Hippocampal gene expression is modulated by hypergravity.

We used the cDNA microarray technique to monitor simultaneously possible changes induced by hypergravity in the expression level of thousands of hippocampal genes. We tested the mRNA level of about 5000 genes in the hippocampus of mice subjected to 1.09 g (1g) or to 1.85 g (2g) for five repeated 1-h daily rotations in a centrifuge (g = 9.81 m/s2). Data were compared with those obtained for mice kept stationary (C). The ratios 1g/C and 2g/C identified genes affected by rotation and rotation + hypergravity, respectively, whereas 2g/1g ratio identified those affected by hypergravity. We found that about 200 genes were affected by rotation and/or rotation + hypergravity. Almost all the genes affected by rotation + hypergravity were up-regulated, only five being down-regulated. The modulated genes code for proteins involved in a wide range of cellular functions (DNA/RNA metabolism, protein processing, intermediate metabolism, cytoskeleton and motility, cell cycle and apoptosis, signal transduction, neuronal structure/function), suggesting that rotation + hypergravity may affect several aspects of the hippocampal function in order to compensate for environmental changes. Six genes directly or indirectly involved in synaptic transmission and plasticity (proSAAS, neuroblastoma ras oncogene, ESTs moderately similar to thymosin beta-10, syndet, inhibin beta E and Ngfi-A binding protein 2) were found to be significantly modulated by hypergravity and unaffected or only slightly affected by rotation. The modulation by hypergravity of these genes suggests that this stimulus might induce plastic remodelling of the hippocampal circuits, possibly both at structural and functional level.

Differential regulation of transcripts for dystrophin Isoforms, dystroglycan, and alpha3AChR subunit in mouse sympathetic ganglia following postganglionic nerve crush.

Previous data suggest that in mouse superior cervical ganglion (SCG) the dystrophin-dystroglycan complex may be involved in the axotomy-induced intraganglionic synapse remodeling. Here we analyzed the levels of mRNAs encoding dystrophins, dystroglycan (Dg), and the alpha3 subunit of the nicotinic acetylcholine receptor (alpha3AChR) in mouse SCG at various postaxotomy intervals. We found that axotomy downregulates the levels of transcripts for molecules related to synaptic transmission (alpha3AChR) and those presumably involved in postsynaptic apparatus organization (dystrophin isoforms) and upregulates the transcript encoding Dg, which, by binding dystrophin, bridges the actin cytoskeleton and several extracellular matrix proteins and may thus be involved in postaxotomy neuronal recovery. The observed transcriptional modulation of the components of dystrophin-dystroglycan complexes indicates their involvement in injury-induced neuronal plasticity and suggests a role in other forms of plasticity such as those required in learning and memory, functions often impaired in Duchenne muscular dystrophy patients.

Dystroglycan distribution in adult mouse brain: a light and electron microscopy study.

Dystroglycan, originally identified in muscle as a component of the dystrophin-associated glycoprotein complex, is a ubiquitously expressed cell-surface receptor that forms a transmembrane link between the extracellular matrix and the cytoskeleton. It contains two subunits, alpha and beta, formed by proteolytic cleavage of a common precursor. In the brain, different neuronal subtypes and glial cells may express dystroglycan in complex with distinct cytoplasmic proteins such as dystrophin, utrophin and their truncated forms. To examine the distribution of dystroglycan in adult mouse brain, we raised antibodies against the recombinant amino- and carboxyl-terminal domains of alpha-dystroglycan. On western blot, the antibodies recognized specifically alpha-dystroglycan in cerebellar extracts. Using light microscopy, alpha-dystroglycan was found in neurons of the cerebral cortex, hippocampus, olfactory bulb, basal ganglia, thalamus, hypothalamus, brainstem and cerebellum, where dystrophin and its truncated isoforms are also known to be present. Electron microscopy revealed that alpha-dystroglycan immunoreactivity was preferentially associated with the postsynaptic specializations. Dystroglycan immunostaining was also detected in perivascular astrocytes and in those facing the pia mater, where utrophin and dystrophin truncated isoforms are present. The cell body and endfeet of astrocytes around blood vessels and the endothelial cells at the blood-brain barrier also expressed dystroglycan. From these data, we suggest that dystroglycan, by bridging the extracellular matrix and the cytoskeleton, may play an important functional role at specialized intercellular contacts, synapses and the blood-brain barrier, whose structural and functional organization strictly depend on the integrity of the extracellular matrix-cytoskeleton linkage.

