ITP splicing pattern and expression of ITP transcript variants in the nervous system of adult male Drosophila.

(A) Drosophila ITP gene can generate 5 transcript variants (ITP-RC, RD, RE, RF and RG). ITP-RC encodes ITPL1 precursor, ITP-RD, RF and RG all encode ITPL2 precursor, and ITP-RE encodes a precursor that produces the amidated ITP (ITPa) peptide. Grey boxes represent exons and lines represent introns (drawn to scale). The regions encoding the open reading frame are colored (pink, green or blue). ITP is located on the second chromosome and numbers on the top indicate the genomic location. ITP-RC-T2A-GAL4 drives GFP (UAS-JFRC81GFP) expression in the (B) brain and (C and D) ventral nerve cord (VNC). B’’ shows another brain preparation (same as in Fig. 2A) where axons of ITP-RC neurons are clearly visible. All images are from male flies. Within the brain, ITP-RC is co-expressed with ITPa in four pairs of lateral neurosecretory cells (L-NSCITP), one pair of diuretic hormone 31 (DH31)-expressing lateral neurosecretory cells (L-NSCDH31), one pair of 5th ventrolateral neurons (5th-LNv) and one pair of dorsolateral neurons (LNdITP). L-NSCITP and L-NSCDH31 are a subset of lateral neuroendocrine cells and the single pairs of 5th-LNv and LNdITP belong to the circadian clock network. Within the VNC, ITP-RC is co-expressed with ITPa in abdominal ganglion neurons (iag), which innervate the rectal pad. In addition, ITP-RC is expressed in a pair of Tv* neurons near the midline in each thoracic neuromere. These neurons are located next to the FMRFamide-expressing Tv neurons (see Fig. 2B). ITP-RD-T2A-GAL4 also drives GFP expression in the (E and F) brain and (G and H) VNC. ITP-RD is expressed in L-NSCITP, 5th-LNv and LNdITP neurons, as well as glia. Within the VNC, ITP-RD is expressed in neurons which are not iag or Tv* neurons. (I) Summary of ITP isoform expression within the nervous system. Grey box indicates presence and white box indicates absence.

ITP is co-expressed with other neuropeptides in the nervous system of adult male Drosophila.

(A) A single pair of ITP-RC > GFP-positive lateral neurosecretory cells in the dorsal brain (marked by an arrowhead) co-express diuretic hormone 31 (DH31). (B) ITP-RC drives GFP expression in a pair of Tv* neurons near the midline in each thoracic neuromere. These neurons are located next to the FMRFamide-expressing Tv neurons. (C and D) ITP-RC and ITPa-expressing peripheral neurons (marked by asterisk) and abdominal ganglion neurons (marked by arrowheads) co-express allatostatin-A (Ast-A) neuropeptide. (E) CCAP > GFP-positive neurons co-express ITPa (and Ast-A by extension) in the abdominal ganglion neurons (marked by arrowheads). (F) ITP-RC and ITPa-expressing neurons are distinct from Ast-A-expressing neurons in the brain. (G) Schematic of the nervous system showing neuropeptides (transcripts or mature peptides) expressed in ITP neurons. Peripheral neurons on one side are marked by arrowheads. Based on previous reports (Kahsai et al., 2010, Hermann-Luibl et al., 2014, Zandawala et al., 2018a) and the present study.

ITP expression in peripheral tissues of adult male Drosophila.

(A) Schematic showing the location of tissues where ITP is expressed. Created with BioRender.com/iceyb7k. (B) t-SNE visualization of single-cell transcriptomes showing ITP expression in different tissues of adult Drosophila. ITP is broadly expressed in peripheral tissues, including (C) trachea, Malpighian tubules (tubule), body, (D) heart, fat body and gut. ITP-RC-T2A-GAL4 drives GFP (UAS-JFRC81GFP) expression in (E) peripheral neurons with axons innervating the heart and (F) abdominal ganglion neurons which innervate the rectum. ITP-RC is not expressed in (G) the fat body, (H) midgut or (I) Malpighian tubules. ITPa immunolabelling is present in (J) a pair of peripheral neurons (cell bodies marked by arrowheads) innervating the heart and (K) in abdominal ganglion neurons which innervate the rectum, but (L) absent in the midgut. ITP-RD-T2A-GAL4 drives GFP expression in (M) the heart and nephrocytes (marked by asterisk), (N) middle midgut, (O) posterior midgut, (P) ureter and (Q) trachea. (R) ITP-RD is not expressed in the fat body.

