Brain Advance Access published online on October 25, 2008
Brain, doi:10.1093/brain/awn281
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Disruption of neurofascin and gliomedin at nodes of Ranvier precedes demyelination in experimental allergic neuritis
Département Signalisation Neuronale, CRN2M, UMR 6231, CNRS, Université de la Méditerranée, Université Paul Cézanne, IFR Jean Roche, Marseille, France
Correspondence to:
Correspondence to: Jérôme J. Devaux, PhD, Département Signalisation Neuronale, CRN2M, UMR 6231, CNRS, Faculté de Médecine, IFR Jean Roche, Bd Pierre Dramard, 13926 Marseille, Cedex 20, France E-mail: jerome.devaux{at}univmed.fr
 |
Summary |
|---|
 Top Summary IntroductionMaterials and MethodsResultsDiscussionSupplementary materialFundingReferences |
|---|
High densities of voltage-gated sodium (Nav) channels at nodes of Ranvier enable the rapid regeneration and propagation of the action potentials along myelinated axons. In demyelinating pathologies, myelin alterations lead to conduction slowing and even to conduction block. In order to unravel the mechanisms of conduction failure in inflammatory demyelinating diseases, we have examined two models of Guillain–Barré syndrome: the experimental allergic neuritis induced in the Lewis rat by immunization against peripheral myelin (EAN-PM) and against a neuritogenic P2 peptide (EAN-P2). We found that Nav channel clusters were disrupted at EAN-PM nodes. Neurofascin and gliomedin, two cell adhesion molecules involved with aggregating Nav channels at nodes, were selectively affected prior to demyelination in EAN-PM, indicating that degradation of the axo-glial unit initiated node alteration. This was associated with autoantibodies to neurofascin and gliomedin. Node disruption was, however, independent from complement deposition at nodes, and deposits of the terminal complement complex (C5b-9) were found on the external surface of Schwann cells in EAN-PM. In these animals, the paranodal junctions were also affected and Kv1 channels, which are normally juxtaparanodal, were found dispersed at nodes and paranodes. Altogether, these alterations were associated with conduction deficits in EAN-PM ventral spinal roots. EAN-P2 animals also exhibited inflammatory demyelination, but did not show alteration in nodal clusters or autoantibodies. Our results highlighted the complex mechanisms underlying conduction abnormalities in demyelinating disorders, and unraveled neurofascin and gliomedin as two novel immune targets in experimental allergic neuritis.
Key Words: GBS; AIDP; multiple sclerosis; paranode; sodium channels
Abbreviations: AIDP, acute inflammatory demyelinating polyneuropathy; AMAN, acute motor axonal neuropathy; AP, action potential; CAP, compound action potential; EAN, experimental allergic neuritis; GBS, Guillain–Barré syndrome; PBS, phosphate-buffered saline; PM, peripheral myelin
Received May 28, 2008. Revised October 2, 2008. Accepted October 6, 2008.
|
Introduction |
|---|
|
TopSummaryIntroductionMaterials and MethodsResultsDiscussionSupplementary materialFundingReferences |
|---|
Guillain–Barré syndrome (GBS) is a group of an acute inflammatory neuropathies affecting humans worldwide (see Hughes et al., 1999
for review). Acute inflammatory demyelinating polyneuropathy (AIDP) is the most common form of GBS in Europe and North America, and is characterized by demyelination of peripheral axons and axonal degeneration in severe cases. Acute motor axonal neuropathy (AMAN) is the predominant form of GBS in China and Japan, and is characterized by extensive axonal degeneration with little evidence of demyelination. The mechanisms leading to GBS are still being elucidated. In AMAN, the lipopolysaccharide of Campylobacter jejuni is thought to stimulate the production of antibodies against the ganglioside GM1 (Yuki and Koga, 2006). These autoantibodies bind to the nodal membrane and fix complement, resulting in nodal dysfunction and even axonal degeneration (Hafer-Macko et al., 1996a; Paparounas et al., 1999; OHanlon et al., 2003; Susuki et al., 2007). In AIDP, the histopathology of affected nerves suggests that T cells and macrophages mediate demyelination (Hughes et al., 1992; Hafer-Macko et al., 1996b; Schmidt et al., 1996). In a widely used animal model of AIDP, experimental allergic neuritis (EAN), P2 myelin protein elicits a similar pathology to AIDP, and the passive transfer of T cells is sufficient to produce the disease (Uyemura et al., 1982; Olee et al., 1990). Many AIDP patients present serum antibodies to peripheral nerve myelin protein (P0, PMP22, P2 and connexin-32) (Gabriel et al., 2000; Kwa et al., 2001; Allen et al., 2005); however, the contribution of the autoantibodies to AIDP pathology is less well established.
Axonal loss and the failure of action potentials (APs) to propagate correctly from node to node along peripheral axons both contribute to the disabilities observed in patients with GBS. The mechanisms leading to conduction failure in GBS are not yet fully understood, but inflammatory processes at nodes of Ranvier have been implicated. Indeed, disruption of the nodal Nav clusters has been reported in animal models of AIDP and AMAN (Novakovic et al., 1998; Susuki et al., 2007). The deposition of anti-GM1 antibodies and complement at nodes are believed to disrupt Nav channel clusters, at least in AMAN rabbits (Susuki et al., 2007). In the AIDP model, the mechanisms underlying the disorganization of Nav channel clusters are unknown. In addition, it remains to be determined whether the loss of Nav channels is common to other models of AIDP.
Important progress has been made in the understanding of the mechanisms leading to node formation. In the PNS, ankyrin-G and Nav channel clustering at nodes is tightly regulated by the overlaying myelinating Schwann cells, especially by the trans-interaction in between gliomedin, expressed on Schwann cell microvilli and neurofascin-186 (NF186), expressed at the nodal axonal membrane (Jenkins and Bennett, 2002; Eshed et al., 2005; Sherman et al., 2005). In contrast, the mechanisms implicated in axonal disorganization during demyelination are less clear. In multiple sclerosis patients, paranodal regions appear to be disrupted prior to node alterations during demyelination (Howell et al., 2006). In AIDP, little is known about the early changes to the axon–glial unit. Herein, we have investigated this issue in two animal models of AIDP, and unraveled the complex mechanisms leading to node alteration. In one model of AIDP, we demonstrated that NF186 and gliomedin are selectively affected before the onset of the disease. These alterations correlate with the presence of serum antibodies against neurofascin and gliomedin, and with the disruption of the Nav channel clusters.
