SEARCH

SEARCH BY CITATION

Keywords:

  • UNC-6/Netrin;
  • UNC-5;
  • UNC-40/DCC;
  • ENU-3;
  • AVM;
  • DA and DB motor neuron axon outgrowth

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. ACKNOWLEDGMENTS
  8. REFERENCES

Background: UNC-6 and SLT-1 guide the migrations of the ventrally directed processes of the AVM and PVM touch receptor neurons and UNC-6 guides the axons of the DA and DB classes of motor neurons in C. elegans. The UNC-6 receptors are UNC-5 and UNC-40. The axon outgrowth defects of a subset of the DB motor neurons in the absence of UNC-5 are enhanced by mutations in enu-3. Results: An enu-3 mutation enhances defects in ventral guidance of the processes of the AVM and PVM touch receptor neurons, the dorsal guidance of the distal tip cell and causes additional architectural defects in axons in unc-40 mutant strains in an UNC-6 dependent manner. These observations suggest that ENU-3 and UNC-40 function in parallel pathways dependent on UNC-6. ENU-3 depends on the presence of UNC-40 for its full effect on motor neuron axon outgrowth. Conclusions: ENU-3 works in an UNC-6 dependent pathway parallel to UNC-40 in ventral guidance of AVM and PVM and in dorsal guidance of the distal tip cells. Motor neuron axon outgrowth defects are caused by the presence of UNC-40 and the absence of functional UNC-5 or UNC-6 and defects are enhanced by the absence of functional ENU-3. Developmental Dynamics 243:459–467, 2014. © 2013 Wiley Periodicals, Inc.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. ACKNOWLEDGMENTS
  8. REFERENCES

Neuronal axon guidance in Caenorhabditis elegans is controlled by guidance cues including Netrin (UNC-6), Slit (SLT-1), Wnts, Ephrins, Semaphorins, as well as the TGFβ-related molecule UNC-129 (reviewed in Killeen and Sybingco, 2008; Kolodkin and Tessier Lavigne, 2011). These same guidance cues also function in other organisms in similar manners (Chilton, 2006; Charron et al., 2007; Kolodkin and Tessier-Lavigne, 2011). During development, migrating axons have a dynamic structure called a growth cone at the leading edge that samples the environment through which the neuron travels (Dent et al., 2011). The growth cones exhibit cytoskeletal alterations in response to guidance cues that result in directed migrations.

UNC-6 is a bifunctional guidance cue produced by ventral epidermoblasts, the ventral cephalic sheaths of the nerve ring and the midline neurons of the ventral cord and the pharynx (Wadsworth et al., 1996). UNC-6 has two well-known receptors, UNC-40/DCC and UNC-5 (Leung-Hagesteijn et al., 1992; Chan et al., 1996). The mammalian homologues of UNC-6 are the Netrins that are expressed in the ventral spinal cord including the floor plate (Serafini et al., 1994; Kennedy et al., 1994). Netrin functions as a chemo-attractive cue for neurons expressing DCC and a chemo-repulsive cue for neurons expressing UNC-5 (Kennedy et al., 1994; Colamarino et al., 1995; Round and Stein, 2007).

UNC-40 is expressed on the surface of motile cells including the AVM and PVM touch receptor neurons, the HSN, PDE, PHA, and PHB neurons that all have processes directed toward ventral sources of UNC-6 (Chan et al., 1996). UNC-40 is also expressed in motor neurons and in the distal tip cell. DCC (deleted in colorectal cancer), the vertebrate homologue of UNC-40 is expressed on commissural neurons and it is a Netrin receptor (Keino-Masu et al., 1996).

The second UNC-6/Netrin receptor UNC-5, is expressed in the dorsally directed axons of the DA, DB, VD, DD, and AS motor neurons, the distal tip cells, and the excretory cells (Leung-Hagesteijn et al., 1992). UNC-5 activity is required cell autonomously for dorsal-ward cell and pioneer motor axon migrations. Ectopic expression of UNC-5 in the AVM touch receptor neuron causes them to migrate in an abnormal dorsal direction and this migration is dependent on UNC-6 (Hamelin et al., 1993; Su et al., 2000). Expression of UNC-5 is also co-incident with dorsal migration of the distal tip cells at the leading edge of the developing gonad and premature expression of UNC-5 results in earlier dorsal migration of the cell (Su et al., 2000; Killeen et al., 2002). UNC-40 is also expressed in many of the same neurons and cells as UNC-5 but in these cases UNC-5 promotes guidance away from sources of UNC-6. UNC-5 was shown to be a Netrin receptor in mammals (Leonardo et al., 1997). It was shown that the intracellular regions of the vertebrate version of UNC-5 and UNC-40/DCC interact and convert Netrin/UNC-6 induced growth cone attraction to repulsion (Hong et al., 1999).