ITP signaling components are found in protostomes.

(A) Multiple sequence alignment of ITP precursor sequences. ITP is homologous to crustacean hyperglycemic hormone (CHH) and molt-inhibiting hormone (MIH). Note the conservation of six cysteine residues (highlighted in red) across all the species. C-terminal glycine which is predicted to undergo amidation is colored in green. Species abbreviations: Drome, Drosophila melanogaster; Locmi, Locusta migratoria; Dapma, Daphnia magna; Carma, Carcinus maenas; Ixosc, Ixodes scapularis; Caeel, Caenorhabditis elegans; Hypdu, Hypsibius dujardini; Prica, Priapulus caudatus; Chala, Charonia lampas. (B) Maximum-likelihood phylogeny of membrane guanylate cyclase receptors identifies two clades that are restricted to protostome phyla which also have ITP. The clade containing D. melanogaster Gyc76C receptor are the putative ITPa receptors. Bootstrap values higher than 200 (based on 500 replicates) are indicated adjacent to the nodes. Drosophila guanylate cyclase alpha and beta subunits were used as outgroups.

Gyc76c expression in adult male Drosophila.

(A) Schematic showing the generation of Gyc76C-T2A-GAL4 knock-in line. Gyc76C-T2A-GAL4 drives GFP (UAS-JFRC81GFP) expression in the (B) anterior midgut, (C) ureter, (D) posterior midgut, (E) Malpighian tubules, (F) ileum, rectum, (G) and adipocytes in the fat body. Gyc76C is expressed in the regions of (H) the anterior midgut and (I) rectal papillae in the rectum that are innervated by ITPa-expressing neurons. Gyc76C is also broadly expressed in the (J) brain and (K) ventral nerve cord. (L) Gyc76C is expressed in glial clock cells and (M) subsets of dorsal clock neurons (both labelled by Period antibody and marked by arrowheads). Gyc76C is not expressed in (N) insulin-producing cells (labelled by DILP2 antibody) and (O) adipokinetic hormone (AKH) producing endocrine cells but is expressed in the corpora allata (CA) (marked in white).

Recombinant Drosophila ITPa inhibits Malpighian tubule secretion via Gyc76C.

(A) Schematic of Ramsay assay used to monitor ex vivo secretion by tubules. (B) Application of Drosophila 500nM ITPa does not affect basal secretion rates by unstimulated tubules. 500nM ITPa inhibits both (C) 10nM leucokinin (LK)-stimulated and (D) 1μM diuretic hormone 31 (DH31)-stimulated secretion rates. Importantly, while 500nM ITPa inhibits (E) 10nM LK-stimulated secretion and (F) 1μM DH31-stimulated by renal tubules from control flies, this inhibitory effect is abolished in tubules where Gyc76C has been knocked down with UAS-Gyc76C RNAi (#106525) in stellate cells using the c724-GAL4 and in principal cells using uro-GAL4. Male Malpighian tubules were used for all experiments. For B-D,* p < 0.05 and **** p < 0.0001 as assessed by unpaired t test. For E and F, within each genotype, different letters denote secretion rates that are significantly different from one another (p < 0.05) as assessed by two-way ANOVA followed by Tukey’s multiple comparisons test.

ITPa activates Gyc76C heterologously expressed in HEK293T cells.

(A) Schematic of the heterologous assay used to functionally characterize Gyc76C. Created with BioRender.com/d44ld9l. Application of (B) 50mM, (C) 250mM or (D) 500nM Drosophila ITPa to HEK293T cells transiently expressing Green cGull (cGMP sensor) and Gyc76C results in a dose-dependent increase in fluorescence compared to control cells which do not express Gyc76C. Graphs represent the mean fluorescence of 17-18 cells. Representative images showing fluorescence in (E) HEK293T cells expressing Gyc76C and (F) those without Gyc76C before and 4 mins after the addition of 500nm ITPa. (G) Area under the curve analysis demonstrates significant differences in Green cGull fluorescence increases between experimental and control conditions; * p < 0.05 and **** p < 0.0001 as assessed by nonparametric one-way ANOVA followed by Dunn’s test for multiple comparisons.

