Peptidergic signaling from clock neurons regulates reproductive dormancy in Drosophila melanogaster

Abstract

With the approach of winter, many insects switch to an alternative protective developmental program called diapause. Drosophila melanogaster females overwinter as adults by inducing a reproductive arrest that is characterized by inhibition of ovarian development at previtellogenic stages. The insulin producing cells (IPCs) are key regulators of this process, since they produce and release insulin-like peptides that act as diapause-antagonizing hormones. Here we show that in D. melanogaster two neuropeptides, Pigment Dispersing Factor (PDF) and short Neuropeptide F (sNPF) inhibit reproductive arrest, likely through modulation of the IPCs. In particular, genetic manipulations of the PDF-expressing neurons, which include the sNPF-producing small ventral Lateral Neurons (s-LNvs), modulated the levels of reproductive dormancy, suggesting the involvement of both neuropeptides. We expressed a genetically encoded cAMP sensor in the IPCs and challenged brain explants with synthetic PDF and sNPF. Bath applications of both neuropeptides increased cAMP levels in the IPCs, even more so when they were applied together, suggesting a synergistic effect. Bath application of sNPF additionally increased Ca2+ levels in the IPCs. Our results indicate that PDF and sNPF inhibit reproductive dormancy by maintaining the IPCs in an active state.

Fig 1. Enhanced activity of PDF-expressing neurons reduces dormancy levels.

(A) Hypersensitization of PDF⁺ neurons (Pdf>Na⁺ChBac) and overexpression of PDF (Pdf>Pdf) both result in a significant reduction of reproductive arrest. (B) Pan-neuronal overexpression of sNPF (elav>2xsNPF) leads to a significant decrease in the proportion of arrested females. (C) Overexpression of sNPF in the PDF-expressing neurons (Pdf>2xsNPF) and in the small ventrolateral neurons (R6>2xsNPF) triggers females to exit from dormancy. Numbers within bars refer to the number of dissected females considered in the dormancy assays. Data are presented as mean ± SEM. ANOVA on arcsine transformations, followed by post-hoc Tukey HSD test. ***p<0.001, n.s. not significant.

Fig 2. Inhibited neuronal activity of PDF⁺ neurons and impairment of sNPF signaling in the IPCs increase dormancylevels.

(A) Expression of the potassium channel Ork (Pdf>Ork) enhances incidence of gonadal arrest in the experimental flies compared to control females. (B) Genetic ablation of PDF-expressing neurons (Pdf>hid) provokes a higher proportion of females entering dormancy compared to controls. (C) Pdf⁰¹ mutant females show enhanced ovarian quiescence levels compared to a congenic wild-type control at both summer and winter photoperiods (see text). (D) PDFR overexpression in IPCs in a han/han mutant background significant reduces dormancy compared to homozygous and heterozygous han/+ controls, but not compared to the wild-type control dilp2 (p)>+, which also appears as one of the controls in Fig 2E. (E) Overexpression of a dominant negative form of sNPFR1 (UAS-sNPFR1-DN) in the IPCs under the control of an early and a later-expressed IPC driver (dilp2(p)-Gal4 and Insp3-Gal4, respectively) increases dormancy levels. (F) Downregulation of sNPFR1 with RNAi using the Insp3-Gal4 driver significantly enhances reproductive arrest compared to controls. Numbers within bars indicate the number of dissected females considered in the dormancy assays. Data are presented as mean ± SEM. ANOVA on arcsine transformations, followed by post-hoc tests. ***p<0.001, **p<0.01, *p<0.05, n.s. not significant.

Fig 3. The terminals of the PDF positive s-LN_vs overlap with dendrites of the insulin producing cells (IPCs).

(A) Representative confocal image (overlay of Z-stacks through the entire brain) of a Hu-S fly brain double-stained with antibodies against PDF (green) and DILP2 (magenta). The brain is outlined by a dashed line. Me = medulla. Scale bar = 100 μm. (B) Representative confocal image of the dorsal part of dilp2(p)>GFP brains double stained with antibodies against PDF (green) and GFP (magenta) (overlay of 10 Z-stacks). Processes of the insulin producing cells (IPCs) overlap with the terminals of the small ventrolateral neurons (s-LN_v term) as can be seen in single confocal stacks of 2 μm thickness (B₁-B₄). Every second single stack of the posterior part of (B) is shown from anterior to posterior. White arrows indicate close proximity between IPCs and s-LN_v terminals. Scale bar = 20 μm. (C-D) Labeling of the IPCs with the dendritic marker DenMark [75] (dilp2(p)>DenMark) indicate that the IPC fibers contacting the s-LN_v terminals are of dendritic origin. Representative confocal images of two different brains are shown. Scale bar 10 μm.

Fig 4. Bath-applied PDF and sNPF induce cAMP increases in the IPCs.

