[
Worm Breeder's Gazette,
1985]
When viewed in thin section, gap juctions are problematic in C. elegans neurons because of their small size. Gap junctions in other cell types are often more extensive, but otherwise look pretty much the same using this technique. The freeze fracture technique offers a different view of gap junctions and we are able to show some differences in their organization when comparing hypodermis, gut and neurons. Similar results were obtained previously in Planaria by Quick and Johnson (1977). One can speculate that the gap junction molecules in different tissues may derive from different genes, a question which C. elegans is particularly well suited to explore. There is molecular evidence that mammalian gjs derive from tissue-specific genes. For instance, 2- D peptide maps of liver and lens gjs show almost no overlap (Hertzberg, et. al., 1982) and comparison of partial sequence data for these two molecules shows major differences (Nicholson, et. al., 1983). Six of these gj protein subunits combine to form a 'hemichannel' which spans one plasma membrane. Two hemichannels must link end-to-end to form a patent channel between two neighboring cells. These channels aggregate into arrays within two apposed plasma membranes to form a typical gap junction. Adult nematodes are fixed in glutaraldehyde, cryoprotected in glycerol and then sandwiched as a monolayer between two gold discs. After rapid freezing in liquid Freon, the gold disc sandwich is placed in a double replica holder and fractured in a Balzer s Freeze/Etch Device. If etching is desired, one waits several minutes to allow water to sublimate from the fractured surfaces before shadowing. The specimen is shadowed with Pt and then coated with carbon to make a Pt/C replica. The replica is cleaned with bleach and mounted on a copper grid for examination by electron microscopy. The fracture plane preferentially travels along membranes, splitting the unit membrane into two opposing halves (the P- and E- faces). Most animals fracture lengthwise, often superficially, to reveal details of the cuticle and hypodermal membranes. Deeper fractures can reveal many recognizable tissues: nerve ring, nerve cords, commissures, gut, muscle, etc. Muscle membranes have been difficult to identify; however, using the freeze etch technique, we have been able to view more deeply into cytoplasm for additional structural detail. Hypodermal membranes facing the cuticle are particularly easy to recognize because of their characteristic regions of parallel infoldings. Gap junctions are all of the A-type , with most of their intramembrane particles adhering to the P-face. An array of pits on the E-face corresponds to the P-face particle array. Each pit presumably represents a hemichannel pulling out of the E-face membrane. Some gap junctions are densely packed and macular, including those found between gut cells. Other gjs have more dispersed groups of particles, particularly those between hypodermal cells. The fraction of particles adhering to the E-face varies in a similar pattern to that noted in Planaria: in gut cells, 6%; in hypodermis, 11%; in neurons, ~25%. Our sample of neuronal gap junctions is very limited so far. No muscle gap junctions have yet been identified, although work with the freeze etch technique shows promise. Gap junction antibodies derived from two mammalian tissues (liver, lens) are now available for immunocytochemistry. It will be interesting to see whether these antibodies recognize C. elegans gap junctions: They might differentiate the gjs in various tissues.