Skip Navigation


Human Molecular Genetics Advance Access originally published online on July 28, 2004
Human Molecular Genetics 2004 13(19):2207-2220; doi:10.1093/hmg/ddh236
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
13/19/2207    most recent
ddh236v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (23)
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by Perez-Olle, R.
Right arrow Articles by Liem, R. K.H.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Perez-Olle, R.
Right arrow Articles by Liem, R. K.H.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

Human Molecular Genetics, Vol. 13, No. 19 © Oxford University Press 2004; all rights reserved

Phenotypic analysis of neurofilament light gene mutations linked to Charcot-Marie-Tooth disease in cell culture models

Raul Perez-Olle, Sidonie T. Jones and Ronald K.H. Liem*

Department of Pathology and Department of Anatomy and Cell Biology, Columbia University College of Physicians and Surgeons, 630 West 168th Street, New York, NY 10032, USA

Received May 5, 2004; Accepted July 12, 2004


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 
Mutations in the neurofilament light (NFL) gene cause Charcot-Marie-Tooth (CMT) disease. There is a wide range of clinical presentations in CMT patients harboring NFL mutations, with patients classified as CMT2E or CMT1F. In this study, we analyzed the effects of five NFL mutations on the assembly and intracellular distribution of intermediate filaments (IFs), and compared the results with those obtained previously for other NFL mutations. Although all NFL mutants affected the formation of IF networks, our data show differential effects on the assembly of IFs depending on the exact nature of the mutation. Defective transport of the mutant NFL subunits was observed for all the CMT-linked NFL mutations, but the characteristics of this defect also depended on the specific mutation. These results show that defects in the assembly and transport of NFs are common to all NFL mutants studied thus far, but the exact nature of the defect appears to be correlated with each mutant genotype.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 
Neurofilaments (NFs) are intermediate filaments (IFs) with important functions in the development and structural maintenance of axons. NFs are composed of three subunits, called heavy (NFH), medium (NFM) and light (NFL) according to their size. NFs are expressed specifically in neurons with higher levels in large myelinated axons. They are synthesized in the cell body and then transported to the axons. NFs have a tri-partite structure typical of all IFs consisting of a conserved central coil-coiled rod domain that mediates dimerization, and globular N-terminal head and C-terminal tail domains. The N-terminal head domains are also important for the formation of filamentous networks (1). NFL can self-polymerize in vitro (2,3) and is required for in vivo assembly of NFs (4,5). Mice null for NFL are viable but display a 15–20% reduction in the number of myelinated axons, decreased axonal diameters and impaired nerve regeneration (6). Alterations of NFs have been identified in a number of neurological diseases, including Alzheimer's disease (7,8), Parkinson's disease (9), amyotrophic lateral sclerosis (ALS; 10), diabetic neuropathy (11) and toxic neuropathies (12).

Charcot-Marie-Tooth (CMT) disease is a slowly progressive bilateral neuropathy with distal predominance. It is an inherited sensory and motor neuropathy, with a reported prevalence of approximately 1 : 2500 (reviewed in 13,14). Symptoms due to loss of muscle control and muscle atrophy include weakness, foot drop, foot deformity and progressive leg deformity. CMT has been subdivided according to nerve conduction velocity studies into a primarily demyelinating form (CMT1) and a primarily axonal form (CMT2). CMT is an extraordinarily heterogeneous disease at both the clinical and molecular levels. Mutations in different genes can present with the same phenotype, while different mutations in the same gene can result in very different clinical symptoms. The overlap at the genetic and clinical levels suggest the possibility that common mechanisms may underlie the development of neuropathy in all CMT patients, and confirms the importance of the interaction and communication between axons and Schwann cells.

The first NFL mutation described in a CMT patient was in the rod domain of the NFL protein (15). A second mutation was subsequently reported in the head domain of NFL (16). Both mutations had autosomal dominant patterns of inheritance. Mutations have not yet been described in the NFM gene, which is adjacent to NFL on chromosome 8p21. Several additional mutations of the NFL gene in other CMT patients have recently been described (17–20). The mutations are located in all three domains of the NFL gene: E7K, P8Q, P8L, P22S, P22T and E89K in the head domain; N97S, A148V and E397K in the rod domain; D469N and a deletion (E528del) in the tail domain (Fig. 1). Some patients have been classified as CMT1F and others as CMT2E, without a clear direct relationship with the specific mutations identified in each patient. The P8 and P22 amino acid residues appear to be mutational hot spots. Previous studies using transfected cells have demonstrated that the P8R and Q333P mutations affect the formation of NF networks and the transport of NFs and mitochondria (21,22). We also demonstrated that D469N NFL was normal with regard to the assembly of IFs (22). D469N NFL was previously identified in a study searching for NF mutations in ALS patients, but was also present in controls (23). Our results confirm that it is not a pathogenic mutation. The E528del variant has also been reported to be a polymorphism (24).



View larger version (14K):
[in this window]
[in a new window]
  		Figure 1. Location of CMT-linked NFL mutations and NFL variants in the different protein domains of NFL. In addition to the five mutations and one variant described in this study, the diagram also shows the NFL mutations/variants that we have previously studied, as well as recently described mutations not included in our study. The different domains of NFL are not drawn to scale.

 
In this study we have characterized newly described NFL mutations that have been linked to additional CMT patients using our previously developed cell culture systems (22). We evaluated the effects of these mutations on the formation of NF networks. We also studied the effects of these mutations on the targeting and distribution of NFs and mitochondria and investigated alterations of the MT network as potential mechanisms of pathogenesis. Our results suggest that these NFL mutations also cause the disruption of NF networks resulting in defective axonal transport. These observations confirm that disruption of axonal transport may be a common pathogenic mechanism involved in a number of neurological and neurodegenerative diseases. Moreover, the cellular models could be useful to assess potential therapeutic options to treat CMT and other diseases characterized by disorganization of the NF networks.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 
We have studied the effects of newly reported CMT-linked NFL mutations on their abilities to assemble and on their transport using cell culture models. We first characterized the effects of NFL mutations on the self-assembly of NFL, then proceeded to analyze their potential dominant-negative effects on the formation of homopolymeric and heteropolymeric IF networks. For these experiments we used SW13 Vim– cells devoid of all endogenous cytoplasmic IFs and SW13 Vim+ that express an endogenous vimentin network. Since our previous studies showed that there were no major differences in IF assembly in a 24–72 h period (22), we chose the 48 h post-transfection time-point for our analysis of IF assembly. These studies have also shown that the mutant phenotypes in the transfected cells could be discerned independent of the level of overexpression by the appropriate comparison with wild-type and non-pathogenic NFL. In the present study, we have observed defects in assembly and transport even when relatively low levels of the transfected protein are expressed. We initially carried out transient transfection experiments in SW13 Vim– cells and collected total cell lysates for analysis of exogenously expressed NFL mutants by Western blotting. All the NFL mutant proteins resulted in the production of a full-size protein of the correct molecular weight.

Assembly characteristics of E7K NFL are indistinguishable from wild-type NFL
One of the reported CMT patients was a compound heterozygote (E7K/P8R) (19). P8R was detected in all affected members of the family, but E7K was not transmitted to the next generation. In our studies, E7K NFL always assembled normally (Table 1). It retained the ability to self-assemble (Fig. 2A), co-assemble with wild-type NFL (Fig. 2G) and form heteropolymeric NFM/NFL networks (Fig. 3A and B). E7K NFL also fully incorporated into the vimentin network when transfected into SW13 Vim+ cells (Fig. 4A and B). Moreover, in cells overexpresssing E7K NFL, we observed a normal MT network (data not shown). These results suggest that E7K is a polymorphism that behaves phenotypically like wild-type NFL and allowed us to use E7K NFL as a control in our experiments.


