In the developing nervous system, axons respond to various guidance cues to find their targets. The effects guidance cues have on an axon may change as an axon undergoes morphological changes, such as branching, turning, and synapse formation. The means by which these changes are regulated are not well understood. In Caenorhabditis elegans, the UNC-40/DCC (deleted in colorectal cancer) receptor mediates responses to the UNC-6/netrin guidance cue. Here, we show that CLEC-38, a protein with predicted transmembrane and C-type lectin-like domains, regulates UNC-40-mediated axon outgrowth as well as the organization of presynaptic terminals. We observe that, in genetic backgrounds sensitized for axon guidance defects, loss of clec-38 function can suppress defects in an UNC-40-dependent manner. Within migrating axons, clec-38 acts cell autonomously. Furthermore, loss of clec-38 function alters UNC-40::GFP (green fluorescent protein) expression. We also observe that loss of clec-38 function disrupts presynaptic patterning in animals with normal axon guidance and that there are genetic interactions between clec-38 and rpm-1, which encodes a protein implicated in regulating presynaptic assembly and axon morphology. We suggest CLEC-38 plays a role in promoting synapse assembly and refining axon outgrowth activity.
In the developing nervous system migrating axons encounter multiple extracellular cues that help control the development of the axon. For example, axons that express the UNC-40/deleted in colorectal cancer (DCC), UNC-5/UNC5, and SAX-3/robo receptors are guided by extracellular UNC-6/netrin and SLT-1/slit cues toward their targets (Tessier-Lavigne and Goodman, 1996; Dickson, 2002). As different local environments are encountered, the response to these cues may change as axons are influenced by additional factors that stimulate new events such as branch formation and synapse development. These factors are mostly unknown, but may include the guidance cues and receptors themselves. For example, Slit and Robo can regulate axon guidance and stimulate branching in vertebrates (Wang et al., 1999; Ma and Tessier-Lavigne, 2007). In Caenorhabditis elegans, changing spatial and temporal patterns of UNC-6 expression by neuroglia and pioneer neurons indicate a hierarchy of netrin cues in the developing nervous system (Wadsworth et al., 1996). The phenotypes of unc-6 mutants, as well as laser ablation studies of UNC-6-expressing cells, indicate that the UNC-6 cues direct local axon migrations as well as interactions among axons and glia cells during the assembly of the axon scaffold (Wadsworth and Hedgecock, 1996; Ren et al., 1999; Hutter, 2003). Within the developing nerve ring, it has been shown that local UNC-6/netrin cues and the UNC-40 receptor help guide axons as well as mediate the assembly of presynaptic terminals (Colon-Ramos et al., 2007).
We present here evidence that CLEC-38 can negatively regulate UNC-40-mediated outgrowth in axons migrating toward their targets. Furthermore, CLEC-38 is required for proper presynaptic development in axons that have reached their targets. To our knowledge, this is the first time that a member of the protein superfamily containing the C-type lectin-like domains (CTLDs) has been directly implicated in these processes. Based on our observations, we suggest that CLEC-38 acts in migrating axons to help mediate signals that both promote synaptogenesis and modify further UNC-40-mediated axon outgrowth.
Materials and Methods
Worms were cultivated according to standard protocol and were maintained at 20°C (Brenner, 1974).
