Antibodies to native myelin oligodendrocyte glycoprotein in children with inflammatory demyelinating central nervous system disease.код для вставкиСкачать
Antibodies to Native Myelin Oligodendrocyte Glycoprotein in Children with Inflammatory Demyelinating Central Nervous System Disease Fabienne Brilot, PhD,1 Russell C. Dale, PhD,1 Rebecca C. Selter,2 Verena Grummel,2 Sudhakar Reddy Kalluri, MSc,2 Muhammad Aslam, M Phil,2 Verena Busch, MD,2,3 Dun Zhou, PhD,2 Sabine Cepok, PhD,2 and Bernhard Hemmer, MD2 Objective: Myelin oligodendrocyte glycoprotein (MOG) is a candidate target antigen in demyelinating diseases of the central nervous system (CNS). Although MOG is encephalitogenic in different animal models, the relevance of this antigen in human autoimmune diseases of the CNS is still controversial. Methods: We investigated the occurrence and biological activity of antibodies to native MOG (nMOG) in 47 children during a first episode of CNS demyelination (acute disseminated encephalomyelitis [ADEM], n ⫽ 19 and clinical isolated syndrome [CIS], n ⫽ 28) by a cell-based bioassay. Results: High serum immunoglobulin G (IgG) titers to nMOG were detected in 40% of children with CIS/ADEM but 0% of the control children affected by other neurological diseases, healthy children, or adults with inflammatory demyelinating diseases, respectively. By contrast, IgM antibodies to nMOG occurred in only 3 children affected by ADEM. Children with high anti-nMOG IgG titer were significantly younger than those with low IgG titer. Anti-nMOG IgG titers did not differ between the ADEM and CIS group, and did not predict conversion from CIS to MS during a mean 2-year follow-up. However, intrathecal IgG anti-MOG antibody synthesis was only seen in CIS children. IgG antibodies to nMOG not only bound to the extracellular domain of nMOG, but also induced natural killer cell-mediated killing of nMOG-expressing cells in vitro. Interpretation: Overall, these findings suggest nMOG as a major target of the humoral immune response in a subgroup of children affected by inflammatory demyelinating diseases of the CNS. Children may provide valuable insight into the earliest immune mechanisms of CNS demyelination. Ann Neurol 2009;66:833– 842 Demyelinating diseases of the central nervous system (CNS) are an important cause of neurological disability in children and young adults. A first CNS demyelinating event in adults is most likely to represent multiple sclerosis (MS). A first demyelinating event in children when accompanied by encephalopathy is termed acute disseminated encephalomyelitis (ADEM) and has a lower risk of progression to MS.1 The other first CNS demyelinating event in children is a clinically isolated syndrome (CIS), which has a higher risk of progression to MS.2–5 A recent study testing new international criteria of pediatric demyelination has found that ADEM has a significantly lower progression to MS than CIS after 2-year follow-up.5 Although the presence of encephalopathy makes the development of MS less likely, it has been recognized that very young children can have encephalopathy during MS presentation.6 Although the cause of demyelinating disease is largely unknown, several lines of evidence suggest an underlying autoimmune process.7,8 Myelin antigens, such as myelin oligodendrocyte glycoprotein (MOG) and proteolipid protein (PLP), have been proposed as targets of the immune response in MS and ADEM.9 –11 MOG is expressed on the surface of oligodendrocytes and readily accessible to autoantibodies. Immunization with MOG induces experimental autoimmune encephalo- From the 1Neuroimmunology Group, Institute for Neuroscience and Muscle Research, the Kids Research Institute at the Children’s Hospital at Westmead, University of Sydney, Sydney, Australia; 2 Department of Neurology, Klinikum rechts der Isar, Technical University Munich, Munich, Germany; and 3Children Hospital, Technical University Munich, Munich, Germany. Potential conflict of interest: Nothing to report. Address correspondence to Dr Hemmer, Department of Neurology, Klinikum rechts der Isar, Technische Universität München, Ismaninger Strasse 22, 81675 Munich, Germany. Received Jun 29, 2009, and in revised form Oct 25. Accepted for publication Oct 30, 2009. Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/ana.21916 © 2009 American Neurological Association 833 myelitis in several animal species.11 In the rat and marmoset model, an antibody response to MOG is observed, which contributes to CNS demyelination.12,13 These antibodies target a conformational epitope of MOG, which is located on the extracellular domain of the protein.14 In humans, immunoglobulin (Ig) G and IgM antibodies to denatured MOG protein have been detected in sera of adult MS patients and controls, although serum levels are rather low and do not seem to correlate with disease progression in patients with CIS.15–17 More recently, cell-based assays have been developed, with MOG being properly expressed on the membrane, which enable quantification of antibodies to native human MOG (nMOG).18 –20 Although IgG antibodies to nMOG are elevated in MS patient sera compared with controls, only a small number of adult MS patients show high antibody titer.18 In this study, we used cell-based bioassays to determine IgG and IgM antibodies to native MOG, PLP, and aquaporin-4 (AQP4), 3 putative autoantigens in demyelinating CNS diseases. We