The short coiled-coil domain-containing protein UNC-69 cooperates with UNC-76 to regulate axonal outgrowth and normal presynaptic organization in Caenorhabditis elegans

Abstract

Background: The nematode Caenorhabditis elegans has been used extensively to identify the genetic requirements for proper nervous system development and function. Key to this process is the direction of vesicles to the growing axons and dendrites, which is required for growth-cone extension and synapse formation in the developing neurons. The contribution and mechanism of membrane traffic in neuronal development are not fully understood, however.

Results: We show that the C. elegans gene unc-69 is required for axon outgrowth, guidance, fasciculation and normal presynaptic organization. We identify UNC-69 as an evolutionarily conserved 108-amino-acid protein with a short coiled-coil domain. UNC-69 interacts physically with UNC-76, mutations in which produce similar defects to loss of unc-69 function. In addition, a weak reduction-of-function allele, unc-69(ju69), preferentially causes mislocalization of the synaptic vesicle marker synaptobrevin. UNC-69 and UNC-76 colocalize as puncta in neuronal processes and cooperate to regulate axon extension and synapse formation. The chicken UNC-69 homolog is highly expressed in the developing central nervous system, and its inactivation by RNA interference leads to axon guidance defects.

Conclusion: We have identified a novel protein complex, composed of UNC-69 and UNC-76, which promotes axonal growth and normal presynaptic organization in C. elegans. As both proteins are conserved through evolution, we suggest that the mammalian homologs of UNC-69 and UNC-76 (SCOCO and FEZ, respectively) may function similarly.

Figure 1

The unc-69 locus encodes a 108-amino-acid protein with a short coiled-coil domain. (a) Genetic and physical maps of chromosome III in the vicinity of the unc-69 locus. unc-69 is close to and left of ced-9. Cosmids and subclones able to rescue the locomotion defect of unc-69(e587) mutants are shown in bold. B: BamHI; H: HindIII; M: MluI; P: PstI; R: EcoRI; S: SacI. Introduction of a frameshift mutation at the BamHI site in the second exon (denoted with an x) abrogated rescue by the minimal PstI-SacI rescuing fragment. Both splice variants, T07A5.6a and T07A5.6b, are contained within this fragment. (b) The UNC-69 protein sequence. The boxed region is predicted to form a coiled-coil domain. Arrows indicate the positions of the three known unc-69 mutations. Additional amino acids encoded by T07A5.6b are shown in italics (see Additional data file 1). (c) Northern-blot analysis of unc-69 revealed a single major transcript of 0.65 kb (arrow).

Figure 2

Schematic diagram of the ALM and AVM neurons in C. elegans. The different parts of the neurons are given designated letters; see Table 1 for details. Anterior is to the left.

Figure 3

UNC-69 is homologous to mammalian SCOCO. (a) Sequence alignment of UNC-69/SCOCO proteins from S. cerevisiae, C. elegans, C. briggsae, mosquito, Drosophila, Fugu, zebrafish, Xenopus, mouse and human. Residues identical in all ten sequences are shaded black; similar residues are shaded gray. The underlined region is predicted in all cases to form a coiled-coil domain. The region boxed in green is acidic, and the region boxed in red is serine/threonine-rich. The bracket indicates the carboxy-terminal basic region. Asterisks mark mutations in unc-69. (b) mRNA of the human unc-69 homolog SCOCO is enriched in fetal brain and is also present in fetal kidney, liver and lung. (c) Expression of human SCOCO rescues the locomotion defect of unc-69 mutant. Movement of the wild type (WT), mutants, and transgenic L4-stage hermaphrodites was scored as complete sine waves per minute. For each genotype n = 10. Error bars represent the standard error of the mean.

Figure 4

UNC-69::GFP is expressed in neurons. Confocal micrographs of mosaic animals expressing a rescuing carboxy-terminal UNC-69::GFP fusion. A 1 μm optical section is shown in (a) all other panels are projections of optical series. (a) Late gastrula (large arrowhead) and early comma-stage embryo (arrow) with widespread expression of UNC-69::GFP. Embryos were still inside the mother. Small arrowheads indicate the maternal VNC; v indicates the maternal vulva. (b) A two-fold-stage embryo with strong UNC-69::GFP expression in VNC neurons (between arrowheads). (c) A three-fold embryo expressing UNC-69::GFP in a growth cone (arrowhead). The arrow indicates a neuronal cell body. (d) An L1-stage larva expressing UNC-69::GFP in neurons and axons in the head (arrow), VNC (small arrowheads) and tail (large arrowhead). The asterisk indicates reporter expression in labial sensory neuronal processes of an adjoining adult animal. (e) An L3 larva expressing UNC-69::GFP in the CAN neuron (large arrow), excretory canal (small arrowheads) and in commissural axons (small arrow). (f) An L4 larva expressing UNC-69::GFP in the CAN (large arrow), HSN (large arrowhead) and ALM (small arrowhead) neurons. Small arrows indicate commissures. All scale bars represent 10 μm. In all cases, anterior is to the left and dorsal is up.

