Demyelinating Myelin Oligodendrocyte Glycoprotein-Specific Response Is Focused on One Dominant Conformational Region in Rodents This information is current as of September 28, 2021. Constanze Breithaupt, Beatrix Schäfer, Hannah Pellkofer, Robert Huber, Christopher Linington and Uwe Jacob J Immunol 2008; 181:1255-1263; ; doi: 10.4049/jimmunol.181.2.1255 http://www.jimmunol.org/content/181/2/1255 Downloaded from

References This article cites 59 articles, 15 of which you can access for free at:

http://www.jimmunol.org/content/181/2/1255.full#ref-list-1 http://www.jimmunol.org/

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 by guest on September 28, 2021

*average

Subscription Information about subscribing to The Journal of 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 © 2008 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Demyelinating Myelin Oligodendrocyte Glycoprotein-Specific Autoantibody Response Is Focused on One Dominant Conformational Epitope Region in Rodents1,2

Constanze Breithaupt,3* Beatrix Scha¨fer,* Hannah Pellkofer,†‡ Robert Huber,*§¶ Christopher Linington,ʈ and Uwe Jacob#

Conformational of myelin oligodendrocyte glycoprotein (MOG) provide a major target for demyelinating autoan- tibodies in experimental autoimmune encephalomyelitis and recent studies indicate that a similar situation may exist in multiple sclerosis. We recently solved the crystal structure of the extracellular domain of MOG (MOGex) in complex with a Fab derived from the demyelinating mAb 8-18C5 and identified the conformational 8-18C5 epitope on MOG that is dominated by the surface exposed FG loop of MOG. To determine the importance of this epitope with regard to the Downloaded from 103 104 polyclonal Ab response to MOGex we investigated the effects of mutating His and Ser , the two central amino acids of the FG loop, on Ab binding. Mutation of these two residues reduced binding of a panel of eight demyelinating conformation- dependent mAbs to <20% compared with binding to wild-type MOGex, whereas substitution of amino acids that do not contribute to the 8-18C5 epitope had only a minor effect on Ab binding. The same restriction was observed for the polyclonal MOG-specific Ab response of MOG DNA-vaccinated BALB/c and SJL/J mice. Our data demonstrate that the pathogenic anti-MOG Ab response primarily targets one immunodominant region centered at the FG loop of MOG. Comparison of the http://www.jimmunol.org/ structure of MOGex with the structures of related IgV-like domains yields a possible explanation for the focused Ab response. The Journal of Immunology, 2008, 181: 1255–1263.