Polysialylated neural cell adhesion molecule is involved in the neuroplasticity induced by axonal injury in the avian ciliary ganglion.

We demonstrated previously in the quail ciliary ganglion, that the immunoreactivity for the neural cell adhesion molecule labeling the postsynaptic specializations of intraganglionic synapses decreases when synaptic remodeling is induced by crushing the postganglionic ciliary nerves. Here we show, in the same experimental conditions, that the immunolabeling for its polysialylated non-stabilizing isoform, which promotes cell plasticity, increases at these subcellular compartments. In control ganglia, poor immunolabeling for the polysialylated neural cell adhesion molecule was occasionally observed surrounding the soma of the ciliary neurons, in correspondence with the calyciform presynaptic ending and the perineuronal satellite cells sheath. At the electron microscope, several neuronal compartments, including some postsynaptic specializations, somatic spines and multivesicular bodies, were immunopositive. Three to six days after ciliary nerve crush, both the number of ciliary neurons labeled for the polysialylated neural cell adhesion molecule and the intensity of their immunolabeling increased markedly. Electron microscopy revealed that, in parallel to the injury-induced detachment of the preganglionic boutons, numerous postsynaptic specializations were found to be immunopositive. Twenty days later, when intraganglionic connections were re-established, polysialylated neural cell adhesion molecule immunoreactivity was comparable to that observed in control ganglia. The increase in immunolabeling also involved the other neuronal compartments mentioned above, the perineuronal satellite cells and the intercellular space between these and the ciliary neurons. From these results we suggest that the switch, at the postsynaptic specializations, between the neural cell adhesion molecule and its polysialylated form may be among the molecular changes occurring in axotomized neurons leading to injury-induced synaptic remodeling. Moreover, from the increase in polysialylated neural cell adhesion molecule immunolabeling, observed at the somatic spines and at the interface between neurons and perineuronal satellite cells, we suggest that this molecule may be involved not only in synaptic remodeling, but also in other more general aspects of injury-induced neuronal plasticity.

Selective reduction in the nicotinic acetylcholine receptor and dystroglycan at the postsynaptic apparatus of mdx mouse superior cervical ganglion.

Our previous data suggested that in mouse sympathetic superior cervical ganglion (SCG) the dystrophin-dystroglycan complex may be involved in the stabilization of the nicotinic acetylcholine receptor (nAChR) clusters. Here we used SCG of dystrophic mdx mice, which express only the shorter isoforms of dystrophin (Dys), to investigate whether the lack of the full-length dystrophin (Dp427) could affect the localization of the dystroglycan and the alpha3 nAChR subunit (alpha3AChR) at the postsynaptic apparatus. We found a selective reduction in intraganglionic postsynaptic specializations immunopositive for alpha3AChR and for alpha- and beta-dystroglycan compared with the wild-type. Moreover, in mdx mice, unlike the wild-type, the disassembly of intraganglionic synapses induced by postganglionic nerve crush occurred at the slower rate and was not preceded by the loss of immunoreactivity for Dys isoforms, beta-dystroglycan, and alpha3AChR. These data indicate that the absence of Dp427 at the intraganglionic postsynaptic apparatus of mdx mouse SCG interferes with the presence of both dystroglycan and nAChR clusters at these sites and affects the rate of synapse disassembly induced by postganglionic nerve crush. Moreover, they suggest that the decrease in ganglionic nAChR may be one of the factors responsible for autonomic imbalance described in Duchenne muscular dystrophy patients.

Disassembly of the cholinergic postsynaptic apparatus induced by axotomy in mouse sympathetic neurons: the loss of dystrophin and beta-dystroglycan immunoreactivity precedes that of the acetylcholine receptor.

In mouse sympathetic superior cervical ganglion (SCG), cortical cytoskeletal proteins such as dystrophin (Dys) and beta1sigma2 spectrin colocalize with beta-dystroglycan (beta-DG), a transmembrane dystrophin-associated protein, and the acetylcholine receptor (AChR) at the postsynaptic specialization. The function of the dystrophin-dystroglycan complex in the organization of the neuronal cholinergic postsynaptic apparatus was studied following changes in the immunoreactivity of these proteins during the disassembly and subsequent reassembly of the postsynaptic specializations induced by axotomy of the ganglionic neurons. After axotomy, a decrease in the number of intraganglionic synapses was observed (t1/2 8 h 45'), preceded by a rapid decline of postsynaptic specializations immunopositive for beta-DG, Dys, and alpha3 AChR subunit (alpha3AChR) (t1/2 3 h 45', 4 h 30' and 6 h, respectively). In contrast, the percentage of postsynaptic densities immunopositive for beta1sigma2 spectrin remained unaltered. When the axotomized neurons began to regenerate their axons, the number of intraganglionic synapses increased, as did that of postsynaptic specializations immunopositive for beta-DG, Dys, and alpha3AChR. The latter number increased more slowly than that of Dys and beta-DG. These observations suggest that in SCG neurons, the dystrophin-dystroglycan complex might play a role in the assembly-disassembly of the postsynaptic apparatus, and is probably involved in the stabilization of AChR clusters.