focused on a welldefined group of children with a first inflammatory demyelinating event who were prospectively followed for a mean of 2 years to monitor ongoing disease activity and conversion to MS. Antibody titers were compared to control groups. We observed a unique antibody response in a subgroup of children with a first demyelinating event. (LN18-PLP) and AQP4 (LN18-AQP4) were established using the same strategy.21 Subjects and Methods Patients and Controls The following bioassay was used to quantify antibody reactivity of sera and cerebrospinal fluid (CSF) to native MOG, PLP, and AQP4. Serum was added to 30,000 LN18-MOG, LN18-AQP4, LN18-PLP, or LN18-CTR cells in RPMI1640, yielding a final serum dilution of 1:100. Cells were incubated on ice on an orbital shaker for 20 minutes and washed twice with washing buffer (1% fetal calf serum [FCS] in phosphate-buffered saline [PBS]). Cells were then stained with Alexa Fluor 488-labeled goat anti-human IgG or IgM secondary antibody (Invitrogen) for 20 minutes on ice, washed again twice, and then resuspended in washing buffer. Analysis of cell surface staining was determined by flow cytometry using CyAn ADP (Beckman Coulter, Fullerton, CA) and Summit software (Beckman Coulter). Levels of antibody titers are expressed by ⌬-median fluorescence intensity (MFI). ⌬MFI was determined by the subtraction of MFI obtained with LN18-CTR cells from the MFI obtained with LN18-MOG, LN18-AQP4, or LN18-PLP cells. To detect anti-nMOG antibodies in CSF, IgG concentration of sera and corresponding CSF were measured by nephelometry (BN ProSpec, Siemens, Erlangen, Germany). Both sera and CSF were adjusted to a final IgG concentration of 5mg/l, and the reactivity to nMOG was determined by flow cytometry as described above. All pediatric patients and controls were recruited in the UK. Adult MS patients were recruited at the neurology departments in Düsseldorf and Munich. Ethics approval for this pediatric study was granted by both the Children’s Hospital at Westmead, Australia, and the Institute of Child Health, Great Ormond Street, NHS Trust, UK ethics committee. Ethics approval for the adult study was obtained by the local ethics committees in Düsseldorf and Munich. Cloning and Expression of Human CNS Proteins The bioassay was established as described previously.18 Briefly, full-length human MOG cDNA, synthesized out of a human brain total RNA (BD Biosciences, San Jose, CA) was used to transduce a human glioblastoma cell line LN18. The primers 5⬘-ATTGAGATCTGAGATGGCAAG-3⬘ and 5⬘-GAGATCTCAGAAGGGATTTCG-3⬘ were used to add BglII restriction sites at 5⬘ and 3⬘ ends of the MOG cDNA. The polymerase chain reaction product was then cloned into the plasmid pLenti6/V5 (Invitrogen, Carlsbad, CA), and a 293FT cell line was transfected by Lipofectamine (Invitrogen) using the pLenti6/V5-MOG. Virus-containing supernatant was used to transduce the LN18 glioblastoma cell line (LN18-MOG). A control cell line (LN18-CTR) was obtained by transducing the LN18 cell line with an empty vector pLenti6/V5. Additional cell lines expressing human PLP 834 Annals of Neurology Vol 66 No 6 December 2009 Assessment of Surface Expression of Human CNS Native Proteins Surface expression of MOG and PLP on transduced LN18 cells was determined after staining with a monoclonal antiMOG antibody (mAb 8-18C5), or anti-PLP antibody (mAb O10) in combination with an Alexa Fluor 488-conjugated goat antimouse IgG antibody (Invitrogen) as secondary antibody. AQP4 expression was confirmed by intracellular staining using polyclonal rabbit anti-AQP4 antibody (1mg/ml, Sigma, St. Louis, MO) and Alexa-488-labeled anti-rabbit (Molecular Probes, Eugene, OR; Invitrogen) (data not shown, Reddy et al, submitted). Cell surface staining was also analyzed by flow cytometry. Immunocytostaining was performed on 3% paraformaldehyde-fixed cells using described protocols.18 Sensitivity and reproducibility of the assay were evaluated by titration experiments using the mAb 8-18C5, mAb O10, or rabbit anti-AQP4 –specific polyclonal antibody, respectively. For titration experiments, transduced LN18 or LN18CTR cells were incubated with diluted mAbs (10 dilutions from 0.001 to 1mg/ml) or anti-AQP4 –specific polyclonal antibody (10 dilutions from 0.02 to 10g/ml), followed by staining with secondary antibodies as described above. Experiments were done in duplicates. Nonspecific binding and polyreactivity of anti-MOG–positive sera to cell surface proteins were excluded by using LN18-CTR. Cell-Based Assay for Quantification of Antibody Reactivity in Sera and Cerebrospinal Fluid In Vitro Antibody Dependent Cellular Cytotoxicity Assay IgG was purified from sera using Protein G-Sepharose (GE Healthcare, Piscataway, NJ) according to the manufacturer’s protocol. The titer of anti-nMOG antibodies in purified IgG was determined by the cell-based assay as described above. Peripheral blood mononuclear cells of healthy donors were separated by density gradient centrifugation (Biocoll Separating Solution, Biochrom, Berlin, Germany). Natural killer (NK) cells were stained with bead-conjugated anti-CD56 antibody (Miltenyi Biotec, Bergisch Gladbach, Germany) according to the manufacturer’s instructions. CD56⫹ cells were then separated by positive selection using an AutoMACS (Miltenyi Biotec). Purity of ⬎95% NK cells was obtained. Six micrograms of purified IgG diluted in RPMI-1640 was added to 30,000 