AWA neurons to inhibit dauer formation in fed animals.

(A) Simplified model of sensory inputs modulating TGF-β and insulin signaling in the regulation of the dauer decision. See text for details. (B) Quantification of dauer formation in wild-type animals in the presence of 6 μM ascr#3 and the indicated amounts of heat-killed (blue circles) or live (red circles) OP50 bacteria. Each filled circle represents one assay; n > 65 animals per assay, three independent experiments. Line represents best fit to the data. ** and *** indicate different from values using 160 μg of heat-killed bacteria at p < 0.01 and 0.001, respectively (ANOVA and Games-Howell post-hoc test). (C) Dauers formed by strains of the indicated genotypes in the presence of 6 μM ascr#3 and 80 μg live OP50. Each data point is the average of ≥3 independent experiments of >65 animals each. Errors are SEM. *, **, and *** indicate different from wild-type at p < 0.05, 0.01, and 0.001, respectively (ANOVA and Games-Howell post-hoc test). (D) Dauer formation in cmk-1 mutants grown with 6 μM ascr#3 and the indicated amounts of live OP50. Each data point is the average of ≥3 independent experiments of >65 animals each. Errors are SEM. ** and *** indicate different from corresponding values using 80 μg OP50 at p < 0.01 and 0.001, respectively (Student's t-test). (E) Dauers formed by strains of the indicated genotypes grown on plates containing 6 μM ascr#3 and 80 μg live OP50. Promoters used to drive wild-type cmk-1 cDNA expression were: cmk-1p—cmk-1 upstream regulatory sequences; ASK—sra-9p; AFD—ttx-1p; ASJ—trx-1p; ASI/AWA—gpa-4p; AWC—ceh-36Δp. Each data point is the average of ≥3 independent experiments of >65 animals each. For transgenic strains, data are averaged from 1–4 independent lines each. Errors are SEM. *, **, and *** indicate different from wild-type at p < 0.05, 0.01, and 0.001, respectively; ### indicates different from cmk-1(oy21) at p < 0.001 (ANOVA and Games-Howell post-hoc test).

https://doi.org/10.7554/eLife.10110.003

Decades of investigation have shown that environmental stimuli detected by sensory neurons modulate neuroendocrine signaling to regulate the choice of larval developmental trajectory in C. elegans (Fielenbach and Antebi, 2008). High pheromone concentrations, low food abundance and high temperature cues downregulate expression of the daf-7 TGF-β ligand and several insulin-like peptide (ILP) genes in subsets of ciliated sensory neurons in the head amphid organs (Ren et al., 1996; Schackwitz et al., 1996; Li et al., 2003; Cornils et al., 2011; Entchev et al., 2015) (Figure 1A). Downregulated TGF-β and insulin signaling in turn decrease biosynthesis of dafachronic acid steroid hormones by neuronal and non-neuronal endocrine cells (Fielenbach and Antebi, 2008). In the absence of these steroid hormones, the DAF-12 nuclear hormone receptor promotes dauer entry, whereas in the ligand-bound form, DAF-12 promotes reproductive development (Antebi et al., 1998, 2000; Ludewig et al., 2004). Ciliated chemosensory neurons required to sense a subset of ascarosides for the regulation of dauer entry have been identified (Schackwitz et al., 1996; Kim et al., 2009; McGrath et al., 2011; Park et al., 2012). However, little is known about how food is sensed, and how food signals are integrated with pheromone signals at the level of endocrine gene expression to influence the dauer decision.

Here, we identify the CMK-1 calcium/calmodulin-dependent protein kinase I (CaMKI) as a key player in the regulation of dauer formation as a function of feeding state. Expression of the daf-7 TGF-β and daf-28 ILP genes are downregulated in well-fed cmk-1 mutants, and we find that CMK-1 acts cell-autonomously in the ASI sensory neurons, and non cell-autonomously in the AWC sensory neurons, to regulate the expression of daf-7 and daf-28, respectively. We show that the subcellular localization of CMK-1 in AWC is feeding state-dependent, and that CMK-1 promotes expression of anti-dauer ILP genes in AWC. Our results indicate that a balance of CMK-1-regulated anti-dauer signals from AWC, as well as pro-dauer signals, regulates dauer entry as a function of feeding state, and that this balance is disrupted in cmk-1 mutants to inappropriately promote dauer entry under well-fed conditions. We also find that basal activity levels in AWC are enhanced upon prolonged starvation in wild-type animals and in well-fed cmk-1 mutants, and that increased activity acts in parallel with CMK-1 to antagonize dauer formation. Together, these results identify CMK-1 CaMKI as a key molecule that encodes information about nutrient availability within a sensory neuron network to regulate neuroendocrine signaling and a polyphenic developmental choice.

Results

cmk-1 mutants form dauers inappropriately in the presence of pheromone and food

To verify that pheromone-induced dauer formation in wild-type animals is suppressed by bacterial food, we quantified dauers formed in the presence of pheromone and different concentrations of non-replicative (heat-killed) as well as replicative (live) bacteria. Although no dauers were observed in the absence of added pheromone (Figure 1—figure supplement 1A; Figure 1—source data 2), >80% of wild-type larvae entered into the dauer stage on plates containing 6 μM ascr#3 (also referred to as asc-ΔC9 or C9) and l60 μg heat-killed OP50 bacteria (Figure 1B; Figure 1—source data 1). Dauer formation decreased upon increasing the amount of heat-killed bacteria and was fully suppressed by only 80 μg of live bacteria (Figure 1B). We could not reliably quantify the effects of lower concentrations of live bacteria since as reported previously, animals arrest development postembryonically when food becomes limiting (Hong et al., 1998; Fukuyama et al., 2006; Baugh, 2013). Food also inhibited dauer formation in daf-22 mutants that are unable to produce most ascarosides (Golden and Riddle, 1985; Butcher et al., 2009) (Figure 1—figure supplement 1B), indicating that under these conditions, dauer formation is not increased simply due to enhanced endogenous pheromone signaling. These observations confirm that food cues modulate pheromone-induced dauer formation.

To begin to explore the mechanisms by which food signals are integrated with pheromone cues to regulate dauer entry, we focused on genes previously implicated in different aspects of nutrient sensing and/or metabolism in C. elegans. These include the aak-1 and aak-2 AMP-activated protein kinases (Apfeld et al., 2004; Narbonne and Roy, 2009; Cunningham et al., 2012), the crh-1 CREB transcription factor and the cmk-1 CaMKI kinase (Kimura et al., 2002; Suo et al., 2006; Suo and Ishiura, 2013), the egl-4 cGMP-dependent protein kinase (Daniels et al., 2000; You et al., 2008), the eat-4 glutamate transporter (Avery, 1993; Hills et al., 2004), the tph-1 tryptophan hydroxylase and cat-2 tyrosine hydroxylase enzymes (Sawin et al., 2000; Hills et al., 2004; Suo et al., 2009; Ezcurra et al., 2011; Entchev et al., 2015), and the skn-1, hlh-30 and mxl-3 transcription factors (Paek et al., 2012; O'Rourke and Ruvkun, 2013) (Figure 1C). Specifically, we reasoned that mutations in genes essential for food signal integration would result in inappropriate entry into the dauer stage in the presence of pheromone and plentiful food but would not lead to constitutive dauer formation (dauer-constitutive or Daf-c) (Hu, 2007).

Mutations in the cmk-1 CaMKI gene fulfilled both these criteria. cmk-1(oy20) missense, as well as cmk-1(oy21) putative null, mutants consistently formed dauers in the presence of 80 μg live bacteria and 6 μM ascr#3—conditions that fully suppressed dauer formation in wild-type animals (Figure 1C). Moreover, few dauers were observed in the absence of added pheromone (Figure 1—figure supplement 1A), indicating that cmk-1 mutants do not form dauers constitutively. cmk-1 mutants also formed dauers in the presence of live food and ascr#2 (also referred to as asc-C6-MK or C6) and icas#9 (also referred to as IC-asc-C5 or C5) (Figure 1—figure supplement 1C,D), indicating that the response was not specific to a particular ascaroside. We could not reliably examine the effects of heat-killed bacteria in this assay since cmk-1 mutants exited the dauer stage prematurely under these conditions (not shown), possibly due to additional metabolic defects that we have not explored further in this study. Increasing the amount of live food to 320 μg decreased but did not fully suppress dauer formation in cmk-1 mutants (Figure 1D), implying that these animals are able to respond to food, but exhibit a shifted threshold of response to feeding. Consistent with the notion that cmk-1 mutants retain the ability to respond to food cues, egg-laying was modulated by bacterial food in both wild-type and cmk-1(oy21) adult animals (Figure 1—figure supplement 1E). In addition to food and pheromone, temperature also regulates dauer formation (Golden and Riddle, 1984b), and we and others previously showed that cmk-1 mutants exhibit altered thermosensory behaviors (Satterlee et al., 2004; Schild et al., 2014; Yu et al., 2014). cmk-1(oy21) mutants retained the ability to respond to temperature in the context of dauer formation, since dauer formation was fully suppressed at a lower temperature of 20°C in the presence of pheromone (Figure 1—figure supplement 1F). We infer that cmk-1 mutants are defective in correctly integrating food signals into the dauer decision pathway.

CMK-1 acts in the AWC and ASI/AWA sensory neurons to regulate dauer formation

cmk-1 is expressed broadly in multiple sensory and non-sensory neuron types in C. elegans (Kimura et al., 2002; Satterlee et al., 2004) (Figure 1—figure supplement 2). We first verified that expression of a cmk-1 cDNA under its endogenous regulatory sequences rescues the dauer formation phenotype in the presence of live bacteria and exogenous ascr#3 (Figure 1E). We next performed cell-specific rescue experiments to identify the site(s) of CMK-1 function in the regulation of dauer entry under these conditions. Expression in the ASK pheromone-sensing, or the AFD thermosensory, neurons did not affect the dauer formation phenotype of cmk-1 mutants (Figure 1E), suggesting, but not proving, that CMK-1 does not act simply by modulating pheromone or temperature sensitivity. However, expression of wild-type cmk-1 sequences in the ASI/AWA or AWC sensory neurons resulted in partial, and nearly complete, suppression of dauer formation, respectively. Expression of cmk-1 in both ASI/AWA and AWC suppressed dauer formation to the same extent as expression in AWC alone (Figure 1E). No rescue was observed upon expression in the ASJ sensory neurons which have also been previously implicated in the regulation of dauer formation (Bargmann and Horvitz, 1991; Schackwitz et al., 1996) (Figure 1E). Thus, CMK-1 acts primarily in AWC but also in ASI/AWA to regulate dauer formation.

CMK-1 acts in both the TGF-β and insulin pathways to regulate dauer formation

The DAF-7 TGF-β and ILP neuroendocrine signaling pathways act in parallel to regulate dauer formation (Fielenbach and Antebi, 2008). We asked whether CMK-1 acts in either or both these pathways to regulate dauer formation. daf-7 TGF-β null mutants are strongly Daf-c, and this phenotype is suppressed upon loss of DAF-3 SMAD or DAF-5 transcription factor function (Golden and Riddle, 1984c; Vowels and Thomas, 1992; Thomas et al., 1993). Despite the presence of multiple ILPs, the C. elegans genome encodes a single insulin receptor encoded by the daf-2 gene. Although loss of function of single ILP genes such as daf-28 results in only weak effects on dauer formation, likely due to redundancy (Li et al., 2003; Cornils et al., 2011; Ritter et al., 2013; Hung et al., 2014), daf-2 insulin receptor mutants are Daf-c (Thomas et al., 1993; Gottlieb and Ruvkun, 1994; Gems et al., 1998). The Daf-c phenotype of daf-2 mutants is suppressed by loss of daf-16 FOXO transcription factor function (Riddle et al., 1981; Gottlieb and Ruvkun, 1994). To determine whether CMK-1 reports food information to either, or both, the TGF-β and insulin pathways, we examined whether daf-3, daf-5, or daf-16 mutations suppress the dauer formation phenotype of cmk-1 mutants. We found that mutations in daf-3 and daf-5 partly suppressed the dauer formation defects of cmk-1 mutants, whereas loss of daf-16 function fully suppressed this phenotype (Figure 2A; Figure 2—source data 1). These results suggest that CMK-1 influences both the TGF-β and insulin pathways to regulate dauer formation.

CMK-1 acts cell-autonomously to regulate daf-7 TGF-β expression in ASI.

