Pharmacological and functional similarities of the human neuropeptide Y system in C. elegans challenges phylogenetic views on the FLP/NPR system

Background The neuropeptide Y system affects various processes, among others food intake, and is frequently discussed in the context of targeting obesity. Studies in model organisms are indispensable to enable molecular studies in a physiological context. Although the NPY system is evolutionarily conserved in all bilaterians, in the widely used model Caenorhabditis elegans there is controversy on the existence of NPY orthologous molecules. While the FMRFamide-like peptide (FLP)/Neuropeptide receptor-Resemblance (NPR) system in the nematode was initially suggested to be orthologous to the mammalian NPY system, later global phylogenetic studies indicate that FLP/NPR is protostome-specific. Methods We performed a comprehensive pharmacological study of the FLP/NPR system in transfected cells in vitro, and tested for functional substitution in C. elegans knockout strains. Further, we phenotypically compared different flp loss-of-function strains. Differences between groups were compared by ANOVA and post-hoc testing (Dunnett, Bonferroni). Results Our pharmacological analysis of the FLP/NPR system including formerly functionally uncharacterized NPY-like peptides from C. elegans demonstrates that G protein-coupling and ligand requirements for receptor activation are similar to the human NPY system. In vitro and in vivo analyses show cross-reactivity of NPY with the FLP/NPR system manifesting in the ability of the human GPCRs to functionally substitute FLP/NPR signaling in vivo. The high pharmacological/functional similarities enabled us to identify C. elegans FLP-14 as a key molecule in avoidance behavior. Conclusions Our data demonstrate the pharmacological and functional similarities of human NPY and C. elegans NPR systems. This adds a novel perspective to current phylogenetic reconstructions of the neuropeptide Y system. NPY and NPR receptors are pharmacologically so similar that the human receptors can functionally compensate for the C. elegans ones, suggesting orthologous relationships. This is also underlined by the presence of NPY-like peptides and parallels in peptide requirements for receptor activation. Further, the results presented here highlight the potential of this knowledge for physiological as well as molecular studies on neuropeptide GPCRs such as the NPY system in the future.


Background
The neuropeptide Y (NPY) family is an intensively studied system due to the essential roles of its members in regulating food intake in humans [1]. Consequently, its potential as target for modulating food consumption and thus, obesity, is intensely investigated [1]. Besides this capacity, the NPY system also has multiple other functions, for instance in the control of mood and anxiety or ethanol intake [2,3]. The hallmark features of NPY and the related peptides PYY and PP are a C-terminal arginine-phenylalanine/tyrosine sequence and an amidated carboxy-terminus (RxRF/ Yamide), which are essential for activation of their four cognate G protein-coupled receptors (GPCRs) in most mammals [4]. Due to the many facets of NPY, numerous insights are based on findings in model organisms (summarized in [5]). Rodent studies established a physiological function for single receptor subtypes. However, the specific roles of the receptors, their distinct cellular effectors, desensitization profiles, and the effects of specific mutations are difficult to address in rodent model systems. In this regard, a more basic model organism that enables higher throughput analyses and ideally offers spatiotemporal control would be desirable.
Such studies are possible in principle in simple model organisms, as several neuropeptidergic signaling systems are conserved in invertebrates. Indeed, a global phylogenetic analysis identified the neuropeptide F/Y (NPF/Y) system as one of nearly 30 ancient neuropeptidergic systems that can be traced back to the common ancestor of protostomes and deuterostomes (urbilaterian) [6][7][8]. For example, a clear NPF ortholog is found in the protostome Drosophila melanogaster [7,[9][10][11]. Likewise, in Caenorhabditis elegans, the receptor NPR-1 (NeuroPeptide receptor Resemblance-1) was initially assigned as NPY receptor ortholog. C. elegans is a valuable model organism due to its translucence and vast genetic toolbox. Importantly for studies on GPCRs, all major signaling pathways are present (G s , G q/11 , G i/o , arrestins). The early classification of C. elegans NPRs as NPY receptor homologs was based on sequence similarities and their role in feeding behavior [12], and was recently supported by a study using vertebrate ancestral chromosome reconstruction [13]. However, in global phylogenetic analyses, NPRs and their ligands appear to form a protostome-specific clade that has diverged from a common bilaterian ancestor and is paralogous to the NPF/Y family, thus challenging a potential use of the nematode as a model system in this context [7,8].
This notion is supported by genome analyses of C. elegans, which show a remarkably expanded repertoire of genes encoding potential NPR receptors comprising 41 members [14]. Likewise, the possible cognate ligands, FMRFamide-like peptides (FLPs), cover 31 genes (reviewed in [14]), each encoding a set of peptides yielding in total more than 70 FLPs. These can be subdivided into FMRFamide (ΩΨRFa; with Ω being an aromatic amino acid and Ψ being a hydrophobic amino acid) and short neuropeptides F (sNPF) harboring a C-terminal ΨΨRFa consensus. In contrast, 'classical' NPF/NPY sequences found in other bilateria are longer and carry a C-terminal RxRF/Yamide. Within the predicted repertoire of FLPs in C. elegans, FLP-27, FLP-33, FLP-34-1 and FLP-34-2 share the greatest similarity with vertebrate NPY and display this C-terminal RxRF/Ya signature. However, only FLP-27 and 33 have been biochemically isolated [15,16] and no function has yet been assigned to these peptides. These observations leave important open questions such as whether NPF/Y orthologs exist in nematodes or have been lost in this phylum, and if so, by which mechanisms their essential physiological functions are compensated.
Here, we demonstrate that the nematode NPR/FLP system is pharmacologically and functionally highly similar to the human NPY system, and the same set of receptors is activated by FLPs of FMRF, sNPF and (long)NPF-type, adding a new perspective to current phylogenetic reconstructions. This resemblance is so profound that human NPY GPCRs are even able to phenotypically rescue NPR function in vivo in C. elegans. The functional homology of both neuropeptide systems was subsequently utilized to identify FLP-14 as a context-specific driver for avoidance behavior in C. elegans.