View this table:
[in this window]
[in a new window]
  		Table 1. Assembly phenotypes of new NFL mutants
 


View larger version (50K):
[in this window]
[in a new window]
  		Figure 2. NFL mutants cannot self-assemble and inhibit the assembly of wild-type NFL. E7K NFL (A and G), E7K/P8R NFL (B and H), P8Q NFL (C and I), P8L NFL (D and J), E89K NFL (E and K), and N97S NFL (F and L) were transfected singly (A–F) or co-transfected with wild-type NFL (G–L) in SW13 Vim– cells. At 48 h post-transfection the cells were fixed and stained with anti-NFL polyclonal antibody plus anti-vimentin monoclonal antibody (data not shown). Dominant assembly defects were observed for all NFL mutations analyzed, except for E7K NFL. Arrows point to discrete inclusions in panels B, D, I and J, and to the ‘sheet’ phenotype in panels F, K and L. Bar=15 µm.

 


View larger version (44K):
[in this window]
[in a new window]
  		Figure 3. Co-transfection with NFM partially rescues some but not all NFL mutants. E7K NFL (A and B), E7K/P8R NFL (C and D), P8Q NFL (E and F), P8L NFL (G and H), E89K NFL (I and J), or N97S NFL (K and L), were co-transfected with NFM in SW13 Vim– cells. At 48 h post-transfection the cells were fixed and stained with anti-NFL monoclonal antibody (A, C, E, G, I and K) plus anti-NFM polyclonal antibody (B, D, F, H, J and L). The staining for both NFL and NFM perfectly co-localized in all cases. In some cases co-transfection of NFM resulted in the formation of filaments, which did not have a normal appearance. Arrows point to curved thick filaments in panels C–F, to straight short thick filaments in panels I–J, and to irregular inclusions of different morphologies in panels G–H. Bar=19 µm.

 


View larger version (55K):
[in this window]
[in a new window]
  		Figure 4. Different degrees of incorporation of NFL mutants in the endogenous vimentin network. E7K NFL (A and B), E7K/P8R NFL (C and D), P8Q NFL (E–H), or N97S NFL (I and J), were transfected in SW13 Vim+ cells. At 48 h post-transfection the cells were fixed and stained with anti-vimentin monoclonal antibody (A, C, E, G and I) plus anti-NFL polyclonal antibody (B, D, F, H and J). Mutants affecting amino acid P8 of NFL had a tendency to form bundled filaments with endogenous vimentin, while thinner filaments were seen after incorporation of E89K and N97S NFL mutants into the endogenous vimentin network. Arrows point to a dense array of parallel thin filaments in B, a filamentous bundle in D and F, a collapsed thick filamentous bundle that also collapses the endogenous vimentin in H, and to a bright core from which thin filaments splay in J. Bar=21 µm.

 
NFL mutations affecting the P8 amino acid result in similar defects of IF assembly
One of the first CMT-linked NFL mutations that has been reported is the P8R mutation. Several additional mutations affecting the P8 amino acid residue of NFL suggest that this residue is a mutational hot-spot. We have analyzed the E7K/P8R compound NFL mutation, as well as two newly described mutations, P8Q and P8L (19).

The compound E7K/P8R mutant was very similar to the P8R mutant (22). The most common phenotype was the formation of one or more discrete cytoplasmic inclusions (Fig. 2B). Co-transfection with wild-type NFL did not result in the formation of filaments, but there were significant numbers of cells displaying a punctate cytoplasmic staining pattern (Fig. 2H). This punctate staining pattern was not observed when E7K/P8R NFL was transfected by itself, suggesting that co-transfection with wild-type NFL either reduces or delays the formation of the inclusions. When the E7K/P8R mutant was co-transfected with NFM (Fig. 3C and D), the resulting filaments were also similar to the P8R mutant, with a predominance of thick, bundled filaments that did not fill the entire cytoplasm (arrows in Fig. 3C and D). Overexpression of the E7K/P8R mutant in SW13 Vim+ cells resulted in the formation of bundled filaments containing both vimentin and NFL (Fig. 4C and D; arrow in Fig. 3D points to a bundled IF structure containing both endogenous vimentin and exogenously expressed E7K/P8R NFL mutant). In cells with NFL inclusions, the endogenous vimentin often appeared to be down-regulated or absent (data not shown). The MT network appeared normal, but tubulin staining was absent from areas where the NFL aggregates were located, and the intensity of tubulin staining was usually increased in the areas adjacent to the NFL inclusions (not shown). Overall, these results suggest that E7K/P8R and P8R are phenocopies.

The effects of the P8Q mutation on the assembly of IFs were also similar to the P8R mutant. P8Q NFL was unable to self-assemble and usually formed aggregates in the form of one or more discrete cytoplasmic inclusions (Fig. 2C; Table 1). When P8Q NFL was co-transfected with wild-type NFL (Fig. 2I, arrows), we observed either punctate staining or the formation of inclusions, indicating that P8Q NFL had a dominant-negative effect on wild-type NFL assembly. Upon co-transfection with NFM (Fig. 3E and F), filaments were observed in nearly all the transfected cells (Table 1). However, these filaments were generally thick and bundled (arrows in Fig. 3E and F). About 10% of transfected cells had NFL inclusions, which were also immunoreactive with anti-NFM antibodies. When SW13 Vim+ cells were transfected with the P8Q mutant, (Fig. 4E and H), we observed similar phenotypes to those obtained for the P8R and E7K/P8R mutants. The P8Q mutant incorporated into the endogenous vimentin network, sometimes resulting in the formation of thick bundles, which were positive for both anti-vimentin and anti-NFL (Fig. 4E and F, arrow in Fig. 4F points to a highly bundled filament). Occasionally, overexpression of P8Q NFL resulted in the collapse of the endogenous vimentin network into a bundled structure (Fig. 4G and H, arrow in Fig. 4H). The P8Q mutant apparently did not affect the MT network, although some MTs were bundled, surrounding the mutant NFL inclusions (data not shown). This bundling of MTs was not observed in areas with punctate staining. In summary, the phenotypes observed upon transfection of the P8Q NFL mutant were similar to those observed for the P8R and E7K/P8R NFL mutants.

Single transfections of P8L NFL in SW13 Vim– cells also resulted in the appearance of inclusions of NFL mutant protein, with predominance of one or more discrete cytoplasmic inclusions (Fig. 2D and Table 1). Co-transfections with wild-type NFL often resulted in punctate staining that served as background for one or more discrete NFL inclusions, although discrete inclusions were also found in the absence of punctate staining (Fig. 2J, arrows). In contrast to the other P8 NFL mutants, the P8L NFL/NFM copolymers did not appear filamentous (Fig. 3G and H, see arrows for different morphologies of inclusions). We also observed a small subset of transfected cells (<10%) displaying a different phenotype with the accumulation of non-filamentous mutant NFL and NFM proteins in a ‘sheet’-like structure (described below in Fig. 2F). The P8L mutant in transfected SW13 Vim+ cells resulted either in aggregates or thick and bundled filaments (data not shown) similar to the P8Q mutant. As observed with other P8 mutants, in some cells with inclusions, there appeared to be little vimentin expression. The MT network appeared normal, although bundling of MTs surrounding mutant NFL inclusions was often seen (data not shown).

