Caveolae preservation in the characterization of human neuromuscular disease.

We have examined freeze-fracture replicas and conventional thin-section images of rat myofibers prepared by perfusion and by conventional immersion fixation protocols, and myofibers of normal and dystrophic human myofibers prepared by similar immersion fixation methods. In both rat and human myofibers, the size and distribution of caveolae was found to differ substantially according to the method of glutaraldehyde exposure, the depth of the myofiber from the surface exposed to the fixative, and if surgically bisected, the distance from the cut end of the myofiber. Conventional immersion fixation resulted in unavoidable but predictable alterations in sarcolemmal caveolae. These reproducible artifacts of fixation technique substantially complicate the use of caveolae as reliable markers for the characterization of human neuromuscular disease.