Inferences of Glia-Mediated Control in Caenorhabditis elegans

Abstract

Astrocytes modulate synaptic transmission; yet it remains unclear how glia influence complex behaviors. Here, we explore the effects of C. elegans astrocyte-like cephalic glia (CEP_glia) and the glia-specific bHLH transcription factor HLH-17 on mating behavior and the defecation motor program (DMP). In C. elegans, male mating has been explicitly described through the male tail circuit and is characterized by coordination of multiple independent behaviors to ensure that copulation is achieved. Furthermore, the sex-specific male mating circuitry shares similar components with the DMP, which is complex and rhythmic, and requires a fixed sequence of behaviors to be activated periodically. We found that loss of CEP_glia reduced persistence in executing mating behaviors and hindered copulation, while males that lacked HLH-17 demonstrated repetitive prodding behavior that increased the time spent in mating but did not hinder copulation. During the DMP, we found that posterior body wall contractions (pBocs) and enteric muscle contractions (EMCs) were differentially affected by loss of HLH-17 or CEP_glia in males and hermaphrodites. pBocs and EMCs required HLH-17 activity in both sexes, whereas loss of CEP_glia alone did not affect DMP in males. Our data suggest that CEP_glia mediate complex behaviors by signaling to the GABAergic DVB neuron, and that HLH-17 activity influences those discrete steps within those behaviors. Collectively, these data provide evidence of glia as a link in cooperative regulation of complex and rhythmic behavior that, in C. elegans links circuitry in the head and the tail.

Graphical abstract.

In this study we show that the invertebrate CEP_glia are important regulators of complex behaviors. C. elegans CEP_glia are similar to astrocytes and assist in temporally regulating bidirectional communication with neurons and non-neuronal cells. Thus, CEP_glia mediated signaling facilitates the fluent transition between independent behaviors in a complex motor program.

1. Introduction

Globally, approximately 20% of individuals have been diagnosed with mood disorders (Sampedro-Piquero & Moreno-Fernandez, 2019; Depression Statistics, 2020), and astrocytes appear to mediate neuronal dysfunction (Harada et al., 2016; Konopaske et al., 2008; Moraga-Amaro et al., 2014; Oliveira et al., 2015; Quesseveur et al., 2013; Yamamuro et al., 2015). Patients with major depressive disorder (MDD) have fewer astrocytes in the brain (Rajkowska & Stockmeier, 2013; Sanacora & Banasr, 2013), and astrocytes in MDD patients have an altered morphology that is associated with reduced functional vigor (Wang et al., 2017). Astrocytes maintain homeostasis in the brain and modulate synapse formation by mediating neurotransmitter and gliotransmitter release (Eroglu and Barres, 2010; Li et al., 2013; Mazaud et al., 2019) and by responding to signals from neighboring cells (Araque et al., 2014; Araque et al., 2002; Bang et al., 2016; Ishibashi et al., 2019; Jourdain et al., 2007; Khan et al., 2001; Ota et al., 2013; Schousboe, 2019; Volterra & Meldolesi, 2005). Astrocytes are thought to modulate synaptogenesis through intercellular communication at tripartite synapses (Halassa et al., 2009; Machado-Vieira et al., 2009; Strauss et al., 2015; Swanson et al., 1999) and have recently been recognized as regulators of behavior (Christensen et al., 2013; Jackson, 2011; Emery & Freeman, 2007; Jackson et al., 2015; Oliveira et al., 2015; Tso et al., 2017) that likely interact at synapses to maintain structure and function. It is still unclear how bidirectional communication between glia and neurons is aligned with coordinating complex and rhythmic behavior circuits, and, currently, studies aim to understand mechanisms by which astrocytes regulate diverse signals and how these regulatory mechanisms coordinate behavior in the proper context (Mederos and Perea, 2019).

C. elegans have four main glia types that are distinguished by morphology, gene expression, and the neurons with which they are associated (Procko et al., 2012; Wallace et al., 2016). Here, we utilize cephalic glia (CEPglia), which are considered morphological and functional homologs of vertebrate astrocytes (Frakes et al., 2020; Oikonomou & Shaham, 2011; Heiman & Shaham, 2007, 2009) and are involved in the regulation of synapse formation and function (Bacaj et al., 2008; Colón-Ramos et al., 2007b; Gibson et al., 2018; Hardaway et al., 2015; Meng et al., 2016; Procko et al., 2011; Rapti et al., 2017; Seifert et al., 2006; Shaham, 2005, 2006; Shao et al., 2013a; Stout et al., 2014; Wallace et al., 2016). CEPglia are bipolar cells with an anterior process that ensheaths four pairs of dopaminergic neurons, and one posterior process that ensheaths the central ganglion neurophile, known as the nerve ring (White et al., 1986). Loss of this glia subtype disrupts the organization of the nerve ring and disorients neurons that subsequently fail to innervate postsynaptic targets (Sulston et al., 1983; Colón-Ramos et al., 2007a; Yoshimura et al, 2008; Shao et al., 2013b; Cook et al., 2019).

In C. elegans, male mating is a complex behavior that is characterized by accurate execution and coordination of many behaviors to ensure copulation (Barr, 2006; Sherlekar & Lints, 2014). Likewise, the defecation motor program (DMP) is a rhythmic and complex behavior in both males and hermaphrodites that requires a sequence of stereotypical behaviors (Branicky & Hekimi, 2006; Thomas, 1990). The synapse architectures that are required for proper execution of the DMP in males and hermaphrodites (Reiner & Thomas, 1995) differ, and in males, the circuitries for the DMP and the male mating are coupled (LeBoeuf & Garcia, 2017; Nagy et al., 2015). Anatomical features in hermaphrodites that promote enteric muscle contractions (EMC) during the DMP are rearranged in males to modulate the time in which the copulatory spicules are protracted. It remains to be determined how these anatomical features facilitate stimulation of neuronal and non-neuronal cells to transmit signals in the proper behavioral context.

Here we show that CEP_glia are needed for proper execution of behaviors in the copulation stage of mating; loss of CEP_glia reduces males’ persistence and motivation to surmount the difficulty of transitioning from prodding to spicule protraction and sperm transfer. We show that loss of CEP_glia causes males to scan beyond the end of the hermaphrodites, losing contact and missing the time in which to execute a ventral turn. In our assessment of independent behaviors in the DMP, we found that loss of CEP_glia results in erratic spicule protraction during mating and after enteric muscle contraction and expulsion in the DMP. Together, our work uncovers possible roles for astrocytes in regulating complex and rhythmic behaviors, in part by modulating inhibitory and excitatory signaling at GABAergic synapses.

