Transfected rat high-molecular-weight neurofilament (NF-H) coassembles with vimentin in a predominantly nonphosphorylated form

A fully encoding cDNA for the high-molecular-weight rat neurofilament protein (NF-H) has been isolated from a lambda gt11 library, sequenced and subcloned into eukaryotic expression vectors. Sequence analysis shows that rat NF-H has an overall homology of 72 and 88% with human and mouse NF-H, respectively. The head and rod domains are almost entirely identical, and the divergences are due to differences in the long C-terminal extensions of the molecule. The consensus phosphorylation sequence for neurofilaments Lys-Ser-Pro (KSP) is present 52 times. The predicted molecular mass of the protein is 115 kDa, 42% lower than that observed by SDS-PAGE. Upon transfection into vimentin-containing fibroblasts, such as L tk-, L929, and 3T6 cells, NF-H is seen distributed with vimentin by light and electron microscopic examinations indicating that copolymers of NF-H and vimentin are formed in these cells. Only a negligible proportion of the cells is positive when stained with a number of antibodies directed against phosphorylated NF-H epitopes. This is in contrast with the middle molecular weight NF protein (NF-M) transfected into L tk- and L929 cells, which can readily be detected by antibodies against phosphorylated neurofilament epitopes. The mobilities of the transfected protein on 1- and 2-dimensional gels confirm that NF-H is predominantly in a nonphosphorylated form. These results indicate that phosphorylation of NF-H, but not NF-M, on the KSP sequence is due to protein kinases, which are not present in fibroblasts and are presumably NF-H specific. The stable NF-H-expressing cell lines can therefore be used to study these putative neurofilament kinases in vitro and in vivo.     

Normal neuronal cell bodies of the nucleus tractus mesencephalici nervi trigemini react with antibodies against phosphorylated epitopes on neurofilaments

Reactivity to antibodies directed against phosphorylated neurofilament epitopes is normally present in axons. Phosphorylated neurofilament immunoreactivity is not generally found in neuronal perikarya, except in abnormal states such as Alzheimer's disease. We found that cytoplasm of neurons of the nucleus tractus mesencephalici nervi trigemini in normal mice and rats reacts with monoclonal antibodies against phosphorylated epitopes on neurofilaments. This suggests either that phosphorylated epitopes on neurofilaments are localized in the perikarya of some normal neurons or that the antibodies that were used (SMI 31 and SMI 34) recognize more than phosphorylated epitopes.     

Immunoreactive phosphorylated epitopes on neurofilaments in neuronal perikarya may be obscured by tissue preprocessing

Monoclonal antibodies (SMI 31 and SMI 34) against phosphorylated epitopes on neurofilaments react almost exclusively with neuronal axons in paraffin sections of perfused animals. Using fresh frozen (unfixed) rodent brain sections the immunoreactivity with these antibodies was also present in numerous neuronal cell bodies. Both antibodies also recognized intracellular antigens in the soma of living cells of a human neuroblastoma clone. These data suggest that phosphorylation of neurofilaments in neuronal perikarya is common in vivo. The phosphorylated neurofilament epitopes in neuronal perikarya may become hidden during tissue preprocessing and fixation.     

cDNA coding for the tail region of the high molecular weight rabbit neurofilament protein NF-H

A cDNA library was prepared from rabbit brain mRNA, in the expression vector, lambda gt11. The library was screened with polyclonal antibodies raised against the neurofilament protein NF-H, and a cloned cDNA (KMRH-1) was selected and characterized. The fusion protein coded for by KMRH-1 includes epitopes for two monoclonal antibodies which react with nonphosphorylated sites in the tail region of NF-H. The selected cDNA includes 891 base pairs. It hybridizes to human genomic DNA, as well as to rabbit genomic DNA, and to a rabbit brain mRNA with a size of 4.7 kilobases (kb). The sequence of KMRH-1 includes extensive repeating regions, including one duplicated 60-base segment. Within the first 196 bases, one 13-base segment is repeated 9 times. The cDNA codes for the carboxy-terminal 184 amino acid residues of NF-H, including a series of 9 serines, each surrounded by a similar group of amino acids: ..Ala.Lys.Ser.Pro.(Glu./Val.).Lys.. Comparison of the derived amino acid sequence for KMRH-1 indicates considerable divergence from the sequence information available for rodent NF-H (Robinson et al.: FEBS Lett 209:203-205, 1986). This diversity in amino acid sequence may account for the failure to induce tangles of neurofilaments in animals, such as rats, following treatment with doses of aluminum which are sufficient to induce such tangles in rabbits and to bring on seizures and behavioral pathology in both species.     

