Modulation of Locomotion and Reproduction by FLP Neuropeptides in the Nematode Caenorhabditis elegans

Neuropeptides function in animals to modulate most, if not all, complex behaviors. In invertebrates, neuropeptides can function as the primary neurotransmitter of a neuron, but more generally they co-localize with a small molecule neurotransmitter, as is commonly seen in vertebrates. Because a single neuron can express multiple neuropeptides and because neuropeptides can bind to multiple G protein-coupled receptors, neuropeptide actions increase the complexity by which the neural connectome can be activated or inhibited. Humans are estimated to have 90 plus neuropeptide genes; by contrast, nematodes, a relatively simple organism, have a slightly larger complement of neuropeptide genes. For instance, the nematode Caenorhabditis elegans has over 100 neuropeptide-encoding genes, of which at least 31 genes encode peptides of the FMRFamide family. To understand the function of this large FMRFamide peptide family, we isolated knockouts of different FMRFamide-encoding genes and generated transgenic animals in which the peptides are overexpressed. We assayed these animals on two basic behaviors: locomotion and reproduction. Modulating levels of different neuropeptides have strong as well as subtle effects on these behaviors. These data suggest that neuropeptides play critical roles in C. elegans to fine tune neural circuits controlling locomotion and reproduction.


Introduction
Neuropeptides are commonly used to modulate behaviors in both vertebrates and invertebrates. They can act synaptically within a neural circuit, as well as extra-synaptically to affect more distant neural circuits. These modes of action, in conjunction with the sheer number of neuropeptides, increase the diversity of behavioral outputs in an organism. Furthermore, a single neuropeptide gene may encode many peptides with similar or different amino acid sequences; these peptides may bind the same or multiple receptors with different affinities to 1(yn2) [11]; LGV: flp-21(pk1601, ok889), flp-6(pk1593); LGX: flp- 18

Isolation of deletion mutants
Libraries of mutagenized animals [32] were screened by polymerase chain reaction (PCR) with primers flanking the coding region of the different flp genes. Populations showing a deletion in the flp gene were sib selected until the flp deletion mutant was isolated. The flp-18(tm2197) mutant was generously donated by the Japanese National BioResource Project. Deletion mutants were backcrossed at least three times into a wild-type background to remove unlinked mutations.

PCR and generation of transgenic animals
PCR was used to confirm the genotype of different strains. The program used was: 94°C for 1 minute, followed by 35 cycles of 94°C for 40 sec, 58°C for 40 sec, and 72°C for 40 sec to 1 min. To generate transgenic lines, the genomic region for representative flp genes was amplified, the product gel purified, and the purified product co-injected with the transgenic markers sur-5:: GFP [33] and/or myo-2p::GFP [34]. The primer pairs used for amplification of genomic regions, sizes of promoter regions, co-injection markers, and concentrations of injected DNA are listed in Table 1. All of the 5' primers used to amplify genomic regions were the same as the primers used in the promoter constructs to determine flp gene expression patterns [22], except for flp-10, where an additional 67 bp was used in the promoter region for the genomic fragment. Extrachromosomal arrays in transgenic lines were integrated into the genome as described [35] with slight modifications.

Behavioral assays
For all behavioral assays, healthy (i.e., non-starved, non-dauered) fourth larval stage animals were plated and used the following day as day 1 adults, except for egg retention assays, when animals were used two days later as day 2 adults. At least 40 animals of each mutant strain and at least 30 animals for each transgenic strain were assayed over multiple trials. All assays were performed at room temperature. Because of the inherent variability among animals and in scoring, multiple, independent researchers scored mutants and the composite averages were tallied.
Swimming assay. Animals were placed into 50 μl of M9 physiological buffer and the number of body bends made in 15 seconds was counted (as modified from [36]). Each individual animal was tested three times and the three values were averaged to give a single mean value for the animal.
Serotonin-induced inhibition. The presence of serotonin inhibits locomotion [37]. Animals were placed into 12.9 mM serotonin in M9 physiological buffer and the number of body bends made in 15 seconds was counted. Each individual animal was tested three times and the three values were averaged to give a mean value for the animal.
Egg-laying rate. Animals placed into physiological buffer are transiently inhibited from egg laying. This inhibition is overridden by the presence of serotonin. Animals were placed into 50 μl of 12.9 mM serotonin in M9 physiological buffer [28] and the number of eggs laid after 60 minutes was counted.
Egg retention. Animals were placed into 50 μl of a weak hypochlorite solution (1.2 ml of Chlorox bleach, 0.5 ml of KOH, water to 10 ml) to dissolve the adult mother. The number of eggs, which are resistant to the hypochlorite solution, was counted.

