Cellular source and circuit context organize functional specificity in the Drosophila NPF system

Madison N Endres; Tarandeep Singh Dadyala; Kevin W Christie; Irina T Sinakevitch; Lisha Shao

doi:10.64898/2026.09.08.750118

This is a preprint.

It has not yet been peer reviewed by a journal.

The National Library of Medicine is running a pilot to include preprints that result from research funded by NIH in PMC and PubMed.

[Preprint]. 2026 Sep 16:2026.09.08.750118. [Version 2] doi: 10.64898/2026.09.08.750118

Cellular source and circuit context organize functional specificity in the Drosophila NPF system

Madison N Endres ^1,^#, Tarandeep Singh Dadyala ^1,^#, Kevin W Christie ¹, Irina T Sinakevitch ¹, Lisha Shao ^1,^2,^*

PMCID: PMC13596182 PMID: 42779640

Abstract

Neuropeptides regulate diverse and sometimes opposing functions, yet how a single peptide achieves functional specificity remains unclear. Receptor diversity can provide specificity, but not in systems with only one receptor. Drosophila neuropeptide F (NPF) acts through a single receptor, providing a model to test whether specificity is organized by the peptide’s cellular source. Using the whole-brain connectome, we defined four adult NPF cell types with no shared first-order synaptic partners and separate networks. These cell types made distinct and partly opposing contributions to reinforcement, feeding, mating, and energy storage. Most peptide-dependent effects were undetectable after population-level NPF reduction, showing that global perturbation can obscure source-specific functions. Changing neuronal activity and reducing NPF in the same cells produced different, sometimes opposing, phenotypes, so neuronal activation cannot be equated with peptide action. Thus, cellular source and circuit context generate functional specificity that population-level perturbation conceals.

Introduction

Neuropeptides act as neuromodulators that reconfigure neural circuits according to an animal’s internal state and social context¹. Yet individual neuropeptides often regulate a remarkably broad range of physiological and behavioral processes. How a shared peptide signal achieves functional specificity across these diverse outputs remains a central problem in neuropeptide biology^2,3. A paradigmatic example is Neuropeptide Y (NPY), an evolutionarily conserved brain-gut peptide with established roles in feeding, energy metabolism, sleep, stress, reward processing, and reproductive physiology^4–7. This breadth poses an immediate conundrum: how can a single peptide regulate so many, and sometimes opposing, functions with specificity?

One explanation is receptor diversity: NPY acts through multiple receptor subtypes with distinct expression patterns and signaling properties⁸, allowing the same ligand to act differently on different targets. This mechanism cannot explain functional diversification in peptide systems that signal through a single receptor, such as nociceptin/orphanin FQ.

Another complementary mechanism is cellular-source specificity: peptide-expressing cell types may occupy distinct circuit positions, allowing the same peptide to influence different downstream networks^2,3. Yet it remains unclear how peptide-dependent functions are partitioned across these cell types, whether their contributions are concordant or opposing, and how they are represented by population-level perturbation.

Addressing this question in mammals is challenging because neuropeptide-expressing neurons are typically broadly distributed and embedded in complex circuits. In contrast, the genetic and anatomical tractability of the fruit fly Drosophila melanogaster offers an opportunity to resolve these questions at cell-type resolution. Drosophila neuropeptide F (NPF) is a homolog of the vertebrate NPY family⁹ but, unlike NPY, signals through a single known receptor, NPFR¹⁰. Receptor-subtype identity therefore cannot account for the diversification of NPF function, making NPF a tractable system in which to test whether cellular source and circuit context contribute to functional specificity.

Despite this simpler receptor architecture, NPF regulates as broad a range of functions as NPY. It is classically orexigenic: npf expression is associated with food attraction in larvae¹¹, and NPF signaling enables food-deprived animals to overcome aversion and consume noxious food¹². Consistent with this role, satiety-associated signals suppress feeding drive by inhibiting NPF neurons^13,14. Beyond feeding, NPF gates hunger-dependent expression of appetitive memory through mushroom-body dopaminergic neurons¹⁵, promotes starvation-induced wakefulness¹⁶, and links sexual experience to reward state, metabolic physiology, stress resistance, and lifespan¹⁷.

These findings also expose an apparent tension in how NPF has been interpreted across reward contexts. In feeding and memory, increased NPF signaling is associated with deprivation-driven food seeking and memory expression, consistent with a state in which reward is lacking^11,12,15. In male sexual experience, by contrast, successful copulation increases brain NPF, whereas the reduced NPF-system activity that results from sexual deprivation increases ethanol seeking¹⁸. Intriguingly, activation of NPF neurons is itself rewarding and can reduce subsequent ethanol seeking^18,19, consistent with a state of reward attainment¹⁸. These observations arise from different sexes, developmental stages, cell populations, behavioral contexts, and modes of perturbation; apparently opposing phenotypes are therefore not directly contradictory. Nevertheless, they are difficult to accommodate within a unified interpretation of NPF signaling. This raises a specific question: are hunger- and mating-related functions carried by distinct NPF cell types with different circuit outputs? Moreover, it remains unknown how adult NPF cell types are organized across the connectome, how their peptide-dependent functions compare across behaviors and physiology, and whether population-level manipulation preserves or obscures their individual contributions.

In this study, we asked how functional specificity is organized across distinct adult NPF cell types within a one-peptide-one-receptor system and how their individual contributions are represented when the peptide system is manipulated at the population level. We constructed a functional map of the NPF population by combining the Drosophila whole-brain connectome with cell-type-specific genetic manipulation and quantitative behavioral and physiological assays. We defined four sexually monomorphic NPF cell types by morphology and connectivity, generated or obtained drivers that provided genetic access to each population, and compared the effects of cell-type-specific NPF knockdown with neuronal activation or silencing across mating, feeding, reward, learning, and metabolic assays.

We found that distinct NPF cell types occupied largely separate connectomic positions and produced separable phenotypes. Population-level NPF loss failed to reveal most cell-type-specific peptide-dependent phenotypes, whereas neuronal activity manipulations frequently produced effects distinct from NPF knockdown. These findings identify cellular source, circuit connectivity, and signaling repertoire as complementary organizers of functional specificity in the absence of known receptor-subtype diversity.

Results

Anatomically distinct NPF subsets have been linked to reward¹⁹, feeding motivation²⁰, larval sugar-reward learning²¹, and satiety¹³. A recent anatomical survey further resolved approximately 50 adult NPF neurons into five morphological clusters, including ventrolateral neurons that project centrifugally to the optic lobes²². Yet despite this progress, a systematic, connectome-guided, cell-type-resolved functional map of the adult NPF population was lacking. To resolve this gap, we combined the whole-brain connectome (FAFB/FlyWire)^23,24 with NeuronBridge²⁵ and available split-GAL4 resources²⁶ to define NPF cell types by morphology and connectivity, establish cell-type-specific genetic access, and compare their functions.

NPF cell types possess distinct connectomic networks

To characterize the organization of the adult NPF population, we identified four sexually monomorphic NPF-expressing cell types (hereafter NPF cell types) in the FAFB whole-brain connectome^23,24 (Table S1). We named them dorsomedial (DM), lateral type 1-large (L1-l, referred to as L1 hereafter), posterior type 1 (P1; unrelated to the fruitless⁺ P1 courtship-command neurons), and posterior type 2 (P2). The four cell types differed in morphology and projection pattern (Figure 1). We detected no synapses between different NPF cell types, although the bilateral P2 neurons formed within-type connections (Figure S1A).

We next mapped the first- and second-order presynaptic and postsynaptic partners of each cell type (Figure 1A, C, D). The four cell types had largely non-overlapping first-order presynaptic partners, with only one partner shared between DM and L1 (Figure 1E), and completely non-overlapping first-order postsynaptic partners (Figures 1F and S1C), whereas their second-order partner sets overlapped only modestly (Figure 1G, H). This segregation was also evident at the neuropil level. The only input neuropils shared by multiple cell types were superior medial protocerebrum (SMP; DM and L1), superior lateral protocerebrum (SLP; DM and P1), superior intermediate protocerebrum (SIP; L1 and P1), and weakly flange (FLA; DM and P1) (Figure 1I), whereas the only shared output neuropils were the SMP (DM and L1) and SIP (L1 and P1) (Figure 1J).

Distinct partner identities do not preclude communication between their broader networks. We therefore asked whether the first-order partners of one NPF cell type also contacted other NPF cell types and whether the partner populations of different NPF cell types synapsed onto one another. First-order partners contacted their corresponding NPF cell type but not the other NPF cell types (Figures S1B and S1C), and synapses between partner populations were sparse and pair-specific (Figure 1K). Pairwise cosine similarity, which incorporates synaptic weight rather than partner membership alone, likewise showed negligible similarity among first-order partner vectors (Figure S1D) and low similarity among second-order vectors, with the highest values between DM and L1 (Figure 1L).

Together, these analyses show that DM, L1, P1, and P2 are embedded in largely distinct input-output networks. DM and L1 were the most similar pair at the second-order level, but their direct partners and overall connectivity remained separable.

Distinct NPF cell types differentially recruit PAM dopaminergic neurons

Anatomical segregation suggests, but does not establish, functional specialization. In previous work using stochastic genetic intersection, we found that optogenetic activation of DM neurons, but not the other NPF populations tested, induced positional preference¹⁹.

