CCAP regulates feeding behavior via the NPF pathway in Drosophila adults

Abstract

The intake of macronutrients is crucial for the fitness of any animal and is mainly regulated by peripheral signals to the brain. How the brain receives and translates these peripheral signals or how these interactions lead to changes in feeding behavior is not well-understood. We discovered that 2 crustacean cardioactive peptide (CCAP)-expressing neurons in Drosophila adults regulate feeding behavior and metabolism. Notably, loss of CCAP, or knocking down the CCAP receptor (CCAP-R) in 2 dorsal median neurons, inhibits the release of neuropeptide F (NPF), which regulates feeding behavior. Furthermore, under starvation conditions, flies normally have an increased sensitivity to sugar; however, loss of CCAP, or CCAP-R in 2 dorsal median NPF neurons, inhibited sugar sensitivity in satiated and starved flies. Separate from its regulation of NPF signaling, the CCAP peptide also regulates triglyceride levels. Additionally, genetic and optogenetic studies demonstrate that CCAP signaling is necessary and sufficient to stimulate a reflexive feeding behavior, the proboscis extension reflex (PER), elicited when external food cues are interpreted as palatable. Dopaminergic signaling was also sufficient to induce a PER. On the other hand, although necessary, NPF neurons were not able to induce a PER. These data illustrate that the CCAP peptide is a central regulator of feeding behavior and metabolism in adult flies, and that NPF neurons have an important regulatory role within this system.

Keywords: dopamine; feeding behavior; hypothalamus; metabolism.

Fig. 1.

Branches of CCAP neurons superimpose NPF neuron cell bodies. (A) Structure of the Drosophila CCAP-R genomic region; blue boxes indicate CCAP-R exons and blue lines indicate introns. Black arrows with corresponding CG numbers above indicate other genes in the region. Sequences cloned to drive GAL4 expression are indicated as black lines with the corresponding FlyLight number above. The genomic sequence used for R64F05-GAL4 is indicated as a green line. (B–D) Dissected R64F05-GAL4;UAS-GFP male adult brains (5 to 9 d posteclosion) were costained with (B) anti-GFP (green) and (C) anti-NPF (magenta). (D) White in the merged picture indicates overlapping GFP (R64F05) and NPF expression. (Scale bar, 50 µm.) Pictures show a representative Z stack, which includes 40 (2-µm) slices. (E–G) Z projection of (E) CCAP (magenta) and (F) GFP (green; NPF-GAL4;UAS-GFP) neurons in dissected male adult brains (10 to 12 d posteclosion). (G) Merged picture showing CCAP (magenta) and GFP (green; NPF) expression. (Scale bar, 50 µm.) The picture is representative of a Z stack, which includes 40 (2-μm) slices. (H–J) Transverse Z projection of (H) CCAP (magenta) and (I) GFP (green; NPF-GAL4;UAS-GFP) neurons in dissected male adult brains (10 to 12 d posteclosion). (J) Merged picture showing CCAP (magenta) and GFP (green; NPF) expression. (Scale bar, 50 µm.)

Fig. 2.

CCAP regulates CCAP-R–expressing NPF neurons. (A–D) Close-up of NPF neurons in whole Drosophila male brains, 5 to 9 d posteclosion, visualizing NPF expression after various times of starvation. (Scale bar, 50 µm.) (E and F) Relative NPF immunofluorescence expression levels in fed flies and after various times of starvation. (E) P1 (dorsal median) NPF neurons. (F) L1-I (dorsal lateral) NPF neurons. NPF neurons of 10 to 12 brains for each condition were investigated (*P < 0.05, ***P < 0.005; initially, a Kolmogorov–Smirnov test of normality was performed before a one-way ANOVA with Tukey’s post hoc test for multiple comparisons between all samples). (G–J) Close-up of NPF neurons in whole Drosophila male brains, control (NPF-GAL4>w¹¹¹⁸) and experimental (NPF-GAL4>UAS-CCAP-R^RNAi), 5 to 9 d posteclosion, visualizing NPF expression after various times of starvation. (Scale bar, 50 µm.) (K and L) Relative NPF immunofluorescence expression levels in fed flies and after various times of starvation. (K) P1 (dorsal median) NPF neurons. (L) L1-I (dorsal lateral) NPF neurons. NPF neurons of 10 to 12 brains for each condition were investigated (*P < 0.05, ***P < 0.005; initially, a Kolmogorov–Smirnov test of normality was performed before a one-way ANOVA with Tukey’s post hoc test for multiple comparisons between samples within the same time point). Error bars represent SEM.