yelin oligodendrocyte glycoprotein (MOG)4 is a but the formation of large demyelinating lesions is dependent on highly conserved transmembrane myelin that Ab-dependent effector mechanisms triggered by MOG-specific au- M is expressed exclusively in the CNS (1, 2). It was iden- toantibodies (5–7). tified originally as an immunodominant target for demyelinating Ab-mediated demyelination in MOG-EAE reproduces many in a guinea pig model of experimental autoimmune features associated with demyelination in MS including the lo- by guest on September 28, 2021 encephalomyelitis (EAE) (3, 4). Subsequently, MOG was shown cal deposition of Igs and complement activation products in to induce a variant of EAE that exhibits many of the clinical and areas of active demyelination. These similarities together with pathological characteristics of multiple sclerosis (MS) in both rats the exposed location of MOG at the surface of CNS myelin and and primates. In these species, inflammation and blood brain bar- the association of MOG-specific autoantibodies with myelin de- rier dysfunction are initiated by the MOG-specific response, bris identify MOG as a prime candidate for demyelinating Abs in MS (8–10). Yet, the frequency and clinical significance of MOG-specific autoantibodies in MS remains controversial. *Max-Planck-Institut fu¨r Biochemie, Abteilung Strukturforschung/Emeritusgruppe Huber, Martinsried, Germany; †Max-Planck-Institut fu¨r Neurobiologie, Abteilung Some laboratories report that MS is associated with signifi- Neuroimmunologie, Martinsried, Germany; ‡Institut fu¨r Klinische Neuroimmunolo- cantly elevated titers of MOG-specific autoantibodies (7, 11– gie, Klinikum Grosshadern, Ludwig Maximilians Universita¨t, Mu¨nchen, Germany; §School of Biosciences, Cardiff University, Cardiff, United Kingdom; ¶Zentrum fu¨r 16), while others find no significant difference between MS pa- Medizinische Biotechnologie, Universita¨t Duisburg, Essen, Germany; ʈGlasgow Bio- tients, patients with other inflammatory neurological diseases, medical Research Centre, University of Glasgow, Glasgow, United Kingdom; and or even healthy control donors (17–21). Recent studies suggest #SuppreMol, Martinsried, Germany that this lack of concordance between different studies is due to Received for publication May 15, 2007. Accepted for publication May 8, 2008. the inability of most assay methods to differentiate between The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance pathogenic (demyelinating) and nonpathogenic MOG-specific with 18 U.S.C. Section 1734 solely to indicate this fact. Abs. Experimental animals mount a complex Ab response to the 1 The atomic coordinates and structure factors of DM2-MOGex presented in this ar- extracellular domain of MOG (MOGex) that targets both linear ticle have been deposited in the Protein Data Bank (PDB; www.rcsb.org) with the as well as discontinuous, conformational epitopes (22–24). Cru- PDB ID code 3CSP. cially, only conformation-dependent MOG-specific Abs are 2 This work was supported by the Gemeinnu¨tzige Hertie Stiftung (1.01.1/04/008, to U.J.). able to initiate demyelination in EAE (22–24). In contrast, Abs 3 Address correspondence and reprint requests to Dr. Constanze Breithaupt at the directed against linear MOGex-peptide epitopes are unable to current address: Institute of Biochemistry and Biotechnology, Martin Luther Univer- bind to the native protein at the membrane surface and fail to sity Halle-Wittenberg, Kurt Mothes Strasse 3, 06120 Halle, Germany. E-mail address: [email protected] induce any significant pathology in vivo. Clinical studies sug- gest that a similar situation exists in humans. MOG-specific Ab 4 Abbreviations used in this paper: MOG, myelin oligodendrocyte glycoprotein; EAE, experimental autoimmune encephalomyelitis; MS, multiple sclerosis; MOGex, extra- responses directed against discontinuous, conformational cellular domain of MOG; hMOG, human MOG; BTN, butyrophilin; ERMAP, ery- epitopes are present in a subset of MS patients (10, 16, 17, 25), throid membrane-associated protein. while autoantibodies to linear peptide MOGex epitopes are Copyright © 2008 by The American Association of Immunologists, Inc. 0022-1767/08/$2.00 found in both MS patients and healthy controls. www.jimmunol.org 1256 SPECIFICITY OF CONFORMATION-DEPENDENT ANTI-MOG Abs

a In view of the importance of conformation-dependent MOGex- Table I. Data collection and refinement statistics specific Abs in immune-mediated demyelination, we recently de- termined the three-dimensional structure of the complex formed Data Collection between MOGex and a Fab derived from the demyelinating mouse Space group P3221 MOG-specific mAb 8-18C5 (26). This structural analysis con- Unit cell dimensions (Å) a ϭ b ϭ 49.24, c ϭ 77.60 firmed that the epitope recognized by 8-18C5 is highly discontin- Resolution (Å) 1.70 (1.70–1.80) Observed reflections 97,717 uous and identified the surface-exposed FG loop of MOGex as a Independent reflections 12,373 major component of the epitope. Completeness (%) 99.3 (99.9) b We have now used site-directed mutagenesis to investigate the Rsym (%) 4.1 (32.7) ϽIϾ/Ͻ␴(I)Ͼ 29.4 (6.7) importance of interactions that involve this region of MOG within 2 Ͼ B value of Wilson plot (Å ) 22.8 the polyclonal MOG-specific Ab repertoire. We report that 80% Refinement c of Ag recognition by conformation-dependent MOGex-specific Rcryst/Rfree (%) 19.9/24.5 mAbs is abolished by mutation of two amino acids (His103 and No. of atoms: protein/water 972/139 104 Rms deviation: bonds (Å)/angles (°)/ 0.011/1.66/1.42 Ser ) at the tip of the FG loop. This highly focused response is 2 d also seen for the MOG -specific Ab response in BALB/c and bonded B’s (Å ) ex Mean B value: protein/water (Å2) 23.10/34.2 SJL/J mice after MOG-DNA vaccination. This study identifies a a Values in parentheses correspond to the highest resolution shell. key immunodominant epitope within the demyelinating MOG- b ϭ⌺ ϪϽ Ͼ ⌺ Rsym hi Ii(h) I(h) / hiIi(h). c ϭ ⌺ ࿣ Ϫ ࿣ ⌺ specific Ab repertoire that is centered at the FG loop of MOG. Rcryst 5 h Fo(h) Fc(h) / h Fo(h). Rfree was determined from 5% of the

