Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2013 Sep 6.
Published in final edited form as: Neuron. 2012 Sep 6;75(5):838–850. doi: 10.1016/j.neuron.2012.06.038

The SOL-2/Neto Auxiliary Protein Modulates the Function of AMPA-Subtype Ionotropic Glutamate Receptors

Rui Wang 1, Jerry E Mellem 1, Michael Jensen 1, Penelope J Brockie 1, Craig S Walker 1, Frédéric J Hoerndli 1, David M Madsen 1, Andres V Maricq 1,*
PMCID: PMC3458792  NIHMSID: NIHMS395814  PMID: 22958824

Summary

The neurotransmitter glutamate mediates excitatory synaptic transmission by gating ionotropic glutamate receptors (iGluRs). AMPA receptors (AMPARs), a subtype of iGluR, are strongly implicated in synaptic plasticity, learning and memory. We previously discovered two classes of AMPAR auxiliary proteins in C. elegans that modify receptor kinetics and thus change synaptic transmission. Here, we have identified another auxiliary protein, SOL-2, a CUB-domain protein that associates with both the related auxiliary subunit SOL-1 and with the GLR-1 AMPAR. In sol-2 mutants, behaviors dependent on glutamatergic transmission are disrupted, GLR-1-mediated currents are diminished, and GLR-1 desensitization and pharmacology are modified. Remarkably, a secreted variant of SOL-1 delivered in trans can rescue sol-1 mutants and this rescue depends on in cis expression of SOL-2. Finally, we demonstrate that SOL-1 and SOL-2 have an ongoing role in the adult nervous system to control AMPAR-mediated currents.

Introduction

The AMPA class of iGluRs is intensely studied because of the critical role these receptors play in excitatory neurotransmission and nervous system function. For example, experience dependent changes in AMPAR properties and number are mechanistically linked to learning and memory (Kerchner and Nicoll, 2008; Kessels and Malinow, 2009). Although glutamate-gated currents can be recorded from heterologous cells that express vertebrate AMPAR subunits, recent studies have conclusively demonstrated that these reconstituted currents are significantly different from native neuronal currents (Jackson and Nicoll, 2011). Neuronal AMPARs associate with multiple classes of transmembrane proteins, which serve important auxiliary functions. Some of the auxiliary proteins function as chaperones; but all have some effect on the kinetics and pharmacology of AMPAR gating, thereby providing additional mechanisms for changes in synaptic strength. The first identified auxiliary subunits were the TARPs (Transmembrane AMPAR Regulatory Proteins) (Chen et al., 2000; Milstein and Nicoll, 2008). This was followed by genetic studies in C. elegans that identified and characterized SOL-1, a CUB-domain transmembrane protein that defined a second class of AMPAR auxiliary protein (Zheng et al., 2004). C. elegans also expresses the TARP proteins STG-1 and STG-2, which have evolutionarily conserved functions (Walker et al., 2006a; Wang et al., 2008). Since then, additional transmembrane proteins have been implicated in AMPAR function (Jackson and Nicoll, 2011; Kalashnikova et al., 2010; Schwenk et al., 2012; Schwenk et al., 2009; von Engelhardt et al., 2010). Two outstanding questions are posed by these studies. First, are there additional auxiliary proteins that contribute to receptor function? Second, how do the auxiliary proteins contribute to synaptic transmission and behavior?

In C. elegans, the GLR-1 AMPAR mediates glutamate-gated current in a subset of interneurons that control movement and the avoidance of noxious stimuli (Hart et al., 1995; Maricq et al., 1995). Genetic and reconstitution studies have demonstrated that GLR-1 is part of a multi-protein, synaptic complex required for GLR-1-mediated currents and behavior (Wang et al., 2008; Zheng et al., 2004). In addition to GLR-1, this complex contains SOL-1 and at least one of the TARPs, i.e., STG-1 and STG-2. SOL-1 is an evolutionarily conserved type I transmembrane protein that contains protein-protein interaction motifs called CUB-domains (Complement, Urchin EGF, BMP). SOL-1 was shown to regulate the rate of GLR-1 desensitization as well as its rate of recovery from desensitization (Walker et al., 2006a; Walker et al., 2006b; Zheng et al., 2006).

More recently, the CUB-domain-containing transmembrane proteins Neto1 and Neto2 were identified in mice. These proteins contribute to signaling mediated by NMDA (N-methyl-D-aspartate) and kainate iGluRs, respectively (Ng et al., 2009; Zhang et al., 2009). However, C. elegans SOL-1 and the vertebrate Neto proteins belong to two different classes of CUB domain proteins. Whereas SOL-1 contains four predicted CUB domains, Neto1 and Neto2 contain two CUB domains and a LDLa domain (low-density lipoprotein receptor class A). This raises the question of whether multiple classes of CUB-domain proteins contribute to the function of specific iGluRs.

In an earlier study (Zheng et al., 2006), we found that a secreted form of SOL-1 (s-SOL-1) that lacked the transmembrane domain was sufficient to rescue the behavioral and synaptic signaling defects of sol-1 mutants. Here, we show that co-expression of s-SOL-1 with GLR-1 and STG-1 in heterologous cells is not sufficient to reconstitute glutamate-gated current. This result led us to the hypothesis that an additional protein, which was missing in heterologous cells, is expressed in neurons and is required for s-SOL-1 function. Presumably, this protein is part of the GLR-1 receptor complex and recruits s-SOL-1 to the complex thus contributing to receptor function. This model also suggests that the protein itself might have a modulatory role in GLR-1 function.

To identify the putative interacting protein, we used an unbiased forward genetic strategy and discovered SOL-2, a CUB-domain transmembrane protein that is the homologue of the vertebrate Neto proteins, with 2 CUB-domains and a LDLa-domain. As predicted, we found that s-SOL-1 function was dependent on SOL-2 and that SOL-2 associates with the GLR-1 signaling complex. We show that surface delivery of GLR-1 and SOL-1 occurs in the absence of SOL-2; however, the stability or function of the complex appears compromised in sol-2 mutants. In sol-1 mutants, the remaining components of the GLR-1 complex are also delivered to the postsynaptic membrane, indicating that SOL-1 does not have an essential role in assembly or trafficking of the signaling complex. We demonstrate that GLR-1-mediated currents depend on both SOL-1 and SOL-2, and that currents in sol-1 and sol-2 mutants can be rescued in adults, thus demonstrating an ongoing role for these CUB-domain proteins in synaptic transmission.

Remarkably, we found that the extracellular domain of SOL-1 secreted in trans is sufficient to rescue glutamate-gated currents in sol-1 mutants. This rescue depends on in cis expression of SOL-2. Finally, we show that glutamate- and kainate-gated currents are differentially disrupted in sol-1 and sol-2 mutants, and that SOL-2 contributes to the kinetics of receptor desensitization. In summary, our results demonstrate that SOL-2 is an essential component of GLR-1 AMPAR complexes at synapses, and contributes to synaptic transmission and behaviors dependent on glutamatergic signaling.

Results

sol-2 encodes a CUB-domain protein required for AMPAR-mediated signaling

AVA interneurons in C. elegans are part of a locomotory control circuit that primarily regulates the direction of a worm's movement. These interneurons receive glutamatergic synaptic inputs and express GLR-1, STG-2 and SOL-1 – essential transmembrane proteins that contribute to a postsynaptic iGluR signaling complex (Brockie et al., 2001a; Maricq et al., 1995; Wang et al., 2008; Zheng et al., 2004). Using in vivo patch-clamp electrophysiology, we recorded rapidly activating and desensitizing currents in wild-type worms in response to pressure application of glutamate (Figure 1A). In sol-1 mutants, glutamate-gated currents rapidly desensitize and consequently we cannot measure the currents using conventional drug application (Figure 1A) (Walker et al., 2006b). A secreted form of SOL-1 that lacks the transmembrane domain (s-SOL-1) can partially rescue the glutamate-gated current when expressed in the AVA neurons of transgenic sol-1 mutants (Figure 1A) (Zheng et al., 2006). This result suggested that s-SOL-1 formed a functional complex with GLR-1 and STG-2.

Figure 1.

