Animals have evolved to maintain homeostasis in a changing external environment by adapting their internal metabolism and feeding behaviour. Metabolism and behaviour are coordinated by neuromodulation; a number of the implicated neuromodulatory systems are homologous between mammals and the vinegar fly, an important neurogenetic model. We investigated whether silencing fly neuromodulatory networks would elicit coordinated changes in feeding, behavioural activity and metabolism. We employed transgenic lines that allowed us to inhibit broad cellular sets of the dopaminergic, serotonergic, octopaminergic, tyraminergic and neuropeptide F systems. The genetically-manipulated animals were assessed for changes in their overt behavioural responses and metabolism by monitoring eleven parameters: activity; climbing ability; individual feeding; group feeding; food discovery; both fed and starved respiration; fed and starved lipid content; and fed/starved body weight. The results from these 55 experiments indicate that individual neuromodulatory system effects on feeding behaviour, motor activity and metabolism are dissociated.
The brain is responsible for maintaining energy homeostasis by monitoring nutritional status and inducing corresponding changes in metabolism, behaviour, and food intake1,2,3,4,5. Many of the brain systems that regulate metabolism and feeding are evolutionarily conserved between mammals and invertebrates, including neuromodulators6, 7. Studying neuromodulators in mammals is made challenging by the expense of experiments and the brain’s immense complexity. These obstacles are overcome with the use of a major invertebrate genetic model system: the vinegar fly, Drosophila melanogaster. Fly neurogenetic tools enable the temporally-resolved manipulation of specific neurons, which minimizes unwanted developmental effects or compensatory mechanisms and permits the analysis of their acute role in energy homeostasis, metabolism, and feeding-related behaviours8,9,10.
Several neuromodulators are believed to have an effect on feeding behaviour: serotonin has been shown to regulate food intake, stomatogastric responses and appetite11,12,13,14; octopamine is believed to modulate feeding and is also regulated by the fly homologs of two obesity-linked genes15; neuropeptide F (NPF) has been shown to respond to feeding-associated signals to help regulate feeding16, 17; and dopamine signalling is required for normal food intake18, 19.
In the present study we examined the effects of silencing neuromodulatory systems on appetitive control, food driven behaviours, metabolism and locomotion. We hypothesized that lesions in these neuromodulatory systems would have coordinated effects on behaviour, feeding and metabolism; for example, that a lesion that increased behavioural activity would also concomitantly increase metabolism and feeding. To examine this hypothesis, we used five transgenic lines to examine the circuit function of five neuromodulators (dopamine, octopamine, tyramine, serotonin, and neuropeptide F) in 11 different assays. We found that feeding phenotypes were not necessarily associated with metabolic or behavioural changes. These results indicate dissociated neuromodulator function—eroding confidence in the hypothesis—and suggesting that the organization of feeding, behaviour and metabolism requires that several modulatory systems act in a coordinated manner.
Systematic literature review of neuromodulators and food intake in Drosophila
A total of 120 articles were identified in the initial search of PubMed; applying the selection criteria reduced this number to three articles (Fig. 1). The three identified articles contained five different experiments that utilized two different methods for assessing food intake: the CAFE assay and the proboscis extension reflex (PER) test. Drosophila feeding behaviour was characterized using these two assays after either silencing or activating dopaminergic, serotonergic, NPF-ergic and octopaminergic (or tyraminergic) neurons (Fig. 2)15, 20, 21. The relative paucity of data relating adult Drosophila neuromodulator function to feeding and metabolism led us to conduct new experiments on this topic.
Silencing of TH-Gal4 neurons reduced activity
We used Gal4 lines with the enhancers of five genes: pale/Tyrosine hydroxylase (TH), Dopa decarboxylase (Ddc), Tyrosine decarboxylase 2 (Tdc2), Tryptophan hydroxylase (Trh) and neuropeptide F (NPF). These driver transgenes were used in combination with the Tub-Gal80 ts conditional repressor transgene22 to drive expression of Kir2.1, an inward-rectifying potassium channel that silences neuronal activity23. Following warm-treatment induction of Kir2.1 expression, motor coordination and activity were compared with two sets of control flies: a driverless line (UAS-Kir2.1; Tub-Gal80 ts / + ) subjected to heat, or flies maintained at 21 °C that had intact Gal80ts repression of Kir2.1 expression but were otherwise genetically identical. Inhibiting electrical activity in TH-Gal4 dopaminergic cells had a profound effect on the activity index, defined as the proportion of time a fly spent moving. The activity index of induced TH > Kir2.1 flies was reduced by −70% compared to uninduced isogenic animals (Fig. 3A). Inhibiting the Ddc-Gal4 cells, which include both dopaminergic and serotonergic neurons, reduced activity by −35%. Flies with silenced Tdc2-Gal4 cells, which include both octopaminergic and tyraminergic neurons, exhibited a +48% increase in activity. Silencing the Trh-Gal4 cells, which includes most of the serotonergic neurons, had only a negligible effect on activity.