(A) Dauers formed by the indicated strains on 80 μg live OP50 and 6 μM ascr#3. Alleles used were: cmk-1(oy21), daf-3(mgDf90), daf-5(e1385), and daf-16(mgDf50). Shown are the averages of ≥3 independent experiments with >65 animals each. Errors are SEM. (B) Representative images of daf-7p::gfp expression in L1 larvae of wild-type or cmk-1(oy21) animals under the indicated conditions. Schematic at top indicates the position of ASI cell body (lateral view); boxed region is shown in panels below. Occasional weak expression is observed in the ADL neurons. Animals were grown with plentiful live OP50 or starved for at least 6 hr in the absence or presence of 1 unit of crude pheromone (see ‘Materials and methods’). White arrowheads indicate cell bodies of ASI. Numbers in bottom left hand corners indicate the percentage of examined larvae that exhibit the shown phenotype; n > 50 each; three independent experiments. Lateral view; scale bar: 10 μm. (C) Scatter plot of fluorescence intensity of daf-7p::gfp expression in ASI in wild-type or cmk-1(oy21) mutants. Median is indicated by a red horizontal line. Animals were grown on ample live OP50 in the absence of exogenous pheromone. Each dot is the fluorescence intensity of a single neuron in a given experiment; n > 60 neurons total each, at least three independent experiments. Promoters driving wild-type cmk-1 cDNA were: ASI—srg-47p; AWC—ceh-36Δp. (D) Dauers formed by shown strains on 80 μg live OP50 and 6 μM ascr#3. The srg-47 promoter was used to drive expression of wild-type daf-7 cDNA in ASI. Shown are the averages of ≥3 independent experiments with >65 animals each. Errors are SEM. Unless indicated otherwise, * and *** indicate different from wild-type at p < 0.05 and p < 0.001, respectively, #, ##, and ### indicate different from cmk-1 at p < 0.05, 0.01, and 0.001, respectively. (ANOVA and Games-Howell post-hoc test).

https://doi.org/10.7554/eLife.10110.008

CMK-1 acts cell-autonomously in ASI, and non cell-autonomously in AWC, to regulate TGF-β and insulin signaling, respectively

Previous work has shown that food and pheromone cues regulate the expression of both daf-7 TGF-β and the daf-28 ILP genes to modulate entry into the dauer stage. daf-7 expression in ASI, and daf-28 expression in both ASJ and ASI, are downregulated upon starvation or upon exposure to high pheromone concentrations (Ren et al., 1996; Schackwitz et al., 1996; Li et al., 2003; Entchev et al., 2015). Since CMK-1 acts in both the TGF-β and insulin pathways to regulate dauer formation, we asked whether the inability of food to fully suppress dauer formation in cmk-1 mutants is in part due to defects in regulation of expression of one or both ligands in cmk-1 mutants.

Since the cmk-1 mutant dauer phenotype is suppressed by downstream mutations in the TGF-β pathway (Figure 2A), we first asked whether CMK-1 acts cell-autonomously in ASI to regulate expression of the daf-7 TGF-β ligand. As reported previously, a daf-7p::gfp reporter gene was expressed strongly and specifically in the ASI neurons of fed wild-type L1 larvae grown under the conditions of low endogenous pheromone concentrations (Ren et al., 1996; Schackwitz et al., 1996; Entchev et al., 2015) (Figure 2B,C). This reporter has been validated to reflect expression of the endogenous daf-7 gene (Ren et al., 1996); the expression level of this gene is one of the components encoding food levels and is variable at all examined food concentrations (Meisel et al., 2014; Entchev et al., 2015). Expression of daf-7p::gfp was strongly decreased upon starvation, or in the presence of crude pheromone (Figure 2B). We found that unlike in wild-type animals, expression of daf-7p::gfp was reduced, but not abolished, in cmk-1 mutant larvae grown on live bacteria and no exogenous pheromone (Figure 2B,C). The strongly reduced expression under fed conditions in cmk-1 larvae precluded us from determining whether starvation or addition of pheromone resulted in a further decrease in daf-7p::gfp expression levels in this mutant background (Figure 2B). Expression of cmk-1 in ASI, but not in AWC, rescued the daf-7p::gfp expression defects of cmk-1 mutant larvae (Figure 2C), indicating that CMK-1 acts cell-autonomously in ASI to regulate daf-7 expression.

We next asked whether CMK-1 acts in parallel to the TGF-β pathway to also regulate ILP gene expression. A daf-28p::gfp reporter is expressed strongly in both ASI and ASJ neurons in fed wild-type larvae (Li et al., 2003) (Figure 3A,B). Under our conditions, addition of crude pheromone decreased expression primarily in ASI, whereas starvation resulted in decreased expression in both ASI and ASJ (Figure 3A). Addition of pheromone to starved wild-type larvae did not appear to further decrease expression in either cell type (Figure 3A). Expression of daf-28p::gfp was unaffected by temperature (Figure 3—figure supplement 1). Interestingly, we observed that in fed cmk-1 mutant larvae grown on live bacteria without exogenous pheromone, daf-28p::gfp expression was strongly decreased in ASJ and more weakly affected in ASI (Figure 3A,B). As in wild-type larvae, addition of pheromone or starvation decreased daf-28p::gfp expression in ASI; we were unable to detect whether expression in ASJ was further reduced under these conditions in cmk-1 mutants (Figure 3A). The daf-28p::gfp expression defect of cmk-1 mutants in ASJ was partly but significantly rescued upon expression of wild-type cmk-1 sequences in AWC but not in ASI/AWA or ASJ (Figure 3B). These results indicate that while pheromone regulates daf-28 expression primarily in ASI, food-dependent regulation of daf-28 is observed in both ASI and ASJ. Moreover, we find that CMK-1 acts non cell-autonomously in AWC to regulate daf-28 ILP gene expression specifically in ASJ under well-fed conditions.

Figure 3 with 1 supplement see all
CMK-1 acts non cell-autonomously in AWC to regulate daf-28 insulin-like peptide (ILP) gene expression in ASJ.

(A) Representative images of daf-28p::gfp expression in L1 larvae of wild-type or cmk-1(oy21) animals under the indicated conditions. Schematic of worm head indicating positions of the ASI and ASJ sensory neuron cell bodies is shown at top; boxed region is shown in panels below. Animals were grown with plentiful live OP50 or starved for at least 6 hr in the absence or presence of 1 unit of crude pheromone (see ‘Materials and methods’). White arrowheads and arrows indicate cell bodies of ASI and ASJ, respectively. Yellow arrowheads indicate expression in an ectopic cell observed in ∼14% of wild-type and ∼50% of cmk-1 mutants under all conditions. Numbers in bottom left hand corners indicate the percentage of examined larvae that exhibit the shown expression patterns; n > 50 each; three independent experiments. Lateral view; scale bar: 10 μm. (B) Scatter plot of fluorescence intensity of daf-28p::gfp expression in ASJ in wild-type or cmk-1(oy21) mutants. Median is indicated by red horizontal line. Animals were grown on ample live OP50 in the absence of exogenous pheromone. Each dot is the fluorescence intensity of a single neuron in a given experiment; n > 60 neurons total each, at least three independent experiments. For transgenic strains, a representative line was selected from experiments shown in Figure 1E and crossed into the reporter strains. Promoters driving wild-type cmk-1 cDNA were: ASI/AWA—gpa-4p; AWC—ceh-36Δp; ASJ—trx-1p. (C) Dauers formed by shown strains on 80 μg live OP50 and 6 μM ascr#3. Promoters used to drive expression of wild-type daf-28 cDNA were: ASI—srg-47p and ASJ—trx-1p. Shown are the averages of ≥3 independent experiments with >65 animals each. (D) Dauers formed by the indicated strains on 80 μg OP50 at 20°C in the absence of exogenous pheromone. Alleles used were: cmk-1(oy21) and daf-28(tm2308). Shown are the averages of ≥3 independent experiments with >65 animals each. Errors are SEM. Unless indicated otherwise, * and *** indicate different from wild-type at p < 0.01 and 0.001, respectively; # and ### indicate different from cmk-1 at p < 0.05 and 0.001, respectively; &&& indicates different from daf-28 at p < 0.001 (ANOVA and Games-Howell post-hoc test).

https://doi.org/10.7554/eLife.10110.010

If reduced expression of daf-7 and daf-28 in cmk-1 mutants is causal to their dauer formation phenotypes, we would predict that increased expression of daf-7 or daf-28 would rescue the dauer formation defects of cmk-1 mutants. We found that constitutive expression of daf-7 and daf-28 in ASI and ASJ, respectively, rescued the dauer formation phenotype of cmk-1 mutants (Figures 2D, 3C; Figure 3—source data 1). daf-28 expression in ASI failed to rescue (Figure 3C) indicating that neuron-specific release properties or spatial diffusion constraints may require DAF-28 expression in ASJ to rescue the cmk-1 phenotype (Cornils et al., 2011; Chen et al., 2013). Taken together, these results indicate that CMK-1 may relay food information into the dauer regulatory pathway by acting cell-autonomously to regulate daf-7 TGF-β expression in ASI, and non cell-autonomously in AWC to regulate daf-28 ILP expression in ASJ.

We further tested the regulatory relationship between CMK-1 and the parallel TGF-β and DAF-28 ILP signaling pathways by performing genetic epistasis experiments. If CMK-1 acts in both the TGF-β and DAF-28 pathways, we predicted that cmk-1; daf-28 double mutants would exhibit increased dauer formation defects compared to either single mutant alone, in part due to reduced daf-7 expression in cmk-1 mutants. As shown in Figure 3D, we found that a significantly larger percentage of cmk-1; daf-28 double mutant animals entered into the dauer stage as compared to cmk-1 or daf-28 null mutants alone in the presence of live food and no added pheromone. Taken together, these results confirm that CMK-1 regulates both TGF-β and insulin signaling to modulate dauer formation.

Feeding state is reflected in the temporal dynamics of CMK-1 subcellular localization in AWC

The AWC neurons have not previously been implicated in dauer formation. We further explored the role of CMK-1 in AWC in the regulation of dauer formation as a function of feeding state. We and others previously showed that CMK-1 shuttles between the nucleus and the cytoplasm in thermosensory neurons based on growth temperature (Schild et al., 2014; Yu et al., 2014), and regulates the expression of thermotransduction genes (Yu et al., 2014). We asked whether CMK-1 subcellular localization in AWC is similarly affected by feeding status. In fed L1 larvae, a functional CMK-1::GFP fusion protein was present mostly uniformly throughout the soma of the AWC neurons (Figure 4A,B). Food withdrawal for 30 min resulted in a transient nuclear enrichment of CMK-1 that persisted for approximately 1 hr (Figure 4A,B). However, after prolonged starvation, CMK-1::GFP was enriched in the cytoplasm of AWC (Figure 4A,B). In contrast, food did not affect subcellular localization of CMK-1::GFP in AFD (Figure 4A), and we previously showed that temperature does not affect CMK-1 localization in AWC (Yu et al., 2014). Overexpression of a constitutively nuclear-enriched CMK-1::GFP protein in AWC strongly rescued the dauer formation phenotype of cmk-1 mutants (Figure 4C; Figure 4—source data 1), whereas a constitutively cytoplasmically enriched protein rescued more weakly (Figure 4C). The subcellular localization of these proteins in AWC was unaffected by genotype or feeding state (Figure 4—figure supplement 1). However, we note that neither the nuclear-enriched nor the cytoplasmically enriched CMK-1::GFP fusion proteins are fully excluded from the cytoplasmic or nuclear compartment, respectively, possibly due to overexpression (Figure 4—figure supplement 1). These results suggest that feeding state regulates CMK-1 subcellular localization in AWC, and that CMK-1 activity in the nuclei of AWC may be important to encode food information in the dauer regulatory pathway.

Figure 4 with 1 supplement see all
Subcellular localization of CMK-1 in AWC is regulated by feeding state.