Novel phenotypes associated with the E89K and N97S NFL mutations
Two of the recently reported NFL mutations are located in the region close to the junction of the head and the rod domains of NFL (Fig. 1). The E89K mutation is located at the end of the head domain of NFL, while the N97S mutation is at the beginning of the rod domain. Our transient transfection experiments indicated that these two mutations resulted in similar assembly defects. The phenotypes were different from those seen for the P8 NFL mutants, but some similarities were observed with the previously described Q333P mutant (22).

Single transfections of E89K NFL in SW13 Vim– cells resulted in punctate staining in 70% of transfected cells (Fig. 2E and Table 1), indicating that this mutant is also incapable of self-assembly. We occasionally observed thin, short filaments along with the punctate staining, but these filaments were never extended as observed for wild-type or non-pathogenic NFL variants (Fig. 2A). When the E89K mutant was co-transfected with wild-type NFL none of the transfected cells showed a clear filamentous network (Table 1). We observed a wide variety of phenotypes: slightly over half the cells had punctate staining, whereas others showed a ‘sheet’-like structure (Fig. 2K, arrow). When we co-transfected the E89K mutant with NFM, ~40% of the co-transfected cells showed filamentous staining (Fig. 3I and J; Table 1), while a similar percentage showed punctate staining. The filaments were generally thick and not organized (arrows in Fig. 3I and J). We also observed some cells displaying the ‘sheet’ phenotype, but less frequently than in co-transfections with wild-type NFL. Upon transfection in SW13 Vim+ cells, the E89K NFL incorporated into the endogenous vimentin network. There were no apparent alterations of the MT network.

N97S NFL was also not able to self-assemble (Fig. 2F) and exerted a dominant-negative effect on the formation of a homopolymeric NFL network (Fig. 2L), since no filaments were formed in either single or co-tranfection experiments. N97S NFL also displayed the ‘sheet’-like phenotype in single transfection experiments (arrows in Fig. 2F), or when co-transfected with wild-type NFL (arrows in Fig. 2L). Cells co-transfected with N97S NFL and NFM generally showed a punctate staining pattern (~70% of all transfected cells), although we also observed a subset of cells with the ‘sheet’-like phenotype (Fig. 3K and L). Thick filamentous bundles were observed in about 10% of the cells co-transfected with N97S NFL and NFM. In SW13 Vim+ cells transfected with N97S NFL, the mutant protein incorporated into the endogenous vimentin network (Fig. 4I and J). The cells sometimes displayed a densely positive core from which filaments splayed out (arrow in Fig. 4J). Some cells were negative for staining with anti-vimentin antibodies and displayed inclusions of mutant N97S NFL protein (data not shown). We did not observe any obvious alterations of the MT network (data not shown). In summary, the N97S NFL mutation also lacks the ability to self-assemble and has a dominant-negative effect on the formation of homopolymeric NFL networks. A unique phenotype of N97S NFL shared with E89K NFL is the formation of ‘sheet’-like non-filamentous structures.

Defective targeting of NFL mutants to the processes of a neuronal cell line
In order to determine whether the mutant NFL proteins were targeted normally to processes, we transfected the NFL mutant constructs in neuronal CAD cells. CAD cells were differentiated in serum-free media after transfection and fixed and analyzed at different times post-transfection (overnight, 48 h, 72 h and 144 h). We co-stained transfected cells with anti-NFL and anti-tubulin antibodies to determine the distribution of NFL proteins in the processes.

Exogenously expressed E7K NFL was distributed throughout the cell body and processes of CAD cells (Fig. 5A), with a tendency to accumulate distally over time (see large arrow in the distal branching point in Fig. 5A). In some cells we saw a filamentous staining pattern in the cell body (small arrow in Fig. 5A). Similar results were obtained for wild-type NFL and the non-pathogenic D469N NFL polymorphism (R. Perez-Olle et al. submitted for publication). Accumulations of high levels of E7K NFL did not prevent its distribution into the processes, and we rarely observed neuritic degeneration in cells overexpressing E7K NFL. These results support the conclusion that E7K is a non-pathogenic polymorphism.



View larger version (53K):
[in this window]
[in a new window]
  		Figure 5. Defective targeting of mutant NFL to the processes of neuronal CAD cells. CAD cells were grown in complete media for 24 h and then transfected with E7K NFL (A), P8L NFL (B and C), N97S NFL (D) and E89K NFL (E). The cells were allowed to differentiate in serum-free media after transfection and fixed at 48 h (B and E) or 72 h (A, C, and D) post-transfection. Staining was carried out using anti-NFL monoclonal and anti-tubulin polyclonal antibodies. The merged images are shown (NFL in green and tubulin in red). NFL mutations resulted in accumulation of mutant NFL and reduced targeting of the mutant proteins to the neuronal processes. Arrows point to the normal distribution of NFL and normal filaments in the cell body in A, and to abnormally transported mutant NFL in B–E, including accumulation of mutant NFL in the cell body, axonal hillock and proximal segments of neurites in B and E, twisted bundles of mutant NFL that fail to be transported in C and neuritic inclusions in D. Bar=12 µm for A, 10 µm for B–D, and 21 µm for E.

 
The results for the transfections with E7K/P8R, P8Q or P8L mutants were similar to those obtained for the P8R mutant (Fig. 5B and C, P8L shown as a representative example). Many transfected cells accumulated the mutant NFL protein in their cell bodies, with preferential accumulations of mutant protein in the initial segment of the process (large arrows in Fig. 5B). Occasionally small amounts of P8L NFL were observed in processes in a discontinuous fashion (small arrows in Fig. 5B). In some transfected cells, we observed structures resembling highly bundled filaments, forming twists and knots (arrows in Fig. 5C).

For the E89K and N97S NFL mutants, we observed small amounts of mutant NFL protein distally in the processes in some cells (small arrows in Fig. 5E, E89K NFL is shown, and similar phenotypes were observed with N97S NFL). The accumulations of mutant NFL often extended to the middle segments of the processes, rather than being limited to the initial segments (large arrows in Fig. 5E). Unlike the P8 NFL mutants, these two NFL mutants frequently formed small isolated focal accumulations in the processes with increased tubulin immunoreactivity surrounding the inclusions (shown for N97S NFL, large arrow in Fig. 5D). The intensity of tubulin staining distal to the axonal inclusions of mutant NFL protein was also decreased (small arrow in Fig. 5D). Lack of processes in cells expressing very high amounts of mutant NFL protein in the cell body was commonly observed with all the NFL mutations.

Altered subcellular distribution of mitochondria in neuronal and non-neuronal cells overexpressing NFL mutants
Alterations of mitochondrial distribution due to the P8R and Q333P NFL mutations linked to CMT have previously been described (21). We therefore investigated the effects of the additional NFL mutations on the distribution of mitochondria in SW13 Vim– and CAD cell lines using the specific rhodamine-labeled dye MitoTracker Red. No alterations in the distribution of mitochondria in SW13 Vim– cells were observed after overexpression of E7K NFL, which does not form protein inclusions (data not shown). Overexpression of the mutations affecting the P8 residue resulted in a normal distribution of mitochondria when cells expressing the P8Q mutant displayed a punctate staining pattern, (Fig. 6A and C), but when the overexpression of the P8Q mutant resulted in inclusions, progressive clumping of mitochondria was observed (Fig. 6D and F, the small arrow shows the co-aggregation of mitochondria with the inclusion, while the large arrow in Fig. 6E points to the remaining mitochondria that still have a normal distribution). The E89K and the N97S NFL mutants had less severe phenotypes, which correlated with their reduced tendency to form discrete inclusions. However, when inclusions were formed, the perinuclear redistribution of a population of mitochondria observed in cells with a punctate NFL staining pattern (shown for N97S NFL; arrow in Fig. 6H) appeared to progress to the complete trapping and clumping of mitochondria (arrows in Fig. 6K).



