Abstract
A neuropeptide (NLP-12) and its receptor (CKR-2) potentiate tonic and evoked ACh release at Caenorhabditis elegans neuromuscular junctions. Increased evoked release is mediated by a presynaptic pathway (egl-30 Gαq and egl-8 PLCβ) that produces DAG, and by DAG binding to short and long UNC-13 proteins. Potentiation of tonic ACh release persists in mutants deficient for egl-30 Gαq and egl-8 PLCβ and requires DAG binding to UNC-13L (but not UNC-13S). Thus, NLP-12 adjusts tonic and evoked release by distinct mechanisms.
Introduction
To become fusion competent, synaptic vesicles (SVs) must physically attach to the plasma membrane (termed docking) and must undergo a process termed priming (Verhage and Sorensen, 2008). Docking and priming are both mediated by the SNARE proteins. Primed vesicles are thought to consist of docked SVs containing partially assembled trans-SNARE complexes (Xu et al., 1999). Several SNARE binding proteins regulate SV docking and priming. UNC-10/RIM, UNC-13/Munc13, and UNC-31/CAPS promote docking and priming while Tomosyn inhibits both processes (Gracheva et al., 2006, 2007, 2008; McEwen et al., 2006; Weimer et al., 2006; Hammarlund et al., 2007). Collectively, these studies suggest that Munc13 (and other priming factors) stimulate exocytosis by promoting SV docking and the initial assembly of trans-SNARE complexes.
SV priming factors are extensively regulated by second messengers (Verhage and Sorensen, 2008). For example, all Munc13 proteins have binding sites for DAG (the C1 domain), calcium (C2B), and calmodulin (Betz et al., 1998, 2001; Shin et al., 2010; Lipstein et al., 2012). Mutations that block DAG binding to Munc13's C1 domain block potentiation of synaptic transmission by synthetic DAG ligands (phorbol esters; Lackner et al., 1999; Rhee et al., 2002; Basu et al., 2007). Similarly, mutations that block calcium and calmodulin binding to Munc13 alter short-term plasticity (Shin et al., 2010; Lipstein et al., 2013). Thus, treatments altering individual second messengers adjust Munc13 priming activity and neurotransmitter release. In contrast, endogenous neuromodulators simultaneously activate multiple second messengers. For example, GPCRs activate protein kinases, Rac and Rho GTPases, as well as phospholipases. Thus, natural neuromodulators may have more complex effects than treatments designed to manipulate individual second messengers.
The Caenorhabditis elegans neuromuscular junction (NMJ) has been used as a model to address these questions. Transmission at this synapse is mediated by graded release of ACh, whereby release varies with the strength of depolarization (Liu et al., 2009). When activity is low, transmission consists of mEPSCs that are evoked by fusion of a single SV (Liu et al., 2005), hereafter designated tonic release. Forced depolarization of motor neurons evokes the synchronous release of several hundred SVs. Prior genetic studies suggest that egl-30 Gαq and its target (egl-8 PLCβ) enhances transmission at NMJs (Hajdu-Cronin et al., 1999; Lackner et al., 1999; Miller et al., 1999). These studies used behavioral (not electrophysiological) assays; consequently, it remains unclear how egl-30 Gαq and egl-8 PLCβ alter ACh release.
We previously showed that a neuropeptide (NLP-12) and its receptor (CKR-2) potentiate tonic and evoked ACh release at NMJs (Hu et al., 2011). Here we show that potentiation of evoked release requires activation of egl-30 Gαq and egl-8 PLCβ, and that distinct UNC-13 proteins mediate the resulting potentiation of tonic and evoked ACh release.
Materials and Methods
Strains.
Animals were cultivated at 20°C on agar nematode growth media seeded with OP50 bacteria. The following strains were used in this study: wild-type N2 bristol, DA1084 egl-30(ad806), JT47 egl-8(sa47), KP6901 unc-13(s69), KP7451 nuEx1677 [Pacr-2::EGL-30];egl-30(ad806), KP7447 nuEx1673 [Pacr-2::EGL-8];egl-8(sa47), KP6893 nuEx1515 [Psnb-1::UNC-13L];unc-13(s69), KP6899 nuIs46 [Punc-13::UNC-13S::GFP];unc-13(s69), KP6893 nuEx1515 [Psnb-1::UNC-13L(H699K)];unc-13(s69), and KP6899 nuIs52 [Punc-13::UNC-13S(H348K)::GFP];unc-13(s69).