Inhibition of TH cells impaired climbing ability
We next assessed climbing ability to determine whether the observed reduction in activity in the transgenic dopamine lines was associated with functional motor deficits24. Temperature-habituated flies expressing Kir2.1 in the TH-Gal4 cells exhibited dramatically worse climbing ability compared to control flies, a climbing index reduction of −0.45. Induced Trh > Kir2.1 flies displayed a moderate decrease in climbing ability, displaying a climbing index reduction of −0.08. All other induced Gal4 > Kir2.1 lines exhibited only small or negligible differences in climbing (Fig. 3B), including a small deficit in Ddc > Kir2.1, which drives in both dopaminergic and serotonergic cells25.
Silencing Trh neurons increased food intake
Neuromodulators are known to affect foraging and feeding behaviours in vertebrates and invertebrates14,15,16, 26, 27. Thus, we measured cumulative food intake in the five driver lines over a period of 6 h in a group of flies using a CAFE assay. Control UAS-Kir2.1; Tub-Gal80 ts /+ flies drank +40% more liquid food after 31 °C warm treatment, which represented the expected change in food intake as a result of warm treatment alone without Kir2.1 induction (Fig. 4). Despite this large change in motor activity in response to temperature, TH > Kir2.1 displayed only a minor change in food intake (Fig. 4A). Interestingly, Trh > Kir2.1 was the only line that displayed a substantial increase (51% over 6 h) in consumption after Kir2.1 induction and the heat recovery period (Fig. 4).
Silencing Ddc, Tdc2 and NPF cells reduced food intake
Data obtained from CAFE assays indicated that the silencing of neuronal cells in several driver lines produced decreases in feeding behaviour in flies that had been starved for 24 h. However, in all cases, the food intake of the uninduced flies was higher than that of the driverless controls: uninduced, Gal80ts-repressed Ddc > Kir2.1, Tdc2 > Kir2.1 and NPF > Kir2.1 flies had +105%, +185% and +166% increased food consumption, relative to UAS-Kir2.1; Tub-Gal80 ts / + controls, respectively (Fig. 4A). One possible explanation is background genotypic differences between controls and experimental lines. Alternatively, since the shared genetic difference between the uninduced experimental flies and the uninduced driverless control is the presence of a driver, the increased intake could also be due to partially incomplete Gal80ts repression at 22 °C20, 28,29,30, though it is unclear why the intake effect might be an increase. Relative to the elevated baseline of uninduced Ddc > Kir2.1 controls, inhibition of Ddc cells reduced their 6 h food intake by 0.11 µl (Fig. 4A). Blockade of neural activity in Tdc2-Gal4 cells also reduced food intake by 0.14 µl; similarly, silencing NPF-ergic neurotransmission decreased food intake 0.15 µl (Fig. 4A). These results indicate that these neuromodulatory systems normally function to promote feeding.
A group feeding assay confirmed two types of feeding changes
The unusual changes in feeding behaviour in the uninduced flies in the single-fly CAFE assay, and large changes in feeding in the induced flies, raised the question as to whether these data were due to isolation of the flies versus group behaviour or might be sporadic results. As such, we replicated all the experiments in a group CAFE assay31. In the group CAFE experiment, the overall effects of genotype and induction were very similar to the original, individual-level data. The increases in the baseline feeding level were also replicated, but were less marked than in the single-fly assay: uninduced Ddc > Kir2.1, Tdc2 > Kir2.1 and NPF > Kir2.1 flies demonstrated a +51%, +77% and +63% increase in feeding, respectively, compared to the uninduced controls (Fig. 4B). This experiment also reproduced the effect of decreased feeding after Kir2.1 induction in these lines, eliciting −29%, −42%, and −25% reductions, respectively (Fig. 4B).
Neuromodulators influence foraging behaviour
To further assess the role of neuromodulators in feeding behaviour we assessed the effects of neural silencing on a food-discovery task32 we refer to as the Small-animal Nutritional Access Chip (SNAC). Liquid food was provided to hungry flies over six sessions and the ability to approach the transient food source was measured. In each epoch, flies had 100 s to find the food outlet alcove. Only induced TH > Kir2.1 animals entered the food alcove markedly less than uninduced controls, with a median of 1/6 possible entries compared to 3/6: Cliff’s delta = −0.62 [95CI −0.47, −0.70], P = 2.5 × 10−10 (Fig. 5A). Induced TH > Kir2.1 flies also travelled a shorter distance before entering the alcove: Δdistance = −65 mm [95CI −33.4, −90.7], g = −0.63, P = 2.6 × 10−10, n = 82, 75 (Fig. 5D). Neither the time to alcove entry nor the path efficiency were affected in any of the lines, although Tdc2 > Kir2.1 induction resulted in a small reduction in the number of alcove entries, from a median of 3 to 2. These data suggest that neuromodulator silencing has a minimal effect on food-discovery in D. melanogaster, with the exception of dopaminergic neuron silencing, as observed in the induced TH > Kir2.1 flies that performed poorly. This effect is likely due to the generally impaired motor function of these flies.
Silencing aminergic circuits affects CO2 production during different nutritional paradigms
Neuromodulators are involved in metabolic homeostasis and energy regulation in both mammals and Drosophila 1,2,3,4,5. We asked, therefore, whether manipulation of neuromodulatory circuits might affect respiration in a nutritional status-dependent manner33. Flies were either allowed to feed ad libitum or were starved for 24 h prior to warm-induced Kir2.1 expression. Warm treatment of control flies at 31°C caused a moderate decrease in the respiration rate of fed control flies decreased with a ΔVCO2 of −0.81 µl/fly/h. Interestingly, in the starved state, silencing cells with either dopaminergic driver (TH-Gal4 or Ddc-Gal4) resulted in decreased respiration and reduced CO2 production where induced TH > Kir2.1 and Ddc > Kir2.1 decreased their CO2 production with 1.63 and 1.32 µl/fly/h, respectively (Fig. 6B). Consistent with their increase in food intake, induced Trh-Gal4 flies also showed an increase of 3.44 µl/fly/h in metabolism after starvation (Fig. 6B).