LN18-CTR or LN18-MOG cells and incubated on ice for 25 minutes under agitation. Cells were then washed twice with fluorescent-activated cell sorter analysis (FACS) buffer (1% FCS in PBS) and once with RPMI1640, and were added to 60,000 CD56⫹ NK cells resuspended in RPMI-1640 in a final volume of 150l. After incubation at 37°C in a humidified CO2 incubator for 12 hours, the supernatant was collected, and remaining cells were detached by trypsin-ethylenediamine tetraacetic acid. Cell number and viability were then analyzed using FACSbased cell counting (CyAn ADP, Beckman Coulter). Table. Clinical Diagnoses of Patients with First Episode of CNS Demyelination (n ⴝ 47), Mean Follow-up Duration, and Progression to MS First CNS Demyelination Episode Diagnosis ADEM, n ⴝ 19 CIS, n ⴝ 28 Sex; mean age, yr (range) 13 boys; 5.2 (1.8–10.5) 12 boys; 9.2 (2.5–15.5) 0/19 8/28a Mean duration of follow-up, yr (range) 1.96 (0.5–4.6) 2.02 (1.0–4.0) Progression to MSb 1/18 12/28 Intrathecal OCB at onset a Six of 8 CIS patients with intrathecal OCB progressed to MS in follow-up. b Diagnosis based on clinical relapse. CNS ⫽ central nervous system; MS ⫽ multiple sclerosis; ADEM ⫽ acute disseminated encephalomyelitis; CIS ⫽ clinically isolated syndrome; OCB ⫽ oligoclonal immunoglobulin G bands. Statistical Analysis Due to the fact the data were not normally distributed, the Mann-Whitney U test and Kruskal-Wallis test were used to compare antibody titer between patients and controls. For correlation analysis between anti-nMOG titer and cytotoxicity, the Spearman rank correlation was applied. Results Patients with First CNS Demyelinating Episode Forty-seven children (25 boys; mean age, 7.63 years; range, 1.8 –15.5 years) were recruited with a first episode of acute onset CNS demyelination between 2002 and 2008 (Table). Patients were retrospectively defined using 2007 International consensus definitions for a first episode of CNS demyelination in children.22 Pediatric ADEM (n ⫽ 19) was defined as a first clinical event of inflammatory demyelination with acute or subacute onset that affected multifocal areas of the CNS. The event was polysymptomatic and included encephalopathy (defined as behavioral change or change in consciousness). Magnetic resonance imaging (MRI) showed focal or multifocal lesions predominantly involving the white matter. Pediatric CIS (n ⫽ 28) was defined as a first event of inflammatory demyelination that was either focal (optic neuritis, n ⫽ 7; hemispheric syndrome, n ⫽ 6; brainstem syndrome, n ⫽ 2; cerebellar syndrome, n ⫽ 2; transverse myelitis, n ⫽ 1) or multifocal (n ⫽ 10), but did not include encephalopathy. The pediatric CIS classification was based on clinical, rather than radiological features.22 Investigation and Management of First CNS Demyelinating Episode All serum samples from children with pediatric ADEM or CIS (n ⫽ 47) were taken within the first 3 weeks of neurological onset (mean, 7.8 days; median, 7 days); 15 presented fulminantly and were sampled within 3 days, 29 presented subacutely and were sampled within 7 days, and 7 presented indolently and were sampled within 3 weeks. All serum and CSF samples were taken at the same time, and before treatment. Patients were routinely tested for CSF/serum oligoclonal bands (see Table). MRI demonstrated typical inflammatory demyelinating lesions disseminated throughout the CNS, but predominantly involving white matter: Forty-five of 47 had supratentorial white matter lesions (1 had isolated radiological optic neuritis; 1 had isolated radiological transverse myelitis). The majority of patients (43/47) received steroid treatment (intravenous methyl-prednisolone 30mg/kg for 3 days, then oral prednisolone over 4 weeks in the case of incomplete recovery). The 4 pediatric patients who did not receive steroid therapy had CIS, and were spontaneously improving at the time of assessment. Three children received 2g/kg body weight of intravenous immunoglobulin due to disease deterioration despite steroid therapy. Follow-up and Multiple Sclerosis Patients were followed-up for a mean duration of 2.0 years (range, 0.5– 4.6 years). The patient classification of the first demyelinating episode was pediatric ADEM Brilot et al: MOG Antibodies in Children 835 or CIS. A child was reclassified as having MS based on the presence of clinical relapse, rather than radiological evidence of new lesions.22,23 Pediatric MS diagnosis required multiple episodes of CNS demyelination separated in time and space, as for adults.22 A relapse involved a new site and occurred ⬎4 weeks after the first event. All patients had repeat magnetic resonance neuroimaging that demonstrated new lesions. During that time, 13 patients relapsed and fulfilled a diagnosis of pediatric MS. Pediatric CIS had a higher risk of relapse and progression to MS.4 Only 1 patient with pediatric ADEM had multiple nonADEM relapses and was diagnosed with MS (see Table). The MS-defining relapse occurred a mean of 10 months and median of 7.5 months after the first event. Ten of 13 pediatric MS patients had their MS-defining relapse within the first year (range, 1– 42 months). Eight of 13 pediatric MS patients are taking betainterferon disease-modifying therapy. Controls The pediatric control group included children with other noninflammatory neurological diseases (ONDs) (n ⫽ 28; 16 boys; mean age, 6.0 years; range, 0.3–14 years), such as developmental