Dystrophin and its isoforms in a sympathetic ganglion of normal and dystrophic mdx mice: immunolocalization by electron microscopy and biochemical characterization.

In normal mouse superior cervical ganglion, dystrophin immunoreactivity is present in ganglionic neurons, satellite cells and Schwann cells. It is associated with several cytoplasmic organelles and specialized plasma membrane domains, including two types of structurally and functionally different intercellular junctions: synapses, where it is located at postsynaptic densities, and adherens junctions. Dystrophin immunostaining can be ascribed to the 427,000 mol. wt full-length dystrophin, as well as to the several dystrophin isoforms present in superior cervical ganglion, as revealed by western immunoblots. In mdx mouse superior cervical ganglion, which lacks the 427,000 mol. wt dystrophin, the unchanged pattern of dystrophin immunolabelling observed at several subcellular structures indicates the presence of dystrophin isoforms at these sites. Moreover, the absence of labelled adherens junctions indicates the presence of full-length dystrophin at this type of junction in the normal mouse superior cervical ganglion. The lower number of immunopositive postsynaptic densities in mdx mouse superior cervical ganglion than in normal mouse ganglion suggests the presence, in the latter, of postsynaptic densities with differently organized dystrophin cytoskeleton: some containing dystrophin isoforms alone or together with 427,000 mol. wt dystrophin, and others containing 427,000 mol. wt dystrophin alone.

Low environmental radiation background impairs biological defence of the yeast Saccharomyces cerevisiae to chemical radiomimetic agents.

Background radiation is likely to constitute one of the factors involved in biological evolution since radiations are able to affect biological processes. Therefore, it is possible to hypothesize that organisms are adapted to environmental background radiation and that this adaptation could increase their ability to respond to the harmful effects of ionizing radiations. In fact, adaptive responses to alkylating agents and to low doses of ionizing radiation have been found in many organisms. In order to test for effects of adaptation, cell susceptibility to treatments with high doses of radiomimetic chemical agents has been studied by growing them in a reduced environmental radiation background. The experiment has been performed by culturing yeast cells (Saccharomyces cerevisiae D7) in parallel in a standard background environment and in the underground Gran Sasso National Laboratory, with reduced environmental background radiation. After a conditioning period, yeast cells were exposed to recombinogenic doses of methyl methanesulfonate. The yeast cells grown in the Gran Sasso Laboratory showed a higher frequency of radiomimetic induced recombination as compared to those grown in the standard environment. This suggests that environmental radiation may act as a conditioning agent.

Activities of 3':5' cyclic nucleotide phosphodiesterases in the superior cervical ganglion of rat: characterization, compartmentalization and observations in young and old animals.

We investigated the presence and features of "low Km" 3'-5' cyclic nucleotide phosphodiesterase activity in the homogenates and extracts of rat superior cervical ganglion. The DEAE chromatographic elution profile of a Triton X-100 extract showed two peaks of cAMP phosphodiesterase activity eluted at 280 and 600 mM sodium acetate and two peaks of cGMP phosphodiesterase activity eluted at 300 and at 500 mM sodium acetate. The activity was poorly stimulated by calcium-calmodulin and neither stimulated or inhibited by cGMP. Both cGMP PDE peaks were inhibited by zaprinast, with IC50's of 1.4 microM and 0.28 microM: their Km values were 4.4 and 3.8 microM, respectively. These features, together with cGMP binding activity, indicate that both enzymes belong to the phosphodiesterase V family. The Km values of the first and second cAMP phosphodiesterase peaks were 1.7 and 3.8 microM. Although both peaks displayed a cAMP specific hydrolysis, only the second peak was inhibited by RO 20-1724, with an IC50 of 8 microM. Preganglionic denervation indicated that the bulk of phosphodiesterase activity is localized in ganglion cells. In order to investigate possible effects of aging on the ganglionic function, phosphodiesterase activity was assayed in the ganglia of young (3 months) and old (25 months) male Fisher rats. The chromatographic profiles and kinetic features revealed no significant differences between young and old rats.