We therefore sought stable genetic access to each cell type for systematic functional comparison. Using the FAFB connectome, NeuronBridge, and GAL4 and split-GAL4 resources^23–26, we generated stable drivers for DM, P1, and P2 and obtained an existing driver for L1²⁷. Immunostaining confirmed that each driver labeled the intended NPF cell type in the central nervous system (Figures 2A and S2A). The drivers showed little or no gut expression, except for weak labeling by the P2 driver in a restricted gut region (Figure S2B). These drivers allowed us to compare the consequences of reducing NPF or manipulating neural activity in each cell type.

Figure 2. — (A) Expression patterns of the split-GAL4 (SS) driver for each NPF subtype in the brain (top) and ventral nerve cord (bottom), revealed by immunostaining of SS-driver>CsChrimson flies. Green, anti-GFP; magenta, anti-nc82 (neuropil); scale bar, 100 μm.

(B) Functional connectivity to PAM dopaminergic neurons (DANs), measured by recording calcium signals from PAM DANs while optogenetically stimulating each NPF subtype (SS-driver>Chrimson88; 58E02>GCaMP7s). Traces show mean ΔF/F per PAM compartment. Activating DM or P2 neurons evoked calcium responses in PAM DANs innervating the γ4 and β′2 compartments, recapitulating the pattern evoked by the full NPF-GAL4 driver.

NPF subtypes were optogenetically activated (SS-driver>CsChrimson) or silenced (SS-driver>A1ACR). In panels C-F, the left subpanel shows activation and the right subpanel shows silencing; on-LED siblings served as controls.

(C) Activating DM neurons produced positional preference for optogenetic stimulation zone, whereas silencing DM produced avoidance.

(D) Pairing an olfactory conditioned stimulus with activation of DM, L1, or P2 generated appetitive memory (positive performance index), whereas silencing did not generate memory.

(E) Activating P1 neurons decreased food consumption, whereas silencing DM or L1 neurons increased it.

(F) Activating DM and P1 neurons reduced female copulation percentage, whereas silencing the individual subtypes did not affect copulation percentage.

(G-J) All NPF-expressing cells were globally activated (NPF-GAL4>CsChrimson) or silenced (NPF-GAL4>A1ACR), with no-LED siblings as controls.

(G) Global activation produced positional preference for the optogenetic stimulation zone, whereas global silencing produced avoidance.

(H) Pairing an olfactory conditioned stimulus with global activation generated appetitive memory; global silencing did not.

(I) Global activation suppressed food intake, whereas global silencing increased it.

(J) Neither global activation nor global silencing affected female copulation rate.

N = 15–25 (positional preference), 11–35 (learning), 22–24 (feeding), n = 29–59 (mating). Data are mean ± SEM. Statistics: Fisher’s exact test (mating); unpaired t-test or Mann-Whitney test (feeding); one-sample t-test or Wilcoxon signed-rank test (positional preference and learning), the criterion for choosing parametric vs non-parametric tests depending on normality by Shapiro-Wilk. *p < 0.05, **p < 0.01, ***p < 0.001.

We first asked whether the four NPF cell types differentially recruit a canonical appetitive reinforcement pathway, the function most consistently attributed to NPF. In adult flies, activation of NPF-expressing neurons can reinforce an associated odor^18,19. NPF signaling also regulates the expression of food-reward memory through mushroom-body-innervating dopaminergic neurons¹⁵, raising the possibility that defined NPF cell types recruit dopaminergic reinforcement pathways. Because protocerebral anterior medial (PAM) dopaminergic neurons mediate positive reinforcement and drive compartment-specific synaptic plasticity in the mushroom body during appetitive learning²⁸, we tested whether activating individual NPF cell types evokes responses in PAM axons.

We optogenetically stimulated each NPF cell type while recording calcium signals from PAM axons that innervate specific compartments of the mushroom body horizontal lobes. Activating DM or P2 neurons evoked calcium responses in PAM axons that innervate the γ4 and β’2 compartments, recapitulating the compartment pattern observed after pan-NPF activation (Figure 2B). In contrast, activating P1 or L1 did not evoke a detectable calcium response in the PAM compartments examined.

These results place DM and P2 neurons functionally upstream of the PAM dopaminergic neurons examined and show that NPF cell types have distinct access to PAM reinforcement circuitry. Because the connectome contained no direct synapses from the four NPF cell types onto PAM neurons, it remains unclear whether DM and P2 recruit PAM through polysynaptic pathways or by peptidergic signaling.

Neural activity manipulations reveal cell-type-specific behavioral functions

We next tested whether differential access to reinforcement circuitry was accompanied by distinct behavioral effects. We first asked whether acutely activating or silencing each NPF cell type carried positive or negative valence in a positional preference assay¹⁹. Flies were allowed to choose between an unstimulated zone and a zone in which a selected cell type was optogenetically activated with CsChrimson²⁹ or silenced with A1ACR³⁰. In this assay, preference for the stimulation zone indicates that the manipulation is appetitive, whereas avoidance indicates that it is aversive.

Activating DM induced positional preference for the stimulation zone, whereas silencing DM induced avoidance. In contrast, manipulating L1, P1, or P2 had no detectable effect on positional preference (Figure 2C). Thus, among the four cell types tested, only DM manipulation altered positional valence.

We then asked whether the valence generated by manipulating neural activity could reinforce an associated olfactory cue. In olfactory conditioning, pairing an odor with activation of DM, L1, or P2 neurons generated appetitive memory (Figure 2D), indicating that each population is sufficient to provide, or recruit, a positive reinforcing signal. By contrast, silencing any individual NPF cell type during odor presentation generated neither appetitive nor aversive memory under these conditions (Figure 2D). Notably, L1 activation supported appetitive memory despite producing no detectable positional preference and no detectable response in the PAM compartments examined. Thus, immediate preference and appetitive reinforcement are supported by overlapping but nonidentical NPF cell types.

The preceding assays used artificial neural stimulation as the valenced or reinforcing event. Because NPF signaling also regulates behaviors directed toward natural rewards, such as food and mating opportunities^11,16,31,32, we next asked whether the four cell types differentially influence food intake and copulation.

We quantified food consumption during cell-type-specific activation or silencing using optoPad^33–35. Activating P1 decreased consumption, whereas silencing DM or L1 neurons increased it (Figure 2E). To assay mating, we paired wild-type males with individual experimental females during optogenetic stimulation. Activating DM or P1 neurons reduced female copulation rate, whereas silencing any individual NPF cell type had no detectable effect (Figure 2F). These results extend the functional segregation of NPF cell types from artificial reinforcement to natural reward-related behaviors.

Having established a cell-type-resolved activity map, we next asked how these contributions combine when the entire NPF population is manipulated. If population-level activity simply aggregated the effects of individual cell types, the direction of cell type-specific phenotypes should be retained; departures from that pattern could instead reflect interactions or nonlinear integration across the population.

Globally activating NPF-expressing cells induced positional preference, whereas global silencing led to avoidance (Figure 2G), matching the directions of DM activation and silencing (Figure 2C). Global activation generated appetitive olfactory memory, whereas global silencing did not support memory formation (Figure 2H), consistent with the effect of activating DM, L1, or P2 individually (Figure 2D). Global activation decreased food intake, whereas global silencing increased it (Figure 2I), matching the directions of P1 activation and DM or L1 silencing, respectively (Figure 2E). In contrast, neither global activation nor silencing affected copulation rate (Figure 2J), despite the reduction produced by activating DM or P1 individually (Figure 2F).

Thus, manipulating activity at the population level preserved the direction of the cell-type-specific effect in positional preference, learning, and feeding, but not mating. The mating exception shows that population-level manipulation does not invariably reveal the effects of its constituent cell types and may instead reflect differential recruitment or nonlinear interactions among populations.

Importantly, however, these activity manipulations do not identify the contribution of NPF itself: optogenetic activation or silencing changes the neurons’ aggregate activity-dependent output, including NPF and any co-released fast-acting transmitters. To distinguish NPF-dependent effects from those of the broader neuronal output, we next reduced NPF selectively within each cell type using cell-type-restricted RNAi against npf.

Cell-type-specific NPF knockdown reveals distinct peptide-dependent functions

The positional-preference and conditioning assays above rely on acute optogenetic stimulation as the valenced or reinforcing event and therefore do not have direct counterparts in constitutive NPF knockdown. We therefore focused the knockdown analysis on feeding and mating, which can be measured without imposed neural stimulation. Because NPF signaling has established roles in energy balance, and because cell-type-specific activity manipulations produced distinct changes in food intake, we also measured triglyceride and glycogen stores to determine whether functional specialization among NPF cell types extends beyond behavior to metabolic physiology.

Knocking down NPF in DM neurons reduced food intake, copulation rate, and glycogen levels (Figure 3A, B, D). Knockdown in L1 neurons reduced food intake, triglyceride, and glycogen (Figure 3A, C, D), whereas knockdown in P1 neurons selectively increased food intake (Figure 3A). Knockdown in P2 neurons selectively reduced glycogen (Figure 3D). The RNAi transgene reduced, but did not eliminate, NPF immunoreactivity in each cell type (Figure S2C); these phenotypes therefore identify functions that are sensitive to partial NPF reduction.

Thus, reducing NPF in different cell types produced distinct, partially overlapping behavioral and metabolic phenotypes.