delay (n ⫽ 8), CNS malformations (n ⫽ 5), neurodegeneration (n ⫽ 8), and epilepsy syndromes (n ⫽ 7). Serum from communityacquired healthy children (HC) (n ⫽ 30; 15 boys; mean age, 11.0 years; range, 9 –13 years) and serum from type 1 diabetes (n ⫽ 15; mean age, 10.9 years; range, 2–16 years) were also used in this study. The pediatric control groups were not age-matched due to the difficulty in obtaining blood samples in healthy young children. However, the pediatric OND group included children with a broad age range, including very young children. The mean, median, and age range of the pediatric OND group were not significantly different from our patient group. The adult MS patient group consisted of 54 patients. One patient had CIS, 33 were relapsing-remitting, 1 was primary, and 19 were secondary progressive MS (mean age, 40 years). High Antibody Titer to Native MOG in Children with a First Episode of CNS Inflammatory Demyelination We studied the presence of autoantibodies to native PLP, AQP4, and MOG in children affected by a first inflammatory demyelinating event compatible with childhood ADEM or CIS. We used a sensitive bioassay for antibody detection based on a glial tumor cell line transduced with full-length MOG (LN18-MOG)-, PLP (LN18-PLP)-, and AQP4 (LN18-AQP4)containing lentiviruses. This assay allows us to determine the binding of antibodies to the extracellular domain of proteins and preserves conformational epitopes, which are essential for antibody binding in 836 Annals of Neurology Vol 66 No 6 December 2009 Fig 1. Cell-based assay to determine antibodies to native central nervous system (CNS) proteins. CNS proteins were expressed in the human glioblastoma cell line LN18. (A) Immunohistochemistry of the LN18-CTR, LN18-MOG, and LN18-PLP cell lines. Expression of native proteins was confirmed after staining of LN18-MOG, LN18-PLP, and LN18-CTR with monoclonal antibody (mAb) 8-18C5 against myelin oligodendrocyte glycoprotein (MOG) (staining shown for LN18-CTR and LN18-MOG) and mAb O10 against proteolipid protein (PLP) (staining shown for LN18PLP) followed by Alexa Fluor 488-conjugated secondary antimouse immunoglobulin G antibodies. Representative images (original magnification, ⫻600) and flow cytometry histograms are shown. Surface expression was confirmed by flow cytometry as shown for mAb anti-MOG 8-18C5 with the LN18-MOG (red line), LN18-PLP (green line), and the LN18-CTR (blue line) cell lines. (B) Titration of mAb demonstrates the correlation between concentration of the antibody and median fluorescence intensity. Flow cytometry on different transfected cells was used to quantify the concentration of the mAbs against MOG (left) and PLP (right) at a wide range of antibody concentrations. Means and standard deviations of 3 measurements for each concentration are shown. vivo. Protein expression was confirmed by immunohistochemistry (shown for MOG and PLP, Fig 1A and B). An empty virus was used as control (LN18-CTR), and was negative for MOG, PLP, and AQP4 stainings (see Fig 1A and B, and data not shown). The intensity of the fluorescence expressed in MFI was correlated to the concentration of antibody used to stain the cells, suggesting that MFI is directly dependent on antibody titer (see Fig 1C). When we applied sera from all children groups at a dilution of 1:100, we observed a significant IgG antibody reactivity to nMOG (Fig 2A and C), but could not detect any significant binding to native AQP4 (nAQP4) and native PLP (nPLP) (see Fig 2C), except in 1 pediatric CIS patient who was weakly positive for anti-nAQP4 antibodies. Furthermore, IgM antibodies to MOG, but not the other autoantigens, were also observed in only 3 of 47 children (see Figs 2B). These findings suggest that a specific immune response to nMOG but not to the other antigens occurs in children. Next, we compared the serum IgG antibody reactivity to nMOG between children with a first inflammatory demyelinating event (DEM) and controls, either children with ONDs or HC (see Fig 3). Antibody titers were higher in DEM children (median ⌬MFI, 24.5; range, 1.0 –1896.3) than pediatric HC (median ⌬MFI, 2.2; range, 0 –10.7), pediatric OND (median ⌬MFI, 3.4; range, 0.3–107.1), and a group of adult patients with MS (median ⌬MFI, 2.1; range, 2.1–21.5) (see Fig 3). Based on the titer observed in pediatric controls, we determined a cutoff for nMOG IgG seropositivity at ⌬MFI ⫽ 38.1 (median ⫹ 95% Š Fig 2. Serum antibody titer to human native myelin oligodendrocyte glycoprotein (nMOG), native proteolipid protein (nPLP), and native aquaporin-4 (nAQP4) in children. Antibody reactivity to human native central nervous system proteins was determined in sera from all patients and controls by incubating LN18-MOG, LN18-PLP, LN18-AQP4, and LN18-CTR cells with serum at a dilution of 1:100. Secondary staining was performed with Alexa Fluor 488-conjugated antihuman immunoglobulin (Ig) G antibody. Representative flow cytometry histograms for negative (upper histogram) and positive (lower histogram) staining for (A) IgG and (B) IgM specific for nMOG are shown (LN18-MOG, red line; LN18PLP, green line; and LN18-CTR, blue line). ⌬MFI is the difference between median fluorescence intensity obtained with the cell line expressing the target protein and the MFI obtained with LN18-CTR cells. (C) IgG antibody titer to nPLP, nAQP4, and nMOG in all childhood patients (children with a first