A neuroendocrine pathway modulating osmotic stress in Drosophila

Environmental factors challenge the physiological homeostasis in animals, thereby evoking stress responses. Various mechanisms have evolved to counter stress at the organism level, including regulation by neuropeptides. Although much progress has been made on the mechanisms and neuropeptides influencing nutritional stress, relatively little is known about the factors and pathways regulating osmotic and ionic stresses. Here, we uncover the neuropeptide Corazonin (Crz) as a neuroendocrine factor that modulates the release of an osmoregulatory peptide, CAPA, to regulate tolerance to osmotic and ionic stress. Both knockdown of Crz and acute injections of Crz peptide impact desiccation tolerance and recovery from chill-coma. Comprehensive mapping of the Crz receptor (CrzR) expression identified three pairs of Capa-expressing neurons (Va neurons) in the ventral nerve cord that mediate these effects of Crz. We further show that Crz is released during dry starvation (desiccation) and acts to restore homeostasis by inhibiting CAPA release via inhibition of cAMP production in Va neurons. Finally, knockdown of CrzR in Va neurons also affects CAPA release, and consequently influences desiccation tolerance and chill-coma recovery, considered proxies for diuretic state. Thus, Crz modulates Va neurons to maintain osmotic and ionic homeostasis, which in turn influences stress tolerance. Taken together with our previous work showing that systemic Crz signaling acts to restore nutrients levels by promoting food search and feeding, we propose that Crz signaling also ensures osmotic homeostasis by inhibiting the anti-diuretic CAPA peptides. Thus, Crz ameliorates stress-associated physiology through systemic modulation of both peptidergic neurosecretory cells and the fat body in Drosophila. Author summary Insects are among the largest groups of animals and have adapted to inhabit almost all environments on Earth. Their success in surviving extreme conditions stems largely from their ability to withstand environmental stress, such as desiccation and cold. However, the neural mechanisms that are responsible for coordinating responses to counter these stresses are largely unknown. To address this, we delineate a neuroendocrine axis utilizing the neuropeptides Corazonin (Crz) and CAPA, that coordinate responses to metabolic and osmotic stress. We show that Crz modulates the release of an anti-diuretic peptide, CAPA from a set of neurosecretory cells. CAPA in turn influences osmotic and ionic balance via actions on the Malpighian tubules (the insect analogs of the kidney) and the intestine. Taken together with earlier work, our data suggest that Crz acts to restore metabolic homeostasis at starvation and as a cosenquence of energy mobilization and ensuing metabolic water production, fluid balance needs to be adjusted and therefore CAPA release is inhibited. Hence, this work provides a mechanistic understanding of the neuroendocrine mitigation of metabolic and osmotic stress by two peptide systems.