View larger version (63K):
[in this window]
[in a new window]
  		Figure 6. Expression of NFL mutants causes defects in the subcellular distribution of mitochondria in non-neuronal and neuronal cells. (A–L) Subcellular distribution of mitochondria in non-neuronal cells. SW13 Vim– cells were transfected with P8Q NFL (A–F) and N97S NFL (G–L), and at 48 h post-transfection mitochondria were labelled with the specific dye MitoTracker Red (B, E, H and K) before being fixed and stained with anti-NFL monoclonal antibody (A, D, G and J). The merged images are also shown (C, F, I and L; NFL in green and MitoTracker Red in red). The formation of mutant NFL inclusions results in the progressive mislocalization of mitochondria to the inclusions. Arrows point to discrete inclusions of mutant NFL protein in D and to mitochondria aggregating within these inclusions in E and K. The big arrow in E points to mitochondria with normal distribution in cells with mutant NFL inclusions. The arrow in H points to the increased perinuclear localization of mitochondria in cells overexpressing mutant NFL but without the formation of discrete inclusions. Bar=15 µm. (M–P) Defective neuritic transport of mitochondria in cells overexpressing mutant NFL. CAD cells were differentiated for 96 h and transfected with E7K NFL (M), N97S NFL (N), P8L NFL (O) and E7K/P8R NFL (P). At 24 h post-transfection mitochondria were labelled with MitoTracker Red and then fixed and stained with anti-NFL monoclonal antibody. The merged images are shown (MitoTracker Red in red and NFL in green). The accumulation of mutant NFL protein prevents the normal distribution of mitochondria to the neurites. In N the small arrows point to reduced neuritic labeling of mitochondria, while the big arrow points to intense mitochondrial labeling co-localizing with a neuritic inclusion of mutant NFL protein. In O and P the big arrows point to the accumulation of mutant NFL protein in the initial segments of neurites, and the small arrows point to the significantly reduced mitochondrial labeling in the neurites, distally to the accumulation of mutant NFL protein. Bar=22 µm for M, N and O, and 18 µm for P.

 
We also observed defective subcellular distribution of mitochondria when we overexpressed mutant NFL proteins in differentiated CAD cells. CAD cells transfected with E7K NFL showed a normal distribution of mitochondrial labeling throughout the cell body and processes (Fig. 6M). In contrast, there was a marked decrease in the distribution of mitochondria in the processes of differentiated CAD cells transfected with any of the other mutant NFL proteins (Fig. 6N–P). This effect correlated with the accumulation of mutant NFL protein in the cell body and the proximal segment of the processes. Expression of E89K or N97S NFL resulted in decreased labeling of mitochondria distal from inclusions of mutant NFL in the processes (small arrows in Fig. 6N for N97S NFL). The inclusions also appeared to trap the mitochondria (large arrow in Fig. 6N). The reduced distribution of mitochondria into the processes was more marked in cells transfected with the P8 NFL mutants. Representative examples of P8L NFL (Fig. 6O) and E7K/P8R NFL (Fig. 6P) are shown. There were accumulations of mutant protein in the initial and proximal segments (large arrows in Fig. 6O and P) and virtually no labeling of mitochondria distally in the processes (small arrows in Fig. 6O and P).

Effects of overexpression of mutant NFL on the MT network
Overexpression of the five NFL mutant proteins in this study did not cause collapse of MTs (Fig. 7A–F). However, we observed displacements of the surrounding MTs by the neurofilamentous inclusions leaving an area devoid of MTs (Fig. 7D–F; the arrow in Fig. 7E points to the area devoid of MTs that is occupied by the NFL inclusion, pointed at by the arrow in Fig. 7F). Similar observations were made in neuronal CAD cells (Fig. 7G–I) when mutant NFL inclusions formed in the process (Fig. 7G, also see arrow in Fig. 7I), leaving an area devoid of MTs (arrow in Fig. 7H). This apparent change in the alignment of MTs could have deleterious effects on axonal transport.



View larger version (92K):
[in this window]
[in a new window]
  		Figure 7. Effects of NFL mutations on the microtubule network. SW13Vim– cells were transfected with N97S NFL (A–C) or P8Q NFL (D–F). CAD cells were transfected with P8Q NFL (G–I). SW13 Vim– cells were fixed at 48 h post-transfection and CAD cells were fixed at 72 h post-transfection. Immunostaining was carried out using anti-NFL polyclonal (A and D) plus anti-tubulin monoclonal (B and E) antibodies, or with anti-NFL monoclonal (G) plus anti-tubulin polyclonal (H) antibodies. Also shown are the merged images (C and F with tubulin in green and NFL in red; I with tubulin in red and NFL in green). The aggregation of NFL mutant proteins appears to alter the normal alignment of MTs but does not result in the collapse of MT networks. The arrows in E and H point to areas in which tubulin staining appears to be displaced away by the formation of inclusions containing mutant NFL, which are pointed at by arrows in F and I. Bar=19 µm for A–F and 5 µm for G–I.

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
MATERIALS AND METHODS
 REFERENCES
 
The number of mutations in the NFL gene in CMT patients is growing, and the discovery of additional mutations requires functional examination of each mutation in order to gain a better understanding of how NFL mutations may result in CMT and to determine if there is a correlation between each specific mutation and the clinical phenotype. The first two reported NFL mutations, Q333P and P8R resulted in defects of IF assembly and intracellular transport (21,22).

Computer analysis of NFL mutations linked to CMT
We used the MultiCoil software (25) to evaluate changes in the coiled-coil probability of the mutant proteins and found only small changes in two of the mutants, E89K and N97S. We also used the Network Protein Sequence @nalysis (NPS@) (26) to evaluate changes in the secondary structure of NFL introduced by the CMT-linked mutations. The E7K/P8R, P8R and P8L mutations introduced a five amino acid stretch with an extended strand in the middle of the random coil structure, while the extended strand stretch introduced by the P8Q mutations comprises four amino acids. These changes in secondary structure could affect the assembly properties of mutant NFL proteins.

Phosphorylation and glycosylation are two of the post-translational modifications of NFs. We used the NetPhos 2.0 prediction server (27) to analyze changes in potential phosphorylation sites introduced by the NFL mutations. The N97S mutation resulted in the introduction of a new potential phosphorylation site for protein kinase C (PKC) at S97. NFL phosphorylated by PKC (at Ser residues 12, 27, 33 and 51) no longer forms filaments (28) and PKC activation in dissociated cultures of spinal cord results in fragmentation of the NF network within 30 min (29). Moreover, treatment with a PKC inhibitor regenerates the endogenous NF network that collapsed by overexpression of NFL in neuronal cells in culture (30). These observations suggest that changes in potential phosphorylation sites resulting from NFL mutations linked to CMT could be biologically important. We also used the NetOGlyc v2.0 Prediction Software from NPS@ (26) to assess changes in potential O-glycosylation sites introduced by NFL mutations and found the loss of a potential glycosylation site (residue S5) in the P8R and P8Q NFL mutants. Glycosylation is another important post-translational modification of NFs (31,32) and the loss of this site could affect the assembly of mutant NFL.

