Nodal proteins are target antigens in Guillain-Barré syndrome

Sera

Patients with GBS (n = 100) and CIDP (n = 50) were referred to the neuroimmunological laboratory at Dokkyo Medical University (Tochigi, Japan) from hospitals throughout Japan for serum anti-ganglioside antibody testing. The Ethical Committee of Dokkyo Medical University approved the study. Written informed consent was obtained from all patients. Sera from GBS patients were collected within 3 weeks of the onset of neurologic symptoms and before the administration of any treatments. Information on patients regarding the age, antecedent illnesses, initial symptoms, and neurologic signs during the course of the illness was obtained. GBS patients were then classified as having AMAN (n = 50) or AIDP (n = 50) according to the electrodiagnostic criteria (Ho et al., 1995). CIDP patients were classified in accordance with the diagnostic criteria published by the American Academy of Neurology AIDS Task Force (1991). Patients' disabilities at nadir were evaluated using the Hughes functional grading scale (Hughes et al., 1978). Serum IgG and IgM antibodies to GM1, GM1b, GD1a, GalNAc-GD1a, GD1b, GT1a, and GQ1b were measured by ELISA as described previously (Odaka et al., 1998). Normal control (NC) sera were obtained from 50 healthy volunteers. Patients with other neurological disorders (OND; n = 80) included patients with acute disseminated encephalomyelitis (n = 20), multiple sclerosis (n = 20), myasthenia gravis (n = 20), and amyotrophic lateral sclerosis (n = 20).

Constructs

Rat gliomedin (NM_181382.2) was amplified by PCR from a rat sciatic nerve cDNA library and the PCR product was sub-cloned into the mammalian expression vector pcDNA3.1 (Invitrogen, Paisley, UK) at KpnI and EcoRV sites. Myc epitope was then inserted at the C-terminus of the protein by site-directed mutagenesis. Full length rat NF186 (NM_001160314.1; HA-tagged), rat contactin (NM_057118), rat neuronal cell adhesion molecule (NrCAM) (NM_013150; HA-tagged) were the kind gift of Drs. Vann Bennett (Duke University Medical Center, Durham, NC), Elior Peles (The Weizmann Institute of Science, Rehovot, Israel), and Catherine Faivre-Sarrailh (CNRS-UMR6231, Marseille, France).

Cell binding assay

Human embryonic kidney (HEK) cells were transiently transfected with NF186, gliomedin, contactin, or NrCAM using JetPEI (Polyplus-transfection, Illkirch, France). Using our protocol, about 60% of the cells expressed NF186, gliomedin, contactin, or NrCAM. One day after transfection, cells were trypsinated, suspended in serum free Opti-MEM medium (Invitrogen), and plated onto poly-l-lysine coated glass coverslips in 24-well plates at a density of 100,000 cells/well. After 1 day, living cells were incubated for 20 min with 50 µl of serum diluted at 1/20 to 1/5,000 in blocking solution with FITC conjugated anti-human IgM (1/50) or TRITC conjugated anti-human IgG (1/50). Cells were washed thrice in PBS, fixed, and blocked as described above. Cells were then incubated for 1 h with primary antibodies - rat monoclonal antibodies against hemagglutinin (HA; 1/200; Roche, Basel, Switzerland), mouse monoclonal antibodies against Myc (1/500; Roche), or a goat antiserum against contactin (1/2,000; R&D Systems, Minneapolis, MN, USA). The cells were then washed and revealed with the appropriate Alexa conjugated secondary antibodies (1/500; Invitrogen). Cells were stained with DAPI, and mounted with Mowiol plus 2% DABCO (Sigma-Aldrich, St. Louis, MO, USA). Patients with serum IgG or IgM that specifically co-localized with NF186, gliomedin, contactin, or NrCAM at cell surface were considered positive. Patients with serum IgG or IgM that bound to non-transfected cells were considered negative. For competition experiments, the extracellular domain of NF186 (NF186-Fc), gliomedin (gliomedin-Fc), and contactin (contactin-Fc) fused to human Fc were obtained as described previously (Charles et al., 2002; Eshed et al., 2005). Sera were pre-incubated for 30 min at room temperature with the Fc fusion proteins (0.5 µg) before incubation with living HEK cells.

