Serotonergic neurons respond to nutrients and regulate the timing of steroid hormone biosynthesis in Drosophila

The temporal transition of development is flexibly coordinated in the context of the nutrient environment, and this coordination is essential for organisms to increase their survival fitness and reproductive success. Steroid hormone, a key player of the juvenile-to-adult transition, is biosynthesized in a nutrient-dependent manner; however, the underlying genetic mechanism remains unclear. Here we report that the biosynthesis of insect steroid hormone, ecdysteroid, is regulated by a subset of serotonergic neurons in Drosophila melanogaster. These neurons directly innervate the prothoracic gland (PG), an ecdysteroid-producing organ and share tracts with the stomatogastric nervous system. Interestingly, the projecting neurites morphologically respond to nutrient conditions. Moreover, reduced activity of the PG-innervating neurons or of serotonin signalling in the PG strongly correlates with a delayed developmental transition. Our results suggest that serotonergic neurons form a link between the external environment and the internal endocrine system by adaptively tuning the timing of steroid hormone biosynthesis.

immunostained for serotonin (magenta) and GFP (green). The serotonin signals were localised at the tips of SE0 PG neurons (arrows), indicating that serotonin is transported to the axon termini in both conditions. Both the distribution of the neurotransmitters and the projection pattern were affected by food conditions. (g-i) TRH>GFP embryos were placed on a 3.0% agar plate and left to hatch. Newly hatched larvae were picked up and dissected.
Serotonergic neurons were labelled with GFP (g) and the PG was labelled with Sro (h). SE0 PG neurons innervated the PG at the 1 st instar stage before food uptake (yellow arrows). The blue arrows indicate the cell bodies of SE0 neurons in the brain behind the PG. (j) The normalised pupariation ratio in response to food condition changes at the 3 rd instar stage. When the 1 st instar larvae were raised on the yeast-poor food (0.2⨉ yeast) and transferred to the yeast-rich food (2.5⨉ yeast) in the early 3 rd instar stage (~6 h after L2-L3 moulting, A3L), the timing of pupariation recovered by 3 days (blue lines). Conversely, when larvae were transferred from regular food (1⨉ yeast) to yeast-free food (0⨉ yeast), the timing of pupariation was delayed by 1.5 days (black lines). These results suggest that the timing of pupariation responds to nutrient conditions after the 3 rd instar stage. The genotypes shown is TRH-GAL4/UAS-GFP; TRH-GAL4/UAS-mCD8::GFP. Bar, 50 m. Each experiment was conducted independently 3 times.
Embryos were collected at 25°C, and newly hatched larvae were raised at 29°C. Most of the control larvae became pupae at 72-96 hours after hatching (hAH), whereas R29H01>TeTxLC n=245 The same result was obtained in Fig. 4a. Note that the difference in the red lines between (a) and (b) is substantial but smaller than the difference in the green lines between (a) and (b).
These data suggest that nutrient-dependent control of pupariation timing is partially, but not fully, mediated by SE0 PG neurons. The percentages were normalised to the final number of living pupae for each genotype: UAS-TeTxLC[G2]/+; R29H01-GAL4/+ (n=68 in a, 160 in b),

UAS-TeTxLC[G2]/+ (n=199 in a, 133 in b)
. Fly food containing 0.2⨉ or 1⨉ yeast was cooked in the same day and the experiments (a and b) were performed side by side. Each experiment was conducted independently 4-5 times and error bars represent SEM.

Supplementary Table 1: The effects of RNAi for serotonin receptor genes in the PG
The PG-specific RNAi was tested against 5 serotonin receptors. In all crosses, RNAi larvae were raised at 29°C to increase GAL4 expression levels. "+" means that the animals showed no obvious phenotypes and purariated during 96-120 hours after hatching (hAH), while "-" means that the animals did not pupariate at 96-120 hAH. n.d., not determined. Only  Several different GAL4 drivers were tested for the tissue-specificity of 5-HT7-RNAi. When crossed with phm-GAL4, 2-286-GAL4, daughterless-GAL4, and tubulin-GAL4, 5-HT7-RNAi larvae exhibited lethality or developmental delay (coloured). Other tested GAL4s produced no visible phenotype. ""+" means that the animals showed no obvious phenotypes and purariated during 96-120 hours after hatching (hAH), while "-" means that the animals did not pupariate at 96-120 hAH. Each experiment was conducted independently 2 times.