data that were not used for refinement. Downloaded from Comparison of the structure of MOGex with the structures of d Root mean square deviation of bonded B’s: root mean square deviation of tem- related Ig-V like domains of other reveals the lack of perature factors of bonded atoms. structural conservation in this epitope region suggesting a pos- sible explanation for its immunodominance within the MOGex- specific Ab repertoire.

group P3221 and contain one molecule per asymmetric unit like the wild- Materials and Methods type crystals. Diffraction data of the crystals frozen in mother liquor sup- http://www.jimmunol.org/ Design of mutant proteins and site-directed mutagenesis plied with 20% butanediol as cryoprotectant were collected at the BESSY synchrotron (beamline MX14-2, ␭ ϭ 0.91841 Å) using a MAR CCD Intermolecular contacts were analyzed using programs of the CCP4 165-mm detector. Data were integrated and scaled using XDS (32). The package (27) and the model building program O (28). Electrostatic structure of DM2-MOGex was determined by molecular replacement using potentials were calculated in GRASP (29) by using atomic charges the program PHASER (33) and wild-type MOGex (Protein Data Bank ID according to Weiner et al. (30). The solvent-accessible surface of 1PKO) as a search model. Refinement was performed by alternated model MOGex was calculated with the utility SURFACE of the CCP4 program building using the program O (28) and crystallographic refinement cycles package. Mutagenesis was conducted using the extracellular domain of rat with CNS (34) that included simulated annealing, positional refinement, MOG (MOGex) subcloned into the His-tag expression vector pQE-12 (26) and restrained temperature factor refinement using the parameters of Engh