N2, Bristol strain, was used as a standard wild-type strain. All mutations used for this study are strong loss-of-function or null alleles unless otherwise indicated. Strains constructed and used for this study are as follows: IM721: clec-38(ur280)V, IM720: clec-38(ur280)V; evIs82a[punc-129::GFP]IV, IM723: clec-38(tm2035)V, IM658: clec-38(ur280)V; unc-6(rh46)X; evIs82aIV, IM660: clec-38(ur280)V; unc-6(e78)X; evIs82a IV, IM661: unc-6(rh46)X; evIs82aIV, IM662: unc-6(e78)X; evIs82aIV, IM207: unc-6(ev400)X; evIs82aIV, IM831: clec-38(ur280)V; unc-6(ev400)X; evIs82aIV, IM832: unc-40(e1430)I; evIs82aIV, IM833: clec-38(ur280)V; unc40(e1430)I; evIs82aIV, IM834: unc-5(e53)IV evIs82aIV, IM835: clec-38(ur280)V; unc-5(e53)IV; evIs82aIV, IM869: clec-38(ur280)V; unc-40(e1430)I; unc-6(rh46)X; evIs82aIV, IM984: clec-38(ur280)V; unc-5(e53)IV; unc-6(rh46)X; evIs82aIV, IM682: clec-38(ur280)V; zdIs5[pmec-4::GFP]I, IM838: unc-6(rh46)X; zdIs5I, IM836: clec-38(ur280)V; unc-6(rh46)X; zdIs5I, IM650: unc-6(ev400)X; zdIs5I, IM837: clec-38(ur280)V; unc-6(ev400)X; zdIs5I, IM647: slt-1(eh15); zdIs5, IM839: clec-38(ur280)V; slt1(eh15); zdIs5I, IM649: unc-6(ev400)X; slt-1(eh15)X; zdIs5I, IM712: sax-3(ky123)X; zdIs5I, IM713: clec-38(ur280)V; sax-3(ky123)X; zdIs5I, IM739: unc-5(e53)IV; zdIs5I, IM648: unc-40(e1430)I; zdIs5I, IM840: clec-38(ur280)V; unc40(e1430)I; zdIs5I, IM843: rpm-1(ur299)V, IM844: rpm-1(ur299)V; evIs82aIV, IM805: rpm-1(ur299)V; unc-6(rh46)X, IM893: clec-38(ur280)V; rpm-1(ur299)V; unc-6(rh46)X; evIs82aIV, IM845: rpm-1(ur299)V; zdIs5I, IM817: rpm-1(ur299)V; slt-1(eh15)X; zdIs5I, IM841: clec-38(ur280)V; rpm-1(ur299)V; zdIs5I, IM828: clec-38(ur280)V; rpm-1(ur299)V; slt-1(eh15)X; zdIs5I, IM868: clec-38(ur280)V; rpm-1(ur299)V; unc-40(e1430)I; zdIs5I, IM823: urEx333 [mec-4::clec-38; flp-20::gfp]; clec-38(ur280)V, IM829: urEx 333[mec4::clec-38; flp-20]; slt-1(eh15)X; clec-38(ur280)V, IM882: evEx66[unc-40::GFP;rol-6]; clec-38(ur280)V, IM981: clec-38(ur280)V;evIs98[unc-5::gfp;+dpy-20], IM982: clec 38(ur280)V;kyEx253[sax-3::gfp;+lin-15], IM977: urEx375[pclec-38::gfp; str-1::gfp], IM898: clec-38(ur280)V;kyEx1212[punc-86::UNC-40::GFP], IM995: clec-38(ur280)V;juIs1[punc-25::SNB-1::GFP]IV.
Transgenes maintained as extrachromosomal arrays included the following: urEx333[mec-4::clec-38; flp-20::gfp], urEx375[pclec-38::gfp; str-1::gfp]. Strains not derived in the Wadsworth Laboratory were kindly provided by the Caenorhabditis Genetics Center (Minneapolis, MN) (NM1455, NL4256, CB4856, and other strains used for mapping), Scott Clark (SK4005; New York University Medical Center, New York, NY), Joe Culotti (evIs82a[punc-129::GFP], evEx66, evIs98; University of Toronto, Toronto, Ontario, Canada), Cori Bargmann (kyEx1212[punc-86::UNC-40::GFP]; The Rockefeller University, New York, NY), and Yishi Jin (juIs-1[punc-25::SNB-1::GFP]; University of California, San Diego, La Jolla, CA). The tm2035 deletion allele was obtained from the Japanese Knock-out Consortium as described by the National BioResource Project (http://www.nbrp.jp/report/reportProject.jsp?project = celegans).
The tm2035 deletion was detected using the following PCR primers: forward, CGTTACAAAACCGTTGAG, and reverse, TAGTTGTTGATGCAATTC; and a 992 bp deletion was confirmed by sequencing. The clec-38(RNAi) phenotype was observed by feeding worms with bacteria expressing double-stranded RNA (Fraser et al., 2000) using the RNA interference (RNAi) feeding bacterial strain V-11O15 (Geneservice, Cambridge, UK). All double or triple mutants generated were confirmed by complementation tests or by PCR genotyping.
Molecular characterization of clec-38(ur280).
Two factor and SNP mapping were used to place the ur280 mutation on the right arm of chromosome V. Cosmids containing sequences for the region were obtained from The Sanger Institute (Cambridge, UK) and were injected into mutant animals at concentrations of 2 ng/μl, along with a coinjection marker pIM175 [Punc-119::gfp] at 80 ng/μl. The animals were examined for rescue of a mild dumpy body phenotype caused by ur280 mutation. Germline transformation with the single cosmid T25E12 rescued the phenotypes of ur280. Furthermore, a genomic PCR product that included the entire sequence of the gene T25E12.10 (i.e., clec-38) rescued the ur280 phenotype. The molecular lesion in clec-38(ur280) was identified by sequencing genomic PCR products from mutant animals and aligning with the reported genomic sequence from C. elegans Genome Sequencing Consortium.