Guidance of the processes of the AVM and PVM touch receptor neurons that migrate in the ventral direction in C. elegans is controlled by two opposing guidance cues; UNC-6/Netrin expression in ventral regions and SLT-1/Slit expression in the dorsal muscle (Ishii et al., 1992; Hao et al., 2001). There are six touch receptor or mechano-sensory neurons called AML/R and PML/R (anterior and posterior lateral microtubule neuron, left and right), AVM and PVM (anterior and posterior ventral microtubule neuron) that sense gentle body touch (Chalfie and Sulston, 1981; Chalfie et al., 1985; Durbin, 1987). PVM is on the left side and has a cell body located laterally in the posterior half of the animal. A single process grows ventrally to form a commissure, joins and runs along the VNC and terminates in the anterior body. AVM is located laterally on the right side of the body in the anterior half. There is a ventrally directed process that enters the VNC (ventral nerve cord) as a commissure. It runs along the VNC beside the process from the PVM until it terminates just after the first bulb of the pharynx. UNC-40 is the only known UNC-6 receptor in AVM and PVM and it guides the processes toward ventral sources of UNC-6.

SLT-1/Slit is the second cue involved in guidance of the ventrally directed processes of AVM and PVM and it functions as a repulsive cue. The earliest SLT-1 expression is in a cap of cells in the anterior region of the comma staged embryo (Hao et al., 2001). By the two-fold stage there is expression in the most anterior epidermal cells, the anterior head muscle cells and some anterior socket cells. At the two-fold stage, expression is seen in the pharynx. SLT-1 is also expressed in the dorsal body wall muscles in the two-fold stage, initially in the posterior and extending in the anterior direction so that by L1 it is expressed by all the dorsal body wall muscles where it continues to be expressed.

SAX-3/Robo is one SLT-1 receptor and it regulates ventral guidance of the AVM and PVM processes cell-autonomously (Zallen et al., 1998). Ectopic expression of SLT-1 in ventral muscles disrupted guidance of AVM and PVM showing that asymmetric expression in the worm body is important (Hao et al., 2001). A second SLT-1 receptor has been identified called EVA-1 that also functions in AVM and PVM guidance (Fujisawa et al., 2007). In the absence of both UNC-6 and SLT-1, the axon exiting the AVM cell body usually migrates in the anterior direction toward the head, rather than toward the ventral cord. This suggests that either cue working through its cognate receptor can polarize the axons.

The guidance of the ventrally directed processes of AVM and PVM is regulated partly by the expression level of the UNC-40 receptor. This was shown when mutations in the receptor tyrosine phosphatase CLR-1 could suppress AVM ventral migration defects in strains with mutations in SLT-1 or SAX-3 but not in a strain lacking UNC-6 (Chang et al., 2004). The mutations in CLR-1 could also suppress AVM migration defects in a strain with a mutation in both SLT-1 and UNC-6, all suggesting that CLR-1 works in an UNC-6 dependent pathway. The experiments indicated that CLR-1 limits the activity of UNC-40 so that an increase in UNC-40 activity could suppress defects caused by loss of SLT-1 repulsion of AVM. Altering the level of UNC-40 can also be achieved by increasing its transcription level by up-regulating the transcription factor LIN-14 production (Zou et al., 2012). The increased UNC-40 can suppress the AVM and PVM guidance defects of slt-1 mutants. LIN-14 translation is controlled by the lin-4 microRNA (Olsen and Ambros, 1999). The lin-4 microRNA is likely to be expressed in AVM and PVM processes that migrate ventrally and not in the ALMs and PLMs that migrate longitudinally and it serves to regulate ventral guidance through its effect on UNC-40 expression (Zou et al., 2012).

ENU-3 is a newly described protein of unknown function that is predicted to have a signal peptide, a trans-membrane domain and a coiled-coil region (Yee et al., 2011). Strains with mutations in the enu-3 gene enhance the axon outgrowth defects of the DB4, 5, and 6 sub-class of motor neuron axons observed in strains lacking functional UNC-5 or UNC-6 (Yee et al., 2011). These findings suggest that ENU-3 works in parallel to the UNC-6/Netrin and UNC-5 pathway for axon outgrowth from the cell body. ENU-3 also plays a role in pathfinding because strains mutant in ENU-3 also have very mild defects in the guidance of the DA and DB axons, although they all reach the dorsal cord successfully (Yee et al., 2011). Mutations in ENU-3 also enhance the axon migration defects of the motor neurons in a hypomorphic strain with a defect in UNC-5, suggesting that ENU-3 and UNC-5 work together in motor axon guidance in an UNC-6 dependent manner. There are five other members of the ENU-3 family in the C. elegans genome called C38D4.1, Y37D8A.12, W03G9.3, W05F2.2, and K01G5.3, all of unknown function.