Figure 5

unc-69 is required for axonal outgrowth, guidance, branching and fasciculation in invertebrates and vertebrates. (a,b) Defect in the migration of the HSN neuron in unc-69 mutant animals. (a) In wild-type animals, the HSN axons (HSNL and HSNR) migrate ventrally until they reach the VNC, which they join and follow rostrally towards the head (arrow in (a)). (b) In unc-69 mutants, HSN axons occasionally fail to grow ventrally and instead project laterally along the body wall (arrow in (b)). Animals were stained with anti-serotonin antibodies to visualize the HSN neurons. Arrowheads indicate the vulva. Dotted lines mark the ventral margin of the body walls. (c,d) Commissures of D-type GABAergic neurons routinely reach the DNC in wild-type animals (c), but often fail in unc-69(e587) animals (d) and prematurely bifurcate (arrow). D-type GABAergic neurons were visualized with the unc-47::gfp transgene oxIs12. Asterisk in (d) marks a gap in the DNC. There are also often ectopic sprouts from the commissures (arrowheads in (d)) in unc-69(e587) mutants. (e,f) Images of the single ALM touch neuron in (e) wild-type and (f) unc-69(e602) animals. Many ectopic neurites branched out from the soma and the axonal shaft of the ALM neuron in unc-69(e602) mutant (arrowheads). (g,h) Tracings of representative electron micrographs of sections through the DNC and VNC. (g) In the wild type, the position and content of the three major fascicles are highly stereotyped (black arrows). (h) In unc-69(e587) mutants, defasciculated axons can often be found migrating separately along the body wall (open arrows). (i,j) Morphology of the bipolar AWC sensory neuron in (i) wild-type and (j) unc-69(e587) animals. Dendrites of AWC neurons in both animals reach the nose (arrows). Axonal shape is normal in wild-type worms, but abnormal in unc-69(e587) mutants, with ectopic bulges occasionally extending from the soma (arrowhead in (j)). (k,l) Expression pattern of SCOCO in stage 26 chick embryos. Sections were incubated with (k) antisense and (l) sense RNA probes for chick SCOCO. SCOCO was highly expressed in neural tissue and was most prominent in DRGs and in motoneurons of both the lateral motor column (LMC) and the medial motor column (MMC). Expression in the notochord (NC) and dermamyotome (DMT) was less pronounced. (m,n) In ovo RNAi of chick SCOCO. Embryos injected and electroporated with double-stranded RNA corresponding to (m) a yfp-containing plasmid or (n) chick SCOCO were immunostained with anti-neurofilament antibodies. (m) In control embryos, the epaxial nerves extending dorsally toward their target, the epaxial muscle, were highly fasciculated. (n) RNAi of SCOCO led to defasciculation of epaxial nerve bundles and extensive branching between muscle segments (arrows). In all panels dorsal is up. Scale bars represent: (a-j) 10 μm, (k,l) 100 μm and (m,n) 500 μm.

Figure 6

unc-69 affects axonal but not dendritic trafficking. (a,c) SNB-1::GFP is seen as evenly spaced puncta along the (a) VNC and (c) DNC in wild-type animals. (b,d,e) In unc-69(ju69) mutants, SNB-1::GFP puncta are on average bigger and often are absent from the VNC (arrowhead in (b)) and the DNC (arrowheads in (d,e)). In addition, SNB-1::GFP sometimes diffuses into the commissure (arrow in (d)). (a,b,e) Lateral views; (c,d) dorsal views of adult hermaphrodites. (f-i) As in (f,h) wild-type animals , neuronal morphology is grossly normal in (g,i) unc-69(ju69) mutants, and commissures still routinely reach the DNC. D-type GABAergic neuron morphology is visualized with the P_unc-25::gfp transgene juIs76. (f,g) Lateral views; (h,i) dorsal views. (j) Distribution of SNB-1::GFP puncta in a stretch of axon labeled with P_unc-25::DsRed monomer in the DNC in a unc-69(ju69) mutant hermaphrodite. SNB-1::GFP puncta are unevenly distributed, even though the DNC anatomy is grossly normal. (k) SNB-1::GFP is not significantly mislocalized into DD dendrites in unc-69(ju69) mutants. Animals carrying an snb-1::gfp transgene were scored at the L1 larval stage. Whereas 90% of unc-16(ju146) L1 larvae (n = 32) show dorsal GFP, 0% of wild-type L1s (n = 47) and 3% unc-69(ju69) L1s (n = 59) show dorsal GFP. Error bars represent the standard error of the mean. (l-n) The diacetyl odorant receptor ODR-10::GFP is targeted efficiently into AWB cilia both in (l) wild-type worms and in (m) unc-69(ju69) mutants. (n) In contrast, ODR-10::GFP becomes diffused in the dendritic targeting mutant unc-101. The arrow indicates the cilia; arrowheads indicate packets of ODR-10::GFP that shuttle in the dendrites. Anterior is to the left and dorsal is up.