2. Materials and Methods

2.1. Strains and maintenance

Strains were acquired from the Caenorhabditis Genetics Center (CGC) at the University of Minnesota (https://cgc.umn.edu). They were maintained at 20°C on solid nematode growth media (NGM) seeded with E. coli OP50 (Lewis and Fleming, 1995; Brenner, 1974).Genetically ablated strains were derived from DCR1337 [nsIs105 (Phlh-17::GFP) I; cima-1(wy84) IV; wyls45(Pttx-3::GFP::rab-3, Punc-122::RFP)X; olaEx805 (Phlh-17::caspase12; Phlh-17::caspase17; Pttx-3::mCherry; Pglr-3::mCherry and Punc-122::GFP)] (Shao et al, 2013) using traditional crossing techniques, including genotypic/phenotypic confirmation by florescence microscopy and PCR. Strains for this study include: DR466 [him-5(e1490) V], OS2649 [hlh-17(ns204) IV]; CMJ4006 [hlh-17(ns204) IV; him-5(e1490)]; CMJ4007 [nsIs105 (phlh-17::GFP) I; him-5 (e1490) V]; CMJ4008 [nsIs105 (phlh-17::GFP) I; hlh-17(ns204) IV; him-5 (e1490) V]; CMJ4009 [nsIs105 (phlh-17::GFP) I; him-5 (1490) V; olaEx805 (Phlh-17::caspase12; Phlh-17::caspase17; Pttx-3::mCherry; Pglr-3::mCherry and Punc-122::GFP)]; CMJ4010 [nsIs105 (phlh-17::GFP) I; hlh-17(ns204) IV; him-5 (1490) V; olaEx805 (Phlh-17::caspase12; Phlh-17::caspase17; Pttx-3::mCherry; Pglr-3::mCherry and Punc-122::GFP)]; CB246 [unc-64(e246)], COP1997 [kNUsI815[pNU1972(F16F9.3p::hlh—17::ttb-2 in ttTi56050]II; unc-119(ed3)II] generated by NemaMetrix, Inc. (https://invivobiosystems.com), CMJ4011 [[nsIs105 (phlh-17::GFP) I; him-5 (1490) V]; [kNUsI815[pNU1972(F16F9.3p::hlh—17::ttb-2 in ttTi56050]II; unc-119(ed3)II]], and CMJ4012[[nsIs105 (phlh-17::GFP) I; hlh-17(ns204) IV; him-5(e1490) V]; [kNUsI815[pNU1972(F16F9.3p::hlh—17::ttb-2 in ttTi56050]II; unc-119(ed3)II]]. We examined the behaviors of males derived from DR466, a him-5 (e1490) background, with intact or ablated (ΔCEP_glia) CEP_glia and that either expressed or lacked (hlh-17) HLH-17. In our assays, him-5 males behaved the same as N2 males (see Figure S1a–d); therefore, we hereafter consider the contribution of the him-5 allele to the described behaviors as negligible, and consider him-5 males with intact CEP_glia and that produce full-length HLH-17 to be phenotypically WT. Specifically, we refer to the phenotypes of strains as follows: CMJ4007 = WT; CMJ4008 = hlh-17(ns204); CMJ4009 = ΔCEP_glia; CMJ4010 = ΔCEP_glia; hlh-17(ns204); CMJ4011 = pAM::hlh-17; and CMJ4012 = hlh-17(ns204); pAM::hlh-17.

The following primers were synthesized by a commercial laboratory (www.idtdna.com) and were used for genotyping:

• hlh-17: 5’ TCC CTG GGG ACT CTC CTC G 3’ and 5’ CGA TTT TTG CTG CTA ATG GGC AAC AC 3’; him-5: 5’ GAC GAT CAC TGT TGA CAA TCA C 3’ and 5’ GTC CAG AAT TCG TTC TAA TAA CG 3’; cima-1: 5’ GAA AAG GAC CAG CCT GTA ATG 3’ and 5’ GAG ATG TTC TAG AAT TCG CAC C 3’.

2.2. Behavioral Assays

Mating

One well-fed adult male was added to a seeded, 60 mm NGM agar plate that contained one drop of an OP50 bacterial lawn that was less than the circumference of a dime (~27 mm). Males were placed with ten partially paralyzed unc-64(e246) early (< = 2-day old) adult hermaphrodites (Correa et al., 2012). Mating events were recorded using AVI acquisition fast quality on a Nikon eclipse 90i digital microscope. Each assay was limited to a 30-minute duration; failure to respond to the hermaphrodite or complete copulation within this period was considered an unsuccessful mating attempt. The assay started when the male pressed his tail onto the hermaphrodite body wall. The assay stopped after a male transferred sperm, swam away, or when the time exceeded 30 minutes. The mating behaviors were assayed in two stages, response stage and copulation stage.

Response Stage: In this stage we assessed the male’s ability to respond to a hermaphrodite and to initiate a vulva search. 1) Contact: We recorded the total time in which the male tail remained in contact with the hermaphrodite, until locating the vulva. Time was stopped if the male lost contact with the hermaphrodite before locating the vulva. 2) Turning: Males that initiate contact on the dorsal side of the body wall must execute turns to the ventral side, while maintaining contact. To assess turning behavior, we calculated the percentage of completed turns divided by the total number of turn attempts, which included completed and missed turns. The quality of turns included: a completed turn, which occurs when the male tail maintains contact or temporarily loses contact but quickly regains it while turning to the opposite side of the hermaphrodite; and a missed turn, which occurs when the male overshoots the turn and completely loses contact with the hermaphrodite. 3)Vulva Location: we assayed ability to locate the vulva based on the number of times a male’s tail scanned beyond the vulva without stopping and executing prodding behavior. Copulation stage: In this stage we assessed the male’s ability to insert his spicules into the vulva by prodding and to transfer sperm. 1) Prodding: We measured prodding behavior as the accumulated time of each prodding attempt, starting from contacting the vulva until the initiation of sperm transfer, until males gave up and swam away, or until 30 minutes elapsed. 2) Transfer: Copulation was considered successful only after we visually detected protraction of the tail spicules and transfer of sperm into the vulva.

Defecation

Defecation behaviors were visualized at 20X magnification using a Nikon eclipse 90i digital microscope. One well-fed adult male or hermaphrodite, grown at 20°C, was placed on a seeded 60 mm NGM plate and allowed to recover for five minutes prior to starting the assay. Time started at the first visually detected posterior body wall contraction and continued for 10 minutes. This period spanned a time that is equivalent to a minimum of 10 cycles of the defecation motor program. One cycle of motor behaviors is initiated by a posterior body wall contraction (pBoc) which is a change in body length at the posterior end of the animal and is visualized as a shrinkage of the tail. Following a pBoc, the anterior body wall muscles contract (aBoc), resulting in a change in body length at the anterior end of the animal and a shrinkage at the head. A similar head movement is depicted during foraging; therefore, aBocs were not assayed. A defecation cycle is complete after enteric muscles contract, which can be visualized as shrinkage at the tip of the tail and quickly followed by expelled gut contents. We assessed periodicity between cycles by calculating the average time between two consecutive pBocs (Mahoney et al., 2008). We scored the average EMCs and pBocs as the number of EMCs/10 and the number of pBocs/10, respectively.

2.3. Data Analysis

We used GraphPad Prism 9 for all statistical analysis, which included generating standard descriptive statistics for each strain, identifying outliers using the ROUT method, and comparing multiple groups via one-way ANOVA followed by Dunnett or Holm–Šidák posthoc analysis, as indicated,to identify statistical differences between data sets. The critical threshold for statistical significance via ANOVA was set as < 0.05, and adjusted P-values based on Dunnet or Holm–Šidák corrections are indicated in the figure legends. Outliers were omitted from the graphs and were not included in the statistical analysis. The number of outliers omitted from each data set is indicated as a number above the appropriate data set.