The mec-4::clec-38 expression construct pIM218 was made by amplifying clec-38 sequence from a C. elegans cDNA library (Invitrogen, Carlsbad, CA) using the following PCR primers: forward, CTACTAGCTAGCATGGCAATATTCTACGAC, and reverse, CGGCGGGGTACCTCAAAAATCAATAGCCCG. The PCR product was digested with NheI and KpnI restriction enzymes and was cloned into the NheI and KpnI sites after the mec-4 promoter sequence of plasmid pIM 207 (Quinn et al., 2006). The promoter fusion (Pclec-38::gfp) construct pIM220 was made by amplifying ∼2 kb sequence upstream of the clec-38 (T25E12.10) ATG codon including the first three codons from C. elegans genomic DNA using the following PCR primers: forward, AAAACTGCAGAGAGGGGAATTTTCAAG, and reverse, CGCCGCGGATCCTATTGCCATTTGTTTTGC. The PCR product was digested with PstI and BamH1 restriction enzymes and was cloned into the PstI and BamH1 sites in the Fire laboratory vector, pPD95.77 (Mello and Fire, 1995).
Transgenic strains were generated by described methods (Mello and Fire, 1995). The pIM218 construct was injected into clec-38(ur280) animals at 50 ng/μl along with a flp-20::gfp coinjection marker [kindly provided by Chris Li (City College of New York, New York, NY)] at 80 ng/μl. The transgenic lines were maintained as extrachromosomal arrays by following the green fluorescent protein (GFP) fluorescence. Three independent lines were established. The array of one line, IM823, was used in genetic crosses with slt-1(eh15) to generate strain IM829. The pIM 220 construct was injected into N2 animals at 5 ng/μl along with str-1::gfp coinjection marker [a kind gift from Cori Bargmann (Rockefeller University, New York, NY)] at 80 ng/μl. The transgenic lines were maintained as extrachromosomal arrays by following the GFP fluorescence. Three independent lines were established and analyzed for clec-38 expression.
Microscopy and axon guidance assay.
Mechanosensory neurons were visualized using a chromosomally integrated mec-4::gfp transgene (zdIs5; gift from Scott Clark), and the DA and DB neurons were observed using an integrated unc-129::gfp transgene evIs82a (a gift from J. Culotti). The animals were mounted on 5% agarose pad with 10 mm levamisole and were analyzed using a Carl Zeiss (Oberkochen, Germany) Axio-imager Z1 microscope with an apotome imager. The ventral guidance of the AVM axon was scored as defective if it failed to reach the ventral cord. Dorsal guidance of the DA and DB axons was scored as defective if none of the axons from the neuron cell bodies situated along the ventral nerve cord in the region between the pharynx and vulva failed to reach the dorsal cord. The PLM axon was scored as overextended if it migrated past the AVM neuron cell body. UNC-40::GFP expression analysis was performed using 63× objective in the wild-type and mutant background under same exposure time. Statistical analysis was done using a two-tailed z test to compare the phenotypes between two different strains. A value of p < 0.05 was considered significantly different. Image analysis was performed using Axio-Vision LE 4.5 software.
Identification and characterization of CLEC-38
A mutation, ur280, was isolated from a genetic screen for mutations that could suppress dorsal guidance defects caused by the unc-6(rh46) mutation. The unc-6(rh46) mutation is a partial loss-of-function allele (Hedgecock et al., 1990; Wadsworth et al., 1996) and mutations were isolated that could improve the ability of DA and DB motor axons to reach the dorsal cord in the unc-6(rh46) background. The ur280 mutation was mapped to a region of chromosome V (supplemental Fig. 1, available at www.jneurosci.org as supplemental material). Using cosmids containing sequence corresponding to this region and DNA-mediated transformation rescue, we found that the ur280 locus is within cosmid T25E12 in the gene clec-38. The clec-38 gene is predicted to encode a protein of 382 aa comprising a short cytoplasmic tail with a putative SH2 binding site, a type-II transmembrane domain and two extracellular CTLDs (supplemental Fig. 1, available at www.jneurosci.org as supplemental material). DNA sequence analysis of ur280 identifies a TTT to GTT change, which causes valine 269 within the second CTLD to be substituted for phenylalanine. We also analyzed tm2035, which was isolated in a PCR-based screen for deletions by the National BioResource Project. This clec-38 mutation can also suppress the dorsal guidance defects of unc-6(rh46) (data not shown). The tm2035 mutation is a deletion removing part of the clec-38 promoter and the first five exons, which correspond to the first 222 aa of CLEC-38 (supplemental Fig. 1, available at www.jneurosci.org as supplemental material). Animals deficient for clec-38 by RNAi show a mild dumpy phenotype similar to that caused by the clec-38(ur280) and clec-38(tm2035) mutations. Furthermore, the ur280 phenotypes were not enhanced when the ur280 allele was placed over the deficiency yDf4, which removes the clec-38 gene (data not shown). Together, these observations indicate that clec-38(ur280) is a strong loss-of-function mutation.