Figure 7

UNC-69 physically interacts with UNC-76, as shown by in vitro GST pull-down assays. (a) Full-length UNC-76 (UNC-76 FL) specifically binds to full-length GST-UNC-69 but not GST-CBP. The E1A-CBP interaction was used as a positive control. (b) Serial deletions of UNC-76: a portion of the carboxy-terminal region (deleted in UNC-76 Δγ but contained within UNC-76 B3 and A3) is necessary for interaction with GST-UNC-69. (c) Point mutation L287P or a small 19-amino-acid deletion (UNC-76 Δ19), which deletes amino acids 281–299, totally abolishes the ability of UNC-76 to bind GST-UNC-69. (d) Summary of the deletion analysis, as well as the results of rescuing experiments. Gray shading indicates conserved regions. Note that UNC-76 Δ19 not only loses its binding ability but also its rescuing activity for the unc-76(e911) mutants. The 19-amino-acid region (green) lies within a conserved region and overlaps with a region we predicted to form a coiled-coil domain (purple). A previously described axonal targeting sequence [14] is in red. The positions of different unc-76 alleles are indicated.

Figure 8

UNC-69 and UNC-76 cooperate to regulate the size and position of synaptic vesicles. (a-d) Lateral view of adult hermaphrodites 52–54 h after hatching, single section. (e,f) Lateral view of the DNC of adult hermaphrodites 52–54 h after hatching, flattened images of confocal z-stack. Anterior is to the left and dorsal is up. (a,e) SNB-1::GFP is evenly distributed along the DNC in wild-type animals. (b) Removing one copy of unc-69 does not affect SNB-1::GFP distribution. (c,d,f) SNB-1::GFP becomes diffused and the puncta becomes larger (arrows) in unc-69(e587)/+; unc-76(e911)/+ and (unc-69(e587)/+; unc-76(n2457)/+ double heterozygotes. Occasionally, SNB-1::GFP is missing altogether from a stretch of the DNC (bracket in (d)). The genotypes are as follows: (a,e) juIs1 [P_unc-25::snb-1::gfp], (b) qC1/unc-69(e587); juIs1, (c) qC1/unc-69(e587); nT1[qIs51]/juIs1; nT1[qIs51]/unc-76(e911), (d,f) qC1/unc-69(e587); nT1[qIs51]/juIs1; nT1[qIs51]/unc-76(n2457). Scale bars represent 10 μm.

