UNC-6/Netrin is an evolutionarily conserved, secretory axon guidance molecule. In Caenorhabditis elegans, UNC-6 provides positional information to the axons of developing neurons, probably by establishing a concentration gradient from the ventral to the dorsal side of the animal. Although the proper localization of UNC-6 is important for accurate neuronal network formation, little is known about how its localization is regulated. Here, to examine the localization mechanism for UNC-6, we generated C. elegans expressing UNC-6 tagged with the fluorescent protein Venus and identified 13 genes, which are involved in the cellular localization of Venus∷UNC-6. For example, in unc-51, unc-14, and unc-104 mutants, the neurons showed an abnormal accumulation of Venus∷UNC-6 in the cell body and less than normal level of Venus∷UNC-6 in the axon. An aberrant accumulation of Venus∷UNC-6 in muscle cells was seen in unc-18 and unc-68 mutants. unc-51, unc-14, and unc-104 mutants also showed defects in the guidance of dorso-ventral axons, suggesting that the abnormal localization of UNC-6 disturbed the positional information it provides. We propose that these genes regulate the process of UNC-6 secretion: expression, maturation, sorting, transport, or exocytosis. Our findings provide novel insight into the localization mechanism of the axon guidance molecule UNC-6/Netrin.
A variety of axon guidance molecules and their receptors are critical for pathfinding axons to reach their precise targets (Tessier-Lavigne and Goodman 1996; Yu and Bargmann 2001; Dickson 2002; Chilton 2006; Killeen and Sybingco 2008). Axon guidance molecules, providing the positional information to axons, are expressed either on the surface of cells or secreted into the extracellular space. The axons, receiving positional information from the axon guidance molecules, express axon guidance receptors at the growth cone.
Netrin is an evolutionarily conserved axon guidance molecule that has both axonal attraction and repulsion activities (Serafini et al. 1994; Colamarino and Tessier-Lavigne 1995). UNC-6 of Caenorhabditis elegans is a member of the Netrin family (Ishii et al. 1992). During C. elegans development, UNC-6 is expressed in the ventral cells, including epidermoblasts, glia, neurons, muscle cells, and vulval precursor cells (VPCs) (Wadsworth et al. 1996; Asakura et al. 2007). UNC-6 is thought to establish a concentration gradient from the ventral to the dorsal side of the animal (Wadsworth 2002), to provide ventral-dorsal positional information to attract some axons ventrally while repelling others to extend dorsally (Hedgecock et al. 1990; McIntire et al. 1992; Wadsworth 2002). In addition, UNC-6 provides positional information for cell migration (Hedgecock et al. 1990), synapse formation (Colón-Ramos et al. 2007; Poon et al. 2008), and cell polarity (Adler et al. 2006; Ziel et al. 2009). However, little is known about the molecular mechanisms of UNC-6 localization.
To examine the localization mechanisms of UNC-6, we generated C. elegans expressing UNC-6 tagged with the fluorescent protein Venus (Asakura et al. 2007) and identified 13 genes required for the cellular localization of Venus∷UNC-6, including unc-51, unc-14, unc-104, unc-18, and unc-68. In addition to being involved in the localization of UNC-6, unc-51, unc-14, and unc-104 mutants also showed defects in UNC-6-mediated axon guidance, suggesting that the inappropriate UNC-6 localization disturbed the positional information available to the axons. Our findings provide novel insight into the localization mechanisms of the axon guidance molecule UNC-6/Netrin.
MATERIALS AND METHODS
To analyze the UNC-6 localization in vivo, we used ghIs9(Venus∷unc-6; str-3p∷dsRed2) (Asakura et al. 2007) as an integrant strain. Each animal was mounted on a 2.5% agarose pad in M9 buffer containing 5% sodium azide and was observed using a fluorescence microscope (Axioplan2, Zeiss). Images were taken using a confocal microscope LSM510 (Zeiss).
Mutagenesis and genetic mapping:
The gh alleles were isolated in a screen performed according to standard protocols (Anderson 1995). Briefly, ghIs9(Venus∷unc-6) was mutagenized with ethylmethane sulphonate (EMS), and the F2 generation was screened for animals that exhibited localization defects of Venus∷UNC-6. We screened ∼3000 haploid genomes. The gh36 mutation was dominant, and the other mutations were recessive. Single nucleotide polymorphism (SNP) mapping was used for genetic mapping in the CB4856 strain (Wicks et al. 2001; Davis et al. 2005). The map position was further refined by a complementation test.