To determine cells that express the clec-38 gene, we designed a promoter fusion construct using 2 kb upstream of clec-38 sequence fused to a green fluorescent protein coding sequence (see Materials and Methods). The transgenic animals carrying this construct show GFP fluorescence in the nervous system (Fig. 1). Fluorescence could be detected in early embryos and persists through adulthood. The expression was detected in head neurons of threefold stage embryos (Fig. 1A). Strong expression was also observed in intestine. In the adults, the fluorescence is detected in the ventral cord motor neurons, PLM touch neurons, head neurons, and the intestine (Fig. 1B,C). Attempts to generate strains expressing a GFP-tagged CLEC-38 translational fusion have been unsuccessful.
CLEC-38 regulates dorsal axon guidance mediated by UNC-40
The clec-38(ur280) allele was isolated as a suppressor of DA and DB motor axon defects caused by the unc-6(rh46) allele. We examined whether the dorsal migration of DA and DB motor neuron axons in different unc-6 mutants are improved by the clec-38(ur280) mutation (Fig. 2). We found that dorsal guidance is improved in unc-6(rh46) and unc-6(e78) mutants, but not in unc-6(ev400) mutants (Fig. 2F). The unc-6(rh46) and unc-6(e78) are reduction-of-function missense mutations that are predicted to encode altered forms of UNC-6, whereas unc-6(ev400) causes an early stop mutation and is considered a null allele (Hedgecock et al., 1990; Wadsworth et al., 1996). These results suggest UNC-6 is required for the improved guidance in the loss of clec-38 function mutant.
The dorsal guidance of DA and DB motor neurons in response to UNC-6 is mediated by UNC-5 and UNC-40 receptors (Leung-Hagesteijn et al., 1992; Chan et al., 1996). To determine whether the improved dorsal guidance observed in the clec-38(ur280) mutants are affected by loss of the receptors, we examined animals with unc-5 and unc-40 loss-of-function mutations (Fig. 2F). Loss of unc-5, but not unc-40, function causes dorsal axon guidance defective by the criteria of our assay, which scored dorsal migration as defective only if none of the axons from the neuron cell bodies situated along the ventral nerve cord in the region between the pharynx and vulva failed to reach the dorsal cord. Results from clec-38; unc-40; unc-6(rh46) and clec-38; unc-5; unc-6(rh46) triple mutants show that the clec-38 mutation cannot suppress the unc-6(rh46) dorsal guidance phenotype in the absence of UNC-40, indicating that the suppression by loss of clec-38 function requires UNC-40 signaling. The dorsal guidance defects are the same in clec-38; unc-5(e53); unc-6(rh46) and unc-5(e53) single mutants; however, this result is not too informative because dorsal guidance is severely impaired by the loss of unc-5 function alone. Together, the results suggest that the loss of clec-38 function affects UNC-40 activity in the DA and DB motor neurons and that this change improves the ability of the axons to respond to the dorsal guidance information provided by UNC-6.
CLEC-38 regulates ventral guidance mediated by UNC-40
The ventral migration of the AVM axon is guided by the UNC-6 and SLT-1 guidance cues through signals mediated by the UNC-40 and SAX-3 receptors (Hedgecock et al., 1990; Wadsworth et al., 1996; Hao et al., 2001; Yu et al., 2002; Gitai et al., 2003). The AVM axon travels away from dorsal sources of SLT-1 and toward ventral sources of UNC-6 (Fig. 3). Loss of both cues results in almost complete failure of AVM axon ventral guidance, whereas the loss of either slt-1 or unc-6 function results in ∼40% failure (Fig. 3D–F). There is no significant difference in the penetrance of the defect between unc-6(ev400) and clec-38;unc-6(ev400) double mutants or between unc-6(rh46) and clec-38;unc-6(rh46) double mutants (Fig. 3F). However, the ventral guidance defect observed in slt-1 mutants is suppressed in clec-38; slt-1 double mutants. These results suggest that loss of clec-38 function can improve the ability of UNC-6 to provide ventral guidance information. Consistent with the finding that UNC-40 is required for improving the dorsal guidance of the DA and DB motor neuron axons, the loss of clec-38 function could similarly improve the AVM response to UNC-6 by enhancing UNC-40 activity. The ability of clec-38(ur280) to improve the ventral AVM axon guidance in the slt-1 loss-of-function background further indicates that the effect of the clec-38 mutation is not dependent on altered forms of UNC-6 but can also improve guidance by the wild type UNC-6 protein.