Figure 1

Open in a new tab

sol-2 encodes a CUB-domain transmembrane protein that is required for s-SOL-1 function. (A) Currents in response to 3 mM glutamate in the AVA neuron of a wild-type worm, a sol-1 mutant, and a transgenic sol-1 mutant that expressed s-SOL-1 in AVA. (B) Glutamate-gated current in muscle cells of transgenic worms (left) and Xenopus oocytes (right) that expressed GLR-1 and STG-1 with either SOL-1 or s-SOL-1. (C) Reversal frequency in transgenic wild-type worms, sol-1 and sol-2 single and double mutants, and mutants that overexpressed specific auxiliary proteins. All worms expressed the GLR-1(A687T) lurcher variant of GLR-1. n=10 for all genotypes. *, significantly different from wild-type, p<0.001. **, p<0.001. (D) Domain organization of transmembrane CUB-domain proteins from C. elegans (SOL-1, SOL-2 and LEV-10) and vertebrates (Neto1 and Neto2). (E and F) Glutamate-gated currents recorded in Xenopus oocytes (E) and the muscle cells of transgenic worms (F). All Cells were voltage-clamped at –60 mV.

See also Figure S1.

To test sufficiency of s-SOL-1, we asked whether we could record glutamate-gated currents from muscle cells that co-expressed GLR-1, STG-1 and s-SOL-1. Muscle cells in C. elegans do not express any known iGluRs, STGs or SOL-1 proteins, and thus are ideal for reconstitution studies. We reliably recorded large, rapidly activating inward currents in response to pressure application of glutamate when full-length SOL-1, STG-1 and GLR-1 were co-expressed in muscle cells (Figure 1B). In contrast, we were unable to record appreciable currents in cells that expressed s-SOL-1 instead of full-length SOL-1 (Figure 1B). We found similar results in co-expression studies with STG-2 (data not shown). We next addressed whether the failure to reconstitute function with s-SOL-1 was specific to muscle cells by reconstitution experiments in Xenopus oocytes. Again, we were able to measure large glutamate-gated currents when full-length SOL-1 was co-expressed with GLR-1 and STG-1 (Figure 1B), but not when s-SOL-1 replaced SOL-1. These results led us to hypothesize that neurons, but not muscle cells or Xenopus oocytes, express a protein that binds to s-SOL-1 and contributes to the function of the GLR-1 complex.

To identify this interacting protein, we turned to a genetic approach that took advantage of the hyper-reversal behavior of transgenic worms that express a gain-of-function variant of GLR-1 (GLR-1(A687T)) (Zheng et al., 1999). The hyper-reversal behavior of these “lurcher” worms is suppressed by mutations in sol-1 and rescued in transgenic sol-1; lurcher mutants that express either full-length SOL-1 or s-SOL-1 (Figure 1C) (Zheng et al., 2006). We hypothesized that mutating the protein predicted to interact with SOL-1 (and s-SOL-1) would also suppress the hyper-reversal phenotype. We therefore mutated lurcher worms, screened approximately 2000 haploid genomes and identified a single mutant, sol-2(ak205), which partially suppressed the hyper-reversal phenotype (Figure 1C). The ak205 mutation complemented mutations in sol-1, stg-1 and stg-2 (data not shown) indicating that we had mutated a new gene required for signaling mediated by the GLR-1 complex.

Using conventional strategies, we mapped the mutation to a small interval on LG I (Figure S1A). We identified an open reading frame (K05C4.11) that rescued the movement of transgenic sol-2; lurcher mutants (Figure 1C). Unlike the case for sol-1; lurcher mutants, s-SOL-1 did not restore hyper-reversal behavior in sol-1; sol-2 double mutants that expressed lurcher. However, s-SOL-1 did rescue hyper-reversal behavior when co-expressed with SOL-2 in transgenic sol-1; sol-2 double mutants suggesting that the function of s-SOL-1 is dependent on SOL-2 (Figure 1C).

By sequencing the genome of the sol-2 mutant, we identified a mutation in the K05C4.11 (sol-2) gene that causes a frame shift and an early stop, suggesting that the mutation is a null (Figure S1B). The sol-2 gene is predicted to encode a 436 amino acid, type I transmembrane protein containing two putative CUB domains and a LDLa domain. The SOL-2 protein has closest sequence identity (approximately 20-21%) to the vertebrate Neto proteins, and significant identity to the C. elegans CUB domain proteins SOL-1 and LEV-10 (Gally et al., 2004; Zheng et al., 2004) (Figure 1D; Figure S1B). Following our mapping experiments, we discovered an existing mutation in sol-2 produced by a deletion (ok1713) (Wormbase.org). sol-2(ok1713) also suppressed the hyper-reversal behavior of lurcher worms, and did not complement the sol-2(ak205) mutation (data not shown).

We next addressed whether SOL-2 was the missing protein required for reconstitution of glutamate-gated current using s-SOL-1. We found that we could reconstitute glutamate-gated currents in Xenopus oocytes or C. elegans muscle cells when s-SOL-1 was co-expressed with SOL-2, STG-1 and GLR-1 (Figure 1E and F), but not in the absence of SOL-2 (Figure 1B). Thus, s-SOL-1 function was dependent on SOL-2. Furthermore, SOL-2 cannot simply replace SOL-1 given that we were unable to reconstitute glutamate-gated current in either oocytes or muscle cells by co-expressing GLR-1, STG-1 and SOL-2 (Figure 1E and F).

GLR-1-mediated behaviors and current are disrupted in sol-2 mutants

Our reconstitution studies demonstrated that SOL-2 and SOL-1 contribute to the function of the GLR-1 signaling complex. In addition, our finding that mutations in sol-2 disrupt the behavior of transgenic lurcher mutants (Figure 1C) predicts that glutamatergic neurotransmission is disrupted in sol-2 mutants. Thus, we evaluated the behavior of sol-2 mutants using two standard assays that depend on GLR-1 function (Hart et al., 1995; Maricq et al., 1995; Mellem et al., 2002). When tested in an osmotic avoidance assay the sol-2 mutants were as slow to recoil from the hyper-osmotic stimuli as glr-1 or sol-1 mutants (Figure 2A). When tested in a touch-avoidance assay (nose touch response) sol-2 mutants were significantly impaired, but not to the extent of glr-1 or sol-1 mutants (Figure 2B). In both assays, sol-1; sol-2 double mutants were no more impaired than sol-1 mutants alone, suggesting that the two gene products act in the same pathway.

Figure 2.

Figure 2

Open in a new tab

SOL-2 is required for GLR-1 dependent behaviors and glutamate-gated current. (A) Quantification of the delay in response to hyper-osmotic stimuli. Wild type, n=41; glr-1, n=21; sol-1, n=9; sol-2, n=24; sol-1; sol-2, n=11; sol-2 rescue, n=15. *, significantly different from wild-type, p<0.01. (B) The response to nose touch stimulation (percentage response of 10 trials per worm). Number of worms, wild-type, n=10; glr-1, n=10; sol-1, n=5; sol-2, n=10; sol-1; sol-2, n=5; sol-2 rescue, n=8. *, significantly different from wild-type, p<0.05. (C) Currents in response to 3 mM glutamate in AVA interneurons of wild-type worms, sol-2 mutants, and transgenic sol-2 mutants that expressed SOL-2::GFP in AVA. (D) Average peak glutamate-gated current in wild-type worms (n=19), sol-1 mutants (n>10), sol-2 mutants (n=10), and transgenic sol-2 mutants that expressed SOL-2::GFP (n=7). § indicates that rapidly activating currents were not detected in sol-1 mutants. *, significantly different from wild-type, p<0.001.

See also Figure S2.

The peak amplitude of the glutamate-gated current in AVA was considerably diminished in sol-2 mutants, and we could only measure a small, rapidly activating and desensitizing current (Figure 2C and D). These currents are distinct from those recorded in sol-1 mutants where we could not detect a rapidly activating inward current under the same recording conditions (Figure 1A; Figure 2D). Only the GLR-1-mediated current was decreased in sol-2 mutants; the slower, outwardly rectifying current is mediated by NMDA receptors (Brockie et al., 2001b) and did not appear appreciably different than wild-type current (Figure 2C). Glutamate-gated currents in the AVA neurons of transgenic sol-2 mutants were rescued by a functional SOL-2::GFP fusion protein that was specifically expressed in AVA using the rig-3 promoter (Feinberg et al., 2008) (Figure 2C and D). We were also able to rescue current in transgenic sol-2 mutants that expressed GFP fused to the extracellular N-terminus of full-length SOL-2 (GFP::SOL-2; Figure S2). These results demonstrate that GFP-tagged SOL-2 is functional and acts cell autonomously. However, unlike the case for SOL-1, we did not observe rescue of transgenic sol-2 mutants that expressed a secreted variant of the fusion protein that lacked the transmembrane domain (GFP::s-SOL-2) (Figure S2).

