Co-transmission of neuropeptides and monoamines choreograph the C. elegans escape response

Co-localization and co-transmission of neurotransmitters and neuropeptides is a core property of neural signaling across species. While co-transmission can increase the flexibility of cellular communication, understanding the functional impact on neural dynamics and behavior remains a major challenge. Here we examine the role of neuropeptide/monoamine co-transmission in the orchestration of the C. elegans escape response. The tyraminergic RIM neurons, which coordinate distinct motor programs of the escape response, also co-express the neuropeptide encoding gene flp-18. We find that in response to a mechanical stimulus, flp-18 mutants have defects in locomotory arousal and head bending that facilitate the omega turn. We show that the induction of the escape response leads to the release of FLP-18 neuropeptides. FLP-18 modulates the escape response through the activation of the G-protein coupled receptor NPR-5. FLP-18 increases intracellular calcium levels in neck and body wall muscles to promote body bending. Our results show that FLP-18 and tyramine act in different tissues in both a complementary and antagonistic manner to control distinct motor programs during different phases of the C. elegans flight response. Our study reveals basic principles by which co-transmission of monoamines and neuropeptides orchestrate in arousal and behavior in response to stress.

Introduction Co-localization of classical neurotransmitters and neuropeptides is a common feature of the animal nervous system [1]. Co-transmission is thought to provide flexibility to the output of hard-wired neural circuits [2]. In mammals, for instance, the acute fight-or-flight response leads to the activation of the sympathetic nervous system (SNS) and the co-release of different "stress hormones" [3]. Adrenalin and noradrenaline release in the SNS trigger an increase in heart rate, blood flow, respiration and release of glucose from energy stores, which prepare the animal for vigorous muscle activity and physical exertion [4,5]. Neuropeptide Y (NPY), one of the most abundant neuropeptides in the mammalian nervous system, is a co-transmitter with noradrenaline (NA) in many neurons of the SNS [6][7][8]. The sympathetic co-transmission of NA and NPY suggest that they may coordinate aspects of the flight response. However, the physiological and behavioral impact of co-transmission can be difficult to dissect. Co-transmitted signaling molecules can activate receptors on a common target (convergence) or different targets (divergence) which induce synergistic or opposing effects, making the functional outcome of co-transmission challenging to anticipate. Studies in mammals have shown that the stress related modulatory functions of NPY and NA co-transmission are complex, with complementary actions in some tissues and antagonistic actions in others. For example, both NA and NPY increase blood pressure through peripheral vasoconstriction, however NPY inhibits presynaptic NA release from sympathetic neurons and opposes the action of NA on cardiac contraction [9][10][11][12][13][14][15][16][17][18][19]. Co-transmission of NPY and catecholamines may induce a longer lasting state of arousal that enhances alertness and the ability to deal with environmental threats. Unraveling the precise effects of co-transmission of neural stress hormones is very challenging in vertebrates given the complexity of the nervous system, the multiple central and peripheral release sites and the diversity of target tissues expressing NA and NPY receptor (NPYR) subtypes.
The nematode Caenorhabditis elegans provides an excellent system to study the co-transmission of aminergic and peptidergic neuromodulators due to its compact and completely defined nervous system and wealth of genetic tools [20,21]. The C. elegans genome encodes a large family of NPY related peptides and G-protein coupled receptors (GPCRs) [22,23]. Like other invertebrates, C. elegans lacks NA, however the structurally related tyramine fulfills a similar role to NA in coordinating stress responses and flight behavior [24][25][26].
In response to a mechanical stimulus C. elegans can engage in a flight-or an escaperesponse, where the worm quickly reverses while suppressing oscillatory head movements. The reversal is followed by a deep ventral bend of the head, and a subsequent slide of the head along the ventral side of the body (omega turn). After the omega turn, the animal moves forward in the opposite direction, away from the noxious stimulus. Tyramine plays a crucial role in the coordination of independent motor programs, which increases the animal's chances to escape from predatory fungi that use hyphal nooses to entrap nematodes [26][27][28][29]. The escape response triggers the release of tyramine from a single pair of neurons called the RIM. Tyramine release activates the inhibitory tyramine-gated chloride channel, LGC-55, in neck muscles, cholinergic head motor neurons and the AVB pre-motor interneuron.
Here we investigate the role of FLP-18 in the orchestration of the C. elegans escape response. We find that FLP-18 plays a central role in coordinating distinct phases of the escape response in C. elegans. FLP-18 is co-released with tyramine from the RIM neurons in response to mechanical stimuli. FLP-18 activates the GPCR NPR-5 in body wall muscle to enhance turning behavior and locomotion speed during the escape response by increasing muscle excitability. Our result show that FLP-18 and tyramine act in different tissues in both a complementary and antagonistic manner to orchestrate distinct motor programs during different phases of the C. elegans flight response.