(A, B) Representative images (A) and quantification (B) of subcellular localization of CMK-1::GFP in AWC neurons following removal from food for the indicated times. Representative images of CMK-1::GFP localization in AFD are also shown in A. Scale bar: 5 μm (A). n > 75 AWC neurons each (B). *** indicates different from distribution at 0 min at p < 0.001 (χ2 test). (C) Dauers formed by the indicated strains on 80 μg live OP50 and 6 μM ascr#3. CMK-1::NLS::GFP and CMK-1::NES::GFP were expressed in AWC under the ceh-36Δ promoter. Shown are the averages of ≥3 independent experiments with >65 animals each. For transgenic strains, data are averaged from two independent lines each. Errors are SEM. *and *** indicate different from wild-type at p < 0.05 and 0.001, respectively; ### indicates different from cmk-1(oy21) at p < 0.001 (ANOVA and Games-Howell post-hoc test).

https://doi.org/10.7554/eLife.10110.013

CMK-1 regulates the expression of ILP genes in AWC to antagonize dauer formation

We next investigated the nature of the CMK-1-regulated signal in AWC that may transmit food information to the dauer regulatory pathway. The transient nuclear localization of CMK-1 in AWC upon food withdrawal suggests that CMK-1 regulates gene expression in AWC as a function of feeding state. A link between nutrient availability and insulin signaling is now well established in many organisms (Erion and Sehgal, 2013; Riera and Dillin, 2015), and in C. elegans, the expression of several ILP genes has been shown to be regulated by food availability (Li et al., 2003; Cornils et al., 2011; Ritter et al., 2013; Chen and Baugh, 2014). Together with the hours-long timescale of integration of sensory cues for the regulation of dauer formation (Golden and Riddle, 1984c; Schaedel et al., 2012), and the fact that AWC may signal feeding state to regulate downstream hormonal signaling, we hypothesized that CMK-1 may regulate the expression of ILP genes in AWC as a function of feeding state.

Of the subset of 40 predicted ILP genes (Pierce et al., 2001; Li and Kim, 2010) (www.wormbase.org) whose expression is reported to be regulated by nutrient availability in C. elegans, only ins-26 and ins-35 are known to be expressed in AWC (as well as in additional cells) (Chen and Baugh, 2014). However, whether their expression in AWC is regulated by food has not been previously determined. We asked whether ins-26 and ins-35 expression in AWC is regulated by food and CMK-1. Under fed conditions, ins-26p::yfp was expressed in ASI, ASE, and a subset of AWC neurons in wild-type L1 larvae; expression in AWC, and to a lesser extent in ASI, was reduced upon starvation (Figure 5A,B). Relative to wild-type, ins-26p::yfp expression was decreased in ASE and AWC in cmk-1 mutants grown with plentiful food, whereas expression in ASI decreased only upon starvation (Figure 5A,B). Similarly, ins-35p::yfp was largely expressed in ASE and AWC in fed wild-type L1 larvae; expression in both neurons was significantly decreased upon starvation, and in fed and starved cmk-1 mutants (Figure 5A,B). Weak expression of ins-35 was also observed in the ASK neurons, particularly in cmk-1 mutants, although expression in this neuron type was not further modulated by starvation (Figure 5B). Expression of a constitutively nuclear-enriched CMK-1 protein in AWC neurons increased ins-35 expression in cmk-1 mutants in both fed and starved conditions, to a slightly greater extent than a constitutively cytoplasmically enriched protein (Figure 5B). These observations suggest that the expression of ins-26 and ins-35 in AWC is regulated by food availability, and that their expression pattern in AWC in fed cmk-1 mutants resembles that of starved wild-type animals. Moreover, nucleocytoplasmic shuttling of CMK-1 may be important for feeding state-dependent regulation of ins-35 gene expression.

Figure 5 with 2 supplements see all
CMK-1 maintains a balance of anti- and pro-dauer signals from AWC as a function of feeding state.

(A) (Top) Schematic of worm head indicating positions of sensory neuron soma. Boxed area is shown in images below. (Below) Representative images of ins-26p::yfp and ins-35p::yfp expression in fed wild-type and cmk-1(oy21) mutants. White and yellow arrowheads indicate ASI and ASE, respectively; white arrow indicates AWC; yellow asterisk marks expression in the intestine. The location of ASK is indicated by a red arrowhead; fluorescence in ASK is weak and not visible at this exposure in shown images. Lateral view; scale bar: 10 μm. (B) Quantification of expression in each neuron type in fed and starved (>6 hr) conditions. Solid and hatched bars indicate strong and weak expression, respectively, in each cell. n > 50 animals each; three independent experiments. (C) Dauers formed by shown strains on 160 μg heat-killed OP50 and 60 nM ascr#2 (left), and 80 μg live OP50 and 6 μM ascr#3 + 600 nM ascr#5 (right). Alleles used were: ins-26(tm1983), ins-35(ok3297) and ins-32(tm6109). n > 4 assays of 65 animals each; at least three independent experiments. (D) Dauers formed by shown strains on 6 μM ascr#3 and 80 μg live OP50. ins-26 and ins-35 cDNAs were expressed in ASE, ASJ, and AWC under che-1, trx-1, and ceh-36∆ regulatory sequences, respectively. At least two independent lines were analyzed for each transgenic strain. Shown are the averages of at least three independent experiments with >65 animals each. (E) Scatter plot of fluorescence intensity of daf-28p::gfp expression in ASJ in the indicated genetic backgrounds. ins-26 and ins-35::SL2::mCherry cDNAs were expressed in AWC under the odr-1 promoter. Only animals expressing mCherry in AWC were scored. Median is indicated by red horizontal line. Each dot is the fluorescence intensity of a single neuron in a given experiment. n > 60 neurons total each, four independent experiments. (F) Quantification of dauer formation in AWC-ablated animals in the presence of 6 μM ascr#3 and the indicated amounts of heat-killed (blue circles) or live (red circles) OP50. Each filled circle represents one assay; n > 65 animals per assay, five independent experiments. Line represents best fit to the data. Dashed line indicates the curve for wild-type animals from Figure 1B shown for comparison. *, **, and *** indicate different from values using 160 μg of heat-killed bacteria at p < 0.05, 0.01, and 0.001, respectively. (G) Dauers formed by the indicated strains on 80 μg live OP50 and 6 μM ascr#3. For each assay: n > 65 animals; four independent experiments. Errors are SEM. Except where indicated, *, **, and *** indicate different from wild-type at p < 0.05, 0.01, and 0.001, respectively; #, ##, and ### indicate different from cmk-1(oy21) at p < 0.05, 0.01, and 0.001, respectively (ANOVA and Games-Howell post-hoc test).

https://doi.org/10.7554/eLife.10110.016

Reduced expression of ins-26 and ins-35 in AWC in starved wild-type or fed cmk-1 mutants suggest that these peptides are candidates for an anti-dauer or pro-growth signal from AWC. If this were the case, we would predict that loss of function of these genes would enhance dauer formation in wild-type but not in cmk-1 mutants. Indeed, we found that ins-26; ins-35 double mutants formed more dauers on heat-killed bacteria in the presence of pheromone than wild-type animals (Figure 5C; Figure 5—source data 1). Dauer formation in cmk-1 mutants on live food was not significantly enhanced upon loss of either ins-26 or ins-35 (Figure 5D). Since live vs heat-killed food provide different sensory inputs that might confound comparisons between wild-type and cmk-1 mutants, we established conditions under which we could force a small fraction of wild-type larvae to enter into the dauer stage on live food (see ‘Materials and methods’). ins-26; ins-35 double mutants also exhibited enhanced dauer formation on live food (Figure 5C), further suggesting that these ILPs may comprise an anti-dauer signal.

To confirm that reduced expression of these ILP genes in AWC in cmk-1 mutants is partly causal to their increased dauer formation phenotype, we asked whether restoration of ILP gene expression specifically in AWC was sufficient to rescue the dauer formation phenotype of cmk-1 mutants. As shown in Figure 5D, we found that overexpression of ins-26 and ins-35 specifically in AWC, but not in ASE or ASJ, rescued the dauer formation defect of cmk-1 mutants. Overexpression of ins-26 and ins-35 in AWC in cmk-1 mutants was also sufficient to partly restore daf-28p::gfp expression in ASJ (Figure 5E). Together, these results suggest that food-dependent regulation of ins-26 and ins-35 in AWC may comprise an anti-dauer signal, and that inappropriate downregulation of these genes in AWC in cmk-1 mutants may in part cause the increased dauer formation phenotype in these animals under fed conditions.

The balance between anti- and pro-dauer signals from AWC is disrupted in cmk-1 mutants

If AWC were in part required to transmit information about food availability to limit dauer formation, we would predict that ablation of AWC would result in increased dauer formation in wild-type animals, and that dauer formation in these animals would be less sensitive to regulation by food. Indeed, we found that AWC-ablated animals formed more dauers overall on heat-killed food than wild-type animals, and moreover, the ability of heat-killed bacteria to modulate dauer formation was reduced (Figure 5F).

Since the ins-26 and ins-35 anti-dauer signals are downregulated in cmk-1 mutants, we expected that ablation of AWC would not further increase dauer formation in cmk-1 mutants. Unexpectedly, we found that ablation of AWC instead suppressed dauer formation in cmk-1 mutants grown on live food (Figure 5G). One interpretation of this observation is that AWC also sends a pro-dauer signal in cmk-1 mutants. Consistent with this notion, ablation of AWC significantly rescued daf-28p::gfp expression in ASJ in cmk-1(oy21) animals (Figure 5—figure supplement 1). We considered the possibility that this pro-dauer signal could be present aberrantly only in cmk-1 mutants. Alternatively, wild-type AWC neurons could also send a pro-dauer signal, but this signal may be masked by the different food conditions (heat-killed vs live food) employed to examine dauer formation in wild-type and cmk-1 mutants, respectively. To address this issue, we asked whether ablation of AWC alters dauer formation of wild-type animals on live food. Indeed, we found that ablation of AWC also decreased dauer formation in wild-type animals grown on live food (Figure 5C), and increased daf-28p::gfp expression in ASJ (Figure 5—figure supplement 1) suggesting that AWC also sends a pro-dauer signal in wild-type animals that is revealed under specific conditions.

We next investigated the identity of this pro-dauer signal. The BLI-4 and EGL-3 proprotein convertases regulate the processing of different subsets of ILP precursors (Leinwand and Chalasani, 2013; Hung et al., 2014). Knocking down bli-4, but not egl-3 in the AWC neurons suppressed dauer formation in cmk-1 mutants (Figure 5—figure supplement 2; Figure 5—source data 2), suggesting that a BLI-4-dependent insulin signal(s) from AWC promotes dauer formation in cmk-1 mutants. Of the nineteen predicted BLI-4 targets (Leinwand and Chalasani, 2013), only the ins-32 ILP gene was previously reported to be expressed in AWC in adult hermaphrodites (Takayama et al., 2010). We found that ins-32 mutants formed fewer dauers on heat-killed food, as well as live food, and pheromone in an otherwise wild-type background (Figure 5C). Loss of ins-32 also suppressed the increased dauer formation phenotype of cmk-1 mutants on live food and pheromone (Figure 5D). We could not detect ins-32 expression in AWC in L1 larvae although we noted that expression of ins-32 reporter gene was weak and dynamic (SJN and PS, unpublished), raising the possibility that ins-32 is expressed in AWC only transiently during the period of dauer signal integration in early postembryonic development (Golden and Riddle, 1984c; Schaedel et al., 2012). Together, these observations suggest that different insulin peptides comprise the food-dependent anti- and pro-dauer signals from AWC, and that the balance between expression and/or release of these antagonistic signals is disrupted in cmk-1 mutants.

The AWC neurons are hyperactive upon long-term starvation in wild-type animals and in fed cmk-1 mutants

CaMKs such as CaMKI and CaMKIV play critical roles in regulating neuronal activity-dependent processes, including activity-dependent gene expression (Wayman et al., 2008, 2011; Cohen et al., 2015). In turn, changes in gene expression feed back to regulate neuronal state and properties. We asked whether AWC activity is altered upon starvation or in cmk-1 mutants. It has previously been shown that the AWC neurons in adult animals are silenced in the presence of food or food-related odorants and are activated upon food/odor withdrawal (Chalasani et al., 2007). Consistent with this observation, the AWC neurons of well-fed wild-type L2 larvae expressing the GCaMP calcium indicator (Nakai et al., 2001) exhibited few, if any somatic calcium transients (Figure 6A,B). However, we found that wild-type AWC neurons became hyperactive following prolonged food deprivation. Both the frequency and amplitude of spontaneous calcium transients increased upon starvation for 2 hr, although no effects were observed after 1 hr (Figure 6A,B). Interestingly, the AWC neurons in well-fed cmk-1 mutant larvae exhibited a high frequency and amplitude of somatic calcium transients similar to the activity pattern observed in starved wild-type larvae (Figure 6A,B); activity in AWC in cmk-1 larvae was not further increased by starvation (Figure 6A,B). Overexpression of wild-type cmk-1 sequences in AWC suppressed basal GCaMP expression (data not shown), implying that AWC may be silenced under these conditions, and precluded examination of cell-specific effects of CMK-1 on AWC activity.

Figure 6 with 1 supplement see all
The AWC neurons exhibit increased basal activity in fed cmk-1, and starved wild-type animals.