Activation of NMDA receptors and Ca2+/calmodulin-dependent protein kinase participate in phosphorylation of neurofilaments induced by protein kinase C

Aberrant phosphorylation of neurofilaments, similar to that occurring in various motor neuron diseases, is produced in cultured motor neurons by activation of protein kinase C (PKC). Following exposure to synthetic diacylglycerol, persistent change in the phosphorylation state of C-terminal domains of neurofilament proteins was detected by increased perikaryal immunoreactivity with the antibody SMI34; this antibody recognizes NF-M/NF-H when C-terminal KSP repeat domains are highly phosphorylated. SMI34 labeling of perikarya and dendrites was prevented by pretreatment with either the NMDA receptor antagonist APV or by the Ca²⁺/calmodulin-dependent protein kinase (CaMK) inhibitor KN-62, but not by antagonists of AMPA/kainate or metabotropic glutamate receptors or by inhibitors of arachidonic acid metabolic pathways. Thus, activation of PKC may induce neurofilament phosphorylation in motor neurons by acting in cooperation with stimulation of NMDA receptors and activation of CaMK. These mechanisms may be relevant to motor neuron disease and other neuronal injuries in which increased PKC activity has been measured. J. Neurosci. Res. 50:514–521, 1997. © 1997 Wiley-Liss, Inc.     

Individual neurofilament subunits reassembled in vitro exhibit unique biochemical, morphological and immunological properties

Purified bovine neurofilament (NF) subunit proteins were reassembled in vitro to form either homopolymeric or heteropolymeric intermediate-sized filaments using single or paired combinations of NF triplet proteins. Using conditions established for the reassembly of bovine NF triplet proteins, we demonstrated that the low Mr NF subunit (NF-L) alone and in combination with the middle Mr NF subunit (NF-M) reassembled very efficiently, i.e. greater than 95% of these proteins formed filaments within 90 min from the start of reassembly. In contra-distinction, the high Mr NF subunit (NF-H) alone and in combination with NF-M or NF-L underwent reassembly to a lesser extent, i.e. 62-88% of these proteins reassembled within 90 min. Immunolabeling of the reassembled NF polymers revealed striking differences in the organization of rod domain determinants. Specifically, antibodies specific for epitopes in the rod domains of NF-H, NF-M and NF-L failed to bind heteropolymeric filaments but recognized rod domains in the homopolymers. In contrast, antibodies specific to head and tail domains of all NF proteins labeled the reassembled hetero- and homopolymeric NFs. Double-labeling of heteropolymers demonstrated that pairs of different NF subunits coassembled into intermediate-sized filaments. Our results also showed that only copolymeric filaments of NF-L and NF-M, but not NF-L/NF-H and NF-M/NF-H were able to form long and stable 10-nm wide filaments. These observations provide new insights into the requirements for stable filament formation from NF subunits. In particular, they support the notion that only NF-L/NF-M, but not NF-L/NF-H or NF-M/NF-H might assemble into a stable filamentous network in vivo.     

Phosphorylation of neurofilament H subunit as related to arrangement of neurofilaments