Mechanical stimulation transiently increases forward velocity
Handling of C. elegans, such as the transfer with a platinum wire, bumping the plate on microscope stage or removing the lid of a plate can induce an increase in locomotion rate that persists for several minutes [45,46]. This indicates that mechanical stimulation can lead to longer lasting changes to the internal state of the animal. To quantify locomotion patterns upon mechanical stimulation we analyzed the locomotion rate of animals subjected to a tap to the side of the plate. Mechanical tap can trigger an escape response, in which C. elegans reverses, turns and resumes forward locomotion in the opposite direction [47,48]. We compared behavior of animals before, during and after spontaneous reversals and tap-induced reversals. In animals that initiated a spontaneous reversal, forward locomotion rate remained the same before and after the reversal (Fig 1A). The absolute velocity during spontaneous reversals was similar to the velocity of forward movement.
When an escape response was triggered with a tap stimulus, animals initiated a rapid reversal; double the speed of spontaneous reversals (Fig 1B and 1C). Furthermore, following a tapinduced reversal, wild-type animals exhibited a markedly elevated forward locomotion rate for approximately 3 minutes (S1A Fig);-a behavior we refer to as the 'forward run' (Fig 1B and  1D). The sustained increase in forward velocity in response to a tap indicated longer lasting changes in the internal state of the animal.

flp-18 mutants have defects in locomotory arousal
Neuropeptides are potent neuromodulators, whose action can lead to longer lasting changes in behavior. We focused our attention on FLP-18 neuropeptides, since the flp-18 gene is known to be expressed in the tyraminergic RIM neurons and the AVA pre-motor interneurons [32,33]. The RIM and AVA play a central role in the escape response to mechanical touch [25,26,47,49]. flp-18 mutants and tyramine deficient tdc-1 mutants initiated reversals in response to tap stimuli similar to wild type ( Fig 1E). This indicates that FLP-18 peptides and tyramine are not required for mechanosensation or the initiation of the escape response. In response to a tap stimulus, flp-18 mutants reversal velocity was slightly reduced compared to wild type and tdc-1 mutant animals (Fig 1B and 1C). Both flp-18 and tdc-1 mutants exhibited shorter reversal duration compared to wild type ( Fig 1B). tdc-1 mutants had a reduced basal Backward-and forward-velocity upon a spontaneous or tap-induced reversal. Wildtype animals show a persistent increase in locomotion speed in response to a strong tap stimulus ("forward run"). flp-18 mutants do not show a robust increase locomotion speed in response to a strong tap stimulus. (A-B) Schematic illustrating behavioral sequence (top) and velocity traces (bottom) from wild type animals, flp-18, and tdc-1 mutants, aligned to reversal events. Negative velocity indicates backward locomotion. Dark lines represent mean velocity and shaded region represents standard error. Black lines at the top of the graph indicate the time periods when reversal and forward run speed were quantified. Reversals velocity, but following a tap, increased velocity during the forward run, similar to the wild type. In contrast, flp-18 mutants had significantly slower forward run velocity compared to the wild type (Fig 1B and 1D). tdc-1; flp-18 double mutants behaved similar to tdc-1 single mutants during the initial phase of the escape response and like flp-18 single mutants failed to maintain forward velocity at later stages of the forward run (S1B Fig).

flp-18 mutants have defects in head and body bending during the escape response
Under regular conditions flp-18 mutants had no obvious defects in locomotion velocity, body curvature, foraging head movements, or spontaneous reversal frequency (Fig 2A-2F) and propagated regular body bends as they moved across the agar surface ( Fig 2G). We analyzed flp-18 mutant animals for defects in other aspects of the escape response, since flp-18 mutants fail to sustain rapid movement during the forward run. During the initial phase of the escape response elicited by gentle anterior touch, animals reverse while suppressing oscillatory head movements. The reversal is often followed by two linked motor programs that together compose an omega turn: a ventral head swing which initiates the omega turn; followed by a deep body bend where the animal slides its head along the ventral side of the body to resume locomotion in the opposite direction. In response to gentle anterior touch flp-18 mutants reversed and suppressed oscillatory head movements like the wild type (Fig 3E). Following a reversal, flp-18 mutants often initiate an omega turn similar to the wild type. However, flp-18 mutants made shallower ventral head bends ( Fig 3B) and executed omega turns (> 90-degree turn) less frequently than wild-type animals (37% vs 57% respectively) ( Fig 3C). Furthermore, when flp-18 mutants did make high-angled omega turns, the head often failed to contact the ventral side of the body during the turn [29] (Fig 3A) resulting in a much higher proportion of open omega turns compared to wild-type animals (62% vs 22% respectively) ( Fig 3D). Expression of a low-copy transgenic flp-18 genomic construct, which includes a 3 kb endogenous promoter ([Pflp-18::FLP-18], referred to as FLP-18(+)), rescued the omega turning defects seen in flp-18 mutants (Fig 3C-3E).