Ultrastructural alterations induced in quail ciliary neurons by postganglionic nerve crush and by Ricinus toxin administration, separately and in combination.

The response to postganglionic nerve crush and Ricinus toxin administration by the ciliary neurons of the quail ciliary ganglion was investigated at the ultrastructural level. The toxin was either applied at the crush site on the postganglionic nerves or injected into the anterior eye chamber without any other operative intervention. Crush of postganglionic nerves without toxin administration and saline injection into the anterior eye chamber served as controls for the two toxin administration procedures. Postganglionic nerve crush caused a distinct chromatolytic reaction, accompanied by massive detachment of the preganglionic axon terminals from the ciliary neurons and loss of most of the synapses, both chemical and electrical. This process does not induce cell death and is reversible. Saline injection in the anterior eye chamber caused a moderate retrograde reaction in some of the ciliary neurons, presumably as a consequence of paracentesis. The changes consisted mainly of an increase of perikaryal neurofilaments with, at most, a minor detachment of the preganglionic boutons from a small portion of the cell body at the nuclear pole. Ricinus toxin administration induced neuronal degeneration following a pattern common to both delivery modes. The degenerative process consisted of disruption and detachment of polyribosomes from the rough endoplasmic reticulum, an increase of smooth cisterns and tubules, a dramatic increase of neurofilament bundles, compartmentalization of the cytoplasmic organelles and, finally, karyorrhexis and cell lysis. The final stages of Ricinus toxin degeneration involve a progressive accumulation of extracellular flocculo-filamentous material and cell lysis. After administration of Ricinus toxin to the crush site, ricin-affected neurons showed withdrawal of the preganglionic boutons from a portion of the ciliary neuron, especially at the nuclear pole. After Ricinus toxin injection into the anterior eye chamber, however, the bouton shell surrounding the affected ciliary neurons remained intact in the early stages of degeneration. Detachment of the preganglionic terminals and disruption of the cell junctions, therefore, is the consequence of nerve crush and not of the toxin itself. This study demonstrated that quail ciliary neurons are a suitable model for experimental neuropathology and neurotoxicology.

The maximum rate of neurofilament transport in axons: a view of molecular transport mechanisms continuously engaged.

Neurofilaments (NFs) were radiolabeled in the optic systems of mice. The leading edge of the radiolabeled NF waveform was distinguished near the injection site (the eye) both by liquid scintillation spectroscopy and visually from fluorographs. The fastest NFs were found to be translocated at rates of between 72 and 144 mm/day. It appears that the continuous (maximal) operation of the slow axonal transport machinery can move polymers intra-axonally at rates one hundred times greater than those previously reported.

Slow axonal transport mechanisms move neurofilaments relentlessly in mouse optic axons.

Pulse-labeling studies of slow axonal transport in many kinds of axons (spinal motor, sensory ganglion, oculomotor, hypoglossal, and olfactory) have led to the inference that axonal transport mechanisms move neurofilaments (NFs) unidirectionally as a single continuous kinetic population with a diversity of individual transport rates. One study in mouse optic axons (Nixon, R. A., and K. B. Logvinenko. 1986. J. Cell Biol. 102:647-659) has given rise to the different suggestion that a significant and distinct population of NFs may be entirely stationary within axons. In mouse optic axons, there are relatively few NFs and the NF proteins are more lightly labeled than other slowly transported slow component b (SCb) proteins (which, however, move faster than the NFs); thus, in mouse optic axons, the radiolabel of some of these faster-moving SCb proteins may confuse NF protein analyses that use one dimensional (1-D) SDS-PAGE, which separates proteins by size only. To test this possibility, we used a 2-mm "window" (at 3-5 mm from the posterior of the eye) to compare NF kinetics obtained by 1-D SDS-PAGE and by the higher resolution two-dimensional (2-D) isoelectric focusing/SDS-PAGE, which separates proteins both by their net charge and by their size. We found that 1-D SDS-PAGE is insufficient for definitive NF kinetics in the mouse optic system. By contrast, 2-D SDS-PAGE provides essentially pure NF kinetics, and these indicate that in the NF-poor mouse optic axons, most NFs advance as they do in other, NF-rich axons. In mice, greater than 97% of the radiolabeled NFs were distributed in a unimodal wave that moved at a continuum of rates, between 3.0 and 0.3 mm/d, and less than 0.1% of the NF population traveled at the very slowest rates of less than 0.005 mm/d. These results are inconsistent with the proposal (Nixon and Logvinenko, 1986) that 32% of the transported NFs remain within optic axons in an entirely stationary state. As has been found in other axons, the axonal transport system of mouse optic axons moves NFs and other cytoskeletal elements relentlessly from the cell body to the axon tip.