We wanted to determine whether clec-38 expression by AVM is important for the phenotype. When the clec-38 promoter was used to express GFP, we did not specifically detect expression in AVM (Fig. 1); however, mosaic or low-level expression by the transgene could cause AVM expression to be missed. To test whether clec-38 acts cell autonomously in AVM, we expressed a clec-38 transgene under the control of the mec-4 promoter to drive expression in the touch neurons, which includes AVM. This expression was able to reverse the suppression of the AVM guidance defects caused by the clec-38 mutation in clec-38; slt-1double mutants (Fig. 3F). This indicates that AVM expression of clec-38 can account for the axon guidance phenotypes.
CLEC-38 is required for presynaptic terminal organization
From the same genetic screen used to isolate the clec-38(ur280) mutation, we isolated an allele of rpm-1, which encodes a protein that regulates presynaptic terminal (Schaefer et al., 2000; Zhen et al., 2000). To determine whether mutations in clec-38 also disrupt presynaptic terminals, we analyzed the presynaptic terminals of the GABAergic DD and VD motor neurons using a transgene expressing synaptobrevin fused to GFP (SNB-1::GFP) and driven by the unc-25 promoter (Hallam and Jin, 1998; Nonet, 1999).
In wild-type animals, SNB-1::GFP puncta are uniform in shape and are evenly distributed along the processes (Fig. 4A,B). In 86% of clec-38 mutants, the distribution of SNB-1::GFP puncta is irregular, leaving clear areas in the dorsal and ventral cord (Fig. 4C,D). The number of dorsal and ventral cord puncta is reduced in clec-38 mutants (Fig. 4E). In addition, the remaining puncta are of irregular shape and size (Fig. 4C,D). To determine whether these synaptic defects could be a result of gross axon guidance defects, we checked the morphology of DD and VD axons in clec-38 mutants. We did not observe mispositioned axons (data not shown), suggesting that axon guidance defects are not the cause of the irregular SNB-1::GFP pattern in clec-38(ur280) mutants.
clec-38 genetically interacts with rpm-1, which encodes a presynaptic terminal protein
RPM-1 is a large protein localized to the presynaptic periactive zone in the mature nervous system. Loss of rpm-1 function causes a range of defects affecting neuronal morphology and synaptic organization (Schaefer et al., 2000; Zhen et al., 2000; Nakata et al., 2005). In both clec-38 and rpm-1 mutants similar GABAergic motor neuron synapse defects are observed. In addition, loss of rpm-1 function similarly affects axon outgrowth mediated by guidance receptors, although RPM-1 negatively regulates the UNC-5 and SAX-3 receptors rather than UNC-40 (Li et al., 2008). Furthermore, RPM-1 appears to regulate the UNC-5 and SAX-3 receptors through a pathway involving the Rab GEF GLO-4; however, the guidance phenotypes of glo-4 and clec-38 mutations are different, suggesting that clec-38 do not act through this pathway.
To further explore a relationship between RPM-1 and CLEC-38 functions, we determined whether there are genetic interactions between clec-38 and rpm-1. For the DA and DB motor neurons, the axon guidance phenotype caused by unc-6(rh46) is suppressed by both clec-38 and rpm-1 mutations (Fig. 5A) (Li et al., 2008). In clec-38; rpm-1; unc-6(rh46) triple mutants, the penetrance is similar to mutants with unc-6(rh46) alone (Fig. 5A). Thus, the loss of clec-38 function reverses the effects of the loss of rpm-1. For the ventral migration of the AVM axon, we find a 90% penetrant defect in the rpm-1;slt-1double mutants, but only a 21% penetrant defect in rpm-1;slt-1;clec-38 mutants (Fig. 5B). Again, the loss of clec-38 function reverses the effects of the loss of rpm-1 function.
We also observe that double mutations have an effect on the control of axon extension. In rpm-1 mutants, the PLM anterior process axon often extends beyond the normal termination point (Fig. 6A–E) (Schaefer et al., 2000; Li et al., 2008). We find that the overextension caused by the loss of rpm-1 function is affected by the loss of clec-38 function. Interestingly, the penetrance of the overextension phenotype in clec-38; rpm-1 mutants is not greater than that of rpm-1 mutants; rather, the length of the extension increases (Fig. 6G,H). Whereas in rpm-1(ur299) mutants only 15% of the PLM axons extend beyond the position of the AVM cell body, 83% of the axons in clec-38; rpm-1 mutants extend beyond the AVM cell body (Fig. 6H). This effect is suppressed by the loss of UNC-40 in unc-40; clec-38; rpm-1 triple mutants, suggesting that the additional extension is likely attributable to the effect on UNC-40 by loss of clec-38 function. The effect is also suppressed by expression of the clec-38 transgene under the control of the mec-4 promoter, suggesting that clec-38 acts cell autonomously in the PLM touch neuron (data not shown).