3. Results

3.1. Mating behavior is elusive in males that lack CEP_glia

C. elegans males perform a mating ritual that occurs in two stages, each of which consists of a series of key behaviors to facilitate fluent and effortless copulation (Barr and Garcia, 2006). The goal of the first stage, referred to here as the response stage, is to execute the vulva search, during which the male contacts and scans the full length of the hermaphrodite, executes a series of turns, and locates the vulva (Figure SM1). The second stage is referred to as the copulation stage. It involves rhythmic prodding at the vulva, spicule protraction to pry open the vulva, and the transfer of sperm (Figure SM2).

Our previous, unpublished microarray data suggest that the glia-specific transcription factor HLH-17 regulates the transcription of genes required for male mating behavior, including lov-1, pkd-2, and flp-10 (Felton, 2014). The transmembrane receptor LOV-1 and polycystin-2 TRP channel, PKD-2 are required for locating the vulva (Penden and Barr, 2005; Barr and Garcia, 2006). FLP-10 is an FMRFamide-like neuropeptide that affects turning behavior; flp-10 males prematurely initiate sharp ventral turns, stop before reaching the ventral side of the hermaphrodite, and then repeat the turn-stop behavior approximately four more times before finally transitioning to the next step (Liu et al., 2007; Peymen et al., 2014). Because HLH-17 mRNA and protein are expressed in the CEP_glia, we hypothesized that CEP_glia are necessary for mating behavior. To examine the role of CEP_glia in modulating mating behavior, and to determine whether that role is dependent on HLH-17 expression, we examined the behaviors of males derived from a him-5 (e1490) background with either intact or genetically ablated (ΔCEP_glia) CEP_glia and that either expressed or lacked (hlh-17) HLH-17. Genetic ablation of the CEP_glia was achieved by driving cell-specific expression of apoptotic caspases under control of the hlh-17 promoter (see Materials and Methods; Chelur and Chalfie, 2008). Expression of the extrachromosomal apoptotic caspase array was visually confirmed by the absence of anterior GFP reporter expression in the proximity of where CEP_glia are normally localized and by the presence of GFP reporter expression in coelomocytes along the dorsal and ventral body walls. Surprisingly, CEP_glia ablation appeared to affect the viability of C. elegans males, which restricted our ability to assay large sample sizes; even in the presence of the him-5 allele which typically produces a 25%−30% male population, fewer than 15% of the population of CEP_glia ablated animals in this study were male, and of those, fewer than 50% survived to adulthood.

In our behavior assays, him-5 males are phenotypically identical to N2 males used as the standard for WT (Figure S1a–d; S1a, F (3,25) = 1.65, p = 0.204; S1b, F (3,26) = 1.576, p = 0.2191; S1c, F (3,21) = 2.280, p = 0.1089; S1d, F (3,24) = 2.975, p = 0.0517); therefore, we refer hereafter to him-5 males with intact CEP_glia and that produce full-length HLH-17 as WT (see Materials and Methods). We found that, overall, males that lack CEP_glia spend less time in the act of mating than WT (11.04 +/− 7.38 minutes, n = 15) (Figure 1a, F (3,37) = 4.4396, p = 0.0096). This generalization was true whether or not the mating attempt was successful and whether (Δ CEP_glia = 9.06 +/−4.71 minutes, n = 5, p = 0.4879 by ANOVA Holm–Šidak post hoc test) or not the ΔCEP_glia males expressed hlh-17 (ΔCEP_glia ; hlh-17(ns204) = 4.84 minutes +/−1.09, n = 6, p = 0.1487 by ANOVA Holm–Šidak post hoc test). In contrast, hlh-17(ns204) males with intact CEP_glia spent more time in the act of mating (17.31 +/−7.18 minutes, n = 14, p = 0.2979 ANOVA Holm–Šidak post hoc test) than WT males, and significantly more time than males that lacked both HLH-17 and the CEP_glia (p = 0.0081 ANOVA Holm–Šidak post hoc test). Loss of CEP_glia also correlated with a reduction in the percentage of ΔCEP_glia males that completed the mating ritual and subsequently copulated with the hermaphrodite: 40% ΔCEP_glia and 40% ΔCEP_glia; hlh-17(ns204), versus 67% WT and 70% hlh-17(ns204) (Figure S3a). We found it noteworthy that loss of hlh-17 alone caused individual males to spend more time in the act of mating (Figure 1a), which corresponded to an increase in the percentage of males that completed the entire behavior to eventually transfer their sperm during copulation (Figure S3a). Taken together, these data led us to question whether a specific step in mating acts as a bottleneck. If such a bottleneck exists, we expected males that persist in overcoming this step would successfully mate, as seen in hlh-17(ns204) animals, and those that do not persist to overcome the bottleneck, would give up and swim away, as seen in ΔCEP_glia animals.

Figure 1:

Mating behavior is affected by loss of CEP_glia and hlh-17 expression. Males were assessed for (a) the total mating time [response and copulation stage]. WT [nsIs105; him-5(e1490)] n = 15, hlh-17(ns204) [nsIs105; hlh-17(ns204); him-5(e1490)⁻] n = 15 , ΔCEP_glia [nsIs105; him-5(e1490); olaEx805] n = 5 , and ΔCEP_glia; hlh-17(ns204) [nsIs105; hlh-17(ns204); him-5(e1490); olaEx805] n = 6. (b) Detection: the number of times males scanned passed the vulva without stopping or pausing and (c) Turning: the percentage of dorsal to ventral turns completed by the male divided by total turn attempted. WT [nsIs105; him-5(e1490)] n = 15, hlh-17(ns204) [nsIs105; hlh-17(ns204); him-5(e1490)⁻] n = 15, ΔCEP_glia [nsIs105; him-5(e1490); olaEx805] n = 5 , and ΔCEP_glia; hlh-17(ns204) [nsIs105; hlh-17(ns204); him-5(e1490); olaEx805] n = 5. (d) Prodding: the time spent prodding until mating or swimming away after 30 minutes. WT [nsIs105; him-5(e1490)] n = 15, hlh-17(ns204) [nsIs105; hlh-17(ns204); him-5(e1490)⁻] n = 15 , ΔCEP_glia [nsIs105; him-5(e1490); olaEx805] n = 5 , and ΔCEP_glia; hlh-17(ns204) [nsIs105; hlh-17(ns204); him-5(e1490); olaEx805] n = 5. The number above a given data set represents the number of outliers. Symbols represent individual males in each assay. For a, c, and d, open symbols represent males that initated contact but did not mate. Closed sysmbols represent males that mated. Lines and bars represent the mean and standard deviation. Astericks represent adjusted P-values after Holm–Šidák posthoc correction: *p<0.05, **p<0.005, and ***p<0.001.

3.2. Turning behavior is affected by loss of CEP_glia

To identify the step of the mating process that is potentially acting as a bottleneck in hlh-17(ns204) and ΔCEP_glia males, we separately analyzed each step of the response and copulation stages of male mating behavior. Based on the influence of HLH-17 on lov-1, pkd-2, and flp-10 expression, we expected these bottlenecks to occur during the response stage, either at the turning step or at the vulva location step.