Materials
All standard chemicals were purchased from Sigma Aldrich or Carl-Roth GmbH unless stated otherwise. Cell culture materials were obtained from Lonza and enzymes from ThermoFisher Scientific.

Cell culture
In all in vitro assays the commercially available cell line HEK293 (Homo sapiens, female, embryonic kidney, DSMZ ACC 305) was used. Cell line identity was confirmed by short tandem repeat profiling at eight different loci (performed by the Leibnitz Institut DSMZ, Braunschweig), and cells were negative for mycoplasma contamination in routine testings. Cells were maintained as a monolayer at 37°C and 5% CO 2 under humidified atmosphere in Dulbecco's modified Eagle's medium (DMEM) with Ham's F12 (1:1; v/v) supplemented with 15% (v/v) heat inactivated fetal calf serum (FCS).

C. elegans strains
C. elegans were maintained at 22°C using standard conditions [17]. Wild-type worms were C. elegans variety Bristol, N2. Strains used in this study are listed in Additional file 1: Table S1. Strains not generated in this study were obtained from the Caenorhabditis Genetics Center (CGC) funded by NIH Office of Research Infrastructure Programs (P40 OD010440).

Generation of plasmids and transgenes Neuropeptide GPCR constructs for in vitro analysis
Expression vectors encoding human NPY and RFamide receptors were kind gifts from A. G. Beck-Sickinger (Y 1 R, Y 2 R, Y 4 R, Y 5 R, NPFF 1 R, NPFF 2 R, QRFPR and PrRPR). These receptor constructs are preceded by a 5′ Kozak sequence (GCCACC) and at 3′ fused to an enhanced yellow fluorescent protein (eYFP) via an ADPPVV linker (GCG GATCCACCGGTCGTG, containing BamHI and AgeI restriction sites). C. elegans receptor sequences (npr-3, npr-6) were amplified from C. elegans cDNA. For cDNA generation, a mixed population of N2 hermaphrodites was harvested, washed in M9 buffer and incubated in TRIzol reagent (ThermoFisher Scientific) for total RNA isolation according to manufacturer's protocol. Reverse transcriptase and oligo-dT primers were used to generate cDNA. cDNA sequences of npr-1, npr-4b, npr-5b, and npr-11 were purchased (GenScript). The coding sequences were fused to a 5′ Kozak sequence, and 3′ DPPVV linker in analogy to the expression vectors for human receptors using the NPR(x)_ Mlu_f and NPR(x)_Linker_r primers. In a second step, these sequences were fused to an eYFP coding sequence (amplified from the Y 2 R-eYFP_pV2 parent construct with the primers Linker-YFP-f and YFP-XbaI-NheI-r) using the PCR overlap extension technique [19]. The resulting DNA fragments were ligated into the pVitro2-hygro-mcs vector (InvivoGen) using a (5′) MluI and (3′) XbaI (NheI in case of npr-1) restriction sites. For primer sequences see Additional file 1: Table S2. Chimeric Gα Δ6qi4myr [20] was a kind gift from E. Kostenis.

Neuropeptide GPCR constructs for transgenesis
All constructs contained the cDNA of a human neuropeptide receptor C-terminally fused to a GFP downstream of a 2 kb npr-1 promoter in vector pPD95.79. The basis for these constructs was npr-1p::npr-1::gfp in pPD95.79 (kind gift of L. Ma). From this vector, the backbone was amplified using primers pPD95.79_f/ pPD95.79_r. In a second step, the npr-1 promoter was amplified from npr-1p::npr-1::gfp using primers npr-1p_ f/ npr-1p_r and ligated back into the obtained backbone via an XmaI restriction site yielding plasmid pSP131.

Inositol phosphate (IP) accumulation assay
For measuring IP accumulation, HEK293 cells grown to 70-80% confluency in 6-well plates were transiently cotransfected using Metafectene Pro (2.5 μl/μg DNA; Biontex) with plasmids containing receptor and/or mock (empty pcDNA3.1) (for endogenous G q -coupling) using 4 μg total DNA in a ratio of 3:1 following the manufacturer's instructions. For testing Gα 16 -and Gα Δ6qi4myr -coupling, plasmids encoding the corresponding Gα protein were co-transfected instead of mock. Gα 16 is naturally promiscuous and stimulates the phospholipase C (PLC) pathway leading to IP production, while the exchange of the C-terminal four amino acids of Gα q for the corresponding residues of Gα i1 in the chimeric Gα Δ6qi4myr confers the ability to couple to G i/o protein-preferring receptors but still stimulate PLC, thus specifically redirecting downstream cellular signaling [21]. 16 h post transfection, the cells were seeded into white 384-well plates at a density of 20,000 cells/well. Media was removed 24 h later and the cells were incubated with the indicated concentration of peptides for 1 h in HBSS + 20 mM LiCl. IP accumulation was measured with the HTRF-based IP-One Gq kit (Cisbio) on a microplate reader (Tecan Spark).