clec-38 regulates the UNC-40::GFP expression pattern
To further explore the relationship between UNC-40 and CLEC-38, we used transgenic strains expressing functional UNC-40::GFP. In the first case, we used a transgene that contains the unc-40 promoter sequence to drive UNC-40::GFP expression (Chan et al., 1996). Although the construct used in making this transgenes may have an in-frame deletion of 40 aa in the third Ig domain of the protein (Joe Culotti, personal communication), it is able to rescue unc-40 null mutant phenotypes (Chan et al., 1996). In this strain, the GFP signal is detectable in all cells during early gastrulation and then gradually decreases, with later expression observed in neurons and motile cells (Chan et al., 1996). We observe that 20% (n = 50) of the clec-38(+) animals with UNC-40::GFP expression have neurons with abnormal morphologies, including multiple processes (Fig. 7). Excess axon outgrowth and the abnormal neuronal morphology is likely caused by higher levels of UNC-40 because similar phenotypes are associated with the expression of constitutively active forms of UNC-40 in neurons (Gitai et al., 2003). Compared with the clec-38(+) animals, 80% (n = 50) of mutants heterozygous for the clec-38 mutation have severe morphological defects (Fig. 7) and 100% (n = 50) of mutants homozygous for the clec-38 mutation die as embryos. In some individual neurons, differences in the intensity of the GFP signal were noticeable. For example, in ventral cord motor neurons, only weak expression was observed in 5% (n = 50) of the clec-38(+) animals, whereas 45% (n = 50) of clec-38 heterozygous mutants showed strong UNC-40::GFP expression (Fig. 7A,B). From these observations, we infer that high levels of UNC-40::GFP cause morphological changes and that loss of clec-38 activity influences the levels or cellular distribution of UNC-40. Analyses of UNC-5::GFP and SAX-3::GFP expression patterns in clec-38 mutants did not reveal major differences from the patterns seen in the wild-type background (data not shown).
We observe that overexpression of the above unc-40::gfp transgene causes severe morphological changes throughout the animal. However, because of the possible effects of the deletion within the unc-40 sequence and because the identification of individual neurons is difficult in these strains (Fig. 7C,D), we also examined the effects of a different unc-40 transgene that is expressed in fewer cells. We used a transgene that expresses UNC-40:GFP under the control of the unc-86 promoter, which drives expression in a subset of cells, which include the HSN neuron (Adler et al., 2006) (Fig. 7E). We observe that expression of this transgene in clec-38 mutants results in strong UNC-40::GFP expression and HSN morphological changes, including multiple axon outgrowths (Fig. 7F). In 78% of clec-38(ur280) mutants (n = 115), there is strong GFP signal in HSN compared with 25% (n = 105) in clec-38(+) animals. Furthermore, in 95% of the mutants, the HSN cell morphology is abnormal and has excessive axon outgrowth. These morphological defects are observed in all larval stages (data not shown). Together, these results suggest that loss of clec-38 function enhances UNC-40::GFP expression. This is consistent with the genetic results that indicate clec-38 negatively regulates UNC-40 activity.
Proteins with C-type lectin-like domains can regulate axon guidance and synaptogenesis
Our results indicate that CLEC-38 regulates axon outgrowth and presynaptic patterning. The ur280 loss-of-function allele has a missense mutation within the second CTLD, suggesting that these CLEC-38 functions are dependent on the activity mediated by this domain. To our knowledge, members of the protein superfamily that contain the CTLD have not been implicated before in axon guidance or synaptogenesis. Indeed, little is known about the function of members of this family in C. elegans, despite the fact that it is considered the second largest family of proteins in C. elegans (Drickamer and Dodd, 1999; Dodd and Drickamer, 2001). The CTLD is distinguished by sequence similarity; it is found in a variety of vertebrate and invertebrate proteins and it is known to mediate binding to carbohydrates, proteins, and inorganic molecules (Drickamer, 1999; Zelensky and Gready, 2003, 2005). Vertebrate family members have been implicated in functions such as extracellular matrix structure, endocytosis, complement activation, pathogen recognition, and cell–cell interactions. The diversity of this superfamily arises because CTLDs are found with a wide variety of other structural modules.
CLEC-38 negatively regulates UNC-40 axon outgrowth-promoting activity
We present genetic evidence that CLEC-38 negatively regulates UNC-40-mediated axon guidance activity. The enhancement of the dorsal guidance of DA and DB motor neuron axons in unc-6(rh46) mutants by loss of clec-38 function is suppressed by the loss of unc-40 function, suggesting that in the clec-38 mutant UNC-40 activity is enhanced and that this improves the ability of the axons to migrate dorsally in response to the unc-6(rh46) product. Furthermore, the ability of clec-38(ur280) to improve the ventral AVM axon guidance in the slt-1 loss-of-function background is consistent with the loss of clec-38 function enhancing the UNC-40-mediated ventral guidance response to UNC-6.