We next asked whether these cell-type-specific phenotypes remain evident after population-level NPF loss. Pan-NPF knockdown using NPF-GAL4, which labels central neurons and gut enteroendocrine cells, did not detectably alter feeding, mating, or triglycerides but reduced glycogen relative to both parental controls (Figure 3E–H).

Glycogen was reduced after NPF knockdown in DM, L1, and P2 individually and after pan-NPF knockdown, indicating that several cell types make directionally aligned contributions to this metabolic readout. By contrast, the feeding, mating, and triglyceride phenotypes revealed by cell-type-specific knockdown were absent after population-level knockdown. Feeding provides a direct example of how opposing cell-type-specific effects can be obscured: NPF knockdown in DM or L1 decreased intake, whereas knockdown in P1 increased it.

Overall, population-level NPF loss produced a substantially narrower phenotype than cell-type-specific knockdown. The results are consistent with opposing, compensatory, or nonlinear interactions among cellular sources and demonstrate that population-level peptide loss does not provide a complete inventory of cell-type-specific NPF functions.

NPF peptide contributes conditionally to phenotypes induced by pan-NPF activation

Activity manipulation and peptide knockdown therefore provide complementary but non-equivalent views of the NPF system (Figures 2 and 3). Because the two perturbations addressed different questions, the assay sets were not fully symmetric. We examined positional preference and appetitive conditioning only with activity manipulation, where optogenetic stimulation served as the valenced (i.e., reinforcing) event. For peptide knockdown, we focused on feeding and mating, which can be measured without imposed neural stimulation, and additionally measured triglyceride and glycogen stores because NPF has established roles in energy balance and cell-type-specific activity manipulation produced distinct feeding phenotypes.

Comparing these complementary maps revealed an asymmetry in how cell-type-specific effects were represented at the population level (Figure 4A). Population-level activity manipulation reproduced the direction of cell-type-specific effects in positional preference, learning, and feeding, but not mating, whereas population-level NPF knockdown retained only the glycogen phenotype (Figure 4A). Notably, glycogen was the only readout for which knockdown in multiple cell types shifted the measure in the same direction; the cell-type-specific feeding, mating, and triglyceride phenotypes were not evident after pan-NPF knockdown. This contrast indicates that the relationship between cell-type-specific and population-level phenotypes depends on whether the aggregate neuronal output or NPF peptide alone is perturbed. We therefore asked directly how much NPF contributes to the robust behavioral effects of pan-NPF activation.

To address this question, we combined pan-NPF activation with npf RNAi and tested the three assays in which global activation produced a phenotype: feeding, positional preference, and olfactory learning. For feeding, results above showed that global activation suppressed fly’s food intake (Figure 2I). We quantified the activation-induced change relative to the no-LED baseline as (Sip_LED − Sip_NoLED)/Sip_NoLED, where −1 denotes complete suppression and 0 denotes no change. NPF knockdown did not detectably alter activation-induced feeding suppression at the higher stimulation intensity but attenuated it at the lower intensity (Figure 4B). Similarly, NPF knockdown attenuated activation-induced positional preference at the lower, but not the higher, stimulation intensity (Figure 4C). In olfactory conditioning, however, NPF knockdown did not detectably reduce appetitive memory generated by high-intensity pan-NPF activation in either fed or starved flies (Figure 4D). We were unable to test lower-intensity reinforcement because weaker stimulation would shift memory toward the assay floor and reduce sensitivity to genotype-dependent differences.

Together, these experiments show that NPF peptide makes a conditional rather than uniform contribution to the behavioral consequences of activating the NPF neuron population. Its contribution was detectable for feeding suppression and positional preference under weaker stimulation intensities, whereas these phenotypes at higher intensity—and appetitive memory under the conditions tested—remained intact after NPF knockdown. The lack of an effect at high intensity does not establish that NPF is dispensable, because strong activation may recruit sufficient NPF-independent output to saturate the behavioral response and thereby mask a peptidergic contribution. Thus, the principal conclusion is that pan-NPF activation cannot be interpreted simply as NPF release: NPF mediates a context- and intensity-dependent component of the response, while additional activity-dependent signals are likely to contribute to the remainder.

Discussion

A neuropeptide system that signals through only one receptor can be nevertheless partitioned into functionally distinct and sometimes opposing cellular functions. This is the principal finding of our study (Figure 4E). By resolving four sexually monomorphic adult NPF cell types and perturbing each at the levels of neuronal activity and peptide expression, we show that DM, L1, P1, and P2 occupy largely separate networks and make separable contributions to reward-related behavior, feeding, reproduction, and metabolism. Reducing NPF within individual cell types uncovered peptide requirements that were mostly absent after population-level perturbation, whereas manipulating neuronal activity revealed functions not explained by NPF alone. Thus, the functional breadth of the NPF system does not arise from a homogeneous population broadcasting a common message. It emerges from source-defined modules that share NPF and NPFR but differ in connectivity, behavioral role, and signaling repertoire.

Cellular source creates specificity within a single-receptor peptide system

Current frameworks of peptidergic neuromodulation propose that peptide function must be resolved at the level of the producing cells and circuits^2,3. Our results provide a cell-type-resolved test of this principle within one peptide system in which receptor multiplicity cannot explain functional diversification.

The four NPF cell types had largely distinct first-order partners, with only a single presynaptic partner shared between DM and L1, and showed only modest second-order similarity, placing them in largely separate input-output networks (Figure 1). Their functions were likewise distinct but partially overlapping. DM and L1 were the most similar pair in the connectome, yet their functional profiles remained separable: both contributed to NPF-dependent control of feeding and glycogen levels, whereas DM additionally affected copulation success and L1 affected triglyceride levels (Figure 3). Circuit position therefore provides a structural basis for source-dependent functional specialization without implying complete independence: NPF cell types may still be functionally linked through shared neuropils and higher-order partners, and NPF released from one source may reach extrasynaptic targets beyond its direct synaptic network.

Population-level perturbation can obscure cell-type-specific functions

Population-level activity manipulation preserved the direction of cell-type-specific effects in positional preference, appetitive learning, and feeding, but not in mating success (Figure 2). Even within feeding, different cell types appeared to dominate the population-level outcome in the two directions: activating all NPF neurons decreased intake, matching P1 activation, whereas silencing all NPF neurons increased it, matching DM or L1 silencing. Strong or directionally compatible source-specific effects can therefore remain visible at the population level, although the mating result shows that this is not guaranteed.

Population-level peptide perturbation, however, produced a substantially narrower profile (Figure 3). NPF knockdown in DM or L1 decreased feeding, whereas knockdown in P1 increased it; pan-NPF RNAi did not produce a detectable feeding phenotype. By contrast, reducing NPF in DM, L1, or P2 shifted glycogen levels in the same direction, and reduced glycogen remained detectable after population-level disruption. These findings show that global peptide perturbation can obscure cell-type-specific functions, particularly when different cellular sources make opposing contributions, whereas directionally aligned effects are more likely to remain detectable.

This pattern should not be necessarily interpreted as simple cancellation. Perturbation efficacy (RNAi did not eliminate the peptide), developmental compensation, nonlinear interactions, and peripheral NPF sources may also shape the phenotype. Thus, a population-level peptide phenotype is a composite readout rather than an inventory of cell-type-specific function. This finding provides an experimental example of the “conglomerate phenotype” problem proposed for broadly distributed neuropeptide systems².

Neuronal activity is not equivalent to peptide action

Activity manipulations test the aggregate output of NPF neurons, whereas npf-RNAi tests the specific contribution of NPF while leaving the rest of the neuronal output intact. Agreement between the two manipulations supports a peptide contribution; divergence may reflect co-released signals, interactions among outputs, or differences in perturbation efficacy and timescale.

In DM and L1, neuronal silencing increased feeding whereas NPF reduction decreased it (Figures 2E and 3A). These divergent effects show that neuronal silencing and peptide depletion are not interchangeable. They are consistent with DM/L1-derived NPF promoting intake while the aggregate activity-dependent output of these neurons suppresses it. In P1, by contrast, activation decreased feeding whereas NPF reduction increased it, supporting a contribution of P1-derived NPF to feeding suppression.

Combining pan-NPF activation with NPF knockdown directly tested whether NPF contributes to phenotypes produced by activating the entire NPF population (Figure 4B–D). If the activation phenotype depends in part on NPF, reducing the peptide should attenuate the response. Reducing NPF attenuated feeding suppression and positional preference at the lower stimulation intensity, supporting a peptide contribution to both responses. The same knockdown did not detectably alter either phenotype at higher intensity and did not reduce appetitive memory at the high intensity used for that assay.

These results do not indicate that NPF acts only at low stimulation intensity. Rather, stimulation intensity determines whether its contribution is experimentally detectable. Under weaker activation, partial loss of NPF is behaviorally consequential. Under stronger activation, residual peptide (Figure S2C) may be released in sufficient amounts, or stronger activation may recruit enough co-released signals to preserve the response despite reduced NPF. These mechanisms are not mutually exclusive and cannot be distinguished here, but both explain why high-intensity activation can conceal the peptide contribution that is detectable at lower intensity.

Together, these experiments establish that NPF is a genuine contributor to the behavioral output: reducing the peptide attenuates phenotypes produced by activating the neurons that release it. They also show that the peptide is not account for the neuron’s complete output: under strong activation, the same reduction had no detectable effect. Residual peptide, co-released signals, or both may preserve the response. Activating an NPF neuron is therefore not equivalent to releasing NPF, and how far the two diverge depends on how strongly the neuron is stimulated.