flp-18 overexpression enhances body bending and omega turning
To further assess the effect of FLP-18 signaling on behavior we overexpressed flp-18 using a high-copy transgene ([Pflp-18::FLP-18], referred to as FLP-18(+++)). flp-18 overexpression caused uncoordinated locomotion with strikingly deep body bends (Fig 2A and 2G) consistent with previous reports [32]. Animals overexpressing flp-18 also displayed hyperactive foraging head movements, increased body curvature, and increased spontaneous reversal frequency (Fig 2C-2F). flp-18 reporters are expressed in the AVA, RIM, AIY, RIG, the M1 and M3 pharyngeal neurons and ventral cord motor neurons [32,33,39]. We overexpressed flp-18 in a subset of neurons of the escape circuit using the cex-1 promoter, which mainly drives expression in the RIM, AVA and AVD neurons [50], but not the AIY, RIG, pharyngeal and ventral cord neurons (S2A Fig). Overexpression of the Pcex-1::FLP-18 transgene in a flp-18 mutant background was sufficient to cause exaggerated body bends and reversals, similar to Pflp-18::FLP-were spontaneous (A) or induced by a tap stimulus (B). (C) Peak backward velocity during the reversal phase was quantified by identifying the maximum backward locomotion rate for each animal during a 1 second period following the tap. (D) Forward run velocity was quantified as the average speed for each animal during the time period between 7 to 10 seconds following the tap. (E) Quantification of the percentage of animals initiating a reversal within 1 second after a tap was delivered. Graphs represent mean ± SEM, significance was calculated using ANOVA with Šidák's (C and D) and Dunnett's (E) multiple comparison test (P> 0.05 = ns, P<0.005 = �� , P<0.0001 = ���� ). Sample sizes: Spontaneous reversals and escape response (A-D), (n = # animals). Spontaneous   Fig and S1 and S2 Videos). This indicates that FLP-18 release from the escape circuit neurons, AVA, RIM and AVD, is sufficient to enhance body bending and reversals. FLP-18 overexpression had an opposite effect on the escape response compared to flp-18 deficiency. In contrast to flp-18 mutants, which have reduced ventral head bend angle, FLP-18 overexpressing animals made high angled head bends ( Fig 3B). Furthermore, overexpression of FLP-18 resulted in an increased omega turn frequency and a reduced proportion of open omega turns (Fig 3C and 3D). Moreover, in response to a mechanical stimulus FLP-18 overexpressing animals failed to suppress head movements during reversals (Figs 3E and S3 and S3 and S4 videos). This phenotype is similar to tyramine deficient mutants (S1C Fig) and mutants that lack a tyramine-gated chloride channel, LGC-55 [26,28,30]. In wild-type animals, suppression of head oscillations during touch induced reversals is mediated by tyraminergic activation of LGC-55 in neck muscles and SMD and RMD head motor-neurons. The observation that animals overexpressing FLP-18 fail to suppress head movements suggests that FLP-18 may oppose the tyraminergic inhibition of neck muscles.

The escape response induces FLP-18 release from the AVA and RIM
flp-18 is expressed in the AVA and RIM neurons [32,33], which are activated during the escape response [25,49,51]. To analyze FLP-18 release, we generated transgenic animals with a fluorescently tagged the FLP-18 pro-peptide with the YFP variant Venus (Pflp-18::FLP-18::Venus). Venus is resistant to quenching in low pH environments and has been used to monitor neuropeptide secretion from dense core vesicles in C. elegans [52][53][54]. FLP-18::Venus fluorescence was observed in the RIM and AVA escape circuit neurons, as well as the AIY and RIG ( Fig  4B). FLP-18::Venus fluorescence was also observed in the coelomocytes, in contrast to a Pflp-18::GFP transcriptional reporter which did not produce detectable coelomocyte fluorescence [32]. This is consistent with the idea that FLP-18::Venus protein is secreted and taken up by coelomocytes, which are endocytic scavenger cells that internalize material from the pseudocoelomic fluid [55]. We measured changes in fluorescent intensity in the AVA and RIM neurons after repeated activation of the escape response (Fig 4A and 4C). Mechanical tap to the plate caused a progressive reduction in FLP-18::Venus fluorescent intensity in the AVA and RIM neurons compared to animals that did not receive the tap stimuli ( Fig 4D and 4E), suggesting that activation of the escape response elicits the release of FLP-18 from the AVA and RIM. Tap treatment did not change the fluorescent intensity in coelomocytes ( Fig 4F). This could be due to protein break down in the coelomocytes and/or the high basal levels of coelomocyte fluorescence due to tonic FLP-18 release from other neurons.

