In a single fusion using the spleen of one mouse, we have raised three hybridomas producing IgG antibodies directed against the native levamisole receptor. We took 1 to 3% pure preparations of the mutant receptor of
lev-1(
x61) prepared as described in the last issue of the C. elegans Newsletter and we injected a mouse twice intraperitoneally with an estimated 5 to 10 g of receptor per injection with a four week interval in between. The first injection was in Freund's complete adjuvant and the second was just straight receptor solution. Receptor solution, initially containing 0.5 M alpha-methylammoside and 1% Triton X-100, was prepared for injection by dialysis to remove alpha-methylammoside and rapid passage over a small column of Amberlite XAD-4 to remove ~95% of the Triton X-100. Triton X-100 removal allowed us to concentrate the receptor solution ~30 fold with sodium carboxymethylcellulose and end up with a final solution for injection containing ~1 to 2% Triton X-100. Supernatants were originally tested for anti-receptor activity by first rotating 5 l of goat anti-mouse IgG derivatized Sepharose beads (anti-heavy and light chains, 1 mg/ml coupled antibody, Cappel Labs) with 50 l of hybridoma supernatant in 0.25 ml total volume for ~6 hours at room temperature (the beads also react with mouse IgM through light chains). The coated beads were spun down, washed with 1 ml of 0. 5% Triton X-100 buffer, and resuspended with 0.2 ml of a crude detergent extract of
x61 mutant receptor in 0.5% Triton X-100 (~10 femtomole of receptor). The beads were rotated overnight at 4 C with the receptor solution. The following day, the beads were spun down and the supernatant solution was assayed for loss of receptor activity to the beads. We have recently found the goat antibody on the beads is sufficiently in excess of the mouse antibody in the hybridoma supernatant that the assay can be done in one step, rotating hybridoma supernatant and receptor solution together and then assaying directly for loss of receptor binding activity to the beads. Using this screening procedure, we found that four of 392 wells produced antibody that absorbed more than 90% of the receptor to the beads. The supernatants from five other wells weakly absorbed receptor (~50 loss of activity). On subculturing, three of the four strong positives and none of the three weak positives were not subcultured). Two of the three good monoclonals cross-react with the wild-type receptor in the bead assay. The third apparently does not. Hybridoma supernatants were also screened histochemically. Nematodes were squashed between two slides, frozen and broken by cracking the slides apart (method of Albertson, Sulston and White with modifications by Culotti, Siddiqui and Francis). The slides were then dipped in methanol and then acetone at -20 C and, with appropriate washes in between, incubated first in hybridoma supernatant and then in fluorescein isothiocynate-conjugated goat anti-mouse IgM and IgG. By this histochemical method, none of the very receptor-positive hybridomas showed any noteworthy specific staining. Roughly half of the 392 wells were screened by this method and 21 wells were identified as containing antibody that recognized regionally localized antigens in the worm. Partly through the inexperience of a technician, only eight of these hybridomas survived subculturing. Among the survivors are several clones with antibody staining an antigen distributed in regularly spaced patches along the hypodermal and/or muscle cell surface, much as desmosomes are known to be. Another survivor appears to stain, amongst other things, a few spots behind the first bulb of the pharnyx, in front of the nerve ring. One of the more interesting lost clones stained only the region of the sensory endings in the worm's nose. The idea of the histochemical screen was to find what sort of other interesting proteins might be co-purifying with our receptor. A more sensitive but less convenient histochemical staining method utilized biotinylated second anti-mouse antibodies (5 g/ml) followed by avidin-peroxidase (10 g/ml, Vector Labs) and subsequent reaction ( 20 minutes at RT) with 0.5 mg/ml diaminobenzidine, 0.04% nickel chloride, 0.01% H2O2, freshly made up in 10 mM Tris-HCl, pH 7.4. With this protocol, we see marvelously patchy staining along the length of the dorsal and ventral nerve cords. No mutants have yet been tested to prove it's for real.