Gαo and Gαq regulate the expression of daf-7, a TGFβ-like gene, in Caenorhabditis elegans.

Caenorhabditis elegans enter an alternate developmental stage called dauer in unfavorable conditions such as starvation, overcrowding, or high temperature. Several evolutionarily conserved signaling pathways control dauer formation. DAF-7/TGFβ and serotonin, important ligands in these signaling pathways, affect not only dauer formation, but also the expression of one another. The heterotrimeric G proteins GOA-1 (Gα(o)) and EGL-30 (Gα(q)) mediate serotonin signaling as well as serotonin biosynthesis in C. elegans. It is not known whether GOA-1 or EGL-30 also affect dauer formation and/or daf-7 expression, which are both modulated in part by serotonin. The purpose of this study is to better understand the relationship between proteins important for neuronal signaling and developmental plasticity in both C. elegans and humans. Using promoter-GFP transgenic worms, it was determined that both goa-1 and egl-30 regulate daf-7 expression during larval development. In addition, the normal daf-7 response to high temperature or starvation was altered in goa-1 and egl-30 mutants. Despite the effect of goa-1 and egl-30 mutations on daf-7 expression in various environmental conditions, there was no effect of the mutations on dauer formation. This paper provides evidence that while goa-1 and egl-30 are important for normal daf-7 expression, mutations in these genes are not sufficient to disrupt dauer formation.

Introduction

Under unfavorable environmental conditions, developing Caenorhabditis elegans enter an alternative stage called dauer. In dauer, growth and feeding arrest. Dauer worms also have sealed orifices and form thickened cuticles. The metabolic and morphological changes that accompany dauer increase the likelihood of the animals’ survival under harsh conditions. The dauer stage is reversible, and larvae resume development when environmental conditions improve (reviewed in [1]).

Dauer formation is controlled in part by the DAF-7/TGFβ-like signaling pathway ([2] and reviewed in [1]). DAF-7 is expressed in the ASI sensory neurons and is required during larval development to inhibit dauer formation [3], [4]. Environmental cues such as starvation and high temperature that trigger dauer formation also downregulate daf-7 expression [3]. While several genes are required for normal daf-7 expression [5]–[7], the signaling pathways that control daf-7 expression and its sensitivity to environmental signals are still not well understood.

One of the genes required for both daf-7 expression and dauer formation encodes tryptophan hydroxylase, TPH-1 [6]. TPH-1 is the rate-limiting enzyme required for serotonin biosynthesis. Serotonin signals through the heterotrimeric G proteins GOA-1 and EGL-30 to control several C. elegans behaviors [8]–[11]. GOA-1 and EGL-30 share a high degree of homology with human Gαo and Gαq [12]. In the human nervous system, Gαo and Gαq act downstream of many neurotransmitters, including serotonin. In C. elegans, GOA-1 and EGL-30 also act upstream of tph-1 to regulate its expression [13]. It is possible then, that goa-1 and egl-30 are important for regulating daf-7 expression and dauer formation, and may do so by regulating either serotonin signaling or biosynthesis. These experiments explore, in a tractable model organism, a new relationship between evolutionarily conserved pathways and proteins important for neuronal signaling and developmental plasticity.

Results and Discussion

Gαo and Gαq do not Affect Morphology of daf-7-expressing Cells

daf-7 is expressed in two head sensory neurons called the ASIs [3], [4]. Structural changes in ASI cilia accompany dauer formation [14]. Two neuronal heterotrimeric G proteins, GPA-2 and GPA-3, affect dauer formation by affecting the sensory cilia of ASIs [15]–[17]. By altering the sensory cilia, GPA-2 and GPA-3 presumably affect the way C. elegans sense the environmental signals that regulate the dauer developmental switch. It is possible that additional G proteins such as GOA-1 and EGL-30 (which, unlike GPA-2 and GPA-3, have homologues in humans) could also primarily affect dauer formation or daf-7 expression through altering the morphology of ASI. To first determine whether ASI morphology was affected by mutations in either goa-1 or egl-30, gross neuronal structure was visualized using DiD. Worms were incubated in DiD, a lipophilic dye, that is only taken up by those sensory neurons making direct contact with the environment. Therefore, if ASI morphology were altered in either goa-1 or egl-30 mutants, the neurons would not fill with dye. ASIs in wild type, goa-1 (n1134) partial loss-of-function, goa-1(sa734) null, egl-30 (n686 and ad805) partial loss-of-function (lf) mutants as well as egl-30 (tg26 and js126) gain-of-function (gf) mutants filled with DiD (Figure 1 and data not shown). These worms also exhibited normal gross morphology of ASIs, suggesting that signaling through either goa-1 or egl-30 is not required for ASI development.

Figure 1

Mutations in goa-1 and egl-30 do not affect ASI development.

