Distinction and Temporal Stability of Conformational Epitopes on Myelin Oligodendrocyte Glycoprotein Recognized by Patients with Different Inflammatory Central This information is current as Nervous System Diseases of September 28, 2021. Marie C. Mayer, Constanze Breithaupt, Markus Reindl, Kathrin Schanda, Kevin Rostásy, Thomas Berger, Russell C. Dale, Fabienne Brilot, Tomas Olsson, Dieter Jenne,

Anne-Katrin Pröbstel, Klaus Dornmair, Hartmut Wekerle, Downloaded from Reinhard Hohlfeld, Brenda Banwell, Amit Bar-Or and Edgar Meinl J Immunol 2013; 191:3594-3604; Prepublished online 6 September 2013;

doi: 10.4049/jimmunol.1301296 http://www.jimmunol.org/ http://www.jimmunol.org/content/191/7/3594

Supplementary http://www.jimmunol.org/content/suppl/2013/09/06/jimmunol.130129 Material 6.DC1

References This article cites 58 articles, 12 of which you can access for free at: by guest on September 28, 2021 http://www.jimmunol.org/content/191/7/3594.full#ref-list-1

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2013 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Distinction and Temporal Stability of Conformational Epitopes on Myelin Oligodendrocyte Glycoprotein Recognized by Patients with Different Inflammatory Central Nervous System Diseases

Marie C. Mayer,* Constanze Breithaupt,† Markus Reindl,‡ Kathrin Schanda,‡ Kevin Rosta´sy,x Thomas Berger,‡ Russell C. Dale,{ Fabienne Brilot,{ Tomas Olsson,‖ Dieter Jenne,#,** Anne-Katrin Pro¨bstel,* Klaus Dornmair,* Hartmut Wekerle,** Reinhard Hohlfeld,*,†† Brenda Banwell,‡‡ Amit Bar-Or,xx,{{ and Edgar Meinl*

Autoantibodies targeting conformationally intact myelin oligodendrocyte glycoprotein (MOG) are found in different inflammatory Downloaded from diseases of the CNS, but their antigenic epitopes have not been mapped. We expressed mutants of MOG on human HeLa cells and analyzed sera from 111 patients (104 children, 7 adults) who recognized cell-bound human MOG, but had different diseases, in- cluding acute disseminated encephalomyelitis (ADEM), one episode of transverse myelitis or optic neuritis, multiple sclerosis (MS), anti--4 (AQP4)–negative neuromyelitis optica (NMO), and chronic relapsing inflammatory optic neuritis (CRION). We obtained insight into the recognition of epitopes in 98 patients. All epitopes identified were located at loops connecting the b strands of MOG. The most frequently recognized MOG epitope was revealed by the P42S positioned in the CC9-loop. http://www.jimmunol.org/ Overall, we distinguished seven epitope patterns, including the one mainly recognized by mouse mAbs. In half of the patients, the anti-MOG response was directed to a single epitope. The epitope specificity was not linked to certain disease entities. Longitudinal analysis of 11 patients for up to 5 y indicated constant epitope recognition without evidence for intramolecular epitope spreading. Patients who rapidly lost their anti-MOG IgG still generated a long-lasting IgG response to vaccines, indicating that their loss of anti-MOG reactivity did not reflect a general lack of capacity for long-standing IgG responses. The majority of human anti-MOG Abs did not recognize rodent MOG, which has implications for animal studies. Our findings might assist in future detection of potential mimotopes and pave the way to Ag-specific depletion. The Journal of Immunology, 2013, 191: 3594–3604. by guest on September 28, 2021 utoantibodies play important roles in different encepha- Only Abs recognizing MOG in its correctly folded form, as on lopathies and inflammatory diseases of the CNS (1–6). the cell surface, can be demyelinating and thus pathogenic (12, 13). A Anti–myelin oligodendrocyte glycoprotein (MOG) IgG Only such conformationally intact MOG—for example, as an is found in subgroups of predominantly pediatric patients with in vitro translated streptavidin-linked tetramer or expressed on the acquired demyelinating diseases, such as acute disseminated en- surface of transfected cells—is suitable to identify proportions of cephalomyelitis (ADEM) and multiple sclerosis (MS). MOG is patients with autoantibodies to MOG. specifically expressed in the CNS and is one of the few myelin Such autoantibodies to MOG are found in a substantial pro- that are localized on the outermost surface of myelin portion (∼20–40%) of children with ADEM, chronic relapsing sheaths (7, 8). This localization makes MOG accessible for patho- inflammatory optic neuritis (CRION), or MS, but they are rarely genic autoantibodies and mAbs to MOG induce demyelination in found in adult MS (14–23). Recently anti-MOG Abs were also rodents (9, 10) and primates (11). found in a few anti–aquaporin-4 (AQP4) negative pediatric and