In the response stage, males initiate the vulva search by using their tails to press against and scan along the body of the hermaphrodite. Males turn from the dorsal body wall of the hermaphrodite by executing sharp turns that propel them towards the ventral body wall, which allows them to continuously scan towards the vulva. Once the vulva is located, males proceed to the copulation stage. We found that the time to complete the steps in the response stage, from initial contact to locating the vulva, was not significantly different across populations (Figure S2, F (3,35) = 1.703, p = 0.183 ANOVA). However, when compared to WT (16.32 +/− 12.10s, n = 14), the total time for the response stage was somewhat shorter in ΔCEP_glia males, whether they expressed (9.94 +/− 5.43s, n = 4, p = 0.5470 ANOVA Holm–Šidak post hoc test) or lacked HLH-17 (9.08 +/− 1.08s, n = 6, p = 0.3759 ANOVA Holm–Šidak post hoc test). We found that the ability to locate the vulva differed significantly among the animals tested (Figure 1b, F (3,32) = 5.8283, p = 0.0027 ANOVA Holm–Šidak post hoc test); in particular, while loss of HLH-17 alone had no affect on the vulva seach (p > 0.9999 ANOVA Holm–Šidak post hoc test), loss of CEP_glia resulted in at least one vulva pass for 40% of both CEP_glia ( p 0.4198 ANOVA Holm–Šidak post hoc test, vs WT or vs hlh-17(ns204)) and ΔCEP_glia; hlh-17(ns204) males (p = 0.0039 ANOVA Holm–Šidak post hoc test, vs WT or vs hlh-17(ns204)).

To assess turning behavior, we calculated the number of turns completed by each animal, as well as the number of animals per strain that were able to execute completed turns (Figure 1c, F (3,34) = 8.466, p = 0.0002, ANOVA). We found that the percentage of completed turns decreased with the loss of CEP_glia. That is, while every WT ( n = 15) and every hlh-17(ns204) (n = 15) male completed 100% of their turns without losing contact with the hermaphrodite, ΔCEP_glia and ΔCEP_glia; hlh-17(ns204) only completed 80% (n = 5, p = 0.0181 ANOVA Holm–Šidak post hoc test) and 26% (n = 5, p = 0.0019 ANOVA Holm–Šidak post hoc test) of their turns, respectively. While loss of hlh-17 alone had no effect on turning behaviors (that is, 0% of hlh-17(ns204) males executed at least one missed turn), 40% ΔCEP_glia and 17% ΔCEP_glia; hlh-17(ns204) males executed at least one missed turn (Figure S3b).

Turning behavior alone was not a predictor of mating success (i.e., successful transfer of sperm during the copulation stage). As shown in Figure 1c, two of five ΔCEP_glia males completed between 30–40% of their turns but never mated. Of the remaining three males that completed all turns, only two eventually mated. Likewise, five of six ΔCEP_glia; hlh-17(ns204) males completed an average of 63.2% of their turns, with one male completing all turns without ever successfully mating. Further, two of those five males that experienced trouble while executing turns were still able to mate successfully (Figure 1c). This finding was interesting for two reasons. First, turning does not appear to be the bottleneck faced by ΔCEP_glia males, since the act of persisting beyond the turning step does not ensure mating success. Turning is also clearly not the bottleneck faced by hlh-17(ns204) males since this step is not affected in these animals. Second, this finding supports that of Loer and Kenyon (1993) which demonstrated that animals lacking dopamine signaling execute “sloppy” or “missed” turns, resulting in loss of contact with the hermaphrodite and failure to copulate. We previously demonstrated that hlh-17(ns204) animals are inefficient at executing dopamine-dependent behaviors due to abnormal dopamine signaling (Felton & Johnson, 2011, 2014), and, importantly, the dopaminergic neurons in C. elegans are ensheathed by the CEP_glia (Sulston et al., 1975; White et al., 1986). Together these data highlight the role of CEP_glia in the execution of turns during mating, while pointing to the potential for HLH-17 activity to influence the dopamine-dependent aspects of CEP_glia function, in part through its influence on the transcriptional activity of flp-10 and genes required for dopamine signaling.

3.3. CEP_glia regulates the copulation stage of mating in an HLH-17 dependent manner

The second stage of mating behavior, the copulation stage, consists of prodding, spicule insertion, and sperm transfer (Figure SM2). As shown in Figure 1a, our data when examining mating as a single comprehensive behavior shows that although a high percentage of WT and hlh-17(ns204) males completed the copulation stage to mate successfully, these males took longer than ΔCEP_glia or ΔCEP_glia; hlh-17(ns204) males to do so. However, we also show that ΔCEP_glia and ΔCEP_glia; hlh-17(ns204) males experience difficulty executing behaviors leading to the prodding step (Figures 1b and 1c). These data suggest that the increased number of failed attempts to locate the vulva decreases the probability that ΔCEP_glia males will proceed to the prodding step, thereby reducing the likelihood of a successful mating outcome. These data, however, do not identify the predicted bottleneck that appears to be the determining factor in whether or not mating is successful for ΔCEP_glia males, nor does it explain why ΔCEP_glia males appear to mate faster than animals with intact CEP_glia.

To address these questions, we first determined the percentage of males to exhibit prodding behavior after locating the vulva and then measured the total time that those males spent prodding. We found that 93% of WT (n = 15), 100% of hlh-17(ns204) (n = 15), 80% of ΔCEP_glia (n = 5) and 60% ΔCEP_glia; hlh-17(ns204) (n = 5) males exhibited prodding behavior after locating the vulva (Figure S3c). While greater than 90% of WT and hlh-17(ns204) prodded, only 69% WT and 67% hlh-17(ns204) males mated successfully (Figure S3c vs Figure 1d). Likewise, 80% ΔCEP_glia , and 60% ΔCEP_glia; hlh-17(ns204) males executed prodding behavior (Figure S3c); of the males that prodded, two of the four (50%) ΔCEP_glia males and two of the of three (67%) ΔCEP_glia; hlh-17(ns204) males subsequently mated (open versus closed circles in Figure 1d). Second, we measured the time that males spent in the act of prodding (Figure 1d, F (3,36) = 2.230, p = 0.104 ANOVA). We found that WT males prodded 179.89 +/−168.16 s (n = 15). hlh-17(ns204) (266.30 +/−223.19 s, n = 15, p = 0.5457 ANOVA Holm–Šidak post hoc test) and ΔCEP_glia males (198.26 +/− 203.14 s, n = 5, p = 0.8559 ANOVA Holm–Šidak post hoc test) prodded slightly longer than WT males, while ΔCEP_glia; hlh-17(ns204) males spent noticeably, but not significantly, less time prodding ( 9.002 +/− 12.31 s, n = 5, p = 0.4001 ANOVA Holm–Šidak post hoc test) before mating or swimming away. These data suggest that loss of HLH-17 or CEP_glia does not prevent males from executing prodding; however, mutant males appear to take longer to complete this step than WT. We found through visual observation that, unlike WT males, hlh-17(ns204) males often prodded relatively close to the vulva but away from the vulva slit, suggesting that loss of hlh-17 expression, which causes aberrant dopamine signaling (Felton and Johnson, 2011, 2014), reduces precision during prodding. This observation is consistent with studies by Correa et al. (2012) demonstrating that dopamine signaling is required for fine-tuning goal-oriented behaviors. The phenotype in ΔCEP_glia males was also visibly abnormal, but more difficult to characterize because they appeared to properly align with the vulva slit but to “linger” in the prodding step. Collectively, our data suggest that prodding is not the bottleneck for males that lack CEP_glia but could be a bottleneck for mating behavior in males that lack HLH-17. We surmised that the time spent in the act of prodding for hlh-17(ns204) and ΔCEP_glia males would increase the total time these animals spent in the act of mating thereby increasing the likelihood of mating success.

