Feeding state regulates pheromone-mediated avoidance behavior via the insulin signaling pathway in Caenorhabditis elegans

Abstract

Animals change sensory responses and their eventual behaviors, depending on their internal metabolic status and external food availability. However, the mechanisms underlying feeding state-dependent behavioral changes remain undefined. Previous studies have shown that Caenorhabditis elegans hermaphrodite exhibits avoidance behaviors to acute exposure of a pheromone, ascr#3 (asc-ΔC9, C9). Here, we show that the ascr#3 avoidance behavior is modulated by feeding state via the insulin signaling pathway. Starvation increases ascr#3 avoidance behavior, and loss-of-function mutations in daf-2 insulin-like receptor gene dampen this starvation-induced ascr#3 avoidance behavior. DAF-2 and its downstream signaling molecules, including the DAF-16 FOXO transcription factor, act in the ascr#3-sensing ADL neurons to regulate synaptic transmission to downstream target neurons, including the AVA command interneurons. Moreover, we found that starvation decreases the secretion of INS-18 insulin-like peptides from the intestine, which antagonizes DAF-2 function in the ADL neurons. Altogether, this study provides insights about the molecular communication between intestine and sensory neurons delivering hunger message to sensory neurons, which regulates avoidance behavior from pheromones to facilitate survival chance.

Keywords: DAF‐2 insulin receptor; avoidance behavior; feeding state; pheromone; synaptic transmission.

Figure 1. An insulin/IGF‐1‐like receptor, daf‐2, is required for increase in ascr#3 avoidance under starvation conditions

1. Experimental scheme of ascr#3 avoidance assay depending upon starvation. 50–100 young adult animals are washed and placed on seeded (fed: colored in black) or non‐seeded (starved: colored in gray) plates for each duration: 3, 6, and 24 h; then, various concentrations of ascr#3 are delivered to the front of a freely moving forward animal to measure avoidance frequencies responding to ascr#3.

2. Fraction reversing of fed and starved animals in response to ascr#3 exposure. n = 50–70. ***P < 0.001 and ****P < 0.0001 (one‐way ANOVA Bonferroni's test).

3. Fraction reversing of re‐fed animals from 6‐h starvation in response to ascr#3 exposure. A 24‐h re‐feeding period reverses ascr#3 avoidances to well‐fed status. n = 40–80. *P < 0.05 (one‐way ANOVA Dunnett's Test).

4. Fraction reversing of daf‐2 mutant animals in fed and starved status in response to ascr#3 exposure. n = 40–60.

5. Fraction reversing of well‐fed wild‐type animals and daf‐2 mutants in response to 100, 200, and 400 nM ascr#3. n = 60. ***P < 0.001 (Bonferroni's test).

6. Fraction reversing of wild‐type animals, daf‐2 mutants, and daf‐2 mutants expressing unc‐14p::daf‐2 cDNA (neurons) or sre‐1p::daf‐2 cDNA (ADL) in response to 500 nM ascr#3. daf‐2 cDNA expression in neuron and ADL restores the defect of ascr#3 avoidance in daf‐2 mutants. n = 60. *P < 0.05 (Dunnett's test).

7. Fraction reversing of wild‐type animals, daf‐2 mutants, and daf‐2 mutants expressing sre‐1p::daf‐2 cDNA (ADL) in fed and starved conditions in response to ascr#3. n = 40–70. **P < 0.01 and ****P < 0.0001 (Bonferroni's test).

Data information: All error bars represent ± SEM.

Figure EV1. Effect of DAF‐2/insulin‐like receptor and feeding status on ascr#3 avoidance

1. Fraction reversing of wild‐type animals exhibiting short reversal (left) and omega turn under different feeding conditions (right). n = 40–80.

2. Fraction reversing of wild‐type animals and eat‐2 mutants, daf‐2 mutants, and eat‐2;daf‐2 double mutants under fed conditions. n = 80–110. **P < 0.01 and ****P < 0.0001 (Dunnett's Test).

3. Fraction reversing of daf‐2 alleles, e1368, and e1370 in response to 500 nM ascr#3 in fed conditions. n = 50–70. **P < 0.01 and ***P < 0.001 (Dunnett's test).

