Paper #10 - Glia promote local synaptogenesis through unc-6 (netrin) signaling in C. elegans

Title: Glia promote local synaptogenesis through unc-6 (netrin) signaling in C. elegans

Year: 2007

Summary: The underlying question in this paper relates to: How do neurons know where to form synapses? Clearly this is a question I'm into, since this is the third paper I've looked into on this question (the first post here, on a paper from the Kaplan lab, and the second one was in yesterday's post, on a paper from the Schafer lab).

In particular, this paper focuses on the neurons AIY and RIA. AIY forms glutamatergic synapses onto RIA in an extremely stereotyped manner. However, why they chose AIY and RIA for this study I don't know, and I don't know how general the findings are expected to be across C. elegans neuron pairs.

Based on three pre-synaptic markers - RAB-3 (a glutamatergic synapse marker), ELKS-1 and SYD-2 - AIY has an extremely stereotyped pattern of pre-synaptic specializiations. Using some combination of these pre-synaptic markers, they then visually screened mutants (how many and why is not discussed in the maintext) for abnormalities in pre-synaptic specializiations. They found a mutant in UNC-40 (also known as DCC, a known netrin receptor) in which AIY's pre-synaptic specializations were quite changed, though the neuron's net morphology was undisturbed. To prove this to the reader, they have beautiful confocal images of WT and unc-40 mutant animals (and here it is easy to see that the overall neuronal morphology in the two different animals is essentially identical).

The mutation in unc-40, though initially noted for its effects on AIY, also messed up RIA's synapses, but via changes in the overall neuronal morphology (the neuron no longer extends a process) rather than purely synaptic changes).

They're curious if the mutation in unc-40 mutation  affects AIY and RIA synapses by affecting AIY and RIA development directly, or if it affects the surrounding cells and AIY and RIA respond to those surrounding cells. Using a technique to obtain mosaic expression (sort of an analogous approach to using a single-cell-specific promoter) of fluorescently tagged wildtype unc-40, they find that synaptic patterning is rescued in AIY only when the wild-type unc-40 is expressed in AIY, and RIA synaptic patterning is rescued by expression of unc-40 only in RIA. They thus conclude that the effects of unc-40 are cell-autonomous (consistent also with previous results showing endogenous expression of unc-40 in AIY).

What about the UNC-40 ligand UNC-6 (netrin)? During neurulation, UNC-6 is "exclusively expressed by ventral cephalic sheath cells (VCSCs) at the nerve ring". Some more beautiful 3-color confocal images show the region where AIY and RIA make synapses (labeled via RAB-3 in AIY for glutamatergic pre-synaptic specializations, and GLR-1 in RIA) in  is sort of enfolded in a groove in a sheath cell.

If they knock out unc-6, the AIY-RIA synapses are abnormal in essentially the same way as the unc-40 mutant, and in unc-6 mutants expressing wildtype unc-6 in a mosaic, they only see rescue if unc-6 is expressed in the sheath cells.

Even a step further, they manipulate the sheath cell morphology using a mutant in an actin regulator, unc-34. In this mutant, the sheath cell extends its process further posteriorly, and the groove region in which AIY and RIA sit and synapse is actually further posterior than previously. Strikingly, AIY and RIA now form synapses much further posterior than previously. Thus the sheath cell morphology and unc-6 (netrin) expression is clearly essential for the formation of synapses.

Questions I still have:

  • why AIY and RIA? This seems like kind of an arbitrary pair.
  • why is unc-6 localized to that one part of the sheath cell? or is it? if it's everywhere, could you ectopically express it in other glia (or even other cell types) to induce synapses?
  • what controls the directionality of the synapse? (why is RIA forming post-synaptic specializations instead of pre-synaptic specializations? how does the axon/dendrite know it is supposed to be an axon/dendrite?)
  • can you directly label unc-6 (netrin)? It would be interesting to know where and how spatial gradients of netrin are located (and temporally how they behave as they dissipate and/or are degraded).