The subperineurial injection of beta,beta'-iminodipropionitrile (IDPN) into rat sciatic nerves resulted in focal disorganization of the axonal cytoskeleton characterized by segregation of neurofilaments and microtubules. Shortly after injection, microtubules clustered together to form a central channel, while neurofilaments became chaotically arrayed between the microtubule channel and axolemma. Electron microscopic autoradiography disclosed that rapidly transported organelles were preferentially associated with the microtubule-enriched central channels. These studies indicate that IDPN acts at the level of the axon to disrupt interactions between cytoskeletal elements and show that rapidly transported constituents are preferentially conveyed in association with microtubules. The model provides an opportunity to dissect the interactions of the cytoskeletal elements and other organelles.

Mice of the strain C57/BL/Ola exhibit a delay of Wallerian degeneration, such that axons survive for several weeks after a nerve transection that separates the axons from the cell bodies. In this Ola strain we have examined the distribution of cytoskeletal proteins in a 5 mm segment of the sciatic nerve for as long as 2 weeks after proximal and distal transections that prevent entry or exit of proteins via axonal transport. By 7 d after transections, there was a marked accumulation of alpha- and beta-tubulin, actin, and nonphosphorylated neurofilament epitopes at the proximal and at the distal ends of the transected axons, and loss of these proteins from the center of the isolated nerve segment. Highly phosphorylated neurofilament epitopes did not redistribute along the nerve, but there was a gradual loss of phosphorylated neurofilament immunoreactivity. These observations indicate the potential for bidirectional transport of a substantial portion of certain cytoskeletal proteins within axons of the PNS.

Accumulations of neurofilaments are observed in a variety of neurological disorders, and their pathogenesis is a fundamental problem of neuropathology. 2,5-Hexanedione (HD) neurotoxicity provides an extensively studied model of axonal neurofibrillary changes in which the pathogenetic mechanisms have been conjectural. Chronic exposure to HD results in neurofilament-filled swellings in the distal regions of large axons of exposed humans and experimental animals. In this report we describe the changes produced by a potent analogue of HD, 3,4-dimethyl-2,5-hexanedione ( DMHD ), in slow axonal transport in the rat sciatic motor axons. Young rats received 0.6 mmol/kg of DMHD for 5 days before [35S]methionine was injected into the lumbar ventral horns. Slow axonal transport of the neurofilament proteins, tubulin, and selected slow component b (SCb) proteins in DMHD -treated animals was compared to the profiles found in age-matched control animals. DMHD administration reduced the rate of transport of the neurofilament proteins 75 to 90%, while tubulin and the SCb proteins were only modestly retarded. No alterations in electrophoretic mobilities of slowly transported proteins were found, nor were any proteins accelerated in transport. These findings were systematically compared to the changes produced by administration of beta,beta'- immino - dipropionitrile (IDPN) (2.0 gm/kg, i.p.), an agent known to impair neurofilament transport. Although slightly less severe, the changes produced by DMHD were nearly identical to those of IDPN. In correlative morphological studies, the neurofilamentous changes were also comparable. The results indicate that DMHD and IDPN share the capacity to interfere selectively with neurofilament transport and thereby share pathogenetic mechanisms. DMHD provides a new agent for exploration of the organization and transport of the neuronal cytoskeleton.

Slow axonal transport is the mechanism by which cytoskeletal proteins are distributed within the axon. This function has traditionally been considered an exclusively unidirectional, anterograde process. Previous observations of cytoskeletal redistribution in surviving, transected axons of the C57BL/Ola mouse led us to hypothesize a retrograde component of cytoskeletal transport. To test this hypothesis against previous methods of measuring slow transport of cytoskeleton, we radioactively pulse-labeled proteins in sensory neurons of C57BL/Ola mice and followed their redistribution by gel fluorography in ligated and unligated sciatic nerves. Slow axonal transport of cytoskeletal proteins proceeded with the same characteristics in C57BL/Ola as in standard C57BL/6 mice. In comparison to the transport profiles from unligated control nerves, in ligated nerves there was redistribution of radiolabeled neurofilament and tubulin proteins back toward the cell body during the 14 d experimental period. These observations demonstrate that pulse-labeled cytoskeletal proteins move bidirectionally in this experimental system, and may provide insight into the normal mechanisms of cytoskeletal maintenance.