Analysis using unc-6 RNAi was performed as described by Asakura et al. (2007). Experiments using RNAi against autophagy-related genes were performed as described by Ogura and Goshima (2006). In this article, the RNAi-hypersensitive double mutant rrf-3(pk1426); hpl-2(ok917) was used (Wang et al. 2005).
We used KOD-Plus (Toyobo, KOD-201) for the PCR experiments. The unc-18 ORF was amplified from pPCR2.1F27D9#F1R1 (Gengyo-Ando et al. 1993), and inserted into the mCherry (McNally et al. 2006) expression vector pNW5 (myo-3 promoter∷mCherry) or pNW19 (H20 promoter∷mCherry), resulting in the myo-3p∷unc-18∷mCherry (pNW7) and H20p∷unc-18∷mCherry (pNW20) constructs.
Transformation of C. elegans:
Transformation was performed as described by Mello et al. (1991). myo-2p∷mRFP (Campbell et al. 2002) (pmy2P-mR) was used as the marker (10 ng/μl). pBluescript SK+ was used to equalize the amount of DNA in the transformations. Mixtures of [pBluescript SK+ (40 ng/μl), pmy2P-mR (10 ng/μl), and pNW7 (50 ng/μl)] or [pBluescript SK+ (40 ng/μl), pmy2P-mR (10 ng/μl), and pNW20 (50 ng/μl)] were injected into the YC81[unc-18(e234); ghIs9] adult gonad, resulting in YC84[unc-18(e234); ghIs9; ghEx20(myo-3p∷unc-18∷mCherry; myo-2p∷mRFP)] or YC85[unc-18(e234); ghIs9; ghEx21(H20p∷unc-18∷mCherry; myo-2p∷mRFP)].
Genetic screening to identify genes that regulate the Venus∷UNC-6 localization:
To identify genes that regulate the UNC-6 localization, we used strain ghIs9, which expresses functional and visible Venus∷UNC-6 (Asakura et al. 2007). On the wild-type background, Venus∷UNC-6 was mainly detected in ventral cells, including epidermoblasts, glia, neurons, muscle cells, and vulval precursor cells (Figure 1 ). Venus∷UNC-6 was detected in dorsal muscle cells in the tail (Figure 1). In male worms, Venus∷UNC-6 was expressed in the ray (data not shown). The general distribution pattern of Venus∷UNC-6 in the wild-type genetic background was similar to that of 3xHA-tagged UNC-6, reported previously (Wadsworth et al. 1996), except for our additional observation of Venus∷UNC-6 expression in P6.p descendants, ventral muscle, dorsal muscle in the tail, and in the ray of the male tail (Figure 1, data not shown). These differences were probably due to the different fixation methods used, because with Bouin's fixative (Nonet et al. 1997), HA staining of urIs1 (3xHA∷UNC-6, Wadsworth et al. 1996) showed an identical pattern to ghIs9 (data not shown). In addition, an unc-6 promoter∷mRFP fusion gene also showed the same pattern (data not shown).
Since UNC-6 is a secreted protein, we expected that some Venus∷UNC-6 would be detected outside of the cells. However, we could not detect any extracellular Venus∷UNC-6, probably owing to its weak fluorescence intensity. Therefore, we focused our analysis on the cellular Venus∷UNC-6 localization, and so in this article, the “localization” of Venus∷UNC-6 refers to not the extracellular but the cellular localization of Venus∷UNC-6. We believe that the observed cellular Venus∷UNC-6 localization largely reflects the process of its secretion.
To identify the genes responsible for the proper localization of Venus∷UNC-6, we took two approaches: (1) we performed EMS mutagenesis screening with ghIs9 to isolate mutant alleles in which the mislocalization of Venus∷UNC-6 was observed, and (2) we examined the localization of Venus∷UNC-6 in existing mutants of genes related to vesicular transport and secretion. From these experiments, we isolated or identified 13 genes required for the proper localization of Venus∷UNC-6 (Table 1). These mutants had no morphological defects on cell shapes except for axons and the penetrance of the localization phenotype in each mutant was 100% (data not shown).