inflammatory demyelinating event) and controls (healthy children and children with other noninflammatory neurological diseases). Antibody titers were compared among groups by the Kruskal-Wallis test. Plain bars display median titer from the whole group, and dotted bars display median titer of positive samples. Brilot et al: MOG Antibodies in Children 837 percentile of pediatric HC and ONDs), and defined a “high” titer as ⌬MFI ⫽ 109.8 (median ⫹ 99% percentile). We detected IgG antibodies to nMOG with ⌬MFI ⬎ 38.1 in 46.8% of DEM children (22 of 47) and 6.9% (2 of 29) of OND children. No pediatric HC or adult MS patient was found to be positive for IgG antibody according to this cutoff. Additionally, we also determined nMOG IgG titers in sera from a cohort of 15 type 1 diabetes children. All type 1 diabetes sera were negative (median ⌬MFI, 2.34; range, 0 –16.52, data not shown), suggesting that IgG to nMOG antibodies were not evidence of general immune dysregulation, but rather a specific finding of CNS demyelination. High titers were only found in 40.4% of DEM children (18 of 47 patients), but were not detected in the 3 childhood and adult control groups. These findings in adult patients were in line with our previous published findings on adult MS and unpublished findings on anti-nMOG titer in adult CIS patients from the Benefit trial.18,24 Titers with MFI ⬎100 were found in ⬍3%, and titers ⬎500 in ⬍1% of adult CIS and MS patients. Anti-nMOG IgG antibodies in some DEM children were still detected at a serum dilution as low as 1:10,000 (data not shown). We also investigated IgM serum antibodies to nMOG (see Fig 3B). Positive IgM antibody reactivity to nMOG was observed in only 3 DEM children (not significant) and no controls. All 3 children also had high IgG serum antibody titer to nMOG. Fig 3. Serum antibody titer to native human myelin oligodendrocyte glycoprotein (nMOG) in children with a first inflammatory demyelinating event (DEM). (A) Immunoglobulin (Ig) G and (B) IgM antibody titers to nMOG were determined in sera of healthy children (HC), children with other noninflammatory neurological diseases (ONDs), children with a first inflammatory demyelinating event (DEM), and adults with multiple sclerosis (MS). Antibody titers were compared among groups by the Kruskal-Wallis. Plain bars display median titer from the whole group, and dotted bars display median titer of positive samples. p Values comparing DEM children with other groups are given. MFI ⫽ median fluorescence intensity; ns ⫽ not significant; nd ⫽ not determined. 838 Annals of Neurology Vol 66 No 6 December 2009 IgG Antibodies to nMOG Are Found in Children Both with CIS and with ADEM, Whereas IgM Antibodies Are Only Present in ADEM Based on current consensus definitions, we stratified DEM children into pediatric CIS or ADEM. Children were followed up for a mean of 2 years after the initial demyelinating event to determine whether they developed MS or had no further disease activity (see Table). Similar IgG antibody titers to nMOG were observed in children with ADEM and CIS (Fig 4A). The frequency of children with high titer did not differ between groups. Furthermore, the presence and titer of antinMOG IgG antibodies did not predict progression to MS, because similar titers were found in the pediatric DEM group that progressed to MS and those that did not progress to MS (see Fig 4B). Next, we compared the age of the patients with high anti-nMOG IgG (⌬MFI ⬎ 109.8) versus low or negative IgG antibody titer to nMOG. Both pediatric ADEM and CIS patients with high titer were significantly younger than children without nMOG serum antibodies or low titer (see Fig 4C). In children with anti-nMOG antibodies (cutoff, ⌬MFI ⫽ 38.1), serum titer and age correlated negatively (r ⫽ ⫺0.46, p ⫽ 0.023, data not shown). Interestingly, IgM antibodies to nMOG were only found in 3 of 19 children affected by ADEM but none of those affected by CIS (see Fig 3B and data not shown). diatric ADEM and 5 pediatric CIS), in whom enough CSF was available for analysis. Anti-nMOG IgG antibodies were detected in all CSF tested. To determine whether IgG anti-nMOG antibodies are produced intrathecally, we adjusted serum and CSF of each patient to the same IgG concentration. Paired samples were then analyzed for nMOG titer. In all children affected by ADEM, the antibody titer was lower in CSF than serum. In 3 of 5 CIS children, an intrathecal IgG antibody production specific for nMOG was observed (titer was higher in CSF than serum), 1 of whom has progressed to MS (Fig 5). IgG Antibodies to Native MOG Are Produced in the Periphery and the CNS To understand whether the antibodies specific for nMOG are produced in the periphery alone, or also locally in the CNS, we determined the antibody titers in the CSF of 8 children from the same cohort (3 pe- Antibodies to nMOG Are Cytotoxic in DEM Children The binding of antibodies to nMOG at the surface of LN18-MOG cells suggests that they might fix complement or induce activation of antibody-dependent cellular cytotoxicity (ADCC) in MOG-expressing cells; this has been proposed as a mechanism by which antibodies can have detrimental effect.25 To determine the biological activity of these antibodies on LN18MOG and LN18-CTR cells, we performed ADCC assays with purified IgG from sera of children with high or low anti-nMOG antibodies. Survival of LN18-CTR was similar after