|
Materials and Methods |
|---|
|
TopSummaryIntroductionMaterials and MethodsResultsDiscussionSupplementary materialFundingReferences |
|---|
Immunization
PNS myelin was prepared from fresh guinea pig peripheral nerves by sucrose density gradient centrifugation (Norton and Poduslo, 1973
; Kadlubowski and Hughes, 1979) and the myelin pellet was resuspended in saline (150 mg of myelin/ml) and stored at –80°C. Neurofascin or gliomedin protein bands were not detected by immunoblots in PNS myelin fraction (Supplementary Fig. 1B and C). The synthetic peptide of bovine P2 myelin protein (amino acids 53–78; Uyemura et al., 1982) was purchased from Bachem (Bubendorf, Switzerland) and dissolved in saline (2 mg/ml). Male inbred adult Lewis rats (6- to 7-weeks old; Elevage Janvier, Le Genest St Isle, France) were sensitized by subcutaneous injection at the base of the tail with 200 µl of an antigen emulsion. The antigens were emulsified with an equal volume of complete Freund's adjuvant (CFA; Sigma, St Louis, MO, USA). Final doses in the inoculum were 100 µg H37RA Mycobacterium tuberculosis and 200 µg P2 antigen (EAN-P2), or 100 µg H37RA M. tuberculosis and 15 mg of wet PNS myelin (EAN-PM). Control animals were injected with 100 µl of saline emulsified with 100 µl of CFA. Animals were weighed and observed daily. Clinical signs were graded as follows: 0 = no illness; 1 = tail tip hanging; 2 = limp tail; 3 = tail paralysis; 4 = gait ataxia; 5 = mild paraparesis; 6 = severe paraparesis; 7 = paraplegia; 8 = tetraparesis; 9 = moribund; 10 = death. All the experiments were in lines with the European Community's guiding principles on the care and use of animals (86/609/CEE).
Histopathological analysis
Rats were euthanized in a CO2 chamber at disease peaks (clinical score = 6). Sciatic nerves and spinal roots were dissected out and placed in 2% paraformaldehyde and 2% glutaraldehyde in 0.1 M phosphate buffer (PB) overnight at 4°C, rinsed in PB, post-fixed in 1% OsO4 in 0.1 M PB for 1 h. After several washes, tissues were dehydrated in an ascending series of ethanol and embedded in epoxy resin. Transverse semi-thin sections were cut (Leica Microsystems GmbH, Wetzlar, Germany), stained with toluidine blue, and examined by light microscopy with a 100 x objective lens. The number of degenerated myelinated axons was measured in five animals for each group (
1500 axons counted for each group). Schwann tubes, >2 µm in diameter, and typically containing myelin debris or infiltrated macrophages but no recognizable axon, were considered to be degenerating myelinated axons. Examples are shown in Supplementary Fig. 1D and E.
Immunolabelling
A 16 amino acid peptide sequence (VGFVGLDPGAPDSTRD), corresponding to amino acids 22–37 from the N-terminal region of KCNQ2 was synthesized (Cooper et al., 2001), conjugated to keyhole limpet haemocyanin, and two rabbits were immunized. The antisera were collected and purified against the peptide immunogen (Eurogentec, Seraing, Belgium). These antisera stained cells that were transfected with a cDNA encoding KCNQ2 but not KCNQ3 (data not shown).
Most experiments were done on fixed tissues, with the exception of C5b-9 and Kv3.1b staining which were done on unfixed tissues. Sciatic and spinal nerves were dissected out and fixed in 2% paraformaldehyde in 0.1 M phosphate-buffered saline (PBS) for 1 h at 4°C, then rinsed in PBS. Axons were gently teased, dried on glass slides and stored at –20°C. Alternatively, fixed nerves were cryoprotected in 30% sucrose in 0.1 M PBS overnight at 4°C, then cut into 5 to 10 µm thick cryosections. Frozen sections and teased fibres were permeabilized by immersion in –20°C acetone for 10 min, blocked at room temperature for 1 h with 5% fish skin gelatin containing 0.1% Triton X-100 in PBS and incubated overnight at 4°C with various combinations of primary antibodies—rabbit antisera against KCNQ2 (1/200), gliomedin (1/500; Eshed et al., 2005), NF186 (1/500; Southwood et al., 2004), Kv3.1b (1/100; Alomone Laboratories, Jerusalem, Israel), Nav1.6 (1/100; Alomone Laboratories), ankyrin-G (1/1000; Bouzidi et al., 2002) or Caspr (1/1000; Menegoz et al., 1997); mouse monoclonal antibodies against PanNav channels (K58/35; 1:500; Sigma), Nav1.2 (1/100; UC Davis/NINDS/NIMH NeuroMab Facility), C5b-9 (1/50; DakoCytomation, Glostrup, Denmark), C5b-9 (1/50; sc-66190; Santa Cruz Biotechnology, Santa Cruz, CA), Kv1.2 (1/100; UC Davis/NINDS/NIMH NeuroMab Facility) or PanNeurofascin (1/100; UC Davis/NINDS/NIMH NeuroMab Facility); goat antibody against contactin (1/200; R&D Systems, Minneapolis, MN, USA). The slides were then washed several times and incubated with the appropriate fluorescein-, rhodamine- and Cy5-conjugated donkey cross-affinity purified secondary antibodies (1:100; Jackson ImmunoResearch Laboratories, West Grove, PA, USA). Slides were stained with DAPI to visualize cell nuclei, mounted with Mowiol plus 2% DABCO, and examined using confocal (Leica TCS SP2) and epifluorescence (Leica DMR) microscopes. Digital images were manipulated into figures with Adobe Photoshop and CorelDraw.
Teased fibers from five to six animals were analyzed for each group (100 axons counted in total). The lengths of individual Caspr-positive paranodes and intercalated nodes were measured using ImageJ software. The mean node length and mean paranodal length were then compared by Student's t-test. For the quantification of demyelinated axons, teased fibers were stained for Caspr, and intercalated nodes >5 µm were counted as paranodally demyelinated. Indeed, EAN nerves show predominantly paranodal demyelination, but also a general node widening. At disease peaks, the node widening presents a near Gaussian distribution for length ranging from 1 to 4 µm in EAN-P2 and for length ranging from 1 to 5 µm in EAN-PM. We thus concluded that node >5 µm more clearly reflects paranodal demyelination. The percentage of nodes >10 µm is also indicated in Figs 2 and 5 for comparison. For measuring Caspr and neurofascin density, a line selection of constant length was centered on the middle of each node (demyelinated axons were rejected) and the intensity was measured with ImageJ. Grey scale values were then normalized and averaged, as shown in Figs 3 and 5.
|
|
Electrophysiology
Recordings were performed at disease peaks (clinical score = 6). After euthanizing, the cauda equina was quickly dissected and transferred into artificial cerebrospinal fluid (ACSF), which contained (in millimolar) 126 NaCl, 3 KCl, 2 CaCl2, 2 MgSO4, 1.25 NaH2PO4, 26 NaHCO3 and 10 dextrose, pH 7.4–7.5. The L5 ventral spinal roots were cut into 2 cm segments. Nerves were then placed in a three compartment recording chamber and perfused (1–2 ml/min) in 36°C ACSF equilibrated with 95% O2–5% CO2. The distal end was stimulated supramaximally (40 µs duration) through two electrodes isolated with Vaseline, and recordings were performed at the proximal end. Signals were amplified, digitized at 500 kHz and stored on a hard disk. Drugs were applied in the central compartment of the chamber (1.4 cm in length); measurements were made once the effects had reached a steady state—typically 30–45 min after application. Nerves were continuously stimulated at a frequency of 0.25 Hz. The delay and duration of compound APs (CAPs) were calculated at half the maximal amplitude; the maximal amplitude and area under the curve were also measured. For recruitment analysis, the amplitude of CAPs was measured and plotted as a function of the stimulation intensity. For refractory period analysis, two stimuli were applied at different intervals, and the amplitude of the second CAP was measured and plotted as a function of the stimulus interval. To ensure that the amplitude of the second response was accurately assessed, the first response was subtracted from all the recordings. The 4-aminopyridine was purchased from Sigma, and XE991 was purchased from Tocris (Ellisville, MI, USA). Histological examinations were also performed on rats used for electrophysiology.