Mutants showing an abnormal localization of Venus∷UNC-6 in neurons:
In the neurons, Venus∷UNC-6 on a wild-type background showed a punctate distribution pattern throughout the cytoplasm and axons and was excluded from the nucleus (Figure 2A ). We identified four genes required for this localization of Venus∷UNC-6 within the neurons (Figure 2, B–E; Table 1). In these mutant worms, Venus∷UNC-6 was expressed abnormally in the neurons, but the expression in other cell types was similar to that in the wild type (data not shown), suggesting that UNC-6 localization in the neurons is regulated by a unique mechanism. In unc-51(e369) and unc-14(e57) mutants, the Venus∷UNC-6 was accumulated unevenly in the neuronal cell body, and little Venus∷UNC-6 was present in the axon (Figure 2, B and C). UNC-51 is a serine/threonine kinase homologous to yeast Atg1, which is required for autophagy (Ogura et al. 1994; Matsuura et al. 1997; Straub et al. 1997; Mizushima 2007). UNC-14, a RUN domain protein, is the binding partner of UNC-51 (Ogura et al. 1997). Although the precise molecular functions are unknown, UNC-51 and UNC-14 have been implicated in membrane trafficking and localization of UNC-5, which is the receptor for UNC-6 (Ogura and Goshima 2006).
UNC-51 has been reported to be involved in the autophagy in C. elegans (Meléndez et al. 2003). However, it is unlikely that defects in the traditional autophagy pathway caused the abnormal Venus∷UNC-6 localization, because the RNAis of other genes required for autophagy (bec-1/atg-6, atg-7, lgg-1/atg-8, and atg-18) in an RNAi-hypersensitive mutant strain showed the normal Venus∷UNC-6 localization (Table 2). In unc-104(e1265) mutants, the Venus∷UNC-6 was accumulated evenly throughout the neuronal cell body, and little Venus∷UNC-6 was present in the axon (Figure 2, D and F). unc-104 encodes a kinesin motor protein, which is homologous to KIF1A in vertebrate (Otsuka et al. 1991; Hirokawa and Noda 2008). These findings therefore suggest that UNC-6 might be transported by the motor protein UNC-104 from the neuronal cell body to the axon. In gh23 mutants, Venus∷UNC-6 in neurons also accumulated evenly in the cell body, with very little appearing in the axon (Figure 2, E and F), which was similar to that in unc-104(e1265) mutants (Figure 2D). Complementation analysis revealed that gh23 was not an allele of unc-104. The similar expression patterns indicated that the responsible gene product of gh23 might be involved in the molecular function of UNC-104 in UNC-6 transport.
Mutants showing an abnormal localization of Venus∷UNC-6 in muscle cells:
In the muscle cells, Venus∷UNC-6 on a wild-type background was distributed throughout the cytoplasm and was excluded from the nucleus (Figure 3A ). We identified seven genes required for this Venus∷UNC-6 localization within muscle cells (Table 1). In these mutant worms, Venus∷UNC-6 was accumulated abnormally in the muscle cells (Figure 3, B–D), but a normal distribution pattern in other cell types was observed (data not shown), suggesting that, like neurons, a unique mechanism for controlling UNC-6 localization exists in muscle cells. Complementation analysis with existing mutants revealed that gh21, gh22, gh28, gh29, gh32, and gh37 were alleles of unc-68. The reference alleles unc-68(e540) and unc-68(r1162) exhibited the same phenotype (Figure 3B). UNC-68 is homologous to ryanodine receptors (RyRs), which regulate body-wall muscle contraction by controlling Ca2+ release from the endoplasmic reticulum (ER) (Maryon et al. 1996; Sakube et al. 1997; Zalk et al. 2007). This release of Ca2+ mediated by UNC-68 is also required for regulating both spontaneous and evoked neurotransmitter release (Liu et al. 2005). These results therefore suggest that the Ca2+ release from the ER mediated by UNC-68 is required for proper UNC-6 localization in muscle.
Complementation analysis revealed that gh27 and gh38 were allelic mutants. Since gh27 and gh38 are mapped to LGV, in which unc-68 is located (supporting information, Figure S1), we performed complementation analysis with unc-68(e540). Venus∷UNC-6 accumulated in the muscle cells of the trans-heterozygous strain e540/gh27, but the level was clearly lower than that of the homozygous strains e540/e540, gh27/gh27, and gh38/gh38 (data not shown). Since unc-68(e540) is thought to be a null allele (Sakube et al. 1997), interallelic complementation of the sort observed in hypomorphic alleles, such as for unc-5 (Merz et al. 2001), was unlikely. In addition, unc-68(e540) shows Unc phenotype, but gh27 and gh38 did not. These findings indicated that gh27 and gh38 interacted genetically with unc-68, but were not allelic to it. Complementation analysis revealed that gh33 and gh26 were alleles of unidentified genes (Table 1).