by following the method of “QuikChange Site-Directed Mutagenesis” by and Huber (35). Analysis of the structure was performed with programs of by guest on September 28, 2021 Stratagene. The oligonucleotides used were: 5Ј-CTTCAGAGACCACGAA the CCP4 package (27). Data collection and refinement parameters are TACCAAGAAGAAGCCGCCG-3Ј (SM1), 5Ј-CACATGCTTCTTCAG summarized in Table I. AGACGGCGAATACCAAG-3Ј (DM1), 5Ј-CACATGCTTCTTCAGAGA CGCTGAATACCAAG-3Ј (DM2), and the corresponding reverse complementary oligonucleotides. The identity of the mutations was veri- DNA vaccination fied by DNA sequencing of the purified plasmids. Large amounts of the expression vector pcDNA3.1(Ϫ) (Invitrogen) that Ϫ Protein expression and refolding of rMOG was used as a control and of pcDNA3.1( ) containing the coding sequence ex of full-length MOG and its signal sequence (Bourquin et al. (36)) were Plasmids containing the extracellular domain of human MOG and the prepared using Qiagen Endotoxin-Free Plasmid kits (Qiagen). Female 42 “humanized” rat MOGex mutant Ser Pro were a gift of N. Ruddle (Yale SJL/J and BALB/c mice and Dark Agouti rats were purchased from University, New Haven, CT) (31). The extracellular domain of rat and Charles River Laboratories. At the age of 4–5 wk, plasmid DNA in PBS human MOG and the mutant proteins were overexpressed in inclusion (mice, 100 ␮g; rats, 400 ␮g) was injected into the tibialis anterior muscle. bodies in Escherichia coli. After disruption of the cells by sonification, Plasma was prepared by centrifugation and stored at Ϫ20°C. All animal the inclusion bodies were purified by repetitive steps of centrifugation experiments were performed according to the Bavarian state regulations for and resuspension in 50 mM Tris-HCl (pH 8.0), 0.3 M NaCl, 0.5% animal experimentation, and were approved by the responsible authorities. lauryldimethylamine oxide. The inclusion bodies were solubilized in solubilization buffer (100 mM NaH2PO4, 10 mM Tris, 6 M guanidinium chloride, 40 mM mercaptoethanol (pH 8.0)). After dilution in mercap- ELISA toethanol-free solubilization buffer, the denatured MOG was bound to ex Ab binding to MOG and to the mutant proteins was measured by ELISA. Ni-NTA Superflow (Qiagen) material and refolded on the column in two ex The mouse mAbs 8-18C5 (4), Y1, Y8, Y9, Y10, Y11, Z2, Z4, Z8, and Z12 steps. At first, a linear gradient from solubilization buffer (1 mM mercap- (37) were purified from hybridoma supernatants by affinity chromatogra- toethanol) to 100 mM NaH PO , 10 mM Tris, 3 mM glutathione (pH 8.0) 2 4 phy on Protein G. Their concentration was estimated by UV/Vis spectros- over 10 h and 80-column volumes was applied, followed by a short linear copy and colorimetrically by the Bradford method. Ninety-six-well plates gradient (2 h, 2-column volumes) to remove the glutathione for complete (Maxisorb; Nunc) were coated with 100 ␮lof10␮g/ml Ag in PBS (1 h, oxidation of the refolded MOG . After elution, unfolded and aggregated ex 30°C), washed three times with PBS containing 0.2% Tween 20 and MOG was removed by a final size exclusion chromatography step. Iden- ex blocked with PBS containing 1% w/v BSA (2 h, 30°C). After washing, the tity and integrity of the proteins were checked by mass spectrometry and plates were incubated with the mAbs (ϳ0.5 ␮g/ml in PBS) or the plasma one-dimensional [1H]NMR. Protein concentrations were determined by samples of the MOG-vaccinated mice diluted 1/250 for1hat30°C. The UV/Vis spectroscopy; relative concentrations were determined by the washing procedure was repeated and anti-mouse IgG (FabЈ) , conjugated Bradford protein assay (Bio-Rad). 2 with HRP (Amersham Biosciences), that was diluted 1/10000 in PBS was Crystallization and structure determination of DM2-MOG added and the plates were incubated for1hat30°C. Ab binding was ex detected by oxidation of o-phenylene diamine and quantified by measuring

Crystals of DM2-MOGex were grown at 20°C with the sitting drop vapor the absorbance at 490 nm after stopping the reaction with H2SO4. The diffusion method by mixing 1.5 ␮l of protein (4 mg/ml, in 1/4 PBS) with displayed values correspond to the means of triplicate (plasma samples) 0.8 ␮l of 100 mM Tris-HCl (pH 7.2), 12% PEG8000, and 100 mM mag- and quadruplicate (hybridoma supernatants) measurements of a represen- nesium acetate. The hexagonal crystals of DM2-MOGex belong to space tative experiment. The Journal of Immunology 1257 Downloaded from http://www.jimmunol.org/ by guest on September 28, 2021

FIGURE 1. The discontinuous 8-18C5 epitope on MOG centered at His103 and Ser104. a, Bar diagram of the solvent-accessible surface of MOG per amino acid in Å2. The total height of each bar corresponds to the solvent-accessible surface of unbound MOG; the part of the surface that is covered by 8-18C5 upon binding (regions A–E) corresponds to the black fraction of the bars. Below the diagram the sequence alignment of rat, mouse, and human MOG and the secondary structural elements of rMOG are shown. b, Stereo view of the MOG-(8-18C5) interface. Amino acids of the 8-18C5 epitope on MOG and amino acids of 8-18C5 that contact His103 and Ser104 are shown in ball-and-stick representation. Apart from MOG amino acids 101–108 (region E) in the epitope centre, the N terminus (A) and three other sequence stretches of MOG (B–D) contribute to binding of 8-18C5.c, Surface-and-ribbon representation of 103 104 42 MOGex. Amino acids of the 8-18C5 epitope are colored green except for His (blue) and Ser (orange). In addition, Ser that corresponds to proline in human MOG is shown in pink, Tyr96 (hMOG: Phe) in yellow, and Val111 (hMOG: Met) in turquoise. d, Binding of His103 and Ser104 (green) at the tip of the FG loop (amino acids 102–105) of MOGex (yellow) to the of 8-18C5. The electrostatic potential of 8-18C5 is mapped onto the molecular surface of the paratope region of 8-18C5 (red: negative charge, blue: positive charge). In the mutated proteins SM, DM1, and DM2, Ser104 and His103 are exchanged for Glu and Gly/Ala, respectively.