Biochemistry
The extracellular domain of neurofascin-155 (NF155-Fc), neurofascin-186 (NF186-Fc), gliomedin (Gldn-Fc) and contactin (Con-Fc) fused to human Fc were obtained as described previously (Charles et al., 2002; Eshed et al., 2005). Sera were collected from five control, five EAN-P2 and eight EAN-PM animals at disease peaks (clinical score = 6). Sciatic nerves from adult rats (100 µg of proteins) and NF155-Fc (1 µg) were directly solubilized in SDS sample buffer containing β-mercaptoethanol and complete protease inhibitors cocktail (Roche Diagnostics GmbH, Mannheim, Germany), then heated for 2 min at 90°C. The samples were centrifuged for 10 min at 750g. Protein concentration was determined using the BioRad kit (Bio-Rad, Hercules, CA, USA). Samples were loaded on a 7.5% SDS–PAGE gel, then transferred onto a nitrocellulose membrane (Laemmli, 1970). Membranes were blocked for 1 h with 5% powdered skim milk 0.5% Tween-20 in Tris-buffered saline (TBS) and incubated with mouse anti-PanNF antibodies (1:2000) or rat sera (1:50) overnight at 4°C. After several washes the blots were incubated in peroxidase-conjugated secondary antibodies against rat IgG or mouse IgG (1:5,000; Jackson ImmunoResearch) for 1 h at RT, washed several times and revealed using ECL plusTM (Amersham, Arlington Heights, IL, USA). For dot blot experiments, Fc fusion proteins (0.5 µg) were deposited onto a nitrocellulose membrane, dried for 20 min, and treated as described above. The integrated density of each spot was measured with ImageJ using a round selection of constant area.
|
Results |
|---|
|
TopSummaryIntroductionMaterials and MethodsResultsDiscussionSupplementary materialFundingReferences |
|---|
Clinical course and histopathology of EAN
We generated two animal models of AIDP, by immunizing Lewis rats with purified myelin (PM) from guinea pig nerves (15 mg; EAN-PM) (Hahn et al., 1988
), or an immunogenic peptide from P2 (200 µg; EAN-P2) (Uyemura et al., 1982; Hahn et al., 1991). The clinical signs of EAN-P2 appeared around 8 days post-injection, and peaked at days 15 to 17 (Supplementary Fig. 1A). The clinical signs of EAN-PM developed more rapidly with a peak at days 12 to 14 (Supplementary Fig. 1A). At disease peaks, the animals from both groups presented a clinical score of 6 characterized by an inability to right the tail, severe hindlimb paraparesis and a paralysis of the hindfeet. Then, the animals progressively improved, although their gait disturbance persisted several weeks.
To examine the histopathology of EAN, teased fibers were prepared from sciatic nerves and the L5 ventral roots sampled at disease peaks (clinical score = 6) and were immunostained for Caspr and contactin, which labels paranodes (Einheber et al., 1997; Rios et al., 2000). Using the distance between the Caspr-positive paranodes of adjacent myelin sheaths to measure the nodal gap, we found abnormally widened nodes in EAN-P2 nerves (up to 20 µm in length); these were more common in EAN-PM nerves (Figs 1 and 2). There were few examples of long unmyelinated segments suggesting segmental demyelination. Thus, EAN produces predominantly paranodal retraction/demyelination. We found similar results using longitudinal cryosections (data not shown). A quantitative analysis (see Materials and methods section) confirmed that demyelinated segments were longer in EAN-PM animals than in EAN-P2 animals (Fig. 2). Because paranodal demyelination was more pronounced in ventral roots than in sciatic nerves in both models, we focused our study on L5 ventral roots. Node length also generally increased in both models of AIDP (Fig. 2B and C), but most prominently in EAN-PM roots (Fig. 2B). Many paranodes were importantly widened in EAN-PM nerves (Fig. 2B and D), indicating that, in addition to paranodal retraction/demyelination, paranodal structures were altered.
|
Examination of semi-thin transverse sections of ventral roots revealed the presence of degenerating myelinated axons (Schwann tubes usually containing myelin debris or infiltrated macrophages but not axon) in both types of EAN (Fig. 2A and Supplementary Fig. 1). A quantitative analysis showed that axonal degeneration was more pronounced in EAN-PM than in EAN-P2.
Altered organization of the nodal components in EAN
To investigate the possibility that nodal Nav channel clusters are affected in these models of GBS, we examined the localization of Nav channels using a monoclonal antibody that recognizes all Nav isoforms (PanNav). In EAN-P2 nerves, Nav channels were properly clustered at normal appearing nodes (Fig. 1Ba). When a demyelinated segment was encountered, Nav staining flanked each paranode; staining was typically absent in the demyelinated gap (Fig. 1Bb and c). We observed the same patterns of immunostaining for ankyrin-G and Nav1.6; both were clustered at normal appearing nodes and formed hemi-nodes (clusters adjacent to each paranodes) at fibers with widened nodes (Supplementary Fig. 2). Nav1.6 appeared to be the major subunit expressed at nodes and hemi-nodes in EAN nerves; we did not detect Nav1.2 subunits at nodes (Table 1).
|
In EAN-PM nerves, most nodes had clustered Nav channels (Fig. 1Ca), but about 26% had clearly disrupted Nav channel clusters (Fig. 1Cb and c and Table 1). These latter presented weak and diffuse Nav staining. Some elongated nodes presented clusters of Nav channels at hemi-nodes flanking the paranodes. However, most elongated nodes had strikingly disrupted Nav clusters (Fig. 1Cb and c). Nav channels were either diffusely localized along the demyelinated segments or undetectable. The same pattern was observed with ankyrin-G and Nav1.6, and no Nav1.2-positive nodes were found in EAN-PM nerves (Supplementary Fig. 2 and Table 1).
We then immunostained teased fibers with a PanNeurofascin (PanNF) monoclonal antibody that recognizes both NF186, which is nodal (Davis et al., 1996), and NF155, which is paranodal (Tait et al., 2000). In normal nerves, the PanNF antibody labeled nodes more intensely than paranodes (Fig. 3A). In EAN-P2 nerves, the same pattern was observed in normal-appearing nodes (Fig. 3Ba), and intense neurofascin-immunoreactivity was maintained at hemi-nodes bordering the paranodes in demyelinated segments (Fig. 3Bb). In EAN-PM nerves, PanNF staining appeared diminished at most nodes compared to paranodes (Fig. 3Ca and Cb) or was even undetectable (Fig. 3Cc). When a demyelinated segment was encountered, neurofascin was clustered at hemi-nodes or undetectable. A quantitative study of EAN-PM nodes revealed that 38% had disrupted neurofascin (Table 1). Worth noting, nodes lacking neurofascin staining also presented disrupted or diminished staining for ankyrin-G and Nav1.6 (Supplementary Fig. 2).