*Institute of Clinical Neuroimmunology, Ludwig-Maximilians-University, 81377 This work was supported by the Deutsche Forschungsgemeinschaft (TR 128), the Munich, Germany; †Institute of Biochemistry and Biotechnology, Martin-Luther- Munich Cluster for Systems Neurology (SyNergy, Munich, Germany), the Verein zur University Halle-Wittenberg, 06120 Halle, Germany; ‡Clinical Department of Neu- Therapieforschung fu¨r Multiple-Sklerose-Kranke, the Bundesministerium fu¨r Bildung rology, Innsbruck Medical University, 6020 Innsbruck, Austria; xDivision of Neuro- und Forschung (Krankheitsbezogenes Kompetenznetz Multiple Sklerose), the pediatrics, Department of Pediatrics I, Innsbruck Medical University, 6020 Gemeinnu¨tzige Hertie Stiftung, and Research Grant I916 from the Austrian Science Innsbruck, Austria; {Neuroimmunology Group, Institute of Neuroscience and Muscle Fund. Research, the Kids Research Institute at the Children’s Hospital at Westmead, Uni- ‖ Address correspondence and reprint requests to Dr. Edgar Meinl, Institute of Clinical versity of Sydney, Westmead, New South Wales 2145, Australia; Neuroimmunology Neuroimmunology, Ludwig-Maximilians-University, Max-Lebsche-Platz 31, 81377 Unit, Department of Clinical Neuroscience, Karolinska University Hospital, 17176 Munich, Germany. E-mail address: [email protected] Stockholm, Sweden; #Comprehensive Pneumology Center, Institute of Lung Biology and Disease, Helmholtz-Zentrum Mu¨nchen, Member of the German Center for Lung The online version of this article contains supplemental material. Research, 81377 Munich, Germany; **Max Planck Institute of Neurobiology, 82152 Abbreviations used in this article: AChR, acetylcholine receptor; ADEM, acute dis- Martinsried, Germany; ††Munich Cluster for Systems Neurology (SyNergy), 80336 seminated encephalomyelitis; AQP4, aquaporin-4; CRION, chronic relapsing inflam- Munich, Germany; ‡‡Division of Neurology, Department of Pediatrics, The Child- xx matory optic neuritis; hMOG, human MOG; MCF, mean channel fluorescence; ren’s Hospital of Philadelphia, University of Pennsylvania, PA 19104; and Neuro- mMOG, mouse MOG; MOG, myelin oligodendrocyte glycoprotein; MS, multiple immunology Unit, Montreal Neurological Institute, McGill University, Montreal, {{ sclerosis; NMO, neuromyelitis optica; mono ADS, patients experiencing only one Quebec H3A 2BA, Canada; and Experimental Therapeutics Program, Montreal acquired demyelinating event (including transverse myelitis, longitudinally extended Neurological Institute, McGill University, Montreal, Quebec H3A 2BA, Canada transverse myelitis and optic neuritis). 1H.W. holds a Hertie Senior Professorship at the Max Planck Institute of Neurobiology. Ó Received for publication May 15, 2013. Accepted for publication July 29, 2013. Copyright 2013 by The American Association of Immunologists, Inc. 0022-1767/13/$16.00 www.jimmunol.org/cgi/doi/10.4049/jimmunol.1301296 The Journal of Immunology 3595 adult patients with neuromyelitis optica (NMO) (16, 18, 21) and in the QuikChange Site-Directed Mutagenesis Kit (Stratagene, Santa Clara, CA), patients at high risk of developing NMO (18). point were introduced into hMOG. The oligonucleotides used were: 9 9 These Abs to MOG are expected to be not only a biomarker; they 5 –CCTGCTTCTTCCGAGATCATGAATACCAAGAGGAGGC–3 (S104E), 59–CCTGCTTCTTCCGAGATGCTGAATACCAAGAGGAGGCAG–39 also contribute to tissue destruction for the following reasons: (H103A/S104E), 59–CATATCTCCTGGGAAGGACGCTACAGGCATGG- human anti-MOG Abs recognized MOG in its native conformation, AGG–39 (N31D), 59–CAGAGTGATAGGACCAGGATACCCTATCCGGG- they are mostly of the complement-activating isotype IgG1 (18– CTCTGG–39 (R9G/H10Y), 59–GGTGACTCTCAGGATCCAGAATGTAA- 9 9 20), and the blood-brain barrier is breached in CNS inflammation, GGTTCTCAGATG–3 (R86Q), 5 –GTGGGGTGGTACCGCTCCCCCTT- CTCTAG–39 (P42S and P42S/H103A/S104E), 59–GTGGGGTGGTACA- allowing anti-MOG IgG to gain access to the CNS. GATCTCCCTTCTCTAGG–39 (P42S), and the corresponding reverse com- Although human autoantibodies to MOG are associated with plementary oligonucleotides. In the case of the P42S/H103A/S104E mutant, inflammatory demyelinating CNS diseases and are presumably the P42S mutation was introduced into the H103A/S104E mutant. The pathogenic, their target epitopes have not been mapped. In this sequences of the purified plasmids were analyzed with DNA sequencing. study, we analyzed 111 sera with anti-MOG Abs obtained from HeLa cells were transfected transiently using Metafectene transfection re- agent (Biontex, Martinsried, Germany), expressing hMOG, mMOG, and hMOG patients with inflammatory CNS diseases, including MS, patients mutants fused C-terminally to EGFP. HEK-293A cells were transfected tran- experiencing only one acquired demyelinating event (including siently using Fugene HD transfection reagent (Promega, Madison, WI), transverse myelitis, longitudinally extended transverse myelitis and expressing hMOG, mMOG, and the P42S mutant fused C-terminally to EmGFP. optic neuritis) (mono ADS), ADEM, anti-AQP4 negative NMO, Surface expression of each of the MOG-constructs was confirmed by FACS-staining using the anti-MOG mAbs 8-18C5 (30) or Y11 (31) at a CRION, and other relapsing ADS cases. concentration of 0.5 mg/ml and detected with a biotin-SP–conjugated goat We addressed the following questions: 1) Can we define distinct anti-mouse IgG 1:500 (Jackson ImmunoResearch, West Grove, PA) and epitopes on conformationally intact MOG recognized by human streptavidin-Dy light 649 1:3300 (Jackson ImmunoResearch; Supple- Downloaded from autoantibodies? 2) Is the response of an individual patient focused mental Fig. 1). The mAb 8-18C5 recognizes MOG on the cell surface, but on a single, individually dominant epitope or is it broadly dis- does not recognize MOG peptides, whereas Y11 recognizes both surface- bound MOG and the MOG peptide aa 76–100 (12). tributed? 3) Are certain epitopes preferentially recognized in certain CNS diseases? 4) In patients with long-term persistence of Cell-bound anti-MOG reactivity assays anti-MOG Abs, do we find evidence for intramolecular epitope For detection of serum Abs, 100,000 cells transiently transfected with MOG spreading, or is the epitope pattern in a given individual stable over variants were suspended in FACS buffer (1% FCS in PBS). The cells were http://www.jimmunol.org/ time? 5) In patients, with rapid decline of anti-MOG IgG, is there incubated with a 1:50 serum dilution for 45 min at 4˚C and washed three evidence for a general inability to mount long-lived plasma cells? times in FACS buffer. The cells were then incubated with a 1:500 dilution of a biotin-SP conjugated goat anti-human IgG (Jackson ImmunoResearch) for 30 min at 4˚C, washed three times, and incubated with streptavidin-Dy Materials and Methods light 649 (Jackson ImmunoResearch) at a dilution of 1:3000. Finally, the Patient samples cells were washed three times and suspended in a 1:500 dilution of propi- dium iodide in PBS. Dead cells were excluded by positive propidium iodide This study included sera of 111 patients with different inflammatory CNS staining. For binding analysis, the cells showing a 100-fold higher FL-1 diseases and Abs to cell-bound MOG: mono ADS, ADEM, MS, NMO, fluorescence intensity as the nontransfected cells were gated (expression CRION, and other relapsing ADS cases (Table I). Of these, 7 patients were level FL-1 102–103, see supplementary Supplemental Fig. 1) and the mean adults (18 or older); 54 of these patients had been recognized as anti-MOG channel fluorescence (MCF) in the FL-4 channel was obtained for these by guest on September 28, 2021 positive in previous studies (18, 20–22); and 57 additional patients were newly cells. The FACS ratio was calculated as MCF(hMOG)/MCF(EGFP only). identified as anti-MOG positive using our cell-bound assay with transiently A cut-off for the FACS ratio was set to 4 SDs above control samples (14 transfected cells from a cohort of 188 pediatric patients with inflammatory healthy controls, 19 other neurologic diseases, and 15 other diseases) as CNS diseases. The proportion of sera identified as positive in this cohort was: 1.7. Cells transfected with the mutants, hMOG, and with EGFP only were ADEM 41%, MS 5%, mono ADS 29%, other relapsing cases 80%, and always measured together in the same experiment to determine the binding MCFðmutantÞ 2 MCFðEGFP onlyÞ CRION 100%. This study was approved by local ethical committees, and percentage as % binding ¼ 3 100%: An informed consent was obtained from all patients, parents, or legal guardians. MCFðWT hMOGÞ 2 MCFðEGFP onlyÞ example is given in supplementary Supplemental Fig. 1. For the dilution experi- Variants and mutants of MOG ments shown in Supplemental Fig. 2, sera were serially diluted in FACS buffer. A subset of 16 patients was also independently analyzed for reactivity to Responses to human MOG (hMOG), mouse MOG (mMOG), and seven hMOG, P42S, and mMOG by titration and immunofluorescence as de- mutants of hMOG (N31D, S104E, H103A/S104E, P42S, P42S/H103A/S104E, scribed in (15). In this case, binding percentage to a mutant was calculated titreðmutantÞ R9G/H10Y, and R86Q) were analyzed. The mutations are shown in Fig. 1. as % binding ¼ 3 100%. The tip of the FG-loop of MOG (Fig. 1) is recognized by the mAb 8- titreðWT hMOGÞ 18C5 (24). Therefore, we mutated the two amino acids H103 and S104 Competition assay to obtain H103A/S104E. We also created the single mutant S104E, because this mutant already reduced binding of rMOG to mAb For competition of the human Abs, the anti-MOG mAbs 8-18C5 and 8-18C5 by .40% (24). Y11 were added to the serum dilution at 240 ng/well, and the transfectants MOG is glycosylated at N31 (25). We were interested in the contribution were incubated with the mixture of serum and mAbs and developed with of the glycosylation for Ab recognition. To this end, we created N31D a 1:500 dilution of biotin-SP conjugated goat anti-human IgG (Jackson hMOG as an unglycosylated MOG mutant. ImmunoResearch) and streptavidin-Dy light 649 1:3000 (Jackson Immuno- When we saw that binding to mMOG was dramatically reduced in the ma- Research) as described above. This anti-human secondary Ab did not cross- jority of patients, we generated three mutants of hMOG with the corresponding react with the murine mAbs (data not shown). murine amino acids. The mutations we chose, based on the sequence and struc- ture analysis of MOG (see Fig. 1), were highly surface exposed in the structure Western blot of mMOG (1PY9) (26) and rat MOG (1PKO) (27) and nonconservatively HeLa cells transfected with MOG-EGFP constructs were lysed at 4˚C for substituted in the sequence of hMOG: P42S, R9G/H10Y, and R86Q. 1 h in RIPA buffer (150 mM NaCl, 1% NP-40, 0.5% sodium deoxycholate, For visualization of the substitutions, a homology model of the Ig-V–like ˚ a 50 mM Tris pH8, 0.1% SDS) containing complete protease inhibitor domain of hMOG (rmsd to mMOG: 0.07 A, 116 aligned C atoms) was mixture (Roche Applied Science, Penzberg, Germany). The lysate was generated with SWISS-MODEL (28) using mMOG as template. Figures then pelleted, and the supernatant was analyzed. For deglycosylation, the were prepared using Pymol (Schro¨dinger, LLC) and Alscript (29). supernatant was digested with PNGaseF (New England Biolabs, Ipswich, Molecular cloning and transfection MA) in Glycoprotein Denaturing Buffer (New England Biolabs), G7 Reaction Buffer (New England Biolabs) and 1% NP40 (New England Full-length hMOG and mMOG were subcloned into the pEGFP-N1 plasmid Biolabs) at 37˚C overnight; 6 mg (digested or undigested) was (CLONTECH Laboratories, Mountain View, CA) or pcDNA 6.2C-EmGFP- loaded onto an SDS gel and separated by gel electrophoresis. The proteins GW/TOPO plasmid (Invitrogen, Carlsbad, CA). These constructs comprise were electroblotted onto a nitrocellulose membrane. The membrane a C-terminal enhanced GFP (EGFP)-tag or emerald GFP (EmGFP)-tag. Using was blocked on PBS containing 3% BSA overnight. The membrane was 3596 DISTINCTION OF MOG EPITOPES incubated with a rabbit anti-GFP mAb (Research Diagnostics, Flanders of expression were obtained with the different constructs, (sup- NJ) at a dilution of 1:5000 for 1 h at room temperature, washed three plementary Supplemental Fig. 1). Cells with the FL1 fluorescence times, and then incubated with a peroxidase-labeled goat anti-rabbit Ab intensity of 102–103 (see Supplemental Fig. 1) were gated to eval- (Dianova, Hamburg, Germany) at a dilution of 1:10,000 for 1 h at room temperature. The blots were developed with ECL. uate the binding to the respective transfected cell. This range was selected to ensure a high expression level and good comparability IgG response to vaccines between the transfectants. The IgG responses to measles and rubella virus were measured by routine Surface expression of each MOG variant was evaluated with two ELISA in the department of clinical chemistry using the Enzygnost Anti- anti-MOG mAbs, 8-18C5 and Y11. We chose these two mAbs Rubella Measles Virus/IgG and Anti-Measles Virus/IgG assays (Siemens because they recognize different epitopes on MOG (24). 8-18C5 Healthcare, Erlangen, Germany). did not recognize the three MOG variants containing the S104E mutation, but bound well to all other mutants we used. All MOG Results variants were recognized by Y11 (supplementary Supplemental Validation of transiently transfected MOG variants and Fig. 1). This shows that the mutations we introduced did not in- reproducibility of binding ratios terfere with MOG surface expression. Furthermore, it was shown hMOG, mMOG, and seven mutants of hMOG (Fig. 1) were an- previously that the introduction of the H103A/S104E mutation did alyzed for recognition by human IgG and mAbs. Similar extents not disturb the overall structure of rat MOG (24). The P42S mu- Downloaded from http://www.jimmunol.org/ by guest on September 28, 2021