4. Fraction reversing of wild‐type and daf‐2 mutant animals on on‐food or off‐food conditions. n = 30–50. *P < 0.05 (Bonferroni's test).

5. The number of transgenic animals expressing sre‐1p::gfp in ADL and ASJ. n = 31.

Data information: All error bars represent ± SEM.

Figure 2. DAF‐2 affects Ca²⁺ transients not in the ADL ascr#3‐sensing neurons but their downstream target neurons

1. Ca²⁺ transients of ADL upon ascr#3 exposure in wild‐type animals and daf‐2 mutants. Ca²⁺ transients to 100 nM ascr#3 of ADL (left), the average peak percentage changes in fluorescence upon 100 nM ascr#3 exposure (middle), and the dose–response curve of the average peak percentage changes in fluorescence Ca²⁺ peaks upon 100 and 500 nM ascr#3 exposure (right). n = 7–12.

2. Ca²⁺ transients of AVA upon 500 nM ascr#3 exposure in wild‐type animals, daf‐2 mutants, and daf‐2 mutants expressing sre‐1p::daf‐2 cDNA (ADL). Ca²⁺ transients in response to ascr#3 in AVA (left), the average peak percentage changes in fluorescence upon 100 nM ascr#3 exposure (right). n = 10. ***P < 0.001 (Dunnett's test).

3. Heat‐map images of Ca²⁺ transients in AVA upon 500 nM ascr#3 exposure in wild‐type animals (left), daf‐2 mutants (middle), and daf‐2 mutants expressing sre‐1p::daf‐2 cDNA (right). Each row represents Ca²⁺ responses of individual animals to ascr#3 exposure. n = 10.

Data information: All error bars represent ± SEM.

Figure 3. DAF‐2 regulates synaptic transmissions in ADL

1. A

   Representative images of wild‐type animals (left) and daf‐2 mutants (right) expressing sre‐1p::snb‐1 cDNA::yfp (top) and sre‐1p::mCherry::rab‐3 cDNA (bottom). The scale bar is 10 μm.

2. B, C

   Relative fluorescence intensity of wild‐type animals and daf‐2 mutant animals expressing sre‐1p::snb‐1 cDNA::yfp (B) and sre‐1p::mCherry::rab‐3 cDNA (C). n = 59–60. ****P < 0.0001 (unpaired Student's t‐test).

3. D

   Relative fluorescence intensity of sre‐1p::snb‐1 cDNA::yfp of wild‐type animals in fed and starved conditions for 3, 6, and 24 h. n = 37–80. **P < 0.01 and ****P < 0.0001 (Bonferroni's test).

4. E

   Relative fluorescence intensity of sre‐1p::snb‐1 cDNA::yfp of daf‐2 mutants in fed and starved conditions. n = 20–30.

5. F

   Fraction reversing of wild‐type animals and daf‐2 mutant animals expressing sre‐1p::TeTx and sre‐1p::pkc‐1(gf). n = 50–90. *P < 0.05, **P < 0.01, and ***P < 0.001 (Dunnett's test). ⁺⁺ P < 0.01 (unpaired Student's t‐test).

Data information: All error bars represent ± SEM. Tops and bottoms of boxes indicate the 25^th and 75^th percentiles, respectively; whiskers represent 10^th–90^th percentile. Median is indicated by a horizontal line and the average is marked by “+” in the box.

Figure EV2. Quantification of SNB‐1::YFP and mCherry::RAB‐3 and gene expression of sre‐1 promoter upon daf‐2 mutation or starvation

1. A

   Relative fluorescence intensity of integrated animals expressing sre‐1p::snb‐1 cDNA::yfp. n = 30. ***P < 0.0001 (unpaired Student's t‐test).

2. B, C

   Quantification of fluorescence intensity of wild‐type and daf‐2 mutant animals expressing sre‐1p::snb‐1 cDNA::yfp (B) and sre‐1p::mCherry::rab‐3 cDNA (C) analyzed using ImageJ software. n = 13–41. ***P < 0.001 (unpaired Student's t‐test).