Constructs and transgenes.
Transgenic strains were isolated by microinjection of various plasmids using either Pmyo-2::NLS-GFP (KP#1106) or Pmyo-2::NLS-mCherry (KP#1480) as a coinjection marker. Integrated transgenes were obtained by UV irradiation and were out-crossed at least six times. EGL-30 (M01D7.7a; wormbase.org) or EGL-8 (B0348.4a; wormbase.org) cDNAs were expressed in cholinergic motor neurons using the acr-2 promoter.
Locomotion assays.
For aldicarb paralysis, between 18 and 25 young adult worms were transferred to plates containing 1.5 mm aldicarb and assayed for paralysis as described previously (Nurrish et al., 1999).
Electrophysiology.
Electrophysiology was done on dissected adults as previously described (Richmond and Jorgensen, 1999; Hu et al., 2011, 2013). For aldicarb exposure, a single adult was transferred to a plate containing 1 mm aldicarb for 60 min before the dissection. Statistical significance was determined using a two-tailed Student's t test.
Results
Aldicarb potentiation of release is decreased in egl-30 Gαq and egl-8 PLCβ mutants
The effects of NLP-12 on release are assessed by recording EPSCs following treatment with a cholinesterase inhibitor, aldicarb (Hu et al., 2011). Aldicarb induces body muscle contraction, which enhances NLP-12 secretion from a stretch sensing neuron (DVA; Hu et al., 2011). Following aldicarb treatment, the mEPSC rate and the total synaptic charge of evoked responses were both approximately doubled (Fig. 1A,B,D–F). The effects of aldicarb on tonic and evoked release are eliminated in mutants lacking NLP-12 and in those lacking an NLP-12 receptor (CKR-2; Hu et al., 2011).
EGL-30/Gαq and EGL-8/PLCβ are required for aldicarb-induced potentiation of evoked release. Mutations inactivating EGL-30/Gαq and EGL-8/PLCβ prevent aldicarb potentiation of evoked ACh release but have little effect on potentiation of tonic release. Evoked EPSCs (A, B) and mEPSCs (D–F) were recorded from adult body wall muscle, with (blue) and without (black) a 60 min aldicarb treatment. Averaged evoked EPSCs (A) and representative mEPSC traces (D) are shown. Summary data for evoked (B) and tonic (E, F) release are shown for wild-type, egl-30, and egl-8 mutants. Rescue indicates mutant animals containing a transgene expressing the indicated gene in cholinergic motor neurons. C, Aldicarb-induced paralysis is compared for the indicated genotypes. Statistically significant differences (***p < 0.001, **p < 0.01, *p < 0.05, and ns not significant), the number of animals analyzed (B, E, F), and the number of replicate experiments (C) are indicated. Error bars indicate SEM.
CKR-2 receptors are coupled to G-proteins containing a Gαq-subunit (Janssen et al., 2008). Consequently, we tested the idea that egl-30 Gαq and egl-8 PLCβ are required for aldicarb-induced potentiation. The aldicarb-induced increase in evoked synaptic charge was eliminated in both egl-30 Gαq and egl-8 PLCβ mutants, and this defect was rescued by constructs expressing the corresponding genes in cholinergic motor neurons (using the acr-2 promoter; Fig. 1A,B). Similarly, restoring egl-30 Gαq and egl-8 PLCβ expression in motor neurons rescued mutant defects in aldicarb-induced paralysis (Fig. 1C). Thus, egl-30 Gαq and egl-8 PLCβ are required for aldicarb-induced potentiation of evoked ACh release, as would be predicted if the NLP-12 receptor (CKR-2) was coupled to Gq (Janssen et al., 2008). Interestingly, the aldicarb-induced increase in mEPSC rate was only modestly reduced in egl-30 Gαq mutants and was unaffected in egl-8 PLCβ mutants (Fig. 1D–F), implying that distinct mechanisms mediate NLP-12 potentiation of tonic and evoked release.