FIGURE 1. Structure of MOG and illustration of the epitopes recognized by human autoantibodies on MOG. (A) Protein sequence alignment of the Ig-V like domain of human MOG (hMOG) and of murine MOG (mMOG). Amino acid residues are highlighted in color according to the sequence conservation of hMOG und mMOG along the spectrum (white indicates identical residues; blue, purple, and red indicate least conservative substitutions). The gly- cosylation site Asn31 is depicted in green. Below the alignment, the secondary structure elements of mMOG are shown. The arrows labeled A to G label the strands of the b-sheet. The structure is displayed such that the green strands are on the front side and the blue strands are on the back side. The FG-loop and neighboring residues that represent the center of the epitope on rat MOG (rMOG) recognized by mAb 8-18C5 are marked by black circles. Residues that have been mutated in this study are marked with a red circle. (B) Sequence conservation of mMOG and hMOG mapped onto the three-dimensional structure of mMOG. Selected residues are displayed in three different sizes. Ca-atoms of residues mutated in this study are shown as largest spheres: R9G/H10Y in red and blue, N31D greenish, H103A/S104E in gray, and P42S and R86Q in purple. Ca-atoms of nonconserved residues, which were not mutated, in this study are shown as midsized spheres: G59A/D60E, D74E/A75T, and G77S. These Ca-atoms are displayed in blue, indicating conservative amino acid differences between mMOG and hMOG. Ca-atoms of nonconserved residues with side chains located inside the protein are shown as small spheres: V20A and F96Y. (C) Ribbon representation of mMOG (26). Residues at positions that were mutated in this study are shown as a stick model. Close-up views of these regions show the superposition of the side chains of hMOG (gray) modeled with SWISS-MODEL (28) and the corresponding side chains of mMOG, the rMOG double mutant H103A/S104E (24) and the modeled N31D mutation, respectively (orange). The Journal of Immunology 3597 tation is also unlikely to change the overall structure of MOG, as (59%) showed reduced or no binding to this mutant (patterns 1, 5, both proline and serine are found in different species at this po- and 7). Moreover, in 41 of 111 patients (37%), the MOG-specific Ab sition. response was focused on the CC9-loop (pattern 1, Fig. 2A). Thirty- The similar expression levels of the mutants, the gating on cells six of these 41 sera also showed reduced or no binding to mMOG, with a defined expression level (102–103), and the demonstration of which was expected because mMOG contains S42. In 21 of these 36 binding of at least one MOG-specific mAb allowed the compari- cases, binding to mMOG was lower than binding to P42S, indicating son of binding to the different mutants. that other differences between human and murine MOG further We assessed the reproducibility of our system. Binding of the 111 decreased binding to mMOG. The CC9-loop is hence identified as sera (Table I) to our mutants was analyzed up to three times in the most frequently recognized epitope on hMOG and as the domi- independent experiments, yielding a good reproducibility of the nant epitope in patient sera that recognize a single epitope. binding percentage. For 36 sera, we compared recognition of the The tip of the FG-loop, which is the main target of conformation- three most important constructs P42S, mMOG, and H103A/S104E dependent MOG-specific Abs in mice, is recognized by 36 of 111 to hMOG (i.e., 108 measurements and each in triplicates) and found (32%) patients. In a subgroup of 10 patients (9%) the immune the following. First, in 33 of 108 measurements, the mutation re- response is strongly focused on this region (pattern 2, Fig. 2B). duced the binding to less than 10%, and we found an absolute SD of These sera showed reduced or no binding to the double mutant 7.8%. Second, in the other 75 measurements, binding was either H103A/S104E, but bound well to the other mutants (P42S, R9G/ strongly reduced (,65%), comparable to hMOG (65–200%) or H10Y, R86Q, N31D, and mMOG). Comparing the reactivity of . strongly increased ( 200%); here, the SD of the binding was 20% these 10 sera to the single mutant S104E and to the double mutant of the binding percentage. Sera representing each of the seven H103A/S104E revealed heterogeneity in recognition of the FG- Downloaded from epitope patterns (see below) were analyzed three times (Fig. 2). loop. In 8 of 10 sera, reduced recognition of both the single mutant Two sera were serially diluted, and Ab binding to the MOG variants S104E and the double mutant was comparable to the binding pattern was assessed by FACS. The binding patterns were comparable at of the mAb 8-18C5. However, one other serum showed no decrease different serum dilutions (Supplemental Fig. 2). in binding to S104E, whereas one serum showed strongly increased A subset of 16 sera was independently analyzed for binding to binding to S104E, but not to H103A/S104E (data not shown). P42S and mMOG by titration in an immunofluorescence assay: the We also analyzed whether the binding to the double mutant http://www.jimmunol.org/ resulting 32 binding percentage values were highly comparable H103A/S104E and the single mutant S104E was different. To this , between the independent assays (Spearman r = 0.7136; p 0.0001). end, 76 sera showing either decreased (34 sera binding less than Recognition of MOG-epitopes analyzed with single and 65%) or good recognition (42 sera) of H103A/S104E were tested multiple amino acid mutants for binding to the single mutant S104E. We found recognition of Anti-MOG positive sera (98/111) showed a reduced binding to at these two mutants to be highly similar (Supplemental Fig. 3; this least one of our variants of MOG. In 52 of 111 samples, the immune figure also shows that some sera bind better to these mutants than to response was clearly reduced by mutations of amino acids posi- the hMOG). tioned in only one loop. In 39 of 52 patients, binding to single One serum did not bind the R86Q mutant anymore (pattern 3, by guest on September 28, 2021 mutated loops was decreased to less than one third, indicating that Fig. 2C) and hence recognized the EF-loop of MOG (Fig. 1). This the IgG response is focused on one epitope. In 32 of 111 samples, serum also did not bind mMOG, as we expected, because mMOG we saw a reduction in binding to multiple mutants. In 14 of 111 contains Q86. This serum recognized all other mutants well. samples, we saw reduced or no binding to mMOG, but were not For 14 of 111 samples, we found a reduced recognition of mMOG, able to assign a specific responsible amino acid. The responses to but were not able to identify the precise recognized epitopes (pattern our mutants allowed us to distinguish seven different patterns of 4, Fig. 2D). These anti-MOG Abs might target regions with rather 99 MOG recognition (Figs. 2, 3). The intensity of the recognition of conservative mutations (C D-loop or DE-loop, see Fig. 1), or their hMOG did not differ significantly between sera recognizing the binding might be influenced by subtle structural differences that seven patterns, as summarized in Fig. 3. The structure of MOG cannot be mimicked by our mutants. with the strands and loops referred to in this work is shown in Fig. 1. In summary, the immune response was directed against one single An overview of these patterns is shown in Fig. 3, and examples epitope in 52 of 111 sera. The most frequently found single epitope is of the seven patterns are shown in Fig. 2. dominated by P42 that is not found in mMOG. The major mAb epitope Patients with anti-MOG IgG directed to one epitope. The most located at the FG-loop is conserved between mMOG and hMOG and frequently recognized MOG-epitope was revealed by the P42S is the second most frequently recognized epitope in humans. mutation positioned in the CC9-loop (Fig. 1); 66/111 patients Patients with anti MOG IgG directed to multiple epitopes. Re- duced binding to multiple mutants was detected in 32 of 111 (29%) patient samples. The most common combination, present in 19 of 32 samples, was reduced binding to P42S and to H103A/S104E Table I. Patient data (pattern 5, Fig. 2E). These mutations are located in loops at op- posite ends of the A9GFCC9C99 b-sheet of MOG (distance of Ca No. of Patients ˚ with Abs to No. of Mean Age in atoms: daverage =23A) and the observed maximum dimensions of Disease hMOG (Adults) Females Years (Range) Ab epitopes account for 21 A˚ 3 28 A˚ (32). Mono ADS other 45 (2) 25 11.0 (1.5–51.8) In the MOG IgD (8-18C5) Fab complex, the FG-loop forms the than ADEM center of the epitope, whereas S42 is located too far away to interact ADEM 40 (1) 19 7.0 (1.4–47.1) with Ab amino acids (27); thus, P42 and H103/S104 cannot si- MS 10 (1) 7 11.3 (3.4–34) multaneously bind to the center of this Ab paratope. In our study, NMO like 2 (1) 1 34.7 (13.5–55.9) CRION 10 (2) 6 15.3 (7.4–32) 64 of 111 (58%) patient Abs recognized either P42 (pattern 1, 7) Other relapsing ADS 4 1 7.9 (3.2–15.7) or H103/S104 (pattern 2, 6). For this reason, sera showing pattern Other relapsing ADS cases: three patients had one ADEM attack plus one non- 5 reactivity (reduced binding to H103A/S104E and to P42S) can ADEM attack, and one patient had a monolesional transverse myelitis. be assumed to contain at least two different Ab populations rec- 3598 DISTINCTION OF MOG EPITOPES