Cell-type specialization reframes an apparent NPF reward paradox

Cell-type specialization provides a framework for reconciling an apparent tension in the NPF literature. In food consumption, the NPF system is engaged when food reward is lacking: NPF promotes food seeking and permits hunger-dependent appetitive memory expression^11,12,15. In sexual experience, the NPF system is engaged when reward has been obtained: copulation increases brain NPF, whereas sexual deprivation reduces NPF-system activity and increases ethanol seeking; activating NPF neurons is itself reinforcing and reduces subsequent ethanol seeking^18,19. Thus, the NPF system has been linked both to pursuing an absent reward and to the state following reward attainment. These studies differ in sex, developmental stage, cell population, and assay, but together they resist a single unitary interpretation of NPF as either a deprivation or attainment signal.

The reinforcement data provide initial evidence that distinct NPF cell types can recruit different reward-related outputs. DM and P2 activation recruited PAM dopaminergic compartments and supported appetitive memory, whereas L1 supported appetitive memory without detectable recruitment of the PAM compartments examined or immediate positional preference (Figure 2), a pattern consistent with unsampled PAM pathways, subthreshold recruitment, or a distinct reinforcement route. The previously described NPF-PPL1 pathway provides one precedent for NPF regulation of an alternative dopaminergic pathway²⁷, although its relationship to L1 remains unresolved. Thus, NPF cell types can access reinforcement through more than one circuit architecture.

Resolving the apparent reward paradox will require measuring natural activity in defined NPF cell types during hunger, feeding, sexual deprivation, and copulation, and determining whether NPF and co-released signals act in the same or opposite directions in each pathway.

A single-receptor architecture with a mammalian counterpart

Together, these findings establish cellular source as a determinant of NPF function at the levels of circuit position, population phenotype, and the relationship between neuronal activity and peptide action. This principle may extend to other peptide systems in which diverse sources signal through a shared receptor.

Mammalian NPY provides a different context. NPY is produced by hypothalamic AgRP neurons, cortical and hippocampal interneurons, brainstem catecholaminergic populations, and other cell types embedded in distinct circuits^5,36–41; its actions are further diversified by multiple receptor subtypes⁸. Our study provides a tractable experimental demonstration that cellular diversity can map onto distinct circuit positions and separable behavioral and physiological functions within a single NPY-family system.

Drosophila NPF isolates the contribution of cellular source because it signals through a single known receptor, NPFR¹⁰; known receptor-subtype identity therefore cannot partition its functions. The mammalian nociceptin/orphanin FQ (N/OFQ) system has a similar architecture as NPF: N/OFQ acts through the broadly distributed NOP/OPRL1 receptor, with no established receptor subtypes⁴², while prepronocicep3n precursor (Pnoc) is expressed in anatomically distinct neuronal populations with sharply different behavioral roles.

Three Pnoc populations illustrate this functional diversity. Central-amygdala Pnoc neurons are required for palatable food consumption and support reward⁴³; paranigral VTA Pnoc neurons constrain reward seeking through local NOP signaling⁴⁴; and extended-amygdala Pnoc neurons encode salient stimuli and promote defensive behavior⁴⁵. Thus, one peptide acting through one receptor can support functions that receptor identity alone cannot distinguish.

Furthermore, many studies of Pnoc-expressing neurons manipulate the cells’ aggregate output, which can include co-released classical transmitters and other peptides, leaving unresolved whether neuronal function can be equated with N/OFQ function. Our matched activity- and peptide-level perturbations show that neuronal function and peptide function can diverge (Figures 2–4). Thus, functional specificity in single-receptor peptide systems may reflect both source-specific circuit organization and the distinct signaling repertoire of each peptide-producing cell type.

Limitations and future directions

The study examines four sexually monomorphic central cell types, not the entire NPF system. Synaptic connectomics cannot identify the complete set of extrasynaptic peptide targets. Characterizing sexually dimorphic and peripheral cell types, validating co-transmitters, mapping NPFR-expressing cell types, and combining cell-type-specific activation with adult-restricted peptide knockdown will be essential next steps.

Some of these unresolved questions are now experimentally tractable because of the cell-type-specific drivers developed here. Stable drivers for DM, L1, P1, and P2, validated anatomically and linked to their connectomic identities, permit these populations to be monitored and manipulated independently. Combined with temporal control and peptide-specific perturbations, these reagents provide a foundation for recording endogenous activity, identifying downstream NPFR-expressing targets, and distinguishing peptide-dependent from co-transmitter-dependent functions across physiological and behavioral states.

Together, these findings establish cellular source as an organizing principle of NPF signaling: one peptide acting through one receptor can produce distinct, and sometimes opposing, behavioral effects depending on the neurons from which it is released. By separating peptide action from the aggregate output of these neurons, our study further shows that peptidergic function is shaped jointly by source-specific circuit organization and the broader signaling repertoire of each cell type. Thus, understanding a peptidergic system requires defining not only the peptide and its receptor, but also which cells release it, which circuits they engage, and what other signals those cells convey.

Materials and Methods

Fly stocks and culture

Drosophila melanogaster stocks (listed in key resources table) were raised at 25°C and 50%-60% relative humidity on commercially prepared standard media (cornmeal/agar/molasses/yeast; Nutri-Fly MF) for anatomical and ‘Con-Ex’ feeding experiments. For optogenetic behavioral and ex vivo functional imaging experiments, flies were raised under constant darkness at 25°C and 50%-60% relative humidity on the same media supplemented with 0.2 mM all trans-retinal. Flies were collected 0–3 days post-eclosion on media containing 0.4 mM all trans-retinal. Flies for anatomical and behavioral assays were 4- to 7-days-old at the time of experiments; flies used for ex vivo functional connectivity calcium imaging were 2- to 5-days-old at the time of dissection. Mated female flies were used in this study, with the exception of copulation experiments, where unmated female flies were used.

Key Resources Table

Reagent Type (Species or Resource)	Designation	Source	Identifier	Assay
Genetic Reagent (D. melanogaster)	UAS-NPF-RNAi (I)	Shao Lab	VDRC:v108772	All
Genetic Reagent (D. melanogaster)	empty-split-GAL4	Shao Lab	BDSC:79603	All
Genetic Reagent (D. melanogaster)	DM-spGAL4/TM6B	This Study	N/A	All
Genetic Reagent (D. melanogaster)	L1-spGAL4	Feng et al., 2021²⁷		All
Genetic Reagent (D. melanogaster)	P1-spGAL4	This Study	N/A	All
Genetic Reagent (D. melanogaster)	P2-spGAL4	Shao et al., 2017¹⁹	N/A	All
Genetic Reagent (D. melanogaster)	UAS-A1ACR-EYFP (attP18)	Bushey et al., 2025³⁰	N/A	Optogenetic Experiments
Genetic Reagent (D. melanogaster)	UAS-CsChrimson-mVenus (attP18)	Klapoetke et al., 2014²⁹	BDSC: 55134	Optogenetic Experiments
Genetic Reagent (D. melanogaster)	UAS-Chrimson88-tdTomato (attP18)	Strother et al., 2017⁴⁶	N/A	Calcium Imaging
Genetic Reagent (D. melanogaster)	LexAop-CsChrimson88-tdTomato (attP18)	Y. Shuai	N/A	Calcium Imaging
Genetic Reagent (D. melanogaster)	R58E02-LexA (attP40)	Pfeiffer et al., 2010⁴⁷	BDSC: 52740	Calcium Imaging
Genetic Reagent (D. melanogaster)	LexAop-jGCaMP7s (Su(Hw)attP8)	A. Wong	N/A	Calcium Imaging
Genetic Reagent (D. melanogaster)	NPF-GAL4	Shao Lab (Wu et al., 2003¹¹)	N/A	All
Genetic Reagent (D. melanogaster)	w ¹¹¹⁸	Shao Lab	N/A	All
Genetic Reagent (D. melanogaster)	Elav-GAL4	Shao Lab		IHC
Antibody	Rabbit anti-GFP	Invitrogen	A-6455; RRID:AB_221570	IHC
Antibody	Mouse nc82	Hybridoma Bank DSHB	NC82; RRID:AB_2314866	IHC
Antibody	Rabbit anti-NPF	Ray Biotech, Inc	RB-19-0001	IHC
Antibody	Mouse anti-GFP	Millipore Sigma	G6539	IHC
Antibody	Alexa 488 goat anti-rabbit	Invitrogen	A-11008	IHC
Antibody	Alexa 488 goat anti-mouse	Invitrogen	A-11001	IHC
Antibody	Alexa 568 goat-anti-mouse	Invitrogen	A-11031	IHC
Antibody	Alexa 568 goat-anti-rabbit	Invitrogen	A-11011	INC
Chemicals	Normal goat serum	Gibco	16210-064	IHC
Chemicals	Paraformaldehyde, 20% aqueous solution	Electron Microscopy Science	15713	IHC
Chemicals	Phosphate buffer saline	Apex BioResearch	18-244	IHC
Chemicals	Schneider’s Drosophila medium	Gibco	21720-024	IHC
Chemicals	Agar Powder	Thermo Scientific	9002-18-0	flyPad/optoPad
Chemicals	All trans-retinal	Spectrim Chemical Mfg. Co.	R3041-1GM	Optogenetic Experiments
Chemicals	Amyloglucosidase	Sigma-Aldrich	10115-1G-F	Metabolism
Chemicals	0.1 M Citric Acid	Fisher	5949-29-1	Metabolism
Chemicals	D-Sucrose	Fisher BioReagents	BP220-212	flyPad/optoPad
Chemicals	Glucose Standard Solution	Sigma-Aldrich	G6918	Metabolism
Chemicals	Glycerol Standard Solution	Sigma-Aldrich	G7793	Metabolism
Chemicals	Infinity Triglyceride Reagent	Fisher BioReagents	TR22421	Metabolism
Chemicals	5 M NaCl	Sigma-Aldrich	7647-14-5	Metabolism
Chemicals	Pointe Scientific Glucose Oxidase	Fisher BioReagents	23-666-290	Metabolism
Chemicals	Protease Inhibitor	Sigma-Aldrich	04 693 124 001	Metabolism
Chemicals	0.1 M Sodium Citrate	Sigma-Aldrich	6131043	Metabolism
Chemicals	1 M Tris-HCl	KD Medical	RGE-3340	Metabolism
Chemicals	Triton-X	Sigma-Aldrich	9002–93-1	Metabolism
Chemicals	Yeast Flakes	Any Lab Supplier	N/A	Choice
Drosophila Medium	Nutri-fly Molasses Formulation Fly Medium	Genesee Scientific	66–123 https://geneseesci.com	All
Equipment	flyPAD and optoPAD High Throughput Feeding Assay	Itskov et al. 2014	https://flypad.rocks/	flyPad/optoPad
Equipment	Microplate Reader	N/A	N/A	Metabolism
Equipment	Modular In-Vivo Multiphoton Microscopy System (MIMMS) v2.0	Janelia Research Campus, HHMI	https://www.janelia.org/open-science/mimms-22-2024; RRID:SCR_016511	Calcium Imaging
Software, Algorithms	BioRender	BioRender	https://biorender.com	Data Analysis
Software, Algorithms	GraphPad Prism 10 Software	GraphPad Software	https://www.graphpad.com/features	Data Analysis
Software, Algorithms	MATLAB R2024a software	The Mathworks, Inc.	https://www.mathworks.com/	Data Analysis
Software, Algorithms	Python Programming Language Software	Python Software Foundation	https://www.python.org; RRID:SCR_008394	Mating; Connectome Analysis