cAMP reporter gene assay
Activation of G s and G i/o was assessed with a cAMP reporter gene assay. HEK293 cells grown to 70-80% confluency in 6well plates were transiently co-transfected with vectors encoding the receptor and the cAMP reporter gene plasmid pGL4.29 [luc2P/CRE/Hygro] (Promega; 4 μg total, 1:1 ratio) using Metafectene Pro (3 μl/μg DNA; Biontex) following the manufacturer's instructions. 16 h post transfection, the cells were seeded into white 384-well plates at a density of 20,000 cells/well. On the next day, the medium was removed and the cells were stimulated with 20 μl of peptide solution (or forskolin as positive control) in serum-free DMEM for 4 h. For measuring G i/o activity, the peptide solutions additionally contained 1 μM forskolin to elevate cellular cAMP levels. After incubation, the luciferase substrate OneGlo in lysis buffer (Promega) was added and the luminescence was measured in a microplate reader Tecan Spark (Tecan). Data were analyzed with GraphPad Prism 5.03 and are shown as fold of basal or fold of forskolin in case of Gα i/o activity. All data are displayed as mean ± SEM of at least three independent experiments performed in triplicate.
Expression of the human neuropeptide receptors in C. elegans was determined by confocal fluorescent microscopy. Nematodes were anesthetized with 125 mM sodium azide and placed on 5% agar pads immediately prior to imaging using a Leica TCS SP8 confocal microscope (HyD photo detector, Leica Microsystems).

Bordering behavior assay
Bordering behavior was determined as described by de Bono and Bargman [12]. In brief, 120 young adults were placed on NGM plates with a thick OP50 lawn (approximately 2.5 cm in diameter). Plates were kept at 22°C for 2 h until scoring. Animals within a 2 mm range of the border of the lawn were counted as bordering (Fig. 3).

MeSa (methyl salicylate) avoidance assay
The MeSa avoidance assay was performed as described previously [24] with some modifications (Fig. 3). Nematode growth medium (NGM) plates (9 cm) were divided into four equal quadrants. Young adults were washed three times with M9 and circa 50 to 120 animals were transferred in 40 μl M9 to the middle of the assay plate. After placing the worms on the plate, either a drop of 2 μl ethanol (EOH) as control or of 2 μl methyl salicylate (MeSa) were positioned on two opposite quadrants equidistant from each other (3.5 cm from the center) (adapted from [25]). To paralyze nearby animals, 2 μl of 0.5 M sodium azide were dripped on each quadrant. Plates were sealed with parafilm and incubated at 22°C until all worms were paralyzed. For measuring avoidance, the animals on each side were counted and the avoidance index was calculated by subtracting the number of worms on the ethanol sides from the number of worms in the methyl salicylate sides and divided by the total worm count.

Statistical analysis and matrix generation
For in vitro activation, data is presented as mean ± SEM of at least three independent experiments conducted in technical triplicate. For a given signal pathway, peptides were considered to activate a particular receptor (and thus be considered in the interaction matrix) if the signal for the highest peptide concentration tested was higher than the threshold, and statistically different from the buffer control as analyzed by a one-way ANOVA with Dunnett's post-hoc test and p values < 0.05. We defined the threshold as 2-fold over the buffer control for cAMP (G s ) and IP (G q ) production, or 20% inhibition of forskolin-induced cAMP levels at this receptor, respectively. For the C. elegans NPRs, we considered the strongest response elicited by any of the peptides downstream of a particular receptor as the receptor's maximal signal. This may or may not be the biological maximum that can be reached by the 'true' endogenous agonist. For the human receptors, the response to the established endogenous agonist reflected the maximal signal. Color-coding of the interaction matrix for the peptide ligands is based on the gradual activities at sub-micromolar (0.1 μM) or micromolar (10 μM) peptide concentrations as a coarse measure of peptide potency (i-v). The two lightest shades (i, ii) reflect a just-above-threshold and~50% of maximal receptor signal in response to 10 μM peptide, respectively, but no response at 0.1 μM concentration (corresponding to an estimated EC 50 of 100-10 μM assuming normal curve steepness and full agonism). The midgrade shade (iii) reflects an above-threshold signal when stimulated with 0.1 μM of peptide but still submaximal response when stimulated with 10 μM of peptide (estimated EC 50~1 μM). The second darkest shade (iv) indicates partial activation by 0.1 μM and full activation by 10 μM of peptide (estimated EC 50~0 .1 μM), while dark green (v) illustrates maximal receptor activation already in response to 0.1 μM of agonist (estimated EC 50 < 0.1 μM).
Of note, there were a few peptide/receptor pairs that displayed equal responses at 0.1 and 10 μM ligand stimulation, but the response was significantly lower than the maximum signal that can be elicited downstream of the given receptor (by another peptide), indicative of high affinity partial agonism. These instances are indicated in Fig. 2 and the supplemental data tables with an asterisk (*). Importantly, our screening and analysis cannot inform about full or partial agonism for ligands with lower potency.
Similarly, for the G protein-coupling preferences, we considered a pathway to be activated if at least one peptide elevated the respective second messenger significantly above control levels using a two-way ANOVA followed by Bonferroni's post-hoc test (variable 1: pathway, variable 2: stimulation). Color-coding of the matrix shows the pathway with highest signal (fold of basal) in dark blue, and secondary pathways in light blue.
In vivo data are reported as mean ± SD. Statistical analysis was performed using GraphPad Prism version 7.0. Data were analyzed using a one-way ANOVA with a Bonferroni post-hoc test or unpaired student's t-test and p values < 0.05 were considered statistically significant. All details are given in the respective Figure legend.