To quantify the decreased neurofascin density at nodes, we selected normal-appearing nodes of similar sizes, and we plotted the grey value of PanNF and Caspr staining as a function of fiber length. The plots of staining intensity were comparable in control and EAN-P2 nerves and peaked at nodes (Fig. 3D). In contrast, PanNF plots dropped at EAN-PM nodes, leaving prominent paranodal staining. Caspr plots were comparable in both control and EAN-P2, although slightly dispersed in EAN-PM nerves (Fig. 3E).
To demonstrate that NF186 is missing in particular, we immunostained teased fibers for NF186 but also for gliomedin, its Schwann cell binding partner (Eshed et al., 2005). NF186 and gliomedin were both detected at nodes in control animals (Fig. 4A and D). In EAN-P2 nerves, gliomedin and NF186 were found clustered at every node and hemi-node (Table 1 and Supplementary Fig. 3), as for Nav channels. In EAN-PM nerves, NF186 and gliomedin were clustered at many nodes or hemi-nodes (Fig. 4B and E), but were often undetectable or diffusely localized in demyelinated segments (Fig. 4C and E). A quantitative study revealed that 37% and 41% of EAN-PM nodes had disrupted NF186 and gliomedin, respectively (Table 1). To determine whether nodes lacking Nav channel clusters also lack gliomedin, we triple-stained teased fibers for gliomedin, contactin, and Nav channels (Fig. 4F–H). We found that 41 nodes out of 209 (19.6%) lacked both Nav channels and gliomedin (Fig. 4 H), and only two nodes (1.0%) exhibited gliomedin aggregates in the absence of Nav channels. The remaining nodes exhibited normal Nav channel and gliomedin aggregates (Fig. 4F; 53.1%) or Nav channel clusters in the absence of gliomedin (Fig. 4G; 26.3%). Altogether, these data suggested that NF186 and gliomedin are more prominently affected, and suggested that disruption of NF186/gliomedin aggregates may initiate Nav channel cluster disorganization.
|
Disruption of NF186 and gliomedin precedes Nav channel dispersion
Gliomedin and NF186 are, indeed, crucial for the formation of Nav channel aggregates at node of Ranvier (Eshed et al., 2005
; Sherman et al., 2005; Zonta et al., 2008). Therefore, we examined whether disruption of gliomedin and NF186 aggregates may occur during the initial stage of the disease. As early as 7 days post-immunization (dpi), when the first clinical signs began to appear, the node length increased significantly in EAN-PM nerves, but few demyelinated axons were found at this stage (Fig. 5A and C). Paranodal demyelination was prominent only at 11 and 13 dpi (Fig. 5B and C).
|
The appearance of the first clinical signs at 7 dpi correlated with the detection of the first nodes lacking gliomedin or NF186 clusters (Fig. 5C). By 9 dpi (mean clinical score = 2.5; Supplementary Fig. 1), the number of nodes showing disrupted gliomedin or NF186 clusters rose significantly. Interestingly, at this stage, most nodes exhibited bright and focal Nav channel staining (Fig. 5D), and only a few showed signs of paranodal demyelination. A few days later, as the clinical signs increased in severity (mean clinical score = 5.1 at 11 dpi), the number of nodes with disrupted gliomedin or NF186 reached a near-maximal level which only varied substantially by 13 dpi (mean clinical score = 5.8). Alterations in Nav channel clusters were significant by 11 dpi as paranodal demyelination rose, and then increased at 13 dpi. Examination of the intensity of the PanNF staining corroborated the above impressions (Fig. 5E and F): neurofascin density significantly decreased at EAN-PM nodes as early as 7 dpi, then further decreased as clinical grade worsen, and reached minimal levels at 11 and 13 dpi. Together, these results revealed that the disorganization of gliomedin and NF186 aggregates precedes demyelination and the dispersion of the Nav channels.
Autoantibodies and complement in EAN
We conjectured that autoantibodies against neurofascin or gliomedin may cause its depletion from nodes in EAN-PM animals. Sera from control, EAN-P2 and EAN-PM animals were first tested on proteins extracts from sciatic nerves. As shown in Fig. 6A, sera from EAN-PM animals, but not from control and EAN-P2 animals, recognized proteins with apparent molecular masses that corresponded to NF155 and NF186. In addition, we found that EAN-PM sera recognized protein bands that may correspond to MAG, P2 and PO (data not shown), as previously reported (Zhu et al., 1994). However, we could hardly resolve by western blot whether gliomedin is specifically recognized. To determine whether autoantibodies recognized gliomedin and the neurofascin isoforms, the extracellular regions of NF155, NF186 and gliomedin fused to human Fc were expressed in HEK cells and immobilized on nitrocellulose membrane. Serum IgG from EAN-PM animals, but not from control or EAN-P2 animals, bound to immobilized NF155 (Fig. 6B). Similarly, EAN-PM serum IgG recognized the native extracellular domain of NF186 and gliomedin (Fig. 6C). However, EAN-PM serum IgG did not bind to immobilized contactin-Fc (Fig. 6C). These data indicated that, in addition to autoantibodies directed against myelin proteins, EAN-PM rats generate autoantibodies toward nodal adhesion molecules.
|
We then examined the deposition of the terminal complement complex (C5b-9) on the EAN-PM ventral roots, as these are potential inflammatory mediators of humoral immunity. C5b-9 staining was almost undetectable at 7 dpi, and few myelinated fibres presented a uniform C5b-9 staining at 9 dpi (data not shown). At 11 and 13 dpi, C5b-9 staining was present along the surface of the Schwann cells, but was not specifically detected at nodes or along the axons (Fig. 6D and E). At any days tested, no correlation could be made in between C5b-9 deposition and neurofascin/gliomedin disappearance. Similar results were obtained with two distinct antibodies against C5b-9 which both labeled degenerating fibers (used as positive control). Thus, Nav channel cluster disruption is independent from complement deposition at nodes in EAN.