incubation with IgG from sera with high anti-nMOG antibodies and negative anti-nMOG antibodies (Fig 6a). On the contrary, IgG from patients with high nMOG antibodies induced NKmediated killing of LN18-MOG cells compared with IgG from patients with negative anti-nMOG antibodies, suggesting that the binding of IgG to nMOG promotes a cytotoxic effect (see Fig 6a). We also observed a strong correlation between the IgG nMOG titer and the extent of ADCC (see Fig 6b). Discussion Several lines of evidence suggest that antibodies and B cells play an important role in autoimmune diseases of Š Fig 4. Relation between disease course, age of onset, and antihuman native myelin oligodendrocyte glycoprotein (nMOG) antibodies. (A) Comparison of immunoglobulin G anti-MOG antibody titer in children with acute disseminated encephalomyelitis (ADEM) and clinically isolated syndrome (CIS), and (B) children who did and who did not progress to multiple sclerosis (MS). Plain bars display median titer from the whole group, and dotted bars display median titer of positive samples. p Values are given. (C) Children with a first inflammatory demyelinating event were stratified into 2 groups, with either high anti-MOG titer in serum (⌬-median fluorescence intensity [MFI] ⬎ 108.9) or negative/low titer, and were compared according to age of disease onset. The p value is displayed on the right side of the graph. Antibody titers were compared between groups using the Mann-Whitney U test. ns ⫽ not significant. Brilot et al: MOG Antibodies in Children 839 against native MOG in children with a first episode of demyelination was performed by O’Connor et al.19 Using a different method, O’Connor found that a significant subgroup of ADEM patients have very high antibodies against native MOG, higher than adult MS patients. Importantly, the O’Connor cohort used pa- Fig 5. Intrathecal synthesis of anti-native human myelin oligodendrocyte glycoprotein (nMOG) antibodies in acute disseminated encephalomyelitis (ADEM) and clinically isolated syndrome (CIS). The extent of specific intrathecal antibody production was determined in individual pediatric ADEM (n ⫽ 3) and CIS patients (n ⫽ 5, including 1 CIS patient who progressed to multiple sclerosis). Cerebrospinal fluid (CSF) and serum immunoglobulin G (IgG) were normalized to 5mg/l, and the ratio of anti-nMOG reactivity was determined. We defined intrathecal IgG production as a ratio above 1.5 (dotted line). the CNS. An increasing number of autoantibodies specific for rare inflammatory diseases of the CNS have been identified such as antibodies to AQP4 in neuromyelitis optica,26 GABA, and amphiphysin in stiff person syndrome,27,28 voltage-gated potassium channels or NMDA-receptor in limbic and paraneoplastic encephalitis.29 The discovery of these antibodies has had an important impact on our understanding of the underlying pathomechanisms, but has also enriched our diagnostic repertoire.30,31 In MS, a number of myelin antigens have been proposed as targets of the autoimmune response. Among them, MOG has been the most persuasive candidate. In particular, its expression on the surface of the myelin sheath and its ability to induce encephalitis in many animal models emphasizes a possible role of MOG in CNS autoimmunity.11 However, the results from many studies have not provided a conclusive picture on the possible role of MOG in autoimmune diseases of the CNS. One of the reasons for the controversial results might have been the different technical approaches to measure antibodies against MOG. Although many studies have shown that biologically relevant antibodies target conformational epitopes of MOG, most studies have measured antibody responses to denatured MOG protein.15–17 Only recently have assays been established that enabled us to measure antibody responses to native MOG reflecting the correct properties of the protein in vivo.18 –20 The only other report investigating antibodies 840 Annals of Neurology Vol 66 No 6 December 2009 Fig 6. Biological activity of anti–native human myelin oligodendrocyte glycoprotein (nMOG) antibodies. Antibodydependent cellular cytotoxicity was performed with purified CD56⫹ natural killer cells and purified immunoglobulin G (IgG) from acute disseminated encephalomyelitis and clinically isolated syndrome serum. (A) Survival rate of LN18-CTR and LN18-MOG cells after overnight incubation with patient IgG. Patients with high IgG anti-nMOG titer (Ab⫹) and patients with negative anti-nMOG IgG antibodies in serum (Ab⫺) were used. Symbols represent cell survival obtained in the presence of individual sera. (B) Comparison between antinMOG antibody titer and specific cytotoxic activity of serum in vitro is shown. Specific cell survival for each serum was determined as follows: % cell survival of LN18-CTR ⫺ % cell survival of LN18-MOG; Spearman rank correlation test was used. r and p values are displayed. DEM ⫽ children with a first inflammatory demyelinating event. tients from multiple countries, did not apply 2007 International Consensus criteria, and did not follow up patients. This is relevant because the 2007 International Consensus is the first published guidelines establishing strict clinical criteria to help in the diagnosis of CIS and ADEM. In our study, we investigated systematically the antibody response in a 2007 International Consensus-defined and prospectively followed group of children affected