Silencing neuromodulator cells elicited at most modest changes in fat and weight
We next assessed whether neuromodulator disruptions caused changes in body weight by measuring the mass and lipid content of fed and starved flies. Group comparisons were made with and without Kir2.1 induction in different sets of flies (Fig. 6C). Only minor differences in lipid levels were observed in the majority of the experimental lines, in both the fed and 24 h-starved states (Fig. 6). An exception was observed in the induced Tdc2-Kir2.1 flies, which had a lipid content reduction of 6.08% compared to uninduced controls after 24 h starvation(Fig. 6C); when compared with the warm-treated wild type control flies (UAS-Kir2.1; Tub-Gal80 ts/ + ) the lipid content of Tdc-Kir2.1 flies was reduced by 4.76% (Fig. 6C). Interestingly, none of the flies showed any substantial changes in overall wet weight under any of the experimental conditions except for NPF flies which displayed a modest increase of + 0.14 mg [95CI 0.082, 0.19], g = + 0.97, P = 2.45 × 10−6 in fed flies (Fig. 6D). However, control flies (UAS-Kir2.1; Tub-Gal80 ts/ + ) also underwent a modest weight increase (+0.9 mg), suggesting that the weight gain attributable to NPF-ergic silencing (as opposed to a heat effect) was even more modest. Moreover, weight changes in NPF > Kir2.1 were seen neither in starved flies (Fig. 6D), nor in dry weight measurements (Fig. 6E). We conclude that none of the neuromodulators silenced are capable of modulating major weight changes in a 24 h interval, possibly because insufficient energy is expended over this time.
Dopaminergic effects on motor function and feeding
Previous studies have shown that activating TH-Gal4 cells in transgenic flies promotes the PER21, whereas inhibiting these cells reduces the PER20, suggesting that dopaminergic activity modulates feeding drive (Fig. 2). However, our data indicate that inhibiting TH-Gal4 cells has only a trivial effect on post-starvation food intake (Fig. 4). Rather, silencing the TH-Gal4 cells resulted in substantial effects on overall activity, climbing, food discovery and respiration—outcomes that are consistent with a disruption in motor function (Table 1). From these data we conclude that all the phenotypes we observed in this transgenic line are due to this generalized motor deficit. These findings confirm previous work that has shown that loss of either dopaminergic neurons or the dopamine synthetic enzyme, tyrosine hydroxylase, results in motor impairments18, 34,35,36,37. Silencing the dopaminergic cells in the Ddc-Gal4 driver did not result in motor deficits; rather, inhibiting the Ddc-Gal4 cells (which include both dopaminergic or serotonergic neurons) robustly and specifically suppressed food intake and reduced respiration. While Ddc-Gal4 targets some serotonergic cells, silencing cells using the nearly comprehensive Trh-Gal4 serotonergic driver, had the opposite effect: induced Trh > Kir2.1 flies increased their food intake and produced more CO2 (Table 1). These distinct results suggest that Ddc-Gal4 contains a subset of dopaminergic cells (or maybe serotoninergic cells) that normally act to promote feeding. One candidate group of feeding-promoting cells is the paired antero-medial cells, which are known to promote appetitive memory38, 39 and are numerous in Ddc-Gal4 cells but sparse in TH-Gal4 cells.
Reduced feeding in animals with silenced Tdc2-Gal4 cells
Tyrosine decarboxylase is required for the synthesis of both octopamine and tyramine40. Octopamine is known to have roles in a range of behaviours41,42,43, including appetitive learning38, and starvation-induced hyperactivity44; less is known about the role of tyramine40. Previous work found that activating the Tdc2-Gal4 cells produces a dramatic increase in capillary feeding, while silencing these cells causes only a minor reduction in PER15, 20. In our experiments, silencing Tdc2-Gal4 cells caused a substantial decrease in food intake, consistent with the earlier evidence that Tdc2-Gal4 cells function to promote feeding. Additionally, inhibition of the Tdc2-Gal4 cells was found to increase lipid depletion during starvation (Fig. 6C), but it is unlikely that this response directly relates to the food intake phenotype as both parameters were measured after 24 h starvation.
Other studies have identified specific neuronal populations to be involved in food intake in Drosophila, for example cells captured by the c673a-Gal4 driver45. The c673a-Gal4 subset is a heterogeneous population containing insulin-releasing, dopaminergic and serotonergic neurons, dispersed throughout the brain. Silencing of c673a-Gal4 neurons caused an increase in food intake and fat storage, while the opposite (extreme leanness) was observed when this neuronal population was hyperactivated. These results are contrary to what was observed in the current study, but unlike the study by Al-Anzi et al. our driver lines are specific for each neuromodulator while c673a-Gal4 is a heterogenous population45. As we found that silencing of dopaminergic neurons to decrease food intake and an increase in food consumption when silencing serotonin, it is likely that the observed phenotype for c673a-Gal4 is mediated by the serotonergic population.