3. D

   Relative fluorescence intensity of daf‐2 mutants expressing sre‐1p::gfp. n = 50.

4. E

   Relative fluorescence intensity of transgenic animals expressing sre‐1p::mCherry in fed and starved conditions. n = 29–35.

5. F

   Relative fluorescence intensity of wild‐type and daf‐2 mutant animals expressing sre‐1p::ocr‐2 genome::mcherry. n = 41.

Data information: All error bars represent ± SEM. Tops and bottoms of boxes indicate the 25^th and 75^th percentiles, respectively; whiskers represent 10^th–90^th percentile. Median is indicated by a horizontal line, and the average is marked by “+” in the box.

Figure 4. PI3K/AKT/FOXO pathway acts downstream of daf‐2 signaling in ADL to mediate ascr#3 avoidance

1. Fraction reversing of wild‐type animals, age‐1 mutants, akt‐1 mutants, akt‐2 mutants, and pdk‐1 mutants in response to 500 nM ascr#3. n = 40–90. *P < 0.05 and **P < 0.01 (Dunnett's test).

2. Relative fluorescence intensity of transgenic animals expressing sre‐1p::snb‐1 cDNA::yfp, including daf‐2 mutants, age‐1 mutants, akt‐1 mutants, daf‐16 mutants, daf‐16;daf‐2 double mutants, and daf‐16 mutants expressing sre‐1p::daf‐16 cDNA (ADL). n = 37–218. *, **, and **** present different from wild type at P < 0.05, P = 0.01, and P < 0.0001 (Dunnett's test). ⁺⁺ P < 0.01 and ⁺⁺⁺⁺ P < 0.0001 (unpaired Student's t‐test). Tops and bottoms of boxes indicate the 25^th and 75^th percentiles, respectively; whiskers represent 10^th–90^th percentile. Median is indicated by a horizontal line, and the average is marked by “+” in the box.

3. Fraction reversing of daf‐16 mutant animals in fed and starved conditions in response to ascr#3 exposure. n = 30–40.

4. Fraction reversing of wild‐type animals, daf‐2 mutants, daf‐16 mutants, and daf‐16;daf‐2 double mutants in response to ascr#3. n = 50–80. *P < 0.05 and **P < 0.01 (Dunnett's test).

5. Fraction reversing of wild‐type animals, daf‐16 mutants, and daf‐16 mutants expressing sre‐1p::daf‐16 cDNA (ADL). n = 70. *P < 0.05 (Dunnett's test).

Data information: All error bars represent ± SEM.

Figure EV3. ascr#3 avoidance behaviors of daf‐16 mutants and of skn‐1 mutants

1. A, B

   Fraction reversing of daf‐16 mutants (A) and skn‐1 mutants (B) in response to 100, 200, and 400 nM ascr#3 under fed conditions. (A) n = 60, *P < 0.05 (Bonferroni's test). (B) n = 40–50.

Data information: All error bars represent ± SEM.

Figure 5. INS‐18, which is secreted in the intestine, inhibits DAF‐2 signaling in ADL

1. Fraction reversing of insulin‐like peptide mutants, ins‐1, ins‐7, ins‐18, ins‐22, ins‐32, ins‐35, and daf‐28 in response to ascr#3. n = 40–170. *P < 0.05 and ***P < 0.001 (Dunnett's test).

2. Fraction reversing of wild‐type animals, ins‐1 mutants, daf‐16 mutants, and daf‐16;ins‐1 double mutants in response to ascr#3. n = 60–70. *P < 0.05 (Dunnett's test).

3. Fraction reversing of wild‐type animals, daf‐2 mutants, ins‐18 mutants, and daf‐2;ins‐18 double mutants in response to ascr#3. n = 70. *P < 0.05 and **P < 0.01 (Dunnett's test).

4. Representative images of wild‐type animals (right) and ins‐18 mutants (left) expressing sre‐1p::snb‐1 cDNA::yfp. Scale bar is 10 μm.