To verify our prediction, we re-examined the total mating time experiments shown in Figure 1a. First, we noted that the mating behavior lasted for 10 minutes or longer for 86% (12/15) hlh-17(ns204) males and 60% (3/5) ΔCEP_glia males (Figure S4). In comparison, mating behavior lasted for 10 minutes or longer in 46% (7/15) WT males and 0% (0/5) ΔCEP_glia; hlh-17(ns204) males. Second, we noted that an increase in mating time correlated with increased mating success for WT males (86% mated successfully when time exceeded 10 minutes) and hlh-17(ns204) males (92% mated successfully when time exceeded 10 minutes). Mating success did not increase when ΔCEP_glia spent more time prodding (33% when time exceeded 10 minutes versus 50 % when under 10 minutes).

Since increased prodding time did not correlate with increased mating success for ΔCEP_glia males, we questioned whether the inability to complete mating could be attributed to abnormal spicule insertion and impaired sperm transfer by examining the effects of CEP_glia and HLH-17 on spicule protraction. As expected, in our assays 100% WT males protract their spicules after localizing the vulva slit during prodding behavior and right before spicule insertion. In contrast, 13% hlh-17(ns204) (n = 15), 60% ΔCEP_glia (n = 5) and 67% ΔCEP_glia; hlh-17(ns204) (n = 6) males prematurely protracted their spicule during the response stage of mating, well before locating the vulva (Figure S5). Based on these data it is not surprising that increase prodding time did not lead to increased mating success in ΔCEP_glia males since early and permanent spicule protraction has been shown to interfere with mating success (LeBoeuf and Garcia, 2012). Collectively these data support a model in which CEP_glia are essential for regulating the copulation stage of mating behavior, potentially relying on HLH-17-dependent dopamine signaling activity to fine tune prodding and other factors to regulate stimulation of the spicule muscles.

3.4. CEP_glia antagonize muscle contractions that are critical for the DMP

To date, mating behavior has only been described in the context of the tail neural network. Our mating behavior data suggest that CEP_glia, which are located in the head, communicate with the tail neural network to modulate behavior. To understand the mechanism by which this long-range communication is modulated, we examined the effect of CEP_glia loss on the defecation motor program (DMP). In both males and hermaphrodites, the DMP is initiated by sequential contractions of the posterior body wall, anterior body wall, and enteric muscles, which ultimately result in the expulsion of pressurized waste in the gut (Branicky & Hekimi, 2006; Thomas, 1990) (FigureSM3). The neuronal connections that drives this program is sexually dimorphic (Figure 2). In hermaphrodites, gap junctions electrically couple the anal depressor to the intestinal and sphincter muscles; all three are characterized as enteric muscles and, when stimulated, promote enteric muscle contractions. In males, however, the anal depressor and the sphincter muscle are not in direct communication, relying instead on intermediary interactions with the male specific protractor muscle (Nagy et al., 2015; Reiner and Thomas, 1995). The anal depressor is connected through gap junctions to the copulation sex muscle, the gubernacular erector muscle, to promote spicule insertion (LeBoeuf & Garcia, 2017). Interestingly, the gubernacular erector muscles functionally link the DMP to mating behavior since they reorient the protracted spicule muscles posteriorly to allow for sperm transfer during mating (Liu et al., 2011). Similarly, the GABAergic DVB and AVL neurons are also sexually dimorphic. DVB, located in the tail, forms a chemical synapse with the anal depressor, which, in hermaphrodites, is required for enteric muscle contractions in the DMP. However, in males, DVB does not synapse with the anal depressor, instead forming a chemical synapse with the protractor muscle that is coupled to the anal depressor. In males, the anal depressor no longer regulates the DMP but instead is required for spicule insertion during mating (LeBoeuf & Garcia, 2017; Hart & Hobert, 2018). The GABAergic AVL neuron, located in the head, synapses with DVB and the enteric intestinal muscle to regulate enteric muscle contractions in hermaphrodites. In males, AVL no longer forms a chemical synapse with the intestinal muscle, however the synapse between AVL and DVB is maintained (LeBouf & Garcia, 2017) (Figure 2a, b).

Figure 2.

Sex differences in DVB synapse formation. (a) Connectivity of the defecation circuitry in hermaphrodites. The DVB is a central player in stimulating the enteric muscles and directly synapses with the intenstinal muscle and anal depressor (b) Connectivity of the defecation and male mating circuitry in males. Changes in anatomical features results in the anal depressor functioning as a copulatory muscle for mating and the DVB directly synapses with the protractor muscle. ger, gubernacular erector muscle. Drawing adapted from LeBoeuf and Garcia, 2017; Hart and Hobert, 2018.

Here, we questioned whether loss of CEP_glia, located in the head, affects the activity of DVB, whose cell body resides in the tail and whose processes terminate approximately mid-body. We rationalized that if CEP_glia are required for modulating the activity of DVB, which synapses with the protractor muscle and regulates the activity of the male spicules, then loss of CEP_glia may result in abnormal spicule protraction. As shown in Figure S5, 38% ΔCEP_glia (n = 8) and ΔCEP_glia;hlh-17(ns204) (n = 8) males protracted and retracted their spicule at least once per DMP cycle while WT males (n = 8) and hlh-17(ns204) males (n = 8) did not. Together, these data suggest that CEP_glia are essential for reducing unorthodox spicule protraction, supporting a model in which CEP_glia in the head can influence tail-mediated behaviors.

Although the DVB neuron no longer synapses with the enteric muscles in males, the sex muscles that regulate spicule protraction are coupled to the sphincter and intestinal muscles, which are both required for enteric muscle contractions during the defecation motor program. Therefore, we expected that the rate of enteric muscle contractions (EMC) would be inhibited by loss of CEP_glia. We tested this prediction by measuring EMC rates as shown in Figure 3a (F (3,24) = 23.94, p < 0.0001 ANOVA). We were surprised to find that the rates of EMCs were similar in WT (0.69 +/− 0.09 EMC/minutes, n = 7) and ΔCEP_glia (0.63 +/− 0.18 EMC/minutes, n = 8, p = 0.6245 ANOVA Dunnett’s post hoc test) males (Figure 3a), while the rates were significantly slower in hlh-17(ns204) males (0.36 +/− 0.09 EMC/minutes, n = 8, p < 0.0001 ANOVA Dunett’s post hoc test) and in ΔCEP_glia;hlh-17(ns204) males (0.20 +/− 0.09 EMC/minutes, n = 7, p < 0.0001 ANOVA Dunnett’s post hoc test). Our results were similar when we examined posterior muscle contractions (pBocs), which follow enteric muscle contractions in response to intestinal signaling (Figure 3b, F (3,26) = 15.17, p < 0.0001 ANOVA). That is, the rate of pBocs in WT (0.71 +/− 0.07 pBoc/minute, n = 7) and in ΔCEP_glia males (0.77 +/− 0.08 pBoc/minute, n = 7, p = 0.5000 Dunnett’s post hoc test) did not differ significantly while the rates were significantly slower than WT in hlh-17(ns204) (0.51 +/− 0.10 pBoc/minute, n = 8, p = 0.0006 ANOVA Dunnett’s post hoc test) and ΔCEP_glia;hlh-17(ns204) (0.54 +/− 0.11 pBoc/minute, n = 8, p <0.0022 ANOVA Dunnett’s post hoc test) males. Taken together, the defecation motor program does not require signaling from the CEP_glia in males; however, hlh-17 expression may be essential. Future studies using additional alleles of hlh-17, or that allow transgenic rescue of the DMP phenotypes in hlh-17(ns204) males will further support a role for HLH-17 in the DMP. Our data shows that the rate of enteric muscle contractions are significantly slower with the loss of both CEP_glia and hlh-17 expression, suggesting that HLH-17 may negatively regulate a signal that antagonizes the muscle contractions inherent to the DMP.