Are feeding, activity and metabolism dissociable?
We originally hypothesized that perturbed neuromodulatory function would have consistent behavioural and metabolic effects. A previous study identified a broad, population of insulin-releasing, dopaminergic and serotonergic neurons that caused an increase in food intake and fat storage when silenced, and fat loss when the neuronal population was hyperactivated45; our data suggest that the coordinated effects may be due to this population’s heterogeneous, complex nature. Both the individual and group assays of post-starvation feeding found that silencing the Trh-Gal4 cells led to an increase in food intake (Table 1). These animals also displayed an increased respiration rate, but this effect was not accompanied by an increase in activity. Silencing NPF-Gal4 cells produced decreases in post-starvation feeding as well as a decrease in metabolism during fed state. In line with the metabolic data, we were also able to observe an increase in wet body weight for the NPF-Gal4 flies, while no difference was seen in dry weight, despite having a reduced food intake. These changes indicate that NPF-silencing affects metabolism. Consistent with previous studies, our results are in agreement regarding NPF and its effects on food intake. Loss of function of NPF in Drosophila larvae has been shown to have a negative effect on food intake as it decreases appetite while overexpression increases food intake producing bigger and heavier flies17. Even though silencing TH-Gal4 cells elicited coordinated reductions in activity, climbing ability and CO2 production, these effects did not translate into reductions in feeding, something that might be expected in an animal with reduced energy expenditure from activity. These isolated phenotypes suggest that incapacitating one neuromodulatory system may have specific, dissociated effects on one or several aspects of feeding, activity and/or metabolism. While it is possible that the diverse effects could have arisen from experimental heterogeneity or sampling error, our findings erode confidence in coordinated neuromodulation of behaviour and metabolism, our original hypothesis. Rather, the data suggest that neuromodulatory subsystems regulate physiological functions in a distinct, separable manner.
The present study is the fourth study in adult Drosophila to specifically examine the effects of silencing specific neuromodulatory circuits15, 20, 21. Recently developed, specific drivers, such as split-Gal4 lines targeting dopaminergic cells46, will allow for new analyses of subsets of the neuromodulatory systems examined here. Such future studies will help determine the extent to which multi-phenotypic effects are due to pleiotropy of single neuromodulator cell types, and which effects are due to dissociable functions within the overall system.
Methods and Materials
Fly stocks and maintenance
The five Gal4 driver lines used were TH-Gal4 47, Ddc-Gal4 25, Tdc2-Gal4 48, NPF-Gal4 49, Trh-Gal4 50. Driver lines were outcrossed with the Canton-Special (CS) wild type strain for five generations. Driver lines and CS were then crossed with UAS-Kir2.1; Tub-Gal80 ts flies to produce the male F1 flies used in all behavioural experiments. The CS flies crossed with UAS-Kir2.1; Tub-Gal80 ts were used as a negative control for each assay and are hereafter referred to as controls. Flies (5–8 days old) were maintained at 22 °C, in 60–70% relative humidity, under 12:12 hour artificial light-dark cycles before the experiment day. Neuronal silencing mediated by over-expression of the potassium Kir2.1 channel was induced in flies carrying a Gal4 driver transgene and UAS-Kir2.1; Tub-Gal80 ts by incubation at 31 °C for 24 h followed by 22 °C for 24 h, 1 day prior to the start of the experiment. Post-heat recovery has been previously used with UAS-Kir2.1; Tub-Gal80 ts, for various intervals at different temperatures: behavioural results with Gal80ts control51,52,53 and with GeneSwitch control54 of UAS-Kir2.1 55 indicate that Kir2.1 inhibition persists for at least 24 h53, and as long as 72 h55, after the removal of induction conditions. Note that, as electrophysiological recordings56 from the neuromodulatory neurons were not possible, the presence of sufficient hyperpolarization to suppress action potentials in any (or all) cells cannot be affirmed. For experiments involving starvation, flies were wet starved for 24 h before the start of the experiment during the heat recovery phase; during wet starvation, flies were deprived of food but not water in a vial containing a strip of filter paper soaked in deionized water. Different sets of flies were used for the uninduced and the induced Kir2.1 groups, as well as for the fed and the food-deprived flies.
Systematic review - database search
A systematic literature search was conducted as previously described57. The search expression [(Drosophila or fruitfly or “vinegar fly” or “fruit fly”) AND (feeding or obesity or foraging or starvation) AND (NPF or “neuropeptide F” or octopamine or serotonin or dopamine or tyramine)] was used to query PubMed in September 2016, which resulted in 120 records (Fig. 1). All the bibliographic information, including Title and Abstract, was exported to a Microsoft Excel spreadsheet, and the titles, abstracts and/or full texts were interrogated. The spreadsheet was used to record the results of the full text screen, which included a detailed screen of experimental design.
Systematic review - study selection
The literature selection process was designed to identify experiments that examined the involvement of neuromodulation on feeding in Drosophila by using various feeding assays. Our analysis focused on experiments that aimed to understand the role of neuromodulators on feeding by using inducible neuronal activation or deactivation in genetically modified flies. The 120 records yielded from the PubMed search were screened in four stages based on title review, abstract reading, full text scan and a detailed review of experimental design, to systematically exclude studies that were not relevant. After systematic review, 117 studies were eliminated (Fig. 1), and data were collected from the remaining three studies, as detailed below.
