Reliable and sensitive tools for labeling specific connections between individual neurons are crucial for our understanding of the functionality of the nervous system. The gold standard for visualizing synaptic connections is electron microscopy (EM), wherein the degree of connectivity (number of sections over which en passant synapses are observed) and ultrastructure can be resolved. However, given the substantial time and resource investment, EM reconstruction is impractical for routine analysis of synaptic connections under distinct conditions (e.g. different time points, developmental stages, genetic backgrounds) and even more so for dynamic studies of neuronal connections. The genetically controlled GRASP system (“GFP Reconstitution Across Synaptic Partners”) has been used to visualize synaptic connections in live, transgenic animals (Feinberg et al., 2008). GRASP is based on two non-fluorescent split-GFP fragments, spGFP1-10 and spGFP11, tethered to synaptic membranes in each of two specific neurons or cell-types. When two neurons, each expressing one of the fragments, are tightly apposed across a synaptic cleft, fluorescent GFP is reconstituted. Actual synapses are labeled by targeting the Split GFP molecules to synaptic membranes using the neuroligin protein. C. elegans Neuroligin localizes both pre- and postsynaptically in neurons.

We have recently used GRASP to label a substantial number of synapses in C. elegans (Oren-Suissa et al., 2016)(E.A.B., M.H., M.M., A.B. and O.H., unpubl. data) and want to share a few lessons that we learned:

1) To best visualize the area of contact between the two cells of choice, the neuronal processes should be labeled with a fluorescent cytoplasmic marker. Using both cellular markers in the same color is an option as long as the area of contact is clearly visible. Optimally, promoters should differ from those driving the split GFP fragments. The cell-specificity is less of a problem and promoters for cytoplasmic labeling could be expressed in additional cells.

2) We perform germ line transformation by microinjection. In cases where the GRASP signal is expected in ventral or dorsal areas of the body (for example, the PAG), we prefer to use the pRF4 (rol-6(su1006)) plasmid as the co-injection marker. It allows for easy visualization of GRASP puncta, as the roller worm is slightly rotated when mounted on a slide for imaging purposes. We co-inject 5 plasmids, two split GFP plasmids, two cytoplasmic markers and a co-injection marker. Alternatively, the GRASP construct could be injected into an already integrated cytoplasmic marker strain, so the marker and GRASP do not need to be on the same array.

3) Initial split GFP concentrations should be minimal and aimed at 10 ng/mcL. If same promoters are used for the cytoplasmic markers, the latter should be injected in extremely low concentrations, to avoid competition for transcription resources. All the heritable F2 worms are maintained for initial live-imaging analysis. As many transgenic lines as possible should be screened to uncover ones with synaptic GFP puncta localization. Imaging should be carried out as described below. Positive transgenic lines are those with discrete GFP puncta between the two cells. If expression is too high, synapses are saturated with GFP and the split-GFP interaction is irreversible. Transgenic animals generated by microinjection carry large extrachromosomal arrays incorporating a high copy number of transgenes, which might cause overexpression and germ line silencing. In addition, the inherit­ance of extrachromosomal arrays is unstable and can create genetic mosaics. We integrate the extrachromosomal arrays into C. elegans chromosomes for stable inheritance, using γ-irradiation. In some cases integrated GRASP strains underwent silencing. Growing these strains at 25C for a couple generations disilenced these arrays.

4) Transgenics lines with good signals are often hard to find. Concentration of split-GFP plasmids should be gradually increased, either of both plasmids or of one of them. A common solution is to swap the split GFP fragment between the pre- and the post-synaptic cells. Alternatively, different neuronal promoters should be tried.

5) GRASP imaging conditions: Compared to our Zeiss Imager Z1 standard compound epifluorescence microscope, we have observed much better signals on a pinhole based confocal microscope (LSM880). Epi-fluorescent microscope images contain too much out-of-focus signals. Puncta were quantified by scanning the original full Z-stack for distinct dots in the area where the processes of the two neurons overlap.