Representative photomicrographs showing DiD filling (first column) and pdaf-7::GFP (second column) in wild type, egl-30, and goa-1 mutant backgrounds. DiD filling was not significantly different between ASI neurons (indicated by white arrows) in wild type and mutant worms. ASI projections, visualized with pdaf-7::GFP, do not appear significantly altered in mutant worms. The last column shows a merge between DiD, GFP, and DIC images. pdaf-7::GFP strains used express GFP under the putative daf-7 promoter, and only express GFP in the ASI neurons [5].

Gαo and Gαq are Required for daf-7 expression but not Dauer Formation

daf-7 expression peaks during the first and second stages of larval development (L1 and L2; [3]) just before the dauer decision is made. When animals were raised in a favorable environment (with food at 20°C), all wild-type L1 larvae exhibited strong daf-7 expression. daf-7 expression was markedly reduced in larvae with (lf) alleles of goa-1 or egl-30 (Figure 2). In addition, daf-7 expression was reduced in larvae with egl-30 gf alleles.

Figure 2

goa-1 and egl-30 regulate daf-7 expression at multiple temperatures.

pdaf-7::GFP levels were measured in late L1 larvae raised at 20°C, 25°C or 27°C. goa-1 (sa734 null and n1134lf), egl-30 (n686lf and ad805lf), and egl-30 (js126gf and tg26gf) mutant larvae exhibited significantly less daf-7 expression when compared to wild type larvae (ksIs2) raised at 20°C. daf-7 expression was also significantly lower in all mutant larvae when compared to wild type larvae at both 25°C and 27°C. The decrease in daf-7 in goa-1(n1134lf) and egl-30(js126gf) mutants was significantly greater than in wild type larvae raised at 27°C. pdaf-7::GFP expression actually increased as temperature increased to 25°C in egl-30(n686lf), egl-30(ad805lf) and egl-30(tg26gf) mutants, and increased again at 27°C in egl-30(n686lf) and egl-30(ad805lf) mutants. * =  significant difference from wild type (ksIs2) worms at the same temperature (Student’s t-test, p<0.05). ** =  significant difference between treatments in the same genotype (Student’s t-test, p<0.05). #  =  significant interaction between genotype and change in temperature, as compared to wild type, ksIs2 larvae (ANOVA, p<0.05).

These data were unexpected for several reasons. First, one would expect that loss-of-function mutations and gain-of-function mutations in the same gene might have opposite effects on their target, in this case daf-7. However, both the (lf) and (gf) mutations in egl-30 decreased daf-7 expression.

The second unexpected result was that daf-7 expression was reduced in both goa-1 and egl-30 (lf) mutants. This was surprising because signaling through GOA-1 is thought to antagonize signaling through EGL-30. These two G proteins are thought to act antagonistically because they have opposite effects on many C. elegans behaviors [9], [18], [19]. In addition, goa-1 and egl-30 have opposite effects on tph-1 expression; goa-1 represses tph-1 expression while egl-30 promotes tph-1 expression [13]. Since tph-1 promotes daf-7 expression [6], one would expect that goa-1 (lf) mutations would cause an increase in daf-7 expression while egl-30 (lf) mutations might case a decrease in daf-7 expression. The data suggest that goa-1 and egl-30 are both required to maintain daf-7 expression, and that any perturbation to either signaling pathway results in decreased daf-7 expression. Because goa-1 and egl-30 are expressed in many cells throughout the worm, it is possible that goa-1 and egl-30 act in distinct subsets of cells that could have opposite effects on daf-7 expression. For instance, goa-1 could be required to activate a set of neurons that promotes daf-7 expression, while egl-30 could act to inhibit the activity of a set of neurons that inhibits daf-7 expression. In a scenario such as this, (lf) mutations in both goa-1 and egl-30 would result in decreased daf-7 expression. The decrease in daf-7 expression seen in goa-1 and egl-30 mutants at 20°C was not sufficient, however, to elicit dauer formation (Figure 3).

Figure 3

Neither goa-1 nor egl-30 inhibit dauer entry at high temperatures.

Dauer formation was assayed in larvae raised at 25°C or 27°C. None of the mutant strains tested exhibited a significant increase in dauer formation at higher temperatures. It was not possible to assay dauer formation in goa-1(sa734) null or egl-30 (tg26gf) mutants at 27°C, because these mutants formed partial dauers or arrested at the L1 stage. Data for these strains at 27°C were therefore not included in the figure. daf-7(e1372) mutants are constitutive dauers at 25°C and 27°C, and served as a positive control. * =  significant difference from wild type (N2) worms at the same temperature (Student’s t-test, p<0.05). ** =  significant difference between treatments in the same genotype (Student’s t-test, p<0.05). #  =  significant interaction between genotype and change in temperature (from 20°C), as compared to wild type larvae (ANOVA, p<0.05).