(A) Heat maps showing the fluorescence intensity (∆F/F0) in the soma of AWC neurons in fed or starved wild-type and cmk-1(oy21) L2 larvae expressing GCaMP 3.0 in AWC under the ceh-36∆ promoter. Animals were cultured at 20°C, starved for 0, 1, or 2 hr and imaged at 20°C. Each horizontal line shows calcium dynamics in a single AWC neuron; n = 20 neurons each. (B) Box-and-whisker plots quantifying total duration of calcium responses >5% above baseline for each genotype and condition shown in A. Median is indicated by a red line. Tops and bottoms of boxes indicate the 75th and 25th percentiles, respectively; whiskers represent fifth and 95th percentiles. Outliers are indicated by + signs. ** indicates different from wild-type at 0 hr at p < 0.01 (Kruskal–Wallis test). n.s.—not significant. (C) Average calcium responses in AWC neurons of fed and starved animals expressing GCaMP 3.0 under the ceh-36∆ promoter in the presence, or upon removal of the odorant isoamyl alcohol (IAA), diluted to 10−3 (blue) or 10−4 (red). Error bars are SEM and are represented by light gray shading. n ≥ 10 for each genotype and condition shown. (D) Scatter plot of the peak fluorescence amplitudes of individual neuron responses following odorant removal for the indicated conditions and genotypes. Horizontal black bars represent the median. * represents different between the indicated values at p < 0.05 (Student's t-test). (E) Dauer formation in the presence of 80 μg live OP50 and 6 μM ascr#3 by non-transgenic and transgenic animals expressing the Drosophila histidine-gated chloride channel 1 (HisCl1) in the presence (black) or absence (gray) of 10 mM histidine. Error bars are SEM. n.s.—not significant; * and *** indicate different between indicated values at p < 0.05 and 0.001, respectively (Student's t-test). (F) Dauer formation in the presence of 80 μg live OP50 and 6 μM ascr#3 by non-transgenic and transgenic animals expressing the constitutively active potassium channel TWK-18(gf). Error bars are SEM. * and ** indicate different between indicated values at p < 0.05 and 0.01, respectively (Student's t-test).

https://doi.org/10.7554/eLife.10110.021

Fed wild-type animals must be transiently starved for the AWC neurons to respond to odor/food removal following a brief exposure to the stimulus (S Chalasani, personal communication). Confirming this observation, we found that removal of either of the attractive odorants isoamyl alcohol or benzaldehyde following a 1 min exposure failed to elicit a response in the AWC neurons of wild-type adult animals transferred directly from food, but led to responses in animals starved on a food-free plate (Figure 6C,D, Figure 6—figure supplement 1). Since the activity state of AWC in fed cmk-1 mutants resembles that of starved wild-type animals, we asked whether AWC neurons in cmk-1 mutant adults are able to respond to odor removal without prior starvation. Indeed, the AWC neurons in cmk-1 adult animals responded robustly to odorant removal regardless of their feeding state (Figure 6C,D, Figure 6—figure supplement 1). Together, these observations imply that the basal activity state of the AWC neurons in fed cmk-1 mutants is similar to the activity state of these neurons in wild-type animals starved for extended periods. Importantly, these results also indicate that AWC retains the ability to respond to food-associated odors in cmk-1 mutants.

We next explored the relationship between increased AWC activity in cmk-1 mutants and their enhanced dauer phenotype. To address this issue, we inhibited activity in AWC neurons and examined the effect of this inhibition on dauer formation. Addition of exogenous histamine has been shown to silence neuronal activity in C. elegans neurons expressing the Drosophila histamine-gated chloride (HisCl1) channel (Pokala et al., 2014). Expression of an activated potassium channel (twk-18(gf)) has also been shown to hyperpolarize C. elegans neurons (Kunkel et al., 2000; Kawano et al., 2011; Zhang and Zhang, 2012). We found that growth on 10 mM histamine increased dauer formation in transgenic wild-type as well as in cmk-1 mutants expressing HisCl1 in AWC but had no effect on non-transgenic animals (Figure 6E; Figure 6—source data 1). Similarly, dauer formation was enhanced in both cmk-1(oy21), and to a weaker extent in wild-type animals expressing twk-18(gf) in AWC (Figure 6F). These results suggests that activity in AWC antagonizes dauer formation in both wild-type and cmk-1 mutants, and acts in parallel with CMK-1 to regulate dauer formation (Figure 7; see ‘Discussion’).

Model for the role of AWC and CMK-1 in the regulation of the dauer decision as a function of feeding state.

CMK-1 acts in AWC to drive expression of the anti-dauer ins-26 and ins-35 ILP genes. CMK-1 may also regulate expression of the ins-32 (or other) pro-dauer signal in AWC; alternatively, the pro-dauer signal could be present tonically regardless of environmental state. Under fed conditions, CMK-1-regulated anti-dauer signals from AWC predominate and drive expression of the daf-28 ILP gene in ASJ. CMK-1 also acts cell-autonomously to regulate expression of daf-7 TGF-β in ASI. Together, daf-7 TGF-β and daf-28 ILP signals promote reproductive development. When starved, anti-dauer signals from AWC are downregulated resulting in decreased expression of daf-28 in ASJ. daf-7 and daf-28 expression in ASI are also downregulated upon starvation. The shifted balance towards pro-dauer signals promotes dauer formation in the presence of pheromone. In cmk-1 mutants, loss of the anti-dauer signals from AWC downregulates daf-28 expression in ASJ under fed conditions. daf-7 expression in ASI is also downregulated in cmk-1 mutants. Consequently, the inappropriate predominance of pro-dauer signals in cmk-1 mutants promotes dauer formation in the presence of plentiful food and pheromone. Feeding conditions also modulate AWC basal neuronal activity; increased activity upon starvation or in cmk-1 mutants may limit dauer formation via a parallel pathway.

https://doi.org/10.7554/eLife.10110.024

Discussion

A balance between anti- and pro-dauer ILP signals from AWC relays food inputs into the dauer decision

Environmental signals such as food availability are encoded in the level of expression of the daf-7 TGF-β, and daf-28 and other ILP genes, to regulate the binary decision between entry into the reproductive cycle or the dauer stage (Ren et al., 1996; Schackwitz et al., 1996; Li et al., 2003; Entchev et al., 2015). We have identified AWC as key neurons that translate food information into changes in daf-28 ILP gene expression to regulate dauer formation. Our results suggest a model in which the AWC neurons transmit both anti- and pro-dauer signals; the balance between these signals is determined by food inputs and feeding state to regulate a critical developmental decision in the nematode life cycle. We show that the anti-dauer signals are comprised in part of the AWC-expressed ins-26 and ins-35 ILP genes, whereas the ins-32 ILP gene may constitute a component of the pro-dauer signal (Figure 7). Expression of ins-26 and ins-35 in AWC is downregulated under starvation conditions that promote dauer formation, and loss of both genes enhances dauer formation (Figure 7). Consistent with a role for these ILPs in antagonizing dauer formation, ins-26 and ins-35 mutations were reported to enhance dauer formation in sensitized backgrounds, and ins-35 was proposed to represent a genetic ‘hub’ regulating dauer entry (Fernandes de Abreu et al., 2014). Conversely, loss of ins-32 suppresses dauer formation, although it remains possible that ins-32 acts in cell types other than AWC to promote dauer formation.

The complex balance between pro- and anti-dauer signals as a function of environmental conditions is highlighted in the effects of AWC ablation on dauer formation. Ablation of AWC enhances dauer formation on heat-killed food but suppresses dauer formation on live food. Bacterial food is a complex cue consisting of many volatile and non-volatile compounds (Orth et al., 2011) that are sensed by multiple neuron types. Live and heat-killed bacteria provide different signals which may be sensed and integrated by distinct networks in the dauer decision. Thus, integration of different food signals from diverse sensory neurons likely differentially influences dauer formation on different food types. Indeed, ASI detects food directly to regulate daf-7 expression (Ren et al., 1996; Schackwitz et al., 1996; Gallagher et al., 2013), and recent work has shown that complex patterns and shapes of expression of daf-7 and the tph-1 tryptophan hydroxylase genes across multiple neuron types represents a ‘neural code’ for food abundance in the context of regulation of adult lifespan (Entchev et al., 2015). Given the consequences of phenotypic plasticity on fitness (Nijhout, 2003; Avery, 2014), the precise tuning of anti- and pro-dauer signals in multiple neurons, including the AWC neurons, is critical to allow animals to make the optimal developmental decision in response to specific environmental conditions.

CMK-1 regulates the balance between anti- and pro-dauer signals in AWC as a function of feeding state

We have shown that the CMK-1 CaMKI plays a critical role in regulating the balance between anti- and pro-dauer signals from AWC. Our results suggest that in cmk-1 mutants, the balance between these signals is decoupled from feeding state, resulting in deregulated dauer formation under conditions that suppress dauer formation in wild-type animals. This hypothesis is based on several lines of evidence. First, cmk-1 mutants form dauers inappropriately on food conditions that fully suppress dauer formation in wild-type animals. Second, the expression patterns of daf-7 in ASI and daf-28 in ASJ in fed cmk-1 mutants are similar to those in starved wild-type animals, indicating that the expression of these key dauer-regulatory hormones does not reflect food abundance accurately in the absence of CMK-1. Third, expression of ins-26 and ins-35 ILP genes in AWC in fed cmk-1 mutants resembles those in starved wild-type animals, further supporting the notion that food signals and gene expression patterns are uncoupled in cmk-1 mutants (Figure 7). Overexpression of ins-26 and ins-35 from AWC, and loss of ins-32 suppress the enhanced dauer formation phenotype of cmk-1 mutants indicating that the altered balance between anti-dauer and pro-dauer signals together with decreased daf-7 expression in ASI, sensitizes the threshold for dauer formation in cmk-1 mutants. Consistent with the notion that this balance is disproportionately shifted towards increased pro-dauer signals from AWC in cmk-1 mutants, ablation of AWC suppresses increased dauer formation and partially restores daf-28p::gfp expression in this background. It is important to note, however, that the expression pattern changes in fed cmk-1 mutants are not identical to those in starved wild-type animals, indicating that additional mechanisms translate food abundance into changes in gene expression patterns.

How might CMK-1 link food stimuli to dauer signals from AWC? We report that the subcellular localization of CMK-1 in AWC is highly dynamic and feeding state-dependent. Starvation results in transient nuclear localization of CMK-1, followed by cytoplasmic enrichment upon prolonged food deprivation. Although we do not yet know the nature of the signal that triggers the initial nuclear translocation, neuronal activity has been shown in multiple contexts to regulate subcellular localization of both mammalian and C. elegans CaMKI (Eto et al., 1999; Ueda et al., 1999; Sakagami et al., 2005; Schild et al., 2014; Yu et al., 2014). Since the AWC neurons must be transiently starved to exhibit responses to odorant removal, one possibility is that this odor removal-induced depolarization promotes CMK-1 nuclear translocation.

Consistent with a well-described role of CaMKs in mediating activity-dependent gene expression (Flavell and Greenberg, 2008; Wayman et al., 2011; West and Greenberg, 2011; Bengtson and Bading, 2012), nuclear localized CMK-1 may then regulate the transcription of anti- and pro-dauer signals. We show that constitutive nuclear localization of CMK-1 is sufficient to suppress dauer formation in the cmk-1 background and increase ins-35 expression under both fed and starved conditions. An intriguing possibility is that CMK-1 initially upregulates anti-dauer signals such as ins-26 and ins-35 upon transient food removal. As food deprivation persists, CMK-1 moves into the cytoplasm resulting in downregulation of anti-dauer signals to promote dauer formation. CMK-1 may also upregulate pro-dauer signals from AWC. Alternatively, pro-dauer signals may be expressed and/or released tonically, and CMK-1-mediated regulation of anti-dauer signals may be sufficient to promote or inhibit dauer formation. C. elegans larvae integrate food signals over an extended period of time to allow accurate assessment of their past, current, and future environment (Schaedel et al., 2012; Avery, 2014). We speculate that not only food abundance but also temporal variability in food availability must be encoded in the balance of anti- and pro-dauer signals. Such mechanisms would allow larvae to ignore relatively transient fluctuations in environmental cues and ensure that irreversible commitment to the dauer stage occurs only under persistently adverse conditions.

