Multiple interactions of aluminum with neurofilament subunits: regulation by phosphate-dependent interactions between C-terminal extensions of the high and middle molecular weight subunits

Neuronal intermediate filaments and neurodegenerative disorders

Intermediate filaments represent the most abundant cytoskeletal element in mature neurons. Mutations and/or accumulations of neuronal intermediate filament proteins are frequently observed in several human neurodegenerative disorders. Although it is now admitted that disorganization of the neurofilament network may be directly involved in neurodegeneration, certain type of perikaryal intermediate filament aggregates confer protection in motor neuron disease. The use of various mouse models provided a better knowledge of the role played by the disorganization of intermediate filaments in the pathogenesis of neurodegenerative disorders, but the mechanisms leading to the formation of these aggregates remain elusive. Here, we will review some neurodegenerative diseases involving intermediate filaments abnormalities and possible mechanisms susceptible to provoke them.

Mechanical and structural properties of in vitro neurofilament hydrogels

Neurofilaments belong to the class of cytoskeletal intermediate filaments and are the predominant structural elements in axons. They are composed of a semiflexible backbone and highly charged anionic sidearms protruding from the surface of the filaments. Here, the rheology of in-vitro networks of neurofilaments purified from pig spinal cord was determined. The mechanical properties of these networks are qualitatively similar to other hydrogels of semiflexible polymers. The low-deformation storage modulus G'(omega) showed a concentration (c) dependence of G' approximately c (1.3) that is consistent with a model for semiflexible networks, but was also observed for polyelectrolyte brushes. A terminal relaxation was not observed in the frequency range investigated (0.007-5 Hz), supporting the notion that sidearms act as cross-links hindering slip between filaments on a time scale of many minutes. The mesh size distribution of the network was measured by analysis of Brownian motion of embedded beads. The concentration dependence of the mesh size follows the same power law behaviour as found for F-actin networks, but shows a significantly wider distribution attributable to the smaller persistence length of neurofilaments. The attractive interaction between filaments is increased by addition of Al(3+) ions resulting in a reduction of the linear response regime from strains bigger than 80% to less than 30%.

Hypophosphorylated neurofilament subunits undergo axonal transport more rapidly than more extensively phosphorylated subunits in situ

Axonal transport of neurofilaments (NFs) has long been considered to be regulated by phosphorylation. We present evidence that in optic axons of normal mice, the rate of NF axonal transport is inversely correlated with the NF phosphorylation state. In addition to 200 kDa NF-H and 145 kDa NF-M, axonal cytoskeletons from CNS contained a range of phospho-variants of NF-H migrating between 160-200 kDa, and of NF-M migrating at 97-145 kDa. While 160 kDa phospho-variants of NF-H have been well characterized, we confirmed the identity of the previously-described 97 kDa species as a hypophospho-variant of NF-M since (1) pulse-chase metabolic labeling confirmed the 97 kDa species to be a new synthesis product that was converted by phosphorylation over time into a form migrating at 145 kDa, (2) the 97 kDa protein reacted with multiple NF-M antibodies, including one specific for hypophosphorylated NF-M, and (3) dephosphorylation converted NF-M isoforms to 97 kDa. Autoradiographic analyses following metabolic radiolabeling demonstrated that hypophosphorylated NF-H and NF-M isoforms underwent substantially more rapid transport in situ than did extensively phosphorylated isoforms, while NF-H subunits bearing a developmentally delayed C-terminal phospho-epitope transported at a rate slower than that of total 200 kDa NF-H. Differential transport of phospho-variants also highlights that these variants are not homogeneously distributed among NFs, but are segregated to some extent among distinct, although probably overlapping, NF populations, indicating that axonal NFs are not homogeneous with respect to phosphorylation state.

