Control of axonal caliber by neurofilament transport

The role of neurofilaments, the intermediate filaments of nerve cells, has been conjectural. Previous morphological studies have suggested a close relationship between neurofilament content and axonal caliber. In this study, the regenerating neuron was used as a model system for testing the hypotheses that neurofilaments are intrinsic determinants of axonal caliber, and that neurofilament content is controlled by the axonal transport of neurofilaments. This system was chosen because previous studies had shown that, after axotomy, axonal caliber was reduced within the proximal stump of the regenerating nerve and, because the relative amount of neurofilament protein undergoing axonal transport in regenerating axons was selectively reduced. The relationship between axonal caliber and neurofilament number was examined in a systematic fashion in both regenerating and control motor axons in rat L5 ventral root. Reconstruction of the spatial and temporal sequences of axonal atrophy in the proximal stump after axotomy showed that reductions in axonal caliber were first detected in the most proximal region of the root and subsequently progressed in a proximal-to-distal direction at a rate of 1.7 mm/day, which is identical to the rate of neurofilament transport in these neurons. Quantitative ultrastructural studies showed that these reductions in caliber correlated with a proportional decrease in the number of axonal neurofilaments but not microtubules. These results support the hypotheses that neurofilament content is a major intrinsic determinant of axonal caliber and that neurofilament content is controlled by the axonal transport of neurofilaments. On this basis, we suggest a role for neurofilaments in the control of axonal volume.     

β,β''-Iminodipropionitrile Impairs Retrograde Axonal Transport

beta,beta'-Iminodipropionitrile (IDPN), a neurotoxin, causes redistribution of neurofilaments in axons followed by the development of proximal axonal swellings and, in chronic intoxication, a distal decrease in axonal caliber. The latter changes are caused by a selective impairment in the slow anterograde axonal transport of neurofilament proteins. To assess the role of retrograde axonal transport in IDPN toxicity, we used [3H]N-succinimidyl propionate ([3H]NSP) to label covalently endogenous axonal proteins in sciatic nerve of the rat and measured the accumulation of radioactively labeled proteins in the cell bodies of motor and sensory neurons over time. IDPN was injected intraneurally 6 h or intraperitoneally 1 day before subepineurial injection of [3H]NSP into the sciatic nerve, and the animals were killed 1, 2, and 7 days after [3H]NSP injection. Neurotoxicity was assessed by electron microscopic observation of the nerves of similarly treated animals. Both intraneural and intraperitoneal injection of IDPN caused an acute reduction in the amount of labeled proteins transported back to the cell bodies. The early appearance of these changes suggests that alterations in retrograde transport may play a role in the production of the neuropathic changes.     

Myosin Va binding to neurofilaments is essential for correct myosin Va distribution and transport and neurofilament density

The identification of molecular motors that modulate the neuronal cytoskeleton has been elusive. Here, we show that a molecular motor protein, myosin Va, is present in high proportions in the cytoskeleton of mouse CNS and peripheral nerves. Immunoelectron microscopy, coimmunoprecipitation, and blot overlay analyses demonstrate that myosin Va in axons associates with neurofilaments, and that the NF-L subunit is its major ligand. A physiological association is indicated by observations that the level of myosin Va is reduced in axons of NF-L-null mice lacking neurofilaments and increased in mice overexpressing NF-L, but unchanged in NF-H-null mice. In vivo pulse-labeled myosin Va advances along axons at slow transport rates overlapping with those of neurofilament proteins and actin, both of which coimmunoprecipitate with myosin Va. Eliminating neurofilaments from mice selectively accelerates myosin Va translocation and redistributes myosin Va to the actin-rich subaxolemma and membranous organelles. Finally, peripheral axons of dilute-lethal mice, lacking functional myosin Va, display selectively increased neurofilament number and levels of neurofilament proteins without altering axon caliber. These results identify myosin Va as a neurofilament-associated protein, and show that this association is essential to establish the normal distribution, axonal transport, and content of myosin Va, and the proper numbers of neurofilaments in axons.     