NFL mutations linked to CMT affect the assembly of IF networks
We demonstrated that E7K NFL is not likely to be a pathogenic mutation, since it behaved identically to the wild-type NFL in assembly assays. The E7K/P8R double mutation in cis was phenotypically similar to the P8R mutant. Both P8Q NFL and P8L NFL mutations have phenotypes similar to the P8R mutant. However, the effect of the P8L mutant is the most severe of the three mutations affecting this residue, because even co-transfection with NFM did not lead to the formation of any filaments. In contrast, E89K NFL was not able to self-assemble or co-assemble with wild-type NFL into filaments, but co-transfection with NFM resulted in almost 40% of cells with filamentous structures. Finally, the N97S NFL mutation also resulted in the loss of the capability to self-assemble and co-assemble with NFL, but co-transfection with NFM led to the formation of heteropolymeric filaments in only ~11% of the cells.

NFL mutations result in defective targeting of NFs into processes
Our previous studies using a neuronal cell line showed that the targeting and distribution of mutant NFL was altered (R. Perez-Olle et al. submitted for publication). The wild-type protein was efficiently targeted to neuronal processes, but the transport of mutant NFL to the processes was inhibited. Our results show that E7K NFL was targeted to processes in the same fashion as wild-type NFL, confirming that the nucleotide change resulting in E7K NFL is a polymorphism not linked to the development of CMT. Mutations affecting amino acid P8 (P8R, E7K/P8R, P8Q, P8L) all resulted in accumulations of mutant NFL protein in the cell body and the proximal segments of the processes. Similar results were obtained for E89K and N97S NFL, but these mutant proteins were generally observed further distally in the processes than the P8 mutants. Interestingly, overexpression of E89K and N97S NFL often resulted in small inclusions that appeared to cause distension or swelling of the processes, as well as accumulation and decreased transport of tubulin (see top neurite in Fig. 5D for an example of N97S NFL).

Effects of mutant NFL overexpression on the MT network
Overexpression of the five mutant NFL proteins in this study did not collapse the MT network. Similar results were obtained for the previously reported NFL mutations (22). However, we observed that discrete inclusions of mutant NFL proteins appeared to displace the surrounding MTs, leaving an area devoid of MTs. This phenomenon was particularly striking in the processes of cultured neuronal cells. These subtle effects could potentially affect the interactions between NFs and MTs, between MTs and MT-associated proteins, and between MTs and molecular motors, thereby affecting axonal transport. Impaired axonal transport caused by overexpression of the MT-associated protein tau has been recently reported (33).

Altered intracellular distribution of mitochondria in cells expressing mutant NFL
One of the effects observed with the first two reported NFL mutations was the disruption of axonal transport of NFs and mitochondria (21). We have extended these observations to five additional NFL mutations. Our results are consistent with a model, where progressive aggregation of NFs resulted in trapping of mitochondria within the neurofilamentous inclusions. Inclusions containing SOD1 have been shown to disrupt trafficking of mitochondria in non-neuronal cells (34). Moreover, trapping of organelles (including mitochondria) in inclusions and vacuolar degeneration has also been shown in animal models of ALS and in human ALS patients (35). The lack of mitochondria in the axons would result in energy deficits that could result in axonal dysfunction and, eventually, axonal degeneration.

Defects of axonal transport in neurological disorders
Both disorganization of NFs and alterations of axonal transport occur in many neurodegenerative diseases. A rare form of axonal hereditary motor and sensory neuropathy in which accumulations of NFs occur in giant axons has been described (36). Accumulation of NFs is also one of the main characteristics of giant axonal neuropathy, linked to mutations in the gigaxonin gene (37). Furthermore, a novel neurofilamentopathy in patients that have early-onset dementia has been reported (38), supporting a more direct role for NF aggregation in neurodegeneration. We have previously observed disrupted transport and accumulation of amyloid precursor protein (APP) in cells overexpressing mutant NFL proteins (R. Perez-Olle et al. submitted for publication). Axonal transport defects and accumulations of NFs have also been observed in mouse models for ALS (10,39). There is also circumstantial evidence linking decreased transport of NFs and CMT. Demyelination results in reduced levels of NF phosphorylation and reduces the slow component of axonal transport and the caliber of axons (40). Altered levels of NF phosphorylation have been reported in CMT1 (41). Thus, the demyelinating forms of CMT could in part be the result of secondary defects of the axonal transport of NFs that in turn could be important in the progression of the neuropathy. Progressive and length-dependent loss of axons has been observed in CMT1 patients and has been suggested to result from defective local interactions between axons and glia (reviewed in 42). Moreover, CMT patients with NFL mutations have been reported to present with a demyelinating phenotype (16,19).

Insights into the mechanisms underlying disease heterogeneity in CMT caused by NFL mutations
All reported CMT patients with NFL mutations have autosomal dominant or sporadic patterns of inheritance. CMT caused by NFL mutations comprises a spectrum of clinical presentations sharing common pathogenic mechanisms. According to their clinical presentation, patients with NFL mutations were initially classified as CMT2 (15), but recently other patients with NFL mutations have been classified as CMT1 (19). There is a high degree of heterogeneity among the different CMT patients with NFL mutations. The heterogeneity of the phenotypes caused by NFL mutations is apparent from the great variation in age of onset, which ranges from the first year of life (19) to the third decade of life (15). This heterogeneity can actually be observed within members of the same family. For example, patients of a three-generation family with an E397K NFL mutation presented with very variable clinical phenotypes and age of onset (20). However, sensory nerve action potentials and sensory conduction velocities were generally well preserved in this family, in contrast to other reported CMT families with NFL mutations that showed small or absent sensory nerve action potentials (16,19). Three de novo NFL mutations (P8L, E89K, and N97S) have been reported in patients with a disease onset at 1 or 2 years of age (19). The authors suggested that these mutations might compromise the reproductive capacity of these patients. However, the N97S mutation has also been reported in another CMT patient, a 35-year-old man with disease onset at 15 years (18). In some families with NFL-linked CMT mutations, younger patients have milder symptoms than older patients, suggesting the possibility of a progressive pathogenic process. In our studies, we found that these three mutations caused defects in the ability of the mutant NFL to coassemble with NFM. However, the Q333P NFL also caused a similar assembly defect, but the disease only manifested between the second and third decade (15).

An explanation for the heterogeneity of mutations linked to CMT could be that the effects of the NFL mutations are context dependent. The in vivo interaction of NFL with the myotubularin-related 2 protein phosphatase (MTMR2) gene, which is mutated in CMT4B patients, has been reported (43). The authors suggested that NFL might serve as a scaffold to place MTMR2 at its required site for activity, or that NFL could itself be a substrate for MTMR2. The different NFL mutations could affect the interaction of NFL with MTMR2 to different extents, resulting in primarily demyelinating or axonal phenotypes. A recent study has also identified mutations in the small heat shock protein chaperone, Hsp27 in CMT2 patients that can affect the ability of NFL to self-assemble (44), further supporting the potential roles of co-factors in the disease.