Neuronal compartments and axonal transport of synapsin I.

Studies on the transport kinetics and the posttranslational modification of synapsin I in mouse retinal ganglion cells were performed to obtain an insight into the possible factors involved in forming the structural and functional differences between the axon and its terminals. Synapsin I, a neuronal phosphoprotein associated with small synaptic vesicles and cytoskeletal elements at the presynaptic terminals, is thought to be involved in modulating neurotransmitter release. The state of phosphorylation of synapsin I in vitro regulates its interaction with both synaptic vesicles and cytoskeletal components, including microtubules and microfilaments. Here we present the first evidence that in the mouse retinal ganglion cells most synapsin I is transported down the axon, together with the cytomatrix proteins, at the same rate as the slow component b of axonal transport, and is phosphorylated at both the head and tail regions. In addition, our data suggest that, after synapsin I has reached the nerve endings, the relative proportions of variously phosphorylated synapsin I molecules change, and that these changes lead to a decrease in the overall content of phosphorus. These results are consistent with the hypothesis that, in vivo, the phosphorylation of synapsin I along the axon prevents the formation of a dense network that could impair organelle movement. On the other hand, the dephosphorylation of synapsin I at the nerve endings may regulate the clustering of small synaptic vesicles and modulate neurotransmitter release by controlling the availability of small synaptic vesicles for exocytosis.

Axonal transport kinetics and posttranslational modification of synapsin I in mouse retinal ganglion cells.

Synapsin I is a neuron-specific phosphoprotein primarily localized at the presynaptic terminals, where it is thought to play an important role in the mechanisms involved in neurotransmitter release. Its interaction with cytoskeletal proteins and with small synaptic vesicles is regulated in vitro by phosphorylation by a calcium/calmodulin-dependent kinase. Here, we present the first evidence that, in the mouse retinal ganglion cells, synapsin I, moving along the axon with the slow component of axonal transport, is phosphorylated in vivo at both the head and tail regions. In addition, our data suggest that, after synapsin I has reached the nerve endings, the relative proportion of differently phosphorylated molecules of synapsin I changes, and that these changes lead to a decrease of the overall content of phosphorus. The more basic forms, here collectively referred to as beta-forms, become predominant at the terminals after 7 d postlabeling, when the bulk of transported synapsin I has entered the superior colliculus. Along the axon, phosphorylation could be functional in preventing synapsin I from forming, with actin, a dense meshwork that would restrict organelle movement. On the other hand, at the terminals, the dephosphorylation-phosphorylation of synapsin I may regulate the clustering of small synaptic vesicles and modulate neurotransmitter release by controlling the availability of small synaptic vesicles for exocytosis.

Quantitative study of neuronal degeneration induced by Ricinus toxin and crush of postganglionic nerves in the ciliary ganglion of quail.

The effects of Ricinus toxin on the neurons of the ciliary ganglia were investigated in the quail. The neuronal death and the morphological alterations of the ganglionic cells were assessed following injection of the toxin in the anterior chamber of the eye or after application of the toxin on the postganglionic nerves at a crush site. A 45% loss of choroid neurons without loss of ciliary neurons was observed after postganglionic nerve crush alone. Injection of the toxin in the anterior chamber of the eye led to a selective loss of ciliary neurons (38%). Application of the toxin to the crushed postganglionic nerves led to a loss from both neuronal populations (40% of total neurons). This work indicates that different procedures result in selective lesion of the different neuronal populations in the ciliary ganglion.

Cytomatrix protein residence times differ significantly between the tract and the terminal segments of optic axons.