Live optical imaging in flies expressing a cAMP sensor in their IPCs (dilp2(p)>Epac1camps). Average inverse FRET traces (CFP/YFP) in IPCs reflecting intracellular changes in cAMP levels. Substances were bath-applied to freshly dissected fly brains at 100 s (indicated by black arrow). Application of 10⁻⁵ M adenylate cyclase activator NKH477 (NKH, dark gray) induced a robust increase in cAMP, indicating that the general procedure was working. As a negative control, hemolymph-like saline (HL3) was applied. (A) Bath-applied PDF (10⁻⁵ M) evokes cAMP increases in the IPCs (light blue), suggesting a possible functional connection between PDF⁺ cells and IPCs. Similar increase was observed when PDF was applied in the presence of 2μM sodium channel blocker tetrodotoxin (TTX; dark blue), indicating a direct connection. (B) Maximum inverse FRET changes quantified for each individual neuron and averaged for each pharmacological treatment between 100–1000 s. (C) A close-up of the immediate changes in cAMP levels occurring from the application point until 200 s. No significant changes can be observed when PDF or PDF+TTX were applied. (D) Maximum inverse FRET changes from 100–200 s. (E) Bath-applied sNPF (10⁻⁵ M) generates cAMP rises in the IPCs (yellow). Similar increase occurs in the presence of TTX (orange), thus suggesting a direct connection. (F) Maximum inverse FRET changes between 100–1000 s. (G) Magnification of the immediate changes between 100–200 s. (H) Maximum inverse FRET changes between 100–200 s. The legend shows the color code of the different treatments and the number of neurons in the dissected brains (in parentheses) considered in this analysis. Data are shown as mean ± SEM. Kruskal-Wallis test followed by Bonferroni-corrected Wilcoxon pairwise-comparisons. ***p<0.001, **p<0.01, *p<0.05, n.s. not significant.

Fig 5. The co-application of PDF and sNPF increase the cAMP response.

(A) Co-application of sNPF and PDF evokes large cAMP increases in the IPCs (red), which remain significant even when concentrations of peptides are halved (rose). (B) Maximum inverse FRET between 100–1000 s. (C, D) Immediate cAMP responses (100–200 s) reveal a large, rapid increase due to the sNPF+PDF co-application. (E) Co-application of sNPF and PDF induces cAMP increases in the IPCs also in the presence of TTX (magenta), and this response seems to be larger in the time interval of 200–600 s than the effect triggered by the application of the single peptides with TTX. (F) Maximum inverse FRET changes from 100–1000 s. (G) The effect of the sNPF+PDF co-application seems to be, at least partially, due to a direct activation of the IPCs. The initial increase is apparently reduced when TTX is present (magenta). (H) Maximum inverse FRET changes from 100–1000 s. The legend shows the color code of the different treatments and the number of neurons in the dissected brains (in parentheses) considered in this analysis. Data are shown as mean ± SEM. Labeling as in Fig 4.

Fig 6. Co-application of sNPF with other Drosophila peptides suggests that sNPF and PDF may have a unique interaction.

(A) Average inverse FRET traces (CFP/YFP) of IPCs reflecting intracellular cAMP changes. 10⁻⁵ M synthetic sNPF was co-applied with PDF, SDNFMRFa, AKH (adipokinetic hormone), DTK (Drosophila tachykinin), Ast-C (Allatostatin-C) /10^-5M for each/ or 10⁻⁵ M Ast-C was added alone. The black arrow indicates the application point of the different substance. NKH: adenylate cyclase activator, used as positive control. HL3: hemolymph-like saline, used as negative control. (HL3) (B) Maximum inverse FRET changes quantified for each individual neuron and averaged for each treatment from 100 s until 1000 s. Statistical comparison revealed that co-application of sNPF with SDNFMRFa, AKH, and DTK resulted in a significant decrease of cAMP levels compared to the sNPF+PDF co-application. sNPF+Ast-C led to a significant increase in the level of cAMP, however similar change was observed in the case of Ast-C application alone. (C) A close up of the immediate cAMP level changes occurring from the application point until 200 s. (D) Maximum inverse FRET from 100 s until 200 s. The legend indicates the color code of the different treatments and the number of neurons in the dissected brains (in parentheses) considered in this analysis. ***p<0.001, **p<0.01, *p<0.05, n.s. not significant.

Fig 7. The responses of the IPCs to PDF and sNPF are absent in the PDF receptor mutant han, but only the response to PDF is mediated by the PDF receptor.

(A-D) Live imaging in han mutants expressing a cAMP sensor in their IPCs (han; dilp2(p)>Epac1camps). In the mutant background, the effects of both neuropeptides are abolished. (A-B) Maximum inverse FRET changes from 100–1000 s. Short-term responses are all abrogated in han flies. (C-D) Maximum inverse FRET changes from 100–200 s. (E-H) Live imaging in han mutants with the PDF receptor rescued in the IPCs. The expression of han in the IPCs induces very strong responses to PDF but not to sNPF (E-F) Maximum inverse FRET changes from 100–1000 s. (G-H) Maximum inverse FRET changes from 100–200 s. Labeling as in Fig 4.

Fig 8. The neuropeptide sNPF induces a small increase in the intracellular Ca²⁺ level, while PDF has no effect.

Left panel: Average changes in GFP fluorescence of IPCs reflecting intracellular changes in Ca²⁺ levels. Substances were bath-applied to freshly dissected fly brains at ~100 s (indicated by a black arrow). The cholinergic agonist carbamylcholine (1 mM CCh) was used as positive control, which induced a robust increase in Ca²⁺, indicating that the general procedure was working. As a negative control, hemolymph-like saline (HL3) was applied. Application of 10–5 M synthetic PDF peptide did not alter Ca²⁺ levels, while 10–5 M sNPF induced a small but significant increase in Ca²⁺ levels. This calcium response seems to be due to direct activation of the IPCs, since it was not blocked in the presence of the sodium channel blocker TTX (2μM). Right panel: Maximum Ca²⁺ changes (%) for each individual neuron were calculated and averaged for each pharmacological treatment from 100 s until 200 s. Statistical comparison revealed significant increases in Ca²⁺ levels compared to the negative control (HL3) for CCh, sNPF and sNPF+TTX, while PDF had no effect. Data are presented as mean ± SEM. The color code of the different treatments and the number of neurons [n] and dissected brains (n) considered in this analysis are shown below the panels. Kruskal-Wallis test followed by Bonferroni-corrected Wilcoxon pairwise-comparisons. ***p<0.001, *p<0.05, n.s. not significant.