FIGURE 2. Examples of the seven iden- tified epitope patterns of anti-MOG re- activity. Depicted is the mean percentage binding value of three independent ex- periments. Error bars represent the SD. In patterns 1–3 (A–C) the anti-MOG IgG response is focused on only one epitope. In pattern 1, the anti-MOG response is directed against the CC9-loop of MOG, binding to P42S and to mMOG is re- duced; binding to P42S/H103A/S104E is also reduced. In pattern 2, the anti-MOG response is directed against the FG-loop of MOG; this causes a reduction in binding Downloaded from to H103A/S104E and to P42S/H103A/ S104E. In pattern 3, the anti-MOG re- sponse is directed against the EF-loop of MOG. Binding to R86Q and to mMOG is reduced. In pattern 4 (D), binding to mMOG is reduced, but the mutations

P42S, R86Q, and R9G/H10Y introduc- http://www.jimmunol.org/ ing individually the murine amino acids had no effect on binding. The immune response of patients recognizing patterns 5–7 (E–G) is directed against multiple distinct epitopes. In pattern 5, the anti- MOG response is directed against the FG-loop and the CC9-loop of MOG. Binding to H103A/S104E, P42S/H103A/ S104E, mMOG, and P42S is reduced. In by guest on September 28, 2021 pattern 6, binding to H103A/S104E and to mMOG is reduced; the mutations P42S, R86Q, and R9G/H10Y had no effect on binding. In pattern 7, binding was di- rected against the CC9-loop (reduced binding to P42S) and to a second loop, either the EF-loop (reduced binding to R86Q) or the AA9-loop (reduced binding to R9G/H10Y).

ognizing distinct epitopes. Four of these 19 samples, however, mMOG was reduced in 4/6), indicating that within this pattern ad- showed strongly reduced binding (,10%) to both H103A/S104E ditional epitopes are recognized to different extents. and to P42S, which indicates that in these sera both loops influ- A subset of 32 of 111 sera showed reduced reactivity to various ence Ab binding simultaneously. Regarding the observed maxi- MOG constructs carrying distant mutations. As discussed earlier, mum dimensions of Ab epitopes and the distance between the two the most frequently recognized epitopes are found at the CC9-loop MOG loops, the epitopes recognized by these samples have to be and at the FG-loop of hMOG. In addition, the reactivity to both of highly extended with hot spots of binding in the two loops at the these two epitopes is the most common combination among sera edges of the epitope. recognizing multiple epitopes. In 7 of 111 patient samples, binding to H103A/S104E and The triple mutant P42S/H103A/S104E. Binding to the triple mutant mMOG was reduced or abolished, but binding to P42S, R9G/H10Y, P42S/H103A/S104E, which combined the two most important and R86Q was good (pattern 6, Fig. 2F). In 6 of 111 samples, epitopes of human anti-MOG IgG, was reduced in 64% of all sera. binding was reduced to the P42S mutant and to the R9G/H10Y or As expected, all 19 sera that showed a reduced binding to both to the R86Q mutant (pattern 7, Fig. 2G). We noted that sera showing P42S and H103A/S104E showed strongly decreased binding to the pattern 7 reactivity differed in reactivity to mMOG (binding to triple mutant (less than 35% binding; pattern 5, Fig. 2E). The Journal of Immunology 3599 Downloaded from http://www.jimmunol.org/

FIGURE 3. Anti-MOG reactivity grouped in seven patterns. We analyzed the epitope recognition of 111 sera with our mutants and were able to assign epitope patterns in 98 samples, which we grouped into seven patterns, indicated by the numbers 1–7 in row 4. Fifty-two of ninety-eight samples showed an immune response focused on one epitope (patterns 1–3). Fourteen of ninety-eight samples showed reduced binding to mMOG, but recognized all other mutants well (pattern 4). Thirty-two samples showed an immune response directed against multiple distant epitopes (patterns 5–7). In the last row, the intensity of the anti-hMOG reactivity of each pattern is indicated. The values shown are the FACS ratios (MCF reactivity to hMOG/MCF reactivity to by guest on September 28, 2021 EGFP; mean 6 SEM). The anti-hMOG reactivity varied within each pattern, but did not differ significantly between the different patterns (difference between any two patterns, one-way ANOVA, p . 0.05).

Within the group of 41 sera with pattern 1 recognition, 13 showed reactivity to the P42S mutant. In 3 of 4 sera, these autoantibodies increased (i.e., more than 200%) binding to H103A/S104E. Nine also bound the triple mutant well, although they did not recognize of these 13 sera also showed reduced binding to the triple mutant H103A/S104E. P42S/H103A/S104, whereas the remaining 4 of 13 bound this triple Higher reactivity to mutated variants of MOG. Thirty-two of 111 mutant comparably strong as hMOG (also see later). Thus, our patients showed a clearly elevated (more than 2-fold) reactivity to triple mutant revealed a further heterogeneity of sera with pattern mutated variants of MOG as compared with hMOG, and some 1 recognition. bound multiple mutants better than hMOG. These 32 patients The triple mutant also allowed further differentiation of sera that included 8 who recognized MOG better in the absence of glyco- recognized the second most important epitope, the FG-loop. Within sylation (N31D mutant, details are discussed later). Twelve patients the subgroup of 17 sera with decreased binding to H103A/S104E recognized mMOG better than hMOG, six of these also bound (patterns 2 and 6), four sera showed increased binding to P42S. better to P42S than to hMOG. This increase in binding to P42S The one serum with pattern 6 recognition and a strongly increased might be explained by the rigidity of proline and the flexibility of P42S recognition showed poor recognition of mMOG. This result serine, which will allow an induced fit of the protein to the Ab. For could be explained by other more subtle structural differences 6 of 12 patients showing higher reactivity to mMOG, we were not between human and mMOG. In only 1 of these 4 sera, binding to able to assign an epitope recognition pattern; 3 of 12 showed pattern the triple mutant was decreased. The remaining 3 sera bound well 2recognitionand3of12recognizedotherpatterns. to the triple mutant. One example is the optic neuritis serum Seventeen of 111 patient sera bound H103A/S104E better than AEB048 that showed only 12% binding to H103A/S104E (pattern hMOG, 4 of which also bound better to mMOG; 13 of these 17 sera 2). This serum showed strongly increased binding to the P42S recognized pattern 1 (reduced binding to P42S). Nine of 13 sera mutant (2057% binding to P42S and 1518% binding to mMOG, with increased binding to H103A/S104E and pattern 1 recognition compared with hMOG). In this case, binding to the triple mutant (reduced binding to P42S) also showed reduced binding to the triple was not reduced, but was also stronger than to hMOG (610%). mutant P42S/H103A/S104E, as described in the previous para- Sera with decreased reactivity to both P42S and to H103A/S104E graph. As explained earlier, these sera could contain at least two also showed decreased reactivity to the triple mutant P42S/H103A/ different kinds of anti-MOG Abs—one recognizing the CC9-loop S104E. Thirteen sera showed increased reactivity to H103A/S104E, and another with improved binding to the artificial mutant H103A/ but did not recognize P42S. In 9 of 13 cases, these autoantibodies S104E. However, we cannot exclude the possibility that an Ab did not recognize the triple mutant. Four sera showed increased recognizes both loops simultaneously at the edge of the epitope. 3600 DISTINCTION OF MOG EPITOPES