The window method of radiolabeled protein analysis was used to study the transport kinetics of axonally transported cytomatrix proteins as they move through segments of mouse optic axons. Three slow component b (SCb) proteins--actin, a 30 kDa protein, and clathrin--were radiolabeled in the eye and were followed for up to 119 days by quantitative one-dimensional gel electrophoresis. These proteins appeared first in the optic nerve, next in the tract, and last in the superior colliculus. All of the radiolabeled proteins had passed through the optic axons and had been effectively removed from the terminals by 119 days. Two different axonal segments ('windows') were examined in detail: a segment of the axon shaft region in the optic tract, and a segment of axon terminal region in the midbrain superior colliculus. The median transit times of the 3 proteins were 53-100% longer in the colliculus than in the tract, and the pulse transients (the total area under the transport curve in each window) were 180-350% larger in the colliculus than in the tract. These results indicate that at least certain cytomatrix and cytoskeletal proteins have longer residence times in the terminal regions than in the axon proper.

Slow component B protein kinetics in optic nerve and tract windows.

The transport kinetics of 3 radiolabeled slow component b (SCb) proteins (a 30 kDa protein, clathrin, and actin) were examined in the axons of mouse retinal ganglion cells. To view the transit of these proteins through the entire optic pathway between the eye and the target cells, we used two different windows: (1) a 2 mm segment from the optic nerve located 3-5 mm from the eye, and (2) a 2 mm segment from the optic tract located past the chiasm 6-8 mm from the eye. The radiolabeled proteins from these windows were separated by 1- and 2-dimensional SDS-PAGE, and the individual radiolabeled bands were quantified. Radiolabeled proteins entered and cleared the optic axons between 1 and 119 days post-labeling. All these proteins had broader transport waves in the more distal optic tract window than in the more proximal optic nerve window. The spreading of transport waves as they advance along the axon appears to be produced by a playing out of the natural heterogeneity of axonal transport rates within each population of labeled proteins. Our results confirm the proposals that clathrin and the 30 kDa protein are transported principally with SCb and that actin is transported both with SCb and with SCa. Although these proteins can be generally classified with SCb, their detailed kinetics differed (for example, their median transit times differed) and, in summary, their characteristic rates of movement can be ordered as: clathrin greater than 30 kDa protein greater than actin.(ABSTRACT TRUNCATED AT 250 WORDS)

Neurofilaments are spaced randomly in the radial dimension of axons.

The organization of the cytoskeleton is compared in the large myelinated parasympathetic and somatic motor axons of the avian oculomotor system. Electron microscopic studies demonstrate that neurofilaments are the chief structural elements in these axons, and quantitative analyses of the distribution of neurofilaments in axonal cross-sections found that the average neurofilament packing density is 25% greater in the parasympathetic axons than in the somatic motor axons. In both types of axon the distributions of neurofilaments matched a randomly generated (Poisson) distribution. In axoplasm, a Poisson distribution could arise if the neurofilaments were distributed in the cross-sectional plane by stochastic forces operating randomly and without significant neurofilament-neurofilament interactions. Thus, in these axons, the neurofilaments behave as if they are inert 'molecules' in a dilute solution-subject to non-specific stochastic forces that tend to distribute them at random. We propose that neurofilaments normally are relatively free to move apart from each other and to fill the available space within the axon.

Axonal transport of cytoskeletal proteins in oculomotor axons and their residence times in the axon terminals.

Cytoskeletal protein transport and metabolism are studied in the somatic motor and parasympathetic axons of the chicken oculomotor system. Kinetic analyses of pulse-labeled proteins indicate that the neurofilaments are transported 2-3 times more rapidly in the somatic motor axons than in the parasympathetic axons. By contrast, the transport rates of the slow component b (SCb) proteins are very similar in these axons. The parasympathetic axons terminate in the ciliary ganglion, and radiolabeling curves from the ciliary ganglion can be used to study the kinetics of cytoskeletal protein removal from the terminals. The rate of removal directly determines the residence time of the cytoskeletal proteins in the ciliary ganglion, and the residence time directly affects the shape and amplitude of the transport curves of the ganglion. A computer model was used to analyze these transport curves and to determine the half-residence time of the cytoskeletal proteins in the terminal regions. From the computer experiments, we estimate that the half-residence times of the neurofilament proteins actin and tubulin are 2, 3.5, and 7 d, respectively. The differences between the half-residence times of the cytoskeletal proteins indicate that the rate of protein removal from the terminals differs for these proteins. On the basis of these results, we propose that the removal mechanisms critically control the composition of the cytoskeleton in the terminal regions. Through their effects on its composition, the removal mechanisms have a crucial role in converting the cytoskeleton of the axon proper into the specialized cytoskeleton of the axon terminal.