Systematic review - data extraction
The following data were collected from the included studies: authors, year of publication, figure and panel numbers, genotype, mean number of flies in the control and experimental groups and the corresponding standard errors of the mean and sample sizes of each group, age and gender of the flies, type of food used during feeding experiments, and the number of hours of starvation prior to starting the experiment. Numeric data were digitized from graphics with the Measuring Tool in Adobe Acrobat Pro.
Small-animal Nutritional Access Control (SNAC) chip food discovery assay
A microfluidic feeding assay was performed as previously described32. Briefly, each microfluidic SNAC chip contained a 20 × 22 mm arena with a feeding alcove connected to a microfluidic channel that is designed to deliver 70 nl of liquid food via an actuator pump. Eight chips were run simultaneously. Experiments were conducted inside an incubator maintained at 22 °C. The flies were given 10 min to habituate to the chip/incubator environment prior to the start of the experiment. A combined stimulus of blue light with an 85 dB, 300 Hz sound was administered at the onset of the 70 nl liquid food reward32. The liquid food (5% sucrose 10% red food dye in deionized water) was retracted automatically when the fly was detected inside the feeding alcove or a timeout of 100 s was reached. Six foraging-feeding epochs were imposed in succession with a 120 s inter-epoch interval. The following parameters were recorded with custom LabView code as previously described: the number of entries into the feeding alcove, time to feed, path efficiency, and the distance travelled during the time the food was presented32. Task performance was calculated by dividing the number of epochs with successful alcove entries by the total number of epochs. We defined “time to feed” as the latency to enter the feeding alcove, and “path efficiency” as the distance of the most direct path to the feeding alcove divided by the actual distance travelled by the fly in each feeding epoch. For the activity data, two color cameras (A601fc, Basler, Germany) were used in conjunction with custom LabVIEW software to record animal motion. With custom MATLAB software, the activity index was then calculated as the proportion of time that an animal spent moving. This calculation used 0.5 s bins; in this time interval, the fly was determined to be immobile if it moved less than 1 mm. A fly was excluded for the epoch if its total distance traveled was less than 200 mm or if the activity index was below 10%. As the number of alcove entries are ordinal data, we plotted the median and reported a nonparametric effect size, Cliff’s delta. Cliff’s delta values were calculated using the orddom package58 in R. While P values from the Mann-Whitney U distribution were reported pro forma, no significance tests were performed.
Male flies were separated from the F1 progeny and maintained in a freshly prepared food vial 24 h before the start of the experiment. No more than 20 flies were housed within the same vial. Five male flies were put inside a 50 ml disposable polystyrene serological pipet (Fisher Scientific, Waltham, USA) that was cut to 50 mm in length. The top and bottom of the tube were sealed with parafilm that was punctured with three small holes to provide ventilation. The tube was then placed flat on a surface for 1 h at 22 °C while the flies acclimatised. At the beginning of the experiment the tube was tapped 2–3 times to encourage the flies to the bottom of the tube. The time for each fly to reach the top marked point of the tube was then recorded. Any flies that could not reach the top mark within 60 s were denoted as failing to climb. The climbing index was calculated as a range between 0 and 1 for the number of flies that managed to reach the top mark within a certain time. If they failed they got a score of 0, and a score of 1 if they succeeded. An average for all flies was calculated for all the trials.
CO2 metabolism assay
Metabolic changes were analyzed in flies by measuring the CO2 absorbance in two sets of experimental flies: control and food deprived. Male flies were transferred to either a vial containing food (control group) or a vial with water soaked filter paper (starvation group) 24 h before the start of the experiment. Kir2.1 heat induced flies were subjected to the heat recovery period before the starvation period. All flies were maintained at 22 °C under a 12:12 hour light-dark cycle. The respiration chamber consisted of a 1 ml syringe (Becton Dickinson, USA) and a glass microcapillary tube (53432–728, VWR, USA) glued to the 16-gauge needle (Becton Dickinson, USA). A thin layer of absorbent and non-absorbent cotton filled the top of syringe. A fly was anesthetized by cooling and loaded into the syringe. The syringe plunger was inserted to close the chamber, leaving a 1–1.5 cm space for the fly to move. A total of 15 µl potassium hydroxide (KOH, Sigma, USA) was loaded on the top of the syringe and the glass microcapillary tube was assembled onto the syringe. The gap between the syringe and holder were wrapped with parafilm. The respiration chambers were incubated under water inside the 22 °C incubator (Sanyo, Japan) for 10 min. A small volume of 30% sucrose solution containing 10% food colour dye filled the top of the glass capillary to isolate the respiration chamber from the outside environment. The position of the coloured solution was measured 15 min after it was filled and its final position was determined after 1 h. The CO2 metabolic volume consumption during a 1 h period was obtained from the increase in volume within the glass capillary.