Comparative structural analysis of MOG relatives Results Searches for related sequences were performed with BLAST (www.ncbi. Two central epitope amino acids of MOG can be mutated nlm.nih.gov) and sequence alignments with GCG (Wisconsin Package ver- without disrupting the overall structure of MOG sion 10.3; Accelrys). The structural alignment was calculated with LSQ The structure of the extracellular Ig-V like domain of MOG MAN (38). Figures were drawn with MOLSCRIPT (39), RASTER3D (40), and GRASP (29). The graphical representation of the sequence alignment (MOGex) in complex with a chimeric Fab derived from mouse was prepared using Alscript (41). mAb 8-18C5 identified a highly discontinuous epitope on MOG 1258 SPECIFICITY OF CONFORMATION-DEPENDENT ANTI-MOG Abs Downloaded from

Ͼ Ϫ FIGURE 2. Structure of mutated MOGex proteins. a, Amide region ( 8 ppm) and aliphatic regions (1.0 to 1.0 ppm) of one-dimensional proton NMR ␣ spectra of wild-type MOGex and of the mutant proteins SM, DM1, and DM2. b, Stereo view of the overlaid C backbone of the MOG mutant DM2 (red), wild-type MOGex (blue, 1PKO) and wild-type MOGex complexed with (8-18C5)-Fab (yellow, 1PKQ). http://www.jimmunol.org/ that is composed of the N terminus, the three upper loops, and two from a size exclusion chromatography column. In addition, one- amino acids of the CCЈ loop (Fig. 1, a and b) (26). In agreement dimensional [1H]NMR spectra of the wild-type and the mutant with the observed cross-reactivity of 8-18C5 between species, proteins are virtually identical and all exhibit typical characteris- those amino acids forming the epitope are strictly conserved be- tics of folded proteins (43), i.e., a large signal dispersion in the tween human, rat, and mouse MOG. Despite the high discontinuity aliphatic region between 1.0 and Ϫ1.0 ppm as well as a signal of the epitope, the major contribution to Ab binding is made by the dispersion beyond 8.5 ppm (Fig. 2a). Thus, as intended, the mu- FG loop and the flanking G-strand amino acids (amino acids 101– tations neither destabilize the three-dimensional structure of