Statistical analysis
All statistical analyses were performed using Prism software (GraphPad). Wild-type animals and mutants were compared using the Mann-Whitney test. Mutants and transgenic mutants containing a genomic transgene were compared as a group with a one-way ANOVA and Neuman Keuls posthoc test.

Results
At least 31 genes encode 72 potential FLP peptides [6,8,9,38]; by cDNA screening all of these genes are expressed [6,8]Wormbase) and 56 of the predicted peptides have been biochemically isolated [39,40,41,42,43,44,45,46,47,48,49]. These genes are expressed in neurons involved in multiple behaviors, including locomotion, reproduction, aggregation, and pharyngeal pumping [22]. However, how these FLP peptides modulate these neural circuits are still unclear. To examine the function of the different FLP peptides, we initiated a project to isolate mutants for each of the flp genes. We screened several mutagenized libraries for flp deletion mutants [32]. Because many of the flp genes encode multiple peptides (Fig 1), we reasoned that a deletion of the flp coding region was more likely to result in a null mutation. Eleven flp mutants were the first to be isolated: flp-3, 4, 6,8,9,10,12,18,19,20, and 21. The complete coding region for flp-3, 6, 8, 19, 20, and 21(ok889) were deleted in the isolated mutants; the peptide coding region was deleted in the flp-4, 9, and 12 mutants and the promoter region with peptide or non-peptide coding regions were deleted in the flp-1, 10, 18, and 21(pk1601) mutants (Fig 1). We also included daf-10 flp-1(yn2) in our assays; the yn2 deletion knocks out two genes, flp-1 and the neighboring daf-10, and will be referred to as flp-1 hereafter [11]. The 12 flp mutants were examined for locomotory and egg-laying behavior. Because a single flp gene can be expressed in multiple types of neurons (e.g., sensory and motoneurons or sensory and interneurons), it is difficult to categorize the flp genes into distinct groups. Among the flp genes being analyzed, a few, flp-4, 6, and 8, are expressed predominantly in sensory neurons, but multiple types of sensory neurons. The more common example is that flp genes are expressed in many types of sensory neurons as well as other types of neurons.  . The expression pattern of one gene among the isolated mutants, flp-9, has not been determined.

The role of FLPs in locomotion
Swimming initiates in C. elegans when animals are immersed in liquid. The body flexures result from the alternating waves of excitation/inhibition of the body wall muscles mediated by the excitatory cholinergic A and B motoneurons [23,26] and the inhibitory GABAergic D motoneurons [23,27]. Multiple sensory neurons synapse directly or indirectly onto the command interneurons PVC and AVB, which synapse onto the ventral cord motoneurons VB and DB for forward movement, and onto the command interneurons AVA, AVD, and AVE, which synapse onto the ventral cord motoneurons VA and DA for backward movement [23,51]. The A and B neurons synapse onto the ventral cord inhibitory motoneurons VD and DD to ensure proper body flexures [23]. Hence, locomotion is the integrated output of multiple sensory inputs. Because the flp genes are expressed in a variety of neurons, we expected that altering many flp genes would affect swimming rates. None of the flp genes examined are expressed in the ventral cord motoneurons. Two flp genes, flp-1 and 18, are expressed in the locomotory command interneurons.
To determine whether these swimming defects were due to loss of the corresponding flp gene, we microinjected the genomic region corresponding to representative genes (flp-3, 4, 10, 19, 20, and 21) into the respective mutants and assayed at least two independent transgenic mutant lines for swimming. flp-4 mutants showed a slightly decreased swimming rate, but in a transgenic mutant flp-4 line, a significant increase in the swimming rate was detected (28.81 ±0.43 body bends/15 sec, n = 98), suggesting that FLP-4 peptides potentiate swimming rate. flp-10 and flp-21(ok889) mutants showed an increased swimming rate and transgene expression of flp-10 and flp-21 caused a significant decrease in the swimming rate, suggesting that FLP-10 and FLP-21 peptides decrease swimming rates.
The presence of serotonin reduces locomotion in wild-type animals [37] through activity of the NSM and HSN neurons [37,57]. The swimming rate, for example, drops to 2.07 ± 0.23 body bends/15 sec (n = 669) in wild-type animals in the presence of serotonin ( Fig 2B,  Table 2). Several flp mutants were insensitive to varying degrees to this serotonin-induced inhibition. For instance, flp-8, 9, and 20 mutants showed slight, but significant resistance, whereas flp-1 and 19 mutants showed significant resistance to the serotonin-induced inhibition ( Fig 2B,  Table 2). This resistance to serotonin-induced inhibition in flp-19 mutants could be rescued by a flp-19 genomic fragment, suggesting that FLP-19 peptides do not affect swimming rate, but inhibit the action of serotonin on swimming. flp-20 mutants had only modest resistance to the serotonin-induced inhibition, and transgenic flp-20 mutants exhibited slight, but not significant serotonin-induced inhibition, suggesting that FLP-20 peptides have a modest effect on the serotonergic circuit.