Nerve staining

Teased fibers from sciatic nerves of adult C57BL/6J mice and adult Wistar rats were prepared as previously described (Lonigro and Devaux, 2009). Teased fibers were permeabilized by immersion in −20°C acetone for 10 min, blocked for 1 h in blocking solution containing 0.1% Triton X-100 and incubated overnight at 4°C with sera diluted at 1/200 to 1/5,000. In some experiments, tissues were double-stained with mouse antibodies against Nav channels (1/500; Sigma-Aldrich) or a rabbit antiserum against Caspr (1/3,000) (Menegoz et al., 1997). The slides were washed and incubated with the appropriate Alexa conjugated secondary antibodies (1/500; Invitrogen). Slides were mounted and examined using an ApoTome fluorescence microscope (ApoTome, AxioObserver and AxioCam MRm, Carl Zeiss MicroImaging GmbH, Jena, Germany). Digital images were manipulated into figures with Adobe Photoshop. For competition experiments, sera were pre-incubated for 30 min at room temperature with the Fc fusion proteins (0.5 µg) before immunolabeling.

Statistical analysis

Differences in group frequencies were compared using GraphPad Prism by chi-square test with Yate's correction or by Fisher's exact test with Bonferroni's adjustment (GraphPad Software, La Jolla, CA, USA). p Values inferior to 0.05 were considered significant. Correlation between the presence of antibodies to gangliosides and the presence of antibodies to NF186/gliomedin/contactin/NrCAM was examined by Spearman's rank correlation test.

GBS and CIDP sera target the nodes of Ranvier

First, we investigated whether IgG from GBS and CIDP patients may bind the nodes of Ranvier or the flanking paranodes. Because of the high background observed when using sera on rat nerves or at a 1/50 dilution on mouse nerves, we tested the GBS, CIDP, and control sera at a 1/200 dilution on teased sciatic nerves from mice. We found that many sera from GBS and CIDP patients labeled the nodal and paranodal regions (Table 1), and particularly co-localized with Nav channels at nodes or with Caspr, a paranodal adhesion molecule (1, 4). Nodes were more frequently stained than the paranodes, and 22% of GBS sera stained both structures. IgG that bound to the nodes were detected at a similar frequency in AIDP and AMAN (34%). Eight AIDP patients showed IgG binding restricted to the nodal region, compared to four AMAN patients (Fig. 1A and 1B). Many patients' sera still labeled the nodal or paranodal regions at a 1/500 and a 1/2,000 dilution. In non-permeabilized tissue, these sera stained the nodes and microvilli (data not shown), suggesting they target extracellular epitopes. By comparison, no sera from healthy controls stained nodes or paranodes at a 1/200 dilution (Table 1). Also, only 3 patients with OND (n = 80), which included immune mediated neurological disorders (acute disseminated encephalomyelitis, multiple sclerosis, and myasthenia gravis), showed IgG binding to the nodal region at a 1/200 dilution. This suggested that nodal staining was specific and not simply due to a rise in immunoglobulin levels. Our results thus emphasized that a high percentage of AIDP (44%), AMAN (42%), and CIDP (30%) patients showed antibodies to nodal antigens. Worth noting, the profiles of IgG deposition were reminiscent of the localization of cell adhesion molecules at nodes and paranodes, suggesting that autoantibodies target the axoglial junctions.

NF186, gliomedin, and contactin are target antigens in GBS patients

Recent genetic investigations identified the constituents of the axoglial junctions at nodes and paranodes (Bhat et al., 2001; Boyle et al., 2001; Charles et al., 2002; Eshed et al., 2005; Zonta et al., 2008; Feinberg et al., 2010; Thaxton et al., 2011). Because these proteins play a central role in saltatory conduction, we examined whether GBS and CIDP patients show antibodies targeting NF186, gliomedin, contactin, or NrCAM. In order to quantify exactly the serum titers, we tested serial dilutions of the sera by immunofluorescence on living cells transfected with these adhesion molecules. This procedure enables staining of cell surface antigens as in flow cytometry analyses, and has the advantage to allow convenient screening of large cohorts of patients. In addition, we found that this technique was more reliable than ELISA to discriminate true antibody reaction against cell adhesion molecules. We found that many sera presented IgG or IgM that bound selectively to NF186, contactin, or gliomedin (Table 2) and co-localized with these at cell surface but did not stain untransfected cells (Fig. 2). The data in Table 2 represent the proportion of patients that complied with these criteria. Sera that stained untransfected cells were considered negative. AIDP patients presented a significantly higher prevalence of antibodies directed against gliomedin (Table 2; p < 0.05; χ² test with Yate's correction or Fisher's exact test with Bonferroni's adjustment) compared to healthy controls and OND. By contrast, antibodies against NF186 were more frequently detected in AMAN patients. The prevalence of antibodies against NrCAM was not significantly increased in any neurological disorders. These results indicated that gliomedin and NF186 are autoantigens in GBS patients. Of interest, many patients showed autoantibodies recognizing a single antigen (Table 2). When examined regardless of the targeted antigen or immunoglobulin class, it appeared that 26% of GBS and 24% of CIDP patients showed autoantibodies to nodal adhesion molecules with titers ranging from 1/50 to 1/1,000. In particular, the titers of anti-NF186 antibodies were higher in AMAN patients (mean titer around 1/200) compared to anti-gliomedin or anti-contactin antibodies (mean titer around 1/50). Only 2 of 50 control sera reacted against NF186, contactin, or NrCAM. Patients with OND or healthy controls did not present an increased humoral response against the tested proteins. The cell binding assay thus suggested that the humoral reactivity toward nodal adhesion molecules was more prevalent in GBS and CIDP.