108). These amino acids not only form 10 of the 12 intermolecular MOGex nor provoke global structural changes. To make sure that hydrogen bonds involved in Ab binding but also provide ϳ65% of the mutations do not induce minor structural changes, for example by guest on September 28, 2021 the total contact area. In the center of the epitope, His103 and in the conformation of loops near the site of mutagenesis, we crys- 104 Ser that form the tip of the FG loop are inserted deeply into a tallized the most strongly mutated MOGex protein DM2 and solved cleft of the paratope of 8-18C5 (Fig. 1, c and d) and lose the the three-dimensional structure of DM2 at a resolution of 1.7 Å highest amount of solvent accessible surface upon Ab binding, (Table I). The overall structure and the loop conformations corre- 104 during which Ser becomes completely buried in the Ag-Ab in- spond to wild-type MOGex (Fig. 2b). Apart from the two mutated terface (Fig. 1a). amino acids, the only slight structural changes are found at the N To investigate whether this region is important for Ag recogni- and C terminus of MOG that are flexible in solution and become tion by other MOG-specific Abs, we generated the following mu- ordered in the crystal due to crystal contacts that differ in wild-type 104 tated versions of MOGex: a single mutant SM1 (Ser Glu) and and mutant MOG crystals because of different crystal packing. For two double mutants DM1 (His103Gly, Ser104Glu) and DM2 the same reason, the CCЈ loop is flexible in the wild-type but 103 104 104 (His Ala, Ser Glu) (Fig. 1d). Substituting Ser by glutamate ordered in the mutant structure and the (8-18C5)-Fab-MOGex should significantly perturb binding of 8-18C5 to MOG as it is structure (Fig. 2b). unlikely that the bulky glutamate side chain will fit into the small binding pocket of 8-18C5 occupied by Ser104 of wild-type MOG. In addition, contacts between the anionic glutamate side chain and Binding of monoclonal MOG-specific Abs is abolished by the partially negatively charged surface of 8-18C5 in the vicinity mutating the tip of the FG loop of Ser104 should be unfavorable. The additional mutation of His103 Analyzing the interaction of the MOG mutants with 8-18C5 by in the double mutants results in the loss of two buried hydrogen ELISA revealed that the single mutation Ser104Glu reduces the bonds between MOG and amino acids of 8-18C5. The introduction binding of MOG to 8-18C5 by Ͼ40% (Fig. 3a). The additional of bulkier amino acids at position 103 was avoided because the exchange of His103 by glycine in DM1 and alanine in DM2 re- side chain of this amino acid determines the strained conformation sulted in reduction of the ELISA signal by 97 and 100%, re- of the FG loop (26, 42) and larger side chains might complicate spectively. This confirms the central role of MOG amino acids refolding of MOG. In contrast to other amino acids that contribute 103/104 in binding 8-18C5 as deduced from the crystal struc- 103 104 to the 8-18C5 epitope, the side chains of His and Ser are not ture of the MOGex-(8-18C5) Fab complex. We then tested involved in intramolecular hydrogen bonds and are thus suitable whether these amino acids also play a significant role in Ag targets for mutations that will not perturb the overall structural recognition by a panel of MOG-specific mAbs that were derived integrity and conformation of the protein. from BALB/c mice immunized with MOG or myelin proteins All three mutant proteins could be expressed and refolded under (37). All these mAbs recognize native glycosylated as well as the same conditions as wild-type MOGex and eluted as single deglycosylated MOG from different species in vitro and bind to peaks with nearly the same retention time as the wild-type protein native human MOG expressed on fibroblasts (22, 37). Except for The Journal of Immunology 1259 Downloaded from http://www.jimmunol.org/

FIGURE 3. Epitope specificity of MOG-specific monoclonal mouse FIGURE 4. Epitope specificity of the Ab repertoire of MOG-DNA vac- cinated rodents. a, Time course of the MOG-specific Ab response after Abs. a, Interaction with MOGex and MOGex with the mutated FG loop, as determined by ELISA. Except for Y11, the binding affinity of all Abs to the vaccination of SJL/J and BALB/c mice. MOG-specific Abs can be detected by ELISA using diluted plasma samples during the whole observation pe- double mutants is drastically decreased. b, Interaction with human MOGex 42 riod (14–73 days after vaccination) with a maximum titer detected after and the “humanized” rat MOGex mutant Ser Pro. Ab binding is only 4–6 wk (F/Œ: MOG-DNA-vaccinated SJL/J/BALB/c mice; E/‚: SJL/J/ slightly reduced compared with wild-type rat MOGex. BALB/c mice) vaccinated with control vector. Data of all mice were nor-