Venus∷UNC-6 was also accumulated in the muscle cells of the unc-18(e234) mutants (Figure 3C). UNC-18, which is homologous to the SM (Sec1/Munc18-like) proteins, regulates a multistep vesicle exocytosis process in neurons, in cooperation with SNARE proteins (Gengyo-Ando et al. 1993; Malsam et al. 2008; Südhof and Rothman 2009). Therefore, we next examined whether neuronal UNC-18 regulates the localization of Venus∷UNC-6 in muscle.
UNC-18 in neurons acts non-cell autonomously to regulate the localization of Venus∷UNC-6 in muscle:
Although Venus∷UNC-6 was accumulated abnormally in the muscle cells of the unc-18(e234) mutant, UNC-18 is expressed by the neurons, not by the muscles (Gengyo-Ando et al. 1993). To analyze the cell autonomy of this gene, we examined whether muscle- or neuron-specific expression of unc-18 could rescue the Venus∷UNC-6 localization defect in unc-18(e234) mutants. In these rescue experiments, we used the myo-3 promoter for muscle-specific expression (Okkema et al. 1993) and the H20 promoter for neuron-specific expression (Shioi et al. 2001).
The muscle-specific expression of unc-18 did not rescue the abnormal Venus∷UNC-6 accumulation in the muscle cells of unc-18(e234) mutants (Figure 4A ), but its neuron-specific expression did (Figure 4B). These results suggested that neuronal UNC-18 cell nonautonomously regulates the localization of UNC-6 in muscle. The possible involvement of presynaptic input in the regulation of UNC-6's localization in muscle is also supported by the phenotypic defects we observed in the syd-1(ju82); rpm-1(ju44) double mutant, in which Venus∷UNC-6 accumulated in the muscle cells (Figure 3D). The syd-1(ju82); rpm-1(ju44) mutant shows a severe defect in locomotion, since the number of synapses is reduced and presynaptic components are disrupted (Nakata et al. 2005). Our findings therefore suggest that synaptic activity is required for the proper localization of UNC-6 in the muscle.
Since unc-18 and unc-68 are known to regulate the exocytosis of synaptic vesicles in neurons, we examined the localization of Venus∷UNC-6 in the mutants of other exocytosis-related genes, such as snt-1/synaptotagmin, snb-1/synaptobrevin, unc-64/syntaxin, and unc-13/Munc13 (Maruyama and Brenner 1991; Nonet et al. 1993; Nonet et al. 1998; Ogawa et al. 1998; Saifee et al. 1998; Malsam et al. 2008; Südhof and Rothman 2009). However, the localization of Venus∷UNC-6 was not altered in these mutants (Table 2). Thus, the mechanism by which UNC-18 regulates the UNC-6 localization remains an unsolved question.
Mutants showing the abnormal localization of Venus∷UNC-6 in VPCs:
We identified one gene required for the localization of Venus∷UNC-6 in VPCs. In gh25 mutants, the intensity of Venus∷UNC-6 was increased in the VPCs and vulval cells (Figure 5, B and C ; Table 1), but its distribution was normal in the other cell types (data not shown), suggesting that VPCs have a specific mechanism for secreting UNC-6.
Since the VPCs are the UNC-6 source for HSN axon guidance (Asakura et al. 2007), we analyzed the HSN axon morphology in gh25 mutants. We found that gh25 mutants had HSN guidance defects (Figure 5D), suggesting that the abnormal Venus∷UNC-6 expression in gh25 mutants result in the impaired UNC-6 secretion and the HSN axon guidance defects.
Mutants showing a high level of Venus∷UNC-6 expression in all the UNC-6-expressing cells:
We identified one gene required globally for the normal expression level of Venus∷UNC-6. In gh36 mutants, Venus∷UNC-6 fluorescence intensity was increased in all of the Venus∷UNC-6-expressing cells (Figure 6, B, D, F, G ; Table 1). However, we found that UNC-5∷GFP (Killeen et al. 2002) fluorescence intensity was not increased in gh36 mutants (Figure S2), indicating that gh36 did not affect transgene expression or fluorescence levels in general. Unlike the other mutants, the distribution of the intracellular Venus∷UNC-6 expression was normal for all cell types. Therefore, the responsible gene product of gh36 may negatively regulate unc-6 expression.
