Aluminum alters cell viability and axonal transport system in Alzheimer's disease pathogenic mutation-bearing cells

Background: We previously reported that pulse exposure of cultured rat neurons to aluminum‐maltol resulted in an abnormal distribution of both neurofilament‐L (NF‐L) and fast axonal transported proteins. It was also found that the presenilin 1 (PS1) missense mutation and aluminum affected early neuronal development of the mouse brain. It has been reported that the presenilin (PS) mutation alters neurite outgrowth, and the axonal transport of the amyloid precursor protein (APP)/PS complex is mediated by kinesin. The present study hypothesizes that aluminum might modulate axonal transport and neurite outgrowth in APP/PS mutant‐bearing cells. Methods: We treated SH‐SY5Y cells and HEK293 cells bearing FAD mutations with aluminum‐maltol (Al‐mal) (0, 250 µm , 500 µm , 1 mm , 2 mm ) for 1 h, followed by propidium iodide (PI) and Calcein‐AM staining, live and dead assay, double staining of NF‐L/synaptophysin or APP/JNK interacting protein‐1 (JIP‐1) 3 days after treatment. Results: Apoptosis was induced Al‐mal treatment in a dose‐dependent manner in all cell lines. In SH‐SY5Y cells bearing PS1 mutations, there were no differences in the rate of cell viability, except for morphological changes observed by Calcein‐AM staining. The distribution of NF‐L and synaptophysin was modified by PS1 mutations and aluminum, suggesting that the PS mutation induces neuronal dysfunction by disturbance of the axonal transport system. The APP mutation‐bearing cell lines showed significant induction of apoptosis compared to that of wild‐type cells. Oxidative stress might thus influence cell viability in APP mutation‐bearing cells. Enhancement of JIP‐1 staining may reflect a disturbance of the intracellular axonal transport system, as well as compensation due to apoptotic changes. Conclusion: The present results, when taken together, show that aluminum alters cell viability and the axonal transport system in FAD pathogenic mutation‐bearing cells.     

Axonal transport of the cytoskeleton in regenerating motor neurons: constancy and change

We have examined slow axonal transport in regenerating motor neurons of the rat sciatic nerve. Using SDS-PAGE (sodium dodecyl sulfate-polyacrylamide gel electrophoresis) we previously found that the slow component is the vehicle for the axonal cytoskeletal proteins, i.e. the neurofilament triplet proteins, tubulin and actin. When these proteins are pulse-labeled by injecting [3H]- or [35S]-amino acids into the spinal cord, they are transported distally in the nerve as two distinguishable waves of radioactivity, SCa and SCb. In normal motor neurons, the neurofilament triplet proteins and the tubulin are transported in SCa at an average velocity of 1.7 mm/day; the less heavily labeled SCb which moves at 2-5 mm/day is the primary vehicle for actin. We now find that during regeneration the velocity of SCa is unchanged in the region of the axon between the cell body and the lesion, but the amount of labeled neurofilament triplet and associated tubulin transported in the axon is decreased in neurons which had been labeled 20 days post-lesion. In contrast, the labeling of the slowly transported proteins moving ahead of the neurofilament triplet is greater in regenerating nerves than in controls. On the basis of our findings, we propose that in motor axons the normal supply of cytoskeletal protein, which is continuously transported in the slow component, is sufficient to support regeneration. Nevertheless, the neuron cell body can alter the supply of these cytoskeletal proteins so as to enhance its regenerative capacity.     

Binding of IDPN (β,β′-iminodipropionitrile) to rat spinal cord: Possible implication in the mechanism of spheroid formation in amyotrophic lateral sclerosis