Early Effects of β,β'-Iminodipropionitrile on Tubulin Solubility and Neurofilament Phosphorylation in the Axon

Abstract: To elucidate the role of neurofilaments in microtubule stabilization in the axon, we studied the effects of β,β'-iminodipropionitrile (IDPN) on the solubility and transport of tubulin as well as neurofilament phosphorylation in the motor fibers of the rat sciatic nerve. IDPN is known to impair the axonal transport of neurofilaments, causing accumulation of neurofilaments in the proximal axon and segregation of neurofilaments to the peripheral axoplasm throughout the nerve. Administration of IDPN at various intervals after radioactive labeling of the spinal cord with l -[³⁵S]methionine revealed that transport inhibition occurred all along the nerve within 1–2 days. Transport of cold-insoluble tubulin, which accounts for 50% of axonal tubulin, was also affected. A significant increase in the proportion of cold-soluble tubulin was observed, reaching a maximum at 3 days after IDPN treatment and returning to the control level in the following weeks. Preceding this change in tubulin solubility, a transient decrease in the phosphorylation level of the 200-kDa neurofilament protein was detected in the ventral root using phosphorylation-dependent antibodies. These early changes agreed in timing with the onset of segregation and transport inhibition, suggesting that interaction between neurofilaments and microtubules possibly regulated by phosphorylation plays a significant role in microtubule stabilization.     

Two Populations of Axonally Transported Tubulin Differentiated by Their Interactions with Neurofilaments

In the sensory fibers of the rat sciatic nerve (fibers of the dorsal root ganglion cells), two components of tubulin transport were observed that differed in the rate of transport, solubility in Triton, and subunit composition. The faster component, migrating ahead of the neurofilament proteins, was soluble in 1% Triton. The slower component, migrating with the neurofilament proteins, was insoluble in 1% Triton and contained a unique polypeptide, "NAP," in the tubulin region that was not present in the faster component. "NAP" was not a subspecies of tubulin as evidenced by peptide mapping. It seems to be a neurofilament-associated protein. When a complete separation of the main tubulin wave from the neurofilament wave was achieved in the motor axons of the same nerve (axons of the ventral motoneurons) under the effect of beta,beta'-iminodipropionitrile, a portion of tubulin was still found associated with the retarded neurofilament wave. The subunit composition of this portion was similar to the slower, neurofilament-associated component in the sensory fibers under normal conditions, i.e., enriched in "NAP" and the most acidic subtype of beta-tubulin. It is suggested that two populations of transported tubulin exist that are differentiated by the extent of their interaction with neurofilaments.     

Increasing neurofilament subunit NF-M expression reduces axonal NF-H, inhibits radial growth, and results in neurofilamentous accumulation in motor neurons

The carboxy-terminal tail domains of neurofilament subunits neurofilament NF-M and NF-H have been postulated to be responsible for the modulation of axonal caliber. To test how subunit composition affects caliber, transgenic mice were generated to increase axonal NF- M. Total neurofilament subunit content in motor and sensory axons remained essentially unchanged, but increases in NF-M were offset by proportionate decreases in both NF-H and axonal cross-sectional area. Increase in NF-M did not affect the level of phosphorylation of NF-H. This indicates that (a) in vivo NF-H and NF-M compete either for coassembly with a limiting amount of NF-L or as substrates for axonal transport, and (b) NF-H abundance is a primary determinant of axonal caliber. Despite inhibition of radial growth, increase in NF-M and reduction in axonal NF-H did not affect nearest neighbor spacing between neurofilaments, indicating that cross-bridging between nearest neighbors does not play a crucial role in radial growth. Increase in NF- M did not result in an overt phenotype or neuronal loss, although filamentous swellings in perikarya and proximal axons of motor neurons were frequently found.     

Binding of γ-Aminobutyric AcidA Receptors to Tubulin

Abstract: The rate of axonal transport of tubulin, actin, and the neurofilament proteins was measured in the peripheral and central projections of the rat L5 dorsal root ganglion (DRG). [³⁵S]Methionine was injected into the DRG, and the "front" of the radiolabeled protein was located 7, 14, and 20 days postinjection. Transport rates calculated for the neurofilament triplet proteins, tubulin, and actin in the peripheral nerve were ∼ 1.5-fold faster than those in the dorsal root. A progressive decrease in the rate of transport was observed from 7 to 20 days after radiolabeling in both the central and peripheral directions (neurofilaments, ∼ 1.7-fold; tubulin/actin, 2.1-fold). A surgical preparation, leaving the peripheral sciatic nerve with predominantly sensory fibers, was the basis for ELISAs for phosphorylation-dependent immunoreactivity of the high-molecular-weight neurofilament protein. In both dorsal roots and peripheral sensory axons the degree of phosphorylation was greater in nerve segments further away from the cell bodies. The degree of phosphorylation-related immunoreactivity correlates with the slowing of transport of radiolabeled cytoskeletal protein.     