Selection of C. elegans RFamide peptides and potential cognate GPCRs
Global phylogenetic reconstructions indicate that the C. elegans NPR/FLP system is not directly orthologous to the human NPY system despite the latter being conserved in all bilaterians [6][7][8], and the NPR and NPY receptors display a considerable degree of homology [12]. To clarify potential similarities between these systems, we first set out to characterize the pharmacological properties of the NPR system in the roundworm and to analyze the pharmacological complementarity with the human neuropeptide NPY and related RFamide systems in vitro. For this purpose, we chose the following eight well-established FMRF-and sNPF-type FLPs, as they have been reported to activate NPR receptors [14]: FLP-1-2, FLP-3-4, FLP-4-2, FLP-5-1, FLP-14, FLP-15-2, FLP-18-5, and FLP-21 (Fig. 1). The respective isoforms were chosen to represent the average length and charge composition. Further, we included the four so far uncharacterized FLPs FLP-27, FLP-33, FLP-34-1, and FLP-34-2, which appear to be more NPY-like. All peptides were synthesized by Fmoc/tert-butyl solid-phase chemistry and purified to > 95% to ensure high assay confidence (Additional file 1: Table S3).
For comparing these C. elegans and the human neuropeptide systems in vitro, we selected the following six well-characterized NPRs: NPR-1, NPR-3, NPR-4b, NPR-5b, NPR-6, and NPR-11. These receptors were chosen as they were identified as most closely related to the human NPY receptors in a recent phylogenetic analysis [27]. Moreover, we considered NPR-4 to be representative also for the GPCR NPR-10 (which was not tested), NPR-5 for NPR-13, and NPR-11 for NPR-12, since these receptors show a close phylogenetic relationship and likely arose from recent gene duplication as suggested by the same study [27]. For NPR-4 and NPR-5, which have multiple isoforms originating from alternative splicing, we chose the longest variants (NPR-4b, NPR-5b).
Human and C. elegans neuropeptide receptors are pharmacologically similar All receptors were expressed heterologously in HEK293 cells and similar expression levels of human and C. elegans receptors as well as their presence in the plasma membrane was confirmed (Fig. 2a). One exception was npr-3, with weaker expression and only a fraction of the receptor being exported to the cell membrane.
We then characterized G protein-coupling abilities as these give first insights into their similarity to the human receptors, and orthologs are expected to have conserved G-protein preferences [28]. Functional studies of C. elegans receptors in human cell lines are possible due to the high conservation of G proteins across the animal kingdom [28] including complete identity in the α5 helix of the Gα subunit that mainly determines G-protein specificity [28][29][30]. G protein-coupling was identified using second messenger assays. Production of cAMP in the presence and absence of the adenylyl cyclase activator forskolin was employed to detect G i/o and G s activity, respectively, and accumulation of intracellular inositol phosphate (IP) concentrations served as indicator of G q/11 protein-coupling. As we did not observe any basal activity in the absence of agonistic peptides (Additional file 1: Table S8), we determined coupling preferences in response to the neuropeptides FLP-5-1, FLP-15-2, FLP-18-5, and FLP-21. Several studies individually tested some peptide-receptor combinations or their biological function (reviewed in [14]), so these peptides were chosen as 'prototypic' agonists to characterize or confirm the G protein-coupling of the selected receptors.
Activation of most C. elegans NPRs led to the inhibition of intracellular cAMP production, indicating coupling to G i/o proteins (Additional file 1: Table S4). Only NPR-5b activation drove IP accumulation and a slight cAMP increase, suggesting coupling to G q and G s (Additional file 1: Tables S4 and S5). To establish a universal screening platform based on a robust signal accumulation, we co-transfected the npr constructs with plasmids encoding Gα 16 or the chimeric Gα Δ6qi4myr protein [20], which both mediate an increase in IP levels. Indeed, Gα Δ6qi4myr efficiently yielded a signal for NPR-3, NPR-4b, NPR-6, and NPR-11 (Additional file 1: Table S5), while co-transfection of Gα 16 exclusively rendered signaling of NPR-4b, albeit with lower signal windows. Interestingly, NPR-1 did not accept the Gα Δ6qi4myr chimera, and activation was only monitored downstream of the endogenous Gα i/o by decrease of cellular cAMP levels. G-protein preferences of all NPRs are summarized in Fig. 2b.
Next, we investigated activation of NPRs towards the entire panel of FLP ligands. We initially tested two peptide concentrations, 100 nM and 10 μM, which is in the range of reported EC 50 values within the C. elegans NPR/FLP system (reviewed in [31,32]) and likely at the upper limit of physiologically relevant concentrations (albeit local concentration in the synaptic cleft may be even up to mM range according to [33]), respectively. We found a high level of FLP promiscuity and consequently, redundancy in C. elegans receptor activation profiles (Fig. 2c). FLP-15-2 and FLP-21 activated all NPRs tested, while the remaining peptides stimulated at least three receptors. The more NPY-like peptides FLP-27, FLP-33, FLP-34-1, and FLP-34-2 appeared somewhat more selective, and mainly activated NPR-11 and NPR-1. Conversely, all NPRs were activated by at least five FLPs, albeit with different activities. One exception was NPR-3, which, in line with its weaker expression, displayed smaller signal windows and significant activation was only detected in response to FLP-15 and FLP-21.
For selected FLP-NPR pairs, we recorded full concentration-response curves to more accurately determine EC 50 values (Additional file 1: Figure S1 and Table S11). As expected, all ligand-receptor interactions color-coded in dark green in Fig. 2 display an EC 50 value below 100 nM. One of the most potent interactions was FLP-21 activating NPR-1 with a potency of 1 nM, in agreement with a previous study [31]. Interestingly, FLP-14 was a potent, but partial agonist at the NPR-1 (EC 50 24 nM, E max 53% of FLP-21 induced response; Additional file 1: Figure S1). The RxRF/ Fig. 2 Basic pharmacological and cross-species reactivity studies of C. elegans and human neuropeptides and receptors in vitro. a Live-cell fluorescence microscopy of C. elegans NPR, human NPY and RFamide receptor::eYFP fusion proteins (yellow) shows similar receptor expression in transfected HEK293 cells and export to the plasma membrane. Images are representative of three independent experiments, and were acquired and processed identically. Nuclei were stained with Hoechst33342 (blue), scale bar = 10 μm. b C. elegans receptors recognize the human G protein repertoire and show distinct coupling preferences. Coupling specificity was deduced from second messenger production (cAMP or inositol phosphate (IP)) in response to FLP-5-1, FLP-15-2, FLP-18-5, and FLP-21, and qualitative data is depicted as color gradient (original numerical data are shown in Additional file 1: Tables S4 and S5). Coupling of NPR-5b to Gα 16 and Gα Δ6qi4myr via inositol phosphate accumulation is not readily accessible due to endogenous Gα q -coupling and was thus not determined (shaded). c Cross-species activation profiles of selected human and C. elegans receptors with different neuropeptides. Activation profiles were based on data from second messenger assays according to the endogenous G protein-coupling (original data can be found in Additional file 1: Tables S6-S9). Shown is a color-coded interaction matrix based on a two-concentration (0.1 μM; 10 μM) peptide screen as a coarse measure of peptide potency. The two lightest shades of green indicate partial receptor activation after stimulation with 10 μM peptide but no response to 0.1 μM of peptide (estimated EC 50 > 10 μM assuming normal curve steepness and full agonism). Mid-grade colors display submaximal activation at 0.1 μM peptide and up to full activation at 10 μM (estimated EC 50 0.1 μM), while the darkest shade represents full activation in response to 0.1 μM of peptide (estimated EC 50 < 100 nM). Few peptide/receptor combinations elicited identical, but submaximal responses at 0.1 and 10 μM concentration, indicative of partial agonism, and are marked with an asterisk (*). For more details on matrix generation in B and C, see Methods Ya-containing peptides FLP-34-1 (carrying a C-terminal tyrosine) and FLP-34-2 also reached nanomolar potencies at the NPR-11 receptor (EC 50 FLP-34-1: 19 nM, FLP-34-2: 0.7 nM, Additional file 1: Figure S1, Table S11).