In this current work, we have examined the role of ENU-3 in the guidance of the ventrally directed AVM and PVM touch receptor neurons and reexamined the roles of UNC-6, UNC-5, and UNC-40 in the outgrowth of the DA and DB motor neuron axons.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. ACKNOWLEDGMENTS
  8. REFERENCES

Mutation in enu-3 Enhances the AVM and PVM Ventral Guidance Defects in an unc-40 Mutant Strain

ENU-3 is a putative trans-membrane protein of 204 amino acids with a predicted signal peptide and coiled-coil regions in the intracellular region. Mutations in the enu-3 gene enhance the motor axon outgrowth defects of a sub-set of the DB axons in strains lacking functional versions of either UNC-5 or UNC-6 (Yee et al., 2011).

Migrations of processes from the AVM or PVM touch receptor neurons toward the ventral nerve cord are affected by mutations in the genes encoding UNC-6/Netrin and its receptor UNC-40/DCC/Frazzled (Hedgecock et al., 1990; Chan et al., 1996). These processes normally migrate from their cell bodies located in the mid-body in a ventral direction and enter the ventral cord where they migrate anteriorly toward the head (Fig. 1A). In mutant animals the processes fail to make the ventral migration and instead migrate directly toward the head following an abnormal trajectory (Fig. 1B).

image

Figure 1. The touch receptor neurons are affected by mutations in both UNC-40 and ENU-3. A: The AVM in a wild-type animal taken under ×400 magnification. B,C: An AVM defect in an unc-40(e1430) mutant (B) and an unc-40(e1430); enu-3(tm4519) mutant (C, at ×200) where the PLM has kinks along its length.

Download figure to PowerPoint

Animals with a deletion in the enu-3 gene are mildly egg laying defective, often suggestive of defects in guidance of ventrally directed neurons. To determine whether a mutation in enu-3 affected ventral guidance of the AVM and PVM touch receptor neurons working with the UNC-6/Netrin pathway members, we constructed strains with mutations in enu-3 and either unc-6 or unc-40. We found that enu-3(tm4519) had no defects in the ventral guidance of the processes (Table 1). We also found that a mutation in enu-3 did not significantly enhance unc-6(ev400), a putative null strain. However, a mutation in enu-3 significantly enhanced the guidance defects of the ventrally directed processes of the AVMs and PVMs in an unc-40(e1430) background (Fig. 1C; Table 1). These observations suggest that ENU-3 has an instructive role for AVM and PVM working with UNC-40 to promote axon outgrowth and/or guidance in a pathway that is UNC-6 dependent. This finding is significant as it has been known for some time that the AVM and PVM axon guidance defects seen in the absence of UNC-6 exceed those for the only known ventral guidance receptor in C. elegans, UNC-40, suggesting the existence of a second receptor involved in ventral guidance. Possibly, ENU-3 itself or a protein in the same pathway is a second UNC-6 receptor.

Table 1. Guidance Defects of the AVM and PVM Mechano-sensory Neurons in the Absence of UNC-40 Are Enhanced by a Mutation in enu-3
Strain (with mec7p::GFP)% AVM Defective% PVM Defectiven
  1. a

    P value of less than 0.05.

  2. b

    P value of less than 0.01.

  3. c

    P value of less than 0.001 on a chi squared test.

enu-3(tm4519)00<100
unc-40(e1430)1927251
unc-40(e1430);enu-3(tm4519)39c50c139
unc-6(ev400)3646200
enu-3(tm4519);unc-6(ev400)374794
slt-1(eh15)403109
enu-3(tm4519);slt-1(eh15)53a7a120

Guidance in the ventral direction of the AVM and PVM processes is also influenced by a transient source of SLT-1 emanating from dorsal muscle. We found that the ventral guidance defects of both AVM and PVM of a strain lacking functional SLT-1 was also enhanced by the absence of ENU-3 as would be expected for two proteins that work in parallel pathways (Table 1). Therefore, ENU-3 could be in a separate UNC-6 dependent pathway parallel to UNC-40, possibly modulating the function of a second UNC-6 receptor.

A role for both UNC-40 and ENU-3 in establishing and or maintaining the overall architecture of the touch receptor neurons was suggested by the observation that many of the axons in the unc-40(e1430);enu-3(tm4519) mutant, especially the longitudinal processes of the PLMs had “kinks” along their length and other types of misdirection problems, particularly in males (Fig. 1C). In the animal shown in Figure 1C, the PLM process also appears to have under-migrated. Occasionally, both the AVM and PVM cell bodies were in the anterior or the posterior of the animals. These kinds of misdirection issues were sometimes seen in unc-40(e1430) but were more frequent in the double mutants.