In the current study, we have found that different NFL mutations result in different assembly phenotypes, which in turn may determine specific defects on axonal transport. The mutations in the P8 residue of NFL generally cause severe motor and sensory NCV slowing (19). In our neuronal cell culture model, mutant P8 NFL proteins tended to accumulate in the initial segment of the processes. These aggregates could cause axonal transport blockage, although further evidence is needed to corroborate this hypothesis. Transgenic mouse models have demonstrated that the accumulations of NFs in the cell body with a concomitant absence of axonal NFs is relatively well tolerated (45). Axonal aggregates caused by transgenic expression of an assembly-defective NFL transgene resulted in motor neuron death (46). We have observed different degrees of accumulations of neurofilamentous inclusions in the processes of the neuronal CAD cell line due to the different mutations. It will be interesting to learn whether there are different degrees of neurofilamentous inclusions due to these CMT-linked NFL mutations and how these inclusions correlate with the observations that we have made in the cell culture systems. Axonal blockage and disruption of axonal transport are obvious mechanisms by which NFL mutations could result in neuronal dysfunction and the clinical manifestations of neuropathy in CMT patients. Our studies add further evidence for disruption of axonal transport as one of the underlying pathogenic mechanisms in the development of human neuropathies and other neuromuscular and neurodegenerative disorders.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
REFERENCES
 
Cell lines and culture conditions
For studies of IF assembly we used the human adrenocarcinoma cell line SW13 Vim– that is devoid of all endogenous cytoplasmic IFs, as well as the SW13 Vim+ cell line that expresses an endogenous network of vimentin (47). Both cell lines were generously provided by Dr Robert Evans at the University of Colorado. They were cultured in DMEM supplemented with 5% fetal bovine serum and 1% antibiotic with 5% CO2.

To study the neuronal targeting of mutant NFs and the subcellular distribution of mitochondria we have employed the catecholaminergic cell line CAD (48), which we established as a neuronal cell culture model to study the effect of NFL mutations on their targeting to the neurites (R. Perez-Olle et al., submitted for publication). CAD cells were generously provided by Dr Dona Chikaraishi at Duke University. CAD cells in culture show morphological heterogeneity. CAD cells were routinely maintained in DMEM/F12 media supplemented by 8% fetal bovine serum and 1% antibiotics in a humid atmosphere with 5% CO2.

Generation of expression constructs
We have previously described the cloning of the cDNAs for wild-type NFL and NFM (22). We employed site-directed mutagenesis techniques using the QuickChange Mutagenesis Kit (Stratagene, La Jolla, CA) following the manufacturer's instructions to generate pCI-based expression constructs of the NFL mutations (E7K, E7K/P8R, P8Q, P8L, E89K and N97S) using the following primers:

  1. E7K NFL polymorphism: E7K (F): 5'-GAGTTCCTTCAGCTACAAGCCGTACTACTCGAC-3' and E7K(R): 5'-GTCGAGTAGTACGGCTTGTAGCTGAAGGAACTC-3'.
  2. E7K/P8R NFL mutant: E7K/P8R(F): 5'-GTTCCTTCAGCTACAAGCGGTACTACTCGACC-3' and E7K/P8R(R): 5'-GGTCGAGTAGTACCGCTTGTAGCTGAAGGAAC-3'.
  3. P8Q NFL mutant: P8Q(F): 5'-CCTTCAGCTACGAGCAGTACTACTCGACCTC-3' and P8Q(R): 5'-GAGGTCGAGTAGTACTGCTCGTAGCTGAAGG-3'.
  4. P8L NFL mutant: P8L(F): 5'-CCTTCAGCTACGAGCTGTACTACTCGACCTC-3' and P8L(R): 5'-GAGGTCGAGTAGTACAGCTCGTAGCTGAAGG-3'.
  5. E89K NFL mutant: E89K(F): 5'-GTCCATCCGCACGCAGAAGAAGGCGCAGCTCCAG-3' and E89K(R): 5'-CTGGAGCTGCGCCTTCTTCTGCGTGCGGATGGAC-3'.
  6. N97S NFL mutant: N97S(F): 5'-CAGCTCCAGGACCTCAGTGACCGCTTCGCCAGC-3' and N97S(R): 5'-GCTGGCGAAGCGGTCACTGAGGTCCTGGAGCTG-3'.

The sequences of all mutant expression vectors were confirmed using the BigDyeTM sequencing kit (Applied Biosystems, Foster City, CA) and a Perkin Elmer/Applied Biosystems Model ABI 377A Sequencer at Columbia University's DNA Sequencing Facility.

Transient transfections
The Lipofectamine Plus kit (Invitrogen, Carlsbad, CA) was used for transfection of both cell lines in serum-free media. Transfections were carried out following the protocols provided by the manufacturer as previously described (22 and R. Perez-Olle et al. submitted for publication).

Indirect immunofluorescence microscopy
For immunostaining, the cells were grown in 18 mm glass coverslips. Immunofluorescence staining was performed according to previously published procedures (22). Briefly, at different times post-transfection, the cells were fixed and permeabilized with cold methanol at –20°C, or were fixed with 4% paraformaldehyde in 1xPBS at room temperature and permeabilized with 0.1% Triton X-100. Cells were washed with 1x PBS prior to immunostaining, blocked with 10% normal goat serum and incubated with the primary antibody. The coverslips were then washed with 1x PBS, incubated with the corresponding fluorescently labeled Alexa-Fluor secondary antibody (Molecular Probes, Eugene, OR), washed again with 1x PBS and mounted. When indicated, the cells were also stained with the specific nuclear dye Hoechst 33342. The specificity of immunostaining was verified by omitting the primary antibodies in parallel cultures. A Nikon Eclipse 800 microscope and a Snap II digital camera were used to obtain the images, which were processed using Adobe Photoshop software. We quantified the phenotypes observed regarding IF assembly. Shown are representative results from one experiment in which single transfection and co-transfection of mutant NFL with wild-type NFL or NFM were carried out simultaneously, with data pooled from three coverslips each being evaluated for each paradigm.

Antibodies used
The antibodies used in our experiments were mouse anti-NFL (clone NR4), mouse anti-NFM (clone NN18), mouse anti-vimentin (clone V9) and mouse anti-{alpha}ß-tubulin (clone 2-28-33) monoclonal antibodies (Sigma, St Louis, MO), rabbit anti-NFL and rabbit anti-NFM polyclonal antibodies (49) and anti-ßIII-tubulin antibody (Covance, Princeton, NJ, kindly provided by Dr Michael L. Shelanski, Columbia University).

Subcellular distribution of mitochondria in neuronal and non-neuronal cells overexpressing NFL mutants
In order to study the intracellular distribution of mitochondria we used the specific mitochondrial dye MitoTracker Red (Molecular Probes). SW13 Vim– or CAD cells were transfected with different NFL constructs and at different times post-transfection, the cells were labeled with 400 nM of MitoTracker Red at 37°C for 60 min, washed, fixed with cold 4% paraformaldehyde in 1xPBS, permeabilized with 0.1% Triton X-100 at room temperature and immunostained following routine procedures using anti-NFL antibodies. In the case of SW13 Vim– cells we labeled the mitochondria at 48 h post-transfection, while neuronal CAD cells were differentiated for 96 h in serum-free media before transfection and the mitochondria were labeled at 24 h post-transfection.


   ACKNOWLEDGEMENTS
 
We thank Dr M.A. Lopez-Toledano for his helpful advice and suggestions. The Charcot-Marie-Tooth Association (USA) supported R.P.-O. with a post-doctoral Fellowship (2002) and the First Carolyn Redell Post-Doctoral Fellowship (2003). He is currently a post-doctoral trainee on NIH training grant AG00189. This research is supported by grant NS15182 from the National Institutes of Health (NIH, USA) to R.K.H.L.