Although separate models can be made based on the phenotypes observed in each individual neuron type, here we discuss a model to explain phenotypes observed in all the neurons. We consider that the underlying guidance functions of CLEC-38 are the same in every neuron, although its function might be modified by the addition or absence of factors specific to a neuron. In this model, axon guidance is a multistep process with the axon guidance receptors involved at different steps (Fig. 8A). In the first step, the association of the ligand enables signaling that upregulates receptor levels and causes the asymmetric localization of receptors to the cell membrane where axon outgrowth will occur. In the second step, the guidance receptors promote axon outgrowth in the direction that was set up during the first step. At this step, the axon outgrowth-promoting activity of a receptor may act independently of its guidance cue ligand. As evidence for this, it was shown that either the attractive or repulsive cue, UNC-6/netrin and Slt-1/slit, respectively, can suppress multipolar AVM axon outgrowth caused by overexpression of the downstream effector MIG-10. In either case, there is enhanced ventral guidance of a single axon, indicating axon outgrowth-promoting activity that acts independently of the guidance cues can be oriented ventrally by either the attractive or repulsive cue (Quinn et al., 2006). Moreover, before the ventral outgrowth of the HSN axon, UNC-6 and UNC-40 trigger the ventral asymmetric localization of UNC-40 as well as proteins that promote actin-based protrusive activity (Adler et al., 2006; Chang et al., 2006; C. Quinn, D. Pfeil, and W. G. Wadsworth, unpublished observations).
The results presented here indicate that clec-38 negatively regulates UNC-40 axon outgrowth-promoting activity rather than the orientation activity. The outgrowth-promoting activity affects the circumferential axon guidance, as well as longitudinal extension. We do not observe evidence of polarity defects in the clec-38 mutants, as might be expected if the orientation of the outgrowth was perturbed. For example, we observe that PLM overextension caused by loss of rpm-1 function is increased in the clec-38; rpm-1 double mutants and that the penetrance of this phenotype is reduced by the loss of unc-40 function. This suggests that loss of clec-38 function enhances an UNC-40-mediated outgrowth activity for the longitudinal extension of PLM. Other genes have also been shown to affect UNC-40-mediated outgrowth in the anterior–posterior direction; however, in contrast to clec-38, these genes appear to affect neuronal polarity, with mutations causing inappropriate UNC-40 localization (Levy-Strumpf and Culotti, 2007; Watari-Goshima et al., 2007). In these mutants, the UNC-40-mediated axon outgrowth activity may be normal, but misdirected.
An UNC-6-independent UNC-40 axon outgrowth-promoting activity also helps explain why, relative to the unc-6(rh46) mutants, the axon guidance defects of the motor neuron axons, but not of the AVM axon, are suppressed in the clec-38(ur280);unc-6(rh46) mutants. We hypothesize that the unc-6(rh46) product, UNC-6 A157P, associates with UNC-40 but does not trigger signaling. Indeed, we have isolated an unc-40 allele that contains a missense mutation altering a single amino acid in the ectodomain sequence and that allows the UNC-40 mutant protein to orient axon outgrowth in unc-6(rh46) mutants (Z. Xu and W. G. Wadsworth, unpublished observations). For the motor neurons, UNC-6 A157P interacts with UNC-5 to polarize axon outgrowth dorsally. Upregulation of UNC-40 in the clec-38(ur280);unc-6(rh46) mutants allows some dorsal UNC-40-mediated axon outgrowth-promoting activity, particularly because the motor neuron growth cones travel away from ventral UNC-6 A157P sources where it is less likely that UNC-40 activity will be inhibited by an association with UNC-6 A157P. For the AVM axon migration in the clec-38(ur280);unc-6(rh46) mutants, upregulation of UNC-40 does not have the same enhancement effect that it has with the motor neuron axons because the AVM axon migrates ventrally, toward the UNC-6 A157P sources, where it is more likely that UNC-40 activity will be inhibited by the association with UNC-6 A157P.
In addition to the axon orientation and outgrowth-promoting activities of the receptors, the results suggest that interactions between receptors can cause silencing. For the guidance of the AVM axon, loss of rpm-1 function in slt-1 mutants causes severe AVM ventral guidance defects, but the addition of the clec-38 mutation reverts the penetrance of the phenotype to that observed in mutants with the slt-1 mutation alone. We recently published evidence that loss of rpm-1 function enhances SAX-3 activity and that this increase in SAX-3 silences the ventral axon guidance mediated by UNC-40 (Li et al., 2008). While this manuscript was under review, a similar model predicting that SAX-3 could inhibit UNC-40-mediated guidance of the AVM axon was suggested (Fujisawa et al., 2007). These observations are consistent with in vitro analyses reporting that an association between the DCC and Robo receptors silences the ability of DCC to mediate turning toward netrin (Stein and Tessier-Lavigne, 2001). We now observe that enhancing UNC-40 activity can reverse the SAX-3 silencing of UNC-40.