β,β′-iminodipropionitrile (IDPN) produces ‘spheroids’ similar to those in certain cases of amyotrophic lateral sclerosis (ALS). Therefore, the target molecule of IDPN could be important to the understanding of the molecular mechanism of spheroid formation in ALS. Wistar rats were injected ip with ¹⁴C-labeled IDPN (¹⁴C-IDPN) and killed at 0.5, 1, 3, 6, 12 and 24 h thereafter. The radioactivity in each organ increased rapidly and reached the maximum at 0.5–1 h after ¹⁴C-IDPN injection. Thereafter, a rapid decrease occurred until 6 h, followed by a gradual decline until 24 h. The radioactivity in the cerebral cortex, diencephalon and cerebellum was higher than in the pons, medulla oblongata and spinal cord. Although high in the visceral organs and skeletal muscles, no or little radioactivity was detected in fat tissue. Autoradiography also confirmed these results. In three rats, ¹⁴C-IDPN was injected to the lumbar enlargement of the spinal cord. Six hours after injection, the segment was removed and homogenized with physiological saline (PS). After centrifugation, the supernatant was obtained (PS fraction). The pellet was resuspended with 4 mol/L urea and the supernatant was obtained (urea fraction). Each fraction was analysed by gel filtration. A peak of radioactivity was observed at the elution fraction Nos 19 and 20 (consistent with free ¹⁴C-IDPN) when PS fraction was applied. On application of urea fraction, another peak was obtained at the elution fractions Nos 8 and 9 (MW 60～80 kDa). The present study demonstrates that ¹⁴C-IDPN does not selectively accumulate to the spinal cord and suggests that an IDPN-binding molecule with an MW of 60–80 kDa is present in the spinal cord. The molecule may be related to the pathological process of spheroid formation in ALS.     

Role of proteasomes in the formation of neurofilamentous inclusions in spinal motor neurons of aluminum‐treated rabbits

We examined the role of the 20S proteasome in pathologic changes, including abnormal aggregation of phosphorylated neurofilaments, of spinal motor nerve cells from aluminum‐treated rabbits. Immunohistochemistry for the 20S proteasome revealed that many lumbar spinal motor neurons without intracytoplasmic neurofilamentous inclusions or with small inclusions were more intensely stained in aluminum‐treated rabbits than in controls, whereas the immunoreactivity was greatly decreased in some enlarged neurons containing large neurofilamentous inclusions. Proteasome activity in whole spinal cord extracts was significantly increased in aluminum‐treated rabbits compared with controls. Furthermore, Western blot analysis indicated that the 20S proteasome degraded non‐phosphorylated high molecular weight neurofilament (neurofilament‐H) protein in vitro. These results suggest that aluminum does not inhibit 20S proteasome activity, and the 20S proteasome degrades neurofilament‐H protein. We propose that abnormal aggregation of phosphorylated neurofilaments is induced directly by aluminum, and is not induced by the proteasome inhibition in the aluminum‐treated rabbits. Proteasome activation might be involved in intracellular proteolysis, especially in the earlier stages of motor neuron degeneration in aluminum‐treated rabbits.     

Immunization does not interfere with uptake and transport by motor neurons of the binding fragment of tetanus toxin

The nontoxic binding domain of tetanus toxin (fragment C or TTC) readily undergoes retrograde axonal transport from an intramuscular injection site. This property has led to investigation of TTC as a possible vector for delivering therapeutic proteins to motor neurons. However, the vast majority of individuals in the developed world have been vaccinated with tetanus toxoid and have circulating antitetanus antibodies that cross‐react with TTC and may block the delivery of a TTC‐linked therapeutic protein. However, it is uncertain whether the immune response is capable of completely neutralizing an intramuscular depot of protein prior to its internalization by presynaptic nerve terminals, where it is inaccessible to antibody. We have evaluated uptake of rhodamine‐labeled TTC following intramuscular injection in normal animals and animals vaccinated with tetanus toxoid prior to injection of fluorescently labeled TTC. All animals demonstrated uptake of TTC, with fluorescence appropriately localized to the hypoglossal nerve and nucleus. The distribution and intensity of fluorescence within neurons and processes were indistinguishable between the two groups and were characteristic of TTC. Vaccinated animals showed levels of uptake of TTC into the brain comparable to those of immunologically naïve animals as measured by quantitative fluorimetry. All vaccinated animals had protective levels of antitetanus antibodies as measured by ELISA. Uptake of TTC by nerve terminals from an intramuscular depot is an avid and rapid process and is not blocked by vaccination associated with protection from tetanus toxin. © 2006 Wiley‐Liss, Inc.     