Egg laying is highly sensitive to levels of different FLPs
The egg-laying motor circuit consists of the serotonergic/peptidergic (flp-19) HSN [22,58] and cholinergic/peptidergic VC motoneurons [26,29], both of which synapse onto the vulval muscles [23] whose contractions allow the release of eggs. Animals alternate between an active and inactive egg-laying state; this temporal switch is regulated by the HSN and VC neurons [59]. Egg-laying rate is sensitive to environmental conditions, such as food availability [28], hypertonic salt concentrations [37], and mechanical vibrations [60], and is coordinated with other behaviors, such as locomotion [28,57]. Hence, sensory input into the egg-laying circuit plays a significant role in determining egg-laying output. The mechanosensory neuron PLM and BDU  Table 1). *, p<0.05, **, p<0.01, ***, p<0.001 significantly different from wild type, Mann-Whitney test;^, p<0.05,^^, p<0.01,^^^, p<0.001 significantly different from mutant, one-way ANOVA, Neuman Keuls posthoc test.   interneuron [22] synapse directly onto the HSN neurons [23], and express the flp-20 and 10 genes, respectively. Because egg laying is affected by multiple sensory stimuli, we expected that loss of multiple flp genes would also indirectly affect egg laying. Wild-type animals were stimulated by serotonin to lay 5.77±0.18 eggs (n = 658) per hour ( Fig 3A; Table 2). We observed decreased egg-laying rates in most flp mutants. flp-19 is expressed in chemo-and aerosensory neurons as well the HSN motoneurons within the egglaying circuit. flp-19 mutants showed a significant decrease in egg-laying rate (3.89±0.46 eggs; n = 93), which was rescued by transgene expression in one line (Fig 3A; Table 2). Whether this egg-laying decrease is due to loss of FLP-19 peptides from HSN or sensory input is unclear. For other flp genes that synapse directly onto the egg-laying circuit, both flp-10 (2.16±0.20 eggs; n = 215) and 20 (0.17±0.05 eggs; n = 162) had significantly decreased egg-laying rates compared to wild type. The flp-10 transgene rescued the egg-laying defect to varying degrees, suggesting that FLP-10 peptides potentiates egg-laying rate. Surprisingly, flp-20 displayed an extremely low egg-laying rate, but presence of its transgene had only a slight, but significant, increase in the egg-laying rate (Fig 3A; Table 2). For flp genes expressed in sensory neurons, there was a variety of responses. flp-6 (4.04±0.36 eggs; n = 106) and 21(pk1601) (3.40±0.37 eggs; n = 112) mutants showed a significant decrease, which could not be rescued by the flp-21 transgene, while flp-8 mutants (7.43±0.60 eggs; n = 83) showed a significant increase in the egg-laying rate (Fig 3; Table 2). Loss of flp-1, which is primarily expressed in interneurons, also showed a significant decrease in egg-laying rate (2.82±0.44 eggs; n = 67). Loss of flp-4 had no effect on egg laying, but its overexpression caused a decrease in egg laying, suggesting that release of FLP-4 peptides from sensory circuits causes an indirect inhibitory effect on egg laying.