malized to the 4-wk plasma sample and values from all mice of one group by guest on September 28, 2021 were averaged with the SD accounting for the individual variability in the the two mAbs Y8 and Z8, all mAbs that we tested enhanced de- development of the MOG-specific Ab response. b, Epitope specificity of myelination in vivo (37). two representative mice (SJL/J, BALB/c; 2 wk after vaccination) and Dark Analysis of the epitope specificity of these mAbs revealed that Agouti rats (4 wk after vaccination) determined by ELISA. Data corre- Յ binding of the mutant MOGex proteins to eight of the nine mAbs spond to the mean of three measurements with SDs A490 0.03. Binding was very similar to that of 8-18C5 (Fig. 3a). Only the interaction to the double mutants DM1 and DM2 is strongly reduced indicating a very with mAb Y11 was not influenced by any of the mutations. For all focused Ab response. the other mAbs, binding to the double mutants was reduced to 0–25% in the ELISA. The strongest effect was seen for Y9 and Յ Y10 whose binding to DM1 as well as DM2 was reduced to 8% nal was reduced by 19% relative to wild-type MOGex for 8-18C5, compared with wild-type MOGex. The single mutant was still rec- 7% for Z2 and only minimally for the remaining mAbs (Fig. 3b). ognized by all eight mAbs although the ELISA signal was weaker The additional substitutions present in hMOG further reduced Ab than that obtained using the wild-type protein, yielding ELISA binding. In the case of 8-18C5, the ELISA signal was decreased by signals in the range of 47–79%. Only Y10 showed no statistically 38%, while the binding of five further mAbs (Z2, Z4, Z8, Z12, Y8) significant reduction in binding. was reduced by 10 to 28%.

42 Ab binding to Ser Pro rat MOGex and human MOGex identifies The response of the whole repertoire of conformation-dependent subtle differences in epitope specificity MOGex-specific Abs in rodents is strongly focused on the FG The amino acid sequences of the IgV-like domains of human and loop of MOG rat MOG exhibit a sequence identity of 92%. The most significant To confirm that the FG loop plays a major role in Ab recognition differences at the molecular surface of the IgV-like domains are the of MOGex, we investigated the epitope specificity patterns of the substitution of rat Gly9 by arginine and rat Ser42 by proline in conformation-dependent polyclonal Ab response to MOG induced 42 human MOGex (Fig. 1a). Ser is positioned directly adjacent to by DNA vaccination in BALB/c and SJL/J mice (36). Whereas 9 the 8-18C5 epitope on MOG whereas Gly is located at a site not active immunization with MOGex generates a mixed repertoire of directly involved in the binding of 8-18C5 (Fig. 1c). To investigate conformation-dependent and peptide-specific anti-MOG Abs that the effects of these structural changes on Ab binding, we deter- greatly complicates analysis of individual components of the rep- mined the ability of the mAbs to bind to the extracellular domain ertoire, and furthermore results in severe EAE, a situation in which 42 of human MOG (hMOG) and the rat MOGex mutant Ser Pro pathogenic conformation-dependent Abs may be preferentially (S42P) (31) by ELISA. In agreement with the structure of the bound within the CNS compartment, MOG-DNA vaccination gen- 42 MOGex-(8-18C5)Fab complex, substitution of Ser by proline erates a strictly conformation-dependent polyclonal Ab response to had only a relatively minor effect on Ab binding. The ELISA sig- native membrane-bound MOG without inducing EAE (36). 1260 SPECIFICITY OF CONFORMATION-DEPENDENT ANTI-MOG Abs

Table II. Epitope specificity of the Ab response of MOG-DNA-vaccinated mice over time, detected by ELISAa

Weeks WT SM DM1 DM2 S42P hM BSA

SJL 2 100 49 (Ϯ22) 17 (Ϯ9) 21 (Ϯ8) 81 (Ϯ8) 63 (Ϯ17) 4 (Ϯ5) 6 100 67 (Ϯ13) 16 (Ϯ13) 17 (Ϯ12) 96 (Ϯ4) 84 (Ϯ5) 5 (Ϯ1) 11 100 70 (Ϯ7) 27 (Ϯ10) 28 (Ϯ10) 101 (Ϯ4) 97 (Ϯ16) 4 (Ϯ6) BALB/c 2 100 33 (Ϯ10) 16 (Ϯ12) 19 (Ϯ10) 84 (Ϯ23) 71 (Ϯ9) 3 (Ϯ2) 6 100 63 (Ϯ13) 21 (Ϯ7) 19 (Ϯ8) 101 (Ϯ3) 83 (Ϯ9) 4 (Ϯ1) 11 100 76 (Ϯ10) 35 (Ϯ8) 29 (Ϯ11) 95 (Ϯ5) 89 (Ϯ8) 1 (Ϯ1)

a Dates correspond to the time after DNA vaccination. Absorption values are normalized to the Ab response targeting wild-type rat MOG and represent the mean of seven mice with the SD accounting for the variability in the epitope specificity of individual mice.