Body weight and lipid measurements
Male flies were starved for 1 h under non-dehydrating conditions to limit the amount of food-derived lipids present in the gut. Each group of flies were then weighed to 0.1 mg accuracy using an analytical balance (Sartorious, Singapore). For lipid measurements, groups of five males were weighed to obtain their wet mass and then dehydrated at 65 °C for 24 h and subsequently weighed to obtain their dry mass. Lipid extraction was performed by placing intact, dehydrated flies in a glass vial containing 9 ml diethyl ether for 24 h at room temperature under gentle agitation. After incubation the diethyl ether was removed and the flies were air dried for an additional 24 h at room temperature. The weight of the flies was then remeasured to obtain lean dry mass. The total lipid content of the flies was considered to be the difference between the dry mass and lean dry mass.
Capillary feeding (CAFE) assays
Male flies were anesthetized by cooling and placed in chambers for the CAFE assay, where capillaries delivered liquid food (5% sucrose 10% red colour food dye in deionized water) to the fly. The experiment was conducted within an incubator that was maintained at 22 °C31. The individual Gal4 lines were tested on different days; all experimental conditions were kept constant for all of the driver lines in between and during each experiment with respect to: temperature, humidity, circadian time, days after eclosion and the sex of the flies. The level of the fluid was noted at the beginning of the experiment and 6 h later. A control experiment, whereby no flies were housed inside the chamber, was also conducted in order to calculate the loss of fluid due to evaporation. The difference in the reduction of fluid level between the experimental and control chambers was assigned as the volume partaken by an individual fly. We used a group assay in which 10 flies were assayed in a vial with a single capillary as previously described31, as well as a single-fly assay, in which each capillary was accessible by a single fly kept in a 12 mm × 12 mm × 2 mm chamber cut from acrylic. The food intake assays for the individual fly lines were not performed concurrently but with identical duration, start time for the experiment, age and sex of flies, temperature and starvation time.
Data were analyzed with custom scripts written in LabView, Matlab and Python, and visualized with GraphPad. The data were analyzed with estimation methods to calculate median, mean, mean differences59, Hedges’ g and Cliff’s delta where appropriate. Standardized effect sizes for Hedges’ g are referred to as ‘trivial’ or ‘negligible’ (g < 0.2), ‘small’ (0.2 < g < 0.5), ‘moderate’ (0.5 < g < 0.8) or ‘large’ (g > 0.8) as per standard practice60. Similarly, the Cliff’s delta uses three different thresholds: < 0.147 as “negligible”, < 0.33 “small”, <0.47 “medium” and >0.47 as “large” for description purposes61. Both Hedges’ g and Cliff’s delta are standardized measures of effect size where the later is used for ordinal and non-parametric data. Hedges’ g is an indication of how much two groups differ with each other, i.e., a g of 1 shows that the two groups differ by 1 standard deviation. Cliff’s delta measures how often values in one group are higher than the values in a second group; it ranges between +1 where all values of one group are higher than the other group, and −1 with the reverse relationship (two identical data sets would have a Cliff’s delta of 0). Both significance testing and power calculations were avoided following recommended practice60, 62; the Mann Whitney U statistic was used to calculate P values for pro forma reporting exclusively. To indicate estimate precision, 95% confidence intervals (95CI) were calculated using bootstrap methods and reported in text and/or as error bars60.
Data availability statement
The data and code that support the findings of this study, along with example video files, are available in the Zenodo repository with the identifier https://doi.org/10.5281/zenodo.495697.
Loftus, T. M. et al. Reduced food intake and body weight in mice treated with fatty acid synthase inhibitors. Science 288, 2379–2381 (2000).
Leibel, R. L., Rosenbaum, M. & Hirsch, J. Changes in energy expenditure resulting from altered body weight. N. Engl. J. Med. 332, 621–628 (1995).
Obici, S. et al. Central administration of oleic acid inhibits glucose production and food intake. Diabetes 51, 271–275 (2002).
Harris-Warrick, R. M. & Marder, E. Modulation of neural networks for behavior. Annu. Rev. Neurosci. 14, 39–57 (1991).
Pfaff, D. W., Kieffer, B. L. & Swanson, L. W. Mechanisms for the regulation of state changes in the central nervous system: an introduction. Ann. N. Y. Acad. Sci. 1129, 1–7 (2008).
Baker, K. D. & Thummel, C. S. Diabetic larvae and obese flies-emerging studies of metabolism in Drosophila. Cell Metab. 6, 257–266 (2007).
Melcher, C., Bader, R. & Pankratz, M. J. Amino acids, taste circuits, and feeding behavior in Drosophila: towards understanding the psychology of feeding in flies and man. J. Endocrinol. 192, 467–472 (2007).
Owusu-Ansah, E. & Perrimon, N. Modeling metabolic homeostasis and nutrient sensing in Drosophila: implications for aging and metabolic diseases. Dis. Model. Mech. 7, 343–350 (2014).
Kaun, K. R., Devineni, A. V. & Heberlein, U. Drosophila melanogaster as a model to study drug addiction. Hum. Genet. 131, 959–975 (2012).
Pandey, U. B. & Nichols, C. D. Human disease models in Drosophila melanogaster and the role of the fly in therapeutic drug discovery. Pharmacol. Rev. 63, 411–436 (2011).
Neckameyer, W. S. A trophic role for serotonin in the development of a simple feeding circuit. Dev. Neurosci. 32, 217–237 (2010).
Shimada-Niwa, Y. & Niwa, R. Serotonergic neurons respond to nutrients and regulate the timing of steroid hormone biosynthesis in Drosophila. Nat. Commun. 5 (2014).