Figure 3.

Steps in the defecation motor program are not affected in ΔCEP_glia males. The rate of (a) enteric muscle contractions and (b) ) posterior body muscle contractions was quantified in males. WT [nsIs105; him-5(e1490)] EMC, n = 7; pBoc, n = 7, hlh-17(ns204) [nsIs105; hlh-17(ns204); him-5(e1490)⁻] EMC, n = 8; pBoc, n = 8 , ΔCEP_glia [nsIs105; him-5(e1490); olaEx805] EMC, n = 8; pBoc, n = 7, and ΔCEP_glia; hlh-17(ns204) [nsIs105; hlh-17(ns204); him-5(e1490); olaEx805] EMC, n = 7; pBoc, n = 8. The number above a given data set represents the number of outliers. Astericks represent adjusted P-values after Dunnett’s posthoc correction: *p<0.05, **p<0.005, ***p<0.001, ****p<0.0001.

3.5. Loss of CEP_glia and HLH-17 reduces pBocs and EMCs in the DMP of hermaphrodites

Recent studies have shown that while the sequence of the DMP is unchanged in males and hermaphrodites, the architecture of the DMP neurons and musculature differs in the two sexes (Reiner and Thomas 1995; Nagy et al., 2015; LeBoeuf and Garcia, 2017). Our mating data and the aberrant spicule protraction phenotype of ΔCEP_glia males suggest that the CEP_glia mitigate these effects via the DVB neuron. Therefore, it is possible that the CEP_glia would have a sexually dimorphic effect on defecation that might not be dependent on hlh-17 activity. To test this possibility, we examined pBocs in hermaphrodites. As shown in Figure 4a (F (5, 123) = 14.43, p < 0.0001 ANOVA), loss of hlh-17 in hermaphrodites had a similar effect as loss of hlh-17 in males; the rate of pBocs in hlh-17(ns204) hermaphrodites (0.88 +/− 0.21 pBoc/minute, n = 30, p < 0.001 Dunnett’s post hoc test) was significantly lower than that in WT (1.17 +/− 0.15 pBoc/minute, n = 30) hermaphrodites. Unlike in males, however, the rates of pBocs were significantly lower in ΔCEP_glia (0.99 +/− 0.22 pBoc/minute, n = 20, p = 0.0022 ANOVA Dunnett’s post hoc test) and ΔCEP_glia; hlh-17(ns204) (0.87 +/− 0.21 pBoc/minute, n = 20, p < 0.0001 ANOVA Dunnett’s post hoc test) hermaphrodites. Together these data suggest that in hermaphrodites, HLH-17 dependent regulation of pBocs also requires functional CEP_glia.

Figure 4.

Steps in the defecation motor program were affected in ΔCEP_glia hermaphrodites. The rate of (a) posterior body muscle contractions and (b) enteric muscle contractions was quantified in hermaphrodites. (c) Quantification of inter-cycle length, the time between enteric muscle contractions (EMC) and posterior body contractions (pBoc). Int = Inter-cycle length. WT [nsIs105; him-5(e1490)] pBoc n = 30 ; EMC, n = 14; Int, n = 5, hlh-17(ns204) [nsIs105; hlh-17(ns204); him-5(e1490)⁻] pBoc, n = 30; EMC, n = 12; Int, n = 5 , ΔCEP_glia [nsIs105; him-5(e1490); olaEx805] pBoc, n = 20 ; EMC, n = 13; Int, n = 5 , and ΔCEP_glia; hlh-17(ns204) [nsIs105; hlh-17(ns204); him-5(e1490); olaEx805] pBoc, n = 21, EMC, n = 10, Int, n = 5. pAmPH hermaphrodites ectopically expressed hlh-17 in the amphid glia. pAmPH pBoc, n = 13; EMC, n = 13, and hlh-17(ns204); pAmPH pBoc, n = 15; EMC, n = 13. The number above a given data set represents the number of outliers. Astericks represent adjusted P-values after Dunnett’s posthoc correction using WT as the control: *p<0.05, **p<0.005, ***p<0.001, ****p<0.0001.

Previous data suggest that posterior contractions are initiated through signals secreted from the intestine, the central pacemaker that modulates the periodic activation of DMPs. Therefore, we questioned whether fewer pBocs in ΔCEP_glia hermaphrodites could be attributed to changes in intestinal signaling by examining periodicity of the DMP, which is measured as the time between two consecutive pBocs (Liu & Thomas, 1994; Thomas, 1990). We expected that the decrease in pBocs in ΔCEP_glia hermaphrodites would correlate with a change in the periodicity of the defecation cycle. Instead, we found that periodicity was unaffected in hlh-17(ns204), ΔCEP_glia, and ΔCEP_glia;hlh-17(ns204) hermaphrodites (Figure S6, F (3,32) = 2.002, p = 0.1335 ANOVA), suggesting that while loss of CEP_glia reduces pBocs/minutes, signaling from the intestine is not altered. These results are in agreement with previous studies suggesting that cycle length is uncoupled to the DMP (Teramoto & Iwasaki, 2006; Thomas, 1990).

In males, loss of the CEPglia alone affects neither the rate of posterior body wall contractions (pBocs) nor the rate of enteric muscle contractions (EMCs); however, both rates decrease in animals that lack HLH-17, with or without CEP_glia (Figure 3). Our data also suggest that CEP_glia influence the activity of the protractor and anal depressor muscles that regulate spicule insertion during mating (Figure S5), possibly through the chemical synapses between the protractor muscle and DVB. In hermaphrodites, the protractor muscle does not develop, and instead the anal depressor chemically synapses with DVB while forming gap junctions with the sphincter and intestinal muscles that are required for EMCs. Therefore, we measured the rate of EMCs in hermaphrodites to determine whether, unlike in males, the rate of EMCs would decrease in hermaphrodites that lack CEP_glia (Figure 4b, F (5, 71) = 8.074 ANOVA). We found that EMC rates in WT hermaphrodites (1.03 +/− 0.23 EMC/minute, n = 14) were significantly higher than in hlh-17 (ns204) (0.81 +/− 0.08 EMC/minutes, n = 12, p = 0.0414 ANOVA Dunnett’s post hoc test) and ΔCEP_glia (0.81 +/− 0.23 EMC/minutes, n = 13, p = 0.0351 ANOVA Dunnett’s post hoc test) hermaphrodites, and higher, but not significantly, than in ΔCEP_glia; hlh-17(ns204) (0.89 +/− 0.20 EMC/minutes, n = 10, p = 0.3821 ANOVA Dunnett’s post hoc test) hermaphrodites, supporting our hypothesis that CEP_glia act through DVB to influence DMP and mating behavior.

