The spacing of dense bodies and M-lines within C. elegans body wall muscle is precise. What we wish to understand is how this subcellular pattern is generated because it is this pattern that ultimately leads to the organization of sarcomeres. The position, length and interdigitation of the thick and thin filaments of a sarcomere are dependent on the position, spacing and depth of the dense bodies and the M-lines. The work of several laboratories has contributed to our understanding of the structure of these key anchoring complexes. We now know that dense bodies and M-lines are both analogs and homologs of vertebrate adhesion complexes. The major structural proteins, integrin, talin and vinculin are present in the anchoring complexes, and recently we, and our collaborators, have identified UNC-97/PINCH, UNC-112, and PAT-4/ILK as components of these structures. Analysis of mutants for each of these genes reveals the order of addition of these components to an adhesion junction. The mutants also yield a hint at the possible functional role of each protein, whether it be in nucleating a complex, spacing of the complex, or further downstream assembly. Our current model has UNC-52/Perlecan and PAT-2/alpha-integrin, PAT-3/beta-integrin as the early nucleating steps with talin and other structural components being added even before correct spacing is organized. The UNC-97/UNC-112/PAT-4 complex appears critical for correct spacing of adhesion complexes but not their initiation, since these mutants have no effect on polarization. UNC-112 and PAT-4 require each other for correct localization and function within the dense body . UNC-97/PINCH which binds to PAT-4/ILK is dependent on this protein for its localization to the dense body. While this smaller complex of UNC-97, PAT-4 and UNC-112 does not impede the addition of talin or vinculin to the dense body, spacing of integrin complexes is altered and no sarcomere with interdigitating thick and thin filaments capable of contraction is generated when any of these three molecules is missing. In all cases the absence of the protein leads to a Pat phenotype similar to what we observe for
unc-52 or
pat-3 null mutations.