To find out what causes differences in phosphorylation states in neurofilaments (NF), we selected two types of dendrite, one provided with very few NFs (Purkinje cell) and the other with relatively many (anterior horn cell). We examined these with four monoclonal antibodies selected by the Western blot analysis, two (NE14 and SMI31) recongnizing only phosphorylated, SMI32 recognizing only nonphosphorylated, and N52 recognizing phosphorylation-independent epitopes of NF-H. The immunoperoxidase labeling of dendrites, and also of perikarya, in both neurons was detectable with all four antibodies. After the tissue was treated with Triton X-100, the labeling was still detectable with SMI32 or N52, but undetectable with NE14 and SMI31. The brain homogenate Triton-extracted supernatant after centrifugation at 100,000g for 1 hr showed the staining of NE14, SMI31, and N52 but not that of SMI32. In Purkinje cell dendrite and perikaryon, NFs always appeared singly. In the immunogold labeling, they were labeled only with SMI32 or N52. Labeling by NE14 or SMI31 was distributed throughout the cytoplasm and hardly associated with NFs. In the anterior horn cell dendrite and perikaryon, NFs appeared both single and in bundles. They were predominantly labeled with SMI31 or N52 when they were single, and with NE14, SMI31, or N52 when they were bundled. Even in one NF, portions that appeared single were labeled mostly with SMI32 or N52, while the remainder, to which other NFs approached closely, were labeled mostly with NE14, SMI31, or N52. Thus, when NFs appear singly, NF-H in their projections or cross-bridges with other organelles is not phosphorylated, while when NFs are bundled, NF-H is phosphorylated in crossbridges between NF core filaments. These data may explain why the NF-H is heavily phosphorylated in axons, where NFs are abundant, and not in dendrites and perikarya, where NFs are sparse. Wiley-Liss, Inc.     

Specificity of human anti-neurofilament autoantibodies

The specificities and isotypes of human antibodies that react with neurofilament (NF) proteins were examined by Western blot analysis. Two-thirds of the subjects tested had antibodies to the 200 kDa high molecular weight neurofilament protein (NF-H), and fewer had antibodies to the low and middle molecular weight neurofilament proteins (NF-L and NF-M respectively). Human autoantibodies bound to both native and dephosphorylated NF-H, but some antibodies bound to dephosphorylated NF-H only, indicating the presence of at least two target epitopes. They also recognized a fusion protein containing a segment of the NF-H protein produced by a cDNA clone in Escherichia coli, indicating that they bind to unmodified peptide epitopes. The anti-NF-H antibodies were mostly IgG, but were frequently complexed to IgA or IgM antibodies, possibly with rheumatoid factor or anti-idiotypic activity. These characteristics of anti-NF-H antibodies are most consistent with a secondary immune response that is antigen driven and T-cell dependent.     

Constitutive expression of the mature array of neurofilament proteins by a CNS neuronal cell line.

Neurofilament protein expression was examined immunochemically in a neuronal cell line derived from postnatal day 21 septal tissue. The SN48.1p cell line was found to constitutively synthesize an array of neurofilament proteins typical of a mature neuron. All three neurofilament subunits (NF-L, NF-M, and NF-H) as well as differentially phosphorylated isoforms (P-, P+, P++, and P ) of NF-M and NF-H were identified by immunoblot analysis. Immunofluorescence studies revealed that the neurofilament proteins were components of discrete, filamentous structures. Abnormal intracellular aggregations of neurofilament proteins were never observed. Some SN48.1p cells apportioned specific isoforms into selected intracellular regions based on the molecular weight and phosphorylation level of the protein. NF-L was preferentially localized to perikarya and proximal neurites; NF-M[P++] and NF-H[P ] were distributed to distal aspects of neurites. The expression of these differentiated features of neurofilament proteins and, presumably, the synthesis of the kinases and phosphatases required for normal neurofilament metabolism occurred in the absence of growth factors, differentiating agents, and specialized culture substrates. In addition, the non-neuronal intermediate filaments glial fibrillary acidic protein and epithelial cytokeratin proteins were absent. These data demonstrate that SN48.1p cells exhibit a neurofilament phenotype characteristic of mature neurons and provide a unique model to examine the expression and function of neurofilaments in differentiated neuronal cells.     

Aluminum-induced chronic myelopathy in rabbits

Young adult New Zealand white rabbits, inoculated intracisternally once monthly with 100 micrograms AlCl3, developed progressive hyperreflexia, hypertonia, gait impairment, weight loss, muscle wasting and abnormal righting reflexes over the course of 8 months. No overt encephalopathic features were present. In spinal motor neuron perikarya, dendrites and axonal processes, argentophilic globular inclusions were extensive. Additionally, neurofibrillary tangle-like argentophilic inclusions were consistently present in the gigantocellularis, reticularis, raphe and trapezoid nuclei, but rarely present in the dorsal and ventral subiculum, parasubiculum and anterior thalamus, and never found in the cerebral cortex, substantia nigra, locus ceruleus, or cerebellum. All neuronal inclusions were immunoreactive with monoclonal antibodies recognizing phosphorylated and nonphosphorylated high and intermediate weight neurofilament proteins (SMI 31, SMI 32). Also, some spinal motor neuron inclusions were immunoreactive with a monoclonal antibody recognizing an 'age-related' phosphorylation state of neurofilament (SMI 34). Ultrastructurally, the inclusions consisted of straight or interwoven skeins of 10 nm filaments. This study demonstrates unique variability in the phosphorylation state of aluminum-induced neurofilamentous inclusions in a predominantly motor system degeneration induced by chronic low dose AlCl3.     