ITPa neurons are active and release ITPa during desiccation.

GFP immunofluorescence, indicative of calcium levels and measured using the CaLexa reporter, is increased in L-NSCITP of (A) male and (B) female flies exposed to desiccation. The GFP intensity returns to control levels in flies that were rehydrated following desiccation. ITPa immunofluorescence, indicative of peptide levels, is lowered in 5th-LNv, LNdITP and L-NSCITP of (C) male and (D) female flies exposed to desiccation. ITPa peptide levels recover to control levels in flies that were rehydrated following desiccation. Lower peptide levels during desiccation indicates increased release. For all panels, * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001 as assessed by one-way ANOVA followed by Tukey’s multiple comparisons test.

Knockdown of ITP in adult female Drosophila impacts metabolic homeostasis, feeding and associated behaviors.

(A) ITPa immunofluorescence is reduced in the L-NSCITP neurons of flies in which ITP was knocked down using ITP-RC-GAL4TS (ITP-RC-T2A-GAL4 combined with temperature-sensitive tubulin-GAL80). Flies with ITP knockdown are (B) less resistant to desiccation tolerance, (C) have reduced water content, (D) increased defecation, and (E) shrunken abdomen. ITP knockdown flies (F) survive less under starvation, (G) have lower levels of circulating glucose, and (H) unaffected glycogen levels. However, reduced ITP signaling results in (I and J) less lipid levels (TAG = triacylglyceride), and (K) smaller ovaries. Moreover, ITP knockdown flies exhibit (L) increased feeding (over 24 hours) and (M and N) defects in preference for nutritive sugars when starved for 18 hours prior to testing. Abbreviations: Luc Ri, luciferase RNAi; ITP Ri, ITP RNAi. For B and F, ** p < 0.01, *** p < 0.001, as assessed by Log-rank (Mantel-Cox) test. For all others, * p < 0.05, ** p < 0.01, and **** p < 0.0001 as assessed by unpaired t test.

ITPa overexpression in adult female Drosophila impacts osmotic and metabolic homeostasis, feeding and related behaviors.

(A) Overexpression of ITPa using ITP-RC-GAL4TS results in increased ITPa immunofluorescence in L-NSCITP. ITPa overexpression results in (B) increased desiccation tolerance, (C) increased water content and (D) a slightly bloated abdomen (marked by an asterisk). ITPa overexpression causes (E) increased starvation tolerance, (F) reduced circulating glucose levels but has no effect on (G) glycogen levels. (H) The size of neutral lipid droplets (stained with Nile red) is increased in flies with ITPa overexpression. (I and J) These flies also exhibit defects in preference for nutritive sugars when starved for 16 hours prior to testing. (K) ITPa overexpression flies have enlarged ovaries. (L) ITPa overexpression has no effect on locomotor activity under fed or desiccating conditions. Black bars indicate night-time and yellow bars indicate daytime. All experiments were performed at 29°C. For B and E, **** p < 0.0001, as assessed by Log-rank (Mantel-Cox) test. For A, * p < 0.05 as assessed by unpaired t test. For all other experiments, * p < 0.05, *** p < 0.001, **** p < 0.0001 as assessed by one-way ANOVA followed by Tukey’s multiple comparisons test. For clarity, significant pairwise differences compared to only the experimental treatment are indicated.

Female Malpighian tubule specific knockdown of Gyc76C impacts osmotic and ionic homeostasis.

Knockdown of Gyc76C in both the (A) principal cells of renal tubules using uro-GAL4 and (D) stellate cells using c724-GAL4 reduces desiccation tolerance. Gyc76C knockdown in (B) principal cells increases survival under salt stress whereas knockdown in (E) stellate cells lowers survival. (C and F) Gyc76C knockdown in principal or stellate cells increases the time taken for recovery from chill-coma. Abbreviation: Gyc76C Ri, Gyc76C RNAi. For all panels, ** p < 0.01, **** p < 0.0001, as assessed by Log-rank (Mantel-Cox) test.

Gyc76C knockdown in the female fat body using yolk-GAL4 impacts metabolic homeostasis, feeding and associated behaviors.