Voltage-gated K+ channels are affected in EAN
We next investigated whether the localization of voltage-gated K+ channels is also affected in EAN nerves. Kv1.1 and Kv1.2 subunits forms heteromeric channels that are concentrated in the juxtaparanodal regions in normal nerves (Mi et al., 1995), but are mislocalized to the paranodal region in mutants that lack paranodal septate-like junctions (Dupree et al., 1999; Bhat et al., 2001; Boyle et al., 2001) and in demyelinating diseases (Arroyo et al., 2004; Devaux and Scherer, 2005). In EAN-P2, Kv1.2 channels remained concentrated at juxtaparanodes even in axons with elongated nodes, and were not detected in demyelinated segments or paranodes (Fig. 7B and Table 1), indicating that paranodes are unaffected in EAN-P2. In EAN-PM, the distribution of Kv1.2 was more heterogeneous (Fig. 7C). Only 51% of the nodes had normal juxtaparanodal Kv1.2 staining (Table 1), 32.2% had diffuse Kv1.2 staining at paranodes, and 11.6% had Kv1.2 in both the paranodes and nodes. Co-staining for contactin demonstrated that most fibers presenting disrupted Kv1.2 also presented slightly altered paranodes.
|
KCNQ2 channels are normally found at PNS and CNS nodes (Devaux et al., 2004
), and share a common anchoring mechanism with nodal Nav channels (Pan et al., 2006). Accordingly, we found that KCNQ2 channel distribution recapitulated that of the Nav channels in both EAN models. In EAN-P2 nerves, KCNQ2 channels were aggregated at every node and hemi-node (Fig. 7B). In EAN-PM, KCNQ2 channels presented a disrupted organization in many nodes (Fig. 7C). As for Nav channels, disrupted KCNQ2 clusters were often associated to demyelinated segments, and in some cases, normal-appearing nodes presented disrupted KCNQ2 channel clusters.
Because Kv3.1b is aberrantly expressed at demyelinated PNS nodes in Trembler-J mice (Devaux and Scherer, 2005), we investigated whether Kv3.1b is aberrantly expressed at PNS nodes in EAN. We did not detect increases in the percentage of Kv3.1b-positive node in both EAN models (data not shown).
Electrophysiology of EAN nerves
To determine whether these morphological changes are correlated to physiological dysfunctions in the two different models of EAN, we recorded extracellular compound APs (CAPs) from L5 ventral roots at disease peaks (clinical score = 6). The amplitudes and areas of CAPs were less affected in EAN-P2 than in EAN-PM (Fig. 8A and Table 2). As compared to EAN-P2 ventral roots, EAN-PM ventral roots presented a more striking conduction slowing, temporal dispersion and a significant shift in the refractory period (Fig. 8B and Table 2). These electrophysiological characteristics are likely the result of the greater degree of paranodal demyelination and of nodal disruption found in EAN-PM ventral roots (Figs 1 and 2
).
|
|
In normal PNS nerves, Kv channels are not implicated in AP repolarization of myelinated axons (Fig. 8C). However, the mis-distribution of Kv1 channels at nodes and paranodes in EAN-PM may likely dampen excitation and affect conduction. To evaluate the possibility that Kv1 channels participate to the reversible conduction loss in EAN nerves, we tested the effects of 4-aminopyridine (4-AP), a Kv channel blocker. If Kv1 channels are involved in reversible conduction loss, then conduction should be ameliorated by 4-AP. We observed that 4-AP had negligible effects on CAPs measured from control and EAN-P2 ventral roots (Fig. 8C). However, 4-AP significantly increased the amplitude (28%) and area (208%), but not the conduction velocity, of CAPs in the ventral roots from EAN-PM animals (Fig. 8C and D). In contrast, we did not find significant effects of XE991 (20 µM), a blocker of KCNQ channels, on conduction in both EAN-P2 and EAN-PM nerves (data not shown). Altogether, these results indicated that 4-AP improves conduction in some but not all AIDP models.
|
Discussion |
|---|
|
TopSummaryIntroductionMaterials and MethodsResultsDiscussionSupplementary materialFundingReferences |
|---|
In two models of human AIDP, we examined the molecular and electrophysiological alterations which follow inflammatory demyelination. Our results provide evidences that disorganization of nodes of Ranvier is not a common feature in EAN, but correlates with humoral response notably against neurofascin and gliomedin. In EAN-P2, autoreactive T cells and macrophages infiltrate the nerves and mediate important paranodal retraction/demyelination. However, autoantibodies are not implicated and Nav channels remain clustered at hemi-nodes, and Kv channels at juxtaparanodes. In contrast, the pathology of EAN-PM also implicates autoantibodies, notably to neurofascin and gliomedin. In the early stage of EAN-PM pathology, we found that NF186 and gliomedin are selectively affected at nodes. Then, paranodal retraction/demyelination occurs and is accompanied by the lateral diffusion of Nav and Kv channels (Supplementary Fig. 4).
Disruption of neurofascin and gliomedin in AIDP model
In agreement with a previous report, we found that nodal aggregates of Nav channels are disrupted in EAN-PM (Novakovic et al., 1998). We have confirmed and extended these results by showing that other nodal proteins (ankyrin-G, NF186, gliomedin and KCNQ2) are also impaired. More importantly, we provided significant clues about the mechanisms underlying node disruption in EAN. Our major finding indicates that neurofascin density is diminished at nodes before the onset of the clinical signs and is closely paralleled by the disappearance of gliomedin. Nodes with disrupted Nav channels and ankyrin-G appeared more lately, and lacked both gliomedin and neurofascin. Thus, Nav channel disruption in EAN is subsequent to the disappearance of adhesive molecules from nodes. Treatment of myelinating cultures with soluble neurofascin-Fc fusion protein also results in the loss of nodal clusters of gliomedin, Nav channels and ankyrin-G (Eshed et al., 2005). Because PanNF monoclonal antibodies recognized the intracellular region of neurofascin, we find it unlikely that autoantibodies in rats may be capping the proteins and thus prevent their detection by the probing antibodies.
In the early stage of the disease, many nodes lacking gliomedin and NF186 still maintained normal Nav channel clusters in EAN-PM nerves. These nodes were flanked by intact paranodes and did not show signs of demyelination, which suggests that paranodes may stabilize Nav channel aggregates in the absence of the neurofascin/gliomedin unit (Supplementary Fig. 4). In the latter stage of the disease, the disruption of the Nav and KCNQ2 clusters became apparent as demyelination rose, and correlated with the mis-localization of Kv1.2 channels. Kv1.2 channel distribution is known to be highly dependent on paranode integrity (Dupree et al., 1999; Bhat et al., 2001; Boyle et al., 2001). We, thus, conclude that Nav channel dispersion is not secondary to paranode disruption alone, but may result from the combination of axo-glial alteration at nodes and paranodal demyelination (Supplementary Fig. 4). A recent study substantiates these finding, and demonstrates that transgenic NF155 rescues the paranode and node formation in the CNS axons from neurofascin deficient mice, but only rescues paranode formation in the PNS (Zonta et al., 2008). Our results also indicate that the pathological process of AIDP models differs from that of multiple sclerosis. Indeed, Nav channel dispersion appears secondary to paranode alterations in multiple sclerosis lesions (Howell et al., 2006).
Autoantibodies to neurofascin and gliomedin in AIDP model
We showed that neurofascin and gliomedin are specifically targeted in EAN-PM. These autoantibodies recognized the native extracellular domains of NF186, NF155, and gliomedin. Autoantibodies were not detected in EAN-P2 animals, in which B-cell response is modest (Fujioka et al., 2000). Because the extracellular domains of NF186 and NF155 differ modestly (Davis et al., 1996), it is not surprising that autoantibodies recognized both isoforms. But why is NF186 selectively affected in EAN-PM nerves? One explanation could be that NF186 is more accessible to circulating antibodies than NF155. Alternatively, the dual autoimmune response against NF186 and gliomedin may potentiate neurofascin disruption at nodes. Indeed, contactin at paranodes was not targeted by autoantibodies.