   FOOTNOTES
 
* To whom correspondence should be addressed. Tel: +1 2123054078; Fax: +1 2123055498; Email: rkl2{at}columbia.edu


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 

  1. Liem, R.K. (1993) Molecular biology of neuronal intermediate filaments. Curr. Opin. Cell Biol., 5, 12–16.[CrossRef][Medline]

  2. Geisler, N. and Weber, K. (1981) Self-assembly in vitro of the 68,000 molecular weight component of the mammalian neurofilament triplet proteins into intermediate-sized filaments. J. Mol. Biol., 151, 565–571.[CrossRef][ISI][Medline]

  3. Liem, R.K. and Hutchison, S.B. (1982) Purification of individual components of the neurofilament triplet: filament assembly from the 70,000-dalton subunit. Biochemistry, 21, 3221–3226.[CrossRef][Medline]

  4. Ching, G.Y. and Liem, R.K. (1993) Assembly of type IV neuronal intermediate filaments in nonneuronal cells in the absence of preexisting cytoplasmic intermediate filaments. J. Cell Biol., 122, 1323–1335.[Abstract/Free Full Text]

  5. Lee, M.K., Xu, Z., Wong, P.C. and Cleveland, D.W. (1993) Neurofilaments are obligate heteropolymers in vivo. J. Cell Biol., 122, 1337–1350.[Abstract/Free Full Text]

  6. Zhu, Q., Couillard-Despres, S. and Julien, J.P. (1997) Delayed maturation of regenerating myelinated axons in mice lacking neurofilaments. Exp. Neurol., 148, 299–316.[CrossRef][ISI][Medline]

  7. Dai, J., Buijs, R.M., Kamphorst, W. and Swaab, D.F. (2002) Impaired axonal transport of cortical neurons in Alzheimer's disease is associated with neuropathological changes. Brain Res., 948, 138–144.[CrossRef][ISI][Medline]

  8. Pigino, G., Morfini, G., Pelsman, A., Mattson, M.P., Brady, S.T. and Busciglio, J. (2003) Alzheimer's presenilin 1 mutations impair kinesin-based axonal transport. J. Neurosci., 23, 4499–4508.[Abstract/Free Full Text]

  9. Trojanowski, J.Q. and Lee, V.M. (1998) Aggregation of neurofilament and alpha-synuclein proteins in Lewy bodies: implications for the pathogenesis of Parkinson disease and Lewy body dementia. Arch. Neurol., 55, 151–152.[Free Full Text]

  10. Williamson, T.L. and Cleveland, D.W. (1999) Slowing of axonal transport is a very early event in the toxicity of ALS-linked SOD1 mutants to motor neurons. Nature Neurosci., 2, 50–56.[CrossRef][ISI][Medline]

  11. McLean, W.G. (1997) The role of axonal cytoskeleton in diabetic neuropathy. Neurochem. Res., 22, 951–956.[CrossRef][ISI][Medline]

  12. Mizisin, A.P. and Powell, H.C. (1995) Toxic neuropathies. Curr. Opin. Neurol., 8, 367–371.[ISI][Medline]

  13. Shy, M.E., Garbern, J.Y. and Kamholz, J. (2002) Hereditary motor and sensory neuropathies: a biological perspective. Lancet Neurol., 1, 110–118.[CrossRef][ISI][Medline]

  14. Suter, U. and Scherer, S.S. (2003) Disease mechanisms in inherited neuropathies. Nat. Rev. Neurosci., 4, 714–726.[CrossRef][ISI][Medline]

  15. Mersiyanova, I.V., Perepelov, A.V., Polyakov, A.V., Sitnikov, V.F., Dadali, E.L., Oparin, R.B., Petrin, A.N. and Evgrafov, O.V. (2000) A new variant of Charcot-Marie-Tooth disease type 2 is probably the result of a mutation in the neurofilament-light gene. Am. J. Hum. Genet., 67, 37–46.[CrossRef][ISI][Medline]

  16. De Jonghe, P., Mersivanova, I., Nelis, E., Del Favero, J., Martin, J.J., Van Broeckhoven, C., Evgrafov, O. and Timmerman, V. (2001) Further evidence that neurofilament light chain gene mutations can cause Charcot-Marie-Tooth disease type 2E. Ann. Neurol., 49, 245–249.[CrossRef][ISI][Medline]

  17. Georgiou, D.M., Zidar, J., Korosec, M., Middleton, L.T., Kyriakides, T. and Christodoulou, K. (2002) A novel NF-L mutation Pro22Ser is associated with CMT2 in a large Slovenian family. Neurogenetics, 4, 93–96.[CrossRef][ISI][Medline]

  18. Yoshihara, T., Yamamoto, M., Hattori, N., Misu, K., Mori, K., Koike, H. and Sobue, G. (2002) Identification of novel sequence variants in the neurofilament-light gene in a Japanese population: analysis of Charcot-Marie-Tooth disease patients and normal individuals. J. Peripher. Nerv. Syst., 7, 221–224.[CrossRef][ISI][Medline]

  19. Jordanova, A., De Jonghe, P., Boerkoel, C.F., Takashima, H., De Vriendt, E., Ceuterick, C., Martin, J.-J., Butler, I.J., Mancias, P., Papasozomenos, S.Ch. et al. (2003). Mutations in the neurofilament light chain gene (NEFL) cause early onset severe Charcot-Marie-Tooth disease. Brain, 126, 590–597.[Abstract/Free Full Text]

  20. Zuchner, S., Vorgerd, M., Sindern, E. and Schroder J.M. (2004) The novel neurofilament light (NEFL) mutation Glu397Lys is associated with a clinically and morphologically heterogeneous type of Charcot-Marie-Tooth neuropathy. Neuromuscul. Disord., 14, 147–157.[CrossRef][ISI][Medline]

  21. Brownlees, J., Ackerley, S., Grierson, A.J., Jacobsen, N.J., Shea, K., Anderton, B.H., Leigh, P.N., Shaw, C.E. and Miller, C.C. (2002) Charcot-Marie-Tooth disease neurofilament mutations disrupt neurofilament assembly and axonal transport. Hum. Mol. Genet., 11, 2837–2844.[Abstract/Free Full Text]

  22. Perez-Olle, R., Leung, C.L. and Liem, R.K. (2002) Effects of Charcot-Marie-Tooth-linked mutations of the neurofilament light subunit on intermediate filament formation. J. Cell Sci., 115, 4937–4946.