Open in a new tab

Connectome analysis

Connectivity of NPF subtype neurons was analyzed using the FlyWire/Codex whole-brain connectome of the FAFB (snapshot 783). Custom Python code, adapted from the analysis framework of Walker et al., 2025⁴⁸, was used to extract each subtype’s direct (first-order) and second-order pre- and postsynaptic partners, applying minimum synapse-count thresholds of 5 and 10 synapses, respectively, to exclude weak connections. Partner overlap among subtypes was visualized with Venn diagrams and quantified using the Jaccard similarity index; the significance of pairwise partner overlap was assessed by hypergeometric tests with Benjamini-Hochberg false discovery rate correction, and differences between input- and output-partner sharing were evaluated using Fisher’s exact test and a paired Wilcoxon signed-rank test across subtype pairs. Additional analyses characterized each subtype’s neurotransmitter composition, neuropil innervation profile, downstream/upstream partner super-class composition, cosine similarity of synapse-weighted partner vectors, and input/output synapse polarity. All analysis code is available at https://github.com/lishashao/NPF-subtype-connectome-analysis.

Behavioral assays

Positional preference

Positional preference assays were performed as described previously at 25°C, 50%-60% relative humidity in darkness¹⁹. Female flies (3–7d old) were loaded into rectangular chambers, 8 flies per chamber. After a 2-min baseline period, one side of the chamber (selected at random) was illuminated by 0.5 Hz 1-second pulsed 625-nm red LED at designated light intensity (indicated in figure legends) for 3 min. Following the illumination period, there was a 2-min recovery period. Experiments were recorded with infrared cameras (BlackFly USB3 camera, Teledyne vision solutions). The frequencies and intensities of LED stimulation were controlled using MATLAB (MathWorks, Inc.). Acquired video was analyzed with custom MATLAB scripts that detected the position of the flies, allowing for the calculation of a Preference Index as the average difference in the fractions of flies on illuminated versus unilluminated sides.

Associative learning

Associative learning assays were performed on female flies with optogenetic stimulation and were conducted in a circular arena that couples optogenetic stimulation with an odor delivery system, as described in Aso et al., 2016²⁸. All tests were performed at 25°C, 50%-60% relative humidity. The olfactory cues used were 4-methylcyclohexanol (1:1000) and 3-octanol (1:1000). Optogenetic stimulation used a 0.5 Hz 625-nm LED light at designated intensity (indicated in figure legends). In optogenetic learning assay, female flies were trained as groups of 12–14 individuals using a single training session consisting of 1-min exposure to air, followed by a 1-min exposure to odor A (CS+) paired with neuronal activation/silencing, followed by a 1-min exposure to air, and finally a 1-min exposure to odor B. Memory was tested 1 min after training, during which the odors A and B were delivered to each pair of opposing quadrants. Experiments were recorded under IR illumination with a camera (Point Gray Research) located above the chamber and equipped with an IR long-pass filter. The subsequent video was analyzed with custom MATLAB scripts that detect the position of flies within the chamber. A Performance Index for odor A (paired with LED, CS+) is the average difference in the fractions of flies on the odor-A versus odor-B quadrants. A reciprocal group received odor B as CS+ (paired with LED) to counterbalance flies’ intrinsic odor bias. The final Performance Index is the mean of PI(A-CS+) and PI(B-CS+).

Mating

The courtship assay assessed copulation behavior between a virgin experimental female fly and a sexually mature CantonS (CS) male fly. Individual female and male flies (4–7d old) were loaded separately using a fly aspirator into a small circular chamber (1-cm in diameter). The flies were initially separated by a thin movable barrier to prevent premature interaction. At the start of the assay, the barrier was removed and behavior was recorded for 70 minutes. In optogenetic experiments, chambers were illuminated with 625-nm red LED light delivered at 1s intervals to stimulate channelrhodopsin-expressing neurons. Mating behavior was analyzed using Fly Flirt Python-based software, which detected copulation events, including copulation onset and duration. Flies that did not mate during the recording period were also identified by the software and included in data analysis.

flyPad/optoPad feeding

This assay was performed as described previously by using the flyPad behavioral monitoring system,25,26 at 25°C, 50%–60% relative humidity³⁵. Warmed, liquefied media were loaded into both probes in each FlyPAD chamber in darkness. Individual female flies (3–7 d old) were placed in each flyPAD chamber and were allowed to feed ad libitum for 50 min total. In the open-loop feeding and choice assay, the flyPAD chambers were illuminated with 625-nm red LEDs to optogenetically stimulate channelrhodopsin-expressing neurons. In the closed-loop feeding assay, the FlyPAD chambers were only illuminated with red LED when flies fed on one of the two probes in each arena. The probe associated with illumination was alternated for the next set of experiments to mitigate potential probe-specific bias. Media containing 7% yeast, 160 mM sucrose, and 0.7% agar was used for open-loop feeding assay. Media containing either 0.7% agar (neutral taste), 100 mM sucrose (sweet taste), or 0.5 mM quinine (bitter taste) were used for open-loop choice assays and closed-loop feeding assays. Analysis of feeding behavior used custom MATLAB scripts that extracted the total number of sips during the feeding period.

Metabolites measurement

Metabolic activity was assessed by quantifying whole-fly triglyceride, glycogen, and free glucose levels using a protocol adapted from Murakami et al., 2016⁴⁹. Two female flies (3–7d old) were euthanized and homogenized in 250 μL of lysis buffer supplemented with protease inhibitors at a 7X concentration. Samples were centrifuged at 13,000 RPM for 15 minutes at 4°C and the resulting supernatant (~200 μL) was used for further metabolite assays. For each metabolic assay, 10 μL of the supernatant was loaded in triplicate into a 96-well plate. To quantify triglycerides, 90 μL of infinity triglyceride reagent was added to the sample-filled wells. Similarly, to quantify free glucose, 90 μL of glucose oxidase reagent was added to additional sample-filled wells on a separate 96-well plate. Reagents were stored at 4°C and incubated with the samples at 37°C for 5 minutes with shaking. To measure glycogen, 20 μL of amyloglucosidase was added to 20 μL of supernatant to hydrolyze the α-1,4 glycosidic bonds, followed by incubation at 37°C for 2 hours. After incubation, samples were loaded in triplicate into 96-well plate, and 90 μL of glucose oxidase reagent was added. Absorbance was measured at 500 nm for all assays. Metabolite concentrations were calculated via standard curves and then converted to usable mass values for data analysis.

Immunostaining and confocal imaging

Immunohistochemical staining and imaging were performed as follow. Briefly, female fly brains and guts were dissected in cold Schneider’s Insect (S2) medium, fixed in 2% paraformaldehyde in S2 at Room Temperature (RT) under nutation for 55–65 min in covered Eppendorf tubes, washed 3–4 times with PBT0.5 (phosphate buffered saline, PBS, containing 0.5% Triton X-100) at RT, blocked with PAT3 (1xPBS,, 3% Normal Goat Serum, 0.5% Triton X-100) for 1 hour at RT, incubated with primary antibody (1:500 rabbit anti-GFP, 1:50 mouse anti-nc82 in PAT3) or (1:250 mouse anti GFP and rabbit anti-NPF Supplementary figure)at 4°C overnight, washed 3–4 times in PBT0.5, incubated with secondary antibody (1:500 Alexa 488 anti-rabbit, 1:500 Alexa 568 anti-mouse in PAT3) or ( 1:500 Alexa 488 anti-mouse, 1:500 Alexa 568 anti-rabbit for NPF staining) ) at 4°C overnight, washed with PBT0.5 and PBS, and mounted in Vectashield. For control of NPF staining when the anti- NPF were omitted while both secondary present, the specific staining was absent.The brains were then imaged on a confocal microscope (Zeiss CellDiscoverer 7, using Zen Blue software, Zeiss LSM880, using Zen Black software).