Fig. 3.

CCAP regulates feeding behavior. (A) A CAFE assay was used to assess total food intake over a 24-h period in 5- to 9-d-old adult males. Five males were used for each replicate and the assay was repeated at least 10 times for each genotype. The food source was 150 mM sucrose. Initially, a Kolmogorov–Smirnov test of normality was performed. Since some of the samples failed the test, a nonparametric Kruskal–Wallis ANOVA was performed with Dunn’s post hoc test for multiple comparisons (*P < 0.05). (B and C) The flyPAD was used to measure the total number of sips over 1 h using males that were either (B) fed ad libitum or (C) previously starved for 18 h. n is between 32 and 64 and a Kolmogorov–Smirnov test of normality was performed. Since some of the samples failed the test, a nonparametric Kruskal–Wallis ANOVA was performed with Dunn’s post hoc test for multiple comparisons, comparing either NPF-GAL4>w¹¹¹⁸, w¹¹¹⁸>UAS-CCAP-R^RNAi, and NPF-GAL4>UAS-CCAP-R^RNAi or CCAP^exc7/+ and CCAP^exc7 nulls (*P < 0.05, **P < 0.01). (D) A starvation resistance test was performed using the DAMS. In short, 5- to 7-d-old male flies were maintained in a DAMS tube with 1% agarose so they had water but no food. Their activity was monitored and these data were used to calculate starvation resistance. n is between 32 and 64. Initially, a Kolmogorov–Smirnov test of normality was performed, followed by a one-way ANOVA performed with Tukey’s post hoc test for multiple comparisons, comparing either NPF-GAL4>w¹¹¹⁸, w¹¹¹⁸>UAS-CCAP-R^RNAi, and NPF-GAL4>UAS-CCAP-R^RNAi or CCAP^exc7/+ and CCAP^exc7 nulls (**P < 0.01, ***P < 0.005). (E) Triglyceride levels were determined in 5- to 7-d-old male flies that were fed ad libitum. n = 30 males per treatment; the assay was repeated at least 10 times for each genotype. Initially, a Kolmogorov–Smirnov test of normality was performed, followed by a one-way ANOVA with Tukey’s post hoc test for multiple comparisons, comparing either NPF-GAL4>w¹¹¹⁸, w¹¹¹⁸>UAS-CCAP-R^RNAi, and NPF-GAL4>UAS-CCAP-R^RNAi or CCAP^exc7/+ and CCAP^exc7 nulls (***P < 0.005). Error bars represent SEM. An asterisk over one sample with no line connecting to another sample indicates it is significantly different from the other samples.

Fig. 4.

CCAP regulates PER behavior. (A and B) Average (±SEM) fraction of adult males 5 to 9 d old fed ad libitum (solid lines) and 24-h wet-starved flies (dashed lines) exhibiting a PER to the indicated concentrations of sucrose (n = 8 to 10, 10 to 20 flies per group). Asterisks denote statistically significant differences between starved flies and flies fed ad libitum (statistical methods can be found in SI Appendix, Materials and Methods; ***P < 0.005). (C) The flyPAD was used for a 2-choice food test between 100 mM l-fucose and 100 mM glucose in flies that were either fed ad libitum prior to testing (blue lines) or after 18-h starvation (orange lines). n is between 32 and 64 and a Kolmogorov–Smirnov test of normality was performed. Since some of the samples failed the test, a nonparametric Kruskal–Wallis ANOVA was performed with Dunn’s post hoc test for multiple comparisons (**P < 0.01, ***P < 0.005). (D–G) Optogenetic control of PER where the channelrhodopsin ReaChR is expressed in (D) CCAP neurons, (E) all NPF neurons using NPF-GAL4, (F) dorsal median NPF neurons using R64F05-GAL4, or (G) dopaminergic neurons (ple-GAL4). Fractions indicate the number of responders out of the number of flies tested. The photostimulation was performed at 620 nm, with 3 pulses at 1 Hz (100-ms pulse width) (n = 5 or 6 flies, 10 to 20 groups of flies for each genotype tested). Two different tests were performed: 1) Similar pulses were compared between various genetic backgrounds and 2) different pulses were compared within the same genetic background. In all instances, a separate Kolmogorov–Smirnov test of normality was performed for D–G, followed by a one-way ANOVA with Tukey’s post hoc test for multiple comparisons (*P < 0.05, ***P < 0.005). Error bars represent SEM.