SOL-2 associates with the GLR-1 signaling complex

Based on our electrophysiological studies, as well as the known expression pattern of GLR-1 and SOL-1 (Brockie et al., 2001a; Zheng et al., 2004), we predicted that SOL-2 would be expressed in the command interneurons. We therefore used confocal microscopy to determine the cellular and subcellular distribution of SOL-2. The sol-2 promoter drives expression of GFP in many head and tail neurons, including neurons that express the GLR-1 subunit, as well as the SOL-1 auxiliary subunit (Figure S3A) (Brockie et al., 2001a; Zheng et al., 2004). SOL-2 is also expressed in neurons that do not express either GLR-1 or SOL-1. With respect to avoidance behavior and locomotion, sol-1; sol-2 double mutants are no more severe than the sol-1 single mutant (Figure 2A and B), indicating that the role SOL-2 plays in these neurons is not directly relevant to these behaviors. We have not investigated whether SOL-2 contributes to the function of additional GLR receptors (Brockie et al., 2001a) or other behaviors. Importantly, SOL-2 is expressed in the command interneurons, as shown by co-expression of mCherry driven by the nmr-1 promoter (Figure S3A).

To determine the subcellular localization of SOL-2, we imaged transgenic worms that co-expressed SOL-2::GFP with GLR-1::mCherry in AVA and found that SOL-2 co-localizes with GLR-1 (Figure 3A). To test whether SOL-2 also co-localizes with SOL-1, we co-expressed GFP::SOL-1 and SOL-2::mCherry in AVA and observed GFP and mCherry puncta that co-localized along the length of the AVA processes (Figure 3B).

Figure 3.

Figure 3

Open in a new tab

SOL-2 associates with the GLR-1 signaling complex. (A and B) Confocal images of the ventral nerve cord in transgenic worms that expressed either SOL-2::GFP and GLR-1::mCherry (A), or GFP::SOL-1 and SOL-2::mCherry (B) in the AVA interneurons. Arrows indicate co-localized GFP and mCherry puncta. (C) SOL-1/SOL-2 BiFC fluorescence and GLR-1::mCherry in the AVA processes. (D) GLR-1/SOL-2 BiFC fluorescence and soluble mCherry in AVA processes. (E) Confocal image showing the absence of NMR-2/SOL-2 BiFC fluorescence in the AVA processes of transgenic worms that also expressed GLR-1::mCherry. (F and G) Anti-HA (red) and GFP fluorescence in the ventral nerve cord of transgenic wild-type worms (F) or sol-2 mutants (G) that expressed HA::GLR-1::GFP. (H and I) SEP::GLR-1 (H) and SEP::SOL-1 (I) fluorescence in the AVA processes of transgenic wild-type worms and sol-2 mutants. Scale bars represent 5 μm.

See also Figure S3 and S4.

The co-localization of SOL-2 with both SOL-1 and GLR-1 suggested that SOL-2 was part of the GLR-1/SOL-1 complex (Walker et al., 2006a). To address this possibility, we used BiFC (Bimolecular Fluorescence Complemenation) to probe possible protein interactions. We tagged SOL-1 with the N-terminal half of the fluorescent protein Venus (a YFP variant) (N-YFP::SOL-1) (Chen et al., 2007; Shyu et al., 2008), and SOL-2 with the C-terminal half (C-YFP::SOL-2) and used the rig-3 promoter to express these constructs along with GLR-1::mCherry in the AVA neurons (Kano et al., 2008). We observed punctate SOL-1/SOL-2 BiFC fluorescence that co-localized with GLR-1::mCherry puncta along the length of the AVA processes in transgenic worms (Figure 3C). We found only minor effects of the BiFC constructs on glutamate-gated current (Figure S3B) and GLR-1::mCherry puncta (Figure S3C), and the intensity of the BiFC signal was somewhat decreased in glr-1 mutants (Figure S3D). We also observed BiFC fluorescence when C-YFP::SOL-2 was co-expressed in AVA with N-YFP::GLR-1 (Figure 3D). No fluorescence signal was detected when N-YFP::SOL-1, C-YFP::SOL-2, or N-YFP::GLR-1 was expressed alone (data not shown). These results indicate that SOL-2 is in close proximity to SOL-1 and GLR-1 given that BiFC interactions are limited primarily by the length and flexibility of the proteins and linkers (Kerppola, 2006).

To address whether SOL-1 and SOL-2 specifically associate with GLR-1 receptors, we co-expressed C-YFP::SOL-2 with both N-YFP::NMR-2 (NMDA-type iGluR subunit) (Kano et al., 2008) and GLR-1::mCherry in AVA of transgenic worms. We did not observe any BiFC fluorescence in these worms indicating that NMR-2 and SOL-2 do not associate (Figure 3E). This result is consistent with earlier studies showing that neither GLR-1 nor SOL-1 co-localize with NMDARs (Brockie et al., 2001b; Zheng et al., 2004).

The BiFC data suggest a model in which SOL-1 and SOL-2 directly interact. To more rigorously test this hypothesis, we asked whether we could detect SOL-1 and GLR-1 after immunoprecipitation of SOL-2. Because of technical limitations (low abundance of auxiliary proteins and receptors in whole worms), we could not reliably detect these proteins in whole worm lysates. Therefore, we co-expressed these proteins in HEK 293 cells and used these transfected cells for biochemical studies. We found that SOL-2 was associated with GLR-1 and SOL-1 (Figure S4A), but did not associate with the unrelated transmembrane protein DCC (Keino-Masu et al., 1996). Together, the biochemical and BiFC data indicate that SOL-1, SOL-2 and GLR-1 physically interact and form a receptor complex.

GLR-1 is expressed at synapses in sol-2 mutants

We had previously demonstrated that GLR-1 surface expression was not appreciably altered in sol-1 mutants (Zheng et al., 2004). Is surface expression of GLR-1 also independent of sol-2? To test this possibility, we double-labeled GLR-1 with a HA-epitope (extracellular) and GFP (intracellular) and expressed this functional construct (HA::GLR-1::GFP) in transgenic worms (Zheng et al., 2004). We assessed the surface expression of GLR-1 by injecting fluorescently labeled anti-HA antibodies into the pseudocoelomic space of transgenic wild-type worms and sol-2 mutants (Gottschalk et al., 2005; Zheng et al., 2004). We observed punctate anti-HA antibody fluorescence along the ventral cord in both wild-type worms (Figure 3F) and sol-2 mutants (Figure 3G) suggesting that GLR-1 is expressed on the cell surface in sol-2 mutants.

We obtained additional evidence to support the claim that GLR-1 surface expression was not appreciably altered in sol-2 mutants by generating a transgenic strain that expressed GLR-1 fused to superecliptic phluorin (SEP, a pH-sensitive variant of GFP) (Miesenbock et al., 1998). SEP was inserted in the extracellular N-terminal domain four amino acids from the mature N-terminus (SEP::GLR-1). In agreement with our antibody studies, we found that the fluorescence intensity of surface expressed SEP::GLR-1 appeared similar in transgenic sol-2 mutants and wild-type worms (Figure 3H). Control experiments (acid wash) indicated that the SEP fluorescence signal represented surface receptors (Figure S4B).

Another possible explanation for the reduced glutamate-gated current in sol-2 mutants is reduced surface delivery of SOL-1. However, we also found no appreciable difference in fluorescence intensity when we compared SEP::SOL-1 surface delivery in transgenic wild-type worms and sol-2 mutants (Figure 3I). Thus, SOL-2 does not appear to have a significant role in the surface delivery of these proteins.

In trans delivery of s-SOL-1 rescues sol-1 mutants

Our reconstitution experiments also allowed us to begin to address the requirements for the delivery and maintenance of the GLR-1 signaling complex. One possibility is that SOL-1 might have an obligate chaperone role for some critical component of the complex, much like that suggested for a subset of the vertebrate TARPs (Milstein and Nicoll, 2008). In this scenario, the delivery or stability of components of the complex would be compromised in the absence of SOL-1. Alternatively, the components might reside stably at the membrane. To help distinguish between these possibilities, we expressed GFP-tagged secreted s-SOL-1 (GFP::s-SOL-1) in muscle cells of transgenic sol-1 mutants using the myo-3 muscle specific promoter. If GLR-1, STGs, SOL-2 and any other necessary components of the complex are stably delivered to the postsynaptic membrane in the absence of SOL-1, then one might predict that s-SOL-1 delivered in trans from muscle cells in transgenic sol-1 mutants would be sufficient to restore a functional signaling complex in AVA.

We first examined whether GFP::s-SOL-1 delivered in trans from muscle cells was co-localized with GLR-1 in the processes of the AVA interneurons. We co-expressed muscle-secreted GFP::s-SOL-1 and AVA-specific GLR-1::mCherry in transgenic mutants. We found that GFP::s-SOL-1 and GLR-1::mCherry co-localized at puncta along the length of the AVA processes in sol-1 mutants (Figure 4A), but not in sol-1; sol-2 double mutants (Figure 4B). We also observed GFP puncta along the AVA processes when muscle secreted GFP::s-SOL-1 was expressed in the absence of the GLR-1::mCherry transgene (Figure S5A). This result indicates that localization of s-SOL-1 to the ventral cord does not require overexpression of GLR-1 or other components of the signaling complex. This localization was also dependent on SOL-2 and thus GFP::s-SOL-1 was not observed along the ventral cord in sol-1; sol-2 double mutants (Figure S5B). We also found that the hyper-reversal movement of sol-1; lurcher mutants was rescued by muscle secreted GFP::s-SOL-1 and that the rescue was dependent on SOL-2 (Figure 4C).

Figure 4.

Figure 4

Open in a new tab

sol-1 mutants can be rescued by in trans delivery of s-SOL-1. (A and B) Confocal images of the ventral nerve cord in transgenic sol-1 mutants (A) and sol-1; sol-2 double mutants (B) that expressed GLR-1::mCherry in the AVA interneurons and GFP::s-SOL-1 secreted by muscle cells. (C) Reversal frequency in transgenic wild-type worms and sol-1 mutants, and transgenic sol-1 mutants or sol-1; sol-2 double mutants with GFP::s-SOL-1 secreted by muscle cells. All worms expressed the GLR-1(A687T) lurcher variant of GLR-1. n=10 for all genotypes; *, significantly different from wild-type, p<0.001. (D) Currents in response to 3 mM glutamate in the AVA neuron of a wild-type worm, sol-1 mutant, sol-1; sol-2 double mutant, a transgenic sol-1 or sol-1; sol-2 mutant with s-SOL-1 secreted by muscle cells, and a transgenic sol-1; sol-2 mutant with s-SOL-1 secreted by muscle cells and SOL-2 expressed specifically in the AVA interneurons. Cells were held at –60 mV. (E) Reversal frequency in transgenic wild-type worms, sol-1 mutants, sol-1; sol-2 double mutants, and transgenic mutants that carried a heat-shock inducible GFP::s-SOL-1 transgene (Phsp16-2::gfp::s-sol-1). All worms expressed the GLR-1(A687T) lurcher variant of GLR-1. (+) and (–) indicate the presence or absence of the transgene and when heat shock was used to induce GFP::s-SOL-1 expression. n=10 for all genotypes; *, p<0.001. (F and G) Reversal frequency (F) and glutamate-gated current (G) in wild-type worms and transgenic sol-2 mutants with a heat-shock inducible SOL-2::GFP transgene (Phsp::sol-2::gfp). n=10 for all genotypes; *, p<0.001.

See also Figure S5.

Our behavioral analysis suggested that s-SOL-1 provided in trans restored GLR-1-mediated signaling in the command interneurons. To more directly examine signaling, we measured glutamate-gated currents in AVA interneurons of sol-1 mutants, sol-1; sol-2 double mutants and transgenic mutants that expressed s-SOL-1 in muscle cells. In either sol-1 or sol-1; sol-2 double mutants we could not detect rapidly activating glutamate-gated currents. However, we found partial recovery of the current in transgenic sol-1 mutants that expressed s-SOL-1 (77.25 ± 28.31 pA, n=4), but not in transgenic sol-1; sol-2 mutants (n=3), indicating that the function of s-SOL-1 was dependent on SOL-2 in AVA interneurons (Figure 4D). We also found partial recovery of the current in transgenic sol-1; sol-2 double mutants that expressed s-SOL-1 in muscle cells and SOL-2 only in the AVA interneurons (96.33 ± 23.07 pA, n=6) (Figure 4D). These results indicate that the binding of exogenously delivered s-SOL-1 is sufficient to reconstitute the function of the receptor complex and that SOL-2 is required cell autonomously to recruit s-SOL-1 to the complex. Our results also suggest that in sol-1 mutants the remaining components of the receptor complex are stably located in the plasma membrane. Presumably, muscle-secreted s-SOL-1 diffuses in the extracellular space and binds to neuronal SOL-2 to reconstitute the GLR-1 receptor complex.

One might imagine that postsynaptic signaling molecules, such as SOL-1, have critical developmental roles in addition to their known signaling functions. Thus, the behavioral defects in sol-1 mutants might also be a consequence of developmental defects in synaptic morphology or function. To test whether SOL-1 has an obligate developmental role, we generated transgenic sol-1; lurcher and sol-1; sol-2; lurcher mutants that expressed GFP::s-SOL-1 under the control of a heat-shock inducible promoter (Phsp::gfp::s-sol-1). Four hours following heat shock of adult worms, we assessed their behavior. In the absence of the Phsp::gfp::s-sol-1 transgene, or in the absence of heat shock, sol-1; lurcher mutants did not reverse nearly as often as wild-type lurcher worms (Figure 4E). In contrast, heat shock induction of GFP::s-SOL-1 rescued reversal behavior in adult sol-1; lurcher mutants, and the rescue was dependent on sol-2 (Figure 4E).

In a complementary set of experiments, we examined whether heat shock driven expression of SOL-2::GFP in adult sol-2 mutants could similarly rescue the behavioral phenotype. Following heat shock induction, reversal behavior in adult sol-2; lurcher mutants was restored to wild-type values (Figure 4F). We extended these studies to examine glutamate-gated currents in heat-shocked worms and found that within four hours of heat shock we could record near wild-type glutamate-gated currents from the transgenic sol-2 mutants (Figure 4G). These experiments demonstrate that the function of s-SOL-1 is dependent on SOL-2, that the remaining components of the receptor signaling complex are stable in the absence of SOL-1, and that SOL-1 and SOL-2 have ongoing roles in synaptic transmission in the adult nervous system.

Glutamate-gated currents are significantly reduced in sol-2 mutants, yet paradoxically, SOL-2 is not essential for reconstitution studies in muscle cells or Xenopus oocytes (Figure 1) (Walker et al., 2006a; Walker et al., 2006b). One possibility is that the overexpression of SOL-1 in reconstitution studies partially compensates for the absence of SOL-2. This hypothesis predicts that overexpressing SOL-1 in sol-2 mutants should rescue glutamate-gated currents. Conversely, overexpressing SOL-2 should not rescue sol-1 mutants. As predicted, we found that current was restored in sol-1; sol-2 double mutants that overexpressed SOL-1, but not in double mutants that overexpressed SOL-2 (Figure 5A and B). The rescue was similar to that observed when SOL-1 was overexpressed in sol-1 single mutants (Figure 5A and B). Furthermore, overexpressing SOL-1 restored the hyper-reversal locomotion of transgenic sol-1; sol-2; lurcher worms, whereas overexpressing SOL-2 did not (Figure 5C). This was in contrast to overexpression of s-SOL-1, which was not sufficient to rescue sol-1; sol-2 mutants (Figure 1C). These data are in agreement with our studies in heterologous cells demonstrating that SOL-1, GLR-1 and STG-1 (or STG-2) constitute the minimal set of proteins required for reconstitution of glutamate-evoked currents (Figure 1) (Walker et al., 2006a; Walker et al., 2006b; Wang et al., 2008).

Figure 5.

Figure 5

Open in a new tab

Overexpression of SOL-1 partially compensates for the loss of SOL-2. (A) Currents in response to 3 mM glutamate in the AVA interneurons of wild-type worms, sol-1 mutants, sol-1; sol-2 double mutants, and transgenic mutants that overexpressed either SOL-1 or SOL-2. Cells were held at –60 mV. (B) Average peak glutamate-gated current in wild-type worms (n=19) and transgenic sol-1 mutants (n=5) or sol-1; sol-2 double mutants (n=4) that overexpressed SOL-1 in AVA. (C) Reversal frequency in transgenic wild-type and mutant worms that expressed the GLR-1(A687T) lurcher variant of GLR-1 with or without overexpression of SOL-1 or SOL-2. n=10 for all genotypes; *, significantly different from wild-type, p<0.01; **, p<0.001.

SOL-2 modifies the pharmacology and kinetics of GLR-1 receptors

Does SOL-2 directly modify the properties of the GLR-1 signaling complex in addition to its role in recruiting and stabilizing SOL-1 to the complex? To begin to address this question, we examined whether agonist gating of GLR-1 was altered in sol-2 mutants. AMPARs are gated open by the partial agonist kainate (Mayer and Armstrong, 2004), but the efficacy of this partial agonist depends on a number of factors, including the association of AMPARs with auxiliary subunits (Kato et al., 2010; Shi et al., 2010). Peak currents evoked by kainate were much smaller in sol-2 mutants than in wild type, indicating that kainate-gating of GLR-1 was also dependent on SOL-2 (Figure 6A and B). We also addressed the question of SOL-1 sufficiency and found that kainate-gated currents were rescued in sol-1 mutants that overexpressed SOL-1, but only marginally rescued in sol-1; sol-2 mutants (Figure 6C and D). This was in stark contrast to glutamate-gated currents, which were rescued in sol-1; sol-2 mutants that overexpressed SOL-1 (Figure 5A and B). These data indicate that not all of SOL-2's effects can be simply explained by recruitment or stabilization of SOL-1. Thus, SOL-2 also appears to directly contribute to the function of the GLR-1 complex.

Figure 6.

Figure 6

Open in a new tab

SOL-2 contributes to the pharmacology and kinetics of GLR-1. (A) Currents in response to 100 μM kainate (KA) in the AVA interneurons of wild-type worms, sol-2 mutants, and transgenic sol-2 mutants that overexpressed wild-type SOL-2 in AVA. (B) Average peak kainate-gated current in wild-type worms (n=14), sol-2 mutants (n=5) and transgenic sol-2 mutants that overexpressed SOL-2 in AVA (n=5). *, p<0.0001. (C) Kainate-gated current in a sol-1 mutant, a transgenic sol-1 mutant, or a sol-1; sol-2 double mutant that overexpressed GFP::SOL-1 in AVA. (D) Average peak kainate-gated current in transgenic sol-1 mutants (n=5) and sol-1; sol-2 mutants (n=9) that overexpressed GFP::SOL-1 in AVA. *, p<0.02. (E) Currents evoked in muscle cells in response to 3 mM glutamate and 100 μM kainate in transgenic worms that ectopically expressed various combinations of GLR-1, STG-1, SOL-1 and SOL-2 in muscles. (F) Currents measured in response to rapid application of 3 mM glutamate to an outside-out patch pulled from an AVA interneuron from a wild-type worm, a sol-1 mutant, or a sol-2 mutant. All cells were voltage-clamped at–60 mV.

We next asked whether we could reconstitute the GLR-1 complex by expression in C. elegans muscle cells. We found glutamate-gated currents of approximately the same magnitude and kinetics in transgenic muscles that expressed GLR-1, STG-1 and SOL-1, or these three proteins along with SOL-2 (Figure 6E). In contrast, we found that the presence of SOL-2 markedly reduced the magnitude of the kainate-gated current (three independent experiments). These experiments indicate that the receptor complex functions differently in muscles compared to neurons, perhaps because additional factors interact with the complex in neurons.

What explains the reduced glutamate-gated currents in sol-2 mutants? To better address the mechanism of SOL-2 function, we used rapid perfusion techniques to study glutamate-gated currents from outside-out membrane preparations. Because the receptor complex behaved differently in muscle, we chose to study patches from AVA. When voltage-clamped at –60 mV, the rapid application of glutamate evoked inward currents that desensitized in the continued presence of glutamate (Figure 6F). In wild type, the average tau of desensitization was approximately 4 ms (n=11). In contrast, we found that the current desensitized in less than 0.4 ms in sol-2 mutants (n=6), which is similar to what we observed in sol-1 mutants (Figure 6F) (Walker et al., 2006b). These rapid rates of desensitization distort the time-course of glutamate-gated currents, leading to a significant decrease in the peak current elicited by pressure application of agonist (Figure 2). Because the rate of desensitization in these mutants was faster than the rate of piezo-driven solution change, we could not determine whether sol-1 and sol-2 mutants exhibit different rates of desensitization.

To better address the functional effects of SOL-2, we turned to reconstitution of GLR-1 function in Xenopus oocytes. We recorded both glutamate- and kainate-gated currents from oocytes in which GLR-1 and STG-2 were co-expressed with either SOL-1, or both SOL-1 and SOL-2 (Figure 7A). The kainate-gated current appeared faster and smaller with co-expression of SOL-2. This can be appreciated by examining the ratio of peak kainate- to glutamate-gated current. SOL-2 decreased this ratio by approximately 50% (Figure 7B). These results suggest that in our reconstitution studies, SOL-2 acts to increase the rate of desensitization. One way to examine this possibility is by studying a GLR-1 variant in which the rate of desensitization is greatly slowed by the introduction of a single amino acid change (Q552Y) in the GLR-1 ligand-binding domain (Brockie et al., 2001b; Stern-Bach et al., 1998; Walker et al., 2006b). The glutamate-gated current recorded from Xenopus oocytes that expressed GLR-1(Q552Y), STG-2, and SOL-1 did not desensitize (Figure 7C). In contrast, there was considerable desensitization when SOL-2 was co-expressed (Figure 7C), indicating that the function of the receptor was modified by SOL-2.

Figure 7.

Figure 7

Open in a new tab

SOL-2 modifies kainate sensitivity and receptor desensitization in reconstitution studies. (A) Currents in response to 100 μM kainate (KA, top) and 1 mM glutamate (Glu, bottom) in Xenopus oocytes that expressed GLR-1, STG-2 and SOL-1 either with (left) or without (right) SOL-2. (B) The ratio of peak kainate to glutamate-gated current in Xenopus oocytes that expressed GLR-1, STG-2 and SOL-1, or GLR-1, STG-2, SOL-1 and SOL-2. For all conditions, n=11. *, p<0.001. (C) Examples of current recorded in response to 1 mM glutamate in Xenopus oocytes that expressed GLR-1(Q552Y), STG-2 and SOL-1 either with (bottom) or without (top) SOL-2. (D) Currents evoked in Xenopus oocytes in response to 1 mM glutamate and 100 μM kainate in either the presence or absence of Conconavalin-A.

See also Figure S6 and S7.

Additional evidence for modification of receptor function by SOL-2 could be observed following treatment by Concanavalin-A, a lectin that strongly blocks the desensitization of kainate receptors, but only weakly blocks desensitization of AMPARs (Partin et al., 1993). In the neuron AVA, Concanavalin-A only weakly modifies glutamate-gated currents (data not shown). However, in reconstitution studies in oocytes, we previously found that receptor desensitization was dramatically slowed by Concanavalin-A (Walker et al., 2006b). We now find that the efficacy of Concanavalin-A depends on the composition of the receptor complex. Thus, the block of desensitization of either glutamate- or kainate-gated currents was greatly diminished if SOL-2 was part of the receptor complex (Figure 7D).

SOL-2 is most closely related to the vertebrate Neto2 protein, which modifies the function of kainate receptors (Zhang et al., 2009), thus raising the question of whether GLR-1 is more closely related to kainate receptors or AMPARs. To address this question we first expressed Neto2 under control of the sol-2 promoter in transgenic sol-2; lurcher mutants. We did not find rescue of the lurcher phenotype nor did we find rescue of glutamate-gated currents in AVA (data not shown). These negative results are not interpretable because it is difficult to evaluate protein expression in transgenic worms. Therefore, we turned to reconstitution experiments in Xenopus oocytes. First, we compared the effects of SOL-2 and Neto2 on GLR-1 mediated currents. Consistent with our transgenic experiments, we did not find an obvious effect of Neto2 on glutamate-gated currents (Figure S6). Although SOL-2 dramatically changed the sensitivity to Concanavalin-A, we observed no such effects with Neto2 (Figure S6). Second, we examined the effects of SOL-2 and Neto2 on vertebrate GluA1(flip). As found previously (Zhang et al., 2009), we did not observe an obvious effect of Neto2 on GluA1-mediated current, and there was no obvious effect of SOL-2 on these currents (Figure S7A). Finally, we examined the effects of SOL-2 and Neto2 on vertebrate GluK2. Again, as previously observed (Zhang et al., 2009), we found that Neto2 dramatically increased GluK2-mediated current. However, we observed no such effect with SOL-2 (Figure S7B). Thus, in contrast to the evolutionarily conserved function of TARP proteins, i.e., vertebrate TARPs can contribute to GLR-1 function and C. elegans TARPs (STG-1, STG-2) can contribute to vertebrate AMPAR function (Walker et al., 2006a; Wang et al., 2008), we do not observe conservation of function with the SOL-2 and Neto2 CUB-domain proteins. Our data suggest that additional interacting proteins might contribute to Neto2, SOL-1 and SOL-2 function.

Discussion

Our study has identified the SOL-2/Neto CUB-domain protein, which is part of the GLR-1 signaling complex, thus defining a third class of AMPAR auxiliary proteins. SOL-2/Neto contributes to the GLR-1 complex by its interactions with SOL-1, and by modifying GLR-1 kinetics and pharmacology. Consequently, in sol-2 mutants GLR-1-mediated current and behaviors are disrupted. Our search for SOL-2 was motivated by our observation that the secreted extracellular domain of SOL-1 (s-SOL-1) was functional when expressed in neurons in vivo, but not in reconstitution studies (Figure 1). These conflicting results suggested that neurons express a specific protein required for s-SOL-1 function that is not expressed in C. elegans muscle cells or Xenopus oocytes. Because of our past success with a genetic strategy to identify components of the GLR-1 complex (SOL-1 and STG-2) (Wang et al., 2008; Zheng et al., 2004), we predicted that this protein could be identified using the same strategy, i.e., screening for mutations that suppress the hyper-reversal phenotype of transgenic lurcher worms. As predicted, we identified a new gene product, the SOL-2 CUB-domain transmembrane protein, that is homologous to vertebrate Neto proteins. This discovery further increases the complexity of the GLR-1 AMPAR postsynaptic signaling complex, which now contains members of at least four classes of proteins: AMPAR subunits, TARPs, SOL-1, and SOL-2/Neto, all of which have been validated by genetic perturbation, electrophysiology, cell biology and behavioral studies.

In support of our model that SOL-2 is part of the GLR-1 receptor complex, we found that SOL-2 co-localized and associated with SOL-1 and GLR-1 at synaptic sites. We also found that overexpressing SOL-1, but not SOL-2, in sol-1; sol-2 double mutants was sufficient to rescue both behavior and glutamate-gated current. These results indicate that SOL-2 likely functions as an adaptor protein that contributes to the interaction between SOL-1 and the receptor complex. However, in reconstitution studies, we also found that SOL-2 modifies relative agonist efficacy and the rate of receptor desensitization. Thus, SOL-2 has at least two roles: interacting with SOL-1 and modifying receptor function.

We were able to exclude an obligate role for SOL-1 in the biosynthesis, trafficking or stability of the GLR-1 signaling complex by demonstrating that s-SOL-1 provided in trans rescues glutamate-gated currents in sol-1 mutants. We also excluded an obligate developmental role for SOL-1 by showing that glutamate-gated current and GLR-1-dependent behavior was rescued in adult sol-1 mutants following heat shock induction of s-SOL-1 in adult worms. This result provides additional evidence that the receptor complex is stable in the absence of SOL-1. We were also able to rescue the behavioral and electrophysiological defects of sol-2 mutants by heat shock induction of SOL-2 in adult worms, indicating that SOL-2 has an ongoing function in adult animals and does not play an essential developmental role.

Components of the complex are also present in the absence of SOL-2 because GLR-1-mediated currents, although diminished, were observed in sol-2 mutants. However, the function of the complex is altered as shown by the differential rescue of glutamate- and kainate-gated currents in sol-2 mutants by overexpression of SOL-1. These data, together with the rapid perfusion experiments, where we could record rapidly desensitizing glutamate-gated currents in the absence of either SOL-1 or SOL-2, indicate that the components of the GLR-1 receptor complex are not degraded in sol-1 or sol-2 mutants. Thus, these proteins do not serve essential chaperone functions, suggesting that the identified components of the signaling complex might be independently regulated. Our results also suggest that dynamic changes in the composition of the complex could modulate the glutamate-gated postsynaptic current.

SOL-2 shares significant domain homology with the CUB-domain protein LEV-10, which is required for clustering of a subset of acetylcholine receptors at the neuromuscular junction in C. elegans (Gally et al., 2004). However, LEV-10 binds the CCP (sushi)-domain protein LEV-9 and does not appear to regulate receptor function (Gally et al., 2004; Gendrel et al., 2009). By sequence identity and domain structure, SOL-2 is homologous to the mammalian CUB-domain-containing transmembrane proteins Neto1 and Neto2. Both Neto proteins serve as auxiliary proteins for kainate receptors (Straub et al., 2011; Tang et al., 2011; Zhang et al., 2009) and modify receptor kinetics and kainate binding. Neto1 also appears to interact with NMDARs (Ng et al., 2009).

C. elegans GLR-1 was first defined as an AMPAR based on sequence identity (Brockie et al., 2001a); however, our demonstration that a Neto protein contributes to its function might suggest that GLR-1 is functionally more similar to kainate receptors. Although GLR-1 appears to share some characteristics with both AMPARs and kainate receptors, the bulk of the evidence indicates that GLR-1 is more like an AMPAR subunit: GLR-1 interacts with TARPs (which are AMPAR-specific auxiliary proteins) (Jackson and Nicoll, 2011; Milstein and Nicoll, 2008); the vertebrate TARP, stargazin, modulates GLR-1 function and C. elegans TARPs modulate vertebrate AMPARs (Walker et al., 2006a; Wang et al., 2008); and a conserved amino acid that dramatically influences AMPAR gating is found in GLR-1 (Brockie et al., 2001b; Stern-Bach et al., 1998; Walker et al., 2006b). Because we have found a homolog of vertebrate Neto proteins (SOL-2) that is required for SOL-1 and thus AMPAR function, we predict that there should also be Neto proteins and SOL-1 homologues in the vertebrate nervous system that function as AMPAR auxiliary proteins.

Our studies revealed several surprises when comparing loss of function in mutant worms to overexpression in reconstitution studies. Thus, in sol-2 mutants, the loss of SOL-2 increased the rate of receptor desensitization. Contrary to our expectations, co-expressing SOL-2 with components of the GLR-1 complex in reconstitution studies also increased the rate of desensitization. This was particularly striking when recording GLR-1(Q552Y)-mediated currents that switched from non-desensitizing in the absence of SOL-2 to desensitizing in the presence of SOL-2. The most likely explanation for these conflicting results is that additional proteins contribute to receptor function and these proteins are not present in the heterologous cells used for reconstitution.

We also found that Concanavalin-A, a drug known to block desensitization of kainate receptors, also blocked desensitization of the GLR-1-mediated currents recorded from Xenopus oocytes in the absence of SOL-2. However, the effect on desensitization was greatly attenuated when SOL-2 was co-expressed with the other known members of the GLR-1 complex. This result suggests that the GLR-1 complex containing SOL-2 behaves more like an AMPAR and is consistent with Concanavalin-A's known differential effect on AMPA and kainate receptors (Partin et al., 1993). Another surprise was that the rates of desensitization measured in outside-out patches from sol-1 and sol-2 mutants were quite similar. Because currents in response to pressure application of glutamate were larger in sol-2 mutants, we would have expected a rate of desensitization that was intermediate between sol-1 mutants and wild type. However, we might not have detected differences between sol-1 and sol-2 mutants due to limitations in the rate at which we could exchange solutions during glutamate application. Alternatively, the similar rates of desensitization might be due to the apparently unstable association between SOL-1 and the receptor complex in the absence of SOL-2. Modification of receptor kinetics by formation of outside-out patches has been previously reported (Li and Niu, 2004). Thus, SOL-1 might dissociate from the complex in outside-out patches from sol-2 mutants.

Our results help provide a new mechanistic view of postsynaptic function, where GLR-1 and the associated TARP proteins interact at the plasma membrane with a protein complex containing SOL-1 and SOL-2. The absence of any one component markedly changes the properties of the receptor suggesting that the presence of AMPARs at the postsynaptic membrane is necessary, but not sufficient, for normal glutamate-gated current. Additionally, our findings suggest that glutamate-gated current might be modified by activity-dependent changes in the relative numbers of auxiliary proteins present at the postsynaptic membrane, or in the association of these auxiliary proteins with AMPAR subunits. SOL-1 has a large extracellular region that could conceivably span the synaptic cleft and interact with active zone presynaptic proteins, thus maintaining the postsynaptic receptor complex in register with presynaptic release sites. In this scenario, SOL-2 and SOL-1 might contribute to functional slots predicted by electrophysiological analysis (Shi et al., 2001). In turn, the number of these slots, and their residency by AMPARs, might be regulated by activity and contribute to synaptic plasticity.

The phenomenon of long term potentiation (LTP) is most simply explained by the movement of receptor complexes from either extracellular regions or intracellular compartments to the synaptic membrane (Jackson and Nicoll, 2011; Kerchner and Nicoll, 2008). In our view, a major challenge for a deeper understanding of LTP is the role of activity-dependent changes in the number of auxiliary proteins. In this view, a subset of AMPARs might be silent because they are not associated with the proper complement of auxiliary proteins.

Experimental procedures

General methods, genetics and plasmids

All C. elegans strains were raised under standard laboratory conditions at 20 °C. Transgenic strains were generated using standard microinjection into the gonad of adult hermaphrodite worms. All fluorescently labeled proteins were found to be functional in transgenic rescue experiments of the relevant mutant phenotype. Plasmids, transgenic arrays and strains are described in Supplemental Experimental Procedures.

Antibody staining, SEP acid washes, and microscopy

Live transgenic worms that expressed HA::GLR-1::GFP in neurons (akIs101) were immunolabelled as previously described (Zheng et al., 2004). Briefly, anti-HA polyclonal sera (Molecular Probes) was diluted (1:1000) in worm injection buffer and injected into the pseudocoelomic space. To test the pH sensitivity of superecliptic phluorin (SEP), transgenic worms that expressed SEP::GLR-1 in AVA neurons were dissected to expose the ventral nerve cord (VNC). Dissected worms were bathed in extracellular fluid pH 7.4 (Mellem et.al., 2002) and the VNC was imaged both before and after solution exchange to ECF pH 6.5. Dissected worms were then bathed in ECF pH 7.4 containing 50 mM NH4Cl to change intracellular pH. Images were acquired using a Roper Cascade 512B CCD camera and a Zeiss 100X 1.0 NA water immersion lens. Confocal images were acquired using Nikon Ti-eclipse equipped with a Yokogawa CSU10 spinning disc head, and captured by a Cascade 1224B EMCCD camera.

Electrophysiological studies

Electrophysiological recordings from Xenopus oocytes were performed using standard two-electrode voltage clamp techniques (Walker et al., 2006a). Plasmids for cRNA are described in Supplemental Experimental Procedures. Recordings from AVA interneurons and muscle cells from dissected transgenic worms were performed as previously described (Mellem et al., 2002). Rapid perfusion experiments were performed on outside-out patches obtained from AVA interneurons in dissected C. elegans preparations. Control extracellular fluid (ECF) and 3 mM glutamate ECF solutions were delivered via theta tubing mounted on a piezoelectric manipulator (MXP ZT-300, Siskiyou). The rate of solution exchange was measured as the 10-90% change in open-tip potential. Statistical significance was determined using Student's T-test.

Behavioral analysis

Nose touch response and osmotic avoidance assays were performed as described in (Mellem et al., 2002). Reversal frequency was recorded manually and quantified using a computer program written in python. A reversal was defined as a switch from forward to backward movement. Statistical significance was determined by using the standard Student's t test. Error bars throughout represent the SEM.

Immunoprecipitation

HEK 293 cells were cultured in DMEM medium with 10% bovine fetal serum and transiently transfected using Lipofectamine 2000 (Life Technologies) in the presence of Opti-MEM. Plasmids for transfection are described in Supplemental Experimental Procedures. 48 hrs post-transfection, cells were lysed in ice-cold immunoprecipitation (IP) buffer (25 mM Tris, pH 7.5; 150 mM NaCl; 10% glycerol; 1% Triton X-100; 1mM EDTA; and Complete Protease Inhibitor Cocktail (Roche Diagnostics)). Cell lysates were spun at 16,100 g for 20 min and 79,000 g for 1 hr. The supernatants were incubated with rabbit polyclonal antibodies (Santa Cruz Biotechnology, Inc.) on ice for 1 hour followed by the addition of Protein A-Agarose (Santa Cruz Biotechnology, Inc.) and gentle mixing at 4 °C for 2 hours in ice-cold IP buffer. Samples were subject to 6 washes in ice-cold IP buffer and then denatured in 5x SDS sample buffer for 5 min at 95 °C. Proteins were resolved on 10% Mini-PROTEAN TGX precast gels (Bio-Rad), transferred to nitrocellulose membranes, and blocked in Odyssey® Blocking Buffer (LI-COR). Blots were probed with mouse monoclonal antibodies (Santa Cruz Biotechnology, Inc.) then IRDye 800CW (LI-COR) as primary and secondary antibodies, respectively. Signal was detected by the Odyssey® Infrared Imaging System (LI-COR).

Supplementary Material

01

Acknowledgements

We thank members of the Maricq laboratory for comments on the manuscript, Dane Maxfield for assistance with microscopy, and the Caenorhabditis Genetics Center (funded by the National Institutes of Health [NIH]) for providing worm strains. This research was made possible by support from NIH Grant NS35812.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  1. Brockie PJ, Madsen DM, Zheng Y, Mellem J, Maricq AV. Differential expression of glutamate receptor subunits in the nervous system of Caenorhabditis elegans and their regulation by the homeodomain protein UNC-42. J Neurosci. 2001a;21:1510–1522. doi: 10.1523/JNEUROSCI.21-05-01510.2001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Brockie PJ, Mellem JE, Hills T, Madsen DM, Maricq AV. The C. elegans glutamate receptor subunit NMR-1 is required for slow NMDA-activated currents that regulate reversal frequency during locomotion. Neuron. 2001b;31:617–630. doi: 10.1016/s0896-6273(01)00394-4. [DOI] [PubMed] [Google Scholar]
  3. Chen B, Liu Q, Ge Q, Xie J, Wang ZW. UNC-1 regulates gap junctions important to locomotion in C. elegans. Curr Biol. 2007;17:1334–1339. doi: 10.1016/j.cub.2007.06.060. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Chen L, Chetkovich DM, Petralia RS, Sweeney NT, Kawasaki Y, Wenthold RJ, Bredt DS, Nicoll RA. Stargazin regulates synaptic targeting of AMPA receptors by two distinct mechanisms. Nature. 2000;408:936–943. doi: 10.1038/35050030. [DOI] [PubMed] [Google Scholar]
  5. Feinberg EH, Vanhoven MK, Bendesky A, Wang G, Fetter RD, Shen K, Bargmann CI. GFP Reconstitution Across Synaptic Partners (GRASP) defines cell contacts and synapses in living nervous systems. Neuron. 2008;57:353–363. doi: 10.1016/j.neuron.2007.11.030. [DOI] [PubMed] [Google Scholar]
  6. Gally C, Eimer S, Richmond JE, Bessereau JL. A transmembrane protein required for acetylcholine receptor clustering in Caenorhabditis elegans. Nature. 2004;431:578–582. doi: 10.1038/nature02893. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Gendrel M, Rapti G, Richmond JE, Bessereau JL. A secreted complement-control-related protein ensures acetylcholine receptor clustering. Nature. 2009;461:992–996. doi: 10.1038/nature08430. [DOI] [PubMed] [Google Scholar]
  8. Gottschalk A, Almedom RB, Schedletzky T, Anderson SD, Yates JR, 3rd, Schafer WR. Identification and characterization of novel nicotinic receptor-associated proteins in Caenorhabditis elegans. Embo J. 2005;24:2566–2578. doi: 10.1038/sj.emboj.7600741. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Hart AC, Sims S, Kaplan JM. Synaptic code for sensory modalities revealed by C. elegans GLR-1 glutamate receptor. Nature. 1995;378:82–85. doi: 10.1038/378082a0. [DOI] [PubMed] [Google Scholar]
  10. Jackson AC, Nicoll RA. The expanding social network of ionotropic glutamate receptors: TARPs and other transmembrane auxiliary subunits. Neuron. 2011;70:178–199. doi: 10.1016/j.neuron.2011.04.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Kalashnikova E, Lorca RA, Kaur I, Barisone GA, Li B, Ishimaru T, Trimmer JS, Mohapatra DP, Diaz E. SynDIG1: an activity-regulated, AMPA-receptor-interacting transmembrane protein that regulates excitatory synapse development. Neuron. 2010;65:80–93. doi: 10.1016/j.neuron.2009.12.021. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Kano T, Brockie PJ, Sassa T, Fujimoto H, Kawahara Y, Iino Y, Mellem JE, Madsen DM, Hosono R, Maricq AV. Memory in Caenorhabditis elegans is mediated by NMDA-type ionotropic glutamate receptors. Curr Biol. 2008;18:1010–1015. doi: 10.1016/j.cub.2008.05.051. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Kato AS, Gill MB, Ho MT, Yu H, Tu Y, Siuda ER, Wang H, Qian YW, Nisenbaum ES, Tomita S, et al. Hippocampal AMPA receptor gating controlled by both TARP and cornichon proteins. Neuron. 2010;68:1082–1096. doi: 10.1016/j.neuron.2010.11.026. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Keino-Masu K, Masu M, Hinck L, Leonardo ED, Chan SS, Culotti JG, Tessier-Lavigne M. Deleted in Colorectal Cancer (DCC) encodes a netrin receptor. Cell. 1996;87:175–185. doi: 10.1016/s0092-8674(00)81336-7. [DOI] [PubMed] [Google Scholar]
  15. Kerchner GA, Nicoll RA. Silent synapses and the emergence of a postsynaptic mechanism for LTP. Nat Rev Neurosci. 2008;9:813–825. doi: 10.1038/nrn2501. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Kerppola TK. Visualization of molecular interactions by fluorescence complementation. Nature reviews Molecular cell biology. 2006;7:449–456. doi: 10.1038/nrm1929. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Kessels HW, Malinow R. Synaptic AMPA receptor plasticity and behavior. Neuron. 2009;61:340–350. doi: 10.1016/j.neuron.2009.01.015. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Li G, Niu L. How fast does the GluR1Qflip channel open? J Biol Chem. 2004;279:3990–3997. doi: 10.1074/jbc.M310410200. [DOI] [PubMed] [Google Scholar]
  19. Maricq AV, Peckol E, Driscoll M, Bargmann CI. Mechanosensory signalling in C. elegans mediated by the GLR-1 glutamate receptor. Nature. 1995;378:78–81. doi: 10.1038/378078a0. [DOI] [PubMed] [Google Scholar]
  20. Mayer ML, Armstrong N. Structure and function of glutamate receptor ion channels. Annu Rev Physiol. 2004;66:161–181. doi: 10.1146/annurev.physiol.66.050802.084104. [DOI] [PubMed] [Google Scholar]
  21. Mellem JE, Brockie PJ, Zheng Y, Madsen DM, Maricq AV. Decoding of Polymodal Sensory Stimuli by Postsynaptic Glutamate Receptors in C. elegans. Neuron. 2002;36:933–944. doi: 10.1016/s0896-6273(02)01088-7. [DOI] [PubMed] [Google Scholar]
  22. Miesenbock G, De Angelis DA, Rothman JE. Visualizing secretion and synaptic transmission with pH-sensitive green fluorescent proteins. Nature. 1998;394:192–195. doi: 10.1038/28190. [DOI] [PubMed] [Google Scholar]
  23. Milstein AD, Nicoll RA. Regulation of AMPA receptor gating and pharmacology by TARP auxiliary subunits. Trends Pharmacol Sci. 2008;29:333–339. doi: 10.1016/j.tips.2008.04.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Ng D, Pitcher GM, Szilard RK, Sertie A, Kanisek M, Clapcote SJ, Lipina T, Kalia LV, Joo D, McKerlie C, et al. Neto1 is a novel CUB-domain NMDA receptor-interacting protein required for synaptic plasticity and learning. PLoS Biol. 2009;7:e41. doi: 10.1371/journal.pbio.1000041. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Partin KM, Patneau DK, Winters CA, Mayer ML, Buonanno A. Selective modulation of desensitization at AMPA versus kainate receptors by cyclothiazide and concanavalin A. Neuron. 1993;11:1069–1082. doi: 10.1016/0896-6273(93)90220-l. [DOI] [PubMed] [Google Scholar]
  26. Schwenk J, Harmel N, Brechet A, Zolles G, Berkefeld H, Muller CS, Bildl W, Baehrens D, Huber B, Kulik A, et al. High-resolution proteomics unravel architecture and molecular diversity of native AMPA receptor complexes. Neuron. 2012;74:621–633. doi: 10.1016/j.neuron.2012.03.034. [DOI] [PubMed] [Google Scholar]
  27. Schwenk J, Harmel N, Zolles G, Bildl W, Kulik A, Heimrich B, Chisaka O, Jonas P, Schulte U, Fakler B, et al. Functional proteomics identify cornichon proteins as auxiliary subunits of AMPA receptors. Science. 2009;323:1313–1319. doi: 10.1126/science.1167852. [DOI] [PubMed] [Google Scholar]
  28. Shi S, Hayashi Y, Esteban JA, Malinow R. Subunit-specific rules governing AMPA receptor trafficking to synapses in hippocampal pyramidal neurons. Cell. 2001;105:331–343. doi: 10.1016/s0092-8674(01)00321-x. [DOI] [PubMed] [Google Scholar]
  29. Shi Y, Suh YH, Milstein AD, Isozaki K, Schmid SM, Roche KW, Nicoll RA. Functional comparison of the effects of TARPs and cornichons on AMPA receptor trafficking and gating. Proc Natl Acad Sci U S A. 2010;107:16315–16319. doi: 10.1073/pnas.1011706107. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Shyu YJ, Hiatt SM, Duren HM, Ellis RE, Kerppola TK, Hu CD. Visualization of protein interactions in living Caenorhabditis elegans using bimolecular fluorescence complementation analysis. Nat Protoc. 2008;3:588–596. doi: 10.1038/nprot.2008.16. [DOI] [PubMed] [Google Scholar]
  31. Stern-Bach Y, Russo S, Neuman M, Rosenmund C. A point mutation in the glutamate binding site blocks desensitization of AMPA receptors. Neuron. 1998;21:907–918. doi: 10.1016/s0896-6273(00)80605-4. [DOI] [PubMed] [Google Scholar]
  32. Straub C, Hunt DL, Yamasaki M, Kim KS, Watanabe M, Castillo PE, Tomita S. Distinct functions of kainate receptors in the brain are determined by the auxiliary subunit Neto1. Nat Neurosci. 2011;14:866–873. doi: 10.1038/nn.2837. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Tang M, Pelkey KA, Ng D, Ivakine E, McBain CJ, Salter MW, McInnes RR. Neto1 is an auxiliary subunit of native synaptic kainate receptors. J Neurosci. 2011;31:10009–10018. doi: 10.1523/JNEUROSCI.6617-10.2011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. von Engelhardt J, Mack V, Sprengel R, Kavenstock N, Li KW, Stern-Bach Y, Smit AB, Seeburg PH, Monyer H. CKAMP44: a brain-specific protein attenuating short-term synaptic plasticity in the dentate gyrus. Science. 2010;327:1518–1522. doi: 10.1126/science.1184178. [DOI] [PubMed] [Google Scholar]
  35. Walker CS, Brockie PJ, Madsen DM, Francis MM, Zheng Y, Koduri S, Mellem JE, Strutz-Seebohm N, Maricq AV. Reconstitution of invertebrate glutamate receptor function depends on stargazin-like proteins. Proc Natl Acad Sci U S A. 2006a;103:10781–10786. doi: 10.1073/pnas.0604482103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Walker CS, Francis MM, Brockie PJ, Madsen DM, Zheng Y, Maricq AV. Conserved SOL-1 proteins regulate ionotropic glutamate receptor desensitization. Proc Natl Acad Sci U S A. 2006b;103:10787–10792. doi: 10.1073/pnas.0604520103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Wang R, Walker CS, Brockie PJ, Francis MM, Mellem JE, Madsen DM, Maricq AV. Evolutionary conserved role for TARPs in the gating of glutamate receptors and tuning of synaptic function. Neuron. 2008;59:997–1008. doi: 10.1016/j.neuron.2008.07.023. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Zhang W, St-Gelais F, Grabner CP, Trinidad JC, Sumioka A, Morimoto-Tomita M, Kim KS, Straub C, Burlingame AL, Howe JR, et al. A transmembrane accessory subunit that modulates kainate-type glutamate receptors. Neuron. 2009;61:385–396. doi: 10.1016/j.neuron.2008.12.014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Zheng Y, Brockie PJ, Mellem JE, Madsen DM, Maricq AV. Neuronal Control of Locomotion in C. elegans is Modified by a Dominant Mutation in the GLR-1 Ionotropic Glutamate Receptor. Neuron. 1999;24:347–361. doi: 10.1016/s0896-6273(00)80849-1. [DOI] [PubMed] [Google Scholar]
  40. Zheng Y, Brockie PJ, Mellem JE, Madsen DM, Walker CS, Francis MM, Maricq AV. SOL-1 is an auxiliary subunit that modulates the gating of GLR-1 glutamate receptors in Caenorhabditis elegans. Proc Natl Acad Sci U S A. 2006;103:1100–1105. doi: 10.1073/pnas.0504612103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Zheng Y, Mellem JE, Brockie PJ, Madsen DM, Maricq AV. SOL-1 is a CUB-domain protein required for GLR-1 glutamate receptor function in C. elegans. Nature. 2004;427:451–457. doi: 10.1038/nature02244. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

01
NIHMS395814-supplement-01.doc (3.2MB, doc)

RESOURCES