Ex vivo calcium imaging

All functional imaging was performed on 2- to 5-day-old female flies. Whole brains were dissected in a Sylgard silicone elastomer (Dow Inc.) coated dish, immersed in artificial fly hemolymph (103 mM NaCl, 3mM KCl, 2 mM CaCl2, 4 mM MgCl2·6 H2O, 26 mM NaHCO3, 1 mM NaH2PO4, 8 mM trehalose, 10 mM glucose, 5 mM TES, bubbled with 95% O2/5% CO2). Brains were placed on a poly-Lysine coated coverslip and imaged with a resonant-scanning 2-photon microscope (MIMMS v2.0, Janelia Research Campus), using a laser with a 920 nm excitation wavelength (Toptica FemtoFiber Ultra 920) and an excitation power of 19–33 mW (2.7–4.3 mW/mm2 intensity). Emitted fluorescence was detected with GaAsP photomultiplier tubes (Hamamatsu) controlled via a Janelia PMT Controller. Images were acquired with an Olympus 40x water-dipping, 0.8 numerical aperture objective at 512 pixels × 512 pixels resolution. Optogenetic stimulation of CsChrimson was provided by a 660-nm red light LED diode (ThorLabs M660L4), fed into a digital mirror device (Texas Instruments DLP Lightcrafter 4500) and projected through the objective. Photostimulation consisted of 4 × 1 s pulses, separated by a 24 s inter-pulse interval, with a stimulation power of 20–21 μW (2.8–3.0 μW/mm2 intensity). Volume stacks were continuously imaged at 2x magnification, and consisted of 45 slices, with an inter-slice interval of 2 μm, thus allowing all the PAM-DANs compartments to be imaged. Stack acquisition began 10s prior to the 1st LED pulse and stopped 24s after the termination of the final pulse. After optogenetic stimulation, brains were imaged for ROI definition of mushroom body compartment anatomy using the same imaging stack dimensions as above, using both 920 nm, and 1050 nm excitation (Toptica FemtoFiber Ultra 1050). Then power intensity for imaging at the 920- and 1050-nm excitation was 3.2–3.4 and 3.1–3.2 mW/mm2 respectively.

Regions of interest (ROI)s for each brain were drawn using custom MATLAB software and the 1050 nm anatomy images described above. The average pixel intensity for each ROI ( $fluo r_{a v g}$ ) was calculated, as was its average baseline intensity ( $fluo r_{bseln}$ ), defined as the averaged intensity for the 5 volumes (approximately 1.5 s) acquired immediately prior to each optogenetic pulse onset. Average background intensity ( $fluo r_{bckgr}$ ) for each volume was calculated from a ROI located in non-brain-tissue containing space. For each PAM compartment, ΔF/F was calculated using the formula:

\frac{(fluo r_{a v g} (t) - fluo r_{bckgr} (t)) - (fluo r_{bseln} (t) - fluo r_{bckgr} (t))}{(fluo r_{bseln} (t) - fluo r_{bckgr} (t))}

Statistical analysis

Analyses were performed at the level of the experimental unit. For single-fly assays, the unit of analysis was the individual fly, such that each fly contributed one value (thus N = number of flies). For group assays (e.g., arena-based population tests), the unit of analysis was an independently loaded cohort of flies tested together in a single chamber/arena; all flies in that cohort were aggregated to yield one summary value (e.g., PI or total sips), and that value contributed N = 1. In these group assays, the number of flies per cohort (n) is reported descriptively in figure legends but was not used for statistical inference to avoid pseudo-replication. When cohorts were measured more than once under different conditions, within-cohort comparisons were analyzed with paired tests; otherwise, unpaired tests were used across independent cohorts.

For each dataset (and, where applicable, model residuals for >2-group designs), distributional normality was assessed with the Shapiro-Wilk test (α = 0.05). All hypothesis tests were two-tailed. When the normality criterion was met, two-group comparisons used paired or unpaired t-tests as appropriate to the design, and >2-group comparisons used one-way ANOVA followed by Tukey’s multiple-comparisons test. When the normality criterion was not met, the behavioral index was tested against 0 using a one-sample Wilcoxon signed-rank test; two-genotype comparisons used the Mann-Whitney U test; and >2-genotype comparisons used the Kruskal-Wallis test with Dunn’s post hoc multiple-comparisons procedure.

The exact test used, sample size (n), and two-tailed p-values are provided in the figure legends. Analyses were performed in GraphPad Prism.

Supplementary Material

Supplement 1

media-1.pdf^{(2.4MB, pdf)}

Acknowledgements

We thank Yoonwoo Park and Sabina Knox for their contributions to the early stage of this project. We thank Drs. Ulrike Heberlein, Amber Krauchunas, Jeremy Bird, and Jon-Michael Knapp for discussion and feedback. We thank We thank Dr. Meet Zandawala for valuable feedback on cell-type iden3fica3on and technical assistance. We thank Drs. Ann-Shyn Chiang and Bloomington Drosophila Stock Center (BDSC) for providing fly strains. We thank Dr. Anita Devineni for sharing the code for connectome analysis and Mr. Sina Mirzaei for initial analysis. We acknowledge the Princeton FlyWire team and members of the Murthy and Seung labs for development and maintenance of FlyWire (supported by BRAIN Initiative grant MH117815 to Murthy and Seung). This work was supported by UD-GUR, UD-UDRF, Delaware CTR program with a grant from the National Institute of General Medical Sciences (NIGMS, U54 GM104941), Delaware INBRE program with a grant from the NIGMS (P20 GM103446), and Maximizing Investigators’ Research Award to L.S. (NIGMS R35GM147504).

Data Availability

Requests for further information and resources should be directed to and will be fulfilled by the lead contact, Lisha Shao (shaol@udel.edu).

Materials availability

Fly lines generated in this study are available upon request from the lead contact.

Data and code availability

The code used for connectome analysis is described in Materials and Methods, and available at github.

Raw data for imaging and behavioral experiments are available upon request from the lead contact.

Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.