Similarly, we propose that the composition of guidance receptors influences the dorsal guidance of the DA and DB motor neuron. Loss of clec-38 function enhances UNC-40 expression in neurons and improves dorsal guidance in the unc-6(rh46) mutant. Loss of rpm-1 function also improves dorsal guidance in the unc-6(rh46) mutant; however, unlike the clec-38 mutations, the rpm-1 mutations suppress the unc-6(rh46) dorsal guidance phenotype in the absence of UNC-40, indicating that the suppression by loss of rpm-1 function is not mediated through UNC-40 (Li et al., 2008). Instead, the loss of rpm-1 function enhances UNC-5 expression in the motor neurons (Li et al., 2008). Interestingly, combining the mutations in a clec-38; rpm-1; unc-6(rh46) triple mutant does not further enhance dorsal guidance; instead, the result is a guidance phenotype similar to mutants with unc-6(rh46) alone. We propose loss of clec-38 function enhances UNC-40 outgrowth-promoting activity, but the upregulation of UNC-40 is not enough to alter UNC-5-mediated polarity effects, and therefore the direction of outgrowth remains dorsal. In the clec-38; rpm-1; unc-6(rh46) triple mutant, however, the enhanced expression of both UNC-5 and UNC-40 leads to the silencing of receptor activity, a situation that is similar to the proposed silencing by the association of SAX-3 and UNC-40 in AVM. There is evidence from experiments using cultured Xenopus spinal cord neurons that UNC-5 and DCC can physically interact, and that the expression of UNC5 in DCC-expressing neurons can convert the response to netrin from attractive to repulsive (Hong et al., 1999). This switch might occur if the introduction of UNC5 into the Xenopus neurons alters trafficking polarity enough to direct axon outgrowth activity toward the side opposite the netrin source. This may coincide with receptor complexes of UNC5 and DCC forming and silencing outgrowth activity.
Finally, consistent with the genetic evidence for the negative regulation of UNC-40-mediated activity by CLEC-38, we find that loss of clec-38 function affects the expression of transgenes encoding UNC-40::GFP. In the clec-38 mutant, morphological defects caused by transgene expression are enhanced. We interpret the more severe phenotypes to be a consequence of increased UNC-40 expression, in part because the expression of constitutively active forms of UNC-40 produce similar morphological defects (Gitai et al., 2003).
A model for CLEC-38 function
Our results indicate that CLEC-38 negatively regulates UNC-40-mediated axon guidance as well as promotes the organization of presynaptic terminals. What is the connection between axon guidance, synaptogenesis, and the functions of clec-38 and rpm-1? An attractive idea is that CLEC-38 and RPM-1 are part of a process that regulates axon outgrowth in response to signals that promote synaptogenesis (Fig. 8B). As a growth cone reaches a target, cues from the target induce CLEC-38 and RPM-1 activity that inhibits the guidance receptors and helps promote the formation of presynaptic structures. CLEC-38 negatively regulates UNC-40, whereas RPM-1 negatively regulates UNC-5 and SAX-3. Our results with double clec-38 and rpm-1 mutants indicate a complex relationship between CLEC-38 and RPM-1 with regards to their guidance functions; the loss of function of one molecule can switch the effects caused by the loss of function of the other molecule. As discussed above, we suggest that this relationship could be attributable to the silencing of receptor activities through the formation of receptor complexes comprising UNC-40 with UNC-5 or SAX-3. However, other models are possible, including more direct regulatory interactions between CLEC-38 and RPM-1. In any case, the results indicate distinct roles for CLEC-38 and RPM-1 in regulating the guidance process, and they suggest interactions that are part of a versatile system to regulate morphology changes of growth cones as they reach their targets and make connections.
This work was supported by National Institutes of Health Grant R01 NS033156 and grants from the New Jersey Commission on Spinal Cord Research. We thank S. Clark, J. Culotti, C. Bargmann, Yishi Jin, the Japanese National BioResource Project, and the Caenorhabditis Genetics Center for strains; we also thank members of the Wadsworth Laboratory for constructive input, and Sunita Kramer, Christopher Quinn, and Martha Soto for critical discussion and comments on this manuscript.
- Correspondence should be addressed to Dr. William G. Wadsworth, Department of Pathology, Robert Wood Johnson Medical School, 657 Hoes Lane West, Piscataway, NJ 08854-5635.