Quantitative Analysis of Axonal Transport of Cytoskeletal Proteins in Chicken Oculomotor Nerve

We studied the axonal transport characteristics of major cytoskeletal proteins: tubulin, the 69,000 molecular weight protein of chicken neurofilaments, and actin. After intracerebral injection of [35S]methionine, we monitored the specific radioactivity of these proteins as they passed through a very short nerve segment of the chicken oculomotor nerve. Specific radioactivities were assessed by quantitative sodium dodecyl sulfate polyacrylamide gel electrophoresis and autoradiography. The transport patterns obtained for tubulin and the neurofilament protein were very similar, corresponding to transport rate ranges of 1-15 and 1-10 mm/day, respectively. A narrower velocity range of 3 to 4.3 mm/day was found for actin. Tubulin and the neurofilament protein appeared to be largely dispersed during the course of their transit along the nerve. The radioactivity associated with the proteins studied persisted in the nerve segment for a long time after the bulk of the labeled molecules had swept down. Finally, none of these proteins was observed to be transported with the fast axonal transport.     

Mutations in the neurofilament light gene linked to Charcot-Marie-Tooth disease cause defects in transport

Neurofilament light gene mutations have been linked to a subset of patients with Charcot-Marie-Tooth disease, the most common inherited motor and sensory neuropathy. We have previously shown that Charcot-Marie-Tooth-linked mutant neurofilament light assembles abnormally in non-neuronal cells. In this study, we have characterized the effects of expression of mutant neurofilament light proteins on axonal transport in a neuronal cell culture model. We demonstrated that the Charcot-Marie-Tooth-linked neurofilament light mutations: (i) affect the axonal transport of mutant neurofilaments; (ii) have a dominant-negative effect on the transport of wild-type neurofilaments; (iii) affect the transport of mitochondria and the anterograde axonal transport marker human amyloid precursor protein; (iv) result in alterations of retrograde axonal transport and (v) cause fragmentation of the Golgi apparatus. Increased neuritic degeneration was observed in neuronal cells overexpressing neurofilament light mutants. Our results suggest that these generalized axonal transport defects could be responsible for the neuropathy in Charcot-Marie-Tooth disease.     

Differential turnover of phosphate groups on neurofilament subunits in mammalian neurons in vivo

The phosphorylation and dephosphorylation of specific proteins was demonstrated directly in the intact vertebrate nervous system in vivo. By exploiting the neurons' ability to segregate a select group of cytoskeletal proteins from most other phosphorylated constituents of the cell by axoplasmic transport, we were able to examine the dynamics of phosphate turnover on neurofilament proteins in mouse retinal ganglion cell neurons simultaneously labeled with [32P]orthophosphate and [3H]proline in vivo. Three [3H]proline-labeled neurofilament protein (NFP) subunits, designated H (160-200 kDa), M (135-145 kDa), and L (68-70 kDa), entered optic axons in a mole:mole ratio similar to that of isolated axonal neurofilaments, supporting the notion that newly synthesized NFPs are transported along axons as assembled neurofilaments. NFP subunits incorporated high levels of 32P before reaching axonal sites at the level of the optic nerve. As neurofilaments were transported along axons, however, many initially incorporated [32P]phosphate groups were removed. Loss of these phosphate groups occurred to a different extent on each subunit. A minimum of 50-60 and 35-40% of the labeled phosphate groups was removed in a 5-day period from the L and M subunits, respectively. By contrast, the H subunit exhibited relatively little or no phosphate turnover during the same period. Dephosphorylation of L in axons is accompanied by a decrease in its net state of phosphorylation; changes in the phosphorylation state of H and M, however, also reflect ongoing addition of phosphates to these polypeptides during axonal transport (Nixon, R.A., Lewis, S.E., and Marotta, C.A. (1986) J. Neurosci., in press). The possibility is raised that dynamic rearrangements of phosphate topography on NFPs represent a mechanism to coordinate interactions of neurofilaments with other proteins as these elements are transported and incorporated into the stationary cytoskeleton along retinal ganglion cell axons.     