To gain further insight into the enhancement of the AVM and PVM defects we performed RNA interference experiments in two different unc-40 strains and one enu-3 mutant strain using clones from the Ahringer RNAi library (Kamath et al., 2003). There was a significant difference between the AVM defects caused by the reduction of the enu-3 mRNA in both unc-40(tm5504) and unc-40(e1430) relative to the strains that received no RNAi treatment or following treatment with the vector alone (Fig. 2). There was also a difference in the PVM defects in unc-40(e1430) but not unc-40(tm5504). Reducing the level of unc-40 mRNA in both the wild-type and enu-3(tm4519) strains also caused a significant increase in the AVM defects relative to the animals treated with vector alone. The level of AVM guidance defects observed was higher in enu-3(tm4519) than in wild-type animals. Reducing the level of unc-40 mRNA also increased the PVM defects of enu-3(tm4519) but not wild-type animals. Both the genetics and the RNA interference experiments suggest that lowering the level of mRNA corresponding to enu-3 in two different unc-40 strains and unc-40 in an enu-3 mutant strain enhances ventral guidance defects of the AVMs and also of the PVMs but to a lesser extent.

image

Figure 2. Knockdown of genes by RNA interference alters AVM and PVM defects. Strains indicated were treated with bacteria containing either pL4440 which is the control plasmid with no insert or with an insert of the gene shown. Each experiment was conducted three times and 100 animals were counted each time. The percentage of affected axons is shown on the Y axis. The mean is plotted for each data set and the standard deviation is shown. A: The AVM defects obtained from treatment of N2, unc-40(tm5504) or unc-40(e1430) carrying muIs32 (mec-7::gfp) with either the control plasmid or the vector containing the enu-3/H04D03.1 RNAi clone. B: The PVM defects obtained in the same animals using the same treatments as A. C: The AVM defects resulting from treatment of either N2 or enu-3(tm4519) both carrying mec-7p::gfp with either the control plasmid or the plasmid containing the unc-40 RNAi clone. D: The PVM defects resulting from the same treatment as shown in Figure 2C. *P value of less than 0.05, **P value of less than 0.01, ***P value of less than 0.001 on a chi squared test.

Download figure to PowerPoint

Mutant ENU-3 Enhances the Distal Tip Cell Defects of a Strain With a Mutation in UNC-40

Strains with mutations in UNC-6, UNC-5, or UNC-40 all have defects in distal tip cell migrations. These cells of mesodermal origin at the leading edge of the developing gonad migrate from either side of the vulva, along the ventral side of the animal, pause, then move toward the dorsal side and finally they travel back toward the vulva on the dorsal side. In unc-6, unc-5, and unc-40 mutant animals the second phase of the migrations is affected such that the cells sometimes fail to make the dorsal migration and instead migrate back toward the vulva on the ventral side, thus creating ventral clear patches (Hedgecock et al., 1990). We looked at the distal tip cell migrations in various strains and found that a mutation in enu-3 enhanced the defects in distal tip cell migration of unc-40(e1430) but had no substantial effect on its own or in addition to either an unc-6 or an unc-5 putative null mutation (Table 2). This experiment supports the idea that ENU-3 works in the UNC-6 pathway, parallel to UNC-40 or ENU-3 could modulate the availability of either UNC-6 or UNC-5.

Table 2. enu-3 Enhances the Distal Tip Cell Migration Defects of unc-40(e1430)a
Strain% Anterior DTCDefective% Posterior DTCDefectiven
  1. a

    Animals were scored under 400× magnification and the gonads were considered defective if the gonad reflexed back toward the vulva on the ventral side creating a ventral clear patch.

  2. aP value of less than 0.05.

  3. b

    P value of less than 0.01.

  4. c

    P value of less than 0.001 on a chi squared test.

N20197
enu-3(tm4519)02152
unc-40(e1430)326100
unc-40(1430);enu-3(tm4519)37c48c193
unc-6(ev400)476592
enu-3(tm4519);unc-6(ev400)4762126
unc-5(e53)2960100
enu-3(tm4519):unc-5(e53)3160101

Enhancement of DA and DB Motor Axon Outgrowth Observed in unc-5 Mutant Animals is Partly UNC-40 Dependent

The idea that ENU-3 works in an UNC-6 dependent pathway for guidance of the touch receptor neurons and the distal tip cell caused us to revisit the role of ENU-3 in motor neuron outgrowth. Strains with mutations in either enu-3 or unc-40 or both together had few motor neuron axon outgrowth defects in a sub-set of the DA and DB classes of motor neurons (Table 3). A mutation in enu-3 was shown to enhance the axon outgrowth defects of a sub-set of motor neurons in a worm strain lacking UNC-5 or its ligand UNC-6 (Yee et al., 2011; Table 3). In enu-3(tm4519);unc-5(e53);evIs82b animals some of the DA and DB motor neurons, particularly DB4, DB5 and DB6 were often defective in axon outgrowth (Yee et al., 2011; Table 3). We showed that unc-6(ev400) (a putative null mutant) had fewer motor axon outgrowth defects than unc-5(e53) and only had significant outgrowth defects in DB5. There were increased outgrowth defects in enu-3(tm4519);unc-6(ev400) relative to unc-6(ev400) particularly in DB5, showing that a mutation in enu-3 also enhanced a mutation in unc-6. We, therefore, concluded that because mutations in enu-3 enhanced the motor axon outgrowth defects of both an unc-6 and an unc-5 mutant, ENU-3 must work in a pathway parallel to the UNC-6/Netrin pathway.

