The self-interaction of myosin.

The first event in thick filament formation must be the interaction of one myosin monomer with another to give a dimer. The energetics of the parallel apposition of the rod portion of myosin were first considered by McLachlan and Karn (Nature 299: 226-231, 1982; J. Mol. Biol. 164: 605-626, 1983). They applied a simple 'point scoring' algorithm to the periodic charge distribution of the myosin rod and suggested that there are peaks in interaction energy when the stagger between parallel rods is close to 14.3 and 43nm. We have modelled the assembly process on a more detailed basis in an attempt to understand the factors governing the structure of the thick filament. We have taken into account both the hydrodynamic and electrostatic properties of myosin. It is likely that the final state of parallel apposition of the monomers is reached via intermediate states. We identify as a likely intermediate a structure in which two monomers are bound at a single point contact to form an extended dimer, and have computed the likelihoods of formation and the stabilities of different forms of this structure. We have also computed the energetics of the pathways leading to the more stable parallel dimer. Our results suggest that selectivity for a 14.3 nm axial stagger is inherent in the pathway for dimerisation but that it is a consequence of the kinetics rather than the energetics of the assembly process. We have identified rat cardiac myosin at minimal ionic strength as a system in which later steps in the assembly process are blocked; small myosin oligomers become stable structures rather than transient intermediates and so can be trapped and characterised. The structures found are in both qualitative and quantitative agreement with the computational predictions.