Phospho-dependent association of neurofilament proteins with kinesin in situ

Recent studies demonstrate co-localization of kinesin with neurofilament (NF) subunits in culture and suggest that kinesin participates in NF subunit distribution. We sought to determine whether kinesin was also associated with NF subunits in situ. Axonal transport of NF subunits in mouse optic nerve was perturbed by the microtubule (MT)-depolymerizing drug vinblastine, indicating that NF transport was dependent upon MT dynamics. Kinesin co-precipitated during immunoprecipitation of NF subunits from optic nerve. The association of NFs and kinesin was regulated by NF phosphorylation, since (1) NF subunits bearing developmentally delayed phospho-epitopes did not co-purify in a microtubule motor preparation from CNS while less phosphorylated forms did; (2) subunits bearing these phospho-epitopes were selectively not co-precipitated with kinesin; and (3) phosphorylation under cell-free conditions diminished the association of NF subunits with kinesin. The nature and extent of this association was further examined by intravitreal injection of (35)S-methionine and monitoring NF subunit transport along optic axons. As previously described by several laboratories, the wave of NF subunits underwent a progressive broadening during continued transport. The front, but not the trail, of this broadening wave of NF subunits was co-precipitated with kinesin, indicating that (1) the fastest-moving NFs were associated with kinesin, and (2) that dissociation from kinesin may foster trailing of NF subunits during continued transport. These data suggest that kinesin participates in NF axonal transport either by directly translocating NFs and/or by linking NFs to transporting MTs. Both Triton-soluble as well as cytoskeleton-associated NF subunits were co-precipitated with kinesin; these data are considered in terms of the form(s) in which NF subunits undergo axonal transport.

C-terminal phosphorylation of the high molecular weight neurofilament subunit correlates with decreased neurofilament axonal transport velocity

We probed the relationship of NF axonal transport of neurofilaments (NFs) to their phosphorylation state by comparing these parameters in two closely-aged groups of young adult mice - 2 and 5 months of age. This particular time interval was selected since prior studies demonstrate that optic axons have already completed axonal caliber expansion and attained adult NF levels by 2 months but, as shown herein, continue to increase NF-H C-terminal phosphorylation. NF axonal transport was monitored by autoradiographic analysis of the distribution of radiolabeled subunits immunoprecipitated from optic axon segments at intervals following intravitreal injection of 35S-methionine. Both the peak and front of radiolabeled NFs translocated faster in 2- vs. 5-month-old mice. This developmental decline in NF transport rate was not due to reduced incorporation of NFs into the cytoskeleton, nor to an overall decline in slow axonal transport. By excluding or minimizing other factors, these findings support previous conclusions that C-terminal NF phosphorylation regulates NF axonal transport.

Regulation of neurofilament interactions in vitro by natural and synthetic polypeptides sharing Lys-Ser-Pro sequences with the heavy neurofilament subunit NF-H: neurofilament crossbridging by antiparallel sidearm overlapping

Neurofilaments are organised into parallel bundles in axons through crossbridges formed by lateral projections of neurofilament subunits. Pure neurofilaments form gels in vitro, consisting of interconnected parallel arrays of filaments regulated by the phosphorylation level of neurofilament subunits. Neurofilament-associated polypeptides sharing phosphorylated epitopes with the repetitive lysine-serine-proline (Lys-Ser-Pro) motifs of the neurofilament heavy subunit sidearm are characterised: they regulate in vitro the neurofilament gelation kinetics in a concentration- and phosphorylation-dependent manner. Studies with synthetic peptides show that interactions between neurofilaments involve both acid and base amino acid residues of neurofilament sidearms and demonstrate the opposite effects of peptides containing either one (inhibition) or two (activation) Lys-Ser-Pro motifs. Electron microscopy reveals an organised network of native neurofilament sidearms, regulated by the phosphorylation level of neurofilament subunits, suggesting a structural transition between intra- and inter-neurofilament sidearm interactions. These results favour the hypothesis of a mechanism of neurofilament crossbridging through the variable antiparallel overlapping of the phosphorylable Lys-Ser-Pro domains of neurofilament sidearms from adjacent filaments, following an equilibrium regulated by neurofilament-associated proteins, bivalent cations and the phosphorylation level of Lys-Ser-Pro motifs from both neurofilament sidearms and neurofilament-associated proteins.

Aluminum inhibits neurofilament assembly, cytoskeletal incorporation, and axonal transport. Dynamic nature of aluminum-induced perikaryal neurofilament accumulations as revealed by subunit turnover