Table 3. Outgrowth Defects of the DA and DB Motor Neuron Axons Are Dependent on the Presence of UNC-40 and ENU-3
Strain (with evIs82b)DB3DA3DB4DA4DB5DA5DB6DA6DB7n
  1. The numbers shown are the percentage of defective animals. Data in lines 6 and 7 are reproduced from Yee et al., 2011. Each double is compared with the appropriate single mutant. The triple mutant in the last line is compared with the double mutant in the line above.

  2. a

    P value of less than 0.05.

  3. b

    P value of less than 0.01.

  4. c

    P value of less than 0.001 on a chi squared test.

enu-3(tm4519)000000000100
unc-40(e1430)001032100100
unc-40(e1430);enu-3(tm4519)001100100100
unc-5(e53)0151133510100
enu-3(tm4519);unc-5(e53)514c24c12c28c626c11100
unc-6(ev400)0121171300100
enu-3(tm4519);unc-6(ev400)529c232c2502130
unc-6(e78)55266406540100
enu-3(tm4519);unc-6(e78)0026058b371095
unc-40(e1430);unc-6(e78)00.70.7c0.7a0.7c1.42.94.30140
unc-40(e1430);unc-5(e53)00302a0200100
unc-40(e1430);enu-3(tm4519);unc-5(e53)219c116c19c01100

A puzzling result was that a second unc-6 mutant allele unc-6(e78) had more defects in outgrowth of DB4 and DB5 than the putative null strain unc-6(ev400) (Yee et al., 2011; Table 3). Those outgrowth defects were enhanced by mutations in enu-3. One potential explanation for the motor neuron axon outgrowth defects observed in unc-6(e78) is that the outgrowth defects in this mutant are dependent on the presence of functional UNC-40. Possibly, the unc-6(e78) allele can facilitate attraction toward the ventral cord mediated by UNC-40 but cannot facilitate repulsion through UNC-5 and UNC-40. If this is so, depleting UNC-40 should decrease motor neuron axon outgrowth defects. To test this prediction, we constructed a double mutant strain of unc-40(e1430);unc-6(e78) and found as predicted that there were much fewer outgrowth defects than in the unc-6(e78) background. We also found that a strain lacking both UNC-5 and UNC-40 had decreased motor axon outgrowth defects in the DB4, DB5, and DB6 motor neurons relative to a strain lacking functional UNC-5, suggesting that the motor axon outgrowth defects at least of DB4 and DB5 might depend on the presence of UNC-40 (Table 3). This observation is consistent with a report that in the absence of UNC-5, UNC-40 promotes attraction of the VD and DD classes of motor neuron axons to the ventral cord (Norris and Lundquist, 2011).

The outgrowth defects in the strain with mutations in all three proteins exceeded those found in the absence of functional UNC-40 and UNC-5, suggesting that much of the axon outgrowth activity seen in the absence of functional UNC-5 and ENU-3 is dependent on the presence of UNC-40. However, because the triple mutant has somewhat more outgrowth defects at least for DB5 and also for DB6 than unc-5(e53) (despite the lack of UNC-40) and many more than unc-5(e53);unc-40(e1430) it appears that ENU-3 has an additional axon outgrowth role that is independent of UNC-5 and UNC-40.

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. ACKNOWLEDGMENTS
  8. REFERENCES

We have shown that ENU-3 functions in a pathway parallel to UNC-40 in an UNC-6 dependent manner for outgrowth and/or guidance of the ventrally directed migrations of the processes of the AVM and PVM touch receptor neurons. RNA interference experiments supported this finding as knockdown of the level of enu-3 mRNA in two different unc-40 mutants and knockdown of unc-40 in an enu-3 mutant caused an enhancement of the defects in ventral guidance, particularly of AVM and PVM to a lesser degree. The migrations of the AVM and PVM processes are also controlled by the SLT-1 pathway in addition to UNC-6 such that when both signaling pathways are inactive, the AVM and PVM processes usually migrate straight toward the head of the animal, thus failing to make the ventrally directed migration (Hao et al., 2001). We also found that mutations in ENU-3 enhanced the AVM and PVM guidance defects of a strain with a mutation in SLT-1 but to a lesser extent than those of a strain with defects in UNC-40. Of interest, the defects caused by the mutation in ENU-3 were mild compared with those seen in the absence of functional UNC-40 and ENU-3, so the mutations in the two genes have synthetic effects, suggesting that they function redundantly.