To confirm the specificity of IgG and IgM binding, positive sera were pre-incubated with the soluble extracellular domains of contactin, gliomedin, or NF186. The binding of IgG and IgM to the surface of transfected cells was completely antagonized by the soluble antigens (Fig. 2), corroborating that nodal adhesion molecules are the targets of antibodies in GBS and CIDP patients.

Autoantibodies from GBS patients target adhesion molecules in peripheral nerves

As indicated in Fig. 1 and Table 1, we found that 58 patients with GBS or CIDP showed IgG deposition at nodes of Ranvier and paranodes. Using cell binding assay, we found that eight of these patients exhibited IgG to NF186, gliomedin, or contactin (Table 2). In addition, 30 patients showed IgM to NF186, gliomedin, contactin, or NrCAM (Table 2). We thus reinvestigated more closely whether autoantibodies against NF186, gliomedin, or contactin may be responsible for IgG deposition. Also, we examined whether serum IgM from positive GBS and CIDP patients bound to the nodal/paranodal regions. We did not detect IgM deposition at nodes or paranodes with any of the tested sera, even at a 1/50 dilution. By contrast, the sera from the eight patients with IgG to NF186, gliomedin, or contactin stained the nodal and/or paranodal regions in permeabilized sciatic nerve fibers (Fig. 3A–D). Both sensory and motor axons were labeled with the sera, but also Remak bundles. In particular, two sera with anti-NF186 IgG stained specifically the nodes and co-localized with Nav channels. Four sera stained the nodes and paranodal junctions, these showed anti-NF186 IgG (Fig. 3) or IgG to multiple antigens. Finally, two sera with anti-contactin IgG labeled specifically the paranodal junctions (Fig. 3A). In order to test the specificity of these autoantibodies, we incubated the sera with the soluble extracellular domains of contactin or NF186. The staining of the nodal or paranodal regions was abolished after incubation of the sera with the soluble antigens (Fig. 3B and 3D), indicating they specifically target these nodal and paranodal adhesion molecules. Nerve fibers were colabeled for Caspr, a marker of paranodal junctions. Caspr staining was not affected by the soluble antigens, confirming the specificity of patients' autoantibodies. Altogether, this indicated that NF186, gliomedin, and contactin are target autoantigens in nerve tissues. Importantly, among the 58 patients with GBS or CIDP that showed IgG fixation at nodes and paranodes, 50 patients did not exhibit IgG to NF186, gliomedin, or contactin. Figure 1 shows labeling obtained with these patients' sera. These results emphasized that autoantibodies may target a variety of nodal antigens. These may be adhesion molecules or other antigens.

Clinical features of GBS and CIDP patients reactive to nodal adhesion molecules

The clinical features of the GBS patients with antibodies to gliomedin, contactin, NF186, and NrCAM are shown in Table 3. A higher proportion of AMAN patients with antibodies to nodal adhesion molecules were older than 60 years (58% vs. 18% in patients lacking antibodies to nodal adhesion molecules). Also, a higher percentage of these patients presented respiratory disturbances. These disturbances were not related to the age of the patients and were observed in both young and elderly patients. No correlation was found between antecedent illness and the presence of antibodies to nodal adhesion molecules. However, we observed that antibodies to NF186/gliomedin/contactin/NrCAM were significantly correlated with the presence of IgM against GM1b, GD1a, and GalNAc-GD1a in AMAN patients (Fig. 4; p < 0.05 using a Spearman correlation test). In AIDP patients, the clinical features were not significantly different between patient groups (Table 3).