5. Relative fluorescence intensity of sre‐1p::snb‐1 cDNA::yfp in wild‐type animals, daf‐2 mutants, ins‐18 mutants, daf‐2;ins‐18 double mutants, and ins‐18 mutants expressing unc‐14p::ins‐18 cDNA (neuron) and acd‐5p::ins‐18 cDNA (intestine). n = 35–75. ****P < 0.0001 (Dunnett's test). ⁺⁺⁺⁺ P < 0.0001 (unpaired Student's test). Tops and bottoms of boxes indicate the 25^th and 75^th percentiles, respectively; whiskers represent 10^th–90^th percentile. Median is indicated by a horizontal line, and the average is marked by “+” in the box.

6. Fraction reversing of wild‐type animals, ins‐18 mutants, and ins‐18 mutants expressing unc‐14p::ins‐18 cDNA (neuron), and ins‐18 mutants expressing acd‐5p::ins‐18 cDNA (intestine). n = 60. *P < 0.05 (Dunnett's test). ⁺⁺⁺ P < 0.001 (unpaired Student's test).

Data information: All error bars represent ± SEM.

Figure EV4. Expression pattern of acd‐5 promoter in the intestine and neuronal rescue of ins‐18 phenotype

1. A representative image of a transgenic animal expressing acd‐5p::gfp. Scale bar is 10 μm.

2. Fraction reversing of wild‐type animals, ins‐18 mutants, and ins‐18 mutants expressing su006(rgef‐1)p::ins‐18 cDNA::gfp (neuron). n = 30–40. *P < 0.05 and **P < 0.01 (Dunnett's test).

Data information: Error bars represent ± SEM.

Figure EV5. Expression of ins‐18 promoter upon starvation

Relative fluorescence intensity of expressing ins‐18p::gfp in the intestine in 6‐h and 24‐h fed and starved conditions. n = 25–40. Tops and bottoms of boxes indicate the 25^th and 75^th percentiles, respectively; whiskers represent 10^th–90^th percentile. Median is indicated by a horizontal line, and the average is marked by “+” in the box.

Figure 6. Intestinal INS‐18 secretion modulates ascr#3 avoidance in a feeding state‐dependent fashion

1. A

   Representative images of transgenic animals expressing acd‐5p::ins‐18 cDNA::gfp at 1‐h fed and starved conditions. Arrows indicate coelomocyte. Scale bar is 10 μm.

2. B, C

   Relative fluorescence intensity of the accumulated GFP of transgenic animals expressing acd‐5p::ins‐18 cDNA::gfp in coelomocytes at different times after feeding, starving (B), and re‐feeding conditions (C). n = 39–50. (B) *P = 0.05 and ****P < 0.0001 (Bonferroni's test). (C) ****P < 0.0001 (Dunnett's test).

3. D

   Relative fluorescence intensity of the accumulated GFP of transgenic animals expressing acd‐5p::ins‐18 cDNA::gfp in coelomocytes under ins‐18 RNAi treatment. n = 36–59. ****P < 0.0001 (Bonferroni's test).

4. E

   Fraction reversing of wild‐type animals, daf‐2 mutants, and daf‐2 mutants expressing sre‐1p::daf‐2 cDNA treated with ins‐18 RNAi for 24 h. n = 80–100. **P < 0.01 (Bonferroni's test).

5. F

   Relative fluorescence intensity of wild‐type animals expressing sre‐1p::snb‐1 cDNA::yfp in 24 h‐ins‐18 RNAi condition. n = 57–58. **P < 0.01 (unpaired Student's t‐test).

Data information: All error bars represent ± SEM. Tops and bottoms of boxes indicate the 25^th and 75^th percentiles, respectively; whiskers represent 10^th–90^th percentile. Median is indicated by a horizontal line, and the average is marked by “+” in the box.

Figure 7. A model for feeding state‐dependent modulation of ascr#3 avoidance behaviors

1. In fed conditions, INS‐18 released from the intestine blocks the activity of DAF‐2 signaling in ADL, which suppresses chemical synaptic transmission in ADL and decreases avoidance behavior to ascr#3.

2. In starved conditions, secretion of INS‐18 from the intestine is decreased, which activates the DAF‐2 signaling of ADL, resulting in increase in synaptic release from ADL to downstream neurons, and promotes enhanced avoidance behavior to ascr#3.