Together, our data show that, in hermaphrodites, loss of either HLH-17 or CEP_glia effects posterior and enteric muscle contractions without affecting DMP cycle length. How is it that ΔCEP_glia and hlh-17 (ns204) hermaphrodites execute fewer contractions than WT when the overall cycle length remains the same? We wondered if hermaphrodites that have difficulty contracting their enteric muscles and expelling waste execute fewer contractions as a result, thereby prolonging the time in which the subsequent pBoc is initiated. Therefore, we measured the inter-cycle length, defined here as the time between an enteric muscle contraction at the end of one cycle and the posterior muscle contraction (pBoc) at the beginning of the subsequent cycle (Figure 4c, F (3, 14) = 10.03, p = 0.0009 ANOVA). We found that the inter-cycle lengths for WT (38.17 +/− 4.23 seconds, n = 4), hlh-17(ns204) (37.58 +/− 1.10 seconds, n = 4 p = 0.9982, ANOVA Dunnett’s post hoc test), and ΔCEP_glia (44.69 +/− 4.27 seconds, n = 4, p = 0.1119 ANOVA Dunnett’s post hoc test) hermaphrodites were not significantly different. In contrast, the inter-cycle length in ΔCEP_glia; hlh-17(ns204) (57.26 +/− 8.65 seconds, n = 5, p = 0.0012 ANOVA Dunnett’s post hoc test) hermaphrodites was significantly longer than that of WT hermaphrodites. These data show that while the DMP cycle length is unaffected by loss of HLH-17 or CEP_glia , enteric muscle contractions are affected in animals that lack CEP_glia, and therefore the inability to execute enteric muscle contractions causes hermaphrodites take longer to initiate a subsequent pBoc.

3.6. Ectopic Expression of HLH-17 Glia specific HLH-17 mediated regulation modulates DMP behaviors

Previous work show that HLH-17 is constitutively expressed in the CEP_glia from embryogenesis to adulthood (McMiller & Johnson, 2005; Yoshimura et al., 2008; Tintori et al., 2016). Additionally, hlh-17 expression is also detected in presumptive CEP_glia cells in the tail, and very faintly in the inner and outer labial glia that are closely associated with the CEP_glia (www.wormatlas.org). Sporadic expression of HLH-17 using a myristoylated GFP marker has also been reported in motor neuron commissures (Yoshimura et al., 2008), a region of the neuron that modulates the trajectory of axons in response to guidance cues such as netrins. Currently, activity of the hlh-17 promoter is the only cell-specific marker for CEP_glia. Collectively these data suggest that hlh-17 activity is tightly restricted to the cephalic sensilla, raising the possibility that HLH-17 is critical for the unique functions of the CEPglia, including potential roles in netrin-dependent signaling and synaptic maintenance (Colon-Ramos et al., 2007). We wondered whether ectopic expression of hlh-17 in other glia cell types could bypass the requirement for HLH-17 activity in the CEP_glia, thereby rescuing the pBoc and EMC phenotypes of hlh-17(ns204) hermaphrodites. We had hoped to also determine whether ectopic expression of hlh-17 could rescue the phenotypes of ΔCEP_glia and ΔCEP_glia; hlh-17(ns204) hermaphrodites; however, we were not able to generate the necessary strains by mating. To test our question, we ectopically expressed hlh-17 in glia cells of the amphid sensilla. Amphid glia (AM_glia) are similar to the CEP_glia in that their cell bodies are located in the head, and that they send a single process to the nose tip; they, and the 12 neuron pairs that they ensheath, form the major chemosensory organs of C. elegans (Bacaj et al., 2008; Wang et al, 2008, 2012; Han et al., 2013; Stout et al., 2014). We hypothesized that HLH-17 function is unique to CEP_glia and that ectopic expression in amphid glia would not be sufficient to rescue the hlh-17(ns204) pBoc and EMC phenotype in hermaphrodites. As shown in Figure 4, ectopic expression of hlh-17 in the amphid glia of WT hermaphrodites that also expression hlh-17 in the CEP_glia had no effect on either pBocs (Figure 4a, vs WT p = 0.9030 ANOVA Dunnett’s post hoc test) or EMCs (Figure 4b, vs WT p = 0.3821). Importantly, ectopic expression of hlh-17 in the amphid glia of hlh-17(ns204) hermaphrodites rescued the pBoc phenotype of hlh-17(ns204) hermaphrodites (Figure 4a) so that the rate of pBocs in hlh-17(ns204);pAM::hlh-17 hermaphrodites (1.15 +/− 0.17 pBoc/minute, n = 15, p = 0.9269 ANOVA Dunnett’s post hoc test) was not significantly different from that in WT hermaphrodites (1.16 +/− 0.15 pBoc/minute, n = 28). To our surprise, ectopic expression of hlh-17 in the amphid glia of hlh-17(ns204) hermaphrodites (0.59 +/− 0.21 EMC/minute, n = 13, p <0.001 vs WT ANOVA Dunnett’s post hoc test) did not rescue the EMC phenotype seen in hlh-17(ns204) hermaphrodites (0.81 +/− 0.08 EMC/minute, n = 12, p = 0.0414 vs WT ANOVA Dunnett’s post hoc test, Figure 4b). Together, these data suggest that in hermaphrodites, hlh-17 expression is required for normal posterior muscle contractions during DMP in hermaphrodites, and that expression from either the CEP_glia or the amphid glia is sufficient. In contrast, these data suggest that in hermaphrodites normal EMCs require hlh-17 activity in the CEP_glia.

4. Discussion

In this study, we show that CEP_glia and hlh-17 expression in the glia are important for regulating complex and rhythmic behaviors in males and hermaphrodites. Complex behaviors integrate multiple sensory modulatory inputs from independent motor programs to orchestrate specific motor outputs. In C. elegans, mating behavior in males and the defecation motor program (DMP) in both sexes are complex behaviors, while DMP is also a rhythmic behavior that utilizes an intrinsic pacemaker to modulate periodic activation of a stereotyped sequence of behaviors. Interestingly, mating and DMP share similarities in anatomical features and circuitry, as well as neurotransmission, making these behaviors ideal for probing glia function.

The anatomical features of male mating and DMP are in the posterior or tail region of the animal. The cell body of the GABAergic DVB neuron is bundled with the DVA and DVC interneurons to make up the dorso-rectal ganglion (White et al., 1986; Li et al., 2006). Processes from DVB extend anteriorly, ending approximately mid-way along the body, while processes from DVA and DVC extend anteriorly, all the way to the nerve ring in the head. We found, by reviewing computer generated reconstructions of the C. elegans neural network from serial electron micrographs (White et al., 1986; Bhatla, 2009; Jarell et al., 2012; Xu et al., 2013; Sammut et al., 2015) that DVC likely forms chemical synapses with DVB, as well as with the CEP_glia and the AM_glia. In C. elegans hermaphrodites, DVB synapses with two enteric muscles that both are, in turn, electrically coupled to the enteric sphincter muscle. In males, DVB synapses with the copulatory protractor muscle, which in turn is electrically coupled to the enteric sphincter muscle and the copulatory anal depressor. Strikingly, the anal depressor is not an enteric muscle in males, and so is not stimulated to contract during the DMP (Figure 2a, b). Thus, DMP and male mating both rely on the modulation of GABAergic signaling from the DVB. Our data suggest that loss of CEP_glia, which reside in the head and extend processes anteriorly towards the mouth, significantly effects male mating and DMP behaviors that are regulated by distally-located, GABAergic DVB. Here we propose a model in which the activities of the DVB neuron during male mating and defecation are mediated through signaling between the CEP_glia and the glutamatergic DVC interneuron, and in which bidirectional communication between the CEP_glia and the intestine generates a feedback loop to control the timing of the DMP (Figure 5).