by a first CNS demyelinating event. We also present the first evidence that these antibodies against nMOG in children can participate in a pathogenic mechanism via antibody-dependent cellularmediated cytotoxicity. In contrast to adults affected by MS, we observed a very strong IgG response to nMOG in a subgroup of these children. In adults, subclinical disease activity may precede the first disease episode by years. By contrast, in children it is conceivable that the onset of clinical symptoms and signs occurs much more rapidly after initiation of CNS demyelination,6 supporting acute inflammation and antibody production. Therefore, children may be a better model of early immunological changes in CNS demyelination. The inverse correlation between antibody titer and age also suggests that profound antibody responses to nMOG are more common in early childhood and do not seem to persist during later stages of CNS autoimmunity. The antibody response against MOG was primarily of the IgG isotype. MOG IgM was observed rarely, and only in ADEM children, possibly related to the hyperacute postinfectious nature of ADEM, compared with CIS and MS. Such high levels of anti-nMOG antibodies were not seen in age-matched controls, in children affected by other neurological diseases, or in type 1 diabetes children. The cohort only had a short mean follow-up, and prolonged follow-up is important not only for diagnosis purposes, but also to define the association between antibodies and first episode of demyelination further. However, it is worth noting that the majority of children (10/13) had their MS-defining relapse within the first year. This is in line with previous data reporting that most children ⬎10 years old experience their second attack within 12 months of the first attack.6 nMOG IgG was also present in the CSF of a limited number of patients tested, although intrathecal production was observed in 3 CIS patients, 1 of whom progressed to MS. However, high intrathecal synthesis of MOG IgG may be an important precondition for relapse of CNS demyelination, as in MS. This is in line with reports in adult MS patients suggesting that antiMOG antibodies are enriched in CNS lesions compared with the periphery.32,33 Further CSF investigation will be needed to clarify this point. Interestingly, the response in patients with pediatric demyelination was specifically against nMOG, because no significant antibodies were observed to other candidate membrane autoantigens, such as PLP or AQP4. Future studies could include other autoimmune control groups, such as type 1 diabetes systemic lupus erythematosus. The detection of antibodies per se does not mean they are important in causing disease, as they can be present as an epiphenomenon to cell damage. However, we demonstrated that anti-nMOG antibodies were associated with NK cell-mediated cytotoxicity. In the human brain, cells such as microglia and macrophages have been shown to be recruited at MS lesions,34 and to have the potential to mediate ADCC and to participate in antibody-induced myelin damage.35 Other mechanisms of antibody pathogenicity, such as complement-mediated cytotoxicity, were not examined in this report, and represent an alternate pathogenic mechanism.18 These findings suggest that a peripheral immune response to nMOG occurs in a subgroup of children affected by a first demyelinating event, which might be pathologically relevant in the demyelination process. The initial events involved in the generation of antibodies to nMOG are not understood. Exposure to viral antigens can lead to autoimmunity via molecular mimicry.36 –38 Further studies are warranted to clarify how antibodies to MOG are generated and whether MOG is the key target of the first inflammatory demyelinating event in children. In summary, we identified a unique and highly specific antibody response to nMOG in a subgroup of children affected by the first demyelinating event. The identification of this early autoimmune response significantly adds to our understanding of the autoantigens involved in early CNS autoimmunity. This study was supported by University of Sydney postdoctoral fellowships (R.C.D., F.B.); the Brain Foundation, Australia (R.C.D., F.B.); the Deutsche Forschungsgemeinschaft (grant He2386/7-1; B.H. and S.C.); the German Ministry of Education, Science, Research, and Technology (krankheitsbezogenes Kompetenznetz MS; B.H., S.C., R.S., V.K., V.G., S.R.K.); the Gemeinnützige Hertie Stiftung (D.Z.); the Düsseldorf chapter of the Multiple Sclerosis Society and the DAAD (M.A.). References 1. Mikaeloff Y, Caridade G, Husson B, et al. Acute disseminated encephalomyelitis cohort study: prognostic factors for relapse. Eur J Paediatr Neurol 2007;11:90 –95. 2. Noseworthy JH, Lucchinetti C, Rodriguez M, et al. Multiple sclerosis. N Engl J Med 2000;343:938 –952. 3. Compston A, Coles A. Multiple sclerosis. Lancet 2002;359: 1221–1231. 4. Menge T, Hemmer B, Nessler S, et al. Acute disseminated encephalomyelitis: an update. Arch Neurol 2005;62: 1673–1680. 5. Dale RC, Pillai SC. Early relapse risk after a first CNS inflammatory demyelination episode: examining international consensus definitions. Dev Med Child Neurol 2007;49:887– 893. Brilot et al: MOG Antibodies in Children 841 6. Banwell B, Krupp L, Kennedy J, et al. Clinical features and viral serologies in children with multiple sclerosis: a multinational observational study. Lancet Neurol 2007;6:773–781. 