Neuronal activity acts in parallel with CMK-1 to antagonize dauer formation

Basal activity in AWC is increased upon prolonged starvation in wild-type animals, and in fed cmk-1 mutants. In wild-type animals, increased AWC activity upon prolonged starvation may be a consequence of feedback from internal state. Similar modulation of peripheral sensory neuron responses as a function of feeding state has been suggested to underlie state-dependent plasticity in sensory behaviors in other organisms (Jyotaki et al., 2010; Palouzier-Paulignan et al., 2012; Sengupta, 2013; Pool and Scott, 2014). In starved wild-type animals, or in the absence of CMK-1 function, this state-dependent feedback may increase AWC activity, and this increased activity in turn, may antagonize dauer formation by regulating an as yet uncharacterized pathway that acts in parallel to the CMK-1 regulated pathway in AWC. Why would animals need to continue to antagonize dauer formation even following prolonged starvation? Pre-dauer L2d larvae integrate environmental cues from 16 hr–33 hr post hatching and make an irreversible commitment to dauer entry only after 33 hr (Schaedel et al., 2012). Given this long timeframe of sensory signal integration, and that entry into the dauer stage is a bet-hedging strategy that maximizes fitness under uncertain environmental conditions (DeWitt and Scheiner, 2004; Avery, 2014), the availability of multiple dauer-regulatory pathways that integrate environmental cues on different timescales may be critical to correctly promote or suppress the dauer decision. A goal for the future will be to define how activity in AWC modulates dauer formation, to identify the circuit mechanisms by which AWC activity is altered as a function of starvation, and to correlate temporal changes in AWC activity with commitment to the dauer stage.

Concluding remarks

Although dauer formation and other forms of polyphenism are the extreme examples of phenotypic plasticity, environmental cues experienced during defined sensitive or critical periods during development also underlie phenotypic variation in mammals (Gluckman et al., 2007). Our results describe how C. elegans integrates and translates environmental cues into hormonal signaling to regulate the dauer decision. We expect that continued investigation of the molecular and neuronal regulation of phenotypic plasticity by sensory cues in different species will lead to insights into the general principles underlying these critical developmental decisions, as well as provide information about the mechanisms of sensory integration that direct the choice of the appropriate developmental pathway.

Materials and methods

Strains and genetics

C. elegans was maintained on nematode growth medium (NGM) agar plates at 20°C, with Escherichia coli OP50 as a food source. Strains were constructed using standard genetic procedures. The presence of mutations was confirmed by PCR-based amplification and/or sequencing. A complete list of strains used in this study is shown in Supplementary file 1.

Molecular biology

Promoter sequences and cDNAs were amplified from genomic DNA or a cDNA library, respectively, from a population of mixed-stage wild-type animals. cDNA sequences were verified by sequencing. The promoters used in this study are as follows: cmk-1p (3.1 kb), sra-9p (ASK, 2.9 kb), ttx-1p (AFD, 2.7 kb), trx-1p (ASJ, 1.0 kb), gpa-4p (ASI/AWA, 2.8 kb), ceh-36Δp (AWC, 0.6 kb), srg-47p (ASI, 1.0 kb), ins-26p (see Chen and Baugh, 2014), ins-35p (see Chen and Baugh, 2014), che-1p (ASE, 0.7 kb), odr-1p (AWC [strong], AWB [weak], 2.4 kb) and odr-3p (AWC [strong], AWA, AWB, ASH, ADF [weak], 2.7 kb). Sense and antisense constructs for RNAi were generated by amplifying exons 2–8 (1.5 kb; bli-4) and 6–8 (568 bp; egl-3) from cDNAs, and cloning into vectors containing cell-specific promoter sequences. Linearized vectors containing either sense or antisense sequences were injected at 100 ng/μl each. The twk-18(gf) allele (Kunkel et al., 2000; Kawano et al., 2011; Zhang and Zhang, 2012) and the Drosophila HisCl1 channel cDNA (Pokala et al., 2014) fused via SL2 to an mCherry reporter were expressed under ceh-36Δp regulatory sequences. ceh-36Δp::HisCl1 and ceh-36Δp::twk-18(gf) were injected at a concentration of 50 ng/μl.

Dauer assays

Dauer assays were performed essentially as described (Neal et al., 2013), using the indicated food sources. Briefly, young adult worms were allowed to lay 65–85 eggs on an assay plate (Neal et al., 2013) containing either ethanol (control) or pheromone, and a defined amount of bacteria. Animals were grown at 25°C unless indicated otherwise. Assays using heat-killed food or live food were examined for dauer and non-dauer larvae approximately 84 hr or 66 hr, respectively, after the midpoint of the egg lay. In order to induce wild-type animals to form dauers on assay plates containing 80 μg live food, a mixture of 6 μM ascr#3 + 600 nM ascr#5 was used. For experiments involving transgenic animals expressing the Drosophila HisCl1 channel, 30 μl of 1 M histidine (Sigma-Aldrich, St. Louis, MO) was mixed with either pheromone or ethanol and was added to the assay plates and overlaid with molten assay agar such that the final concentration was 10 mM. At least three independent trials were conducted for each condition with two technical replicates each. Statistical analyses of dauer data were performed in SPSS (IBM, Armonk, NY), as described in the figure legends.

Quantification of fluorescence intensity

Strains expressing fluorescent reporters were growth-synchronized by hypochlorite treatment, and embryos were allowed to develop for 20–24 hr at 20°C to the end of the L1 larval stage on OP50, in the presence or absence of crude pheromone (Golden and Riddle, 1982; Zhang et al., 2013). Crude pheromone plates were prepared by spreading 20 μl of 1:4 crude pheromone (∼1 unit, defined as the amount required for forming ∼33% dauers on heat-killed OP50 at 25°C) on the agar surface and allowing it to dry completely prior to seeding with bacteria. For starvation conditions, worms were washed from growth plates in M9 buffer and were transferred to assay plates for 4–6 hr. Prior to imaging, worms were collected by centrifugation, transferred to a 2% agarose pad on a microscope slide, and immobilized using 10 mM levamisole (Sigma-Aldrich). Animals were visualized on a Zeiss Axio Imager.M2 microscope using either a 40× (NA 1.3) or 63× (NA 1.4) oil objective, and images were captured using a Hamamatsu Orca camera. Neurons were identified by position relative to the subset of neurons filled with DiI (Sigma-Aldrich).

For quantification of fluorescence intensity, images were acquired from a single focal plane. The exposure time for each fluorescent reporter was adjusted in the wild-type background to ensure that pixel intensity in the cell of interest was in the linear range. Pixel intensities for the soma (daf-7p::gfp reporter) or the nucleus (daf-28p::gfp reporter) were measured in ImageJ (NIH) by calculating the mean pixel intensity for the entire region of interest. All measurements within a single experiment were normalized to the median wild-type expression value (set at 1) to account for variation in conditions across trials. All imaging and subsequent quantification was performed blind to the genotype.

For shown representative images, z-stacks (0.5 μm per slice) were acquired through the head of the animal, and a sub-stack containing all GFP-expressing cells was rotated and maximally projected in ImageJ. All images within a panel were collected using the same exposure times. Adjustments to levels and/or contrast for optimal viewing were applied equally to images within each panel. Images used for quantification were not processed.

Calcium imaging

Imaging of spontaneous calcium dynamics in AWC was performed essentially as previously described (Biron et al., 2008). Briefly, L2 larvae were glued to an NGM agar pad on a cover glass, bathed in M9, and mounted under a second cover glass for imaging. The edge of the sandwiched cover glasses was sealed with a mixture of paraffin wax (Fisher Scientific, Pittsburgh, PA), and Vaseline, and the sample was transferred to a slide placed on a Peltier device on the microscope stage. The elapsed time from removal of the animal from the incubator to initiation of imaging was <3 min. The temperature was maintained at 20°C via temperature-regulated feedback using LabView (National Instruments, Austin, TX) and measured using a T-type thermocouple. Individual animals were imaged for 90 s at a rate of 2 Hz. Images were captured using MetaMorph (Molecular Devices, Sunnyvale, CA) and a Hamamatsu Orca digital camera. Data were analyzed using custom scripts in MATLAB (The Mathworks, Natick, MA) (Source code 14). A neuronal response was defined as the percent change of the relative fluorescence of the neuron from its baseline fluorescence level after background subtraction. A fluorescence change of >5% in AWC was considered a response. The duration of calcium events was calculated as the sum of all events in each animal, and averaged for each genotype and condition.

Odor-evoked imaging was performed as described previously (Jang et al., 2012; Ryan et al., 2014) using custom microfluidics devices. Imaging was conducted on an Olympus BX52WI or Carl Zeiss Axio Observer A1 microscope equipped with a 40X oil objective and a Hamamatsu Orca CCD or a Zeiss Axiocam 506 mono camera. Animals were exposed to a 1 min pulse of diluted odorant in S-basal. Recording was performed at 4 Hz during the last 10 s of the pulse and 50 s following removal of the stimulus. Starved worms were transferred to an assay plate with no food before imaging. Recorded image stacks were aligned using the StackReg plugin (Thevenaz et al., 1998) for ImageJ (rigid body option) and were cropped to a region containing the AWC cell body. Relative changes in fluorescence, following background subtraction, were calculated using custom MATLAB scripts (Source code 5, 6). Individual traces were normalized to their average baseline value for the five seconds prior to odorant removal.

References

  1. 1
    Dauer formation induced by high temperatures in Caenorhabditis elegans
    1. M Ailion
    2. JH Thomas
    (2000)
    Genetics 156:1047–1067.
  2. 2
    daf-12 regulates developmental age and the dauer alternative in Caenorhabditis elegans
    1. A Antebi
    2. JG Culotti
    3. EM Hedgecock
    (1998)
    Development (Cambridge, England) 125:1191–1205.
  3. 3
    daf-12 encodes a nuclear receptor that regulates the dauer diapause and developmental age in C. elegans
    1. A Antebi
    2. WH Yeh
    3. D Tait
    4. EM Hedgecock
    5. DL Riddle
    (2000)
    Genes & Development 14:1512–1527.
  4. 4
  5. 5
    The genetics of feeding in Caenorhabditis elegans
    1. L Avery
    (1993)
    Genetics 133:897–917.
  6. 6
  7. 7
    Control of larval development by chemosensory neurons in Caenorhabditis elegans
    1. CI Bargmann
    2. HR Horvitz
    (1991)
    Science (New York, N.Y.) 251:1243–1246.
  8. 8
  9. 9
  10. 10
    Nuclear calcium signaling
    1. CP Bengtson
    2. H Bading
    (2012)
    Advances in Experimental Medicine and Biology 970:377–405.
    https://doi.org/10.1007/978-3-7091-0932-8_17
  11. 11
  12. 12
  13. 13
  14. 14
  15. 15
  16. 16
  17. 17
  18. 18
  19. 19
  20. 20
  21. 21
  22. 22
    egl-4 acts through a transforming growth factor-beta/SMAD pathway in Caenorhabditis elegans to regulate multiple neuronal circuits in response to sensory cues
    1. SA Daniels
    2. M Ailion
    3. JH Thomas
    4. P Sengupta
    (2000)
    Genetics 156:123–141.
  23. 23
    Phenotypic Plasticity: Functional and Conceptual Approaches
    1. TJ DeWitt
    2. SM Scheiner
    (2004)
    Oxford University Press.
  24. 24
  25. 25
  26. 26
  27. 27
  28. 28
  29. 29
  30. 30
  31. 31
  32. 32
  33. 33
    ASI regulates satiety quiescence in C. elegans
    1. T Gallagher
    2. J Kim
    3. M Oldenbroek
    4. R Kerr
    5. YJ You
    (2013)
    The Journal of Neuroscience : The Official Journal of the Society for Neuroscience 33:9716–9724.
    https://doi.org/10.1523/JNEUROSCI.4493-12.2013
  34. 34
    Two pleiotropic classes of daf-2 mutation affect larval arrest, adult behavior, reproduction and longevity in Caenorhabditis elegans
    1. D Gems
    2. AJ Sutton
    3. ML Sundermeyer
    4. PS Albert
    5. KV King
    6. ML Edgley
    7. PL Larsen
    8. DL Riddle
    (1998)
    Genetics 150:129–155.
  35. 35
  36. 36
  37. 37
  38. 38
  39. 39
  40. 40
  41. 41
    daf-2, daf-16 and daf-23: genetically interacting genes controlling Dauer formation in Caenorhabditis elegans
    1. S Gottlieb
    2. G Ruvkun
    (1994)
    Genetics 137:107–120.
  42. 42
  43. 43
  44. 44
    Developmental regulation of a cyclin-dependent kinase inhibitor controls postembryonic cell cycle progression in Caenorhabditis elegans
    1. Y Hong
    2. R Roy
    3. V Ambros
    (1998)
    Development (Cambridge, England) 125:3585–3597.
  45. 45
  46. 46
  47. 47
  48. 48
  49. 49
  50. 50
  51. 51
  52. 52
  53. 53
    Mutants of a temperature-sensitive two-P domain potassium channel
    1. MT Kunkel
    2. DB Johnstone
    3. JH Thomas
    4. L Salkoff
    (2000)
    The Journal of Neuroscience : The Official Journal of the Society for Neuroscience 20:7517–7524.
  54. 54
  55. 55
  56. 56
  57. 57
  58. 58
    Ascaroside signaling in C. elegans
    1. AH Ludewig
    2. FC Schroeder
    (2013)
    WormBook pp. 1–22.
  59. 59
    Animal Species and Evolution
    1. E Mayr
    (1963)
    Cambridge, MA: Belknap Press of Harvard University Press.
  60. 60
  61. 61
  62. 62
    Social polymorphism in Hymenoptera
    1. CD Michener
    (1961)
    Symposium of the Royal Entomological Society of London pp. 43–56.
  63. 63
  64. 64
  65. 65
  66. 66
  67. 67
  68. 68
  69. 69
  70. 70
  71. 71
  72. 72
  73. 73
  74. 74
  75. 75
  76. 76
  77. 77
  78. 78
  79. 79
  80. 80
  81. 81
  82. 82
  83. 83
  84. 84
  85. 85
  86. 86
  87. 87
  88. 88
  89. 89
  90. 90
  91. 91
  92. 92
  93. 93
  94. 94
  95. 95
  96. 96
    A pyramid approach to subpixel registration based on intensity
    1. P Thévenaz
    2. UE Ruttimann
    3. M Unser
    (1998)
    IEEE Transactions on Image Processing : A Publication of the IEEE Signal Processing Society 7:27–41.
    https://doi.org/10.1109/83.650848
  97. 97
    Evidence for parallel processing of sensory information controlling dauer formation in Caenorhabditis elegans
    1. JH Thomas
    2. DA Birnby
    3. JJ Vowels
    (1993)
    Genetics 134:1105–1117.
  98. 98
    Distribution and intracellular localization of a mouse homologue of Ca2+/calmodulin-dependent protein kinase Ibeta2 in the nervous system
    1. T Ueda
    2. H Sakagami
    3. K Abe
    4. I Oishi
    5. A Maruo
    6. H Kondo
    7. T Terashima
    8. M Ichihashi
    9. H Yamamura
    10. Y Minami
    (1999)
    Journal of Neurochemistry 73:2119–2129.
  99. 99
    Genetic analysis of chemosensory control of dauer formation in Caenorhabditis elegans
    1. JJ Vowels
    2. JH Thomas
    (1992)
    Genetics 130:105–123.
  100. 100
  101. 101
  102. 102
  103. 103
  104. 104
  105. 105
  106. 106
  107. 107
  108. 108