7
GAL4 T11a line (referred to as CrzR-GAL4 from hereon) was used for all the subsequent 9 preparations had no impact on Ca 2+ levels in Va neuron as measured using GCaMP 247 fluorescence ( Figures 3E-F). These results suggest that Crz inhibits cAMP levels in Va 248 neurons, most likely by direct action on the CrzR. 249 250 We next asked what the neuronal source of Crz is and whether its mode of action on Va 251 neurons is systemic via the circulation or synaptic/paracrine within the abdominal 252 neuromeres. To address this, we examined the morphology of Crz and CAPA neurons. As 253 described above, Crz is expressed in the DLPs, and in males, in abdominal ganglion 254 interneurons (S1 Figure). The DLPs do not have axonal projections descending into the 255 VNC but they send axons to the corpora cardiaca (CC), anterior aorta and intestine (9, 43).  CAPA as an anti-diuretic. In addition, we observed increased survival under starvation 332 ( Figure 5F), possibly due to effects on food intake as a result of CrzR knockdown in non-333 CAPA expressing neurons (S10 Figure). Hence, we knocked down the CrzR more 334 selectively using CAPA-GAL4. The CAPA-GAL4 driven CrzR-RNAi resulted in increased 335 survival under desiccation ( Figure 5G). However, it did not have any impact on chill-coma 336 recovery ( Figure 5H). Similar results were also obtained by knocking down CrzR in CAPA 337 neurons using an independent CrzR-RNAi construct CrzR-RNAi (2). The lack of effect on 338 chill-coma recovery following CrzR knockdown with CAPA-GAL4 could be due to it being a 339 very week driver. Taken together, our data indicate that Crz acts on Va neurons to inhibit 340 the release of an anti-diuretic hormone CAPA, and thereby affecting tolerance to osmotic 341 and ionic stresses.
to maintain osmotic and ionic homeostasis ( Figure 6). The Crz-CAPA signaling thereby also 348 influences tolerance to osmotic and cold stress. An earlier study suggested that Crz is 349 released during nutritional stress to mobilize energy stores from the fat body to fuel food 350 search behavior (14). Furthermore, that study suggests that increased Crz signaling 351 compromises resistance to starvation, desiccation and oxidative stress (14). We confirm 352 these findings here, and also find that Crz inhibits a set of CrzR expressing Va neurons in 353 the abdominal ganglia that produce anti-diuretic CAPA peptides. conditions when flies are exposed to dry starvation (desiccation without food) the two 364 signaling systems act in tandem to restore homeostasis. The exposure to desiccation 365 results in increased hemolymph osmolarity, and the fly responds acutely by releasing anti-366 diuretic CAPA (and maybe other anti-diuretic hormones) to preserve water (20). As the dry 367 starvation continues, the nutrients diminish in the fly, and nutrient sensors record this 368 deficiency, which triggers release of hormones that act to restore metabolic homeostasis. 369 As indicated above, one such hormone is Crz, known to provide energy for food search and 370 induce feeding (14, 16). Mobilization of energy, especially in the form of stored lipids, yields 371 metabolic water and metabolites (see 51, 52, 53) that need to be excreted and thus diuresis 372 needs to be diminished. In insects, increased food intake also yields water (see 54), which 373 requires post-feeding diuresis to restore water homeostasis. The increased need for 374 excretion means that the anti-diuretic action of CAPA needs to be terminated. We propose 375 that onset of Crz release is likely to be induced later than that of CAPA peptides at dry 376 starvation and results in inhibition of the Va neurons and hence diminished CAPA release. 377 method sensitive enough to monitor the timing of changes in hemolymph levels of Crz and 379 CAPA in small organisms such as Drosophila. Indirect measurements, like monitoring levels 380 of peptide-immunolabel in neurons of interest are not necessarily accurate, since these 381 levels reflect the "balance" between peptide release and production and therefore not useful 382 for resolving onset and duration of release with accuracy.  To quantify peptide levels in flies exposed to various stressors, adult males were transferred 486 to either an empty vial (desiccation) or a vial containing aqueous 0.5% agar (starvation) or 487 artificial diet (normal food) and incubated for 18 hours. In addition, one set of flies were 488 desiccated for 15 hours and then transferred to a vial containing 0.5% agar (re-watered) for 489 3 hours. These flies were then processed for immunohistochemistry as described above. 490 Cell fluorescence was quantified as described previously (33)

Chill-Coma Recovery 534
Flies were taken from the Drosophila vials and lightly anesthetized by CO 2 . Once 20 held over ice. For each treatment, 9-15 males were selected per treatment. After injections, 537 flies were transferred individually into 7.5 mL glass vials that were then submerged in a 0°C 538 ice-water slurry for 1 hour. Afterwards, vials were removed from the slurry and gently dried. 539 Flies were left to recover at room temperature without being disturbed for the duration of the 540 experiment. Recovery was recorded by measuring the time when a fly stands on all legs. 541 Experiments were repeated in at least two independent biological replicates. 542 543 Desiccation 544 As above, male flies were isolated and placed over ice within a white plastic weighing dish. 545 For each treatment, 11-15 males were used. After injection with saline injection buffer alone 546 or buffer containing Crz, all flies in that treatment group were transferred en masse into 547 empty Drosophila vials without any food or water. Survival under this desiccation treatment 548 was monitored at regular intervals. This experiment was repeated in three replicates. 549 550

Calcium and cAMP imaging of Va neurons 551
The whole CNS of feeding 3 rd instar larvae expressing the calcium sensor GCaMP6m 552 The mRNA levels were normalized to rp49 levels in the same samples. Relative expression 590 values were determined by the 2 −∆∆CT method (86). See Table S1 for the primers used for 591 qPCR.

Defecation assay 667
To quantify defecation following TRPA1 based thermogenetic activation of CAPA neurons, 668 flies (mixed sex) were kept on food containing bromophenol blue for 24h at 18°C. One hour 669 before the experiments, flies were transferred to a petri dish lined with wet filter paper. 670 Then, each fly was put in a small glass tube at 29°C and video recorded for 60 min.