FLPs modulate the number of eggs retained in the uterus
The altered egg-laying rate in the flp mutants could be caused by a number of factors, including a decreased number of eggs generated, an incorrect sensing of the number of eggs in the uterus, or a failure of egg release due to muscle insensitivity to serotonin. We examined whether the altered egg-laying rates were due to a decreased pool of eggs available to be released or an increased retention of eggs in the uterus. We and others [61] have found that wild-type animals have 10-15 eggs in their uterus at any one time (two day adults: 13.28±0.29 eggs (n = 701); Fig  3B; Table 2).
Loss of flp-19 caused a decreased egg-laying rate and an increased egg retention rate (16.46 ±0.79 eggs; n = 138), indicating that the defective egg-laying rate in flp-19 mutants is not due to decreased egg production, but due to decreased egg release. However, transgene expression of flp-19 in mutants only exacerbated the egg retention defect (Fig 3B; Table 2), suggesting that FLP-19 peptides released from sensory or interneurons also affect reproduction. Loss of flp-20, which is expressed in neurons with direct synaptic input onto the egg-laying motor circuit, is similarly correlated. Among all the flp mutants examined, flp-20 showed the lowest egg-laying rate and the highest number of retained eggs (32.69±1.05 eggs; n = 240), which could be GFP and/or sur-5::GFP were used as the transgenic markers (see Table 1). *, p<0.05, **, p<0.01, ***, p<0.001 significantly different from wild type, Mann-Whitney test;^, p<0.05,^^, p<0.01,^^^, p<0.001 significantly different from mutant, one-way ANOVA, Neuman Keuls posthoc test. partially rescued by a flp-20 transgene. By contrast, flp-10, which is also expressed in a neuron with synaptic input into the egg-laying motor circuit, had a decreased egg-laying rate, and showed a decreased number of eggs retained in the uterus, suggesting that lack of egg production is responsible for the decreased egg laying rate; transgene expression of flp-10 in mutants caused an extremely high number of retained eggs in the uterus, indicating that FLP-10 peptides may influence activity of the sensor that determines the optimal number of eggs in the uterus or serotonergic input onto vulval muscles. Among the other flp mutants, the inverse correlation between egg-laying rate and number of eggs retained in the uterus was seen in flp-1, 4, 12, and 21 mutants, again suggesting that decreased egg production is not responsible for the decreased egg-laying rate; the increased egg retention was rescued with the transgene in flp-4 and 21 mutants. Like flp-10 mutants, flp-6 mutants showed a decreased egg-laying rate and a decreased number of eggs in the uterus (7.17±0.29 eggs; n = 174) ( Table 2), suggesting that egg production may factor into the decreased egg-laying rate. These results reinforce our findings that multiple FLP peptides modulate the output of the egg-laying circuit.