Reference

1.Nässel D. R. & Zandawala M. Recent advances in neuropeptide signaling in Drosophila, from genes to physiology and behavior. Progress in Neurobiology https://doi.org/10.1016/j.pneurobio.2019.02.003 (2019) doi: 10.1016/j.pneurobio.2019.02.003. [DOI] [PubMed] [Google Scholar]
2.Nässel D. R. & Zandawala M. Endocrine cybernetics: neuropeptides as molecular switches in behavioural decisions. Open Biol. 12, 220174 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
3.Nässel D. R., Pauls D. & Huetteroth W. Neuropeptides in modulation of Drosophila behavior: how to get a grip on their pleiotropic actions. Curr. Opin. Insect Sci. 36, 1–8 (2019). [DOI] [PubMed] [Google Scholar]
4.Cowley M. A. et al. Integration of NPY, AGRP, and Melanocortin Signals in the Hypothalamic Paraventricular Nucleus. Neuron 24, 155–163 (1999). [DOI] [PubMed] [Google Scholar]
5.Sajdyk T. J. et al. Neuropeptide Y in the Amygdala Induces Long-Term Resilience to Stress-Induced Reductions in Social Responses But Not Hypothalamic–Adrenal–Pituitary Axis Activity or Hyperthermia. J. Neurosci. 28, 893–903 (2008). [DOI] [PMC free article] [PubMed] [Google Scholar]
6.Tanaka M., Yamada S. & Watanabe Y. The Role of Neuropeptide Y in the Nucleus Accumbens. Int. J. Mol. Sci. 22, 7287 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
7.Gao J., Ghibaudi L. & Hwa J. J. Selective activation of central NPY Y1 vs. Y5 receptor elicits hyperinsulinemia via distinct mechanisms. Am. J. Physiol.-Endocrinol. Metab. 287, E706–E711 (2004). [DOI] [PubMed] [Google Scholar]
8.Pedragosa-Badia X., Stichel J. & Beck-Sickinger A. G. Neuropeptide Y receptors: how to get subtype selectivity. Front. Endocrinol. 4, 5 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
9.Brown M. R. et al. Identification of a Drosophila brain-gut peptide related to the neuropeptide Y family. Peptides 20, 1035–1042 (1999). [DOI] [PubMed] [Google Scholar]
10.Garczynski S. F., Brown M. R., Shen P., Murray T. F. & Crim J. W. Characterization of a functional neuropeptide F receptor from Drosophila melanogaster. Peptides 23, 773–780 (2002). [DOI] [PubMed] [Google Scholar]
11.Wu Q. et al. Developmental control of foraging and social behavior by the Drosophila neuropeptide Y-like system. Neuron 39, 147–161 (2003). [DOI] [PubMed] [Google Scholar]
12.Wu Q., Zhao Z. & Shen P. Regulation of aversion to noxious food by Drosophila neuropeptide Y- and insulin-like systems. Nature neuroscience 8, 1350–1355 (2005). [DOI] [PubMed] [Google Scholar]
13.Kim D.-H., Shin M., Jung S.-H., Kim Y.-J. & Jones W. D. A fat-derived metabolite regulates a peptidergic feeding circuit in Drosophila. PLoS Biol. 15, e2000532 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
14.Beshel J., Dubnau J. & Zhong Y. A Leptin Analog Locally Produced in the Brain Acts via a Conserved Neural Circuit to Modulate Obesity-Linked Behaviors in Drosophila. Cell metabolism 25, 208–217 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
15.Krashes M. J. et al. A neural circuit mechanism integrating motivational state with memory expression in Drosophila. Cell 139, 416–427 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]
16.Chung B. Y. et al. Drosophila Neuropeptide F Signaling Independently Regulates Feeding and Sleep-Wake Behavior. Cell reports 19, 2441–2450 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
17.Gendron C. M. et al. Drosophila life span and physiology are modulated by sexual perception and reward. Science 343, 544–548 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
18.Shohat-Ophir G., Kaun K. R., Azanchi R., Mohammed H. & Heberlein U. Sexual Deprivation Increases Ethanol Intake in Drosophila. Science 335, 1351–1355 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
19.Shao L. et al. Dissection of the Drosophila neuropeptide F circuit using a high-throughput two-choice assay. Proceedings of the National Academy of Sciences of the United States of America 114, E8091–E8099 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Pu Y., Zhang Y., Zhang Y. & Shen P. Two Drosophila Neuropeptide Y-like Neurons Define a Reward Module for Transforming Appetitive Odor Representations to Motivation. Scientific reports 8, 11658 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Rohwedder A., Selcho M., Chassot B. & Thum A. S. Neuropeptide F neurons modulate sugar reward during associative olfactory learning of Drosophila larvae. Journal of Comparative Neurology 523, 2637–2664 (2015). [DOI] [PubMed] [Google Scholar]
22.Wang J. & Lehmann F.-O. Neuropeptide F-expressing neurons in Drosophila constitute centrifugal pathway to optic lobes. PLOS One 21, e0343221 (2026). [DOI] [PMC free article] [PubMed] [Google Scholar]
23.Dorkenwald S. et al. Neuronal wiring diagram of an adult brain. Nature 634, 124–138 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
24.Schlegel P. et al. Whole-brain annotation and multi-connectome cell typing of Drosophila. Nature 634, 139–152 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
25.Meissner G. W. et al. A searchable image resource of Drosophila GAL4 driver expression patterns with single neuron resolution. Elife 12, e80660 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
26.Meissner G. W. et al. A split-GAL4 driver line resource for Drosophila neuron types. eLife 13, RP98405 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
27.Feng K.-L. et al. Neuropeptide F inhibits dopamine neuron interference of long-term memory consolidation in Drosophila. iScience 24, 103506 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
28.Aso Y. & Rubin G. M. Dopaminergic neurons write and update memories with cell-type-specific rules. eLife 5, 156 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
29.Klapoetke N. C. et al. Independent Optical Excitation of Distinct Neural Populations. Nat. methods 11, 338–346 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
30.Bushey D. et al. RubyACRs Enable Red-Shifted Optogenetic Inhibition in Freely Behaving Drosophila. https://doi.org/10.7554/elife.108210.1 (2025) doi: 10.7554/elife.108210.1. [DOI] [Google Scholar]
31.Kim W. J., Jan L. Y. & Jan Y. N. A PDF/NPF neuropeptide signaling circuitry of male Drosophila melanogaster controls rival-induced prolonged mating. Neuron 80, 1190–1205 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
32.Liu W. et al. Neuropeptide F regulates courtship in Drosophila through a male-specific neuronal circuit. eLife 8, 31 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
33.Itskov P. M. et al. Automated monitoring and quantitative analysis of feeding behaviour in Drosophila. Nat. Commun. 5, 4560 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
34.Moreira J.-M. et al. optoPAD, a closed-loop optogenetics system to study the circuit basis of feeding behaviors. eLife 8, e43924 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
35.Collins N. J., Endres M. N., Sinakevitch I. T. & Shao L. Protocol for quantifying Drosophila feeding behavior using flyPAD and optoPAD. STAR Protoc. 7, 104711 (2026). [DOI] [PMC free article] [PubMed] [Google Scholar]
36.Aponte Y., Atasoy D. & Sternson S. M. AGRP neurons are sufficient to orchestrate feeding behavior rapidly and without training. Nat. Neurosci. 14, 351–355 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
37.Krashes M. J. et al. Rapid, reversible activation of AgRP neurons drives feeding behavior in mice. J. Clin. Investig. 121, 1424–1428 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
38.Tremblay R., Lee S. & Rudy B. GABAergic Interneurons in the Neocortex: From Cellular Properties to Circuits. Neuron 91, 260–292 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
39.Pelkey K. A. et al. Hippocampal GABAergic Inhibitory Interneurons. Physiol. Rev. 97, 1619–1747 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
40.Sawchenko P. E. et al. Colocalization of neuropeptide Y immunoreactivity in brainstem catecholaminergic neurons that project to the paraventricular nucleus of the hypothalamus. J. Comp. Neurol. 241, 138–153 (1985). [DOI] [PubMed] [Google Scholar]
41.Heilig M. The NPY system in stress, anxiety and depression. Neuropeptides 38, 213–224 (2004). [DOI] [PubMed] [Google Scholar]
42.Toll L. et al. Nociceptin/Orphanin FQ Receptor Structure, Signaling, Ligands, Functions, and Interactions with Opioid Systems. Pharmacol. Rev. 68, 419–457 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
43.Hardaway J. A. et al. Central Amygdala Prepronociceptin-Expressing Neurons Mediate Palatable Food Consumption and Reward. Neuron 102, 1037–1052.e7 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
44.Parker K. E. et al. A Paranigral VTA Nociceptin Circuit that Constrains Motivation for Reward. Cell 178, 653–671.e19 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
45.Rodriguez-Romaguera J. et al. Prepronociceptin-Expressing Neurons in the Extended Amygdala Encode and Promote Rapid Arousal Responses to Motivationally Salient Stimuli. Cell Rep. 33, 108362–108362 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
46.Strother J. A. et al. The Emergence of Directional Selectivity in the Visual Motion Pathway of Drosophila. Neuron 94, 168–182.e10 (2017). [DOI] [PubMed] [Google Scholar]
47.Pfeiffer B. D. et al. Refinement of Tools for Targeted Gene Expression in Drosophila. Genetics 186, 735–755 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]
48.Walker S. R., Peña-Garcia M. & Devineni A. V. Connectomic analysis of taste circuits in Drosophila. Sci. Rep. 15, 5278 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
49.Murakami K. et al. translin Is Required for Metabolic Regulation of Sleep. Curr. Biol. 26, 972–980 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplement 1

media-1.pdf^{(2.4MB, pdf)}

Data Availability Statement

Requests for further information and resources should be directed to and will be fulfilled by the lead contact, Lisha Shao (shaol@udel.edu).

Fly lines generated in this study are available upon request from the lead contact.

The code used for connectome analysis is described in Materials and Methods, and available at github.

Raw data for imaging and behavioral experiments are available upon request from the lead contact.

Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.

[R1] 1.Nässel D. R. & Zandawala M. Recent advances in neuropeptide signaling in Drosophila, from genes to physiology and behavior. Progress in Neurobiology https://doi.org/10.1016/j.pneurobio.2019.02.003 (2019) doi: 10.1016/j.pneurobio.2019.02.003. [DOI] [PubMed] [Google Scholar]

[R2] 2.Nässel D. R. & Zandawala M. Endocrine cybernetics: neuropeptides as molecular switches in behavioural decisions. Open Biol. 12, 220174 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R3] 3.Nässel D. R., Pauls D. & Huetteroth W. Neuropeptides in modulation of Drosophila behavior: how to get a grip on their pleiotropic actions. Curr. Opin. Insect Sci. 36, 1–8 (2019). [DOI] [PubMed] [Google Scholar]

[R4] 4.Cowley M. A. et al. Integration of NPY, AGRP, and Melanocortin Signals in the Hypothalamic Paraventricular Nucleus. Neuron 24, 155–163 (1999). [DOI] [PubMed] [Google Scholar]

[R5] 5.Sajdyk T. J. et al. Neuropeptide Y in the Amygdala Induces Long-Term Resilience to Stress-Induced Reductions in Social Responses But Not Hypothalamic–Adrenal–Pituitary Axis Activity or Hyperthermia. J. Neurosci. 28, 893–903 (2008). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R6] 6.Tanaka M., Yamada S. & Watanabe Y. The Role of Neuropeptide Y in the Nucleus Accumbens. Int. J. Mol. Sci. 22, 7287 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R7] 7.Gao J., Ghibaudi L. & Hwa J. J. Selective activation of central NPY Y1 vs. Y5 receptor elicits hyperinsulinemia via distinct mechanisms. Am. J. Physiol.-Endocrinol. Metab. 287, E706–E711 (2004). [DOI] [PubMed] [Google Scholar]

[R8] 8.Pedragosa-Badia X., Stichel J. & Beck-Sickinger A. G. Neuropeptide Y receptors: how to get subtype selectivity. Front. Endocrinol. 4, 5 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R9] 9.Brown M. R. et al. Identification of a Drosophila brain-gut peptide related to the neuropeptide Y family. Peptides 20, 1035–1042 (1999). [DOI] [PubMed] [Google Scholar]