7. O’Connor KC, Bar-Or A, Hafler DA. The neuroimmunology of multiple sclerosis: possible roles of T and B lymphocytes in immunopathogenesis. J Clin Immunol 2001;21:81–92. 8. Hemmer B, Archelos JJ, Hartung HP. New concepts in the immunopathogenesis of multiple sclerosis. Nat Rev Neurosci 2002;3:291–301. 9. Sospedra M, Martin R. Immunology of multiple sclerosis. Annu Rev Immunol 2005;23:683–747. 10. Greer JM, Pender MP. Myelin proteolipid protein: an effective autoantigen and target of autoimmunity in multiple sclerosis. J Autoimmun 2008;31:281–287. 11. Iglesias A, Bauer J, Litzenburger T, et al. T- and B-cell responses to myelin oligodendrocyte glycoprotein in experimental autoimmune encephalomyelitis and multiple sclerosis. Glia 2001;36:220 –234. 12. Linington C, Bradl M, Lassmann H, et al. Augmentation of demyelination in rat acute allergic encephalomyelitis by circulating mouse monoclonal antibodies directed against a myelin/ oligodendrocyte glycoprotein. Am J Pathol 1988;130:443– 454. 13. Genain CP, Nguyen MH, Letvin NL, et al. Antibody facilitation of multiple sclerosis-like lesions in a nonhuman primate. J Clin Invest 1995;96:2966 –2974. 14. Breithaupt C, Schubart A, Zander H, et al. Structural insights into the antigenicity of myelin oligodendrocyte glycoprotein. Proc Natl Acad Sci U S A 2003;100:9446 –9451. 15. Reindl M, Linington C, Brehm U, et al. Antibodies against the myelin oligodendrocyte glycoprotein and the myelin basic protein in multiple sclerosis and other neurological diseases: a comparative study. Brain 1999;122:2047–2056. 16. Berger T, Rubner P, Schautzer F, et al. Antimyelin antibodies as a predictor of clinically definite multiple sclerosis after a first demyelinating event. N Engl J Med 2003;349:139 –145. 17. Kuhle J, Pohl C, Mehling M, et al. Lack of association between antimyelin antibodies and progression to multiple sclerosis. N Engl J Med 2007;356:371–378. 18. Zhou D, Srivastava R, Nessler S, et al. Identification of a pathogenic antibody response to native myelin oligodendrocyte glycoprotein in multiple sclerosis. Proc Natl Acad Sci U S A 2006;103:19057–19062. 19. O’Connor KC, McLaughlin KA, De Jager PL, et al. Selfantigen tetramers discriminate between myelin autoantibodies to native or denatured protein. Nat Med 2007;13:211–217. 20. Lalive PH, Menge T, Delarasse C, et al. Antibodies to native myelin oligodendrocyte glycoprotein are serologic markers of early inflammation in multiple sclerosis. Proc Natl Acad Sci U S A 2006;103:2280 –2285. 21. Bennett JL, Lam C, Kalluri SR, et al. Intrathecal pathogenic anti-aquaporin-4 antibodies in early neuromyelitis optica. Ann Neurol 2009;66:617– 629. 842 Annals of Neurology Vol 66 No 6 December 2009 22. Krupp LB, Banwell B, Tenembaum S. Consensus definitions proposed for pediatric multiple sclerosis and related disorders. Neurology 2007;68:S7–S12. 23. McDonald WI, Compston A, Edan G, et al. Recommended diagnostic criteria for multiple sclerosis: guidelines from the International Panel on the diagnosis of multiple sclerosis. Ann Neurol 2001;50:121–127. 24. Hemmer et al. Mult Scler 2008;14:S48. 25. Antel J, Bar-Or A. Roles of immunoglobulins and B cells in multiple sclerosis: from pathogenesis to treatment. J Neuroimmunol 2006;180:3– 8. 26. Lennon VA, Kryzer TJ, Pittock SJ, et al. IgG marker of opticspinal multiple sclerosis binds to the aquaporin-4 water channel. J Exp Med 2005;202:473– 477. 27. Raju R, Rakocevic G, Chen Z, et al. Autoimmunity to GABAA-receptor-associated protein in stiff-person syndrome. Brain 2006;129:3270 –3276. 28. Sommer C, Weishaupt A, Brinkhoff J, et al. Paraneoplastic stiff-person syndrome: passive transfer to rats by means of IgG antibodies to amphiphysin. Lancet 2005;365:1406 –1411. 29. Dalmau J, Gleichman AJ, Hughes EG, et al. Anti-NMDAreceptor encephalitis: case series and analysis of the effects of antibodies. Lancet Neurol 2008;7:1091–1098. 30. Vincent A, Lang B, Kleopa KA. Autoimmune channelopathies and related neurological disorders. Neuron 2006;52:123–138. 31. Wingerchuk DM, Lennon VA, Pittock SJ, et al. Revised diagnostic criteria for neuromyelitis optica. Neurology 2006;66: 1485–1489. 32. O’Connor KC, Appel H, Bregoli L, et al. Antibodies from inflamed central nervous system tissue recognize myelin oligodendrocyte glycoprotein. J Immunol 2005;175:1974 –1982. 33. Genain CP, Cannella B, Hauser SL, et al. Identification of autoantibodies associated with myelin damage in multiple sclerosis. Nat Med 1999;5:170 –175. 34. Bruck W, Porada P, Poser S, et al. Monocyte/macrophage differentiation in early multiple sclerosis lesions. Ann Neurol 1995;38:788 –796. 35. Marta CB, Bansal R, Pfeiffer SE. Microglial Fc receptors mediate physiological changes resulting from antibody crosslinking of myelin oligodendrocyte glycoprotein. J Neuroimmunol 2008;196:35– 40. 36. Fujinami RS, Oldstone MB. Amino acid homology between the encephalitogenic site of myelin basic protein and virus: mechanism for autoimmunity. Science 1985;230:1043–1045. 37. Wucherpfennig KW, Strominger JL. Molecular mimicry in T cell-mediated autoimmunity: viral peptides activate human T cell clones specific for myelin basic protein. Cell 1995;80: 695–705. 38. Hemmer B, Fleckenstein BT, Vergelli M, et al. Identification of high potency microbial and self ligands for a human autoreactive class II-restricted T cell clone. J Exp Med 1997;185: 1651–1659.