FLP-18's modulation of the escape response depends on npr-5
How does FLP-18 signaling modulate locomotion and the escape response? FLP-18 peptides activate the GPCRs: NPR-1, NPR-4 and NPR-5 in vitro [32,36,44]. We analyzed npr-1, npr-4, and npr-5 loss-of-function mutants for defects in the escape response. npr-1, npr-4 and npr-5 mutants displayed largely normal locomotion and initiated an escape response upon anterior touch. While loss of npr-1 or npr-4 had no impact on the frequency of omega turns, npr-5 mutants executed fewer omega turns compared to the wild type ( Fig 5A). The decrease in omega turning of npr-5 mutants was comparable to flp-18 mutants. Furthermore, like flp-18 mutants, npr-5 mutants also display a higher proportion of open omega turns ( Fig 5B). This suggests that FLP-18 peptides released during the escape response enhance head and body bending to promote turning through activation of the GPCR NPR-5.

PLOS GENETICS
Co-transmission of neuropeptides and monoamines choreograph the C. elegans escape response Next, we examined whether neuropeptide receptor mutants could suppress the hyper bending phenotype of the FLP-18 overexpression strain. Mutations in either npr-1, npr-4 or npr-5 alone had no effect on body curvature during basal locomotion (S4A Fig). In the FLP-18(+++) background, deletion of npr-5, but not npr-1 or npr-4, restored body curvature to wild-type levels ( Fig 5C). Loss of npr-5 also suppressed the hyper bending during the omega turn caused by FLP-18 overexpression (Figs 5D and S4B). npr-5; FLP-18(+++) animals make omega turns similar to wild type ( Fig 5F). However, the proportion of open omega turns made by npr-5; FLP-18(+++) animals is significantly increased and similar to npr-5 single mutants ( Fig 5G).
npr-1; FLP-18(+++) and npr-4; FLP-18(+++) were indistinguishable from FLP-18(+++) in their failure to suppress head oscillations in response to touch. In contrast, npr-5; FLP-18(++ +) animals suppressed head movements in response to anterior touch like wild-type animals ( Fig 5E). Loss of both npr-1 and npr-5 reduced the high spontaneous reversal frequency caused by FLP-18 overexpression (Fig 5H). npr-1 single mutants displayed significantly fewer spontaneous reversals compared to wild-type animals (S4C Fig), likely due to the fact that npr-1 mutants avoid ambient oxygen concentrations [56]. Under low oxygen conditions loss of npr-1 failed to suppress the increased reversal frequency caused by FLP-18 overexpression (S4D Fig). This suggests that the effect of npr-1 on reversal frequency reflects its role in oxygen sensation independent of npr-5. We further analyzed npr-5 mutants for locomotion defects during the escape response. Tap-stimulated peak reversal velocity was slightly reduced in npr-5 mutants. In addition, there was a marked decrease in forward run speed (Fig 6A-6D), closely resembling the defect observed in flp-18 mutants. Taken together these data indicate that FLP-18 modulates locomotory arousal in the escape response primarily through the activation of the NPR-5 receptor.

NPR-5 acts in head-, neck and body wall muscle to drive body bending
To analyze where NPR-5 acts we generated transgenic animals that express a GFP tagged NPR-5 translational fusion [Pnpr-5::NPR-5::GFP]. NPR-5::GFP expression was detected in head, neck, and body-wall muscle (Fig 6E), and in a few neuronal cell bodies in the head ( Fig  6E, lower right panel), consistent with previously described transcriptional reporters for npr-5 [36]. NPR-5::GFP expression was particularly strong in head-and neck-muscles. We found that NPR-5::GFP localizes to dense bodies within muscle cells and is highly enriched in muscle arms and the muscle plate around the nerve ring ( Fig 6E). The neck muscle arms around the nerve ring receive direct synaptic inputs from the FLP-18 expressing RIM neurons [20]. Overexpression of NPR-5 under control of the endogenous promoter (Pnpr-5::NPR-5) did not cause elevated curvature during basal locomotion ( Fig 6F). However, in response to a mechanical stimulus, Pnpr-5::NPR-5 animals displayed increased body curvature comparable to FLP-18(+++) animals ( Fig 6G). This suggests that the effect of NPR-5 overexpression driven by the endogenous promoter is uncovered when FLP-18 is released during the escape response. Cellspecific overexpression of NPR-5 in all body wall muscles (including head-and neck-muscles;

PLOS GENETICS
Co-transmission of neuropeptides and monoamines choreograph the C. elegans escape response Pmyo-3::NPR-5), on the other hand, did increase body curvature during both basal and tapinduced locomotion, similar to FLP-18 overexpressors (Fig 6F and 6G). The discrepancy in basal curvature between the Pnpr-5::NPR-5 and Pmyo-3::NPR-5 transgenic animals may be due to differences in NPR-5 expression pattern and/or level. Pmyo-3 is expressed strongly in all body wall muscle, while expression of the endogenous Pnpr-5 promoter is restricted to a subset of neurons and is high in head-and neck-muscles. Overexpression of NPR-5 under control of its endogenous promotor increased spontaneous reversal frequency, similar to FLP-18 overexpressing animals. Muscle specific expression of NPR-5 did not increase reversals ( Fig  6H). Consistent with previous observations, this may suggest that neuronal expression of NPR-5 modulates reversal behavior [42]. Our results indicate that FLP-18 acts through NPR-5 in muscle to enhance body bending.

