Using the unusual properties of class C acetylcholinesterase (AChE), the third AChE class described by Dennis Kolson at the last C. elegans meetings, -we devised an assay which is specific for this AChE class, as judged by the fact that the same AChE activity level is detected in wild type as in either single or double mutants affected in the previously described genes
ace-1 and
ace-2 (Johnson et al., Genetics 97, 261-279; Culotti et al., Genetics 97, 281-305). This assay was then used to screen for mutants affected in class C AChE level, among the F2 progeny of cloned F1's from EMS-mutagenized N2 parents. Of 400 such progeny clones examined (2 animals each), 2 independent isolates proved to have extremely low levels of AChE by this assay. Enzymatic assay of male cross progeny from a cross between the two new strains indicates that they harbored allelic defects producing the enzymatic deficiency. The new mutations were given the names
dc2 and
dc3, and the gene they identify was given the name
ace-3. Scoring always by enzymatic assay,
ace-3 was mapped to the right end of chromosome II, near
unc-52 and
cad-1; unlinked to either
ace-1 or
ace-2. The strain harboring
dc2 was backcrossed twice to N2, then to a strain containing
rol-1, to produce the
rol-1 utant, and finally to N2, to reisolate
ace-3 (final strain name PR1300). The strain harboring
dc3 was backcrossed twice to N2 (final strain name CD14). Extracts of the backcrossed
ace-3 strain PR1300 were examined by velocity sedimentation and ion exchange chromatography, using the conditions and the selective assay which previously identified class C AChE in N2; no evidence for any class C AChE was found, and the lower detection limit was less than 5% of the wild type level. Whether ace- 3 is the structural gene for class C AChE is uncertain; gene dosage experiments have not yet been done, and we have not yet detected any residual activity for qualitative characterization in a mutant strain. Neither of the two backcrossed
ace-3 strains PR1300 and CD14 exhibits any obvious abnormalities of either behavior or development; thus class C AChE, like class A and class B AChE, is dispensable in the presence of the other AChE classes.
ace-3; omozygotes are also behaviorally and developmentally normal, and
ace-2; omozygotes are behaviorally grossly normal but have a slightly slower growth rate than wild type; thus in the absence of class C AChE, either class A or class B AChE alone is fully or almost fully adequate for normal growth and behavior. By contrast with these marginal deficiencies, apparent
ace-2;
ace-1 triple homozygotes are much more markedly affected. These have been segregated from parents of two different genotypes, viz.
ace-2;
ace-1/+ and
ace-2;
ace-1, and are recognized by the fact that they constitute approximately one-fourth of the progeny and are both developmentally and behaviorally arrested. Such animals accumulate as eggs in which it is difficult to decide whether hatching has already occurred; the eggs are much less refractile than usual, but still have an apparent outer membrane, and give the impression of having been enzymatically digested from within without being punctured. Attempts to move these eggs invariably rupture any outer membrane that was there originally, revealing inside a coiled offspring that only a worm's mother could love. This poorest of attempts at an L1 is still bent into three parts, as required to fit into the egg shell, and remains similarly bent for several days thereafter; since it is essentially motionless when released (moving less than 1 worm diameter in 24 hours), it has also presumably been motionless in the egg, accounting for the failure of the shell to be mechanically ruptured. Surprisingly, these animals survive for several days, as judged by the production of a very occasional pharyngeal pump, but show absolutely no growth. Internally, by electron microscopy, they exhibit the expected types of cells, at least broadly speaking, but structure is much distorted by the overall bent nature of the whole animal. The difference in phenotype between these animals and
ace-2; omozygotes indicates that in the absence of AChE classes A and B, class C AChE is essential for normal development, and that the amount normally present is adequate for partial but not full coordination. In support of this conclusion, animals of the genotype
ace-2;
ace-1 (one of the two parental types from which triple homozygotes have been segregated) are themselves small, considerably less fecund than normal, and even more uncoordinated than
ace-2; omozygotes. If, as we suspect,
ace-3 is a structural gene for class C AChE, such animals should by gene dosage have only half the normal level of class C AChE, perhaps accounting for their more severe phenotype. Given the severity of the triple homozygote phenotype, we hope that with
ace-3 we have identified the last of the structural genes for C. elegans AChE.