We did not detect neurofascin or gliomedin in the myelin fraction used for sensitization (Supplementary Fig.1); however, we cannot exclude that autoantibodies may, in some part, arise from a primary immune reaction against trace amounts of antigens in the myelin fraction. In addition, autoantibodies may arise in vivo from a secondary immune response. Neurofascin is palmitoylated and is found with GM1 into the lipid rafts at nodes (Ren and Bennett, 1998). We, thus, speculate that immunity toward glycosphingolipids may trigger secondary immune reaction against neurofascin and its partners. Epitope spreading of autoreactive B cells has been documented in models of multiple sclerosis (Robinson et al., 2003) and may reasonably account for autoimmunity toward neurofascin and gliomedin in EAN. The fact that autoantibodies to NF155 and NF186 are also detected in sera from multiple sclerosis patients (Mathey et al., 2007) further substantiates this hypothesis. Our study with others, thus, highlights that autoantibodies to nodal proteins might contribute to disease progression in inflammatory demyelinating pathologies. Plasma exchange and intravenous immunoglobulin administration are the most widely used treatments for AIDP. Immunoglobulin administration also ameliorates clinical recovery in EAN-PM rats (Gabriel et al., 1997). In this context, autoantibodies to neurofascin and gliomedin may impair remyelination or node reformation during recovering.
Commonly, antibodies mediate their action through inflammatory cells or through the complement pathway. We did not find deposition of the terminal complement complex (C5b-9) at EAN-PM nodes at any clinical stages. Complement components of the membrane attack complex were instead detected on the surface of the Schwann cells, which is in agreement with previous reports made in AIDP patients and models (Stoll et al., 1991; Hafer-Macko et al., 1996b; Putzu et al., 2000). Thus, in contrast to AMAN, Nav cluster disruption is not mediated by complement deposition at nodes in EAN. Interestingly, Mathey et al. (2007) demonstrated that xenogeneic transfer of high doses of mouse anti-NF186 antibodies exacerbates adoptive transfer EAE in rats, notably by promoting axonal injury. In their model, anti-NF186 antibodies intensely bind the nodes and result in deposition of complement. Axonal degeneration is also detected in EAN-PM nerves; however, we did not find clear evidence of IgG deposition at nodes (data not shown). Because NF186 and gliomedin were absent or decreased at most EAN-PM nodes, we conjectured that the autoantibodies may favour their relocation or destruction, thus explaining the lack of IgG and complement deposit at nodes (Supplementary Fig. 4). Moreover, the low autoantibody titers found in EAN-PM sera may not suffice to activate the complement pathway, in contrast to acute injections of high titers of anti-NF186 antibodies. The dynamic and strength of the humoral response are important factors which may influence the physiopathology of inflammatory demyelinating diseases, and should be taken into consideration.
What other mechanisms may underlie node disorganization? As mentioned above, a direct action of the autoantibodies on neurofascin/gliomedin interaction is plausible. In pemphigus diseases, autoantibodies to desmogleins induce a breakdown of the epidermis layer in a complement-independent manner (Sitaru et al., 2007). Similarly, soluble neurofascin-Fc fusion protein induces the disruption of nodal clusters of gliomedin and Nav channels in myelinating cultures (Eshed et al., 2005). Alternatively, cell-secreted proteases may affect node structure. In particular, the matrix metalloproteinases remodel the Schwann cell basal lamina in GBS (Hughes et al., 1998) and may release gliomedin which is incorporated to the extracellular matrix (Eshed et al., 2007). In keeping, mice defective for laminin 2 present abnormal Nav channel clusters (Occhi et al., 2005). The complex physiopathology of EAN makes it difficult to privilege one mechanism rather than another. Further investigations are, therefore, required to define the pathogenicity of autoantibodies to neurofascin and gliomedin, and the mediators implicated in node disruption.
Conduction failure in inflammatory neuropathies
Electrodiagnostic is currently the most efficient way to differentiate the axonal forms of GBS (AMAN) from the demyelinating forms (AIDP) (Hughes et al., 1999). We herein describe important conduction slowing and conduction block in EAN-PM ventral roots, but modest abnormalities in EAN-P2 roots. These deficits reflected well the heterogeneity found in AIDP patients (Albers et al., 1985; Hiraga et al., 2005). Our data indicate that node disorganization may play a significant function in the physiopathology of AIDP. Indeed, axonal degeneration and demyelination were found in both AIDP models; however, prominent conduction deficits were found in EAN-PM nerves which exhibited node disruption. Lowering the density of Nav channels at nodes should logically shift the APs threshold to more depolarized potentials. We did not find significant changes in axon recruitment in EAN-PM nerves. Similarly, nerve thresholds are normal in AIDP patients (Kuwabara et al., 2002). Thus, we conclude that disrupted nodes might not conduct APs properly and might participate to conduction loss in AIDP models.
Paranodal alterations and Kv channel disorganization also participate in conduction slowing and in the abnormal refractory period in EAN-PM rats. Paranodal alterations might logically increase nodal capacitance and decrease paranodal resistance. In addition, the effects of 4-AP in demyelinated axons of EAN-PM nerves indicated that Kv channels may participate in the conduction abnormalities by shunting-down the depolarizing currents. Because Kv3.1b was not detected at EAN nodes, we suspect that 4-AP effects are mediated through the blockade of Kv1.1 and Kv1.2 channels at nodes and paranodes. 4-AP may, thus, be of some benefits for restoring conduction in AIDP patients.
|
Supplementary material |
|---|
|
TopSummaryIntroductionMaterials and MethodsResultsDiscussionSupplementary materialFundingReferences |
|---|
Supplementary material is available at Brain online.
|
Funding |
|---|
|
TopSummaryIntroductionMaterials and MethodsResultsDiscussionSupplementary materialFundingReferences |
|---|
Association Française contre les Myopathies (MNM2 2006-12180); National Multiple Sclerosis Society (RG 3839A1/T).
|
Acknowledgements |
|---|
We thank Drs Steven Scherer, Catherine Faivre-Sarrailh and José Boucraut for comments; Drs Laurence Goutebroze, Gisele Alcaraz, Alex Gow, Elior Peles and Peter Brophy for generous gift of antibodies and constructs; Dr Nadine Clerc for guinea pig tissues; and Axel Fernandez for technical assistance. The monoclonal antibodies against Nav1.2, Kv1.2 and PanNeurofascin were obtained from the UC Davis/NINDS/NIMH NeuroMab Facility, supported by NIH grant U24NS050606 and maintained by the Department of Pharmacology, School of Medicine, University of California, Davis, CA 95616.