ENU-3 and UNC-40 also appear to co-operate in adhesion or establishment of the architecture of axons to the substratum where they are migrating. The evidence is that in unc-40(e1430);enu-3(tm4519) double mutants, many of the processes had kinks and other types of either migration or attachment defects. The attachment defects were more evident in males than hermaphrodites (data not shown). DCC is known to play a role in adhesion of cells by binding to Netrin/UNC-6 present in the substratum (Shekarabi et al., 2005). unc-40(e1430) animals sometimes appear to have attachment defects but these are much more pronounced in the double mutant background, indicating that in C. elegans the two protein pathways co-operate to promote adhesion of axons to their substrata or possibly to maintain adhesion. Perhaps both UNC-40 and a receptor in the ENU-3 pathway binds to UNC-6 to promote outgrowth and/or guidance. An alternative possibility is that mutations in or reduction of ENU-3 results in a reduction of UNC-6 which would be expected to result in decreased interaction between UNC-40 and UNC-6. The findings are consistent with the idea that ENU-3 is in the UNC-6 pathway for AVM and PVM guidance rather than the SLT-1 pathway and appears to work parallel to UNC-40. We were unable to construct a strain with a second unc-40 allele unc-40(tm5504) along with enu-3(tm4519) due to synthetic lethality. This unc-40 mutant is predicted to have a deletion after amino acid 427 just before the first fibronection domain and the remainder of the mRNA is likely to be out of frame.

ENU-3 also functions in a pathway parallel to UNC-40 in guidance of the migrations of the distal tip cells and the defects were enhanced in the absence of UNC-40. A strain with a mutation in ENU-3 had very few defects in the dorsal guidance of the mesodermal cell, suggesting that it is not likely to have any significant role in the migrations of these cells (Table 2). A mutation in ENU-3 enhanced the distal tip cell defects of a strain with a mutation in UNC-40, although the overall level of defects was lower than that seen in the absence of functional UNC-6. UNC-5, the second UNC-6 receptor is instructive for guidance of the dorsal migrations of the distal tip cell. The expression of UNC-5 is co-incident with dorsal migrations of the distal tip cells and premature expression of UNC-5 leads to premature dorsal migrations of the distal tip cells (Hamelin et al., 1993; Su et al., 2000; Killeen et al., 2002). UNC-40 in contrast is expressed throughout the migration of the distal tip cell (Chan et al., 1996). The effect of ENU-3 on UNC-40 suggests that ENU-3 could function parallel to UNC-40 in the UNC-6 pathway for dorsal distal tip cell migrations, possibly by modifying UNC-5 receptor activity.

We have shown previously that a strain lacking ENU-3 and either UNC-5 or UNC-6 has enhanced axon outgrowth defects in a subset of the DA and DB motor neurons over strains lacking either UNC-5 or UNC-6 (Yee et al., 2011). In this study, we demonstrate that the failure of the DA and DB classes of motor neuron axons to exit the ventral cord in strains with mutations in UNC-5 or UNC-6 is largely dependent on the presence of UNC-40, because motor neuron axon outgrowth defects are much reduced in the absence of both UNC-5 and UNC-40 or both UNC-6 and UNC-40. Most of the motor axon outgrowth defects except for those of DB5 are also UNC-6 dependent as they disappear in unc-6(ev400), a putative null strain relative to unc-5(e53), also a putative null strain. However, the outgrowth defects of DB5 appear to be UNC-6 independent but depend on the presence of UNC-40, suggesting that these outgrowth defects may depend on another UNC-6 independent pathway.

The motor axon outgrowth defects seen in the unc-6(e78) mutant were greater than those in unc-6(ev400) or unc-5(e53) (both presumptive nulls), particularly in DB4 and DB5 (Yee et al., 2011). It was reported that the unc-6(e78) mutant has a cysteine to tyrosine alteration in the V-3 domain of UNC-6 (Lim and Wadsworth, 2002). The effect of the mutation is that there are a high percentage of DA and DB motor axon outgrowth migration defects but little effect on distal tip cell migrations or the defective egg laying phenotype characteristic of unc-6(ev400), a putative null strain. Because there are dorsal but few ventral guidance defects, it is likely that the version of UNC-6 encoded by unc-6(e78) can interact appropriately with UNC-40 but not UNC-5.