  23. Vechio, J.D., Bruijn, L.I., Xu, Z., Brown, R.H., Jr and Cleveland, D.W. (1996) Sequence variants in human neurofilament proteins: absence of linkage to familial amyotrophic lateral sclerosis. Ann. Neurol., 40, 603–610.[CrossRef][ISI][Medline]

  24. Yamamoto, M., Yoshihara, T., Hattori, N. and Sobue, G. (2004) Glu528del in NEFL is a polymorphic variant rather than a disease-causing mutation for Charcot-Marie-Tooth disease in Japan. Neurogenetics, 5, 75–77.[CrossRef][ISI][Medline]

  25. Wolf, E., Kim, P.S. and Berger, B. (1997) MultiCoil: a program for predicting two- and three-stranded coiled coils. Protein Sci., 6, 1179–1189.[Abstract]

  26. Combet, C., Blanchet, C., Geourjon, C. and Deleage, G. (2000) NPS@: network protein sequence analysis. Trends Biochem. Sci., 25, 147–150.[CrossRef][ISI][Medline]

  27. Blom, N., Gammeltoft, S. and Brunak, S. (1999) Sequence and structure-based prediction of eukaryotic protein phosphorylation sites. J. Mol. Biol., 294, 1351–1362.[CrossRef][ISI][Medline]

  28. Gonda, Y., Nishizawa, K., Ando, S., Kitamura, S., Minoura, Y., Nishi, Y. and Inagaki, M. (1990) Involvement of protein kinase C in the regulation of assembly-disassembly of neurofilaments in vitro. Biochem. Biophys. Res. Commun., 167, 1316–1325.[CrossRef][ISI][Medline]

  29. Doroudchi, M.M. and Durham, H.D. (1996) Activation of protein kinase C induces neurofilament fragmentation, hyperphosphorylation of perikaryal neurofilaments and proximal dendritic swellings in cultured motor neurons. J. Neuropathol. Exp. Neurol., 55, 246–256.[ISI][Medline]

  30. Carter, J., Gragerov, A., Konvicka, K., Elder, G., Weinstein, H. and Lazzarini, R.A. (1998) Neurofilament (NF) assembly; divergent characteristics of human and rodent NF-L subunits. J. Biol. Chem., 273, 5101–5108.[Abstract/Free Full Text]

  31. Dong, D.L., Xu, Z.S., Chevrier, M.R., Cotter, R.J., Cleveland, D.W. and Hart G.W. (1993) Glycosylation of mammalian neurofilaments. Localization of multiple O-linked N-acetylglucosamine moieties on neurofilament polypeptides L and M. J. Biol. Chem., 268, 16679–16687.[Abstract/Free Full Text]

  32. Dong, D.L., Xu, Z.S., Hart, G.W. and Cleveland, D.W. (1996) Cytoplasmic O-GlcNAc modification of the head domain and the KSP repeat motif of the neurofilament protein neurofilament-H. J. Biol. Chem., 271, 20845–20852.[Abstract/Free Full Text]

  33. Mandelkow, E.M., Stamer, K., Vogel, R., Thies, E. and Mandelkow, E. (2003) Clogging of axons by tau, inhibition of axonal traffic and starvation of synapses. Neurobiol. Aging, 24, 1079–1085.[CrossRef][ISI][Medline]

  34. Tobisawa, S., Hozumi, Y., Arawaka, S., Koyama, S., Wada, M., Nagai, M., Aoki, M., Itoyama, Y., Goto, K. and Kato, T. (2003) Mutant SOD1 linked to familial amyotrophic lateral sclerosis, but not wild-type SOD1, induces ER stress in COS7 cells and transgenic mice. Biochem. Biophys. Res. Commun., 303, 496–503.[CrossRef][ISI][Medline]

  35. Sasaki, S. and Iwata, M. (1996) Impairment of fast axonal transport in the proximal axons of anterior horn neurons in amyotrophic lateral sclerosis. Neurology, 47, 535–540.[Abstract/Free Full Text]

  36. Vogel, P., Gabriel, M., Goebel, H.H. and Dyck, P.J. (1985) Hereditary motor sensory neuropathy type II with neurofilament accumulation: new finding or new disorder? Ann. Neurol., 17, 455–461.[CrossRef][ISI][Medline]

  37. Bomont, P., Cavalier, L., Blondeau, F., Ben Hamida, C., Belal, S., Tazir, M., Demir, E., Topaloglu, H., Korinthenberg, R., Tuysuz, B. et al. (2000) The gene encoding gigaxonin, a new member of the cytoskeletal BTB/kelch repeat family, is mutated in giant axonal neuropathy. Nat. Genet., 26, 370–374.[CrossRef][ISI][Medline]

  38. Cairns, N.J., Perry, R.H., Jaros, E., Burn, D., McKeith, I.G., Lowe, J.S., Holton, J., Rossor, M.N., Skullerud, K., Duyckaerts, C. et al. (2003) Patients with a novel neurofilamentopathy: dementia with neurofilament inclusions. Neurosci. Lett., 341, 177–180.[CrossRef][ISI][Medline]

  39. Collard, J.F., Cote, F. and Julien, J.P. (1995) Defective axonal transport in a transgenic mouse model of amyotrophic lateral sclerosis. Nature, 375, 61–64.[CrossRef][Medline]

  40. De Waegh, S.M., Lee, V.M. and Brady, S.T. (1992) Local modulation of neurofilament phosphorylation, axonal caliber, and slow axonal transport by myelinating Schwann cells. Cell, 68, 451–463.[CrossRef][ISI][Medline]

  41. Watson, D.F., Nachtman, F.N., Kuncl, R.W. and Griffin, J.W. (1994) Altered neurofilament phosphorylation and beta tubulin isotypes in Charcot-Marie-Tooth disease type 1. Neurology, 44, 2383–2387.[Abstract/Free Full Text]

  42. Scherer, S. (1999) Axonal pathology in demyelinating diseases. Ann. Neurol., 45, 6–7.[CrossRef][ISI][Medline]

  43. Previtali, S.C., Zerega, B., Sherman, D.L., Brophy, P.J., Dina, G., King, R.H., Salih, M.M., Feltri, L., Quattrini, A., Ravazzolo, R. et al. (2003) Myotubularin-related 2 protein phosphatase and neurofilament light chain protein, both mutated in CMT neuropathies, interact in peripheral nerve. Hum. Mol. Genet., 12, 1713–1723.[Abstract/Free Full Text]

  44. Evgrafov, O.V., Mersiyanova, I., Irobi, J., Van Den Bosch, L., Dierick, I., Leung, C.L., Schagina, O., Verpoorten, N., Van Impe, K., Fedotov, V. et al. (2004). Mutant small heat-shock protein 27 causes axonal Charcot-Marie-Tooth disease and distal hereditary motor neuropathy. Nat. Genet., 36, 602–606.[CrossRef][ISI][Medline]

  45. Eyer, J. and Peterson, A. (1994) Neurofilament-deficient axons and perikaryal aggregates in viable transgenic mice expressing a neurofilament-beta-galactosidase fusion protein. Neuron, 12, 389–405.[CrossRef][ISI][Medline]

  46. Lee, M.K., Marszalek, J.R. and Cleveland, D.W. (1994) A mutant neurofilament subunit causes massive, selective motor neuron death: implications for the pathogenesis of human motor neuron disease. Neuron, 13, 975–988.[CrossRef][ISI][Medline]

  47. Sarria, A.J., Nordeen, S.K. and Evans, R.M. (1990) Regulated expression of vimentin cDNA in cells in the presence and absence of a preexisting vimentin filament network. J. Cell Biol., 111, 553–565.[Abstract/Free Full Text]

  48. Qi, Y., Wang, J.K., McMillian, M. and Chikaraishi, D.M. (1997) Characterization of a CNS cell line, CAD, in which morphological differentiation is initiated by serum deprivation. J. Neurosci., 17, 1217–1225.[Abstract/Free Full Text]

  49. Kaplan, M.P., Chin, S.S., Macioce, P., Srinawasan, J., Hashim, G. and Liem, R.K. (1991) Characterization of a panel of neurofilament antibodies recognizing N- terminal epitopes. J. Neurosci. Res., 30, 545–554.[CrossRef][ISI][Medline]