To further investigate EGL-30's role in tonic release, we analyzed eat-16 RGS mutants (Fig. 2). EAT-16 has GTPase-activating activity for EGL-30 Gαq (Hajdu-Cronin et al., 1999); consequently, eat-16 mutants can be used to assess the effects of enhanced EGL-30 activity. In eat-16 mutants, mEPSC rates were significantly increased (Fig. 2A,B). This effect was abolished in egl-30 eat-16 double mutants (Fig. 2A,B). Thus, increased EGL-30 Gαq activity produces a corresponding increase in tonic release.
Increased EGL-30 Gαq activity enhanced tonic release. A, B, The eat-16 RGS mutants have enhanced EGL-30 activity (Hajdu-Cronin et al., 1999) and a corresponding increase in mEPSC rate. This effect was abolished in egl-30 mutants, and was diminished in nlp-12 and ckr-2 mutants. Representative traces (A) and summary data (B) for mEPSCs are shown. C, The egl-30 and ckr-2 single mutants were both resistant to aldicarb-induced paralysis but double mutants did not exhibit additive defects. These results suggest that NLP-12 stimulates ACh release by activating EGL-30. Statistically significant differences (***p < 0.001, **p < 0.01, *p < 0.05, and ns, not significant), the number of animals analyzed (B), and the number of replicate experiments (C) are indicated. Error bars indicate SEM.
EGL-30 Gαq is required for synaptic potentiation by NLP-12
NLP-12 and EGL-30 Gαq are both required for synaptic potentiation by aldicarb, consistent with the idea that NLP-12 potentiates release via activation of EGL-30. We did two further experiments to test this idea. First, we analyzed tonic release in nlp-12 eat-16 and eat-16; ckr-2 double mutants. The enhanced mEPSC rate exhibited by eat-16 single mutants was significantly reduced in both nlp-12 eat-16 and eat-16; ckr-2 double mutants (Fig. 2A,B). The mEPSC rate of nlp-12 and ckr-2 single mutants is indistinguishable from wild-type controls (Hu et al., 2011). Thus, inactivating NLP-12 and CKR-2 decreased mEPSC rate only when EGL-30 Gαq activity was enhanced (in eat-16 mutants). Second, we analyzed aldicarb-induced paralysis of double mutants. As previously reported, both ckr-2 and egl-30 single mutants are resistant to aldicarb-induced paralysis; however, additive effects on aldicarb sensitivity were not observed in egl-30; ckr-2 double mutants (Fig. 2C). Together, these results support the idea the NLP-12, CKR-2, and EGL-30 Gαq act together to potentiate ACh release.
Aldicarb-potentiated tonic release is mediated by UNC-13L
Activation of EGL-8 PLCβ stimulates PIP2 hydrolysis, producing the second messengers DAG and IP3. Prior studies suggested that DAG binding to UNC-13/Munc13 promotes ACh release in C. elegans and glutamate release in rodent neurons (Lackner et al., 1999; Rhee et al., 2002). For this reason, we tested the idea that UNC-13 is required for aldicarb-induced potentiation of ACh release.
The unc-13 gene encodes two isoforms (UNC-13S and L), which have different N-terminal domains but share a 1200 aa C-terminal domain (Kohn et al., 2000; Hu et al., 2013). To determine which UNC-13 protein is required for NLP-12's effects on tonic release, we analyzed aldicarb potentiation of mEPSC rate in unc-13(s69) mutants that express either UNC-13L or S transgenes (Fig. 3). Aldicarb-treatment increased the mEPSC rate of UNC-13L-rescued animals and wild-type controls to similar levels, and this effect was eliminated by a mutation (H699K) that disrupts DAG binding to UNC-13L (Fig. 3A–C; Betz et al., 1998). In contrast, aldicarb had no effect on the mEPSC rate of UNC-13S-rescued animals (Fig. 3A–C). Thus, NLP-12's effects on tonic release require DAG binding to UNC-13L.
Aldicarb potentiation of tonic release required DAG binding to UNC-13L but not UNC-13S. Aldicarb potentiated tonic release in animals expressing UNC-13L but not in those expressing UNC-13S. Potentiation of tonic release was blocked by a mutation that prevents DAG binding to UNC-13L (H699K). UNC-13S or UNC-13L transgenes were expressed in unc-13(s69) mutants. mEPSCs were recorded from adult body wall muscle of the indicated genotypes, with (blue) and without (black) a 60 min aldicarb treatment. Representative traces (A), average mEPSC rates (B), and aldicarb potentiation of mEPSC rates (C) are shown. Statistically significant differences (***p < 0.001 and ns, not significant) and the number of animals analyzed are indicated. Error bars indicate SEM.
Aldicarb potentiates evoked release mediated by both UNC-13L and S
Which UNC-13 protein potentiates evoked release? Transgenes expressing either UNC-13S or UNC-13L partially rescue the baseline-evoked EPSC defect of unc-13(s69) mutants (Fig. 4A,B), consistent with our prior study (Hu et al., 2013). Partial rescue is expected because the wild type-evoked response is a composite of both UNC-13S and UNC-13L-mediated release (Hu et al., 2013). Aldicarb potentiation of evoked release in UNC-13S- and UNC-13L-rescued animals was similar to that in wild-type controls (Fig. 4A–C). Potentiation of evoked responses was eliminated by C1 domain mutations that disrupt DAG binding to UNC-13S (H348K) or UNC-13L (H699K; Betz et al., 1998; Nurrish et al., 1999; Fig. 4A–C). Thus, NLP-12 potentiates evoked release via DAG binding to both UNC-13 proteins.
DAG binding to UNC-13S and L potentiated evoked release. Aldicarb-potentiated evoked release was analyzed in animals expressing UNC-13S or UNC-13L. For both UNC-13 proteins, potentiation of evoked release was blocked by mutations that prevent DAG binding. UNC-13 transgenes were expressed in unc-13(s69) mutants. Averaged evoked responses (A) and summary data (B, C) are shown. D, The time course of charge transfer during evoked responses is compared. The number of animals analyzed is indicated for each genotype. Statistically significant differences (***p < 0.001, **p < 0.01, *p < 0.05, and ns, not significant) and the number of animals analyzed are indicated. Error bars indicate SEM.
We previously showed that the kinetics of evoked release mediated by UNC-13L is faster than that mediated by UNC-13S (Hu et al., 2013). To determine whether aldicarb alters release kinetics, we analyzed the charge transfer kinetics of evoked responses. Aldicarb treatment significantly slowed the charge transfer of UNC-13L-mediated evoked responses and this effect was eliminated by the UNC-13L(H699K) mutation (Fig. 4D). In contrast, aldicarb had no effect on the kinetics of UNC-13S-mediated evoked responses (Fig. 4D). These results suggest that DAG binding to UNC-13L slows ACh release during evoked responses.
Discussion
Here we show that NLP-12 stimulates ACh release via DAG binding to UNC-13S and UNC-13L. These results identify an endogenous neuromodulator that potentiates release through the PLCβ/UNC-13 pathway. UNC-13/Munc13 proteins have been implicated in phorbol ester effects on secretion and in several forms of short-term plasticity. Our results suggest that GPCRs coupled to Gαq, and in particular neuropeptide receptors, potentiate transmission through changes in DAG liganding of UNC-13. Consistent with this idea, CCK has been shown to stimulate glutamate release in hippocampal neurons via a Gαq-coupled receptor (Breukel et al., 1997; Deng et al., 2010). Because the NLP-12 receptor (CKR-2) is most similar to mammalian CCK receptors (Janssen et al., 2008), our results suggest that potentiation of neurotransmitter release by NLP-12/CCK-like neuropeptides is ancient and likely acts via changes in UNC-13 activity. Neuropeptides are broadly expressed in the brain in both vertebrate and invertebrate animals. Consequently, neuropeptide regulation of UNC-13/Munc13 proteins provides a potential mechanism for modulating circuit function and behavioral states.
Prior studies showed that different Munc13 isoforms mediate different forms of short-term plasticity in rodent neurons (Rosenmund et al., 2002). Following high-frequency stimulus trains, synapses relying on Munc13-1 are depressed whereas those using Munc13-2 are potentiated. UNC-13 proteins containing the C2A domain (UNC-13L, Munc13-1, and ubMunc13-2) mediate a fast form of release whereas UNC-13 proteins lacking the C2A domain (UNC-13S and bMunc13-2) mediate slow release (Chen et al., 2013; Hu et al., 2013; Zhou et al., 2013). Thus, differential expression of Munc13 isoforms endows synapses with different patterns of release, and different forms of plasticity.
Here we extend these studies by showing that NLP-12 potentiates tonic and evoked ACh release by distinct mechanisms. In particular, our results indicate that different UNC-13 isoforms regulate tonic (UNC-13L) and evoked (both UNC-13L and S) release. These results provide genetic evidence that neuromodulators engage different UNC-13 proteins to regulate different forms of release.
Aldicarb caused UNC-13L-mediated evoked release to become significantly slower and more prolonged, whereas it had no effect on UNC-13S release kinetics. These results suggest that DAG binding to UNC-13L loosens the coupling of primed SVs to the calcium channel driving release. This could be mediated by altered binding of UNC-13L to UNC-10/RIM or by changes in the kinetics of calcium binding to UNC-13L primed SVs. Further experiments are required to distinguish between these possibilities. These results suggest that neuromodulators like NLP-12 provide a means to adjust release kinetics.
NLP-12 potentiation of tonic and evoked release also differed in their sensitivity to egl-30 Gαq and egl-8 PLCβ mutations. The egl-30(ad106) mutation is a partial loss of function; consequently, residual aldicarb potentiation of tonic release in this mutant could be mediated by residual EGL-30 activity. It is not possible to test the effect of egl-30-null mutations, as these mutants are not viable (Brundage et al., 1996). Alternatively, the residual potentiation of tonic release in egl-30 mutants could be mediated by other Gα-subunits. Further experiments are required to distinguish between these possibilities. DAG binding to UNC-13L is required for potentiation of tonic release, as this effect is blocked by the UNC-13L (H699K) mutation. Nonetheless, EGL-8 PLCβ-null mutations had no effect on aldicarb-induced potentiation of mEPSC rate. These results imply that DAG produced by other phospholipases potentiates tonic release.
How are tonic and evoked release differentially regulated? Aldicarb potentiates tonic release mediated by UNC-13L but not by UNC-13S. UNC-13L has a C2A domain that is absent in UNC-13S (Kohn et al., 2000; Hu et al., 2013). The C2A domain binds UNC-10/RIM (Lu et al., 2006), localizing UNC-13L to the center of the active zone (adjacent to the dense projection; Weimer et al., 2006). UNC-13 proteins lacking C2A exhibit a more diffuse presynaptic distribution (Chen et al., 2013; Hu et al., 2013; Zhou et al., 2013). RIM proteins bind voltage-activated calcium channels (CaVs) thereby concentrating CaV channels at active zones (Han et al., 2011; Kaeser et al., 2011; Graf et al., 2012; Müller et al., 2012). Thus, our results suggest that aldicarb selectively promotes tonic release of SVs that are adjacent to presynaptic CaV channels.
Several prior studies support the idea that different forms of release are mediated by distinct sets of synaptic proteins. Mouse DOC2 is required for spontaneous but not evoked neurotransmitter release (Groffen et al., 2010; Pang et al., 2011). Inactivating synaptotagmin I or complexin decreases synchronous-evoked release but enhances spontaneous neurotransmitter release (DiAntonio and Schwarz, 1994; Littleton et al., 1994; Pang et al., 2006; Hobson et al., 2011; Martin et al., 2011). Collectively, these results suggest that the different patterns of release and different forms of synaptic plasticity are dictated by the expression and function of distinct synaptic proteins.
Footnotes
This work was supported by a National Institutes of Health research grant (GM54728 to J.K.). We thank members of the Kaplan lab for critical comments on this manuscript. Some strains were provided by the C. elegans Genetics Stock Center.
The authors declare no competing financial interests.
- Correspondence should be addressed to Joshua M Kaplan, Department of Molecular Biology, Massachusetts General Hospital–Simches Research Building, 7th Floor, 185 Cambridge Street, Boston MA 02114. kaplan{at}molbio.mgh.harvard.edu