Figure 5.

Mode of regulation for CEP_glia. Proposed mechanisms by which CEP_glia transmit signaling molecules to direct behavior in (a) hermaphrodites. The DMP is initiated (1) by signals released from the intestine to the posterior body wall muscles to initiate pBocs. Signals from the intestine through the (2) AVL neuron may regulate anterior body wall muscles (aBocs) and through the (3) DVB to promote the release of excitatory GAB and stimulate the enteric muscle and initiate enteric muscle contraction (EMC) and expulsion of waste. The CEP_glia respond to activity from the (4) intestine and (5) signal through DVC to DVB to promote the release of inhibitory GABA onto the enteric muscles to allow for the transition to the next pBoc. The behaviors modulated in (b) males include the DMP and mating. The mode of regulation for the DMP in males is similar to hermaphrodites. During mating inhibitory signals released from the DVB neuron, by the way of the CEP_glia, modulates the time in which the SPC neuron stimulates the copulatory muscles and facilitates the transition from prodding to spicule insertion and sperm transfer.

During the DMP in hermaphrodites, DVB transmits excitatory GABA to stimulate enteric muscle contractions (Figure 5a). This process is initiated through proton signaling from the intestine to the posterior body wall muscles, which initiates posterior body contractions (pBocs) (Beg et al., 2008), and by signaling from the intestine to DVB, which promotes release of excitatory GABA to stimulate enteric muscle contraction (McIntire et al.,1993b; Schuske et al., 2004; Mahoney et al., 2008). In males, this same excitatory GABAergic signaling stimulates the copulatory protractor muscle (Jobson et al., 2015), and indirectly, the enteric muscles and the (copulatory) anal depressor. During mating, DVB switches between excitatory and inhibitory GABAergic signaling to fine-tune the execution of spicule protraction, while inhibitory signaling during DMP promotes the transition from EMC to pBoc (Figure 5b). Our data support a model in which the CEP_glia mediates inhibitory GABAergic signaling from DVB by way of the DVC interneuron to inhibit the excitation and contraction of the enteric and copulatory muscles. Further, our model predicts that bidirectional signaling between the intestine and CEP_glia promotes a feedback loop that coordinates the transition between behaviors within the defecation motor program via a mechanism that relies on transcriptional regulation by HLH-17. Based on our model the influences of CEPglia and HLH-17 are sexually dimorphic, which is supported by reduced viability of ΔCEP_glia and ΔCEP_glia; hlh-17(ns204) males (see Materials and Methods) and by the sex-specific impact of CEPglia on posterior body wall (pBoc) and enteric muscle (EMC) contractions in hermaphrodites.

In C. elegans males, precision and accuracy during the mating ritual relies on coordinated signaling between glia, neurons, and non-neuronal cells. After the vulva is detected during the response stage, for example, the post cloacal neurons ensure precise positioning of the male tail (Liu and Sternberg, 1995; Loer et al., 1999) and subsequently release acetylcholine onto the protractor muscle. The cholinergic SPC neuron then activates the protractor muscle, thereby promoting prolonged contraction (Garcia et al., 2001; Liu et al., 2007). DVB is thought to regulate the activation and contraction of the spicule muscles (Hart and Hobert, 2018) through synapses with SPC and the protractor muscle. The spicule protraction phenotypes of ΔCEP_glia and ΔCEP_glia; hlh-17(ns204) males (Figure S5) supports our model which predicts that signaling from the CEP_glia modulates cholinergic signaling from SPC by temporally mediating the transmission of inhibitory GABAergic signaling by DVB. Others have demonstrated the importance of intercellular signaling between glia, neuronal, and non-neuronal interactions in the modulation of behavior. Hart and Hobert (2018), for example, found that expression of the neuroligin protein, NLG-1, in the postsynaptic copulatory muscles and expression of neurexin protein, NRX-1, in the DVB neuron regulates neurite outgrowth. We speculate that, given the ability of neuroligins and neurexins to modulate synaptic transmission (Südhof, 2008), cell autonomous expression of NRX-1 in the DVB neuron may be required for the synaptic connectivity with the DVC interneuron.

Our study suggests that hlh-17 expression in CEP_glia modulates and integrates neurotransmitter signaling, which contributes to the mechanism by which CEP_glia fine-tune the fluent coordination of complex behaviors. We previously demonstrated that dopamine signaling is disrupted in hlh-17(ns204), and that HLH-17 is required for normal dopaminergic behaviors (Felton and Johnson, 2011, 2014). Here, hlh-17(ns204) males experienced difficulty in transitioning from prodding to spicule insertion, a step in which accuracy and precision require dopamine signaling (Loer & Kenyon 1993; Correa et al., 2012). Our findings are similar to work in zebrafish which suggest that inhibiting dopamine signaling reduces persistence in swimming behavior (Tran et al., 2015), partly through the action of radial astrocyte glia that sense ineffective behavior and initiate changes in behavioral state to suppress swimming behavior (Mu et al., 2019). Our data also show that loss of hlh-17 expression negatively effects pBocs and EMCs in the DMP, and that the pBoc phenotype can be rescued by ectopic expression of HLH-17 in the AM_glia (Figure 4a). We interpret these data to mean that the HLH-17 regulatory network includes genes that are required for CEP_glia mediated control of the DMP. Thus, ectopic HLH-17 activity in the AM_glia promotes AM_glia expression of genes that are normally restricted to the CEP_glia, thereby rescuing the pBoc phenotype in hlh-17(ns204) animals. In support of this interpretation, expression of pbo-5 and pbo-6, genes that encode a proton-gated ion channel required for posterior body wall contractions in response to intestinal signaling, is affected in hlh-17 animals (unpublished microarray data, Felton, 2014). Unfortunately, we could not confirm pbo-5/pbo-6 expression in FACs enriched CEPglia using RT-qPCR because we could not completely eliminate contamination of closely associated neurons. Our future studies will further explore signaling between the CEP_glia and the intestine during the DMP. The failure to rescue the EMC phenotype of hlh-17(ns204) animals by ectopic HLH-17 expression in the AMglia (Figure 4b) was surprising, since AM_glia and CEP_glia both synapse to DVC. We surmise either that HLH-17 is required, but not sufficient to drive signaling between the CEP_glia and DVC or that HLH-17 expression outside of the CEPglia may influence the release of inhibitory molecules that affect signaling between DVC and AVL, a GABAergic neuron that functions redundantly to DVB to regulate EMCs.

In summary, we have provided the first set of behavioral evidence that outlines mechanisms by which the tail neural network communicates with the invertebrate CEP_glia , through complex and diverse modes of regulation, to modulate behaviors.

Significance Statement.

Mechanisms by which astrocytes play essential roles in bidirectional signaling with neuronal and non-neuronal cells has been a burgeoning topic over the last decade. Yet, despite our current understanding of astrocytic signaling and response, it is still unclear how glia regulate complex behaviors. This work highlights communication between astrocyte-like CEP_glia and interneurons and explores the mechanism by which that information is translated into a specific behavioral output in Caenorhabditis elegans. More importantly, these new insights into CEP_glia function provide clues for understanding the physiological and pathological relevance for astrocytic responses in mammals.