Cell, Vol. 15, 405-411, October 1978, Copyrigh! @ 1978 by MIT Myosins Exist as Homodimers of Heavy Chains: Demonstration with Specific Antibody Purified by Nematode Mutant Myosin Affinity Chromatography Fred Schachat, Robert L. Garcea and Henry F. Epstein Department of Pharmacology Stanford University School of Medicine Stanford, California 94305 Summary The body-walls of Caenorhabditis elegans con- tain two different myosin heavy chains (Epstein, Waterston and Brenner, 1974) that associate to form at least two species of myosin (Schachat, Harris and Epstein, 1977a). To better define the distribution of these heavy chains in myosin mol- ecules, we have characterized the myosin of C. elegans by immunochemlcal methods. Specific, precipitating anti-myosin antibody has been pre- pared in rabbits using highly purlfled nematode myosin as the lmmunogen. The difference in reactivity of the anti-myosin antibody with wild- type myosln containing both kinds of heavy chains (designated uric-54 and non-uric-54 heavy chains on the basis of genetic specification) and myosln from the mutant El90 that lacks uric-54 heavy chains indicates that there are antigenlc differences between myosln molecules contain- ing uric-54 heavy chains and myosin molecules containing only non-uric-54 heavy chains. Anti- body specific for the uric-54 myosin determinants has been prepared by the immunoadsorption of anti-myosin antibody with El90 myosin. This spe- cific anti-uric-54 myosln antibody precipitates myosln that contains only uric-54 heavy chains. At the limits of resolution of our immunoprecipi- tation techniques, we could detect no heterodi- merit myosin molecules containing both uric-54 and non-uric-54 heavy chains. The body-wall myosins of C. elegans therefore exist only as homodimers of either class of heavy chain. This specific anti-uric-54 myosin antibody prom- ises to be a valuable tool In elucidating the role of two myoslns In body-wall muscle and In molecular characterizations of mutant myosins in C. ele- gans. We report here the use of this antibody to detect antigenic differences between uric-54 myosln from the wild-type and the muscle mutant E676. in conjunction with the original anti-myosin antibody, other studies show that both uric-54 and non-uric-54 myosins exist within the same body- wall muscle cells (Mackenzie, Schachat and Ep- stein, 1976) and that both myosins are coordi- nately synthesized during muscle development in C. elegans (Garcea, Schachat and Epstein, 1976). We discuss the implications of the self-associa- tion of uric-54 and non-uric-54 myosin heavy chains fnto homodlmeric myosins within the same body-wall muscles with respect to the assembly of thick filaments and their organization into a regular lattice. Introduction The observation that the body-wall muscle of Cae- norhabditis elegans produces two myosin heavy chain polypeptides encoded by different genes (Epstein et al., 1974; Epstein, Schachat and Wolff, 1977) led Schachat, Harris and Epstein (1977a) to propose that the presence of two myosin forms homogeneous with respect to these heavy chains may be important in the assembly and organization of thick filaments. That suggestion was based on the well-known role of the heavy chain-containing myosin rods in the formation of thick filaments (Lowey et al., 1969), the finding of uric-54 mutants in heavy chains which disrupt the filament lattice in C. elegans (Epstein et al., 1974; Harris and Epstein, 1977) and the observation that rabbit fast skeletal muscle also produces two different heavy chains (Starr and Offer, 1973). We were able to fractionate nematode myosin, which is dimeric with respect to heavy chains, on hydroxyapatite and show that in the uric-54 mutant E675 at least 85% of myosin was a homodimer with respect either to the heavy chains encoded by the uric-54 gene (in this mutant, 203,000 daltons) or a 210,000 dalton (210 kd) heavy chain. The latter myosin can be defined as a single chromatographic species present in E190, an uric-54 mutant without detectable uric-54 gene product in its myosin (Ep- stein et al., 1977; MacLeod et al., 1977b; Schachat et al., 1977a). In the studies of E675 myosin, about 15% of the myosin did not fractionate as one or the other homodimer, and evidence was presented suggesting that this material was probably an ag- gregate of homodimers rather than heterodimers. Knowledge of the number of different myosin species generated by the two different heavy chains is clearly important in determining any func- tional consequences of this multiplicity. In this study, we further characterize the two experimen- tally distinguishable myosins of C. elegans by im- munochemical techniques. We show that there are antigenic differences between the homodimeric myosins, and purify antibody specific to homodi- mers whose heavy chains are encoded by the unc- 54 gene (uric-54 myosin). Using this specific anti- uric-54 myosin antibody, we cannot detect any heterodimeric species in C. elegans. Using the same reagents, immunocompetition experiments can measure small differences in antigenicity be- tween uric-54 myosin of the wild-type N2 and the altered uric-54 myosin of the E675 mutant, in which about 5% of the heavy chain mass has been lost