Schoofs, A. et al. Selection of motor programs for suppressing food intake and inducing locomotion in the Drosophila brain. PLoS Biol. 12 (2014).
Gasque, G., Conway, S., Huang, J., Rao, Y. & Vosshall, L. B. Small molecule drug screening in Drosophila identifies the 5HT2A receptor as a feeding modulation target. Sci. Rep. 3 (2013).
Williams, M. J. et al. Obesity-linked homologues TfAP-2 and Twz establish meal frequency in Drosophila melanogaster. PLoS Genet. 10 (2014).
Shen, P. & Cai, H. N. Drosophila neuropeptide F mediates integration of chemosensory stimulation and conditioning of the nervous system by food. J. Neurobiol. 47, 16–25 (2001).
Lee, K.-S., You, K.-H., Choo, J.-K., Han, Y.-M. & Yu, K. Drosophila short neuropeptide F regulates food intake and body size. J. Biol. Chem. 279, 50781–50789 (2004).
Riemensperger, T. et al. Behavioral consequences of dopamine deficiency in the Drosophila central nervous system. Proc. Natl. Acad. Sci. USA 108, 834–839 (2011).
Zhou, Q. Y. & Palmiter, R. D. Dopamine-deficient mice are severely hypoactive, adipsic, and aphagic. Cell 83, 1197–1209 (1995).
Marella, S., Mann, K. & Scott, K. Dopaminergic modulation of sucrose acceptance behavior in Drosophila. Neuron 73, 941–950 (2012).
Inagaki, H. K. et al. Visualizing neuromodulation in vivo: TANGO-mapping of dopamine signaling reveals appetite control of sugar sensing. Cell 148, 583–595 (2012).
McGuire, S. E., Mao, Z. & Davis, R. L. Spatiotemporal gene expression targeting with the TARGET and gene-switch systems in Drosophila. Sci. STKE 2004, l6 (2004).
Baines, R. A., Robinson, S. G., Fujioka, M., Jaynes, J. B. & Bate, M. Postsynaptic expression of tetanus toxin light chain blocks synaptogenesis in Drosophila. Curr. Biol. 9, 1267–1270 (1999).
Bartholomew, N. R., Burdett, J. M., VandenBrooks, J. M., Quinlan, M. C. & Call, G. B. Impaired climbing and flight behaviour in Drosophila melanogaster following carbon dioxide anaesthesia. Sci. Rep. 5, 15298 (2015).
Li, H., Chaney, S., Roberts, I. J., Forte, M. & Hirsh, J. Ectopic G-protein expression in dopamine and serotonin neurons blocks cocaine sensitization in Drosophila melanogaster. Curr. Biol. 10, 211–214 (2000).
Zhang, T., Branch, A. & Shen, P. Octopamine-mediated circuit mechanism underlying controlled appetite for palatable food in Drosophila. Proc. Natl. Acad. Sci. USA 110, 15431–15436 (2013).
Vucetic, Z. & Reyes, T. M. Central dopaminergic circuitry controlling food intake and reward: implications for the regulation of obesity. Wiley Interdiscip. Rev. Syst. Biol. Med. 2, 577–593 (2010).
Masek, P. & Keene, A. C. Drosophila fatty acid taste signals through the PLC pathway in sugar-sensing neurons. PLoS Genet. 9, e1003710 (2013).
Pool, A.-H. et al. Four GABAergic Interneurons Impose Feeding Restraint in Drosophila. Neuron 83, 164–177 (2014).
Jeong, Y. T. et al. Mechanosensory neurons control sweet sensing in Drosophila. Nat. Commun. 7, 12872 (2016).
Ja, W. W. et al. Prandiology of Drosophila and the CAFE assay. Proc. Natl. Acad. Sci. USA. 104, 8253–8256 (2007).
Navawongse, R. et al. Drosophila learn efficient paths to a food source. Neurobiol. Learn. Mem. 131, 176–181 (2016).
Zhang, L. et al. TMC-1 attenuates C. elegans development and sexual behaviour in a chemically defined food environment. Nat. Commun. 6, 6345 (2015).
Riemensperger, T. et al. A single dopamine pathway underlies progressive locomotor deficits in a Drosophila model of Parkinson disease. Cell Rep. 5, 952–960 (2013).
Shaltiel-Karyo, R. et al. A novel, sensitive assay for behavioral defects in Parkinson’s disease model Drosophila. Parkinsons Dis. 697564, 25 (2012).
Islam, R. et al. A neuroprotective role of the human uncoupling protein 2 (hUCP2) in a Drosophila Parkinson’s disease model. Neurobiol. Dis. 46, 137–146 (2012).
Coulom, H. & Birman, S. Chronic exposure to rotenone models sporadic Parkinson’s disease in Drosophila melanogaster. J. Neurosci. 24, 10993–10998 (2004).
Burke, C. J. et al. Layered reward signalling through octopamine and dopamine in Drosophila. Nature 492, 433–437 (2012).
Liu, C. et al. A subset of dopamine neurons signals reward for odour memory in Drosophila. Nature 488, 512–516 (2012).
Roeder, T. Tyramine and octopamine: ruling behavior and metabolism. Annu. Rev. Entomol. 50, 447–477 (2005).