Next, we studied possible reactivity of C. elegans FLPs on human neuropeptide GPCRs and vice versa. We included NPY receptors, but also a group of related human RFamide receptors (PrRPR, QRFPR, and NPFFR). These display an equally high homology in pairwise alignments to NPRs and similar peptide ligands (Fig. 1), but have not yet been considered as potential orthologs since they were de-orphanized only after the seminal study on npr-1 [12]. The ability of ligands to activate receptors of the other species can support an orthologous relationship as this usually goes along with pharmacological/functional conservation. However, the timewise evolutionary distance between C. elegans and human is large, reducing chances of cross-species reactivity. For this reason, we investigated all human NPY and RFamide receptors, including Y 4 R (and its ligand PP) and NPFF 2 R, which have evolved more recently during early vertebrate evolution [13,34]. In this regard, we consider cross-species reactivity of a set of FLP-ligands (from local duplications during C. elegans evolution) to at least one receptor of a human receptor family (NPY or PrRPR or QRFPR or NPFFR) as functional similarity between these particular families, reflecting essential ligand requirements for receptor activation. Cross-species reactivity of an evolutionary 'younger' human peptide ligand to the ancient C. elegans NPR receptors might also occur, but is expected to be less likely. This is because a coevolution of peptide and receptor typically creates more refined (and selective) ligand-receptor interactions over the long evolutional timescale. For instance, ligand length or surface electrostatics vary, which are key factors determining receptor-(subtype)-selectivity of peptide ligands [35,36]. Thus, the 'original' binding pockets may be partially incompatible with or inaccessible for the resulting peptides.
Conversely, human neuropeptides also had activating abilities on C. elegans, albeit to a much more limited extent ( Fig. 2c and Additional file 1: Table S6). NPR-11 displayed the most distinct activation by human NPY and RFa peptides, and we confirmed robust agonistic activity of PYY (EC 50 24 nM, Emax 86% of FLP-21-induced response) and partial agonism for NPY (EC 50 256 nM, E max 49% of FLP-21-induced response) in full concentration-response curves (Additional file 1: Figure S1 and Table S11).
To check which residues are crucial for this cross-species activation, we applied peptides with the penultimate arginine exchanged for an alanine as this residue is essential for receptor-binding of human NPY and RFamide peptides [38,39]. R8A control peptides of the universal ligands FLP-15-2 and FLP-21 drastically lost activity on all receptors (Additional file 1: Table S10 and Figure S1). Similarly, human [R35A]-PYY was inactive in two-point concentration screens and full concentration-response curves on C. elegans receptors NPR-6 and NPR-11 (Additional file 1: Table S10 and Figure S1), indicating that receptor-activation is mediated by the penultimate arginine residue of the peptide similar to the human system.
These data show that human NPY and NPFF and C. elegans FLP/NPR signaling systems display significant pharmacological overlap in activating preferentially the G i/o pathway, cross-species activity of certain ligands and the requirement of the conserved penultimate arginine for receptor activation.
Several human neuropeptide Y and FF receptors can phenotypically rescue physiological function in a C. elegans npr-1 knockout strain The potential of C. elegans and human ligands to crossactivate neuropeptide receptors of the other species in in vitro studies suggests that both neuropeptide systems share significant similarities and may adopt each other's function. To elucidate this possibility, we used the neuropeptide receptor NPR-1 in C. elegans as a model. This GPCR has been reported to be activated by FLP-18 and FLP-21 [24,31,40], which is confirmed and extended by our study. As our data indicate that these neuropeptides are also capable to activate human Y 2 R, NPFF 1 R, and NPFF 2 R to elicit the same signal as NPR-1 (G i/o ) (Fig. 2c), it is conceivable that these GPCRs can phenotypically rescue endogenous neuropeptide signaling in the npr-1 knockout. Similar to members of the mammalian NPY receptor family, NPR-1 controls feeding, but also social behavior [12,[40][41][42]. Further functions of the receptor include response to heat [43] and substances such as ethanol [44] and methyl salicylate (MeSa) [24]. Consistently, a putative null mutation of the receptor, npr-1(ky13) [12], displays a multitude of phenotypic abnormalities, among others forming clumps and accumulating at the edge of the bacterial lawn as a result of social behavior [40] and reduced avoidance to MeSa [24]. These characteristics formed the basis for a read-out assay to test the ability of human neuropeptide GPCRs to functionally compensate loss of NPR-1. Two assays were established: one to measure accumulation of worms at the edge of the bacterial lawn (bordering) (Fig. 3a), the other to analyze avoidance to MeSa (Fig.  3b). As previously described [45], npr-1(ky13) mutants show a strong bordering phenotype (Fig. 3c) and a greatly reduced MeSa avoidance (Fig. 3d) compared to wild-type nematodes. An npr-1(ky13) strain expressing an npr-1 promoter-driven npr-1 from a transgene (construct was a kind gift of L. Ma) (Additional file 1: Figure  S3A) rescued both phenotypes (Fig. 3c, d). However, the effect of the transgene to ameliorate the bordering phenotype was small, rendering this assay suboptimal. Thus, we focused on MeSa avoidance for testing heterologous neuropeptide GPCRs.
GFP-tagged human neuropeptide receptors Y 1 R, Y 2 R, Y 4 R, Y 5 R, NPFF 1 R, NPFF 2 R, and PrRPR were transgenically expressed under the control of the npr-1 promoter in npr-1(ky13) nematodes (Additional file 1: Figure S3) and their capability to rescue the MeSa avoidance phenotype was analyzed (Fig. 3e). As hypothesized, Y 2 R, NPFF 1 R and NPFF 2 R were able to increase MeSa avoidance of npr-1(ky13) mutant nematodes to a similar level as transgenic npr-1. Interestingly, we observed that also Y 4 R and Y 5 R showed the ability to rescue the mutant phenotype, with Y 5 R to a lesser extent. Importantly, expression of Y 1 R or PrRPR did not rescue function, as expected from the lack of signaling response to any of the tested FLP in vitro, underlining the specificity of this rescue. These data show that the human neuropeptide receptors Y (Y 2 R, Y 4 R, Y 5 R) as well as FF (NPFF 1 R, NPFF 2 R) are able to phenotypically rescue the function of the C. elegans neuropeptide GPCR NPR-1.