Ultrastructure of a new microtubule-neurofilament coupler in nerves

A new structure associated with the surfaces of neuronal microtubules is described which connects microtubules to neurofilaments in the axonal processes of cultured chick sensory ganglia. These couplers consist of a spherical core particle (15 nm in diameter) from which radiate several thin filaments (4 nm in diameter). Connection of adjacent microtubules and neurofilaments is achieved by thin filaments radiating from core particles positioned between these cytoskeletal elements. Couplers are most conspicuous in regions of axonal processes containing widely separated microtubules and neurofilaments. The structure and distribution of these couplers suggests that they are directly involved in intra-axonal organelle movements, possibly by modulating the spatial separation of adjacent microtubules and neurofilaments, thereby allowing the passage of transported organelles.     

Cyclin-dependent kinase 5 increases perikaryal neurofilament phosphorylation and inhibits neurofilament axonal transport in response to oxidative stress

Cyclin-dependent kinase 5 (cdk5) phosphorylates the high molecular weight neurofilament (NF) protein. Overexpression of cdk5 inhibits NF axonal transport and induces perikaryal accumulation of disordered phospho-NF cables. Experimental and clinical motor neuron disease is characterized by oxidative stress, increased cdk5 activity, and accumulation of phospho-NFs within perikarya or proximal axons. Because oxidative stress increases cdk5 activity in experimental motor neuron disease, we examined whether oxidative stress induced cdk5-mediated NF phosphorylation. Treatment of cultured neuronal cells with hydrogen peroxide inhibited axonal transport of green fluorescent protein-tagged NF subunits and induced perikaryal accumulation of NF phosphoepitopes normally confined to axons. These effects were prevented by treatment with the cdk5 inhibitor roscovitine or transfection with a construct expressing the endogenous cdk5 inhibitor peptide. These findings indicate that oxidative stress can compromise NF dynamics via hyperactivation of cdk5 and suggest that antioxidants may alleviate multiple aspects of neuropathology in motor neuron disease.     

Regulation of neurofilament dynamics by phosphorylation

Neurofilament (NF) phosphorylation has long been considered to regulate axonal transport rate and in doing so to provide stability to mature axons. Studies utilizing mice in which the C-terminal region of NF subunits (which contains the vast majority of phosphorylation sites) has been deleted has prompted an ongoing challenge to this hypothesis. We evaluate the collective evidence to date for and against a role for NF C-terminal phosphorylation in regulation of axonal transport and in providing structural support for axons, including some novel studies from our laboratory. We present a few suggestions for further experimentation in this area, and expand upon previous models for axonal NF dynamics. Finally, we address how C-terminal phosphorylation is regionally and temporally regulated by a balance of kinase and phosphatase activities, and how misregulation of this balance can contribute to motor neuron disease.     

Kif1Bbeta isoform is enriched in motor neurons but does not change in a mouse model of amyotrophic lateral sclerosis

The kinesin superfamily motor protein Kif1B is expressed in two isoforms, Kif1Balpha and Kif1Bbeta, with distinct cargo-binding domains. We examined the mRNA distribution of the two isoforms in adjacent sections of brain and spinal cord of adult mice using in situ hybridization analysis. Kif1Bbeta mRNA is enriched in several regions of brain and spinal cord. Its levels are four to five times higher than that of the alpha isoform, which was barely detectable. The highest mRNA levels of Kif1Bbeta were found in the cortex, hippocampus, cerebellum and the grey matter of the spinal cord. At the cellular level the highest signal was found in motor neurons in the motor nuclei of medulla oblongata and the ventral horn of spinal cord. Because expression of other Kif genes is altered in amyotrophic lateral sclerosis (ALS) models, we examined the expression level of Kif1Bbeta mRNA in the spinal cord of transgenic mice carrying the SOD1G93A mutation, a model of familial ALS, at presymptomatic and early stages of the disease. No changes were observed in Kif1Bbeta mRNA in motor neurons or in other regions of the spinal cord. These findings indicate that Kif1Balpha, which modulates the transport of mitochondria, may play a major role in tissues other than the central nervous system. Instead Kif1Bbeta, responsible for the transport of synaptic vesicle precursors, seems to play an important role in the nervous system, particularly in the lower motor neurons. The absence of changes of Kif1Bbeta mRNA in transgenic SOD1G93A mice suggests that other molecular mechanisms may play a role in the disruption of axonal transport occurring in the motor neurons of these mice.     

Alterations in slow transport kinetics induced by estramustine phosphate, an agent binding to microtubule-associated proteins.