Changes in cytoskeletal protein synthesis following axon injury and during axon regeneration

Injury to the axons of facial motoneurons stimulates increases in the synthesis of actin, tubulins, and GAP-43, and decreases in the synthesis of neurofilament proteins: mRNA levels change correspondingly. In contrast to this robust response of peripheral neurons to axotomy, injured central nervous system neurons show either an attenuated response that is subsequently aborted (rubrospinal neurons) or overall decreases in cytoskeletal protein mRNA expression (corticospinal and retinal ganglion neurons). There is evidence that these changes in synthesis are regulated by a variety of factors, including loss of endoneurially or target-derived trophic factors, positive signals arising from the site of injury, changes in the intraaxonal turnover of proteins, and substitution of target-derived trophic support by factors produced by glial cells. It is concluded that there is, as yet, no coherent explanation for the upregulation or downregulation of any of the cytoskeletal proteins following axotomy or during regeneration. In considering the relevance of these changes in cytoskeletal protein synthesis to regeneration, it is emphasized that they are unlikely to be involved in the initial outgrowth of the injured axons, both because transit times between cell body and injury site are too long, and because sprouting can occur in isolated axons. Injuryinduced acceleration of the axonal transport of tubulin and actin in the proximal axon is likely to be more important in providing the cytoskeletal protein required for initial axonal outgrowth. Subsequently, the increased synthesis and transport velocity for actin and tubulin increase the delivery of these proteins to support the increased volume of the maturing regenerating axons. Reduction in neurofilament synthesis and changes in neurofilament phosphorylation may permit the increased transport velocity of the other cytoskeletal proteins. There is little direct evidence that alterations in cytoskeletal protein synthesis are necessary for successful regeneration, nor are they sufficient in the absence of a supportive environment. Nevertheless, the correlation that exists between a robust cell body response and successful regeneration suggests that an understanding of the regulation of cytoskeletal protein synthesis following axon injury must be a part of any successful strategy to improve the regenerative capacity of the central nervous system.     

The neurofilament middle molecular mass subunit carboxyl-terminal tail domains is essential for the radial growth and cytoskeletal architecture of axons but not for regulating neurofilament transport rate

The phosphorylated carboxyl-terminal "tail" domains of the neurofilament (NF) subunits, NF heavy (NF-H) and NF medium (NF-M) subunits, have been proposed to regulate axon radial growth, neurofilament spacing, and neurofilament transport rate, but direct in vivo evidence is lacking. Because deletion of the tail domain of NF-H did not alter these axonal properties (Rao, M.V., M.L. Garcia, Y. Miyazaki, T. Gotow, A. Yuan, S. Mattina, C.M. Ward, N.S. Calcutt, Y. Uchiyama, R.A. Nixon, and D.W. Cleveland. 2002. J. Cell Biol. 158:681-693), we investigated possible functions of the NF-M tail domain by constructing NF-M tail-deleted (NF-MtailDelta) mutant mice using an embryonic stem cell-mediated "gene knockin" approach that preserves normal ratios of the three neurofilament subunits. Mutant NF-MtailDelta mice exhibited severely inhibited radial growth of both motor and sensory axons. Caliber reduction was accompanied by reduced spacing between neurofilaments and loss of long cross-bridges with no change in neurofilament protein content. These observations define distinctive functions of the NF-M tail in regulating axon caliber by modulating the organization of the neurofilament network within axons. Surprisingly, the average rate of axonal transport of neurofilaments was unaltered despite these substantial effects on axon morphology. These results demonstrate that NF-M tail-mediated interactions of neurofilaments, independent of NF transport rate, are critical determinants of the size and cytoskeletal architecture of axons, and are mediated, in part, by the highly phosphorylated tail domain of NF-M.     

Abnormal neurofilament transport caused by targeted disruption of neuronal kinesin heavy chain KIF5A

To test the hypothesis that fast anterograde molecular motor proteins power the slow axonal transport of neurofilaments (NFs), we used homologous recombination to generate mice lacking the neuronal-specific conventional kinesin heavy chain, KIF5A. Because null KIF5A mutants die immediately after birth, a synapsin-promoted Cre-recombinase transgene was used to direct inactivation of KIF5A in neurons postnatally. Three fourths of such mutant mice exhibited seizures and death at around 3 wk of age; the remaining animals survived to 3 mo or longer. In young mutant animals, fast axonal transport appeared to be intact, but NF-H, as well as NF-M and NF-L, accumulated in the cell bodies of peripheral sensory neurons accompanied by a reduction in sensory axon caliber. Older animals also developed age-dependent sensory neuron degeneration, an accumulation of NF subunits in cell bodies and a reduction in axons, loss of large caliber axons, and hind limb paralysis. These data support the hypothesis that a conventional kinesin plays a role in the microtubule-dependent slow axonal transport of at least one cargo, the NF proteins.     