The FLP system acts context-specific with FLP-14 being essential for GPCR function in MeSa avoidance behavior
Given the multitude of FLPs activating NPR-1, as reported by us and others (reviewed in [14]), it is difficult to resolve which ligand(s) mediate specific functions in vivo, such as MeSa avoidance. We took advantage of the narrower pharmacological spectrum of human NPY receptors that were able to substitute NPR-1 function in vivo to investigate potential flp loss-of-function mutants. As the only neuropeptides that activate both NPR-1 and all GPCRs capable of rescuing its function are FLP-14 and FLP-21 (Fig. 2c), these were prime candidates. However, since NPR-1 has been described to bind to FLP-18 and FLP-21 [24,31,40] in various contexts and FLP-15 had the potential to strongly activate NPR-1 (Fig. 2c), we also tested strains deficient for these FLPs (flp-14(gk3039), flp-15(gk1186), flp-18(gk3063), flp-21(ok889)) for their bordering and MeSa avoidance capabilities. Surprisingly, none of the mutants had any effect in the bordering assay (Fig. 4a). Likewise, neither flp-18;flp-21 double mutants nor a flp-15;flp-18;flp-21 triple mutant showed any phenotype similar to the one of npr 1(ky13) (Fig. 4a) suggesting the involvement of a different FLP and thus, a context-specific activation of NPR-1. Interestingly, flp-14(gk3039) mutants had a reduced MeSa avoidance while all other tested flp loss-of-function mutants were indistinguishable from wild-type (Fig. 4b), indicating an involvement of this so far nearly uncharacterized neuropeptide in chemosensation. As previous studies showed that flp-18(gk3063) mutants have a reduced MeSa avoidance [24], we analyzed flp-18;flp-21 double and flp-15;flp-18;flp-21 triple mutants to exclude any redundancies between these neuropeptides. However, none of them displayed any altered MeSa avoidance behavior.
Taken together, enabled by the pharmacological pattern of functional NPR-1 homologs from human we identified FLP-14 as a candidate for playing a role in avoidance of MeSa and interacting with NPR-1. Further, the FLP system seems to be highly context-specific.