How precisely EGL-30 and GOA-1 regulate daf-7 expression may be difficult to elucidate because of a complex feedback loop that exists between daf-7 and tph-1. While TPH-1 upregulates daf-7 expression [6], DAF-7 downregulates tph-1 expression [20]. tph-1 expression is also elevated in dauer larvae [21] when daf-7 expression is low. It is unlikely that GOA-1 or EGL-30 act downstream of daf-7 to regulate tph-1 expression (and then daf-7 expression) because egl-30 and goa-1 are not necessary for the increase in tph-1 seen in dauer larvae [21]. GOA-1 and EGL-30 may instead be acting downstream or independently of tph-1 to regulate daf-7 expression.

Gαo and Gαq Alter Temperature-induced Changes in daf-7 Expression but not Dauer Formation

High temperatures can induce dauer formation in some mutants that do not readily form dauers at moderately high temperatures [22], [23]. It was possible that while goa-1 and/or egl-30 mutants did not form dauers at favorable temperatures (20°C, Figure 3), they would enter dauer at high temperatures. Moderately high (25°C) or high (27°C) temperatures were both insufficient to induce dauer formation in any of the G protein mutants tested (Figure 3). While most non-dauer worms developed into full adults, goa-1(sa734) null and egl-30(tg26gf) mutants did not. These non dauer worms appeared to arrest as larvae; either L1 or partial dauers (determined by SDS sensitivity). The larval arrest in these mutants occurred prior to the time in development when the dauer decision is made, and suggests that normal development at 27°C was disrupted by the sa734 and tg26 alleles.

Despite the absence of dauer formation seen at 25°C or 27°C, daf-7 expression was significantly altered in all mutant strains tested (Figure 2). In almost all strains, there was a significant difference in the way temperature affected daf-7 expression. These data suggest that GOA-1, and EGL-30 in particular, relay some sensory cues important for daf-7 expression. EGL-30 appears to be important for downregulating daf-7 in response to high temperatures. In fact, the effect of temperature on daf-7 expression is reversed in the egl-30(lf) mutants and exaggerated in the egl-30 (js126gf) mutant. Other signaling pathways likely contribute to the behavioral/developmental response to high temperature since the decrease in daf-7 expression in egl-30 (gf) mutants was not sufficient to induce dauer formation at high temperatures.

Gαo and Gαq Alter the Response to Limiting Amounts of Food

In addition to temperature, food availability affects the course of C. elegans development. When larvae hatch from eggs in the absence of food, their development arrests at the L1 stage (prior to the time in development when the dauer decision is made) and daf-7 expression is reduced in the arrested L1s [3]. Signaling through GOA-1 and EGL-30 modulates food sensitivity in adult C. elegans [24]–[26], so it is possible goa-1 and egl-30 are required for mediating the effect of food in larvae. As in wild-type L1s, starvation caused a significant reduction in daf-7 in goa-1 (lf) mutants (Figure 4). As expected, an equivalent reduction was seen in the egl-30(tg26gf) mutant and an increase was seen in the egl-30(n686lf) mutant. While these mutants still appeared to be in the arrested L1 stage, the data suggest that egl-30 mediates the daf-7 response to starvation. These results and those from studies in adult worms [25] suggest that EGL-30 plays the same role in the response to starvation throughout development.

Figure 4

goa-1 and egl-30 regulate daf-7 expression in response to starvation.

pdaf-7::GFP levels were measured in well-fed late L1 larvae raised at 20°C or starved L1s. daf-7 expression was lower with starvation in all genotypes except egl-30(n686lf), egl-30 (ad805lf) and egl-30(tg26gf), in which pdaf-7::GFP expression was unchanged or increased with starvation. The decrease in daf-7 in goa-1(sa734) null, and egl-30 (js126gf) mutants was significantly greater than in wild type larvae. * =  significant difference from wild type (ksIs2) worms at the same temperature (Student’s t-test, p<0.05). ** =  significant difference between treatments in the same genotype (Student’s t-test, p<0.05). #  =  significant interaction between genotype and change in temperature, as compared to wild type, ksIs2 larvae (ANOVA, p<0.05).

Reduced daf-7 expression was not seen in worms expressing the other egl-30 (gf) mutant allele (js126), however. The difference between the egl-30 (gf) phenotypes may be caused by differences in the way each mutation affects the EGL-30 protein. The tg26 (gf) allele is thought to contain a mutation that alters guanine nucleotide binding [27]. The js126 (gf) allele is thought to contain a mutation that alters GTPase activity [28]. Both tg26 and js126 mutants have (gf) phenotypes with respect to other EGL-30-dependent behaviors such as egg laying and movement [28], [29], however it is not clear whether it is reasonable to predict that both alleles would affect all ELG-30-dependent processes in the same way.