Another possible explanation for increased recognition of MOG recognized pattern 2; 1/8 recognized pattern 7, and 1/8 bound well variants could be that high-affinity anti-MOG Abs do not recognize to all mutants.) a MOG variant allowing multiple other lower affinity anti-MOG MOG epitopes recognized by patients with different disease Abs to bind to this variant. Theoretically, the sum of binding of entities these low-affinity Abs could result in an increased FACS ratio. A subset of 32 of 111 sera showed increased reactivity to mutated Distinct epitope patterns were recognized by anti-MOG Abs in the variants of MOG. The most common mutants to elevate autoan- serum of patients with six different clinical entities: ADEM, mono tibody binding were H103A/S104E, mMOG and N31D. ADS, MS, CRION, NMO, and other relapsing ADS cases. Each epitope pattern was found in several clinical presentations (Table Comparison of MOG epitope analysis with mutated variants II). No statistically significant association between epitope rec- versus competition with defined mAbs ognition and diagnosis was found (x2 test, p = 0.27). Nevertheless, We compared our single amino acid mutation assay to blocking a larger sample size might potentially reveal such an association. assays with defined mAbs. The mAbs 8-18C5 and Y11 recognize It is interesting to note that in the CRION group, 4 of 10 and only different epitopes on MOG (24). We analyzed IgG binding of 15 4 of 101 other patients showed more than 200% binding to the sera in competition with either 8-18C5 or Y11; we found both N31D mutant. mAbs to compete with the human IgG in 15 of 15 sera. The tested We analyzed whether patients with chronic inflammatory dis- sera recognized different epitopes; 13 of them recognized patterns eases had an anti-MOG response directed against more epitopes 1, 2, 4, 5, 6 and 7; and two bound well to all mutants. Fig. 4 shows than did patients with a monophasic disease. Of 26 cases of chronic two sera that were blocked by the mAbs 8-18-C5 and Y11 to inflammatory diseases—namely MS, NMO, CRION, and other Downloaded from a similar extent, although they recognized different epitopes as relapsing ADS cases—35% recognized multiple epitopes (patterns revealed by our mutants; the mutation H103A/S104E abrogated 5, 6, and 7) and 31% recognized a single epitope (patterns 1, 2, the binding of one, but did not interfere with the binding of the and 3). Of 85 patients with a monophasic disease (ADEM and other serum. This result shows that mAbs can block binding of mono ADS), 27% recognized multiple epitopes (patterns 5–7) and human anti-MOG IgG, even when a different epitope is recog- 44% recognized a single epitope (patterns 1–3). These results are

nized by the human Abs. Therefore, systematic mutation of amino summarized in Table II, and they indicate that recognition of http://www.jimmunol.org/ acids gives more detailed information about the epitopes recog- a single epitope of multiple epitopes is not linked to monophasic nized by human anti-MOG Abs than competition assays do with or chronic disease. defined mAbs. Long-term analysis of individual epitope patterns on MOG The role of MOG-glycosylation in autoantibody binding Follow-up sera of 11 anti-MOG Ab positive patients were analyzed Deglycosylation with PNGaseF yielded MOG proteins with the with the different MOG mutants. In 9 of 11 patients, we were able same size as the N31D mutant (Fig. 5), indicating that N31D is the to assign one of the aforementioned epitope patterns. The patterns only N-glycosylation site used in our constructs. The “no glyco- stayed constant in 9 of 9 analyzed cases for an observation period

sylation” mutant N31D did not significantly lower binding in any of up to 50 mo (MS) without evidence for intramolecular epitope by guest on September 28, 2021 of the sera. Examples of recognition of N31D are shown in Figs. 2 spreading (Fig. 6 and Supplemental Fig. 4). Constant epitope and 6 and in Supplemental Fig. 4. We conclude that the glyco- patterns were found in MS, CRION, and ADEM; this was espe- sylated part of MOG is not recognized by autoantibodies. Instead, cially remarkable for patients with MS. In pediatric patients with 8 sera recognized the unglycosylated MOG better than the hMOG. MS, anti-MOG IgG persists with fluctuations (20). For example, Five of these patients showed low recognition of hMOG, with the anti-MOG reactivity of the MS patient ACJ-162 (Supple- a FACS ratio less than 2.0. Enhanced recognition of the ungly- mental Fig. 4H) decreased below detection level after 12 mo, but cosylated mutant was not linked to recognition of a certain epitope clear anti-MOG reactivity was seen after 24 mo (20). The anti- pattern (3/8 recognized pattern 4, 2/8 recognized pattern 1; 1/8 MOG Abs still recognized the same pattern (pattern 1) after a follow-up period of 36 mo. Different epitope patterns, 1, 2, 4, 5, 6 and 7, stayed constant over years. Additional details of this follow-up part of our study are presented in Supplemental Fig. 4.

FIGURE 4. Mutated variants of MOG allow more precise insight into recognized epitopes than blocking with mAbs does. The two patient sera FIGURE 5. N31D mutation completely abrogates MOG-glycosylation. displayed here showed a different reactivity to mutated variants of MOG: Cell lysates of HeLa cells transiently transfected with MOG-variants were ACJ-022 recognizes the FG-loop containing H103/S104 and the CC9-loop digested with PNGase F as indicated; 6 mg total protein were loaded into containing P42S (pattern 5); ACX053 does not bind the FG-loop, but each well of an SDS gel, separated by gel electrophoresis, blotted onto recognizes the CC9-loop containing P42 (pattern 1). Nevertheless, both nitrocellulose, incubated with a rabbit anti-EGFP mAb and developed with sera are blocked by the two anti-MOG mAbs 8-18C5 and Y11 to a similar a peroxidase-labeled goat anti-rabbit Ab and ECL. After digestion with extent. The mAb 8-18C-5 recognizes the FG-loop of MOG (24). Y11 PNGase F, all MOG-mutants had the same size as N31D. Digestion of recognizes cell-bound MOG and the hMOG peptide aa76-100 (12). N31D did not alter the size of this particular mutant. The Journal of Immunology 3601

specific recognition pattern is different from the features of anti- AQP4 autoantibodies. Human anti-hAQP4 Abs cross-react with mAQP4; staining of mouse tissue was even used to identify NMO IgG (34). Our study shows that other human autoantigens might be missed when screening with rodent tissue. The pathogenic potential of human autoantibodies is best demonstrated in transfer studies into experimental animals as done with anti-AQP4 Abs (35–38). The human Abs to MOG have all the characteristics of pathogenic autoantibodies: they recognize MOG in its correct conformation, they are mostly of the complement fixing isotype IgG1 (18–20), and they activate Ab-dependent cel- FIGURE 6. Temporal stability of recognition of MOG-epitopes. De- picted is a case of pediatric MS with a pattern 1 recognition that stayed lular cytotoxicity (14, 18). Their pathogenic activity, however, has constant over the observation period of 50 mo. An increased recognition of not yet been shown with affinity-purified Abs. Transfer experi- hMOG after three months is marked with an asterisk. ments with concentrated human sera (39) are difficult to interpret, because human sera could have pathogenic compounds beyond anti-MOG IgG. Our study shows that only a minority of human Comparison of the dynamic of anti-MOG IgG with sera with anti-MOG IgG are suitable for transfer experiments in anti-measles virus and anti-rubella virus IgG mice. The recognized proline at position 42 is not present in mice

We selected three patients with a rapid decline of anti-MOG IgG and rats, not even in the New World primate Callithrix jacchus, Downloaded from (20) for a comparative analysis of the dynamic of IgG produced in but appears in the rhesus monkey (Macaca mulatta). response to vaccines. All three patients had been vaccinated The major autoantibody epitope found here is different from the against measles and rubella virus. IgGs against these vaccines immunodominant and pathogenic epitope in rodents (24). Human typically persist. We found that these patients mounted a perse- anti-MOG Abs mainly recognized the CC9-loop around P42 of vering IgG response against both measles and rubella virus, but hMOG, whereas most mouse mAbs to MOG recognized the FG-