unc-104, unc-18, and unc-68 are required for anterior ventral microtubule cell ventral guidance:
Given that UNC-6 is required for dorso-ventral axon guidance, we predicted that the altered localization of Venus∷UNC-6 in these identified mutants probably caused or reflected UNC-6 secretion defects, which should result in dorso-ventral axon guidance defects. In support of this hypothesis, unc-51 and unc-14 mutants show defective dorsally directed axon pathfinding by DD/VD neurons (McIntire et al. 1992). In 2006, we reported that unc-51 and unc-14 interact genetically with unc-6 to influence DD/VD axon guidance (Ogura and Goshima 2006). To verify this hypothesis further, we examined ventrally directed anterior ventral microtubule cell (AVM) axon guidance in unc-104, unc-18, and unc-68 mutants.
The AVM cell body is located laterally, and its axon grows ventrally to the ventral nerve cords (Figure 7, A and B ). The pathfinding of the AVM axon is affected by two parallel guidance cues, UNC-6/Netrin and Slt-1/Slit. Ventral UNC-6 attracts and dorsal SLT-1 repels the AVM axon (Figure 7A; Hao et al. 2001). The AVM axon grows ventrally at L1 stage. We confirmed that unc-104, unc-18, and unc-68 mutants had localization defects of Venus∷UNC-6 at the L1 stage as well (Figure S3).
In unc-104(e1265), unc-18(e234), and unc-68(e540) mutants, minor defects of the AVM axon guidance were observed (Figure 7G). Although unc-104; unc-6 double mutants exhibited no enhancement in AVM defects compared to unc-6 single mutants, unc-104; slt-1 double mutants exhibited enhanced defects compared to slt-1 single mutants. These results suggest that unc-104 has a role in the unc-6-pathway and is consistent with our hypothesis that the UNC-6 localization defect reflects the axon guidance defect of the AVM neuron.
Interestingly, unc-68; unc-6 double mutants exhibited enhanced defects compared to unc-6 single mutants (Figure 7G). unc-18; slt-1 and unc-68; slt-1 double mutants exhibited suppressed defects compared to slt-1 single mutants (Figure 7G). The enhancement in unc-68; unc-6 double mutants may result from the accumulation of other axon guidance molecules, such as SLT-1, in unc-68 mutants. The absence of enhancement in unc-18; slt-1 and unc-68; slt-1 double mutants suggests that UNC-6 expressed by the muscle cells do not participate in the AVM ventral guidance.
UNC-6 is also required for synaptic development of the DA9 neuron (Poon et al. 2008). We analyzed the synaptic development of the DA9 neuron in unc-18 and unc-68 mutants. However, we did not find the defects. It is also reported that synaptic development of the DA9 neuron is normal in unc-18 mutants (data not shown; Zhao and Nonet 2000). These suggest that UNC-6 expressed by muscle cells does not participate in the synaptic development of the DA9 neuron. UNC-6 is also required for dorsal migration of distal tip cells (Hedgecock et al. 1990). However, we did not find the dorsal migration defects of the distal tip cells in unc-18 and unc-68 mutants as well (data not shown). We could not find clear defects on the UNC-6 function in unc-18 and unc-68 mutants.
Identification of genes required for proper Venus∷UNC-6 localization:
The secretory axon guidance molecule UNC-6/Netrin provides positional information for axon guidance (Tessier-Lavigne and Goodman 1996; Yu and Bargmann 2001; Dickson 2002; Chilton 2006; Killeen and Sybingco 2008). The temporal and spatial expression of UNC-6/Netrin has been well documented and plays an important role in neural network formation (Wadsworth et al. 1996; Watanabe et al. 2006; Asakura et al. 2007). A model for the patterning mechanism, in which the Netrin receptor frazzled rearranges secreted Netrin in Drosophila melanogaster, has been proposed (Hiramoto et al. 2000). However, little is known about the localization/secretion mechanisms of UNC-6/Netrin.
In this study, we identified 13 genes required for the cellular localization of Venus∷UNC-6 in C. elegans. Four genes were required specifically for the proper localization of Venus∷UNC-6 in neurons, 7 for its localization in muscle, 1 for its localization in VPCs, and 1 for controlling the global expression level of Venus∷UNC-6. We propose that, in these cells, these 13 genes regulate the processes associated with the secretion of UNC-6: expression, maturation, sorting, transport, and exocytosis, and that each of these cell types has a specific mechanism for regulating UNC-6 localization.