UNC-40 expression is required in motor neuron axons because in the absence of UNC-40 the axons fail to reach their final targets (Hedgecock et al., 1990). The role of UNC-40 is mainly in long-range migrations of the motor neurons (MacNeil et al., 2009). They showed that the TGFβ like molecule UNC-129 destabilizes UNC-5 homodimers and favours the appearance of UNC-40:UNC-5 hetero-dimers to steer axons to the dorsal cord. One possibility is that ENU-3 plays a similar role in destabilizing UNC-40 homo-dimers, most likely in favour of UNC-40:UNC-5 hetero-dimers. If the UNC-6 protein made in unc-6(e78) cannot bind to UNC-5, perhaps UNC-40 homo-dimers are more likely to form in this strain. Because UNC-40 is chemo-attracted by UNC-6, UNC-40 homo-dimers expressed in the motor neurons should cause the axons to fail to exit the ventral cord resulting in axon outgrowth defects. This idea is supported by our data because most of the outgrowth defects of a strain lacking UNC-5 depend on the presence of UNC-40. Similarly, it has been shown in the VD and DD classes of motor neurons that UNC-6 controls protrusion of axons through UNC-5 and UNC-40 (Norris and Lundquist, 2011). They suggest that the severe guidance defects in unc-5 mutants may be due to latent UNC-40 attractive signaling that steers the growth cone back toward the ventral source of UNC-6 by chemo-attraction. Our findings support this idea, at least for the DB4 and DB5 axons. However, the VD and the DD neurons are entirely dependent on UNC-5 for outgrowth in contrast to the DAs and DBs that usually exhibit outgrowth even in the absence of UNC-5, suggesting the existence of an additional pathway for outgrowth of these axons (Yee et al., 2011). Our data suggest that ENU-3 is likely to belong to this other pathway.

If ENU-3 functions only to prevent UNC-40 homo-dimer formation, the absence of ENU-3 in addition to UNC-5 and UNC-40 would be expected to have no effect on axon outgrowth. This is not the case as the strain lacking fully functional UNC-40, UNC-5, and ENU-3 had more defects than one lacking the two receptors UNC-40 and UNC-5 and resembled a strain lacking UNC-5. In addition, a mutation within enu-3 increased the outgrowth defects of DB5 in unc-6(ev400), and DB5 and DB6 in unc-6(e78), so ENU-3 appears to have an additional role in motor neuron axon outgrowth that may be UNC-6 independent.

The enu-3 mutant has very few defects in distal tip cell and AVM guidance unless UNC-40 is absent. However, the DA and DB motor neuron axons exit the ventral cord very well unless both ENU-3 and UNC-5 or ENU-3 and UNC-6 are compromised and the data indicate that these defects mostly depend on the presence of UNC-40. How is it possible to account for these observations with a coherent, consistent model regarding the role of ENU-3? In the case of the DA and DB motor neurons it is likely that most of the axon outgrowth defects in unc-5 and unc-6 mutant animals or as doubles with enu-3 depend on the presence of UNC-40. It is possible that part of the function of ENU-3 may be to localize UNC-40 or something in its pathway to the site of axon outgrowth. UNC-40 is known to be important in axon outgrowth from the HSN and AVM axons (Gitai et al., 2003). It was shown that UNC-40 and MIG-10/Lamellipodin are both localized to the ventral side of the HSN in the L2 stage of growth before exit of the axon and this localization is followed by selection of a single neurite in the correct location as an axon (Adler et al., 2006). They showed that UNC-40 localization was dependent on UNC-6. Therefore, the polarization of UNC-40 and MIG-10 causes a break in the symmetry of the cell. Both UNC-6 and SLT-1 are able to polarize axons in a MIG-10/RIAM/Lamellipodin dependent mechanism that is triggered by the asymmetric distribution of CED-10/Rac1 (Quinn et al., 2006, 2008). Although MIG-10 and UNC-40 are localized to the site of axon outgrowth within the HSN cell body (Adler et al., 2006; Quinn et al., 2008), it is not clear if these proteins can entirely account for all axon outgrowth. In the case of the motor neurons, both UNC-5 and UNC-40 are expressed and their locations within the cell body are not known. If UNC-40 is responsible for breaking symmetry in the motor neurons to allow axon exit, it must be located on the dorsal side away from UNC-6 sources. Therefore, perhaps ENU-3 plays some role in establishing the asymmetry of the motor neuron cell body to allow motor axon outgrowth to occur properly.

In the case of the touch receptor neurons, the true role of ENU-3 is less clear probably due to the fact that both outgrowth and guidance of the AVM and PVM processes are controlled by multiple pathways including UNC-6, SLT-1 that control ventral guidance (Hedgecock et al., 1990; Hao et al., 2001). Anterior–posterior outgrowth and guidance of AVM and PVM are controlled by Wnts and VAB-8 (Wightman et al., 1996; Hilliard and Bargmann, 2006; Levy-Strumpf and Culotti, 2007). Perhaps ENU-3 also has a bona-fide axon outgrowth activity of its own in addition to a role in UNC-40 dependent motor axon outgrowth but this would be challenging to demonstrate. In the AVMs the presence of UNC-40 allows the axon to exit in the correct place and in its absence some of the axons exit inappropriately. In the absence of both UNC-40 and ENU-3 or UNC-6 the axons are no longer attracted to UNC-6 and, therefore, exit inappropriately. ENU-3 could localize either UNC-40 or MIG-10 to the correct location in the AVMs and PVMs to allow axon outgrowth. Alternatively, ENU-3 may have some ability to cause outgrowth in the correct location but this ability must be weaker than UNC-40 because ENU-3 mutants have few AVM guidance defects. Further experiments will need to be conducted to determine the true role of ENU-3 in the touch receptor neurons.