Figure 9

UNC-69 and UNC-76 colocalize as puncta in neuronal processes. (a-o) Functional P_unc-69::cfp::unc-69 and P_unc-76::unc-76::yfp constructs were coinjected at 5 ng/μl each into unc-76(e911) mutants, and worms rescued for locomotion were selected. Note that the unc-76(e911) mutation was removed from the background in (g-o). (d-o) are deconvoluted single-layer images. (a-c) Lateral view of an adult hermaphrodite from one line of transgenic animals with a wild-type phenotype. Both CFP::UNC-69 and UNC-76::YFP form discrete, large puncta in the DNC, as well as in the commissure (arrow). Vignette in (c) shows an enlarged image of colocalized puncta in the DNC from the rectangle. (d-f) Lateral view of an adult hermaphrodite from a second line of transgenic animals with a wild-type phenotype. Note that CFP::UNC-69 and UNC-76::YFP are both cytoplasmic and punctate, and the puncta are present in lateral and sublateral processes. (g-i) In unc-116(rh24) mutants, UNC-76::YFP puncta became diffuse in a stretch of axon in the VNC, and failed to colocalize with CFP::UNC-69 (arrows in (h,i)). (j-l) CFP::UNC-69 and UNC-76::YFP colocalize in perinuclear structures in the soma of a neuron in the tail ganglia. (m-o) In unc-116(rh24) mutants, both UNC-76::YFP and CFP::UNC-69 often appear as partially overlapping or non-overlapping aggregates in the soma of a (i) preanal and (ii-iii) two tail ganglion neurons. (p-u) Expression pattern of (p-r) opIs124 (P_unc-69::unc-69::gfp) and (s-u) opIs130 (P_unc-76::unc-76::yfp). Both transgenes were integrated into the genome to ensure stable gene expression. All pictures show the CAN neuron soma (arrowhead) and its vicinity. (p,s) In wild-type worms, the CAN neuron extended its bipolar processes along the excretory canal, and the CAN neurites were filled with UNC-69::GFP and UNC-76::YFP. Note that puncta cannot be seen in these integrants owing to overexpression of the transgenes. (q) In unc-104(e1265) mutants, UNC-69::GFP accumulated near the CAN soma as well as in its neuronal processes (asterisks), giving it a notched appearance. (t) UNC-76::YFP localization appeared to be grossly normal in unc-104(e1265) mutants. (r,u) In unc-116(rh24) mutants, both UNC-69::GFP and UNC-76::YFP accumulated in CAN neurites (asterisk). UNC-69::GFP accumulation was prominent near the CAN soma and was accompanied by ectopic branches. In contrast, UNC-76::YFP aggregated and was evenly distributed along the CAN processes. The scale bar represents 20 μm. (v,w) The CAN neuron visualized by the integrated transgene kyIs4 (P_ceh-23::ceh-23::unc-76_1–197::gfp). (v) In wild-type worms, GFP appeared as string of dots, reminiscent of endogenous UNC-76 expression pattern. (w) In unc-116(rh24) mutants, GFP dots became larger and more dispersed. (x,y) CAN neuron visualized by an extrachromosomal array opEx901 (P_unc-69::gfp). Unlike the UNC-69::GFP fusion (r), GFP itself did not accumulate significantly around CAN soma in unc-116(rh24) mutants (y), although ectopic branches were frequently observed. (p-y) are confocal z-stack images. Anterior is to the left in (a-i) and (p-y); anterior is up and ventral is to the right in (j-o). All scale bars except in (p-u) represent 10 μm.

Figure 10

UNC-69 does not interact with ARL-1, ARL-3 or ARFRP. (a) Plasmids containing LexA-unc-69 or LexA-human SCOCO were cotransformed into yeast cells with vector alone or vectors containing GAD-unc-76γ, GAD-arl-1, GAD-arl-3, or GAD-arfrp. Protein-protein interactions were measured as β-galactosidase activity by using ONPG liquid assays. UNC-69 did not interaction with any of the three ARL proteins. SCOCO did not interact with any of the three ARL proteins either (data not shown). (b) Auxotrophic growth assays for interactions between LexA-UNC-119 and GAD-ARL-3 or GAD-UNC-76γ. Cells (3 × 10⁴) were plated onto +His or -His plates, and serial tenfold dilutions of cells were then subsequently plated. Cells were grown at 30°C for 48 h before images were taken. Note that the -His plate did not contain 3-amino triazol. The strength of interaction between UNC-119 and ARL-3Δ17 was only a fifth of that between UNC-119 and ARL-3 FL, as assayed by β-galactosidase activity (data not shown).

Figure 11

UNC-69 does not colocalize with the Golgi marker mansII. P_unc-69::cfp::unc-69 and P_unc-69::mansII::yfp plasmids were coinjected at 5 ng/μl each into unc-69(e587) mutant hermaphrodites, and worms rescued for locomotion were selected for analysis. (a-c) Subcellular localization of CFP::UNC-69 and mansII::YFP in a stretch of axon in the DNC in an adult hermaphrodite animal. The scale bar represents 10 μm. (d-i) Subcellular localization of CFP::UNC-69 and mansII::YFP in the cell bodies of a tail neuron (d-f) and a neuron in the VNC (g-i) in adult hermaphrodite animals. UNC-69 and mansII usually occupied distinct subcellular regions (arrows in (c,f,i)); only occasionally did the expression patterns of the two proteins overlap (arrowheads in (c,i)). Anterior is to the left and dorsal is up in all pictures. All pictures are deconvoluted single focal plane images. (j) Mean fluorescence intensity of CFP::UNC-69 and mansII::YFP along a 17 μm distance inside the bracketed region in (c).