Estramustine phosphate (EP) disassembles microtubules by binding to microtubule-associated proteins (MAPs) rather than tubulin. In this study, EP-induced alterations of MAP integrity caused a unique form of axonal atrophy in rats. Initially, EP-induced axonal atrophy occurred in both proximal and distal axons of the sciatic nerve, characterized by an increase in neurofilament packing density, associated with a decrease in axonal area. In chronic exposure, distal axonal atrophy was associated with decreased numbers of microtubules, while the neurofilament number remained unaltered for the myelin spiral length. Continued exposure caused enlargement of proximal axons associated with an increase in neurofilament content. Correlative slow transport studies done at two different times, 7 and 14 days after [35S] methionine injection showed that EP retards the transport of cytoskeletal proteins migrating with both components of slow transport (SCa and SCb). However, there was a differential effect on SCb which showed progressive slowing along the nerve while the rate of SCa stayed relatively constant. In this model, the early occurring distal axonal atrophy can best be explained by reduced cytoskeletal components, particularly those traveling in SCb. Later in the course of intoxication, a relatively constant rate of SCa permitted continuous transport of neurofilament triplets, accounting for unaltered numbers of neurofilaments in distal axons with increased packing density. This model of axonal atrophy is unique because spacing of neurofilaments, not numbers determined axon size. Furthermore, EP-induced dissociation of the SCa and SCb kinetics suggests that MAPs play a role in the orderly, cohesive migration of slow transport components, essential for the normal organization of cytoskeleton.     

Impairment of retrograde axonal transport in wobbler mouse motor neuron disease

The earliest horseradish peroxidase (HRP) neuronal labeling (the fastest retrograde transport) was determined by histochemical techniques at various intervals after intramuscular HRP injection in wobbler mice and normal littermates. In the clinically impaired forelimb system, the retrograde transport rate was 150-170 mm/day in wobbler mice and 170-230 mm/day in controls. However, there was no statistical difference between the two groups. The neuronal HRP accumulation at the early intervals was significantly less in wobbler mice than controls, suggesting that the amount of HRP transport was diminished in each axon. For the clinically intact hindlimb nerves, the rate was normal in wobbler mice, but the amount of neuronal HRP was significantly increased. Retrograde axonal transport appeared to be affected in a differential fashion, depending on the extent of disease.     

Normal dendritic arborization in spinal motoneurons requires neurofilament subunit L

Neurofilaments, composed of three polypeptide subunits, NF-L, NF-M, and NF-H, are major cytoskeletal elements in large neurons with long axons. Neurofilaments play a critical role in the development of axonal diameter; however, their role in the development of dendrites is largely unknown. By overexpressing different neurofilament subunits, we previously demonstrated that alteration of neurofilament subunit composition resulted in dramatic changes in dendritic arborization. To further determine the role of neurofilaments in dendritic growth, we examined and compared the dendritic architecture of spinal cord neurons in young NF-L knockout (-/-), heterozygous (+/-), and wild-type (+/+) mice. We show that an absence or reduction in the expression of NF-L inhibited dendritic growth most dramatically in large motoneurons, mildly in medium neurons, but had no effect on small neurons. We also reveal that a decrease in NF-L leads to an increase in NF-M and NF-H subunits in cell bodies and their reduction in dendrites. These results demonstrate that NF-L is a critical intrinsic factor for dendritic growth in large motoneurons.     

Retrograde intraaxonal transport of horseradish peroxidase by neurons in octopus

While retrograde axonal transport is the basis of a widely used neuroanatomical method, it has been rigorously demonstrated in vivo only in a few vertebrate species and not yet in an invertebrate. Evidence is presented that motor neurons of the octopus stellate ganglion are capable of retrograde intraaxonal transport of horeseradish peroxidase. This demonstration shows that retrograde transport occurs in widely divergent groups of animals, and may be a general property of neurons.     

Sexual dimorphism of pleopod motor neurons in lobster

Sexual dimorphism of the efferent neurons that innervate the pleopods in the first abdominal segment of the lobster was examined with nickel/cobalt backfilling techniques. An average of 28.8 somata (maximum of 31) were stained in the female while the mean number of male efferent somata was 25.8 (maximum of 28). These differences are statistically significant. All except one of these somata are situated close to the base of the ipsilateral first root. They are organized into large anterior (13–21 somata) and small posterior (8–13) clusters. One soma is located on the contralateral side of the ganglion.     