Decision letter

  1. Oliver Hobert
    Reviewing Editor; Columbia University, United States

In the interests of transparency, eLife includes the editorial decision letter and accompanying author responses. A lightly edited version of the letter sent to the authors after peer review is shown, indicating the most substantive concerns; minor comments are not usually included.

[Editors’ note: a previous version of this study was rejected after peer review, but the authors submitted for reconsideration. The first decision letter after peer review is shown below.]

Thank you for choosing to send your work entitled “Food-dependent regulation of developmental plasticity via CaMKI and neuroendocrine signaling” for consideration at eLife. Your full submission has been evaluated by a Senior Editor, a Reviewing Editor and two peer reviewers and the decision was reached after discussions between the reviewers. Based on our discussions and the individual reviews below, we regret to inform you that your work will not be considered at this time for publication in eLife.

There was an overall sense that the manuscript is an assembly of too many “loose ends” that emerge from multiple interesting observations-the dauer phenotype of cmk-1, the AWC ablation cmk-1 suppression data, the correlation of apparent increased basal neuronal activity of AWC in starved WT and fed cmk-1 mutants, cell-autonomous activity of cmk-1 in ASI and possible cell-non-autonomous regulation of insulin expression in ASJ. These observations did not coalesce into a coherent picture with a clear take-home message of broad interest. As you can see in the reviews, a number of suggestions were made for additional experiments. If additional data can shed further light on the phenomenon, we do encourage you to submit the manuscript to eLife again, but it would be a new submission.

Reviewer #1:

The authors use genetics and imaging to probe how C. elegans neural circuits integrate food signals into the dauer developmental decision.

1) I am wondering if cmk-1 mutants might have altered sensitivity to pheromones and not just altered responses to food. Is there an additional assay that can be used to probe cmk-1 responses to pheromones?

2) The starvation-driven responses in AWC are very interesting. These seem to be much much weaker than those initiated by odors. Do these signals have any significance? It is very interesting that cmk-1 mutants have altered patterns. Moreover, starvation is still able to modulate responses in cmk-1 mutants (Figure 2). This circuit is likely to be more complex.

3) The imaging responses are highly variable. I would suggest testing different odor concentrations to get a better handle on the responses.

4) Is the nuclear localization of cmk-1 required to drive expression of insulins (35 and 26)?

5) Is daf-2 acting downstream of ins-35 and 26? Also, where is daf-2 required? The model predicts that daf-2 might be required in ASJ to integrate food signals?

6) Fluorescence measurements of daf-28-gfp and daf-7-gfp are highly variable. I am wondering if the authors could use an independent assay to test the genetic interactions (perhaps quantitative PCR).

Reviewer #2:

The manuscript by Neal et al. describes the characterization of CMK-1 (calcium/calmodulin-dependent protein kinase I) and its role in regulating the dauer developmental decision in C. elegans. The authors find that the expression of a number of different neuroendocrine peptide signals, which are known to influence the dauer decision by modulating insulin and TGF-β signalling pathways, is attenuated in the cmk-1 mutant. The authors use neuron-specific rescue experiments to implicate the function of cmk-1 in two pairs of sensory neurons, AWC and ASI, in the regulation of the dauer decision. The data are presented clearly with relevant statistical analysis. The authors' experimental system provides an interesting opportunity to dissect the mechanisms of neuroendocrine physiology.

The basic narrative, highlighted in the Abstract and interpretation of the results, is that the expression of neuroendocrine peptides is attenuated in subsets of sensory neurons in the cmk-1 mutant (ins-26, ins-35 in AWC, daf-7 in ASI, daf-28 in ASJ). The similarity in expression changes that are also observed in wild type animals under starvation conditions lead the authors to suggest that CMK-1 functions in a food-dependent manner to regulate neuroendocrine responses. However, this basic narrative is at odds with the strong evidence provided by the authors that in the cmk-1 mutant, there is an aberrant AWC neuronal state that gives rise to the production of pro-dauer signals. Genetic ablation of the AWC neurons suppresses the cmk-1 dauer phenotype (while ablation of AWC, if anything, enhances dauer formation in wild type animals, under the specific assay conditions employed). This casts considerable doubt on the functional relevance of the data on the ins-26 and ins-35 peptides, and moreover, suggests that instead of relaying food-dependent signals to the C. elegans nervous system, CMK-1 may have an essential role in maintaining the function of AWC.

The authors show that knockdown of the proprotein convertase bli-4 in AWC suppresses the cmk-1 dauer phenotype, and the authors suggest that insulin signals from AWC promote dauer formation under limiting food. In addition, the authors show that the AWC neurons are more basally active and more responsive to IAA in starved animals or well-fed cmk-1 mutants as compared to fed wild type animals (Figure 2A and 2C), and subsequently conclude, “CMK-1 regulates AWC activity as a function of food, and that deregulated AWC activity in fed cmk-1 mutants promotes inappropriate entry into the dauer stage.” This seems to get in the way of the overall narrative that CMK-1 has as key role in maintaining anti-dauer signals in response to food. Might ablation of AWC also suppress the diminished daf-28::GFP expression observed in the ASJ neurons of the cmk-1 mutant? Does blocking this aberrant pattern of activity, without ablating AWC itself, suppress or enhance the cmk-1 mutant?

If in fact the aberrant AWC activity leads to the production of pro-dauer signals, then one is not questioning that there are also real differences in the anti-dauer signals in the cmk-1 mutant. It's just that their functional relevance, and correlating expression in the cmk-1 mutant with what is observed in starved wild type animals, may not yield an accurate picture of what CMK-1 actually does in normal physiology. That is, if the absence of CMK-1 simply gives rise to dysfunctional AWC neurons that produce increased levels of pro-dauer signals, then it is more difficult to infer, based on the genetic analysis alone, a pivotal physiological role for CMK-1 in the expression of neuroendocrine signals in response to food.

[Editors’ note: what now follows is the decision letter after the authors submitted for further consideration.]

Congratulations: we are very pleased to inform you that your article, “Feeding state-dependent regulation of developmental plasticity via CaMKI and neuroendocrine signaling”, has been accepted for publication in eLife. The Reviewing Editor for your submission was Oliver Hobert.

Reviewer #1:

The revised manuscript of Neal et al. has satisfactorily addressed my principal concern of the initial manuscript that in cmk-1 mutants, an aberrant AWC state promotes dauer formation in a non-physiological manner. The authors show that wild type worms on live bacteria, in the presence of presumably high doses of pheromones (ascr#3 and 5), can form dauers, allowing them to show that the ablation of AWC can suppress dauer formation in wild type animals, and thus implying that AWC generates pro-dauer signals under physiological conditions, in a manner that appears to be particularly inducible by/sensitive to pheromone levels. The authors indicate these pro-dauer signals. They then go a step further and provide evidence that ins-32 may even be this signal-function, in parallel to the original and overriding subject of their manuscript, which is that CMK-1 regulates both ILP expression to modulate daf-28 levels in ASJ, as well as daf-7 levels in ASI. Given the relatively large effects of removing daf-7 and ASI on dauer formation, compared with the modest effects of removing daf-28, genetically dissecting the contribution of AWC signalling to the cmk-1 phenotype was not straightforward, but the authors have put together a thoughtful explanation of how CMK-1 functions as part of the sensorineural response to food to modulate entry into dauer diapause. The authors show that the intriguing elevated basal activity of the AWC neurons in the cmk-1 mutant, which partially mimic what the authors see upon starvation, essentially attenuates dauer formation; importantly this is shown in wild type animals, as the authors use the HisCl1 system to show that dauer formation is enhanced in both wild type and cmk-1 mutant backgrounds on live food (and pheromone). I would add that the meaning of this elevated basal activity remains to be determined, but the observation is quite striking! The authors provide an accompanying text that is precise in stating what has been shown and what has not been shown, which builds a credible case that CMK-1 functions in at least two different pairs of neurons to regulate a feeding state-dependent developmental decision.

Reviewer #2:

The authors use genetics and imaging to identify a role for cam kinase in distinct sensory neurons to drive a dauer developmental decision. This is a beautiful study and I only have a few comments before publication.

1) The authors use the missense allele oy21 instead of the putative null oy20 in most of their experiments. I would not suggest repeating the experiments, but would recommend some clarification. Perhaps, the missense is non-functional and can be identified bioinformatically.

2) I am wondering if AWA rescue of cmk-1 has any effect. The promoter the authors use in Figure 1 expresses in both AWA and ASI.

3) The authors state that ins-26 and ins-35 is expressed in AWC. Maybe I am not reading panel 5B correctly, but it looks like these genes are expressed by 50-70% of the animals. I would suggest that the authors discuss this issue in the text.

4) The authors show that cmk-1 is localized to the AWC nuclei after 30 minutes of starvation. However, the increase in AWC calcium occurs after 2h of starvation. I am wondering about the differences between the two assays and how this affects their model. Perhaps this could be included in the Discussion.

https://doi.org/10.7554/eLife.10110.034

Author response

[Editors’ note: the author responses to the first round of peer review follow.]

We are now submitting a completely revised manuscript incorporating data from new results. Based on issues raised by the reviewers, we have now performed several additional experiments that have led us to further develop and refine the proposed model and manuscript. As detailed in the attached ‘Response to Reviewers’, we have now addressed almost all comments from the reviewers, have included data from many new experiments, and completely reorganized and rewritten parts of the manuscript including a new Discussion.

Reviewer #1:

The authors use genetics and imaging to probe how C. elegans neural circuits integrate food signals into the dauer developmental decision.

1) I am wondering if cmk-1 mutants might have altered sensitivity to pheromones and not just altered responses to food. Is there an additional assay that can be used to probe cmk-1 responses to pheromones?

The reviewer raises a valid point. As discussed in the manuscript, food and pheromone signals are weighted and integrated in the dauer decision. Since pheromone and food antagonize each other in this developmental decision, decreased food sensitivity can also be read out as higher pheromone sensitivity (since animals will form more dauers at a given pheromone concentration if their threshold for food sensitivity is increased) and vice versa.