Genes required for the UNC-6 localization in neurons:
Venus∷UNC-6 is expressed in neurons, including PVT, AVG, RIF, AVA, AVB, PVQ, VA, and VB(Wadsworth et al. 1996; Asakura et al. 2007). In unc-51, unc-14, unc-104, and gh23 mutants, Venus∷UNC-6 accumulated in the neuronal cell bodies, but there was little fluorescence in the axons, suggesting that these genes regulate the transport of UNC-6 from the neuronal cell body to the axon. In unc-51, unc-14, and unc-104 mutants, UNC-6/Netrin-mediated dorso-ventral axonal guidance is defective (McIntire et al. 1992; this study). In addition, unc-51, unc-14, and unc-104 interact genetically with unc-6 (Ogura and Goshima 2006; this study). These findings suggest that the defects in UNC-6 transport in these mutants disrupted the normal secretion of UNC-6 from the neurons.
UNC-104 is a homolog of the kinesin motor protein, KIF1A, which transports the precursors of synaptic vesicles (SVs) and dense core vesicles (DCVs) from neuronal cell bodies to synapses (Hall and Hedgecock 1991; Otsuka et al. 1991; Yonekawa et al. 1998; Zahn et al. 2004; Hirokawa and Noda 2008). The pattern of accumulation of Venus∷UNC-6 was very similar to that of SVs and DCVs in unc-104 mutants (Hall and Hedgecock 1991; Nonet 1999; Zahn et al. 2004), supporting our hypothesis that UNC-104 transports vesicles containing UNC-6 from the neuronal cell body to the axon. In addition, UNC-6 may be secreted by neurons at the synapse, since UNC-104/KIF1A transports the precursors of synaptic vesicles (Hirokawa and Noda 2008). The phenotype of the gh23 mutant was very similar to that of unc-104, therefore the responsible gene product of gh23 may be involved in UNC-104 function. In mutants of unc-116 and osm-3, which encode the kinesin motor proteins UNC-116/KIF5 and OSM-3/KIF17, respectively (Patel et al. 1993; Shakir et al. 1993), the localization of Venus∷UNC-6 was identical to that in wild-type animals, indicating that these kinesin motor proteins are not involved in the transport of UNC-6. This is consistent with the results of a series of recent studies showing that the kinesin and dynein motor proteins use specific adaptor or scaffold proteins to recognize and bind different cargoes (Hirokawa and Takemura 2005).
The Venus∷UNC-6 accumulation phenotype in unc-51 and unc-14 mutants was different from that in unc-104 mutant. In unc-51 and unc-14 mutants, Venus∷UNC-6 accumulated unevenly in the cell body, whereas in the unc-104 mutant, it accumulated evenly throughout the cell body except in the nucleus. The difference may reflect the roles of these gene products in the regulation of UNC-6 localization. UNC-51 is a serine/threonine kinase, which binds UNC-14, a RUN domain protein (Ogura et al. 1994, 1997). UNC-51 and UNC-14 are predicted to play important roles in vesicle trafficking. UNC-51 can bind VAB-8, a kinesin-like protein, and phosphorylate VAB-8 in vitro (Wolf et al. 1998; Lai and Garriga 2004). UNC-51 and VAB-8 cooperatively regulate the posterior axonal outgrowth of CAN neurons. UNC-14 can bind UNC-16/JIP, and together they cooperate with a kinesin motor protein UNC-116/KIF5 to transport synaptic vesicles (Sakamoto et al. 2005). The UNC-51 of D. melanogaster phosphorylates UNC-76/FEZ1, a kinesin heavy chain adaptor protein, to regulate axonal transport (Toda et al. 2008). In addition, UNC-51 and UNC-14 regulate the localization (trafficking) of UNC-5, a receptor for UNC-6, in neurons (Ogura and Goshima 2006). Therefore, UNC-51 and UNC-14 may regulate the processing involved in UNC-6's localization in neuronal cell bodies, including UNC-6's maturation, selection, or transport (Figure 8A ). Furthermore, UNC-104 may function as a motor protein in concert with the activities of UNC-51 and UNC-14 to transport vesicles containing UNC-6 from cell bodies to axons.
UNC-51 is also required for autophagy in C. elegans (Meléndez et al. 2003). However, the traditional autophagy pathway probably does not participate in the Venus∷UNC-6 localization, since the RNAi of other genes required for autophagy resulted in normal Venus∷UNC-6 localization.