The mechanism by which aluminum induces formation of perikaryal neurofilament (NF) inclusions remains unclear. Aluminum treatment inhibits: 1. The incorporation of newly synthesized NF subunits into Triton-insoluble cytoskeleton of axonal neurites; 2. Their degradation and dephosphorylation; 3. Their translocation into axonal neurites. It also fosters the accumulation of phosphorylated NFs within perikarya. In the present study, we addressed the relationship among these effects. Aluminum reduced the assembly of newly synthesized NF subunits into NFs. During examination of those subunits that did assemble in the presence of aluminum, it was revealed that aluminum also interfered with transport of newly assembled NFs into axonal neurites. Similarly, a delay in axonal transport of microinjected biotinylated NF-H was observed in aluminum-treated cells. Aluminum also inhibited the incorporation of newly synthesized and microinjected subunits into the Triton-insoluble cytoskeleton within both perikarya and neurites. Once incorporated into Triton-insoluble cytoskeletons, however, biotinylated subunits were retained within perikarya of aluminum-treated cells to a greater extent than within untreated cells. Notably, these subunits were depleted in the presence and absence of aluminum within 48 h, despite the persistence of the aluminum-induced perikaryal accumulation itself, suggesting that individual NF subunits undergo turnover even within aluminum-induced perikaryal accumulations. These findings demonstrate that aluminum interferes with multiple aspects of neurofilament dynamics and furthermore leaves open the possibility that aluminum-induced perikaryal NF whorls may not represent permanent structures, but rather may require continued recruitment of cytoskeletal constituents.

Spinal motor neuron neuroaxonal spheroids in chronic aluminum neurotoxicity contain phosphatase-resistant high molecular weight neurofilament (NFH)

It has previously been shown that a single intracisternal inoculum of AlCl3 in young adult New Zealand white rabbits will induce a dose-dependent phosphatase resistance of high molecular weight neurofilament protein (NFH) that is proportionate to the extent of neurofilamentous inclusion formation (Strong and Jakowec, 1994). To determine if the potential for dissolution of aluminum-induced neurofilamentous inclusions was dependent on the degree of NFH phosphatase resistance, we have examined NFH phosphatase sensitivity in a reversible chronic model of aluminum neurotoxicity. Rabbits receiving repeated intracisternal inoculums of 100 microgram AlCl3 at 28 day intervals until day 267 develop spinal motor neuron perikaryal and neuroaxonal neurofilamentous aggregates in a stereotypic, dose-dependent fashion. In the rabbits receiving inoculums until day 156 with survival until day 267 without further aluminum exposure, neuroaxonal spheroids remained prominent while perikaryal inclusions largely resolved. Immunoreactivity to a monoclonal antibody recognizing phosphorylated NFH (SMI 31) was abolished in perikaryal aggregates at each time interval by dephosphorylation with bovine alkaline phosphatase. However, neuroaxonal spheroids maintained their immunoreactivity. Using time-course dephosphorylation studies of spinal cord homogenates, we observed a significant reduction in the rate of dephosphorylation of NFH following 267 days of AlCl3 exposure (P < 0.05). These observations suggest that neuroaxonal spheroids contain phosphatase-resistant NFH isoforms and that the potential for resolution of intraneuronal neurofilamentous inclusions correlates with the susceptibility of NF within these inclusions to enzymatic dephosphorylation.

Inhibition of proteolysis enhances aluminum-induced perikaryal neurofilament accumulation but does not enhance tau accumulation

As observed for neurons in situ, phosphorylated neurofilament (NF) epitopes are normally segregated within the axonal cytoskeleton of NB2a/d1 cells. However, accumulations of phosphorylated NFs develop in NB2a/d1 perikarya following exposure to aluminum salts and following inhibition of proteolysis. In the present study, we observed that perikarya of cells exposed to both aluminum and the protease inhibitor C1 (also known as "AllNal") were more intensely labeled by monoclonal antibodies directed against both nonphosphorylated and phosphorylated epitopes than were cells treated with either aluminum or protease inhibitor alone. Since these monoclonal antibodies crossreact with tau, we also immunostained cells treated under these conditions with monoclonal antibodies directed against phosphate-insensitive (5E2) and phosphorylated (PHF-1) epitopes of tau. Aluminum treatment, but not C1 treatment, induced accumulation of total tau isoforms as judged by an increase in 5E2 immunoreactivity. Neither treatment, either separately or in combination, induced an increase in PHF-1 immunoreactivity. These findings suggest that alterations in immunoreactivity with SMI antibodies reflected increases in NF epitopes. This was confirmed by immunoblot analyses. Since proteolysis is apparently instrumental in maintaining the normal distribution patterns of phosphorylated NF epitopes, these findings implicate deficiencies in proteolytic mechanisms in the development of neurofibrillary pathology, and underscore the possibility of a multiple etiology in human neuropathological conditions.