The structure of the largest murine neurofilament protein (NF-H) as revealed by cDNA and genomic sequences

The complete primary structure of the largest mammalian neurofilament component, NF-H, is predicted from mouse cDNA and genomic clones, revealing a protein of molecular weight ca. 115,000. A central filament-forming domain structurally typical of all intermediate filament proteins is present, but anomalies are noted which may place constraints on the mechanism of NF-H assembly into filaments. The COOH-terminal portion of the protein is extremely long (661 amino acids) by comparison to non-neuronal intermediate filament components and has a remarkably monotonous, highly charged composition (Glu and Lys at 20% each). Its most remarkable feature is a tandem repeat of a 6 amino acid sequence containing the motif Lys-Ser-Pro that extends for more than half the length of the COOH-terminus. The Lys-Ser-Pro motif appears 48 times and since it is now known that the serine therein is a target for in vivo kinases, the massive axonal phosphorylation of NF-H is explained. Comparison of mouse and human NF-H reveals that otherwise conserved proteins have been subjected to evolutionary mutation within their multiphosphorylation repeat domains, although the Lys-Ser-Pro motif has been conserved.     

Alzheimer's disease antibodies bind specifically to a neurofilament protein in Torpedo cholinergic neurons

Alzheimer's disease (AD) is characterized by neurofibrillary tangles and neuritic plaques and by the degeneration of central cholinergic neurons. Recent studies indicated the presence of antibodies in the sera and cerebrospinal fluid of AD patients which react with neuronal tissue and which recognize cholinergic neurons. In order to identify the cholinergic antigens against which the AD antibodies are directed, we have recently used the purely cholinergic electromotor neurons of the electric fish Torpedo which are chemically homogenous and cross-react antigenically with mammalian cholinergic neurons. This study revealed that immunoglobulins (IgG) from sera of AD patients bind specifically to an antigen in Torpedo electromotor neurons with an apparent molecular weight of 200 kDa. In the present report we attempt to characterize this antigen. The similarity in size of this protein to that of the heavy neurofilament subunit (NF-H) and the association of neurofilaments with plaques and tangles prompted us to examine the possibility that it is a neurofilament protein. Our findings show that IgG from sera of AD patients bind to the NF-H protein of Torpedo cholinergic neurons. Comparison of the binding of AD and control IgG to Torpedo cholinergic NF-H revealed that AD IgG bind to this neurofilament protein more readily than do control IgG. In contrast, AD and control IgG bind similarly to NF-H obtained from the chemically heterogenous Torpedo spinal cord and rat brain. These findings suggest that AD sera contain a repertoire of anti-NF-H IgG and that a subpopulation of these antibodies whose levels are significantly elevated in AD binds to epitopes highly enriched in Torpedo cholinergic NF-H.(ABSTRACT TRUNCATED AT 250 WORDS)     

Monoclonal antibodies to the heavy neurofilament subunit (NF-H) of Torpedo cholinergic neurons.

Previous studies from our laboratory suggest that Alzheimer's disease sera contain a repertoire of antibodies to the heavy neurofilament subunit (NF-H) and that a subpopulation of these antibodies bind specifically to epitopes highly enriched in NF-H isolated from the purely cholinergic electromotor neurons of Torpedo. In the present study, we prepared and characterized monoclonal antibodies (MAbs) that bind to epitopes specifically enriched in Torpedo cholinergic neurons. This was performed by a differential enzyme-linked immunosorbent assay (ELISA) in which MAbs were selected that bind to epitopes much more abundant in the NF-H protein of Torpedo cholinergic neurons than in NF-H from the chemically heterogeneous Torpedo spinal cord. This yielded four MAbs, three of which (TC4, TC8, and TC21) were found to be specific to NF-H and one (TC15) that reacts with both NF-H and the medium-size neurofilament subunit NF-M. Dephosphorylation abolishes the binding of MAbs TC4 and TC15 to Torpedo cholinergic NF-H, partially reduces that of MAb TC21 and has no effect on the binding of MAb TC8. This suggests that the antigenic sites specific to Torpedo cholinergic NF-H contain phosphorylated as well as non phosphorylated epitopes. All the MAbs cross-react with rat brain NF-H.     