|
References |
|---|
|
TopSummaryIntroductionMaterials and MethodsResultsDiscussionSupplementary materialFundingReferences |
|---|
Albers JW, Donofrio PD, McGonagle TK. Sequential electrodiagnostic abnormalities in acute inflammatory demyelinating polyradiculoneuropathy. Muscle Nerve (1985) 8:528–39.[CrossRef][Web of Science][Medline]
Allen D, Giannopoulos K, Gray I, Gregson N, Makowska A, Pritchard J, et al. Antibodies to peripheral nerve myelin proteins in chronic inflammatory demyelinating polyradiculoneuropathy. J Peripher Nerv Syst (2005) 10:174–80.[CrossRef][Web of Science][Medline]
Arroyo EJ, Sirkowski EE, Chitale R, Scherer SS. Acute demyelination disrupts the molecular organization of peripheral nervous system nodes. J Comp Neurol (2004) 479:424–34.[CrossRef][Web of Science][Medline]
Bhat MA, Rios JC, Lu Y, Garcia-Fresco GP, Ching W, St Martin M, et al. Axon-glia interactions and the domain organization of myelinated axons requires Neurexin IV/Caspr/Paranodin. Neuron (2001) 30:369–83.[CrossRef][Web of Science][Medline]
Bouzidi M, Tricaud N, Giraud P, Kordeli E, Caillol G, Deleuze C, et al. Interaction of the Nav1.2a subunit of the voltage-dependent sodium channel with nodal Ankyrin(G) - in vitro mapping of the interacting domains and association in synaptosomes. J Biol Chem (2002) 277:28996–9004.
Boyle MET, Berglund EO, Murai KK, Weber L, Peles E, Ranscht B. Contactin orchestrates assembly of the septate-like junctions at the paranode in myelinated peripheral nerve. Neuron (2001) 30:385–97.[CrossRef][Web of Science][Medline]
Charles P, Tait S, Faivre-Sarrailh C, Barbin G, Gunn-Moore F, Denisenko-Nehrbass N, et al. Neurofascin is a glial receptor for the paranodin/Caspr-contactin axonal complex at the axoglial junction. Curr Biol (2002) 12:217–20.[CrossRef][Web of Science][Medline]
Cooper EC, Harrington E, Jan YN, Jan LY. M channel KCNQ2 subunits are localized to key sites for control of neuronal network oscillations and synchronization in mouse brain. J Neurosci (2001) 21:9529–40.
Davis JQ, Lambert S, Bennett V. Molecular composition of the node of Ranvier: identification of ankyrin-binding cell adhesion molecules neurofascin (mucin + third FNIII domain-) and NrCAM at nodal axon segments. J Cell Biol (1996) 135:1355–67.
Devaux JJ, Kleopa KA, Cooper EC, Scherer SS. KCNQ2 is a nodal K + channel. J Neurosci (2004) 24:1236–44.
Devaux JJ, Scherer SS. Altered ion channels in an animal model of Charcot-Marie-Tooth disease type IA. J Neurosci (2005) 25:1470–80.
Dupree JL, Girault JA, Popko B. Axo-glial interactions regulate the localization of axonal paranodal proteins. J Cell Biol (1999) 147:1145–51.
Einheber S, Zanazzi G, Ching W, Scherer SS, Milner TA, Peles E, et al. The axonal membrane protein Caspr/Neurexin IV is a component of the septate-like paranodal junctions that assemble during myelination. J Cell Biol (1997) 139:1495–506.
Eshed Y, Feinberg K, Carey DJ, Peles E. Secreted gliomedin is a perinodal matrix component of peripheral nerves. J Cell Biol (2007) 177:551–62.
Eshed Y, Feinberg K, Poliak S, Sabanay H, SarigNadir O, Spiegel I, et al. Gliomedin mediates Schwann cell-axon interaction and the molecular assembly of the nodes of Ranvier. Neuron (2005) 47:215–29.[CrossRef][Web of Science][Medline]
Fujioka T, Purev E, Kremlev SG, Ventura ES, Rostami A. Flow cytometric analysis of infiltrating cells in the peripheral nerves in experimental allergic neuritis. J Neuroimmunol (2000) 108:181–91.[CrossRef][Web of Science][Medline]
Gabriel CM, Gregson NA, Hughes RA. Anti-PMP22 antibodies in patients with inflammatory neuropathy. J Neuroimmunol (2000) 104:139–46.[CrossRef][Web of Science][Medline]
Gabriel CM, Gregson NA, Redford EJ, Davies M, Smith KJ, Hughes RA. Human immunoglobulin ameliorates rat experimental autoimmune neuritis. Brain (1997) 120(Pt 9):1533–40.
Hafer-Macko C, Hsieh ST, Li CY, Ho TW, Sheikh K, Cornblath DR, et al. Acute motor axonal neuropathy: an antibody-mediated attack on axolemma. Ann Neurol (1996a) 40:635–44.[CrossRef][Web of Science][Medline]
Hafer-Macko CE, Sheikh KA, Li CY, Ho TW, Cornblath DR, McKhann GM, et al. Immune attack on the Schwann cell surface in acute inflammatory demyelinating polyneuropathy. Ann Neurol (1996b) 39:627–37.[Web of Science]
Hahn AF, Feasby TE, Steele A, Lovgren DS, Berry J. Demyelination and axonal degeneration in Lewis rat experimental allergic neuritis depend on the myelin dosage. Lab Invest (1988) 59:115–25.[Web of Science][Medline]
Hahn AF, Feasby TE, Wilkie L, Lovgren D. P2-peptide induced experimental allergic neuritis - a model to study axonal degeneration. Acta Neuropathol (1991) 82:60–5.[CrossRef][Medline]
Hiraga A, Kuwabara S, Ogawara K, Misawa S, Kanesaka T, Koga M, et al. Patterns and serial changes in electrodiagnostic abnormalities of axonal Guillain-Barre syndrome. Neurology (2005) 64:856–60.
Howell OW, Palser A, Polito A, Melrose S, Zonta B, Scheiermann C, et al. Disruption of neurofascin localization reveals early changes preceding demyelination and remyelination in multiple sclerosis. Brain (2006) 129:3173–85.
Hughes PM, Wells GM, Clements JM, Gearing AJ, Redford EJ, Davies M, et al. Matrix metalloproteinase expression during experimental autoimmune neuritis. Brain (1998) 121(Pt 3):481–94.
Hughes R, Atkinson P, Coates P, Hall S, Leibowitz S. Sural nerve biopsies in Guillain-Barre syndrome: axonal degeneration and macrophage-associated demyelination and absence of cytomegalovirus genome. Muscle Nerve (1992) 15:568–75.[CrossRef][Web of Science][Medline]
Hughes RA, Hadden RD, Gregson NA, Smith KJ. Pathogenesis of Guillain-Barre syndrome. J Neuroimmunol (1999) 100:74–97.[CrossRef][Web of Science][Medline]
Jenkins SM, Bennett V. Developing nodes of Ranvier are defined by ankyrin-G clustering and are independent of paranodal axoglial adhesion. Proc Natl Acad Sci USA (2002) 99:2303–8.