FLP-18 increases calcium levels in body wall muscle
How does FLP-18 stimulate body bending? Muscle contractions are triggered by the rapid entry of calcium into the muscle cytosol [57,58]. To determine if FLP-18 signaling affects calcium levels in body wall muscle we generated a strain expressing the fluorescent calcium indicator GCaMP6s in body wall muscles [Pmyo-3::GcaMP6] (Fig 7A). C. elegans locomotion is driven by alternating waves of contraction and relaxation in dorsal and ventral muscles. During locomotion the highest GCaMP fluorescence was observed when muscles were contracted and faded upon relaxation, consistent with calcium influx driving muscle contraction ( Fig  7A). flp-18 mutants showed a reduction in GCaMP fluorescence in the muscle (Fig 7B). In sharp contrast, overexpression of FLP-18 resulted in a striking increase of GCaMP fluorescence in body-wall muscles (Fig 7A and 7B). This indicates that FLP-18 increases calcium transients in muscles that facilitate body bending.
Like flp-18 mutants, npr-5 mutants displayed a similar decrease in GCaMP fluorescence in body wall muscles. Furthermore, an npr-5 mutation completely suppressed the increase in GCaMP fluorescence of FLP-18 overexpression (Fig 7B). In vitro experiments indicate that NPR-5 signals primarily through the Gα q pathway [44]. Activation of Gα q coupled receptors can cause the release of calcium from intracellular stores [59,60]. egl-30 encodes the main C. elegans Gα q ortholog [61]. To determine whether in vivo NPR-5 signaling is Gα q dependent we examined the effect of mutations in egl-30 on muscle calcium levels. Loss of egl-30 substantially decreased GCaMP fluorescence in muscle. GCaMP fluorescence was also strongly reduced in the muscles of egl-30; FLP-18(+++) animals and indistinguishable from egl-30 single mutants (Fig 7B). Mutations in FLP-18 signaling genes did not affect intensity of a coexpressed Ca 2+ insensitive fluorophore (S5A and S5B Fig) indicating that the differences in GCaMP fluorescence are due to altered intracellular Ca 2+ levels. Our results support the idea

PLOS GENETICS
Co-transmission of neuropeptides and monoamines choreograph the C. elegans escape response that FLP-18 activates NPR-5 and EGL-30/Gα q signaling cascades, resulting in increased calcium levels in the muscle (Fig 8B).

FLP-18 enhances motor output through the activation of NPR-5
In this study we identify FMRFamide-like peptides encoded by the flp-18 gene as key modulators of the distinct phases in the C. elegans escape response. We showed that FLP-18 neuropeptides are released from the RIM and AVA neurons during an escape response elicited by a mechanical stimulus. FLP-18 promotes head and body bending during the omega turn and increases forward run velocity following the turn. We find that the effects of FLP-18 on

PLOS GENETICS
locomotion are largely mediated through activation of the Gα q coupled NPR-5 receptor. NPR-5 is highly expressed in head and body wall muscles which receive direct synaptic inputs from the RIM neurons. Studies in both vertebrates and C. elegans have shown that EGL-30/Gα q activates phospholipase Cβ (PLCβ) [62,63]. EGL-8/PLCβ increases IP 3 levels which can lead to the

PLOS GENETICS
Co-transmission of neuropeptides and monoamines choreograph the C. elegans escape response release of Ca 2+ from internal stores [64,65]. We find that FLP-18 activation of NPR-5 increases calcium levels in muscles in vivo in an EGL-30/Gα q dependent manner. In vitro studies that show that NPR-5 activation by FLP-18 peptides resulted in rises in intracellular Ca 2+ levels [44]. Our data support a model in which FLP-18 release during the flight response stimulates Ca 2+ release in head and body wall muscle through the NPR-5-EGL-30/Gα q signaling pathway (Fig 8B). Increases in calcium levels would stimulate muscle tension, required for head and body bending the during the omega turn.
Previous studies have indicated that FLP-18 acts in part through NPR-5 in body wall muscles to decrease the convulsion phenotype of nicotinic acetylcholine receptor acr-2(gf) mutants [39]. Our data raise the possibility that a FLP-18 dependent increase in intracellular Ca 2+ concentrations may reduce the Ca 2+ gradient across the muscle membrane and thus provides a homeostatic response in muscles to the elevated neuronal activity of cholinergic motor-neurons in acr-2(gf) mutants. Recent reports have shown FLP-18 acts through NPR-5 in sensory neurons to control reversal frequency [42]. We found that NPR-5 overexpression using the endogenous promoter (neurons and muscle) increases reversal frequency, while muscle-specific overexpression had no effect. This observation is consistent with a neuronal contribution of NPR-5 in the modulation of reversals. We hypothesize that FLP-18 and NPR-5 mediated increases in muscle Ca 2+ levels and excitability contribute to the increase in forward run velocity following the turn. However, we cannot exclude a role for neuronal FLP-18/NPR-5 signaling in the regulation of locomotion rate.
Together, our results show that FLP-18 signaling functions to enhance motor output during the escape response and locomotory arousal through activation of NPR-5 in muscle. FLP-18 thus may contribute a locomotory arousal state that enhances vigilance and the ability to escape from predation. Other studies have shown that FLP-18 activation of NPR-1, dampens the activity of sensory neurons to promote the sleep like quiescent period associated with larval molts [37,38,66,67]. Interestingly, activation of the escape response also induces the release of a distinct RFamide peptide, FLP-20, from mechanosensory neurons. FLP-20 release enhances arousal and motor output across multiple sensory modalities through activation of NPYR like receptor FRPR-3 in the neurosecretory RID neurons [46]. These findings highlight the multilevel nature of RFamide neuropeptides in the modulation of arousal in C. elegans.

Co-transmission of FLP-18 and tyramine orchestrate the flight response
Many neurons contain both small molecule transmitters and neuropeptides. In addition to FLP-18, the RIM neurons also contain additional neuropeptides and the neuromodulator/ neurotransmitter tyramine and glutamate [26,[68][69][70]. We have previously shown that tyramine release from the RIM during the escape response coordinates independent motor programs of the escape response [26,28,29]. Our findings indicate that activation of the escape response results in the co-transmission of tyramine and FLP-18 from the RIM. The RIM neurons contain dense core vesicles [71] which can store both neuropeptides and monoamines [72,73]. The RIM also contains regular synaptic vesicles (possibly for the release of glutamate); thus, it is currently unclear whether tyramine and FLP-18 are co-released from the same vesicle. The well-defined escape circuit and the role of the RIM in this response provides the opportunity to dissect of how tyramine and FLP-18 co-transmission choreograph independent motor outputs in the execution of the escape response. Co-transmission of FLP-18 and tyramine from the RIM are convergent in some aspects of the escape response but divergent in others. For example, FLP-18 and tyramine both act to facilitate deep bending during the omega turn but do so through distinct mechanisms in different tissues. Tyramine inhibits the GABAergic VD motor-neurons through activation of the Gα o coupled receptor SER-2 [29].
The inhibition of GABA release onto ventral body wall muscle facilitates the deep ventral contraction during the omega turn. FLP-18 on the other hand activates NPR-5 in the neck-and body wall-muscles resulting in increased calcium levels and enhanced head and body bending. Thus, co-transmission of FLP-18 and tyramine exert excitatory and inhibitory effects in different cells resulting in functional synergism that enhances omega bending (Fig 8A and 8B). In contrast to the complementary function of tyramine and FLP-18 on body bending, we observe convergent but antagonistic effects of these neuromodulators on exploratory head oscillations. During the reversal phase of the escape response, wild-type C. elegans suppress head oscillations to increase its chances to escape encounters with hyphal nooses of predatory fungi [27]. Tyramine release from the RIM mediates the suppression of head oscillations through the activation of a tyramine-gated chloride channel, LGC-55, in neck muscles and cholinergic motor neurons that synapse onto neck muscles [28]. Animals that overexpress FLP-18 fail to suppress head oscillations during reversals, but this defect is rescued by loss of npr-5. The simplest explanation is that in FLP-18 overexpressing animals, the tyramine induced hyperpolarization of neck muscles is no longer sufficient to overcome the excess of FLP-18 induced depolarization. In the behavioral sequence of the escape response, head oscillations are reinitiated during the forward run that follows the omega turn. The tyraminergic activation of a fast-acting ionotropic LGC-55 receptor and FLP-18 activation of the slow-acting metabotropic NPR-5 receptor may contribute to coordinate the timing of suppression and re-initiation of head oscillations.
The use of genetically tractable model organisms with simple nervous systems like C. elegans allows systematic interrogation of the mechanisms underlying co-transmitter modulation of behavior. Our work has shown examples of convergent/antagonistic as well as divergent/ complementary actions of co-transmitters released from a single neuron pair. Using a similar genetic and behavioral approach, other functional properties of co-transmission have been elucidated in C. elegans. Co-transmission of serotonin and the neuropeptide NLP-3 coordinate the activity of the C. elegans egg-laying circuit [74]. Foraging behavior and olfactory adaptation are shaped by convergent / antagonistic co-transmission of glutamate and NLP-1 from the AWC sensory neurons to dampen olfactory responses [75].
The complexity of mammalian system makes it more difficult to dissect the in vivo consequences of neuropeptide and co-transmission. Nevertheless, there are intriguing similarities in mammals where neuropeptide Y (NPY) is co-released with noradrenaline (NA) in response to stress [6,7,15,16,76,77]. In mammals, co-transmission of NA and NPY in the mammalian sympathetic nervous system has both complementary and antagonistic actions. For example, both NA and NPY produce a pressor response through contraction in vascular smooth muscle [10,12,16], while NPY and NA have opposing inotropic and chronotropic effects on cardiac muscle [11,13,14,78]. The similarities between vertebrate NA/NPY and C. elegans tyramine/ FLP-18 signaling indicate that co-transmission monoamines and neuropeptides is a central feature of the signaling pathways that underlie arousal in response to stress. In the presence of danger, heightened arousal can prepare an animal to respond more effectively to a threat and increase the chance of survival. The stereotyped C. elegans escape response allows us to elucidate how co-transmission from a single neuron can shape circuit activity on different time scales to orchestrate arousal and physiology in response to stress.

Molecular biology and transgenics
Standard molecular biology methods were used. A list of strains used is this manuscript is included as Table 1. Most transgenic strains were made through microinjection of plasmid DNA into lin-15(n765ts). A lin-15 rescuing plasmid (PL15EK) was used as a co-injection marker (80 ng/μl) and transgenics were identified based on rescue of the lin-15 multivulva phenotype. FLP-18 transgenic strains were created by amplifying a 5.9kb sequence of genomic DNA beginning 3.3kb upstream of the open reading frame and containing the flp-18 coding region as well as the 3' UTR. The sequence was amplified using the forward primer 5'-GTG ATGTAGGTAGCGCGGAAC-3' and 5'-TTCATGATCCCAGATTCAATTAATC-3' reverse primer. This amplicon was ligated into a plasmid containing an SL2 linked mCherry transcriptional reporter to visualize expression. This plasmid was injected along with pL15EK and transgenics were identified based on lin-15 rescue. The extrachromosomal array was stably integrated into the genome using X-ray irradiation and were backcrossed to wild type N2 animals 4 times. NPR-5 transgenics were created using Gibson Assembly (Gibson Assembly Master Mix (E2611), New England BioLabs inc., [79]) to insert the npr-5 genomic locus into a pSP72 plasmid backbone. The insert consisted of a 7.7kb sequence of genomic DNA beginning 3kb upstream of the npr-5 open reading frame and including the npr-5 coding sequence and 3' UTR. The sequence was amplified using 5'-cagatctgatatcatcgatGGCCCCGTGATTCTTGTC A-3' as a forward primer and 5'-atccccgggtaccgagctcgaatTCTTGCTCGGTGTGCATTTTC-3' as a reverse primer. The lowercase letters indicate the homology arms that anneal to pSP72 sequence, and the uppercase letters anneal to the npr-5 genomic locus for amplification. pSP72 was linearized for insertion using EcoRI. The resulting plasmid was injected along with pL15EK and transgenics were identified based on lin-15 rescue.

Behavioral experiments
All behavioral experiments were carried out using young adult animals which had been staged as L4s the previous day. Behavioral experiments were performed on medium (60mm) thinlawn NGM plates where a single drop of OP50 (~30μl) was spread across the plate and allowed to grow at room temperature overnight. Thin lawn plates older than 24hrs were discarded. Prior to behavioral experiments, animals were transferred from standard plates to thin lawn plates and allowed to acclimate for at least 5 minutes. All experiments were conducted with 20 worms to a plate unless otherwise stated and were run in parallel with wild type controls across at least 3 days.

Worm tracking
Quantification of locomotion was performed using one of two different worm tracking systems. For quantification of foraging speed (Fig 2C), the Worm Tracker 2.0 system was used (Eviatar Yemini and Tadas Jucikas, https://www.mrc-lmb.cam.ac.uk/wormtracker, [80]) and "absolute foraging speed" was reported. A single animal on a thin lawn plate was tracked for 5 minutes at 30fps using a USB digital microscope (DinoLite Pro AM413T, Dino-Lite US) on a motorized stage (Zaber TNA08A50 linear actuator and Zaber TSB60-M motorized stage, Zaber Technologies Inc.). Quantification of velocity, body curvature and reversal frequency were measured using the Multi-Worm Tracker (Rex Kerr, https://sourceforge.net/projects/mwt/). When tracking experiments included a tap stimulus, the stimulus was delivered by a tubular push solenoid (Saia-Burgess/Ledex #195205-127) controlled by the Multi-Worm Tracker software. The tap stimulus consisted of a train of 20 taps with an inter tap interval of 10 milliseconds. Tap stimuli were always delivered after 5 minutes of tracking had elapsed.
Experiments were analyzed using custom MATLAB (The MathWorks, Inc.) scripts to interface with the Multi-Worm Tracker feature extraction software Choreography. Analysis was limited to objects that had been tracked for a minimum of 20 seconds and had moved a minimum of 5 body lengths. All experiments were conducted on thin lawn plates with 20 animals on each plate.

Body curvature
For comparisons of basal versus tap stimulated curvature animals were tracked using the Multi-Worm Tracker. The body curvature of the population was averaged over 5 seconds either just prior to the tap stimulus for basal curvature or just after the tap stimulus for stimulated curvature.

Reversal frequency
20 animals were tracked while freely moving on a thin lawn plate with the Multi-Worm Tracker. The total number of reversals made by the population in a 3-minute window was extracted using the Multisensed plugin for Choreography (Rex Kerr) and divided by the average number of worms tracked during that time to get the number of reversals per worm in 3 minutes. For simplicity this value was divided by 3 and plotted as number of reversals per worm per minute.

Touch assays
Experiments quantifying omega turning, suppression of head oscillations and ventral turn head angle were conducted by gently touching an animal just posterior to the pharynx with a fine paint brush bristle taped to a glass pipette as previously described [47]. Animals were transferred to a thin lawn plate and allowed to acclimate for at least 5 minutes. 20 animals were assayed in each experiment and the population average was reported for each behavior. Animals were touched a single time and animals that failed to reverse were ignored. Animals that moved their head from side to side during a reversal were scored as defective in the suppression of head oscillations.

Omega turns
Omega turns were defined as any turn which results in a reorientation greater than 90˚from their initial forward bearing before the reversal. Turns that were greater than 90˚where the animal failed to touch its body with its head were considered open omega turns. Percent omega turns for a given experiment was calculated as

Ventral turn angle
To calculate the angle of the head during the initiation of a ventral turn animals were touched with an eyelash while video was being recorded through a dissecting microscope (SZ6TR1 Olympus) with a digital camera (AVT Pike F421-b, Allied Vision Technologies) using Fire-I image acquisition software (Unibrain inc.). Videos of animals that made omega turns were loaded into ImageJ / Fiji [81] and the first frame of forward movement following the touch initiated reversal was identified. To calculate the head deflection angle two lines were drawn, one running from the tip of the nose to the beginning of the intestine and the second running from the beginning of the intestine to a point on the midline 1 pharynx length further posterior. The angle between these two lines was measured using the angle tool in Fiji.

Confocal imaging
Animals were immobilized with 60mM sodium azide and placed on a 2% agarose pad under a coverslip. Images were acquired using a Zeiss LSM700 confocal microscope (Carl Zeiss AG)

Venus quantification
A population of animals expressing the Pflp-18::FLP-18::Venus transgene were subjected to a plate tap every 2 minutes for up to 2 hours. Individuals were removed at 0, 1 and 2 hours and immediately mounted and anesthetized as described above. Z-stacks were acquired with a confocal microscope (LSM 700, Carl Zeiss Microscopy GmbH) using a 40x objective (C-Apochromat 40x/1.20 W Korr, Carl Zeiss Microscopy GmbH). Images were loaded into ImageJ and the soma of the RIM and AVA were identified from their anatomical location. A circular region of interest (ROI) with a diameter of 13.5μm was placed around the soma or coelomocyte to measure the mean grey value of each neuron. An additional measurement was taken adjacent to the soma to quantify background fluorescence; this value was subtracted from the gray values of the RIM, AVA, and coelomocyte in that animal. The values obtained for each condition were normalized to the average intensity of the relevant neuron or coelomocyte in the control condition.

Calcium imaging
Young adult animals expressing a Pmyo-3::GCaMP6::SL2::mCherry transgene were placed into a drop of M9 buffer on a 2% agarose pad atop a microscope slide and a coverslip was gently placed onto the pad. Animals were able to move freely under the coverslip and were imaged on an inverted fluorescent microscope (Axio Observer A.1, Carl Zeiss Microscopy GmbH) using a Hammamatsu ORCA-Flash4.0 camera within 10 minutes of mounting. Images were analyzed using ImageJ software by drawing a 1075μm x 630μm rectangular ROI which could be rotated or translated to fit the position of an animal and the mean gray value was measured. The same ROI was placed in a position containing no animals to measure the mean grey value of the background which was then subtracted from the grey value of the ROI containing the animal. Wild type animals expressing the Pmyo-3::GCaMP6::SL2::mCherry transgene were imaged in parallel with mutants and the average background-subtracted grey value of the wild type animals for that day was used to normalize the values obtained from imaging the mutants. mCherry imaging was performed in a similar manner using the same rectangular ROI and the raw pixel values were quantified. S3 Video. Touch response in wild type. The escape response to anterior touch in wild type animals consists of a reversal during which head movements are suppressed followed by an omega turn and resumption of head movements and forward locomotion in the opposite direction. An eyelash is used to gently touch the animal behind the head to trigger the escape response. (MP4) S4 Video. Touch response in FLP-18(+++). The escape response in animals overexpressing FLP-18 results in deeper body bends and a failure to suppress exploratory head movements while reversing. An eyelash is used to gently touch the animal behind the head to trigger the escape response. (MP4)