Discussion
Environmental stimuli are integrated through multiple sensory modalities to determine behavior. Hence, we expected that disruption of any neural circuit would ultimately have an effect on most behavioral outputs, such as locomotion and reproduction. In this paper, we explored the effects of FLP neuropeptides released from different neural circuits [23,62]. Because neuropeptides are generally co-localized with a small molecule transmitter [10,22] and have modulatory effects, we predicted that modulating levels of many of the FLP peptides would affect locomotion and reproduction and found this prediction to be correct. Although FLP peptides are expressed in the locomotory motoneurons, none of the flp genes examined are expressed in the locomotory motoneurons. Five of the 11 flp mutants showed defective swimming rates, which could be at least partially rescued by re-introduction of the respective flp gene (Fig 2). The expression patterns of the affected flp genes were not specific for certain types of neurons. Instead, the affected mutants included genes expressed in interneurons (flp-1 and 18) (36,39) and sensory neurons (flp-4, 10, and 21). flp-8, 10, 12, and 19 are expressed in the oxygen sensor neurons BAG and/or URX; however, only loss of flp-10 affected swimming rates. Similarly, flp-4, 6, 10, and 20 are expressed in the chemosensory ASE neuron, but only flp-4 and 10 knockouts showed modulated swimming rates (Table 2). flp-4 is also expressed in one mechanosensory neuron, but other flp genes expressed in mechanosensory neurons (flp-8, 12, and 20) showed no change in swimming rate. Because flp-4, 10, and 21 are expressed in so many different types of sensory neurons, it is difficult to ascribe their effects on swimming to a particular circuit. However, regulation of neuropeptide levels must be under very tight homeostatic control, because loss of many flp genes affect swimming rates and, conversely, transgene overexpression of flp genes also affect swimming rate. For instance, flp-4 overexpression increased swimming rate, while flp-10 and 21 overexpression decreased swimming rate, supporting the indirect role of these peptides in modulating swimming behavior. Among the characterized FLP peptides, many bind multiple receptors and the receptors through which these peptides signal show widespread expression. For instance, FLP-18 and 21 peptides bind at nM affinities to NPR-1, 3, 4, 5, 6, 10, and 11 [13,14,16]; for the NPR receptors that have been characterized, NPR-1 and 4 show widespread expression [16,63], making it difficult to draw conclusions as to which downstream neurons and circuits are being activated by the FLPs to affect locomotion.
The transgenic lines often gave a wide variety of responses, which we attribute to variation due to the mosaic nature of transgene expression. Unlike other organisms where the transgene is inserted into the genome, C. elegans transgenes are present as extrachromosomal arrays, which can be lost at any cell division. Hence, even within one non-integrated C. elegans transgenic line responses are variable, so multiple lines were analyzed. In addition, varying lengths of the promoter and 3'UTR regions (Table 1) for the different genomic fragments were used; longer lengths of these regions may contribute to better stability of the transcript, which could have resulted in better and/or less variable responses.
Within the reproductive motor circuit, none of the examined flp genes are expressed in the VC motoneurons or AVF command interneuron [22], but flp-19 is expressed in the HSN motoneurons, the critical neuron regulating the active egg-laying state [59]. Not surprisingly, modulating levels of FLP-19 peptides affected egg-laying rates. Loss of flp-19 decreased egg-laying rates, which could be rescued by re-introduction of FLP-19 peptides, suggesting that FLP-19 peptides released from HSN potentiates egg laying. No FLP-19 receptor has been identified thus far. Similarly, flp-10 is expressed in the BDU interneuron that synapses onto HSN, and again, not surprisingly, decreasing and increasing levels of FLP-10 peptide decreased or increased the egg-laying rate, respectively. The single FLP-10 peptide signals through the EGL-6 receptor, which is expressed by HSN neurons [15], suggesting that FLP-10 promotes egg laying by activating HSN. These results are in contrast to those of Ringstad and Horvitz [15], who reported that overexpression of flp-10 inhibited egg laying, while knockout of egl-6 had no effect on egg laying, as scored by the stage of the egg laid as opposed to egg-laying rate; in addition, these authors did not report flp-10 expression in BDU neurons with their expression vector, which may account for our differing results. Ringstad and Horvitz [15] also reported that flp-17 activity within the oxygen sensor BAG neuron inhibited egg laying. In addition to flp-10 and flp-17, flp-12 and 19 are expressed in the BAG neuron. Like flp-10 and 19, loss of flp-12 also decreased the egg-laying rate. Because flp-10 and 19 are expressed in multiple neurons within the egg-laying neural circuits, it is unclear from which neurons the peptides are exerting their effects on egg laying. Among other flp genes expressed in sensory neurons, loss of flp-6, 20, and 21 decreased the egg-laying rate, while flp-8 knockouts, as well as knockout of flp-9, whose expression pattern is unknown, showed an increased egg-laying rate; transgenic copies of flp-20 and 21 in the mutants caused the converse phenotype, indicating that, as previously reported [61], multiple sensory stimuli affect egg-laying rate. RNAi knockdown of FLP-21 receptors NPR-3, 6, and 11 also decreased brood size [64], suggesting that FLP-21 signals through these receptors to modulate egg laying. The receptors through which FLP- 6,9,12,19, and 20 peptides signal have not been identified.
We hypothesized that if there was a problem with egg release, egg-laying rates would be inversely correlated to egg retention in the uterus. Specifically, we expected that animals that displayed high egg-laying rates would have low numbers of eggs in the uterus and, conversely, animals with low egg-laying rates would have high numbers of eggs retained in the uterus. This inverse correlation was seen among several mutants. flp-1, 12, 19, 20, and 21 mutants showed low egg-laying rates and high numbers of eggs in the uterus, suggesting that eggs were being generated, but were not being released, perhaps because the neurons were not signaling to the muscles properly or the muscles were not responsive to serotonin. Conversely, flp-9 mutants showed an increased egg-laying rate and a decreased number of eggs retained, suggesting that eggs were laid as soon as they were generated and flp-9 is involved in the timing of egg release. There were also mutants, such as flp-4, whose egg-laying rate was unaffected, yet it retained a significant number of animals in its uterus; this egg retention defect was rescued with a wild-type copy of flp-4 ( Table 2) and may indicate a disruption of a sensor that regulates the optimal number of eggs in the uterus. Several mutants showed a positive correlation in egg-laying rate and egg retention. For instance, flp-3, 6, 10, and 19 mutants showed significant decreases in the egg-laying rate as well as a decreased number of eggs retained, suggesting that there is a disruption in egg formation. Hence, the FLP peptides affect reproductive behavior in a multitude of ways.

Conclusions
We examined the effects of loss of FLP neuropeptides on locomotory and reproductive behavior. Because environmental stimuli activate multiple sensory systems that are integrated to determine locomotory and reproductive behavior, we have found that any perturbation of sensory or motor systems, such as by altering the levels of neuropeptide signaling, will modulate these behaviors. Furthermore, we have not explored, because of technical difficulties, whether neuropeptides exert their effects extrasynaptically, such that neuropeptides released within the nerve ring may affect neural circuits for which they do not have a direct synaptic connection. Our work has shown the interconnections between all neural circuits in the control of critical behaviors in an organism. Nematodes have the largest known family of FLPs in the animal kingdom [65]. The role of each specific FLP peptide in locomotion and reproduction awaits further studies.