Segregation of viral structural proteins in cultured neurons of rat spinal ganglia and cord

Cultured spinal ganglion and spinal cord neurons were used to examine the intraneuronal distribution of five structural proteins of Sendai virus by immunohistochemistry. In spinal ganglion cells the internal, cytosolic viral proteins (the nucleocapsid, polymerase and matrix proteins) were confined to the perikarya, while the envelope glycoproteins (the haemagglutinin-neuraminidase and fusion proteins) also appeared in the axon-like processes. All five proteins occurred in the dendrite-like processes of spinal cord neurons. In both types of neuron the cytosolic viral proteins showed a pattern of distribution similar to that observed for the microtubule-associated protein MAP2. The segregated occurrence of the viral envelope and cytosolic proteins in axons may prevent virus assembly in axons and limit long-distance spread of paramyxoviruses in the nervous system.     

Calcium/calmodulin-dependent protein kinase IIbeta isoform is expressed in motor neurons during axon outgrowth and is part of slow axonal transport

Previously, we identified calcium/calmodulin-dependent protein kinase IIbeta (CaMKIIbeta) mRNA in spinal motor neurons with 372 bp inserted in what corresponds to the "association" domain of the protein. This was interesting because known additions and deletions to CaMKIIbeta mRNA are usually less than 100 bp in size and found in the "variable" region. Changes in the association domain of CaMKIIbeta could influence substrate specificity, activity or intracellular targeting. We show that three variations of this insert are found in CNS neurons or sciatic motor neurons of Sprague-Dawley rats. We used PCR and nucleic acid sequencing to identify inserts of 114, 243, or 372 bases. We also show that addition of the 372 bases is associated with outgrowth of the axon (the standard CaMKIIbeta downregulates when axon outgrowth occurs). Radiolabeling, immunoblots, and 2D PAGE identified this larger CaMKIIbeta as part of the group of soluble proteins moving at the slowest rate of axonal transport (SCa) in sciatic motor neurons (similar1 mm/day). This group is composed mainly of structural proteins (e.g., tubulin) used to assemble the cytoskeleton of regrowing axons.     

Formation of 10-nanometer filaments from the 150K-dalton neurofilament protein in vitro

In the present study we report self-assembly of individual neurofilament (NF) triplet proteins (70K, 150K, and 200K daltons) isolated by anion exchange chromatography from bovine spinal cord. Formation of smooth 10-nm filaments by both NF 150K and NF 70K is shown. Optimal conditions for NK 150K filament formation were incubation in 100 mM MES, 0.2 M NaCl, 1 mM DTT, 0.5 mM EGTA, pH 6.5, at 37 degrees C for 24 hr. Under the same assembly conditions, NF 200K formed 7-nm coiled structures. These thin filaments were similar to those formed by NF 70K and 150K under less than optimal conditions. Our results indicate that NF 150K is an integral part of the filament (self-assembly of NF 70K was previously demonstrated by others). We suggest that the optimal conditions resulting in the formation of a 10-nm 200K homopolymer remain to be determined and that the thin coiled structures formed by all three NF proteins are protofilaments that coalesce to form a double helical 10-nm filament.     

Ubiquitin and phosphorylated neurofilament epitopes in ballooned neurons of the extraocular muscle nuclei in a case of Werdnig-Hoffmann disease

Summary The extraocular muscle nuclei in one case of Werdnig-Hoffmann disease were examined immunocytochemically using antibodies against phosphorylated neurofilament (pNF) and ubiquitin (UBQ). The oculomotor and trochlear nuclei showed several chromatolytic ballooned neurons. All ballooned neurons contained epitopes of pNF and UBQ. pNF were present mainly in the periphery of the cell in a ring-like shape and were occasionally seen in the center of some cells. On the other hand, the structures stained by the antibody to UBQ were small vesicles or granules and most of them were aggregated in the center of the cell. These distribution patterns of pNF and UBQ may be unique in Werdnig-Hoffmann disease, since similar patterns were reported in other types of neurons of Werdnig-Hoffmann disease but were not seen in two other motor neuron diseases: classical amyotrophic lateral sclerosis, and familial amyotrophic lateral sclerosis with posterior column and spinocerebellar tract involvement.