[R10] 10.Garczynski S. F., Brown M. R., Shen P., Murray T. F. & Crim J. W. Characterization of a functional neuropeptide F receptor from Drosophila melanogaster. Peptides 23, 773–780 (2002). [DOI] [PubMed] [Google Scholar]

[R11] 11.Wu Q. et al. Developmental control of foraging and social behavior by the Drosophila neuropeptide Y-like system. Neuron 39, 147–161 (2003). [DOI] [PubMed] [Google Scholar]

[R12] 12.Wu Q., Zhao Z. & Shen P. Regulation of aversion to noxious food by Drosophila neuropeptide Y- and insulin-like systems. Nature neuroscience 8, 1350–1355 (2005). [DOI] [PubMed] [Google Scholar]

[R13] 13.Kim D.-H., Shin M., Jung S.-H., Kim Y.-J. & Jones W. D. A fat-derived metabolite regulates a peptidergic feeding circuit in Drosophila. PLoS Biol. 15, e2000532 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R14] 14.Beshel J., Dubnau J. & Zhong Y. A Leptin Analog Locally Produced in the Brain Acts via a Conserved Neural Circuit to Modulate Obesity-Linked Behaviors in Drosophila. Cell metabolism 25, 208–217 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R15] 15.Krashes M. J. et al. A neural circuit mechanism integrating motivational state with memory expression in Drosophila. Cell 139, 416–427 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R16] 16.Chung B. Y. et al. Drosophila Neuropeptide F Signaling Independently Regulates Feeding and Sleep-Wake Behavior. Cell reports 19, 2441–2450 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R17] 17.Gendron C. M. et al. Drosophila life span and physiology are modulated by sexual perception and reward. Science 343, 544–548 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R18] 18.Shohat-Ophir G., Kaun K. R., Azanchi R., Mohammed H. & Heberlein U. Sexual Deprivation Increases Ethanol Intake in Drosophila. Science 335, 1351–1355 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R19] 19.Shao L. et al. Dissection of the Drosophila neuropeptide F circuit using a high-throughput two-choice assay. Proceedings of the National Academy of Sciences of the United States of America 114, E8091–E8099 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R20] 20.Pu Y., Zhang Y., Zhang Y. & Shen P. Two Drosophila Neuropeptide Y-like Neurons Define a Reward Module for Transforming Appetitive Odor Representations to Motivation. Scientific reports 8, 11658 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R21] 21.Rohwedder A., Selcho M., Chassot B. & Thum A. S. Neuropeptide F neurons modulate sugar reward during associative olfactory learning of Drosophila larvae. Journal of Comparative Neurology 523, 2637–2664 (2015). [DOI] [PubMed] [Google Scholar]

[R22] 22.Wang J. & Lehmann F.-O. Neuropeptide F-expressing neurons in Drosophila constitute centrifugal pathway to optic lobes. PLOS One 21, e0343221 (2026). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R23] 23.Dorkenwald S. et al. Neuronal wiring diagram of an adult brain. Nature 634, 124–138 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R24] 24.Schlegel P. et al. Whole-brain annotation and multi-connectome cell typing of Drosophila. Nature 634, 139–152 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R25] 25.Meissner G. W. et al. A searchable image resource of Drosophila GAL4 driver expression patterns with single neuron resolution. Elife 12, e80660 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R26] 26.Meissner G. W. et al. A split-GAL4 driver line resource for Drosophila neuron types. eLife 13, RP98405 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R27] 27.Feng K.-L. et al. Neuropeptide F inhibits dopamine neuron interference of long-term memory consolidation in Drosophila. iScience 24, 103506 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R28] 28.Aso Y. & Rubin G. M. Dopaminergic neurons write and update memories with cell-type-specific rules. eLife 5, 156 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R29] 29.Klapoetke N. C. et al. Independent Optical Excitation of Distinct Neural Populations. Nat. methods 11, 338–346 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R30] 30.Bushey D. et al. RubyACRs Enable Red-Shifted Optogenetic Inhibition in Freely Behaving Drosophila. https://doi.org/10.7554/elife.108210.1 (2025) doi: 10.7554/elife.108210.1. [DOI] [Google Scholar]

[R31] 31.Kim W. J., Jan L. Y. & Jan Y. N. A PDF/NPF neuropeptide signaling circuitry of male Drosophila melanogaster controls rival-induced prolonged mating. Neuron 80, 1190–1205 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R32] 32.Liu W. et al. Neuropeptide F regulates courtship in Drosophila through a male-specific neuronal circuit. eLife 8, 31 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R33] 33.Itskov P. M. et al. Automated monitoring and quantitative analysis of feeding behaviour in Drosophila. Nat. Commun. 5, 4560 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R34] 34.Moreira J.-M. et al. optoPAD, a closed-loop optogenetics system to study the circuit basis of feeding behaviors. eLife 8, e43924 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R35] 35.Collins N. J., Endres M. N., Sinakevitch I. T. & Shao L. Protocol for quantifying Drosophila feeding behavior using flyPAD and optoPAD. STAR Protoc. 7, 104711 (2026). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R36] 36.Aponte Y., Atasoy D. & Sternson S. M. AGRP neurons are sufficient to orchestrate feeding behavior rapidly and without training. Nat. Neurosci. 14, 351–355 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R37] 37.Krashes M. J. et al. Rapid, reversible activation of AgRP neurons drives feeding behavior in mice. J. Clin. Investig. 121, 1424–1428 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R38] 38.Tremblay R., Lee S. & Rudy B. GABAergic Interneurons in the Neocortex: From Cellular Properties to Circuits. Neuron 91, 260–292 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R39] 39.Pelkey K. A. et al. Hippocampal GABAergic Inhibitory Interneurons. Physiol. Rev. 97, 1619–1747 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R40] 40.Sawchenko P. E. et al. Colocalization of neuropeptide Y immunoreactivity in brainstem catecholaminergic neurons that project to the paraventricular nucleus of the hypothalamus. J. Comp. Neurol. 241, 138–153 (1985). [DOI] [PubMed] [Google Scholar]

[R41] 41.Heilig M. The NPY system in stress, anxiety and depression. Neuropeptides 38, 213–224 (2004). [DOI] [PubMed] [Google Scholar]

[R42] 42.Toll L. et al. Nociceptin/Orphanin FQ Receptor Structure, Signaling, Ligands, Functions, and Interactions with Opioid Systems. Pharmacol. Rev. 68, 419–457 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R43] 43.Hardaway J. A. et al. Central Amygdala Prepronociceptin-Expressing Neurons Mediate Palatable Food Consumption and Reward. Neuron 102, 1037–1052.e7 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R44] 44.Parker K. E. et al. A Paranigral VTA Nociceptin Circuit that Constrains Motivation for Reward. Cell 178, 653–671.e19 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R45] 45.Rodriguez-Romaguera J. et al. Prepronociceptin-Expressing Neurons in the Extended Amygdala Encode and Promote Rapid Arousal Responses to Motivationally Salient Stimuli. Cell Rep. 33, 108362–108362 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R46] 46.Strother J. A. et al. The Emergence of Directional Selectivity in the Visual Motion Pathway of Drosophila. Neuron 94, 168–182.e10 (2017). [DOI] [PubMed] [Google Scholar]

[R47] 47.Pfeiffer B. D. et al. Refinement of Tools for Targeted Gene Expression in Drosophila. Genetics 186, 735–755 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R48] 48.Walker S. R., Peña-Garcia M. & Devineni A. V. Connectomic analysis of taste circuits in Drosophila. Sci. Rep. 15, 5278 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R49] 49.Murakami K. et al. translin Is Required for Metabolic Regulation of Sleep. Curr. Biol. 26, 972–980 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

This is a preprint.

Cellular source and circuit context organize functional specificity in the Drosophila NPF system

Madison N Endres

Tarandeep Singh Dadyala

Kevin W Christie

Irina T Sinakevitch

Lisha Shao

Abstract

Introduction

Results

NPF cell types possess distinct connectomic networks

Figure 1. NPF cell types have distinct connectivity in the connectome.

Distinct NPF cell types differentially recruit PAM dopaminergic neurons

Figure 2. Cell-type-specific and population-level activity manipulations reveal functional specialization within the NPF system.

Neural activity manipulations reveal cell-type-specific behavioral functions

Cell-type-specific NPF knockdown reveals distinct peptide-dependent functions

Figure 3. Cell-type-specific NPF knockdown reveals functions not apparent after global NPF loss.

NPF peptide contributes conditionally to phenotypes induced by pan-NPF activation

Figure 4.

Discussion

Cellular source creates specificity within a single-receptor peptide system

Population-level perturbation can obscure cell-type-specific functions

Neuronal activity is not equivalent to peptide action

Cell-type specialization reframes an apparent NPF reward paradox

A single-receptor architecture with a mammalian counterpart

Limitations and future directions

Materials and Methods

Fly stocks and culture

Connectome analysis

Behavioral assays

Positional preference

Associative learning

Mating

flyPad/optoPad feeding

Metabolites measurement

Immunostaining and confocal imaging

Ex vivo calcium imaging

Statistical analysis

Supplementary Material

Acknowledgements

Data Availability

Materials availability

Data and code availability

Reference

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases