SEARCH

SEARCH BY CITATION

Keywords:

  • Caenorhabditis elegans;
  • exploratory behavior;
  • gentle touch response;
  • neurexin;
  • neuroligin;
  • sinusoidal postural movement;
  • synapse

Abstract

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. References
  7. Acknowledgments
  8. Supporting Information

Neurexins are cell adhesion proteins that interact with neuroligin and other ligands at the synapse. In humans, mutations in neurexin or neuroligin genes have been associated with autism and other mental disorders. The human neurexin and neuroligin genes are orthologous to the Caenorhabditis elegans genes nrx-1 and nlg-1, respectively. Here we show that nrx-1-deficient mutants are defective in exploratory capacity, sinusoidal postural movements and gentle touch response. Interestingly, the exploratory behavioral phenotype observed in nrx-1 mutants was markedly different to nlg-1-deficient mutants; thus, while the former had a ‘hyper-reversal’ phenotype increasing the number of changes of direction with respect to the wild-type strain, the nlg-1 mutants presented a ‘hypo-reversal’ phenotype. On the other hand, the nrx-1- and nlg-1-defective mutants showed similar abnormal sinusoidal postural movement phenotypes. The response of these mutant strains to aldicarb (acetylcholinesterase inhibitor), levamisole (ACh agonist) and pentylenetetrazole [gamma-aminobutyric (GABA) receptor antagonist], suggested that the varying behavioral phenotypes were caused by defects in ACh and/or GABA inputs. The defective behavioral phenotypes of nrx-1-deficient mutants were rescued in transgenic strains expressing either human alpha- or beta-NRXN-1 isoforms under the worm nrx-1 promoter. A previous report had shown that human and rat neuroligins were functional in C. elegans. Together, these results suggest that the functional mechanism underpinning both neuroligin and neurexin in the nematode are comparable to human. In this sense the nematode might constitute a simple in vivo model for understanding basic mechanisms involved in neurological diseases for which neuroligin and neurexin are implicated in having a role.

The neurexins are a family of synaptic adhesion proteins that have an important role in synaptic formation and maintenance. In vertebrates there are three genes encoding neurexins, each with two promoters that generate two different transcripts named alpha and beta variants (Missler et al. 1998; Ushkaryov et al. 1992). The neurexins have a high degree of evolutionary conservation in vertebrates and are essential for postnatal survival (Li et al. 2007; Tabuchi & Sudhof 2002). Triple knockout mice lacking NRXN-1α, NRXN-2α and NRXN-3α, died on the first day of birth (Missler et al. 2003). In vertebrates, neurexins are synthesized in both excitatory and inhibitory neurons (Ichtchenko et al. 1995, 1996; Ullrich et al. 1995). Although different neurexin isoforms are differentially distributed in distinct brain areas, there are overlapping distributions of neurexin isoforms in different types of neurons (Ullrich et al. 1995).

The extracellular sequence of α-neurexins has six LNS (laminin/neurexin/sex hormone-binding globulin) domains with three intercalated epidermal growth factor (EGF)-like domains (Missler et al. 1998). On the other hand, the extracellular sequence of β-neurexins contains a single LNS domain. However, the transmembrane and intracellular domains of alpha and beta isoforms are identical. Neuroligins bind both α- and β-neurexins (Boucard et al. 2005) and this binding may allow the correct alignment of presynaptic with the postsynaptic terminals, facilitating the interaction between neurotransmitter release and receptor sites (Song et al. 1999).

Neurexins alone are sufficient to induce gamma-aminobutyric (GABA) and glutamate postsynaptic specializations in contacting dendrites. Therefore neurexins, along with neuroligins, seems to be essential for the homeostasis of inhibitory GABAergic and excitatory glutamatergic synaptogenesis (Graf et al. 2004). Supporting this observation is the fact that mutations in genes coding for neurexins and neuroligins have been associated with a wide range of mental disorders (Sudhof 2008). In particular, some copy number variations and point mutations in neurexin-encoding genes have been linked to neurodevelopmental disorders including autism (Gauthier et al. 2011; Kim et al. 2008; Szatmari et al. 2007; Vaags et al. 2012).

Caenorhabditis elegans has only one gene, nrx-1, encoding neurexin. The nrx-1 gene product, NRX-1, shows the same functional domains of vertebrate neurexins, and also α- and β-neurexin transcript variants have been described (Haklai-Topper et al. 2011). In addition, it has been reported that nrx-1 is expressed in almost all the neurons of the nervous system of C. elegans during all larval and adult stages at the presynaptic sites. However, expression is not observed in glial cells (Haklai-Topper et al. 2011). Recently it has been suggested that in C. elegans NRX-1 and NLG-1 mediate a retrograde synaptic signal that inhibits neurotransmitter release at neuromuscular junctions (Hu et al. 2012). Thus, in particular situations it has been proposed that NLG-1 and NRX-1 might be pre- and postsynaptic respectively, which is contrary in polarity to their arrangement in other synapses where has been originally characterized. Cases of reverse polarity of neuroligins has been reported in Celegans (Feinberg et al. 2008), and of neurexins in mouse and Drosophila (Chen et al. 2010; Kattenstroth et al. 2004; Taniguchi et al. 2007).

Here we show that nrx-1-deficient mutants of C. elegans are defective in several behaviors including exploratory capacity, sinusoidal postural movements and gentle touch response, probably because they have defects in connectivity which lead to altered neural network function. In addition, we present results showing that transgenic strains expressing the human alpha- or beta-NRXN-1 isoforms, rescue these abnormal phenotypes found in nrx-1-defective mutants. These results complement a previous report showing that human neuroligin was also functional in C. elegans (Calahorro & Ruiz-Rubio 2012), suggesting that the functional mechanism underpinning both neuroligin and neurexin is conserved throughout evolution. In this sense the nematode might constitute a useful tool for understanding basic mechanisms of human neurological diseases in which neuroligins and neurexins are implicated in having a role.

Materials and methods

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. References
  7. Acknowledgments
  8. Supporting Information

Strains and maintenance

All nematodes were grown and maintained at 20°C under standard conditions (Brenner 1974). Table 1 shows the C. elegans strains used in this study. OP50 Escherichia coli strain was obtained from the Caenorhabditis Genetic Center (University of Minnesota, Minneapolis, MN, USA).

Table 1. Caenorhabditis elegans strains used in this study
Strain nameGenotypeSource
  1. a

    Caenorhabditis Genetics Center.

  2. b

    National Bioresource Project.

  3. c

    After outcrossing FX01961 strain with N2 at least six times.

  4. d

    After outcrossing VC1416 strain with N2 at least six times.

  5. e

    The cDNA human NRXN-1α-coding region was obtained from clone KIAA0578 (bf00990), Kazusa DNA Research Institute, Japan.

  6. f

    Human NRXN-1β::CFP was a gift from Dr. Ehud Isacoff, University of California, Berkeley.

N2Wild type, DR subclone of CB originalCGCa
FX01961nrx-1 (tm1961) VNBPb
VC1416nrx-1 (ok1649) VCGC
VC228nlg-1 (ok259) XCGC
CRR1nlg-1 (ok259) XCalahorro & Ruiz-Rubio (2012)
CRR3nrx-1 (tm1961) VThis studyc
CRR4nrx-1 (ok1649) VThis studyd
CRR112nrx-1 (tm1961) V; crrEx12 [pPD95.75 (Pnrx-1::nrx-1β); pDD04 NeoR (pmyo-2::GFP)]This study
CRR113nrx-1 (tm1961) V; crrEx13 [pPD95.75 (Pnrx-1::NRXN-1α); pDD04 NeoR (pmyo-2::GFP)]This studye
CRR114nrx-1 (tm1961) V; crrEx14 [pPD95.75 (Pnrx-1::NRXN-1β::CFP); pDD04 NeoR (pmyo-2::GFP)]This studyf

To study the behavioral phenotype of C. elegans nrx-1-deficient mutants we examine the effect of two different mutant alleles, ok1649 and tm1961 of the nrx-1 gene (Fig. S1). The ok1649 allele consists of a deletion of 861 pb which results in the complete removal of exon 9 and partially introns 8 and 9. The exon 9 includes 50 residues in the NRX-1-LNS2 domain. The tm1961 allele consist of a deletion of 428 bp including partially intron 15 and exon 13 and completely introns 13 and 14, and exons 14 and 15. This deletion in tm1961 allele generates a truncated protein without the LNS4, LNS5, LNS6, transmembrane domains and C-terminal region (Fig. S1). In spite of these differences, the phenotypes of nrx-1 mutant strains with tm1961 and ok1649 alleles did not present significant behavioral differences. A possible explanation is that the ok1649 deletion, although only affecting the NRX-1-LNS2 domain, causes a significant change in the three-dimensional structure of the protein (Fig. S1). Thus, while the NRX-1-wild-type protein has a repetitive pattern of β-sheet/α-helix tandem in this region, NRX-1-ok1649 protein shows an ongoing series of β-sheets, which may affect the protein functionality. Therefore, it is possible that in the mutant alleles ok1649 and tm1961 the NRX-1 activity is equally abolished.

Cloning and transgenic methods

cDNA cloning of the C. elegans nrx-1β gene

Total RNA was extracted from mixed stage Bristol N2 animals by the TRIzol® extraction method as described by the manufacture (Life Technologies S.A., Madrid, Spain). The short spliced form nrx-1β (Haklai-Topper et al. 2011) was detected by RT-PCR analysis using the following primers: nrx-1β forward 5′-CCggatccGATGGTATTCCTTGAGC-3′ and nrx-1β reverse 5′-AAgaattcGATTATACGTA-3′. The putative nrx-1β-coding sequence was predicted using the longer spliced form nrx-1α sequence (C29A12.4b transcript, from Wormbase version WS233). The nrx-1β ATG putative start codon is located at the end of nrx-1α intron 22, and the entire coding region comprises nrx-1α exons 23, 25, 26, 27 and 28 (Fig. S2; see complete coding sequence in Fig. S3).

cDNA cloning of NRXN-1α human gene

NRXN-1 alpha precursor genomic sequence was used to design the following specific primers: NRX1 forward 5′-TTCTTCGTCACGggatccAGGACC-3′ and NRXN-1 reverse 5′-TTTCTATAgaattgTCCATTTAAGATC-3′. These primers were used to amplify by PCR, the coding region from the cDNA clone KIAA0578 (bf00990) (obtained from Kazusa DNA Research Institute, Japan).

cDNA cloning of NRXN-1β human gene

cDNA cloning of NRXN-1β-mCFP was obtained of a generous gift from Dr. Ehud Isacoff, University of California, Berkeley, USA. In this construct monomeric CFP is inserted in NRXN between LNS and transmembrane domains (Pautot et al. 2005). The entire NRXN-1β-mCFP sequence was amplified using the following primers, NRXN-1β-CFP forward 5′-AGCTCAAggatccCATGTACCAG-3′ and NRXN-1β-CFP reverse 5′-TCGCgaatccACTAGTTCAGACATAC-3′, and amplified products were cloned into pPD95.75 vector.

Transgenic procedure

Transgenic strains were generated by coinjecting ‘marker’ (0.5 µg/ml) and ‘rescue’ (30–50 µg/ml) plasmids into the germ line of adult hermaphrodites, as previously described (Mello & Fire 1995).

Rescue constructions and transgenic methods

Translational constructions were generated fusing nrx-1β, NRXN-1α or NRXN-1β-mCFP cDNAs with the C. elegans nrx-1 promoter (Fig. S4), in the BamHI/EcoRI site of the pDD95.75 vector. The promoter region of nrx-1 was amplified using the Pnrx-1forward 5′-ACATTTTTAAACAtctagaTTTCTAGG-3′ and Pnrx-1reverse 5′-CGTTAAGTATGGTCCAggatccATCAAAC-3′ from C429A12 cosmid clone (from Sanger Institute, UK). FX1961 nrx-1 (tm1961) V animals were microinjected with the ‘marker’ plasmid pDD04NeoR (Pmyo-2::GFP) (Fig. S4a) together with nrx-1β, NRXN-1α or NRXN-1β-mCFP ‘rescue’ constructions (Fig. S4b–i).

Behavioral assays

The behavioral experiments were carried out with two strains deficient in nrx-1 gene, CRR3 (tm1961) and CRR4 (ok1649). The phenotype of strains with tm1961 allele was assessed in more detail, and the strain with ok1649 allele was assayed to confirm that they shared similar phenotypes. In the behavioral experiments, at least two different lines of each transgenic strain were analyzed to check if they showed similar phenotypes. All the experiments were repeated at least three times in different days. The experiments were carried out blind except the reversal and the exploratory capacity assay and the exploratory assays.

Reversals and omega turns measurements

In the absence of food, C. elegans disrupts its forward movement with spontaneous changes of direction including reversals and omega turns. They have been classified depending on their length and degree of reorientation: r1, reverse movement with a head swing with ≈40° change direction; r2, reverse movement with two head swings with ≈70° change direction; R3, reverse movement with three head swings with ≈90° change direction; R4, reverse movement with four head swings followed by an omega turns with a ≈ 170° change direction (Gray et al. 2005). In the case of r1, r2 and R3 movements, sometimes they can be followed by an omega turn.

To quantify the spontaneous reversal behavior, the number of reversals and omega turns were measured based on the procedure described previously (Gray et al. 2005). Briefly, animals were grown on non-crowded conditions at 20°C in Nematode Growth Medium (NGM) plates seeded with the bacterial strain OP50. Young adult hermaphrodites (at late L4 larval stage) were gently transferred with a platinum wire onto a NGM plate without bacteria, and after 20 seconds were transferred again onto a new NGM plate without bacteria where reversals and omega turns were measured. These foodless NGM plates were obtained dispensing 7.5 ml of NGM on 50 mm Petri dishes that were air-dried at room temperature for 90 min, and then incubated overnight at 37°C and air-dried at room temperature for 90 min just before carrying out the assay. The observation of reversal and omega turns began exactly 1 min after transfer to the second foodless NGM plate, because the transfer suppresses turning for around 1 min (Zhao et al. 2003). The spontaneous reversal assays lasted for 40 min; within this period, four measurements of 1 minute each were made at intervals of 1–2, 3–4, 9–10 and 39–40 min.

Exploratory capacity

To quantify the exploratory capacity of the worms, about 20 L4 larval stage animals of each genotype grown on OP50 Escherichia coli in fresh NGM plates were picked and placed at the same time in the middle of a 9 cm plate containing 8 ml of agar medium (2% agar, 0.25% v/v Tween 20, 10 mm HEPES pH 7.2) without bacteria. Plates were incubated from 5 to 7 days at 20°C. The tracks of the worms in the agar were recognized by sight as a consequence of their roaming, and they were quantified by scanning the surface of agar dishes at 600-ppp resolution and analyzing the images in ‘Image J’ software (National Institute of Health, USA) using custom scripts. Images were turned into binary format converting them through a predetermined threshold, and finally a particles analysis was performed. Data were normalized with respect N2 wild-type strain.

Wavelength and amplitude measurements

Young adult animals were freely moving on NGM plate on an OP50 E. coli lawn during 60 min incubation at 20°C. The bacterial lawn was obtained by spreading 30 µl of an OP50 stationary culture and incubating a 37°C the day before the experiments. Tracks left by animals on the bacterial lawn were recorded through snapshots covering the whole plate. Wavelength and amplitude were measured in 10 independent tracks on 10 consecutive waves for each plate. The measurements were carrying out using Image J software and data were normalized with respect N2 wild-type strain.

Drug response assays
Aldicarb assay

A total of 25 young adult hermaphrodites were transferred to NGM plates containing 0.5 mm aldicarb (Sigma, Barcelona, Spain) and paralysis rate was measured as described (Mahoney et al. 2006). Briefly, the first day, 40–50 L4 larval stage animals of each genotype were picked to a fresh NGM plates containing OP50 E. coli. The second day, aldicarb plates (one per strain) were brought out from 4°C and temperate at 20°C. A small spot of about 4–6 mm in diameter of OP50 E. coli was placed on to each plate. Then 25 young adults (at late L4 larval stage) of a single genotype were placed on to an aldicarb plate and left at room temperature. After 30 min, the worms were observed for paralysis in periods of 30 min under the stereomicroscope. The paralysis was defined by the absence of movement when prodded three times with a platinum wire on the head and tail, and when pharyngeal pumping was absent.

Levamisole assay

Paralysis assays were performed as previously described (Gottschalk et al. 2005) by adding 0.2 mm levamisole (Sigma, Barcelona, Spain) to NGM plates. Preparation of the drug plates was achieved as described for aldicarb assay. A total of 25 young adult hermaphrodites (at late L4 larval stage) of each strain were transferred to levamisole plates. Paralysis was followed by visual inspection under the stereomicroscope every 15 min and defined as lack of movement in response to prodding three times with a platinum wire.

Pentylenetetrazole assay

Assays were performed as it has been described previously (Locke et al. 2008), adding 70 mm pentylenetetrazole (PTZ) (Sigma, Barcelona, Spain) on NGM plates. Paralysis was followed by visual inspection under stereomicroscope every 15 min and was defined by the absence of movement when prodded three times with a platinum wire on the head and tail, and when pharyngeal pumping was absent.

Gentle touch assay

This phenotype was tested by stroking the worm 10 times with an eyebrow hair attached to a toothpick, alternating the anterior (just behind the pharynx) and posterior (just before the anus) part of the body. A positive response causes the animal to move backward or forward respectively (Bounoutas & Chalfie 2007; Chalfie et al. 1985)

Statistical analysis

Comparisons shown in each experiment were done by one-way anova, using spss (IBM, Madrid, Spain) statistical tool.

Results

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. References
  7. Acknowledgments
  8. Supporting Information

The number of reversals and omega turns increases in nrx-1-deficient mutants respect to N2 wild-type strain

Caenorhabditis. elegans feed continuously and a brief food deprivation is an alerting signal that stimulates it to explore the environment. Thus, in the absence of food, the worm changes its forward movement with spontaneous reversals and omega turns (Gray et al. 2005).

In nrx-1-deficient mutants the frequency of reversal and turn events was increased with respect to the wild-type strain (Fig. 1a). This augmentation was observed over the time of navigation of the worm on plates without food (1–2 min: F1,226 = 7.527, P = 0.007; 3–4 min: F1,226 = 8.728, P = 0.003; 9–10 min: F1,226 = 16.825, P = 0.000; 39–40 min: F1,226 = 13.033, P = 0.000). Interestingly, nrx-1 mut-ants had specifically an elevated number of r1–r2 reversals and omega coupled r1–r2 movements (r1–r2: 1–2 min, F1,36 = 7.977, P = 0.008; 3–4 min, F1,36 = 7.412, P = 0.010; 9–10 min, F1,36 = 34.791, P = 0.000; 39–40 min, F1,36 = 7.648, P = 0.009. r1–r2Ω: 1–2 min, F1,36 = 14.552, P = 0.001; 3–4 min, F1,36 = 11.694, P = 0.002; 9–10 min, F1,36 = 13.423, P = 0.00; 39–40 min, F1,36 = 3.521, P =0.069; Fig. 1b,c). In contrast, nlg-1-deficient mutants presented a decrease in spontaneous reversal rate respect to the wild-type strain (1–2 min: F1,148 = 6.802, P = 0.010; 3–4 min: F1,148 = 5.835, P = 0.017; 9–10 min: F1,148 = 2.755, P = 0.099; 39–40 min: F1,148 = 0.381, P = 0.538; Fig. 1a). This decrease was observed in the frequency of total events and in all kinds of reversals and turns. These results are consistent with those previously described that reported a general decrease in spontaneous reversal rate of nlg-1-deficient mutants (Hunter et al. 2010).

image

Figure 1. Increased number of reversals and omega turns events in neurexin-deficient mutants of Caenorhabditis elegans, and rescue by expression of human α- and β-NRXN-1 cDNAs. Frequency of total reversals, short and long reversals (r1–R4) and omega turns are shown. The spontaneous reversal assays was scored in NGM plates without food assays and lasted for 40 min. Within this period four measurements of 1 min each were carried out at intervals of 1–2, 3–4, 9–10 and 39–40 min. At least 20 animals of each strain were analyzed. The values of short–long reversals (r1–R4+) (a) and omega turns (r1Ω–R4+Ω) (b) were distinguished of the total reversals (c). R4+ indicates four or more head swings. Error bars show mean ± SEM. Statistical significance was calculated by anova/Fisher's exact test. *P ≤ 0.05, **P ≤ 0.01 and ***P ≤ 0.001 compared with N2 wild-type strain.

Download figure to PowerPoint

The exploratory capacity is reduced in nrx-1-deficient mutants of C. elegans

In C. elegans, dwelling is a state in which the worms make small movements and stay in the same spot and conversely roaming is movement with long forward ‘runs’ exploring a wide area. These two different behaviors are controlled by the presence or absence of food in the environment (Fujiwara et al. 2002). During the dwelling phase worms have an intermittent backward and forward movement and present a low speed of locomotion. However, in the roaming phase worms have predominantly forward movement with few changes of direction (Fujiwara et al. 2002; Shtonda & Avery 2006).

Figure 2 shows that the roaming behavior of nrx-1-deficient mutants was almost absent when worms were transferred from OP50 bacteria plates to a plate without food (F1,19 = 94.011, P = 0.000 vs. wild type, Fig. 2b). However, nlg-1-deficient mutants showed an increase in roaming movement (F1,13 = 26.522, P = 0.000 vs. wild type, Fig. 2b). These opposing results indicate that nrx-1 and nlg-1-deficient mutants are impaired in a behaviorally antonymous fashion with respect navigation phenotypes in response to changes in the food environment. This is in agreement with the results of Fig. 1, where reversals and omega turns were increased and decreased in nrx-1 and nlg-1-defective mutants, respectively.

image

Figure 2. Abnormal exploratory capacity of neurexin-deficient mutants of Caenorhabditis elegans, and rescue by expression of human α- and β-NRXN-1 cDNAs. (a) Assay plates showing the tracks of roaming patterns of each strain. A representative experiment is shown. (b) About 20 L4 larval stage animals were placed in the middle of a 9-cm Tween-20 agar plates. After 5 days, the tracks were quantified by scanning the surface of agar dishes at 600-ppp resolution and analyzed with the ‘Image J’ software. (c) Quantification of the explored area by the worm in the assay plates. At least three experiments with each strain were carried out. Error bars show mean ± SEM. Statistical significance was calculated by anova/Fisher's exact test. ***P ≤ 0.001 compared with N2 wild-type strain.

Download figure to PowerPoint

The nrx-1-deficient mutants of C. elegans showed abnormal sinusoidal patterns during locomotion

Caenorhabditis elegans crawl on the surface of agar plates seeded with a bacterial lawn generating a characteristic sinusoidal wave pattern path behind it. This is produced by complementary dorsal and ventral muscular contractions provided by internal hydrostatic pressures of the worm (Park et al. 2008). Figure 3 shows that neurexin- and neuroligin-deficient mutants showed a decrease in both, wavelength and amplitude of waves during the crawling motion compared with wild type (wavelength: nlg-1 F1,13 = 91.986, P = 0.000, nrx-1 F1,14 = 31.711, P = 0.000; amplitude: nlg-1 F1,13 = 241.628, P = 0.000, nrx-1 F1,14 = 216.098, P = 0.000). This suggests that this neuromuscular circuit may be impaired in both mutants.

image

Figure 3. Anomalous sinusoidal pattern of locomotion of nrx-1-defective mutants of Caenorhabditis elegans, and rescue by expression of human α- and β-NRXN-1 cDNAs. (a) Tracks shaped by freely moving individual animals feeding on food after 60 min. Wavelength and amplitude of waves are indicated by white and black lines, respectively. Representative snapshot images of moving wild type, nlg-1 and nrx-1 mutants, transgenic strains expressing NRXN-1β human cDNA and transgenic strain expressing worm nrx-1β cDNA as control are shown. Image above panel shows a detail of each wave-track. (b) Quantification of wavelength and amplitude of waves. The measurements were made on selected tracks in each image taken, and the length of wave/amplitude indicated with arrows in image was plotted. At least 20 animals of each strain were scored, and a total of 200 tracks were measured—10 individual tracks in each animal. Error bars show mean ± SEM. Statistical significance was calculated by anova/Fisher's exact test. *P ≤ 0.05 and ***P ≤ 0.001 compared with N2 wild-type strain.

Download figure to PowerPoint

Evidence for defective neuromuscular signaling in nrx-1-deficient mutants

In C. elegans, dorsal and ventral body muscles are innervated by both, cholinergic and GABA motor neuron synapses. Release of acetylcholine (ACh) leads to contraction of the body wall muscle on one side and stimulates GABA release onto muscles on the opposite side (Jorgensen 2005; Schuske et al. 2004). This pattern of alternating dorsal and ventral contractions, produced by interactions between excitatory and inhibitory motor neurons, generates the sinusoidal pattern of movement and leads to coordinated locomotion (Jorgensen 2005). The abnormal wave pattern observed in nrx-1 and nlg-1 mutants might be due to defects in synaptic transmission related to ACh and/or GABA neurotransmitters functionality.

To test whether the impairment in wave pattern behavior of nlg-1- and nrx-1-deficient mutants were caused by defects in ACh and/or GABA inputs, we examined the susceptibility of these strains to levamisole (ACh agonist), aldicarb (acetylcholinesterase inhibitor) and PTZ (GABA receptor antagonist), to specifically characterize signaling at neuromuscular junctions (Locke et al. 2006, 2008). The nrx-1-deficient mutant showed a similar sensitivity to 0.2 mm levamisole compared to wild type (Fig. 4a), but in contrast the nlg-1-deficient mutant was sensitive to levamisole after 15 min of exposure (15 min: F1,8 = 21.479, P = 0.002; 30 min: F1,8 = 46.225, P = 0.000; 45 min: F1,8 = 52.563, P = 0.000; 60 min: F1,8 = 14.516, P = 0.005; 75 min: F1,8 = 23.273, P = 0.001; 90 min: F1,8 = 8.333, P = 0.020, Fig. 4a). Interestingly, nlg-1- and nrx-1-defective mutants were resistant and sensitive respectively to 0.5 mm aldicarb compared to wild type (Fig. 4b). The increase of sensitivity of the nrx-deficient mutant was from 30 to 210 min of exposure (30 min: F1,8 = 54.382, P = 0.000; 60 min: F1,8 = 39.763, P = 0.000; 90 min: F1,8 = 61.728, P = 0.000; 120 min: F1,8 =37.236, P = 0.000; 150 min: F1,8 = 53.778, P = 0.000; 180 min: F1,8 = 17.308, P = 0.000; 210 min: F1,8 = 10.125, P = 0.000 vs. wild type, Fig. 4b). Resistance to aldicarb is associated with an ACh release reduction in presynaptic terminals (Locke et al. 2006; Mahoney et al. 2006). Finally, both mutants were more sensitive to 70 mm PTZ from 15 to 90 min exposure (nrx-1 vs. wild type—15 min: F1,10 = 10.292, P = 0.009; 30 min: F1,10 = 254.297, P = 0.000; 45 min: F1,10 = 269.017, P = 0.000; 60 min: F1,10 = 328.904, P = 0.000; 75 min: F1,10 = 366.416, P = 0.000; 90 min: F1,10 = 96.429, P = 0.000. nlg-1 vs. wild type—15 min: F1,10 = 12.273, P = 0.006; 30 min: F1,10 = 22.237, P = 0.001; 45 min: F1,10 = 172.404, P = 0.000; 60 min: F1,10 = 144.118, P = 0.000; 75 min: F1,10 = 92.418, P = 0.000; 90 min: F1,10 = 7.529, P = 0.021, Fig. 4c), consistent with the suggestion that there is reduced GABAergic signaling in these mutants.

image

Figure 4. Hypersensitivity of nrx-1-defective mutant to levamisole and PTZ, and rescue by expression of human α- and β-NRXN-1 cDNAs. Genomic deletion of nrx-1 (tm1961) causes an increase of sensitivity to 0.5 mm aldicarb and 70 mm PTZ, respectively in paralysis assays. Whereas nlg-1 mutants show a resistance to 0.5 mm aldicarb and 70 mm PTZ, and an increase in sensitivity to 0.2 mm levamisole, respectively. Induced paralysis to 0.2 mm levamisole (a), 0.5 mm aldicarb (b) and 70 mm PTZ (c) are shown. Each data point represents the mean of at least three trials with 25 animals in each trial. Error bars show mean ± SEM. Statistical significance was calculated by anova/Fisher's exact test. *P ≤ 0.001 significant differences with wild type.

Download figure to PowerPoint

Gentle touch response is altered in nrx-1-defective mutants of C. elegans

When the nematode receives a tactile stimulus with an eyebrow hair in the anterior or posterior part of its body, it changes the direction of motion inducing movement back or forward respectively (Chalfie et al. 1985). Neuroligin-deficient mutants had a defective capability of the mechanosensory response in both the anterior and posterior part of the body with respect to the wild-type strain (Calahorro & Ruiz-Rubio 2012). Figure 5 shows that the nrx-1-deficient mutant is also defective in gentle touch response losing a significant capability of the mechanosensory response in both the anterior and posterior part of the body with respect to the wild-type strain (anterior: F1,51 = 45.779, P = 0.000; posterior: F1,51 = 84.361, P = 0.000 vs. wild type).

image

Figure 5. Altered gentle touch response of nrx-1-defective mutants of Caenorhabditis elegans, and rescue by expression of human α- and β-NRXN-1 cDNAs. Data are quantified as percentage of positive response to gentle touch. Animals were touched 10 times, alternating the anterior and posterior part of the body. The measurement was carried out by stroking an eyebrow hair across the body just behind the pharynx for the anterior touch response (a), or just before the anus for the posterior touch response (b). At least three different experiments were carried out with each strain (approximately 50 worms per experiment). Error bars show mean ± SEM. Statistical significance was calculated by anova/Fisher's exact test. ***P ≤ 0.001 compared between indicated groups.

Download figure to PowerPoint

Expression of human alpha- or beta-NRXN1 isoforms in neurexin-deficient mutants strain of C. elegans rescues the behavioral defects

The defective behavior phenotypes found in the nrx-1-deficient mutants were rescued in transgenic strains expressing cDNAs-coding human NRXN-1α or NRXN-1β, both of them driven by the C. elegans nrx-1 promoter. In these experiments transgenic nrx-1-deficient mutants expressing NRX-1β from the nematode was used as control.

Figure 1 shows that the frequency in the total number of reversal and turn events were restored in transgenic strains to the level of N2 wild type [1–2 min: worm nrx-1β, F1,112 = 0.019, P = 0.890; Human NRXN-1α, F1,106 = 0.532, P = 0.467; Human NRXN-1β, F1,106 = 0.074, P = 0.786; 3–4 min: worm nrx-1β, F1,112 = 0.008, P = 0.928; Human NRXN-1α, F1,106 = 0.691, P = 0.408; Human NRXN-1β, F1,106 = 0.018, P = 0.894; 9–10 min: worm nrx-1β, F1,112 = 1.491, P = 0.225; Human NRXN-1α, F1,106 = 0.058, P = 0.808; Human NRXN-1β, F1,106 = 0.024, P = 0.877; 39–40 min: worm nrx-1β, F1,112 = 0.038, P = 0.845; Human NRXN-1α, F1,106 = 0.204, P = 0.652; Human NRXN-1β, F1,106 = 2.056, P = 0.155). The defective exploring capacity of nrx-1-deficient mutants was also rescued in transgenic strains (worm nrx-1β, F1,10 = 156.959, P = 0.000; Human NRXN-1α, F1,16 = 59.786, P = 0.000; Human NRXN-1β F1,10 = 260.298, P = 0.000; vs. nrx-1-deficient mutant, Fig. 2b).

The expression of human NRXN-1α and NRXN-1β rescues the defective wavelength and amplitude of waves during crawling motion (wavelength: worm nrx-1β, F1,13 = 0.403, P = 0.536; Human NRXN-1α, F1,14 = 0.517, P = 0.484; Human NRXN-1β, F1,12 = 5.826, P = 0.033; vs. wild-type strain, and amplitude: worm nrx-1β, F1,13 = 4.180, P = 0.062; Human NRXN-1α, F1,14 = 8.396, P = 0.012; Human NRXN-1β, F1,12 = 1.682, P = 0.219; vs. wild-type strain, Fig. 3). The sensitivity to aldicarb and PTZ was also rescued in the transgenic strains (Fig. 4).

Finally, Fig. 5 shows that the transgenic strains expressing human NRXN-1α and NRXN-1β restored the wild-type behavior of nrx-1-deficient mutant relating to the gentle touch response in both the anterior and posterior part of the body [anterior: worm nrx-1β, F1,49 = 18.780, P = 0.000; Human NRXN-1α, F1,60 = 29.616, P = 0.000; Human NRXN-1β, F1,50 = 17.652, P = 0.000; posterior: worm nrx-1β, F1,49 = 25.523, P = 0.000; Human NRXN-1α, F1,60 = 20.719, P = 0.000; Human NRXN-1β, F1,50 = 28.737, P = 0.000 vs. nrx-1 (tm1961)].

Discussion

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. References
  7. Acknowledgments
  8. Supporting Information

Neurexins are a family of neuronal plasma membrane proteins which have a key role as trans-synaptic receptors (Ushkaryov et al. 1992). The binding of presynaptic neurexins to postsynaptic proteins, such as neuroligins, has been proposed to participate in signaling pathways regulating synapse formation, homeostasis and plasticity (Yamagata et al. 2003). In fact, mutations in neurexin genes have been associated with autism, mental retardation, schizophrenia and other brain disorders. The nematode C. elegans has a very well-defined and genetically tractable nervous system which offers a simple model to explore basic mechanistic pathways involved in autism and other neurological diseases (Calahorro & Ruiz-Rubio 2011). Here we showed that Celegans nrx-1-deficient mutants are impaired in behaviors associated with exploratory capacity, sinusoidal postural movements and gentle touch response. These results indicate that neurexins might be necessary for correct synapse functionality and synchronization of a neuronal network.

Neurexin- and neuroligin-deficient mutants showed opposing roaming behavior

The circuits mediating C. elegans ‘spontaneous’ reversal behaviors have been reported in considerable detail (Tsalik & Hobert 2003; Zheng et al. 1999). Control of reversal requires input from different sensory neurons, followed by processing through AIY, AIZ, RIB and RIM interneurons. It has been reported that neuroligin is only expressed in about 40 neurons of the nervous system; but includes the AIY neurons (Hunter et al. 2010). The lack of neuroligin in this circuit may weaken the synaptic connections preventing in most cases the backwards movement originating a ‘hypo-reversal’ phenotype. This behavior contrasts with the neurexin-deficient mutants, which displayed a ‘hyper-reversal’ phenotype. The nrx-1 gene, contrary to nlg-1, has been shown to be expressed in most neurons of the nervous system of the nematode (Haklai-Topper et al. 2011). It is possible that the phenotype observed in neurexin-deficient mutants is due to wide spread aberrant connectivity in this circuit. Supporting this suggestion is the fact that laser ablation of the AIY neurons leads also to a ‘hyper-reversal’ phenotype (Tsalik & Hobert 2003).

In relation to these observations, we have described previously that while the nlg-1-deficient mutant was defective in detecting a 4 m fructose solution, the nrx-1-deficient mutants had a wild-type response (Calahorro et al. 2009). However, in the assay that we performed, the worm had to move forwards about 5 ml to reach the sugar solution. With the high frequency of forward–backward movements we have detected in these mutants (Figs. 1, 2), they could not make contact and reach the sugar solution. Therefore the fructose assay, as it was performed does not permit one to distinguish to know whether the change of direction is due to the detection of a high osmotic strength or to the intrinsic behavior of these nrx-1-defective strains.

The neurexin- and neuroligin-deficient mutants exhibit major defects in the amplitude and wavelength of sinusoidal postural movements on an E. coli lawn

In C. elegans, dorsal and ventral body muscles are innervated through both, cholinergic and GABA motor neurons. Release of ACh leads to contraction of the body wall muscle on one side and stimulates GABA release onto muscles on the opposite side (Jorgensen 2005; Schuske et al. 2004). This pattern of alternating dorsal and ventral contractions, produced by interactions between excitatory and inhibitory motor neurons, generates the sinusoidal pattern of movement and leads to coordinated locomotion (Jorgensen 2005).

The abnormal wave pattern observed in nrx-1 and nlg-1 mutants (Fig. 3) might be due to defects in synaptic transmission related to ACh and/or GABA neurotransmitters functionality. Supporting this idea we observed that nlg-1- and nrx-1-defective mutants were resistant and sensitive respectively to the cholinesterase inhibitor aldicarb, when compared with the wild-type strain. The resistance to aldicarb has been associated with an ACh release reduction in presynaptic terminals (Locke et al. 2006; Mahoney et al. 2006). Both mutants were hypersensitive to the GABA antagonist PTZ. In relation to this response C. elegans synaptic transmission mutants have been classified by complementary aldicarb and PTZ exposure response (Locke et al. 2008). On this basis it is likely that the nlg-1-deficient mutant is deficient in ACh transmission because it is aldicarb resistant and showed no PTZ-induced anterior convulsions. On the other hand, the nrx-1-deficient mutant likely fails to negatively regulate ACh transmission because it is aldicarb sensitive and also showed no PTZ-induced anterior convulsions. However, the increase in sensitivity to PTZ in these mutants may also suggest that the release, reuptake or distribution of GABA receptor may be altered in these strains.

The neurexin-deficient mutants are defective in gentle touch response

Gentle body touch is sensed by the mechanosensory neurons ALML/R, AVM and PLML/R (Chalfie & Sulston 1981; Chalfie et al. 1985). The sense of gentle touch is based on the capability of these touch receptor neurons to translate mechanical inputs into ionic currents which activate a neural circuit that drives a locomotory response (O'Hagan et al. 2005). ALM-AVM and PLM are responsible of anterior and posterior touch sensitivity, respectively. While ALM and AVM axons enter the nerve ring, PLM enters the ventral nerve cord. In all cases these neurons form connections with many other neurons providing direct input to the interneurons that control locomotion. However, the pattern of these connections is very asymmetric. Thus, the touch cells neurons form gap junctions with agonist interneurons and chemical synapses with the antagonist interneurons. This pattern of connectivity suggested that the gap junctions are excitatory and the synaptic connections are inhibitory (Chalfie et al. 1985). Consequently, as in the case of the abnormal sinusoidal pattern, the defective gentle touch response observed in nrx-1 mutants might be due to defects in synaptic transmission and/or reception related to GABA or ACh in neurons responsible of the nematode movement.

The mutant phenotype C. elegans neurexin-deficient mutants are rescued by the human NRXN-1α and NRXN-1β genes

The expression of human NRXN-1α and NRXN-1β isoforms in neurexin-deficient mutants of the nematode demonstrated that both were able to rescue the abnormal phenotype behaviors found in nrx-1-deficient mutants. This observation is consistent with the homology between human and nematode neurexins. Both proteins have the same functional domains and exhibit an overall 23% identity (Tabuchi & Sudhof 2002). Furthermore, human transmembrane proteins have been expressed successfully in the nematode (Calahorro & Ruiz-Rubio 2012; Treusch et al. 2004), indicating that transmembrane domains from human proteins are functional in C. elegans.

In mammals, the interaction between neurexins and neuroligins seems to follow a connection code based on alternative splicing of neurexins. Thus, neurexin variants which have exon 20 at AS4 generates NRXN 4(+), and alternatively spliced form without these 30 amino acid insertion in this site produces NRXN 4(−) (Iijima et al. 2012). The regulation of different splicing forms is proposed to modulate the appropriate excitatory or inhibitory specialization across the synaptic cleft (Boucard et al. 2005; Chih et al. 2006). In our case, for rescuing the phenotype of nematode nrx-1-defective mutants we used NRXN-1β 4(−) and NRXN-1α 4(+) variants. Both variants were able to rescue with similar efficacy the abnormal exploratory phenotype, suggesting that both isoforms were are able to interact with the worm neuroligin.

To date, in C. elegans, an interaction code between neurexin and other ligand–proteins has not been described. In this work, we have isolated a short neurexin-coding transcript (Fig. S2b) and provided evidence that it corresponds to an nrx-1β isoform because it rescued the phenotype described in the nematode. The identity and similarity percentages between human NRXN-1β and our putative C. elegans NRX-1β were 11.73 and 23.25, respectively (Fig. S5). It would be interesting in the future to investigate if, as in the mammalian protein, an interaction code exists between neurexin and other proteins. This could be investigated by expressing, with neuronal promoters, distinct neurexin and/or neuroligin isoforms and analyzing their capabilities for rescuing behavioral phenotypes. In this sense, given that C. elegans and humans seem to share basic synaptic molecular mechanisms, the nematode could be a useful in vivo genetic tool for studying mechanisms underpinning human neurological diseases.

References

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. References
  7. Acknowledgments
  8. Supporting Information

Acknowledgments

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. References
  7. Acknowledgments
  8. Supporting Information

This work was supported by grant PI0197, Consejería de Salud, Junta de Andalucía, Spain. We thank the Caenorhabditis Genetics Center, the Japanese National Bioresource Project, Antonio Miranda-Vizuete, Julián Cerón, Sandra Wiese, Ehud Isacoff, Peter Askjaer, Juan Cabello, Rosina Giordano and Denis Dupuy for sharing worm strains and plasmids. We are grateful to Lindy M. Holden-Dye for critical reading of the manuscript, Patricia G. Izquierdo for the valuable help with the statistical analysis, Francisco Gómez-Scholl and Encarna Alejandre for useful comments and suggestions, and Raquel Romero Porras and Isabel Caballero for laboratory assistance. We also thank Justo P. Castaño for the support with the microinjection station in his laboratory. None of the authors has any conflict of interest with regard to the research reported here.

Supporting Information

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. References
  7. Acknowledgments
  8. Supporting Information
FilenameFormatSizeDescription
gbb12046-sup-0001-figures1.docWord document3758KFigure S1: Hypothetical models of NRX-1 proteins of C. elegans. Three-dimensional structure models of the wild-type NRX-1 (a), NRX-1-ok1649 (b) and NRX-1-tm1961 (c) proteins are shown. Conformational changes in the proteins coded by both mutant alleles are observed. Models were powered by Swiss-Model Proteomic Serve (Arnold et al. 2006; Guex & Peitsch 1997; Schwede et al. 2003). (D) Structure of C. elegans nrx-1 locus on chromosome V and identification of the nrx-1 (ok1649) and nrx-1 (tm1961) deletion alleles. Exons of the transcript are shown as numbered shaded boxes connected by lines representing the introns. The inset shows a PCR amplification of N2 wild type, CRR3 (tm1961) and CRR1 (ok1649) genomic DNA using the following primers flanking the deletion: nrx-1(ok1649) forward 5′-CGGAAGCAAAGAAACCAAAG-3′, nrx-1(ok1649) reverse 5′-GTTGAGCATTTGCAATCGAA-3′, nrx-1(tm1961) forward 5′-ATCTGGCCGATCAAAGTTAC-3′ and nrx-1(tm1961) reverse 5′-TCTAACCTCCCGTTGAGCAT-3′. The expected sizes were as follows: 2459 bp (ok1649 allele), 3320 bp (wild type control); 1398 bp (tm1961 allele), 1826 bp (wild type control).
gbb12046-sup-0002-figures2.docWord document3758KFigure S2: Exonic organization and protein domains of C. elegans NRX-1β isoform. (a) nrx-1α transcript isoform (coding transcript C29A12.4b) and nrx-1β isoform. The exons are indicated with yellow boxes and exon numbers are labeled above or below each exon. (b) Domains organization of the C. elegans neurexin proteins. The α-neurexins contain a N-terminal signal peptide (SP) that is removed from the final form functional protein, followed by three repeats units consisting of two laminin-G domains (LNS) flanking an EGF domains, a transmembrane domain (TM) and a short cytoplasmatic domain (not shown). The putative worm NRX-1β protein is also shown.
gbb12046-sup-0003-figures3.docWord document3758KFigure S3: Caenorhabditis elegans nrx-1β isoform sequences. (a) nrx-1β isoform-coding sequence corresponding to isolated cDNA from Bristol N2 wild-type strain (see Materials and Methods section). (b) Putative NRX-1β full-length protein sequence translated.
gbb12046-sup-0004-figures4.docWord document3758KFigure S4: GFP or CFP fluorescence in transgenic strains CRR114 expressing human NRXN-1β cDNAs in neurexin deficient mutants of C. elegans. Expression of GFP in pharynx muscles (a), and expression of CFP in neurons of head ganglia (b) or embryo (f). The transgenic strain contains the extra-array {crrEx14 [pDD04 NeoR (pmyo-2::GFP); pPD95.75 (Pnrx-1::NRXN-1β::CFP)]}. The Bristol N2 wild-type strain was used as a control for unspecific fluorescence signals (d, h). Asterisks indicate autofluorescence signal. The inset in (b) shows a detail of extranuclear expression patterns in body cell of head neurons (arrows). The images (c), (e), (g) and (i) correspond to DIC. (j) Translational construction used for human NRXN-1β expression.
gbb12046-sup-0005-figures5.docWord document3758KFigure S5: Comparative amino acid sequences of C. elegans NRX-1β and human NRXN-1β isoforms, showing a full length protein sequence (a), extracellular (b) and intracellular domains (c) alignments. Accession numbers for nucleotide and protein sequences are as follow: Homo sapiens, NM_1387535.2 and NP_620072.1, consensus CDS: CCDS1845.1. The extra- and intracellular regions of C. elegans and human protein isoforms were predicted using TMHMM server v. 2.0 (Center for Biological Sequence Analysis, Technical University of Denmark, DTU). The nrx-1β-coding transcript sequence was translated and alignment of protein sequences was performed using the Clustal W method (BLOSUM62 Similarity Matrix). Identity (I) and similarity (S) percentages: I = 11.73/S = 23.25 with full-length proteins; I = 8.19/S = 16.93 among extracellular domains; I = 17.89/S = 32.63 among intracellular domains.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.