Kadlubowski M, Hughes RA. Identification of the neuritogen for experimental allergic neuritis. Nature (1979) 277:140–1.[CrossRef][Web of Science][Medline]
Kuwabara S, Ogawara K, Sung JY, Mori M, Kanai K, Hattori T, et al. Differences in membrane properties of axonal and demyelinating Guillain-Barre syndromes. Ann Neurol (2002) 52:180–7.[CrossRef][Web of Science][Medline]
Kwa MSG, vanSchaik IN, Brand A, Baas F, Vermeulen M. Investigation of serum response to PMP22, connexin 32 and P-0 in inflammatory neuropathies. J Neuroimmunol (2001) 116:220–5.[CrossRef][Web of Science][Medline]
Laemmli U. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (1970) 227:680–5.[CrossRef][Web of Science][Medline]
Mathey EK, Derfuss T, Storch MK, Williams KR, Hales K, Woolley DR, et al. Neurofascin as a novel target for autoantibody-mediated axonal injury. J Exp Med (2007) 204:2363–72.
Menegoz M, Gaspar P, Le Bert M, Galvez T, Burgaya F, Palfrey C, et al. Paranodin, a glycoprotein of neuronal paranodal membranes. Neuron (1997) 19:319–31.[CrossRef][Web of Science][Medline]
Mi HY, Deerinck TJ, Ellisman MH, Schwarz TL. Differential distribution of closely related potassium channels in rat Schwann cells. J. Neurosci. (1995) 15:3761–74.[Abstract]
Norton WT, Poduslo SE. Myelination in rat brain. Method of myelin isolation. J Neurochem (1973) 21:749–58.[Web of Science][Medline]
Novakovic SD, Levinson R, Schachner M, Shrager P. Disruption and reorganization of sodium channels in experimental allergic neuritis. Muscle Nerve (1998) 21:1019–32.[CrossRef][Web of Science][Medline]
Occhi S, Zambroni D, DelCarro U, Amadio S, Sirkowski EE, Scherer SS, et al. Both laminin and Schwann cell dystroglycan are necessary for proper clustering of sodium channels at nodes of Ranvier. J Neurosci (2005) 25:9418–27.
OHanlon GM, Humphreys PD, Goldman RS, Halstead SK, Bullens RWM, Plomp JJ, et al. Calpain inhibitors protect against axonal degeneration in a model of anti-ganglioside antibody-mediated motor nerve terminal injury. Brain (2003) 126:2497–509.
Olee T, Powell HC, Brostoff SW. New minimum length requirement for a T cell epitope for experimental allergic neuritis. J Neuroimmunol (1990) 27:187–90.[CrossRef][Web of Science][Medline]
Pan ZM, Kao TC, Horvath Z, Lemos J, Sul JY, Cranstoun SD, et al. A common ankyrin-G-based mechanism retains KCNQ and Na-V channels at electrically active domains of the axon. J Neurosci (2006) 26:2599–613.
Paparounas K, O’Hanlon GM, O’Leary CP, Rowan EG, Willison HJ. Anti-ganglioside antibodies can bind peripheral nerve nodes of Ranvier and activate the complement cascade without inducing acute conduction block in vitro. Brain (1999) 122(Pt 5):807–16.
Putzu GA, Figarella-Branger D, Bouvier-Labit C, Liprandi A, Bianco N, Pellissier JF. Immunohistochemical localization of cytokines, C5b-9 and ICAM-1 in peripheral nerve of Guillain-Barre Syndrome. J Neurol Sci (2000) 174:16–21.[CrossRef][Web of Science][Medline]
Ren Q, Bennett V. Palmitoylation of neurofascin at a site in the membrane-spanning domain highly conserved among the L1 family of cell adhesion molecules. J Neurochem (1998) 70:1839–49.[Web of Science][Medline]
Rios JC, MelandezVasquez CV, Einheber S, Lustig M, Grumet M, Hemperly J, et al. Contactin-associated protein (Caspr) and contactin form a complex that is targeted to the paranodal junctions during myelination. J Neurosci (2000) 20:8354–64.
Robinson WH, Fontoura P, Lee BJ, de Vegvar HE, Tom J, Pedotti R, et al. Protein microarrays guide tolerizing DNA vaccine treatment of autoimmune encephalomyelitis. Nat Biotechnol (2003) 21:1033–9.[CrossRef][Web of Science][Medline]
Schmidt B, Toyka KV, Kiefer R, Full J, Hartung HP, Pollard J. Inflammatory infiltrates in sural nerve biopsies in Guillain-Barre syndrome and chronic inflammatory demyelinating neuropathy. Muscle Nerve (1996) 19:474–87.[CrossRef][Web of Science][Medline]
Sherman DL, Tait S, Melrose S, Johnson R, Zonta B, Court FA, et al. Neurofascins are required to establish axonal domains for saltatory conduction. Neuron (2005) 48:737–42.[CrossRef][Web of Science][Medline]
Sitaru C, Mihai S, Zillikens D. The relevance of the IgG subclass of autoantibodies for blister induction in autoimmune bullous skin diseases. Arch Dermatol Res (2007) 299:1–8.[CrossRef][Web of Science][Medline]
Southwood C, He C, Garbern J, Kamholz J, Arroyo E, Gow A. CNS myelin paranodes require Nkx6-2 homeoprotein transcriptional activity for normal structure. J Neurosci (2004) 24:11215–25.
Stoll G, Schmidt B, Jander S, Toyka KV, Hartung HP. Presence of the terminal complement complex (C5b-9) precedes myelin degradation in immune-mediated demyelination of the rat peripheral nervous system. Ann Neurol (1991) 30:147–55.[CrossRef][Web of Science][Medline]
Susuki K, Rasband MN, Tohyama K, Koibuchi K, Okamoto S, Funakoshi K, et al. Anti-GM1 antibodies cause complement-mediated disruption of sodium channel clusters in peripheral motor nerve fibers. J Neurosci (2007) 27:3956–67.
Tait S, Gunn-Moore F, Collinson JM, Huang J, Lubetzki C, Pedraza L, et al. An oligodendrocyte cell adhesion molecule at the site of assembly of the paranodal axo-glial junction. J Cell Biol (2000) 150:657–66.
Uyemura K, Suzuki M, Kitamura K, Horie K, Ogawa Y, Matsuyama H, et al. Neuritogenic determinant of bovine P2 protein in peripheral nerve myelin. J Neurochem (1982) 39:895–8.[CrossRef][Web of Science][Medline]
Yuki N, Koga M. Bacterial infections in Guillain-Barre and Fisher syndromes. Curr Opin Neurol (2006) 19:451–7.[Web of Science][Medline]
Zhu J, Link H, Weerth S, Linington C, Mix E, Qiao J. The B cell repertoire in experimental allergic neuritis involves multiple myelin proteins and GM1. J Neurol Sci (1994) 125:132–7.[CrossRef][Web of Science][Medline]
Zonta B, Tait S, Melrose S, Anderson H, Harroch S, Higginson J, et al. Glial and neuronal isoforms of Neurofascin have distinct roles in the assembly of nodes of Ranvier in the central nervous system. J Cell Biol (2008) 181:1169–77.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  M. Iijima, M. Tomita, S. Morozumi, Y. Kawagashira, T. Nakamura, H. Koike, M. Katsuno, N. Hattori, F. Tanaka, M. Yamamoto, et al. Single nucleotide polymorphism of TAG-1 influences IVIg responsiveness of Japanese patients with CIDP Neurology, October 27, 2009; 73(17): 1348 - 1352. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||