Tubule and neurofilament immunoreactivity in human hairy skin: markers for intraepidermal nerve fibers

The cytoplasmic protein gene product 9.5 (PGP 9.5) is considered a reliable marker for intraepidermal nerve fibers (IENFs). However, PGP 9.5 expression has never been compared with antibodies against the main components of the cytoskeleton. We compared the density of PGP 9.5-positive IENF at the leg with that obtained using a panel of antibodies specific for certain cytoskeletal components, namely, anti-unique beta-tubulin (TuJ1), anti-nonphosphorylated microtubule-associated protein-1B (MAP1B), anti-70 and 200 KDa neurofilament (NF), and antiphosphorylated neurofilament (SMI 312), in 15 healthy subjects and in 10 patients with painful neuropathy. We also performed colocalization studies and investigated the relationship between IENFs and Schwann cells. In both controls and neuropathies, the density of IENF labeled by PGP 9.5, TuJ1, and MAP1B did not differ, whereas that of NF and SMI 312 was significantly lower. Double-staining studies confirmed that antibodies against cytoskeletal markers can be used to reliably stain skin nerve fibers, suggesting that they might provide insight into specific axonal impairment in peripheral neuropathies.     

Induction of neurofilament triplet proteins in PC12 cells by nerve growth factor

The localization of neurofilament triplet proteins in PC12 cells grown in the absence of (PC12−) or maintained in the presence of (PC12+) nerve growth factor (NGF) was studied using indirect immunofluorescence and monospecific, immunosorbent purified antibodies to 68,000 (P68), 150,000 (P150) and 200,000 (P200) dalton neurofilament proteins. The intensity of immunofluorescent staining of the triplet protein was always greater in PC12+ compared with PC12− cells. Neuritic staining was seen in PC12+ cells with all 3 monospecific antibodies to neurofilament proteins. However, the perikaryal distribution of each of the neurofilament proteins differed in both PC12+ and PC12− ells. Monospecific antibodies to P68 protein yielded a ‘ball-like’ cytoplasmic staining pattern whereas monospecific antibodies to P150 protein stained in a stippled pattern. Monospecific antibodies to P200 on the other hand diffusely stained the perikaryal cytoplasm with very faint but detectable foci of f ‘ball-like’ configurations and stippling. Electron microscopic study of PC12+ and PC12− cells revealed intermediate filaments in the cell bodies of both as well as in the processes of the former. ‘Ball-like’ clusters of such filaments were rarely seen. However, these filaments lacked the three-dimensional organization typical of intact neurofilaments.It is concluded that PC12 cells contain dissociated or incompletely assembled immunoreactive neurofilament triplet proteins and that these proteins can be induced by NGF. The PC12 cells are therefore an attractive model system not only for studies of neuronal differentiation but also for studies of neurofilament metabolism and disorders thereof.     

Expression of two developmentally regulated brain-specific proteins is correlated with late outgrowth of the pyramidal tract