Genes required for appropriate UNC-6 localization in muscle cells:
We identified seven genes required for the proper localization of Venus∷UNC-6 in muscle cells. These were unc-18, unc-68, syd-1, rpm-1, and the responsible genes of gh33, (gh27, gh38), and gh26. In the unc-18 mutant, Venus∷UNC-6 accumulated in muscle cells. This finding was unexpected, because UNC-18 belongs to the SM (Sec1/Munc18-like) protein family (Gengyo-Ando et al. 1993; Malsam et al. 2008; Südhof and Rothman 2009), which regulates vesicle exocytosis by interacting with syntaxin, a member of the SNARE proteins in neurons (Sassa et al. 1999; Weimer et al. 2003; Malsam et al. 2008; Südhof and Rothman 2009). Indeed, UNC-18 is expressed in neurons but not in muscle (Gengyo-Ando et al. 1993). We showed that the muscle-specific expression of unc-18 did not rescue the Venus∷UNC-6 accumulation in unc-18 mutants, whereas the neuron-specific expression of UNC-18 did. These results suggested that UNC-18 in neurons regulates the UNC-6 localization in muscle.
How does UNC-18 regulate the Venus∷UNC-6 localization in muscle? We showed that, except for UNC-18, mutations in the genes encoding SNARE proteins and other proteins that are essential for the exocytosis of neurotransmitters (Malsam et al. 2008; Südhof and Rothman 2009) did not cause the abnormal localization of Venus∷UNC-6 (Table 2), indicating that the traditional machinery for neurotransmitter release is not involved in the UNC-18 function. In addition, there were no defects in the localization of Venus∷UNC-6 in the muscle of the unc-13, unc-25, and unc-31 mutants (Table 2). UNC-13/Munc13, UNC-25/glutamate decarboxylase 1, and UNC-31/CAPS are, respectively, required for acetylcholine (ACh) secretion (Maruyama and Brenner 1991), GABA synthesis (Jin et al. 1999), and neuropeptide secretion (Berwin et al. 1998; Speese et al. 2007). These findings therefore indicate that traditional neurotransmitters such as ACh, GABA, or neuropeptides are also not involved in the UNC-18 function. The normal Venus∷UNC-6 localization observed in the unc-25 mutant indicated that the muscle homeostasis regulated by GABA (Garcia et al. 2007) is not involved in the UNC-18 function. It is also unlikely that the abnormal Venus∷UNC-6 localization in unc-18 resulted from the severe defect in muscle contraction observed in this mutant, since the unc-13 mutant, which also displayed a paralyzed phenotype, showed normal localization of Venus∷UNC-6.
We propose that an unknown signal from neurons, mediated by UNC-18, regulates the UNC-6 localization in muscle (Figure 8B). The unknown signal is probably secreted at synapses, since Venus∷UNC-6 also accumulated in the muscles of syd-1; rpm-1 mutants, in which presynaptic components are reduced and disrupted (Nakata et al. 2005). The secretion machinery associated with this unknown signal is probably different from that of traditional neurotransmitters. Although the signal and its secretory machinery remain unidentified, the characterization of gh27, gh33, or gh26 may reveal the mechanisms responsible for the proper localization of UNC-6 in muscle.
Venus∷UNC-6 also accumulated in muscle in the unc-68 mutant. UNC-68 is homologous to ryanodine receptors (RyRs), a class of Ca2+ channels (Maryon et al. 1996; Sakube et al. 1997; Zalk et al. 2007). UNC-68 plays important roles in muscle contraction, mediating Ca2+-induced Ca2+ release (CICR) from the endoplasmic reticulum (ER). In addition, CICR in neurons is required for regulating both spontaneous and evoked neurotransmitter release (Liu et al. 2005). Therefore, the simplest explanation for the UNC-68 function is that, in the muscle cells and/or neurons, CICR from the ER mediated by UNC-68 is required for the proper localization of UNC-6 in muscle. Unfortunately, we could not make an UNC-68 construct that expressed specifically in the muscle or neurons, because of the large size of the ORF (15.6 kb). It therefore remains unclear where and how UNC-68 regulates the localization of UNC-6 in muscle cells.