Aluminum treatment of intact neuroblastoma cells alters neurofilament subunit phosphorylation, solubility, and proteolysis

Addition of 400 microM AlCl3 to the culture medium for 72 h has been previously shown to induce perikaryal whorls of intermediate-sized filaments in intact mouse NB2a/d1 neuroblastoma cells. Immunoblot analyses demonstrated that in vivo treatment of cells with aluminum induced the de novo appearance of extensively phosphorylated NF-H isoforms in cytoskeletons of undifferentiated cells and increased levels of these isoforms in differentiated cells. Neurofilament subunits isolated from intact cells treated with aluminum were resistant to dephosphorylation in vitro by alkaline phosphatase and to in vitro degradation by endogenous calcium-dependent protease(s). These alterations were accompanied by a greater tendency of neurofilaments to form insoluble aggregates after isolation. These findings demonstrate direct effects of aluminum on neurofilament subunits within intact neuronal cells similar to those previously demonstrated following in vitro exposure of isolated neurofilaments to aluminum.

Calcium modulates aluminum neurotoxicity and interaction with neurofilaments

We examined the influence of calcium on neurotoxicity of AlCl3 and Al-lactate toward differentiated NB2a/d1 cells. Apart from induction of perikaryal neurofibrillary inclusions, AlCl3 at 1 mM induced no obvious additional signs of toxicity when added to culture medium in the presence of the normal medium CaCl2 content of 1.8 mM, nor when extracellular calcium was decreased by the addition to the medium of 0.9 mM EDTA. Increasing the extracellular CaCl2 concentration by fivefold was only marginally toxic, but in the presence of AlCl3, reduced viable cell numbers by well over 50% as compared to control cultures, and by approximately 50% vs fivefold CaCl2 alone. A twofold increase in extracellular CaCl2 did not increase the percentage of cells exhibiting Bielschowsky-positive perikarya, but induced a near doubling in the percentage of cells exhibiting accumulations in the presence of 1 mM Al-lactate. AlCl3 (1 mM) retards the electrophoretic migration of NF subunits on SDS-gels. This effect was eliminated by withholding CaCl2 from the incubation mixture and including 5 mM EDTA during incubation of cytoskeletons with AlCl3. The presence of CaCl2 alone did not alter NF migration. These findings underscore the possibility that multiple factors, including those that compromise general neuronal homeostasis, may contribute to neurofibrillary pathology.

Relative susceptibility of cytoskeleton-associated and soluble neurofilament subunits to aluminum exposure in intact cells. A possible mechanism for reduction of neurofilament axonal transport during aluminum neurotoxicity

Previous studies have demonstrated the appearance of phosphorylated neurofilament (NF) subunits within perikaryal cytoskeletons following aluminum exposure. In order to examine the mechanisms leading to this altered distribution of NF subunits, we carried out biochemical analyses of NF subunits in Triton-insoluble and -soluble fractions derived from aluminum-treated NB2a/d1 cells. In addition to increases in the Triton-insoluble cytoskeleton, increases in all three NF subunits were also detected within the Triton-soluble fraction of aluminum-treated cells. To address the nature of this increase in Triton-soluble subunits, aluminum-treated and untreated cultures were harvested in the absence of Triton and fractionated by established procedures to yield fractions greatly enriched for perikarya and neurites, respectively. Each of these subcellular fractions was then subjected to further homogenization in the presence of 1% Triton and centrifugation to yield Triton-insoluble cytoskeletons and Triton-soluble material derived from perikarya and axonal neurites, respectively. Resulting Triton-soluble fractions were "clarified" by high-speed centrifugation to eliminate oligomeric assemblies or soluble neurofilaments. Immunoblot analysis demonstrated quantitative recovery of the aluminum-induced increase in Triton-soluble NF subunits in the perikaryal fraction. Additional aluminum-treated and untreated cultures were pulse-chase radiolabeled with [35S]methionine and fractionated into Triton-insoluble and soluble fractions from isolated perikarya and axonal neurites. Autoradiographic analysis of immunoprecipitated NF subunits revealed that aluminum treatment delayed the translocation of newly synthesized subunits into neurites and resulted in the accumulation of radiolabeled subunits within the Triton-soluble fraction of perikarya. These findings suggest that aluminum may exert a relatively greater effect on NF subunits that have not yet undergone axonal transport and/or incorporation into Triton-insoluble structures vs those that have already deposited into axons. This possibility was supported by the observation that a higher concentration of aluminum was required to alter the electrophoretic migration of in vitro reassembled neurofilaments vs that required for unassembled NF subunits. These findings provide possible mechanisms for the accumulation of NF subunits in perikarya during aluminum intoxication.