they lost the anti-MOG IgG rapidly (Fig. 7). loop. Animal experiments have shown that not all Abs against http://www.jimmunol.org/ MOG are pathogenic (12, 40, 41). Mouse mAbs, which recognize Discussion MOG on the cell surface and are pathogenic, can recognize dif- In this study, we define epitopes of conformationally intact MOG ferent epitopes: both the mAb 8-18C5 and the mAb Y11 are recognized by human autoantibodies. The mutants of MOG we pathogenic (31). Thus, one would expect that not only those hu- applied allowed us to obtain insight into recognized epitopes in 98 man anti-MOG Abs recognizing the FG-loop (patterns 2, 5, and of 111 patients. Based on the tested mutants, half of the patients 6), but that also other Abs recognizing another part of the surface showed an immune response directed against one single epitope, of MOG, e.g., the CC9-loop, are pathogenic. The CC9-loop is closer the other half recognized multiple epitopes. to the membrane than the FG-loop. Because this loop is recog- Mutation of the single amino acid P42 in the CC9-loop abrogated nized by Abs when displayed on the surface of transfected cells, by guest on September 28, 2021 or reduced recognition of MOG in the majority of anti-MOG– we anticipate that it is also recognized on the surface of myelin. It positive patients. The second most frequently recognized epitope is evident from features of the anti-MOG mAb Y11 that the same is located at the tip of the FG-loop (H103; S104), which is bound Ab can recognize both a linear peptide and the cell-bound con- by the mAb 8-18C5 (27). Overall, we distinguished seven patterns formationally intact MOG protein (12). Thus, it is likely that some of Ab recognition. All epitopes identified in this work are located of the anti-MOG Abs in patients recognizing cell-bound MOG at loops that connect the b-strands of the IgV-like fold of MOG. also recognize linear peptides. This observation is in harmony with the concept that antigenicity The identification of precise epitopes of autoantibodies can pro- correlates with solvent accessibility and flexibility in proteins vide the basis for an Ag-specific depletion of relevant B cells. A proof (33). It is currently unknown whether the serum anti-MOG Ab of concept for such an Ag-specific therapy has recently been obtained response is polyclonal. Our data provide direct evidence for the in an animal model of diabetes (42). In a different approach, in- polyclonality of at least a subgroup of anti-MOG sera, because we tracerebral injection of competing nonpathogenic anti-AQP4 Abs observed reduced binding to multiple mutants in about a third of reduced AQP4 and myelin loss in a mouse model of NMO (43). all donors. Our study shows that the application of mutant variants of MOG Most of the patients recognizing hMOG did not recognize allows a more precise insight into epitope recognition than blocking mMOG, largely because the majority of sera did not bind to P42S, with defined mAbs does. In agreement with previous studies, we which is also found in mMOG. Other amino acid differences found that the anti-MOG mAb 8-18C5 competes with binding of between the two species also contribute to the differential recognition human Abs to cell-bound MOG (19, 20). We show that this mAb of human and murine MOG (patterns 3, 4, and 7). This species- also inhibits sera that recognize different epitopes. This is not

Table II. Recognition of epitope patterns in different disease groups

No. of No Epitope Disease Patients Pattern 1 Pattern 2 Pattern 3 Pattern 4 Pattern 5 Pattern 6 Pattern 7 Found Mono ADS other than ADEM 45 18 404822 7 ADEM 40 16 515533 2 MS 103001210 3 NMO like 2 0 000110 0 CRION 10 3 104001 1 Other relapsing ADS 4 1 000300 0 Other relapsing ADS cases: three patients had one ADEM attack plus one non-ADEM attack, and one patient had a monolesional transverse myelitis. 3602 DISTINCTION OF MOG EPITOPES

This study analyzes epitopes of human anti-MOG Abs with mutated cell-bound MOG. Earlier studies using peptide ELISA assays (47–49) would not give information on epitopes of con- formationally intact MOG. In this study, we show that human anti- MOG Abs recognize the loops of structurally intact MOG, which should not be provided in a peptide ELISA, and indeed two studies failed to identify these epitopes in an ELISA assay (47, 48). An- other study reported that linear epitopes aa 37–48 and aa 42–53 are immunodominant in a peptide ELISA assay, but these peptides were also recognized by controls at a lower frequency (49). The donors assessed in (49) were adult patients with MS, who rarely have IgG against conformationally intact MOG. In addition, because a sec- ondary Ab recognizing IgG, IgA, and IgM was used in that study, it is likely that mainly low-affinity IgM was detected. To study the potential relevance of MOG-glycosylation in our study, we applied the N31D mutant (23). This application com- pletely abolished N-glycosylation of MOG, indicating that in our MOG-constructs N31 is the only used N-glycosylation site. A

second potential glycosylation site N52, which lacks the consen- Downloaded from sus N-glycosylation sequence N-!P-[S|T], was found in mouse brain by tandem mass spectrometry (50). This site was not con- sidered a high-confidence glycosylation site (50) and was not used to glycosylate MOG in our case. In our study, unglycosylated MOG was recognized well by all human anti-MOG Abs, in agree-

ment with O’Connor et al. (23). We noted that 8 of 111 sera even http://www.jimmunol.org/ showed increased reactivity to unglycosylated MOG, which might be due to the better accessibility of MOG lacking the polycarbo- hydrate chain at its upper, very exposed edge of its extracellular domain. This effect is reminiscent of observations made in HIV, in which deglycosylation of the HIV envelope glycoprotein gp120 led to increased recognition by neutralizing Abs (51). In addition to unglycosylated MOG, other mutated variants of MOG were also recognized at least twice as good as hMOG: 17 of 111 sera recognized H103A/S104E better, and 6 of 111 recognized by guest on September 28, 2021 P42S better than hMOG. Mutation of serine to glutamic acid, as in the H103A/S104E mutant is used experimentally to mimic phos- FIGURE 7. Dynamic of anti-MOG IgG compared with anti–measles phorylation, a strategy called “pseudo-phosphorylation” (52). It is virus and anti–rubella virus IgG. The IgG recognition of measles virus, possible that these Abs are generated against phosphorylated rubella virus, and MOG was analyzed longitudinally. The response at MOG. We have used NetPhos 2.0 (53) to predict phosphorylation disease onset was set as 100% for each Ag, and the subsequent responses sites on hMOG and found S104 to be a likely site for phosphor- were calculated. In these three patients, IgG against MOG declined ylation, with a score of 0.994 (data not shown). Further experi- quickly, whereas the IgG response against measles and rubella virus vac- ments are required to confirm whether MOG in the human CNS is cine was stable for the observation period of 5–6 y. All three patients had indeed phosphorylated at S104 or other positions. A a monophasic CNS inflammation. ( ) Patient ACX-053 experienced a Although patients, in particular with ADEM, only show a transient B C single episode of optic neuritis. ( and ) Patients ACJ-013 and ACJ-181 IgG response to MOG, others (in particular with MS) tend to have had monophasic ADEM. a persisting Ab response to MOG (15, 20). We have addressed two issues related to the dynamic of anti-MOG IgG. First, we analyzed surprising, because MOG is a relatively small protein, with only whether there was epitope spreading. Second, we examined whether 125 extracellular amino acids (44). At the outermost periphery of those patients with a rapid decline of anti-MOG IgG were still able the epitope, R52 of the 8-18C5 H chain contacts the MOG amino to mount a persisting IgG response to other Ags. acids Y40 and S45 (27), making clashes with Abs that bind the We found that the recognized epitopes remained constant. This nearby amino acid 42 to the center of their paratope highly likely. was seen not only in patients with ADEM who rapidly lose their For this reason, binding of an mAb can inhibit binding of anti- anti-MOG Abs, but also in childhood patients with MS and CRION MOG autoantibodies, even if they recognize different distant who have anti-MOG Abs persisting for years. Therefore, we find loops of MOG. Therefore, specific mutation of single amino acids that for MOG-Abs there is neither intramolecular epitope spreading gives a much more detailed and correct insight into epitope rec- nor epitope loss. Autoantibody epitope spreading has been reported ognition, whereas competition analysis might be misleading. for a number of autoimmune targets, among them anti-AChR Abs Similar results were found for anti-acetylcholine receptor (AChR) in myasthenia gravis (54), anti-mitochondrial Abs in primary autoantibodies. Anti-AChR mAbs compete with human autoanti- biliary cirrhosis (55), and anti-citrullinated protein Abs in rheu- bodies (45). Further epitope mapping revealed that some human matoid arthritis (56). anti-AChR autoantibodies recognized different epitopes than the The maintenance of serum IgG is crucial for our ability to fight competing mAbs; the area covered by one single mAb was large pathogens (57). The persistence of serum IgG is based on long- enough to allow blocking of human autoantibodies recognizing lived plasma cells that find a survival niche in the bone marrow or distant epitopes (46). inflamed tissue (58). We found that those patients, who rapidly The Journal of Immunology 3603 lost their anti-MOG IgG, were still able to mount a persisting IgG 16. Kitley, J., M. Woodhall, P. Waters, M. I. Leite, E. Devenney, J. Craig, J. Palace, and A. Vincent. 2012. Myelin-oligodendrocyte glycoprotein antibodies in adults response against two pathogens: measles and rubella virus. This with a neuromyelitis optica phenotype. Neurology 79: 1273–1277. finding shows that these patients do not have a general inability to 17. Lalive, P. H., M. G. Ha¨usler, H. Maurey, Y. Mikaeloff, M. Tardieu, H. Wiendl, generate long-lived plasma cells. Instead, it suggests that the IgG- M. Schroeter, H. P. Hartung, B. C. Kieseier, and T. Menge. 2011. Highly reactive anti-myelin oligodendrocyte glycoprotein antibodies differentiate demyelinating secreting cells generated during their anti-MOG response are less diseases from viral encephalitis in children. Mult. Scler. 17: 297–302. competent to seed the survival niches than are plasma cells gen- 18. Mader, S., V. Gredler, K. Schanda, K. Rostasy, I. Dujmovic, K. Pfaller, erated after vaccination. A. Lutterotti, S. Jarius, F. Di Pauli, B. Kuenz, et al. 2011. Complement activating antibodies to myelin oligodendrocyte glycoprotein in neuromyelitis optica and This study shows that mutation of the single amino acid P42 in related disorders. J. Neuroinflammation 8: 184. the membrane proximal CC9-loop disrupts MOG recognition in the 19. McLaughlin, K. A., T. Chitnis, J. Newcombe, B. Franz, J. Kennedy, S. McArdel, J. Kuhle, L. Kappos, K. Rostasy, D. Pohl, et al. 2009. Age-dependent B cell majority of patient sera. We dissect seven patterns that describe autoimmunity to a myelin surface antigen in pediatric multiple sclerosis. J. the epitopes of 98 of 111 patients with anti-MOG Abs. Notably, Immunol. 183: 4067–4076. the individual epitope patterns remained constant over time without 20. Pro¨bstel, A. K., K. Dornmair, R. Bittner, P. Sperl, D. Jenne, S. Magalhaes, A. Villalobos, C. Breithaupt, R. Weissert, U. Jacob, et al. 2011. Antibodies to evidence for intramolecular epitope spreading. This study has MOG are transient in childhood acute disseminated encephalomyelitis. Neu- implications for the design of transfer studies in animals, anti- rology 77: 580–588. hMOG detection assays, and future therapies aiming at Ag-specific 21. Rosta´sy, K., S. Mader, E. Hennes, K. Schanda, V. Gredler, A. Guenther, A. Blaschek, C. Korenke, M. Pritsch, D. Pohl, et al. 2013. Persisting myelin depletion of MOG-reactive Abs. oligodendrocyte glycoprotein antibodies in aquaporin-4 antibody negative pe- diatric neuromyelitis optica. Mult. Scler. 19: 1052–1059. 22. Rostasy, K., S. Mader, K. Schanda, P. Huppke, J. Ga¨rtner, V. Kraus, M. Karenfort, Acknowledgments D. Tibussek, A. Blaschek, B. Bajer-Kornek, et al. 2012. Anti-myelin oligoden- We thank Robert Bittner, Petra Sperl, and Heike Ru¨bsamen for excellent drocyte glycoprotein antibodies in pediatric patients with optic neuritis. Arch. Downloaded from technical assistance. Dr. Gurumoorthy Krishnamoorthy generously pro- Neurol. 69: 752–756. 23. O’Connor, K. C., K. A. McLaughlin, P. L. De Jager, T. Chitnis, E. Bettelli, vided the mMOG DNA and together with Dr. Naoto Kawakami gave C. Xu, W. H. Robinson, S. V. Cherry, A. Bar-Or, B. Banwell, et al. 2007. Self- valuable comments on the manuscript. antigen tetramers discriminate between myelin autoantibodies to native or denatured protein. Nat. Med. 13: 211–217. 24. Breithaupt, C., B. Scha¨fer, H. Pellkofer, R. Huber, C. Linington, and U. Jacob. Disclosures 2008. Demyelinating myelin oligodendrocyte glycoprotein-specific autoantibody