We found that unc-18 and unc-68 mutants had localization defects of Venus∷UNC-6 in muscle cells. However, our results suggest that unc-18 and unc-68 do not appear to participate in known UNC-6 functions including AVM axon guidance, DA9 synaptic development, and cell migration of distal tip cells. We think that UNC-6 expressed by muscle cells may have unknown functions, or that functions of UNC-6 expressed by muscle cells may be masked by UNC-6 expressed by neurons. Another possibility is that, although unc-18 and unc-68 mutants showed Venus∷UNC-6 localization defects in muscle cells, UNC-6 secretion from the muscle cells could be normal in these mutants.
Genes required for the UNC-6 localization in vulval precursor cells:
UNC-6 expressed by the VPCs is essential for proper HSN axon guidance (Asakura et al. 2007). Venus∷UNC-6 was accumulated in the VPCs in the gh25 mutant. In addition, HSN axon guidance defects were observed in the gh25 mutants. Taken together, these findings suggest that the responsible gene of gh25 plays an important role in the UNC-6 secretion by the VPCs.
The Venus∷UNC-6 accumulation in the gh25 mutant resembled the EGL-17/FGF accumulation in EGL-17-secretion defective mutants (Kamikura and Cooper 2003, 2006). EGL-17 is a secreted protein that is expressed in the VPCs and attracts sex myoblasts (Burdine et al. 1998). The UNC-6 secretion by the VPCs may be mediated by a mechanism similar to that for EGL-17 secretion.
Genes required for UNC-6 expression:
In mice, the transient expression of Netrin 1 at the dorsal spinal cord is required for the accurate axon guidance of primary sensory axons (Watanabe et al. 2006), and transcription factors such as Runx3 are also involved in the axon guidance of primary sensory axons (Inoue et al. 2002). In C. elegans, the upregulation of UNC-6 in the vulval precursor cells is essential for the complex axon guidance of the HSN neurons (Asakura et al. 2007). These observations suggest that accurate construction of the nervous system requires axon guidance molecules to be expressed at the proper time, place, and concentration.
In mutant gh36, the Venus∷UNC-6 fluorescence intensity was increased in all the cells that expressed Venus∷UNC-6, without any alteration in its intracellular distribution. Therefore, the responsible gene of this mutant may negatively regulate unc-6 expression. The analysis of the gh36 mutant may provide information about how UNC-6 expression is regulated.
Other mechanisms of Netrin localization:
In D. melanogaster, a model for Netrin's patterning mechanism has been proposed, in which Netrin's localization is regulated by interaction with its receptor, Frazzled/UNC-40 (Hiramoto et al. 2000). In C. elegans, we could not find any alteration in the localization of Venus∷UNC-6 in unc-40 mutants (data not shown). Since the nervous system of C. elegans is extremely simple compared to that of D. melanogaster, such a redistribution mechanism of UNC-6 might not be required for it to form. Determining whether such a redistribution mechanism for Netrin is conserved in mammalian species is an important issue to be addressed in the future.
Finally, we used strong loss-of-function alleles or RNAi to examine the localization of Venus∷UNC-6 on the known mutants or genes. Since all of them are not null alleles, it remains possible that the genes listed in Table 2 could be involved in regulating the localization of UNC-6. In addition, all the new mutants except for gh36 were recessive alleles, therefore, we think that these mutants except for gh36 are loss-of-function alleles. However, they could be gain-of-function alleles, since we did not identify the responsible genes.
We thank Takeshi Ishihara for the H20 promoter clone, Roger Y. Tsien for the mRFP clone, Jon Audhya for the mCherry clone, Andrew Fire for the C. elegans expression vectors, Keiko Gengyo-Ando for the unc-18 clone, Yuji Kohara for the C. elegans EST clones, Joseph G. Culotti for nuIs9, Takako Okada for technical support, Sandy Chen and Ayako Asakura for reading the manuscript, and members of the Goshima laboratory for suggestions and helpful discussion. Some nematode strains used in this study were provided by the Caenorhabditis Genetics Center, which is funded by the National Institutes of Health National Center for Research Resources. This work was supported by Core Research for Evolutional Science and Technology (CREST), Japan Science and Technology Agency (JST) (Y.G.), grants-in-aid for scientific research in a priority area (Y.G. and K.O.) from the Ministry of Education, Science, Sports and Culture, and the Yokohama Foundation for Advancement of Medical Science (T.A. and K.O.).
Supporting information is available online at http://www.genetics.org/cgi/content/full/genetics.110.116293/DC1.
Communicating editor: O. Hobert
- Received March 3, 2010.
- Accepted April 2, 2010.
- Copyright © 2010 by the Genetics Society of America