Axonal transport of class II and III beta-tubulin: evidence that the slow component wave represents the movement of only a small fraction of the tubulin in mature motor axons.

Pulse-labeling studies demonstrate that tubulin synthesized in the neuron cell body (soma) moves somatofugally within the axon (at a rate of several millimeters per day) as a well-defined wave corresponding to the slow component of axonal transport. A major goal of the present study was to determine what proportion of the tubulin in mature motor axons is transported in this wave. Lumbar motor neurons in 9-wk-old rats were labeled by injecting [35S]methionine into the spinal cord 2 wk after motor axons were injured (axotomized) by crushing the sciatic nerve. Immunoprecipitation with mAbs which recognize either class II or III beta-tubulin were used to analyze the distributions of radioactivity in these isotypes in intact and axotomized motor fibers 5 d after labeling. We found that both isotypes were associated with the slow component wave, and that the leading edge of this wave was enriched in the class III isotype. Axotomy resulted in significant increases in the labeling and transport rates of both isotypes. Immunohistochemical examination of peripheral nerve fibers demonstrated that nearly all of the class II and III beta-tubulin in nerve fibers is located within axons. Although the amounts of radioactivity per millimeter of nerve in class II and III beta-tubulin were significantly greater in axotomized than in control nerves (with increases of +160% and +58%, respectively), immunoassay revealed no differences in the amounts of these isotypes in axotomized and control motor fibers. We consider several explanations for this paradox; these include the possibility that the total tubulin content is relatively insensitive to changes in the amount of tubulin transported in the slow component wave because this wave represents the movement of only a small fraction of the tubulin in these motor fibers.     

Slow transport of the cytoskeleton after axonal injury.

The delivery of cytoskeletal proteins to the axon occurs by slow axonal transport. We examined how the rate of slow transport was altered after axonal injury. When retinal ganglion cell (RGC) axons regenerated through peripheral nerve grafts, an increase in the rate of slow transport occurred during regrowth of the injured axons. We compared these results to axonal injury in the optic nerve where no substantial regrowth occurs and found a completely different response. Slow transport was decreased approximately tenfold in rate in the proximal segment of crushed optic nerves. This decreased rate of slow transport was not induced immediately, but occurred about 1 week after injury. To explore whether a decrease in the rate of slow transport was induced when the regeneration of peripheral nerves was physically blocked, we examined slow transport in motor neurons after the sciatic nerve was transected and ligated. In this case, no change in the rate of the comigrating tubulin and neurofilament (NF) radioactive peaks were observed. We discuss how the changes in the rate of slow transport may reflect different neuronal responses to injury and speculate about the possible molecular changes in the expression of tubulin which may contribute to the observed changes.     

Slow transport of the cytoskeleton after axonal injury

The delivery of cytoskeletal proteins to the axon occurs by slow axonal transport. We examined how the rate of slow transport was altered after axonal injury. When retinal ganglion cell (RGC) axons regenerated through peripheral nerve grafts, an increase in the rate of slow transport occurred during regrowth of the injured axons. We compared these results to axonal injury in the optic nerve where no substantial regrowth occurs and found a completely different response. Slow transport was decreased approximately tenfold in rate in the proximal segment of crushed optic nerves. This decreased rate of slow transport was not induced immediately, but occurred about 1 week after injury. To explore whether a decrease in the rate of slow transport was induced when the regeneration of peripheral nerves was physically blocked, we examined slow transport in motor neurons after the sciatic nerve was transected and ligated. In this case, no change in the rate of the comigrating tubulin and neurofilament (NF) radioactive peaks were observed. We discuss how the changes in the rate of slow transport may reflect different neuronal responses to injury and speculate about the possible molecular changes in the expression of tubulin which may contribute to the observed changes. © 1992 John Wiley & Sons, Inc.     

Neurofilament subunit NF-H modulates axonal diameter by selectively slowing neurofilament transport

To examine the mechanism through which neurofilaments regulate the caliber of myelinated axons and to test how aberrant accumulations of neurofilaments cause motor neuron disease, mice have been constructed that express wild-type mouse NF-H up to 4.5 times the normal level. Small increases in NF-H expression lead to increased total neurofilament content and larger myelinated axons, whereas larger increases in NF-H decrease total neurofilament content and strongly inhibit radial growth. Increasing NF-H expression selectively slow neurofilament transport into and along axons, resulting in severe perikaryal accumulation of neurofilaments and proximal axonal swellings in motor neurons. Unlike the situation in transgenic mice expressing modest levels of human NF-H (Cote, F., J.F. Collard, and J.P. Julien. 1993. Cell. 73:35-46), even 4.5 times the normal level of wild-type mouse NF-H does not result in any overt phenotype or enhanced motor neuron degeneration or loss. Rather, motor neurons are extraordinarily tolerant of wild-type murine NF-H, whereas wild-type human NF-H, which differs from the mouse homolog at > 160 residue positions, mediates motor neuron disease in mice by acting as an aberrant, mutant subunit.     