UNC-40, UNC-5, ENU-3, and UNC-6 are unlikely to be the only proteins responsible for motor neuron outgrowth in C. elegans because the only significantly affected axons in most of our strains are DB4, DB5, and DB6, although a strain lacking both UNC-5 and ENU-3 has significant outgrowth defects in DA3, DA4 also. There is no significant defect in the outgrowth of the remainder of the DB axons or the DA axons in most of the other strains. At best, these proteins appear to account for the outgrowth of up to 58% of the DB5 axons. Other outgrowth proteins will undoubtedly be described as it is clear from this work that in the absence of the known members of the entire Netrin pathway, most of the motor neuron axons exit the ventral cord perfectly well so there must be at least one other motor axon outgrowth pathway or there must be additional molecules that prevent UNC-40 directed motor neuron axon outgrowth defects in the other DA and DB motor neurons.

EXPERIMENTAL PROCEDURES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. ACKNOWLEDGMENTS
  8. REFERENCES

Strains and Handling

Stains were handled according to standard procedures and grown at 20°C unless stated otherwise (Brenner, 1974; Hope, 1999). The strains used were N2, regarded as the wild-type strain. In addition, the following alleles were used: LGI: unc-40(e1430); unc-40(tm5504); LGII: [mec-7p::gfp; lin-15(+)], LGIII: enu-3(tm4519); LGIV: unc-5(e53); evIs82b [unc-129p::gfp;dpy-20], LGV: him-5(e1490); LGX: unc-6(ev400); unc-6(e78). Transgenes used were evIs82b (Colavita et al., 1998) and muIs32 (Ch'ng et al., 2003).

Axon Outgrowth

Motor axon outgrowth was assessed as described in Yee et al. (2011) using the evIs82b transgene which is made from the unc-129 neuronal promoter driving green fluorescent protein (Colavita et al., 1998). Briefly, animals were mounted on agarose pads in levasole and examined under epifluorescence on a compound, upright Leica DM5000B microscope under magnification of ×200, ×400, or sometimes ×630. Cell bodies without detectable axons exiting were scored as axon outgrowth defective.

The touch receptor neurons were observed using the muIs32 (mec-7p::gfp lin-15(+)] transgene (LG II) (Ch'ng et al., 2003). Animals were examined under ×200 and ×400 on the Leica DM5000B microscope. L4 animals were scored as ventral outgrowth defective if the AVM or PVM axon exiting the cell body did not reach the ventral cord directly or after a short anterior of posterior migration. Many defects were observed in unc-40(e1430);enu-3(tm4519) including positioning of the AVM and PVM cell bodies, abnormal extension of the processes particularly of the PLMs and extra branches but they were not scored.

Distal Tip Cell Defects

The migrations of the distal tip cells were scored when animals were mounted on agarose pads in levamasole under the compound, upright Leica DM5000B microscope under magnification of ×400 as previously described (Hedgecock et al., 1990). Distal tip cells were scored as defective if the migration of the third stage of the distal tip cell migration occurred along the ventral side causing a ventral clear patch extending from the vulva.

RNA Interference

Animals from various strains were grown on bacterial clones containing double stranded RNA corresponding to C. elegans cDNAs (Kamath et al., 2003). Axon guidance was assessed in strains expressing mec-7::gfp (Ch'ng et al., 2003) as described. The samples were N2;muIs32, unc-40(tm5504);muIs32, unc-40(e1430);muIs32, enu-3(tm45119);muIs32. Each experiment was performed three times and 100 animals were counted each time. The controls included a no treatment sample and a sample treated with the vector lacking an insert (pL4440). There was no significant difference between the untreated sample and the sample with the vector without an insert, so vector only data is plotted in Figure 2.

ACKNOWLEDGMENTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. ACKNOWLEDGMENTS
  8. REFERENCES

Our thanks to Dr. J. Culotti and members of his research group for strains and helpful suggestions throughout the course of this work. Thanks to Drs. Marc Perry, Naomi Levy-Strumpf, Kazuko Fujisawa and Costin Antonescu for review of the manuscript. Thanks to Dr. Andrew Laursen and Anna Farman for help with statistics. Karmen Lam, Anna Bosanac, Tiffany Lee and Anna Farman were undergraduates who provided technical help. This research project was supported by D.Gs. from NSERC for M.K. and J.F. and by FEAS Deans Research Fund award for M.K. Some strains were provided by the National the National Bioresource Project in Japan and others by CGC, which is funded by NIH Office of Research Infrastructure Programs (P40 OD010440).

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. ACKNOWLEDGMENTS
  8. REFERENCES