Discussion
Neuropeptidergic signaling is highly complex with many essential functions in mammals and deep evolutionary conservation. Global phylogenetic analyses [6,7,13] established a common neuropeptidergic signaling repertoire in the urbilaterian: About 30 different systems diversified prior to the split of protostomes and deuterostomes, including the NPF/Y, QRFPR, and NPFF families [6][7][8]. In the corresponding phylogenetic trees [6,7], C. elegans NPRs are, however, placed next to the common bilaterian NPYR clade as representatives of a sNPF/FMRFa group, suggestive of a paralogous relationship to NPY receptors, but with a closer relationship to NPYR compared to other RFa receptors. However, reconstruction of vertebrate ancestral chromosomes [13] suggest that C. elegans NPRs have evolved from the most ancient NPY2/7R gene prior to further local gene duplications generating the NPY1/4/6R and NPY5R genes during early deuterostome evolution [13,34] indicating that human NPYR and C. elegans NPRs are coorthologs of the NPY2/7R ancestor. However, it cannot be ruled out that RFa receptors evolved later from a common ancestor in early deuterostome history followed by significant sequence divergence [13]. NPRs would then be expected to be equally distantly related to any of the human systems.
Here, we present the most comprehensive pharmacological and functional study of neuropeptidergic signaling in C. elegans to date gaining insights into relations between this system and the human NPY and RFamide systems PrRPR, QRFPR and NPFF 1/2 R, respectively. Our data show functional similarities between C. elegans NPRs and the human NPY receptor system as well as NPFF receptors as human receptors recognize several FLPs and share the G i/o protein preference with the C. elegans GPCRs. Further, human Y 2 R, Y 4 R as well as NPFF 1 R and NPFF 2 R rescue MeSa avoidance in npr-1(ky13) null mutants. It has to be noted that in contrast to other studies [46], bordering behavior is not rescued to a great extent by npr-1, which can be explained by the use of a different-sized npr-1 promoter (2 kb versus > 3 kb).
Conserved G protein preference can be one indication for orthologous receptors [28]. Although G protein specificity might change during evolutionary timescales, this is not expected to be the case for all receptor subtypes. Mutations usually retain the ancient G-protein binding pocket and add new binding interfaces with distinct epitopes [28]. We note that we cannot fully exclude that NPRs also couple to other G proteins not detected in our assays despite decent expression levels and the fact that heterologous tissue culture systems have been proven valid for characterization of nematode GPCRs [47,48]. As expected by a co-evolutionary scenario of the peptide-receptor interface outlined above, we find robust activation of human NPY and NPFF receptors by FLPs, but limited cross-species reactivity of human peptides at C. elegans NPRs. The most NPY-like FLPs FLP-27, FLP-33, FLP-34-1 and FLP-34-2 do not generally display higher potencies at the C. elegans or human receptors, but the second arginine apparently increases receptor specificity, also against the NPFF receptors.
Interestingly, PrRPR and QRFPR showed a distinct pharmacological profile: They were only activated by their endogenous ligands, and showed no phenotypic rescue in vivo. Moreover, these receptors couple to G q proteins rather than G i/o like most of the NPRs tested, arguing against an orthologous relationship.
Thus, our findings highlight functional similarities of C. elegans NPRs and human NPY receptors. Of note, we demonstrate for the first time that longer, more NPY-like peptides with a C-terminal RxRF/Ya sequence encoded by (See figure on previous page.) Fig. 3 The effect of human neuropeptide receptors on bordering and MeSa avoidance behavior of npr-1 null nematodes. a Assay layout for determining the bordering behavior of C. elegans nematodes. Adult hermaphrodites were transferred on a plate seeded with a defined bacterial lawn. Wild-type animals are dispersed randomly on the lawn (left) whereas npr-1 null mutant worms form aggregates at the border of the bacterial lawn (right). b Schematic depiction of the assay set-up determining MeSa avoidance. Assay plates lacking bacteria are divided into four quadrants, containing a drop of ethanol (EOH) and sodium azide, or MeSa and sodium azide, respectively. Nematodes are initially placed in the middle of a plate (left). In the time-course of the assay, they crawl to the quadrants with MeSa or EOH and get paralyzed by the sodium azide (right). EOH serves as control substance, as it does not have any effect on npr-1 mutant nematodes ( [24] and Additional file 1: Figure S2). Worms in each quadrant are scored and MeSa avoidances is calculated. c The bordering phenotype of npr-1 mutant individuals can be slightly ameliorated by transgenic expression of npr-1 promoter-driven npr-1. Data are show as mean ± SD in at least four independent experiments (n ≥ 480). *p < 0.05; ***p < 0.001 compared to npr-1 mutants. d MeSa avoidance of npr-1 mutants is greatly reduced compared to wild-type animals. This phenotype is rescued by transgenic expression of npr-1 driven by a npr-1 promoter (npr-1; Ex [npr-1]). Data are shown as mean ± SD in at least six independent experiments (n ≥ 300). ***p < 0.001 compared to npr-1 mutants. e Reduced MeSa avoidance of npr-1 mutants is rescued by expression of several human neuropeptide receptors driven by a npr-1 promoter from a transgene (Ex [receptor]). Construct Ex [npr-1] (also shown in (d)) served as a positive control. Data are shown as mean ± SD in at least six independent experiments (n ≥ 300). *p < 0.05; ***p < 0.001 compared to npr-1 mutants Fig. 4 Loss-of-function of C. elegans FLP-14 phenocopies the npr-1 null mutant in avoidance assay. Bordering behavior (a) and MeSa avoidance (b) of different flp loss-of-function mutants. Flp-14 mutants show an effect in MeSa avoidance, but not in bordering behavior. Data are presented as mean ± SD in at least three independent experiments (n ≥ 150). *p < 0.05; ***p < 0.001 compared to wild-type the C. elegans genome activate endogenous receptors, most prominently NPR-11 and NPR-1, and display crossspecies activity at human NPY receptors. In this regard, also the presence of a C-terminal tyrosine (typical for vertebrate NPY-like peptides) in FLP-34-1 is tolerated for the activation of these C. elegans NPRs. This corresponds well to the recent suggestion that NPR-11 is the NPYR-ortholog based on sequence similarity to Drosophila NPFR and human NPY receptors [49], and is responsible for recognition of long NPY/NPF peptides. However, we find that NPR-11, but also the other NPRs were robustly activated by FLPs of FMRFa-as well as sNPF-and partially (long) NPF-type. Thus, our data challenge the concept of distinct C. elegans receptors for FMRFa, sNPF, and (long) NPF, respectively.