When larvae hatch in the presence of limiting amounts of food, they progress through the L1 stage and then enter dauer [30]. If egl-30 and goa-1 are important for relaying the food cues important for normal development, one would expect that mutations in either goa-1 or egl-30 would disrupt dauer formation caused by low food levels. When goa-1 and egl-30 mutant larvae were exposed to limiting amounts of food, they did arrest at an early stage of development (Figure 5). However, based on size, most larvae appeared to be arrested at the L1 or partial dauer stages and not as full dauers. Most of the small larvae exhibited pharyngeal pumping and were sensitive to SDS, indicating that they did not arrest as full dauers [31]. These data suggest that goa-1 and egl-30 mutant worms are still sensitive to alterations in food availability, because they did not fail to arrest development in the presence of limiting food.

Figure 5

Neither goa-1 nor egl-30 are required for dauer formation in response to reduced food.

Dauer formation was assayed in larvae grown at 25°C on several concentrations of E. coli OP50. Wild type larvae formed dauers as food concentration decreased. In at least one assay, non-dauer worms of egl-30 (tg26gf), egl-30 (n686lf), egl-30 (ad805lf), or goa-1 (n1134lf) genotype did not develop into adults when grown on 1% food. Instead, these larvae were arrested as L1s or partial dauers. Data for these strains at 1% food concentration were therefore not included in the figure. * =  significant difference from wild type (N2) worms at the same food concentration (Student’s t-test, p<0.05). ** =  significant difference between treatments in the same genotype (Student’s t-test, p<0.05).

Overall, the experiments in this study showed that goa-1 and egl-30 regulate daf-7 expression in early development. While goa-1 and egl-30 mutations significantly decreased daf-7 expression, they did not affect dauer formation. These results suggest that other signaling pathways act in concert with GOA-1 and EGL-30 to decrease daf-7 expression to levels sufficient to induce dauer formation.

Materials and Methods

Worm Strains

C. elegans worm strains were maintained on NGM plates with Escherichia coli OP50 as the food source [32]. Strains were provided by the Caenorhabditis Genetic Center (CGC) and were derived from the wild-type N2 Bristol strain. Strains used were as follows: N2, JT734 goa-1(sa734), KO96 goa-1(n1134lf), MT1434 egl-30(n686sd), NM1380 egl-30(js126gf), KY26 egl-30(tg26gf), DA823 egl-30 (ad805), and CB1372 daf-7(e1372ts). Strains containing G protein mutations were crossed into the FK181 ksIs2 [pdaf-7::GFP, rol-6(su1006)] strain for pdaf-7::GFP analysis.

Microscopy

For all assays, the developmental stages of larvae were carefully synchronized. For temperature assays, gravid adults laid eggs on NGM plates for 4 hours. Adults were removed and eggs were grown on NGM plates for 18–24 hours. For starvation assays, gravid adults were bleached to isolate eggs. Eggs were grown on NGM plates for 18–24 hours (fed) or in M9 medium for 48 hours (starved). Larvae were transferred to a 4% agarose pad on a microscope slide, immobilized with 10 mM levamisole, and viewed using a Leica DM5500 microscope. ASI images were captured with a fixed exposure time using a Hammamatsu Orca ER camera and Leica Microsystems Image capture software. GFP intensity was quantified using NIH Image J software version 1.44o. The intensity of GFP in each ASI cell body was quantified. The intensity of a similarly sized background selection was subtracted from the ASI GFP intensity to get the adjusted GFP intensity. Approximately ten larvae of each genotype were imaged in each experiment. Experiments were performed in triplicate, on three separate days. Dye filling was performed using 0.1 mg/ml DiD (Molecular Probes) as described [33].

Dauer Assays

Dauer assays were done similarly to those previously described [30]. Modified NGM plates were prepared without peptone, and Noble agar (Difco) was used. 3 ml of modified NGM was used in each dauer assay plate. Plates were seeded with 20 µl of 4% (w/v) OP50, unless otherwise noted. E. coli OP50 was resuspended in S Medium with 50 µg/ml streptomycin. Larvae were synchronized by first isolating eggs from bleached, gravid adults. Eggs were resuspended in S Medium. Approximately 100 eggs were pipetted onto seeded dauer plates that were incubated for 60–72 hours. Dauer worms were scored visually, and scoring was confirmed using SDS. Worms were considered dauers if they survived a several minute incubation in 1%SDS. Larvae were considered partial dauers if they were the same size and shape as dauer larvae, but exhibited pharyngeal pumping and did not survive SDS treatment.

Assays were performed with 100–200 worms for each genotype. All genotypes were tested in an assay. Assays were performed in triplicate, on three separate days.