The authors have no financial conflicts of interest. response is focused on one dominant conformational epitope region in rodents. http://www.jimmunol.org/ J. Immunol. 181: 1255–1263. 25. Gardinier, M. V., and J. M. Matthieu. 1993. Cloning and cDNA sequence analysis of myelin/oligodendrocyte glycoprotein: a novel member of the im- References munoglobulin superfamily. Schweiz. Arch. Neurol. Psychiatr. 144: 201– 1. Vincent, A., C. G. Bien, S. R. Irani, and P. Waters. 2011. Autoantibodies asso- 207. ciated with diseases of the CNS: new developments and future challenges. 26. Clements, C. S., H. H. Reid, T. Beddoe, F. E. Tynan, M. A. Perugini, T. G. Johns, Lancet Neurol. 10: 759–772. C. C. Bernard, and J. Rossjohn. 2003. The crystal structure of myelin oligo- 2. Lancaster, E., and J. Dalmau. 2012. Neuronal autoantigens—pathogenesis, as- dendrocyte glycoprotein, a key autoantigen in multiple sclerosis. Proc. Natl. sociated disorders and antibody testing. Nat. Rev. Neurol. 8: 380–390. Acad. Sci. USA 100: 11059–11064. 3. Iorio, R., and V. A. Lennon. 2012. Neural antigen-specific autoimmune dis- 27. Breithaupt, C., A. Schubart, H. Zander, A. Skerra, R. Huber, C. Linington, and orders. Immunol. Rev. 248: 104–121. U. Jacob. 2003. Structural insights into the antigenicity of myelin oligoden-

4. Krumbholz, M., T. Derfuss, R. Hohlfeld, and E. Meinl. 2012. B cells and antibodies drocyte glycoprotein. Proc. Natl. Acad. Sci. USA 100: 9446–9451. by guest on September 28, 2021 in multiple sclerosis pathogenesis and therapy. Nat. Rev. Neurol. 8: 613–623. 28. Schwede, T., J. Kopp, N. Guex, and M. C. Peitsch. 2003. SWISS-MODEL: An 5. Srivastava, R., M. Aslam, S. R. Kalluri, L. Schirmer, D. Buck, B. Tackenberg, automated protein homology-modeling server. Nucleic Acids Res. 31: 3381– V. Rothhammer, A. Chan, R. Gold, A. Berthele, et al. 2012. 3385. KIR4.1 as an immune target in multiple sclerosis. N. Engl. J. Med. 367: 115– 29. Barton, G. J. 1993. ALSCRIPT: a tool to format multiple sequence alignments. 123. Protein Eng. 6: 37–40. 6. Quintana, F. J., M. F. Farez, V. Viglietta, A. H. Iglesias, Y. Merbl, G. Izquierdo, 30. Linnington, C., M. Webb, and P. L. Woodhams. 1984. A novel myelin-associated M. Lucas, A. S. Basso, S. J. Khoury, C. F. Lucchinetti, et al. 2008. Antigen glycoprotein defined by a mouse monoclonal antibody. J. Neuroimmunol. 6: microarrays identify unique serum autoantibody signatures in clinical and 387–396. pathologic subtypes of multiple sclerosis. Proc. Natl. Acad. Sci. USA 105: 31. Piddlesden, S. J., H. Lassmann, F. Zimprich, B. P. Morgan, and C. Linington. 18889–18894. 1993. The demyelinating potential of antibodies to myelin oligodendrocyte 7. Brunner, C., H. Lassmann, T. V. Waehneldt, J. M. Matthieu, and C. Linington. glycoprotein is related to their ability to fix complement. Am. J. Pathol. 143: 1989. Differential ultrastructural localization of myelin basic protein, myelin/ 555–564. oligodendroglial glycoprotein, and 29,39-cyclic nucleotide 39-phosphodiesterase 32. Ramaraj, T., T. Angel, E. A. Dratz, A. J. Jesaitis, and B. Mumey. 2012. Antigen- in the CNS of adult rats. J. Neurochem. 52: 296–304. antibody interface properties: composition, residue interactions, and features of 8. Mayer, M. C., and E. Meinl. 2012. Glycoproteins as targets of autoantibodies in 53 non-redundant structures. Biochim. Biophys. Acta 1824: 520–532. CNS inflammation: MOG and more. Ther. Adv. Neurol. Disord. 5: 147–159. 33. Thornton, J. M., M. S. Edwards, W. R. Taylor, and D. J. Barlow. 1986. Location 9. Linington, C., M. Bradl, H. Lassmann, C. Brunner, and K. Vass. 1988. Aug- of ‘continuous’ antigenic determinants in the protruding regions of proteins. mentation of demyelination in rat acute allergic encephalomyelitis by circulating EMBO J. 5: 409–413. mouse monoclonal antibodies directed against a myelin/oligodendrocyte gly- 34. Lennon, V. A., D. M. Wingerchuk, T. J. Kryzer, S. J. Pittock, C. F. Lucchinetti, coprotein. Am. J. Pathol. 130: 443–454. K. Fujihara, I. Nakashima, and B. G. Weinshenker. 2004. A serum autoantibody 10. Schluesener, H. J., R. A. Sobel, C. Linington, and H. L. Weiner. 1987. A marker of neuromyelitis optica: distinction from multiple sclerosis. Lancet 364: monoclonal antibody against a myelin oligodendrocyte glycoprotein induces 2106–2112. relapses and demyelination in central nervous system autoimmune disease. J. 35. Bennett, J. L., C. Lam, S. R. Kalluri, P. Saikali, K. Bautista, C. Dupree, Immunol. 139: 4016–4021. M. Glogowska, D. Case, J. P. Antel, G. P. Owens, et al. 2009. Intrathecal 11. Genain, C. P., M. H. Nguyen, N. L. Letvin, R. Pearl, R. L. Davis, M. Adelman, pathogenic anti-aquaporin-4 antibodies in early neuromyelitis optica. Ann. M. B. Lees, C. Linington, and S. L. Hauser. 1995. Antibody facilitation of multiple Neurol. 66: 617–629. sclerosis-like lesions in a nonhuman primate. J. Clin. Invest. 96: 2966–2974. 36. Bradl, M., T. Misu, T. Takahashi, M. Watanabe, S. Mader, M. Reindl, 12. Brehm, U., S. J. Piddlesden, M. V. Gardinier, and C. Linington. 1999. Epitope M. Adzemovic, J. Bauer, T. Berger, K. Fujihara, et al. 2009. Neuromyelitis optica: specificity of demyelinating monoclonal autoantibodies directed against the human pathogenicity of patient immunoglobulin in vivo. Ann. Neurol. 66: 630–643. myelin oligodendrocyte glycoprotein (MOG). J. Neuroimmunol. 97: 9–15. 37. Saadoun, S., P. Waters, B. A. Bell, A. Vincent, A. S. Verkman, and 13. von Bu¨dingen, H. C., S. L. Hauser, J. C. Ouallet, N. Tanuma, T. Menge, and M. C. Papadopoulos. 2010. Intra-cerebral injection of neuromyelitis optica im- C. P. Genain. 2004. Frontline: Epitope recognition on the myelin/oligodendrocyte munoglobulin G and human complement produces neuromyelitis optica lesions glycoprotein differentially influences disease phenotype and antibody effector in mice. Brain 133: 349–361. functions in autoimmune demyelination. Eur. J. Immunol. 34: 2072–2083. 38. Kinoshita, M., Y. Nakatsuji, T. Kimura, M. Moriya, K. Takata, T. Okuno, 14. Brilot, F., R. C. Dale, R. C. Selter, V. Grummel, S. R. Kalluri, M. Aslam, A. Kumanogoh, K. Kajiyama, H. Yoshikawa, and S. Sakoda. 2009. Neuro- V. Busch, D. Zhou, S. Cepok, and B. Hemmer. 2009. Antibodies to native myelin myelitis optica: Passive transfer to rats by human immunoglobulin. Biochem. oligodendrocyte glycoprotein in children with inflammatory demyelinating Biophys. Res. Commun. 386: 623–627. central nervous system disease. Ann. Neurol. 66: 833–842. 39. Zhou, D., R. Srivastava, S. Nessler, V. Grummel, N. Sommer, W. Bru¨ck, 15. Di Pauli, F., S. Mader, K. Rostasy, K. Schanda, B. Bajer-Kornek, R. Ehling, H. P. Hartung, C. Stadelmann, and B. Hemmer. 2006. Identification of a patho- F. Deisenhammer, M. Reindl, and T. Berger. 2011. Temporal dynamics of anti- genic antibody response to native myelin oligodendrocyte glycoprotein in mul- MOG antibodies in CNS demyelinating diseases. Clin. Immunol. 138: 247–254. tiple sclerosis. Proc. Natl. Acad. Sci. USA 103: 19057–19062. 3604 DISTINCTION OF MOG EPITOPES

