Cytoskeletal organelles and myelin structure of beaded nerve fibers.

Freshly removed cat peripheral nerves and lumbar spinal cord roots were prepared by freeze-substitution to study the form changes, cytoskeletal alterations and myelin structure in beaded nerve fibers. Fibers of unstretched nerves so prepared were close to cylindrical. When lightly stretched with tensions of 2-10 g before being rapidly frozen, beading appeared as a series of constrictions between the more normally expanded regions of the internodes with the paranodal regions spared. Beading also was seen in the fibers of sciatic and radial nerves fast-frozen in situ with the limbs placed in full extension to cause stretching. The cross-sectional area of the axon in the constrictions of beaded fibers was reduced by as much as 95%. The compaction of the microtubules and neurofilaments in the constructions was accounted for by the movement of axoplasmic fluid from the constrictions axially into the nearby regions where the axon and fiber diameters are close to normal. The electron-lucid area approximately 5 nm thick around the microtubules appeared to hinder their close approach in the constrictions although some microtubules touch. The neurofilaments are generally separated at a mean distance of 8-10 nm and approach to a mean distance of 4 nm in the constrictions. Neither the beading nor the reversal of beading, which occurs on relaxation from stretch, was blocked by periods of anoxia lasting several hours. Deletion of calcium from the incubating medium initiated some small amount of beading and additionally greatly augmented the beading on stretch. Beading also was present in some of the myelinated fibers of the dorsal columns of the spinal cord where stretch would not be present. These findings suggest that beading is due to a contractile process in the axon initiated by stretch and by other changed states of the fiber. Concomitantly with the contraction of the axon in the beading constrictions, the myelin sheath in that region was greatly reduced in circumference, to as much as 1/3 to 1/5 of normal. The decrease of the sheath diameter was not accompanied by a change in its thickness or in its lamellar fine structure. A repeat distance of the dense lines of 14 nm was measured in both the constricted and nonconstricted regions. To account for these findings lipid, and most likely other components of the myelin lamellar membranes, must move longitudinally from the constrictions in the plane of the lamellar membranes, and do this within 5-10 s.(ABSTRACT TRUNCATED AT 400 WORDS)