Defective Neurofilament Transport in Mouse Models of Amyotrophic Lateral Sclerosis: A Review

Neurofilament proteins synthesized in the cell body of neurons are assembled and transported into axons, where they influence axon radial growth, axonal transport, and nerve conduction velocities. In diseased states, neurofilaments accumulate in cell bodies and proximal axons of affected neurons, and these lesions are characteristic of many neurodegenerative diseases, including amyotrophic lateral sclerosis (ALS), Alzheimer's disease (AD), spinal muscular atrophy (SMA), Charcot-Marie-Tooth disease type 2 (CMT2), and hereditary sensory motor neuropathy. Although the molecular mechanisms that contribute to these accumulations are not yet identified, transgenic mouse models are beginning to provide insight into the role of neurofilament transport in disease-related dysfunction of neurons. This review addresses axonal transport in mouse models of ALS and the special significance of neurofilament transport in this disease.     

Gene replacement in mice reveals that the heavily phosphorylated tail of neurofilament heavy subunit does not affect axonal caliber or the transit of cargoes in slow axonal transport

The COOH-terminal tail of mammalian neurofilament heavy subunit (NF-H), the largest neurofilament subunit, contains 44-51 lysine-serine-proline repeats that are nearly stoichiometrically phosphorylated after assembly into neurofilaments in axons. Phosphorylation of these repeats has been implicated in promotion of radial growth of axons, control of nearest neighbor distances between neurofilaments or from neurofilaments to other structural components in axons, and as a determinant of slow axonal transport. These roles have now been tested through analysis of mice in which the NF-H gene was replaced by one deleted in the NF-H tail. Loss of the NF-H tail and all of its phosphorylation sites does not affect the number of neurofilaments, alter the ratios of the three neurofilament subunits, or affect the number of microtubules in axons. Additionally, it does not reduce interfilament spacing of most neurofilaments, the speed of action potential propagation, or mature cross-sectional areas of large motor or sensory axons, although its absence slows the speed of acquisition of normal diameters. Most surprisingly, at least in optic nerve axons, loss of the NF-H tail does not affect the rate of transport of neurofilament subunits.     

Phosphorylation on carboxyl terminus domains of neurofilament proteins in retinal ganglion cell neurons in vivo: influences on regional neurofilament accumulation, interneurofilament spacing, and axon caliber

The high molecular weight subunits of neurofilaments, NF-H and NF-M, have distinctively long carboxyl-terminal domains that become highly phosphorylated after newly formed neurofilaments enter the axon. We have investigated the functions of this process in normal, unperturbed retinal ganglion cell neurons of mature mice. Using in vivo pulse labeling with [35S]methionine or [32P]orthophosphate and immunocytochemistry with monoclonal antibodies to phosphorylation- dependent neurofilament epitopes, we showed that NF-H and NF-M subunits of transported neurofilaments begin to attain a mature state of phosphorylation within a discrete, very proximal region along optic axons starting 150 microns from the eye. Ultrastructural morphometry of 1,700-2,500 optic axons at each of seven levels proximal or distal to this transition zone demonstrated a threefold expansion of axon caliber at the 150-microns level, which then remained constant distally. The numbers of neurofilaments nearly doubled between the 100- and 150- microns level and further increased a total of threefold by the 1,200- microns level. Microtubule numbers rose only 30-35%. The minimum spacing between neurofilaments also nearly doubled and the average spacing increased from 30 nm to 55 nm. These results show that carboxyl- terminal phosphorylation expands axon caliber by initiating the local accumulation of neurofilaments within axons as well as by increasing the obligatory lateral spacing between neurofilaments. Myelination, which also began at the 150-microns level, may be an important influence on these events because no local neurofilament accumulation or caliber expansion occurred along unmyelinated optic axons. These findings provide evidence that carboxyl-terminal phosphorylation triggers the radial extension of neurofilament sidearms and is a key regulatory influence on neurofilament transport and on the local formation of a stationary but dynamic axonal cytoskeletal network.