We specifically addressed this issue in the manuscript. We and others previously have shown that the ASK neurons are required to sense ascr#2 and ascr#3 in the dauer decision (1, 2). If CMK-1 acts simply by altering the sensitivity of ASK to either ascaroside, we might expect that CMK-1 would act in ASK to regulate the dauer decision. In Figure 1E, we showed that this is not the case, suggesting that CMK-1 does not act simply by modifying pheromone sensitivity at the level of sensory neuron responses.

However, to test this issue further, we performed another experiment. Although there are no other readouts for pheromone sensitivity in larvae, adult hermaphrodites have previously been shown to avoid ascr#3 via the ADL sensory neurons (3). We examined ascr#3-induced intracellular calcium dynamics in adult ADL neurons in wild-type and cmk-1 mutants, and found that cmk-1 mutants are actually somewhat less sensitive to ascr#3 than wild-type animals (Image 1). We chose to not include these data in the manuscript since we wished to keep the work focused on mechanisms of dauer development, and the relationship of the ADL ascr#3 responses in adult hermaphrodites to pheromone responses in dauer development are unclear. Thus, although we cannot completely rule out that pheromone sensitivity is increased in cmk-1 mutants in the dauer decision, taken together with additional data presented in the manuscript suggesting that the state of fed cmk-1 mutants is similar to that of starved wild-type animals, we conclude that CMK-1 alters sensitivity to food but not to pheromones.

Author response image 1

Responses of ADL neurons in WT and cmk-1(oy21) mutants to ascr#3. Panels show average calcium responses of ADL to the indicated concentrations of ascr#3. Errors are SEM. Also shown is a scatter plot of peak fluorescence amplitudes of individual neuron responses in WT and cmk-1 mutants to ascr#3. Horizontal black bars are the median. * and *** indicate P<0.05 and 0.001, respectively.

https://doi.org/10.7554/eLife.10110.032

2) The starvation-driven responses in AWC are very interesting. These seem to be much much weaker than those initiated by odors. Do these signals have any significance? It is very interesting that cmk-1 mutants have altered patterns. Moreover, starvation is still able to modulate responses in cmk-1 mutants (Figure 2). This circuit is likely to be more complex.

It is possible that the AWC neurons are more depolarized in cmk-1 mutants or in starved wild-type animals, leading to increased basal stochastic responses. However, verification of this hypothesis will require quantification of resting membrane potential which is beyond the scope of this work. However, starvation does not further modulate cmk-1 responses in AWC (revised Figure 6B, 6C, 6D, new Figure 6–figure supplement 1; please also see response to point 3 below).

The reviewer is correct in surmising that the circuit is more complex. We have now further examined the consequences of increased neuronal activity on dauer formation. We find that suppressing activity by growing animals expressing the Drosophila histamine-gated chloride channel in AWC (HisCl1) (4) on histamine enhances dauer formation in both wild-type and cmk-1 mutants (new Figure 6E). This result suggests that increased activity acts in parallel with the CMK-1 pathway to antagonize dauer formation. We speculate that this increased activity upon prolonged starvation in wild-type animals is a consequence of feedback from internal state that is deregulated in cmk-1 mutants, and that this activity also serves to integrate environmental signals on longer timescales to correctly regulate dauer formation (please see revised Figure 7, revised Discussion, and response to Reviewer 2).

3) The imaging responses are highly variable. I would suggest testing different odor concentrations to get a better handle on the responses.

Variability in AWC responses to odorant stimuli has been observed previously [see for example Figure S1b, S1h in (5)]. As suggested by the reviewer, we have now examined responses in AWC to the removal of two concentrations of isoamyl alcohol (new Figure 6C, 6D). In addition, we examined responses to the removal of one concentration of benzaldehyde, another AWC-sensed odorant (new Figure 6–figure supplement 1). Responses remain variable regardless of the odorant or its concentration, indicating that this variability may be a feature of AWC responses. Similarly, Luo et al. recently observed variability in the responses of the ASE sensory neurons to salt concentration changes that are not dependent on the specific salt concentration tested (6). To provide a clearer description, we have now included a statistical analysis of the peak odorant response amplitudes and number of responding cells in the main figure (revised Figure 6D – replaces previous Table S1; also see new Figure 6–figure supplement 1).

4) Is the nuclear localization of cmk-1 required to drive expression of insulins (35 and 26)?

We addressed this issue by expressing CMK-1+NLS and CMK-1+NES in AWC in cmk-1(oy21) mutants and quantifying ins-35p::yfp expression in this neuron type under fed and starved conditions. We find that expression of CMK-1+NLS suppresses the downregulated ins-35p::yfp expression phenotype of cmk-1(oy21) mutants in AWC in both fed and starved conditions to a slightly greater extent than expression of CMK-1+NES (revised Figure 5B). These results suggest that nuclear localization of CMK-1 is required to drive expression of this ILP gene.

5) Is daf-2 acting downstream of ins-35 and 26? Also, where is daf-2 required? The model predicts that daf-2 might be required in ASJ to integrate food signals?

This is a good question. As the reviewer is no doubt aware, mutations in daf-2 lead to pleiotropic effects including effects on dauer formation. Determining whether DAF-2 acts downstream of ins-26/35 in ASJ requires cell-specific knockdown of daf-2 in ASJ. However, we have had some issues (no knockdown, non-specific effects) knocking down a subset of genes in a neuron-specific manner. We are confident of the data presented in current Figure 5–figure supplement 2 since we observed effects with bli-4 but not egl-3 knockdown, and moreover, we observed effects upon bli-4(RNAi) using multiple promoters. Our model also does not require that INS-26/35 act directly on ASJ; these ILPs could act via intermediate neurons and other signaling pathways. We have now made this point clear in the model presented in revised Figure 7. We also expect that multiple ILPs may signal to ASJ, complicating interpretation of the dauer phenotypes of ASJ-specific daf-2 knockdown. Given these caveats, at this time, we would prefer to address the site of INS-26/35 action in future experiments.

6) Fluorescence measurements of daf-28-gfp and daf-7-gfp are highly variable. I am wondering if the authors could use an independent assay to test the genetic interactions (perhaps quantitative PCR).

We cannot easily measure expression differences via qPCR from whole animals since daf-28 is expressed in at least two neuron types, and effects are observed only in ASJ in cmk-1 mutants. Variability in daf-7 expression has been reported previously (7, 8), and this variability is thought to be one component of the neural code reporting food abundance (7). However, to address this very valid point, we performed single molecule FISH (smFISH) to examine daf-7 expression in wild-type and cmk-1 mutants. As shown in Figure 2 here, we find that expression of daf-7 remains variable, suggesting that consistent with previous observations, this variability in expression is a feature of the system. Thus, we have opted to retain the fluorescent reporter measurements in the manuscript.

Author response image 2

Expression of daf-7 in ASI analyzed via smFISH. A) Maximum and mean pixel intensity of daf-7 expression in ASI neurons. Median is indicated by red bars. B) Representative images showing the range of daf-7 smFISH signals in ASI. Yellow arrowheads point to ASI cell bodies which are magnified in panels at right. Anterior is at left.

Wild-type L1 larvae were fixed and hybridized with a daf-7 probe as described previously (8). Fluorescence intensities were quantified from maximum projection images and background subtracted. n=19 neurons; 10 animals.

https://doi.org/10.7554/eLife.10110.033

Reviewer #2:

[…] The basic narrative, highlighted in the Abstract and interpretation of the results, is that the expression of neuroendocrine peptides is attenuated in subsets of sensory neurons in the cmk-1 mutant (ins-26, ins-35 in AWC, daf-7 in ASI, daf-28 in ASJ). The similarity in expression changes that are also observed in wild type animals under starvation conditions lead the authors to suggest that CMK-1 functions in a food-dependent manner to regulate neuroendocrine responses. However, this basic narrative is at odds with the strong evidence provided by the authors that in the cmk-1 mutant, there is an aberrant AWC neuronal state that gives rise to the production of pro-dauer signals. Genetic ablation of the AWC neurons suppresses the cmk-1 dauer phenotype (while ablation of AWC, if anything, enhances dauer formation in wild type animals, under the specific assay conditions employed). This casts considerable doubt on the functional relevance of the data on the ins-26 and ins-35 peptides, and moreover, suggests that instead of relaying food-dependent signals to the C. elegans nervous system, CMK-1 may have an essential role in maintaining the function of AWC.

The authors show that knockdown of the proprotein convertase bli-4 in AWC suppresses the cmk-1 dauer phenotype, and the authors suggest that insulin signals from AWC promote dauer formation under limiting food. In addition, the authors show that the AWC neurons are more basally active and more responsive to IAA in starved animals or well-fed cmk-1 mutants as compared to fed wild type animals (Figure 2A and 2C), and subsequently conclude, “CMK-1 regulates AWC activity as a function of food, and that deregulated AWC activity in fed cmk-1 mutants promotes inappropriate entry into the dauer stage.” This seems to get in the way of the overall narrative that CMK-1 has as key role in maintaining anti-dauer signals in response to food. Might ablation of AWC also suppress the diminished daf-28::GFP expression observed in the ASJ neurons of the cmk-1 mutant? Does blocking this aberrant pattern of activity, without ablating AWC itself, suppress or enhance the cmk-1 mutant?

If in fact the aberrant AWC activity leads to the production of pro-dauer signals, then one is not questioning that there are also real differences in the anti-dauer signals in the cmk-1 mutant. It's just that their functional relevance, and correlating expression in the cmk-1 mutant with what is observed in starved wild type animals, may not yield an accurate picture of what CMK-1 actually does in normal physiology. That is, if the absence of CMK-1 simply gives rise to dysfunctional AWC neurons that produce increased levels of pro-dauer signals, then it is more difficult to infer, based on the genetic analysis alone, a pivotal physiological role for CMK-1 in the expression of neuroendocrine signals in response to food.

The insightful points raised by the reviewer prompted us to perform a number of additional experiments to clarify our model and, we hope, to present a more nuanced and developed manuscript. If we may paraphrase the reviewer’s comments, the major concern is that despite the overall similarities in phenotypes between starved wild-type and fed cmk-1 mutants suggesting a role for CMK-1 in maintaining anti-dauer signals from AWC in response to food, AWC may instead simply be ‘dysfunctional’ in cmk-1 mutants, resulting in the production of an aberrant pro-dauer signal. In particular, the fact that AWC ablation increases dauer formation in wild-type (suggesting the presence of an anti-dauer signal), whereas AWC ablation suppresses dauer formation in cmk-1 mutants (suggesting a possibly aberrant pro-dauer signal) is of concern. As a consequence the precise role of CMK-1 in the regulation of food-dependent neuroendocrine signals is unclear.

In brief, based on the reviewer’s points, we set out to determine whether AWC sends pro- and anti-dauer signals in both wild-type and cmk-1 mutants as a function of feeding state, and whether it is this balance between these antagonistic signals that is disrupted in cmk-1 mutants, as opposed to the cmk-1 phenotypes arising from an aberrant state that may not provide accurate information about the wild-type condition. We also further analysed the role of the increased basal activity observed in AWC in cmk-1 mutants.

a) We previously suggested that ins-26/35 represent the anti-dauer signal from AWC based in part on the observed downregulation of expression of both genes in AWC in starved wild-type or fed cmk-1 mutants. Based on this hypothesis, we would predict that:

i) ins-26/35 mutants should form dauers under conditions in which wild-type animals do not

ii) loss of ins-26/35 in the cmk-1 mutant background should not further enhance dauer formation (since both genes are already downregulated in cmk-1 mutants)

iii) overexpression of ins-26/35 from AWC should rescue the enhanced dauer formation phenotype and restore daf-28p::gfp expression in ASJ in cmk-1 mutants.

We previously presented data supporting ii) and iii) (current Figures 5D, 5E). We now also show that ins-26/35 double mutants do indeed increase dauer formation (new Figure 5C). We also show that CMK-1+NLS promotes ins-35 expression in AWC in both the fed and starved state (revised Figure 5B), and rescues dauer formation to a greater extent than expression of CMK-1+NES (current Figure 4C). Together, these results support the hypothesis that ins-26/35 represent the anti-dauer signal in AWC that is downregulated upon starvation in wild-type, and inappropriately down-regulated in fed cmk-1 mutants.

b) From the above model, we would then predict that ablation of AWC should increase dauer formation in wild-type animals (since anti-dauer signals would be abolished), but that this ablation in cmk-1 mutants should have no further effect on dauer formation (since the anti-dauer signals are already downregulated in cmk-1 mutants). As we reported previously, we observed that dauer formation is indeed enhanced in wild-type animals grown on heat-killed food (current Figure 5F). However, dauer formation is instead suppressed in cmk-1 mutants grown on live food (current Figure 5G), indicating that AWC also produces a pro-dauer signal. (As requested by the reviewer, we also now show that ablation of AWC suppresses the diminished daf-28p::gfp expression in ASJ phenotype of cmk-1 mutants.New Figure 5–figure supplement 1).