40. von Bu¨dingen, H. C., S. L. Hauser, A. Fuhrmann, C. B. Nabavi, J. I. Lee, and 49. Khalil, M., M. Reindl, A. Lutterotti, B. Kuenz, R. Ehling, C. Gneiss, P. Lackner, C. P. Genain. 2002. Molecular characterization of antibody specificities against F. Deisenhammer, and T. Berger. 2006. Epitope specificity of serum antibodies myelin/oligodendrocyte glycoprotein in autoimmune demyelination. Proc. Natl. directed against the extracellular domain of myelin oligodendrocyte glycopro- Acad. Sci. USA 99: 8207–8212. tein: Influence of relapses and immunomodulatory treatments. J. Neuroimmunol. 41. Marta, C. B., A. R. Oliver, R. A. Sweet, S. E. Pfeiffer, and N. H. Ruddle. 2005. 174: 147–156. Pathogenic myelin oligodendrocyte glycoprotein antibodies recognize glycosy- 50. Zielinska, D. F., F. Gnad, J. R. Wisniewski, and M. Mann. 2010. Precision lated epitopes and perturb oligodendrocyte physiology. Proc. Natl. Acad. Sci. mapping of an in vivo N-glycoproteome reveals rigid topological and sequence USA 102: 13992–13997. constraints. Cell 141: 897–907. 42. Henry, R. A., P. L. Kendall, and J. W. Thomas. 2012. Autoantigen-specific B-cell 51. Koch, M., M. Pancera, P. D. Kwong, P. Kolchinsky, C. Grundner, L. Wang, depletion overcomes failed immune tolerance in type 1 diabetes. Diabetes 61: W. A. Hendrickson, J. Sodroski, and R. Wyatt. 2003. Structure-based, targeted 2037–2044. deglycosylation of HIV-1 gp120 and effects on neutralization sensitivity and 43. Tradtrantip, L., H. Zhang, S. Saadoun, P. W. Phuan, C. Lam, M. C. Papadopoulos, antibody recognition. Virology 313: 387–400. J. L. Bennett, and A. S. Verkman. 2012. Anti-aquaporin-4 monoclonal antibody 52. Xiao, Q., K. Yu, Y. Y. Cui, and H. C. Hartzell. 2009. Dysregulation of human blocker therapy for neuromyelitis optica. Ann. Neurol. 71: 314–322. bestrophin-1 by ceramide-induced dephosphorylation. J. Physiol. 587: 4379– 44. Kroepfl, J. F., L. R. Viise, A. J. Charron, C. Linington, and M. V. Gardinier. 4391. 1996. Investigation of myelin/oligodendrocyte glycoprotein membrane topology. 53. Blom, N., S. Gammeltoft, and S. Brunak. 1999. Sequence and structure-based J. Neurochem. 67: 2219–2222. prediction of eukaryotic protein phosphorylation sites. J. Mol. Biol. 294: 1351– 45. Tzartos, S. J., M. E. Seybold, and J. M. Lindstrom. 1982. Specificities of anti- 1362. bodies to acetylcholine receptors in sera from myasthenia gravis patients mea- 54. Vincent, A., L. Jacobson, and P. Shillito. 1994. Response to human acetylcholine sured by monoclonal antibodies. Proc. Natl. Acad. Sci. USA 79: 188–192. receptor alpha 138-199: determinant spreading initiates autoimmunity to self- 46. Luo, J., P. Taylor, M. Losen, M. H. de Baets, G. D. Shelton, and J. Lindstrom. antigen in rabbits. Immunol. Lett. 39: 269–275. 2009. Main immunogenic region structure promotes binding of conformation- 55. Mori, T., H. Ohira, M. Kuroda, M. Kato, Y. Yamaguchi, and H. Kochi. 2012. dependent myasthenia gravis autoantibodies, nicotinic acetylcholine receptor Characterization of Autoantibodies against the E1a Subunit of Branched-Chain conformation maturation, and agonist sensitivity. J. Neurosci. 29: 13898–13908. 2-Oxoacid Dehydrogenase in Patients with Primary Biliary Cirrhosis. Int. J. 47. Xu, Y., Y. Zhang, C. Y. Liu, B. Peng, J. M. Wang, X. J. Zhang, H. F. Li, and Hepatol. 2012: 369740. L. Y. Cui. 2012. Serum antibodies to 25 myelin oligodendrocyte glycoprotein 56. Sokolove, J., R. Bromberg, K. D. Deane, L. J. Lahey, L. A. Derber, Downloaded from epitopes in multiple sclerosis and neuromyelitis optica: clinical value for diag- P. E. Chandra, J. D. Edison, W. R. Gilliland, R. J. Tibshirani, J. M. Norris, et al. nosis and disease activity. Chin. Med. J. (Engl.) 125: 3207–3210. 2012. Autoantibody epitope spreading in the pre-clinical phase predicts pro- 48. Haase, C. G., J. Guggenmos, U. Brehm, M. Andersson, T. Olsson, M. Reindl, gression to rheumatoid arthritis. PLoS ONE 7: e35296. J. M. Schneidewind, U. K. Zettl, F. Heidenreich, T. Berger, et al. 2001. The fine 57. Amanna, I. J., N. E. Carlson, and M. K. Slifka. 2007. Duration of humoral im- specificity of the myelin oligodendrocyte glycoprotein autoantibody response in munity to common viral and vaccine antigens. N. Engl. J. Med. 357: 1903–1915. patients with multiple sclerosis and normal healthy controls. J. Neuroimmunol. 58. Manz, R. A., A. E. Hauser, F. Hiepe, and A. Radbruch. 2005. Maintenance of 114: 220–225. serum antibody levels. Annu. Rev. Immunol. 23: 367–386. http://www.jimmunol.org/ by guest on September 28, 2021