Crocker, A. & Sehgal, A. Octopamine regulates sleep in drosophila through protein kinase A-dependent mechanisms. J. Neurosci. 28, 9377–9385 (2008).
Crocker, A., Shahidullah, M., Levitan, I. B. & Sehgal, A. Identification of a neural circuit that underlies the effects of octopamine on sleep:wake behavior. Neuron 65, 670–681 (2010).
Zhou, C. & Rao, Y. A subset of octopaminergic neurons are important for Drosophila aggression. Nat. Neurosci. 11, 1059–1067 (2008).
Yang, Z. et al. Octopamine mediates starvation-induced hyperactivity in adult Drosophila. Proc. Natl. Acad. Sci. USA 112, 5219–5224 (2015).
Al-Anzi, B. et al. Obesity-blocking neurons in Drosophila. Neuron 63, 329–341 (2009).
Aso, Y. et al. The neuronal architecture of the mushroom body provides a logic for associative learning. Elife 23, 04577 (2014).
Friggi-Grelin, F. et al. Targeted gene expression in Drosophila dopaminergic cells using regulatory sequences from tyrosine hydroxylase. J. Neurobiol. 54, 618–627 (2003).
Cole, S. H. et al. Two functional but noncomplementing Drosophila tyrosine decarboxylase genes: distinct roles for neural tyramine and octopamine in female fertility. J. Biol. Chem. 280, 14948–14955 (2005).
Wu, Q. et al. Developmental control of foraging and social behavior by the Drosophila neuropeptide Y-like system. Neuron 39, 147–161 (2003).
Alekseyenko, O. V., Lee, C. & Kravitz, E. A. Targeted manipulation of serotonergic neurotransmission affects the escalation of aggression in adult male Drosophila melanogaster. PLoS One 5, e10806 (2010).
Shih, H.-W. et al. Parallel circuits control temperature preference in Drosophila during ageing. Nat. Commun. 6, 7775 (2015).
Kim, W. J., Jan, L. Y. & Jan, Y. N. Contribution of visual and circadian neural circuits to memory for prolonged mating induced by rivals. Nat. Neurosci. 15, 876–883 (2012).
Mann, K., Gordon, M. D. & Scott, K. A Pair of Interneurons Influences the Choice between Feeding and Locomotion in Drosophila. Neuron 79, 754–765 (2013).
Osterwalder, T., Yoon, K. S., White, B. H. & Keshishian, H. A conditional tissue-specific transgene expression system using inducible GAL4. Proc. Natl. Acad. Sci. USA 98, 12596–12601 (2001).
Depetris-Chauvin, A. et al. Adult-specific electrical silencing of pacemaker neurons uncouples molecular clock from circadian outputs. Curr. Biol. 21, 1783–1793 (2011).
Baines, R. A., Uhler, J. P., Thompson, A., Sweeney, S. T. & Bate, M. Altered electrical properties in Drosophila neurons developing without synaptic transmission. J. Neurosci. 21, 1523–1531 (2001).
Yildizoglu, T. et al. Estimating Information Processing in a Memory System: The Utility of Meta-analytic Methods for Genetics. PLoS Genet. 11, e1005718 (2015).
Rogman, J. J. Ordinal Dominance Statistics (Orddom): an R Project for Statistical Computing Package to Compute Ordinal, Nonparametric Alternatives to Mean Comparison. Version 3.1 (2013).
Claridge-Chang, A. & Assam, P. N. Estimation statistics should replace significance testing. Nat. Methods 13, 108–109 (2016).
Cumming, G. Understanding the new statistics effect sizes, confidence intervals, and meta-analysis. (Routledge, 2012).
Cliff, N. Dominance statistics: Ordinal analyses to answer ordinal questions. Psychol. Bull. 114, 494 (1993).
Altman, D., Machin, D., Bryant, T. & Gardner, S. Statistics with confidence: confidence interval and statistical guidelines. Bristol: BMJ Books (2000).
We thank Nicola Poh Si En for technical assistance with the single-fly CAFE, and Joses Ho for help plotting the systematic review results. We thank Jessica Edwards of Insight Editing London for editing of the manuscript. This work received major support for AE, RN, MR, DC, ZPW, JCS and ACC from A*STAR Joint Council Office grant 1231AFG030 awarded to ZPW and ACC. The authors were supported by a Biomedical Research Council block grant to the IMCB, and a Science and Engineering Research Council block grant to SIMTech. ACC received support from Duke-NUS Medical School, Joint Council Office grant 1431AFG120 and Ministry of Education grant MOE-2013-T2-2-054.
The authors declare an ongoing patent application related to this research.
Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Eriksson, A., Raczkowska, M., Navawongse, R. et al. Neuromodulatory circuit effects on Drosophila feeding behaviour and metabolism. Sci Rep 7, 8839 (2017). https://doi.org/10.1038/s41598-017-08466-0
This article is cited by
State and physiology behind personality in arthropods: a review
Behavioral Ecology and Sociobiology (2022)
Assessing the cognitive status of Drosophila by the value-based feeding decision
npj Aging and Mechanisms of Disease (2021)
Satiation state-dependent dopaminergic control of foraging in Drosophila
Scientific Reports (2018)
Using Drosophila behavioral assays to characterize terebrid venom-peptide bioactivity
Scientific Reports (2018)
By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.