The reviewer raised the valid question of whether this pro-dauer signal is only observed as an aberrant feature of the cmk-1 mutation, or whether this signal is also present in wild-type animals. We realized that wild-type AWC neurons could also send a pro-dauer signal, but this signal may be masked by the different food conditions (heat-killed vs. live food) employed to examine dauer formation in wild-type and cmk-1 mutants, respectively. Dauer assays for cmk-1 mutants are performed on live food since they rapidly exit dauer on heat-killed food. Wild-type animals do not generally form dauers on live food allowing us to demonstrate the shifted food sensitivity threshold for cmk-1 mutants. However, live vs. heat-killed food provide different signals which may be sensed and integrated by distinct networks in the dauer decision, potentially confounding comparisons between wild-type and cmk-1 mutants.

We have now succeeded in establishing conditions in which a small number of wild-type animals form dauers on live food. As now shown in new Figure 5C, we find that under these conditions, ablation of AWC does indeed suppress dauer formation in wild-type animals. Consistent with this result, we also find that as in cmk-1 mutants, ablation of AWC increases daf-28p::gfp expression in ASJ in wild-type animals grown on live food (new Figure 5–figure supplement 1). Together, these observations indicate that AWC sends both a pro- and anti-dauer signal even in wild-type animals, with the balance between these signals being regulated by food and feeding state, and that this feeding state-dependent balance is disrupted in cmk-1 mutants. Please see the rewritten Discussion for further analysis.

c) We have now also identified a potential candidate for this pro-dauer signal. We noted previously that this pro-dauer candidate was likely to be a target of the BLI-4 proprotein convertase. Of the 19 predicted BLI-4 ILP targets (10), only ins-32 was shown to be expressed in AWC in adult hermaphrodites (11). We have now shown (new Figure 5C), that mutations in ins-32 not only suppress the enhanced dauer formation phenotype of cmk-1 mutants, but that they also inhibit dauer formation in wild-type animals on both live and heat-killed food. These new results imply that ins-32 is a candidate for the pro-dauer signal from AWC, and further imply that a balance between CMK-1-regulated anti- and pro-dauer signals from AWC regulate dauer formation.

d) Finally, we explored the relationship between increased activity in AWC in wild-type animals upon prolonged starvation, or in fed cmk-1 mutants, and dauer formation. As suggested by the reviewer, we asked whether suppression of AWC activity (not ablation) modifies dauer formation in wild-type and cmk-1 mutants. If the aberrant activity in cmk-1 mutants was causal to their increased dauer formation phenotype, we might expect that inhibition of this activity would suppress dauer formation in cmk-1 mutants but may not affect dauer formation in wild-type animals.

We suppressed AWC activity by growing transgenic wild-type and cmk-1(oy21) animals expressing the Drosophila histamine-gated chloride (HisCl1) channel in AWC on 10 mM histamine (4). Suppression of AWC activity in either wild-type or cmk-1 mutants resulted in increased dauer formation on live food in the presence of pheromone (new Figure 6E). This result suggests that increased AWC activity upon prolonged starvation in wild-type animals, or in cmk-1 mutants, antagonizes dauer formation. Moreover, since reducing activity enhances dauer formation similarly in both wild-type and cmk-1 mutants, these results also indicate that increased activity acts in parallel with the CMK-1 pathway to limit dauer formation.

We now suggest that the increased activity in AWC is a consequence of feeding state-regulated feedback from internal state that is deregulated in cmk-1 mutants. Given the long timescale of sensory signal integration in the dauer decision (12), we suggest that the availability of multiple pathways (CMK-1, activity) that integrate environmental cues on different timescales to regulate anti- and pro-dauer signals may be critical to correctly promote or suppress the dauer decision based on a temporally varying environment (please see revised Discussion).

To summarize, our new data indicate that AWC sends both pro- and anti-dauer signals based on environmental conditions, and the balance between these signals is altered in cmk-1 mutants leading to inappropriate dauer formation under well-fed conditions. Thus, CMK-1 does indeed have an essential role in maintaining AWC function in the context of correct integration of food signals into the dauer decision.

Summary of changes: Based on the new results stemming from the reviewers’ comments, we have revised the entire manuscript including reorganizing the Results section, and have completely rewritten the Discussion. We have also now performed several new experiments, and in addition to new data provided in this document (Figures 1 and 2), data from these experiments are included in the following new and revised Figures: Revised Figure 5B – new data added; New Figure 5C; Revised Figure 5D – new data added; New Figure 5–figure supplement 1; Revised Figure 6C, new Figure 6D – new data and analyses added; New Figure 6–figure supplement 1; New Figure 6E; Revised Figure 7.

References

1. Park D, O'Doherty I, Somvanshi RK, Bethke A, Schroeder FC, Kumar U, et al. Interaction of structure-specific and promiscuous G-protein-coupled receptors mediates small-molecule signaling in Caenorhabditis elegans. Proc Natl Acad Sci USA. 2012;109: 9917-9922.

2. Kim K, Sato K, Shibuya M, Zeiger DM, Butcher RA, Ragains JR, et al. Two chemoreceptors mediate developmental effects of dauer pheromone in C. elegans. Science. 2009;326: 994-998.

3. Jang H, Kim K, Neal SJ, Macosko EZ, Kim D, Butcher RA, et al. Neuromodulatory state and sex specify alternative behaviors through antagonistic synaptic pathways in C. elegans. Neuron. 2012;75: 585-592.

4. Pokala N, Liu Q, Gordus A, Bargmann CI. Inducible and titratable silencing of Caenorhabditis elegans neurons in vivo with histamine-gated chloride channels. Proc Natl Acad Sci USA. 2014;111: 2770-2775.

5. Chalasani SH, Chronis N, Tsunozaki M, Gray JM, Ramot D, Goodman MB, et al. Dissecting a neural circuit for food-seeking behavior in Caenorhabditis elegans. Nature. 2007;450: 63-70.

6. Luo L, Wen Q, Ren J, Hendricks M, Gershow M, Qin Y, et al. Dynamic encoding of perception, memory, and movement in a C. elegans chemotaxis circuit. Neuron. 2014;82: 1115- 1128.

7. Entchev EV, Patel DS, Zhan M, Steele AJ, Lu H, Ch'ng Q. A gene-expression-based neural code for food abundance that modulates lifespan. Elife. 2015;4: e06259.

8. Meisel JD, Panda O, Mahanti P, Schroeder FC, Kim DH. Chemosensation of bacterial secondary metabolites modulates neuroendocrine signaling and behavior of C. elegans. Cell. 2014;159: 267-280.

9. Yu YV, Bell HW, Glauser DA, Goodman MB, Van Hooser SD, Sengupta P. CaMKI- dependent regulation of sensory gene expression mediates experience-dependent plasticity in the operating range of a thermosensory neuron. Neuron. 2014;84: 919-926.

10. Leinwand SG, Chalasani SH. Neuropeptide signaling remodels chemosensory circuit composition in Caenorhabditis elegans. Nat Neurosci. 2013;16: 1461-1467.

11. Takayama J, Faumont S, Kunitomo H, Lockery SR, Iino Y. Single-cell transcriptional analysis of taste sensory neuron pair in Caenorhabditis elegans. Nucleic Acids Res. 2010;38: 131-142.

12. Schaedel ON, Gerisch B, Antebi A, Sternberg PW. Hormonal signal amplification mediates environmental conditions during development and controls an irreversible commitment to adulthood. PLoS Biol. 2012;10: e1001306.

https://doi.org/10.7554/eLife.10110.035

Article and author information

Author details

  1. Scott J Neal

    Department of Biology, National Center for Behavioral Genomics, Brandeis University, Waltham, United States
    Contribution
    SJN, SJN, Conception and design, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article
    Competing interests
    The authors declare that no competing interests exist.
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-1031-077X
  2. Asuka Takeishi

    Department of Biology, National Center for Behavioral Genomics, Brandeis University, Waltham, United States
    Contribution
    AT, AT, Acquisition of data, Analysis and interpretation of data
    Competing interests
    The authors declare that no competing interests exist.
  3. Michael P O'Donnell

    Department of Biology, National Center for Behavioral Genomics, Brandeis University, Waltham, United States
    Contribution
    MPO, MPO'D, Analysis and interpretation of data, Drafting or revising the article
    Competing interests
    The authors declare that no competing interests exist.
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0001-8313-8969
  4. JiSoo Park

    Department of Brain and Cognitive Sciences, Daegu Gyeongbuk Institute of Science and Technology (DGIST), Daegu, Republic of Korea
    Contribution
    JSP, JSP, Acquisition of data, Analysis and interpretation of data
    Competing interests
    The authors declare that no competing interests exist.
  5. Myeongjin Hong

    Department of Brain and Cognitive Sciences, Daegu Gyeongbuk Institute of Science and Technology (DGIST), Daegu, Republic of Korea
    Contribution
    MH, MH, Acquisition of data, Analysis and interpretation of data
    Competing interests
    The authors declare that no competing interests exist.
  6. Rebecca A Butcher

    Department of Chemistry, University of Florida, Gainesville, United States
    Contribution
    RAB, RAB, Contributed unpublished essential data or reagents
    Competing interests
    The authors declare that no competing interests exist.
  7. Kyuhyung Kim

    Department of Brain and Cognitive Sciences, Daegu Gyeongbuk Institute of Science and Technology (DGIST), Daegu, Republic of Korea
    Contribution
    KK, KK, Analysis and interpretation of data, Drafting or revising the article
    For correspondence
    khkim@dgist.ac.kr
    Competing interests
    The authors declare that no competing interests exist.
  8. Piali Sengupta

    Department of Biology, National Center for Behavioral Genomics, Brandeis University, Waltham, United States
    Contribution
    PS, PS, Conception and design, Analysis and interpretation of data, Drafting or revising the article
    For correspondence
    sengupta@brandeis.edu
    Competing interests
    The authors declare that no competing interests exist.

Funding

NIH Office of the Director (OD) (F32 DC013711)

  • Michael P O'Donnell

National Institutes of Health (NIH) (T32 NS007292)

  • Michael P O'Donnell

Human Frontier Science Program (HFSP) (RGY0042/2010)

  • Rebecca A Butcher
  • Piali Sengupta

DGIST R&D Programn of the Ministry of Science, ICT and Future Planning (MSIP) (15-BD-06)

  • Kyuhyung Kim

POSCO T.J. Park Foundation (TJ Park Science Fellowship)

  • Kyuhyung Kim

National Research Foundation of Korea (2012R1A1A2009385)

  • Kyuhyung Kim

National Science Foundation (IOS 1256488)

  • Piali Sengupta

National Institutes of Health (GM 081639)

  • Piali Sengupta

National Institutes of Health (GM 103770)

  • Piali Sengupta

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Acknowledgements

This work was supported in part by the NSF (IOS 1256488—PS), the NIH (GM 081639 and GM 103770 - PS, F32 DC013711 and T32 NS007292—MPO'D), the Human Frontiers Science Program (RGY0042/2010—PS and RAB), the DGIST R&D Program of the Ministry of Science, ICT and Future Planning (15-BD-06—KK), the National Research Foundation of Korea (NRF-2012R1A1A2009385—KK), and the T.J. Park Science Fellowship of POSCO T.J. Park Foundation (KK).

Reviewing Editor

  1. Oliver Hobert, Columbia University, United States

Publication history

  1. Received: July 15, 2015
  2. Accepted: July 31, 2015
  3. Version of Record published: September 3, 2015 (version 1)
  4. Version of Record updated: March 30, 2016 (version 2)

Copyright

© 2015, Neal et al.

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.

Metrics

  • 2,188
    Page views
  • 484
    Downloads
  • 10
    Citations

Article citation count generated by polling the highest count across the following sources: Scopus, Crossref, PubMed Central.

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Download citations (links to download the citations from this article in formats compatible with various reference manager tools)

Open citations (links to open the citations from this article in various online reference manager services)

Further reading

    1. Cancer Biology
    2. Developmental Biology
    Katarina Valoskova et al.
    Research Article