We reported previously that EMS-induced mutations in two genes, srf- 2(I) and
srf-3(IV), appear to unmask antigenic determinants on the cuticle surface that are present, but hidden, in the wild-type cuticle (WBG Vol. 10, #2, p. 69). Although the mutant phenotypes detected by surface immunofluorescence represent an apparent gain in surface antigenicity, all of the mutant phenotypes (for three alleles of
srf-2 and one allele of
srf-3) are recessive. A plausible explanation for both the phenotype and the unexpected relationship between genotype and phenotype in these mutants is that components normally covering the wild-type surface are missing, thereby exposing components of internal cuticle layers. As an attempt to analyze biochemically the phenotypes defined initially by differential surface antigenicity, live wild type (N2),
srf-2(
yj262), and
srf-3(
yj10) adult worms were surface radiolabeled by Iodogen-mediated ionination with [125I]. Between 93 and 98% of the radioactivity bound to the worms could be solubilized by boiling labeled worms in SD6-2-mercaptoethanol-urea buffers. This material may therefore be considered inclusive of all molecules accessible to and reactive with [125I]. Such extracts were analyzed on 10-20% linear gradient SD6 polyacrylamide slab gels. Striking differences were found in the surface composition of the three strains at Mr below 30,000. Wild-type worm extracts showed intense bands running at Mr 19, 400 (a doublet of Mr 19,400-20,000 in same separations), 15,000, 9,700 and 5,600. In contrast, extracts of radiolabeled
srf-2(
yj262) or srf- 3
(yj10) mutants showed only single intense bands at Mr 6,200 and 6,700, respectively. The band at Mr 6,200 in
srf-2(
yj262) was less intense than either the wild-type or
srf-3(
yj10) patterns in the low Mr region, even though all lanes combined the same amount of trichloroacetic acid precipitable radioactivity. The 9,700 and 5,600 Da molecules seen in SD6-2-ME-urea extracts could also be extracted by homogenization in PES containing protease inhibitors (PBS-Pi). These molecules were immunoprecipitated from PBS- Pi extracts with the rabbit anti-cuticle antiserum used, in conjunction with adsorption experiments, to define the mutation, but were not precipitated in a negative control with normal rabbit serum. However, no bands were differentially precipitated by immune and normal sera from PBS-Pi extracts of
srf-2(
yj262) . Immunoprecipitation of PBS-Pi extracts of
srf-3(
yj10) has not yet been attempted. Thus, surface molecules present on wild type, but not mutant worms are antigenic with respect to the same parent antiserum used to define the mutation. These results are consistent with the model proposed above, and indicate that mutations likely to be point mutations can cause striking changes in nematode surface composition, with both loss and gain of components occurring simultaneously. the phenotypes here do not affect gross morphology and may be lesions specific to surface composition. Perhaps the layered structure of the cuticle provides a phylogenetic resource to nematodes. Similar phenotypic variations occurring in parasite populations might be selected differentially by host immune systems, for example.