The regulation of axon outgrowth is not well understood. In previous studies, however, axon elongation has been well correlated with expression of a small number of growth-associated proteins (GAPs). To identify other proteins whose expression could be correlated with axon outgrowth during development of CNS pathways, monoclonal antibodies were raised against growth cone particles isolated from neonatal hamster brains. Two of these antibodies recognized a brain-specific 33 kDa protein associated with intracellular membranes of axons and growth cones. Immunoblotting demonstrated a sharp developmental decline in levels of the protein in hamster brain during the first postnatal week and a more gradual decline thereafter. Immunocytochemical studies with the antibodies revealed ubiquitous staining of the neuropil during the first several days, which by the end of the first week became restricted to a few later-maturing pathways. Staining was most intense in the pyramidal tract and was well correlated with axon outgrowth, which continues until 14 d in this pathway. These results suggest that the 33 kDa protein may, like previously identified GAPs, play a role in axon elongation. Late outgrowth of the hamster pyramidal tract is also correlated with expression of another developmentally regulated protein, the high-molecular-weight neurofilament subunit (NF-H). Immunostaining with a monoclonal antibody that recognized phosphorylated NF-H demonstrated that this subunit does not begin to appear in the late-maturing pyramidal tract fibers until several weeks after birth, in striking contrast to intense immunoreactivity of other spinal cord pathways from postnatal day 1. This finding suggests that specific pathways may have a highly idiosyncratic time course for expression of neurofilament subunits.     

Epitopes located in spatially separate domains of each neurofilament subunit are present in Parkinson's disease Lewy bodies.

Subcortical Lewy bodies are the pathological hallmark of idiopathic Parkinson's disease. This study sought to determine the extent to which each neurofilament subunit [low (NF-L), mid (NF-M), or high (NF-H)] was present in Lewy bodies by using light, confocal, and electron microscopy. A battery of 37 antineurofilament antibodies, characterized as to subunit specificity, epitope domain, and phosphorylation status, was employed to probe substantia nigra Lewy bodies from 15 Parkinson's disease cases. All 37 antibodies labelled Lewy bodies. The epitopes recognized by these antibodies included those in the NF-L rod and tail domains; the NF-M head, rod, and tail domains, as well as epitopes within, and flanking, the multiphosphorylation repeat site; and the NF-H rod domain and multiphosphorylation repeat sites. With these probes, nearly the entire length of each subunit could be demonstrated in Lewy bodies. However, the staining pattern of the Lewy bodies suggested that the tail domains of NF-M and NF-H were present in the periphery of the Lewy body core and in the Lewy body corona, but they appeared to be altered or missing in the center of the Lewy body core. In contrast, the head domain of NF-M, the tail domain of NF-L, and the rod domains of all three subunits are present throughout the Lewy body. These results strongly suggest that the entire extent of each neurofilament subunit is found in Lewy bodies but that the neurofilament subunits may be altered during the processing of these filaments into Lewy bodies.     

Neurofilament protein in clonal lines of mouse neuroblastoma.

As indicated by immunofluorescence with neurofilament antiserum and by electron microscopy two neuroblastoma clones in suspension culture contained a pool of neurofilament subunits which could be induced to assemble into filaments following exposure to vinblastine and colchicine. Under the same culture conditions neuroblastoma cells treated with aluminum extended thick processes with many filaments. The processes were much smaller in nontreated cells but still contained bundles of filaments. These filaments persisted after inhibition of protein synthesis by cycloheximide. In the long neurites formed by cells attached to plastic immunofluorescence with neurofilament antiserum was particularly intense at the growth cone.     

Mice with disrupted midsized and heavy neurofilament genes lack axonal neurofilaments but have unaltered numbers of axonal microtubules.

Mammalian neurofilaments are assembled from the light (NF-L), midsized (NF-M), and heavy (NF-H) neurofilament proteins. While NF-M and NF-H cannot self-assemble into homopolymers, the data concerning NF-L has been more contradictory. In vitro bovine, porcine, and murine NF-L can homopolymerize in the absence of other subunits. However, in vivo studies suggest that neither rat nor mouse NF-L can form filaments when transfected alone into cells lacking endogenous intermediate filaments. By contrast, human NF-L forms homopolymers in similar cell lines. Recently we generated mice with null mutations in the NF-M and NF-H genes. To determine if mouse NF-L can homopolymerize in mouse axons, NF-M and NF-H null mutants were bred to create a line of double mutant animals. Here we show that axons in NF-M/H double mutant animals are largely devoid of 10-nm filaments. Instead, the axoplasm is transformed to a microtubule-based cytoskeleton-although the lack of any increase in tubulin levels per unit length of nerve or of increases in microtubule numbers relative to myelin sheath thickness argues that microtubules are not increased in response to the loss of neurofilaments. Thus in vivo rodent neurofilaments are obligate heteropolymers requiring NF-L plus either NF-M or NF-H to form a filamentous network.