Flies with fat body specific Gyc76C knockdown with UAS-Gyc76C RNAi (#106525) are (A) extremely susceptible to starvation and (B) have reduced glucose levels. (C) Glycogen levels are unaltered in flies with fat body specific Gyc76C knockdown. (D and E) However, lipid levels (TAG = triacylglyceride) are drastically reduced. Gyc76C knockdown flies exhibit (F) increased feeding (over 24 hours), (G) a preference for yeast over sucrose, and (H and I) defects in preference for nutritive sugars when starved for 4 hours prior to testing. Flies with Gyc76C knockdown in the fat body have (J) smaller ovaries, they (K) defecate more and have (L) reduced water content than the controls. For K, number of excreta counted over 2 hours. Gyc76C knockdown also impacts (M) DILP2 peptide levels (N) but not ITPa levels in the neurosecretory cells. CTCF = Corrected Total Cell Fluorescence. (O) Representative confocal stacks showing DILP2 and ITPa immunostaining. Gyc76C knockdown flies also display reduced daytime locomotor activity under (P) fed and (Q) and starved conditions compared to controls. Black bars indicate night-time and yellow bars indicate daytime. (R) Average night and daytime activity over one day under fed and starved conditions. For A, **** p < 0.0001, as assessed by Log-rank (Mantel-Cox) test. For M and N, * p < 0.05 as assessed by unpaired t test. For all others, * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001 as assessed by one-way ANOVA followed by Tukey’s multiple comparisons test. For clarity, significant pairwise differences compared to only the experimental treatment are indicated.

Inputs and outputs of ITP neurons based on connectomics and single-cell transcriptomics.

(A) Reconstruction of ITPa-expressing neurons using the complete electron microscopy volume of the adult female brain (data retrieved from the FlyWire platform). Four pairs of lateral neurosecretory cells (L-NSCITP) are grey, 5th ventrolateral neurons (5th-LNv) are cyan, and dorsolateral neurons (LNdITP) are yellow. Diuretic hormone 31 (DH31)-expressing lateral neurosecretory cells (L-NSCDH31) are not shown since it is unclear which of the three pairs of L-NSCDH31 co-expresses ITPa. (B) Location of input and output synapses are colored according to the ITP neuron type. (C) Proportion of input synapses (grouped by super class annotations for the FlyWire connectome (Schlegel et al., 2024)) to each ITP neuron type. (D) Reconstructions of neurons from different super classes providing inputs to (left) and receiving outputs from (right) 5th-LNv, LNdITP and L-NSCITP. Only the top 10 cell types are shown here. (Middle) Number of neurons, categorized by super class, providing inputs to and receiving outputs from 5th-LNv, LNdITP and L-NSCITP. (E) Thermo/hygrosensory input pathway to LNdITP. (F) Output from L-NSCITP to other osmoregulatory hormone-producing cells. (G) Identification of single-cell transcriptomes representing different subsets of ITPa-expressing neurons in the adult brain dataset (Davie et al., 2018). Since both the 5th-LNv and LNdITP co-express ITP, cryptochrome (cry) and neuropeptide F (NPF), these cells are grouped as LNITP. All three sets of neurons express genes required for neuropeptide processing and release (amon, svr, Pal2, Phm and Cadps) and were identified based on the neuropeptides (ITP, NPF, Dh31, sNPF and Tk) they express. Dot plots showing expression of (H) monoamine, (I) neuropeptide and (J) neurotransmitter receptors in different sets of ITPa neurons.

A schematic depicting ITP signaling pathways modulating metabolic and osmotic homeostasis in Drosophila.

Different subsets of ITP neurons in the brain have been color-coded. LNdITP and 5th-LNv are part of the circadian clock network and regulate clock-associated behaviors and physiology. L-NSCITP release ITPa into the circulation following dehydration and information regarding this internal state is likely conveyed to L-NSCITP by other neuromodulators. Following its release into the hemolymph, ITPa activates a membrane guanylate cyclase receptor Gyc76C on the adipocytes in the fat body, principal and stellate cells in the renal tubules, as well as other targets. These signaling pathways affect diverse behaviors and physiology to modulate metabolic and osmotic homeostasis. Dashed arrows depict pathways that remain to be clarified, solid arrows represent direct effects, and red bars represent inhibition. Created with BioRender.com/i8nevmx.