Surprisingly, functional similarities between NPR and NPFF receptors were found despite the suggestion that the latter form a more distant phylogenetic group orthologous to protostomian SIFamide receptors [6][7][8]. However, phylogenetic relationships in this case are dissimilar for receptors and ligands [7], with NPFFs forming a clade with protostomian FMRFa/sNPF peptides, consistent with our functional data. Likewise, similarities between NPFF and NPY receptors are supported by high sequence homology in humans [50] and shared ligand recognition via a conserved aspartate or glutamate at the top of transmembrane helix 6 (D/E6.59 Ballesteros and Weinstein nomenclature [51]) forming a salt bridge to the peptide's penultimate arginine [39]. We identified this arginine, which is missing in SIFamides, also to be critical for FLPs to activate C. elegans and human receptors. NPR-11 does not display the typical acidic residue at position 6.59, but slightly displaced at E6.61, which is also enriched in acidic residues among NPY and RFa receptors. Similarities in peptide recognition are also underlined by the fact that residues E5.24 and Q3.32, which are important for binding of NPY and NPFF peptides [38,[52][53][54], are present in NPRs (Additional file 1: Figure S4), raising the question whether this is a case of convergent evolution or if orthologous relations need to be re-considered. Indeed, NPFF receptors appear quite permissive, in particular for short peptides. The only instances with negligible activity are NPY, PYY, PP and FLP-34-1, which carry a C-terminal RYamide. The phenylalanine to tyrosine exchange also reduces potency in the short endogenous NPFF ligand [37]. This lack of tolerance towards a C-terminal tyrosine is not seen in the FLP/NPR system, tentatively arguing against an orthologous relationship. In contrast, the interaction pattern of FLPs with NPYRs appeared more specific and a C-terminal tyrosine residue is accepted in the context of the FLP-34-1 peptide. In conjunction with the phylogenetic reconstructions and in vivo data, this more likely reflects an orthologous relationship. Due to our study design, we can, however, not rule out that the functional similarities between NPR and human NPY receptors, although striking, result from convergent evolution.
The comparative and comprehensive design of our study allowed to clarify several physiological aspects of C. elegans neuropeptide signaling and thus, highlights the value of such an approach. The activation pattern of NPRs displayed considerable redundancy as one receptor is activated by multiple FLPs. We suggest that this redundancy is a key in realizing context-specific receptor activation in distinct cells expressing distinct sets of FLPs. Contextdependent activation of one NPR by different FLPs might be one mechanism to orchestrate the diverse functions, among them roles in feeding and social behavior [12,[40][41][42], regulation of aerotaxis [45,55,56], response to ethanol [44], MeSa [24] and innate immunity [57,58], all related to NPR-1. While NPR-1 is activated by FLP-21 to control social behavior [31,45], we did not observe any effect of FLP-21 on MeSa avoidance indicating that this NPR-1 function is mediated by a different FLP. The combination of in vitro activation assays and rescue analyses using human GPCRs identified FLP-14 as a candidate for exerting the NPR-1 effect in response to MeSa. Interestingly, FLP-14 was a high-affinity partial agonist in our in vitro screening, which may contribute to a unique pharmacological profile. By maintaining a submaximal response, receptor desensitization (by arrestin-mediated or alternative mechanisms) is minimized, which helps to maintain the responsiveness of the addressed receptors and/or a certain 'tone' of an important biological signal. The role of this so far only poorly characterized FLP was confirmed by the flp-14(gk3039) null mutant, which phenocopies the MeSa avoidance seen in npr-1(ky13) null mutants. It needs to be noted that previous studies describe FLP-18 as an interactor of NPR-1 in MeSa avoidance [24]. Although a flp-18(gk3063) mutant shows a trend in avoidance behavior, we did not observe any significant effect, possibly originating from a slightly different assay set-up.
The concept of context-specific activation is supported by the distinct expression patterns of different flps (WormBase, release WS269): While FLP-14 is localized to the nerve ring, flp-18 is expressed mainly in head, nerve ring and nerve cord neurons, whereas FLP-21 is present in neurons of the head, intestine and tale. There are only very few neurons in which overlapping expression of two of these neuropeptides has been described. Thus, context-specific expression of flps (and potentially their receptors) in combination with different specificities for NPRs might be one way of creating specificity in the highly complex FLP/NPR system. This somewhat resembles the situation in the human NPY system, in which the biological specificity also seems to arise from both, pharmacological (ligand) preferences and specific receptor expression and hormone secretion profiles. For instance, the selectivity between PP and NPY in activating the Y 4 R is backed up by the predominant expression of this receptor subtype in the gastrointestinal tract, where it senses comparably low levels of circulating of peptide hormones as part of the gut-brain axis [1,4]. Moreover, the biological functions of the Y 1 and Y 2 receptors are divergent despite very similar functional affinities of NPY and PYY to activate the G i pathway at these receptors [4]. Thus, the differential physiological functions must be encoded in the cellular expression in the brain, which is evidenced by expression studies (e.g. reviewed in [1]), and/or additional cellular pathways.

Conclusions
In the present study, we provide evidence for a great extent of pharmacological and functional similarity between the C. elegans RFamide system (FLP/NPR) and human NPY as well as NPFF receptors. For the first time, we characterize NPY-like peptides (FLP-27, FLP-33, FLP-34-1, FLP-34-2) of C. elegans which underlines parallels in the peptide repertoire and requirements for receptor activation. This adds a functional perspective to current phylogenetic reconstructions and corroborates the suggested orthology between these systems. These findings demonstrate that they can serve as models to gain insights into the biology and molecular mechanisms of the respective other system. While several other organisms have been successfully employed to investigate functions of neuropeptides and their receptors such as Drosophila melanogaster (summarized in [10]), with the knowledge of our study on hand, the nematode offers distinct advantages for more generic studies. For instance, its transparency enables tracking of multiple fluorescently labeled peptides and monitoring their dynamics. This approach can also be conducted in a highthroughput manner, up to a 384-well format.
Thus, our study offers the foundation to investigate structure-function aspects of human receptors in an easily manageable in vivo model that we expect to be highly valuable for future studies.