In BALB/c as well as in SJL/J mice, MOG-specific Abs were MOGex also involves a dominant interaction with the FG loop, Y1 detected 2 wk after vaccination. The Ab titer reached a maximum and Y10 additionally recognize linear epitopes delineated by the 4–6 wk after vaccination and then fell to ϳ40% of this value by MOG-derived peptides 63–87 and 76–100 (22). The overlapping 11 wk (Fig. 4a). Analysis of epitope specificity by ELISA using region of these peptides contains the sequence of the tip of the the mutated MOGex proteins revealed that both strains developed exposed DE loop (amino acids 76–79), which is directly adjacent Downloaded from a specificity pattern that closely resembles that of the mAbs (Fig. to the 8-18C5 epitope (26) (Fig. 1c). Besides, the binding of both 4b, Table II). Two weeks after vaccination, the mean ELISA val- mAbs is not influenced by the Ser42Pro mutation (Fig. 3b). Merg- ues obtained using the double mutants were reduced to 19 Ϯ 8% ing these data generates a probable epitope of Y1 and Y10 that (SJL/J) and 17 Ϯ 11% (BALB/c) compared with the wild-type includes the FG loop amino acids His103 and Ser104 and thus over- protein. At this early time point, these data match the averaged laps with the 8-18C5 epitope, but which is shifted toward the back

value for the binding of all 10 mAbs to the double mutants relative sheet of the Ig domain to cover the tip of the DE loop. Interest- http://www.jimmunol.org/ to wild-type protein which was 18%. At the same time the ELISA ingly, the combined antigenic region includes the N-glycosylation signals obtained using the single mutant, the Ser42Pro mutant and site (Asn31) indicating that glycosylation may also influence the the human protein, were significantly reduced. Among these pro- composition of the demyelinating MOG-specific Ab repertoire. teins, binding to the single mutant was affected most, resulting in It should be mentioned that we used rat MOGex to study binding

ELISA signals ranging from 22 to 94% relative to wild-type of mouse Abs. The amino acid sequences of rat and mouse MOGex Ͼ MOGex. However, with time, Ab binding to MOGex mutant pro- are 95% identical (Fig. 1a). All three nonconserved surface res- teins exhibited a definite trend to increase. After 11 wk, the aver- idues that are well-accessible to Abs are replaced by chemically aged absorption values for the double mutants had increased to similar, slightly larger residues in mouse compared with rat

Ϯ Ϯ by guest on September 28, 2021 28 9% (SJL/J) and 32 10% (BALB/c) of the wild-type values MOGex. Therefore, it is most likely that mouse MOGex-specific

(Table II). Similarly, cross-reactivity of hMOG and S42P MOG Abs also bind to rat MOGex which is confirmed by the cross- with wild-type MOGex significantly increased between 2 and 11 reactivity of all 10 mouse mAbs (37), especially of Y1 and Y10 wk postvaccination. These observations suggest that maturation of that seem to bind to the DE loop that harbors two of the three the MOG-specific Ab response is associated with epitope diversi- mutations (Ser75Thr; Gly77Ser). fication, although the FG loop remains the dominant target. Anal- To confirm that the results obtained using MOG-specific mAbs ysis of sera obtained from MOG-DNA-vaccinated Dark Agouti were representative of the polyclonal response to MOGex,wein- rats indicates that the dominant role of the FG loop as a target for vestigated the ability of polyclonal MOG-specific antisera to bind 103 MOGex-specific Abs is not restricted to the mouse. The specificity to the mutated proteins. We found that mutation of His and 104 pattern of the polyclonal response to MOGex in the DA rat resem- Ser had a dramatic effect on Ag recognition by polyclonal bled that of the mouse mAbs; binding to the single mutant was MOGex-specific Abs induced by MOG-DNA vaccination. In two reduced to 70% and the average binding to the double mutants was different mice strains, Ͼ80% of binding was abolished by mutation 103 104 reduced to 30% of wild-type MOGex (Fig. 4b). of His and Ser . These results confirm that the antigenic re- gion centered on the tip of the FG loop is the immunodominant

Discussion target of the MOGex-specific Ab response in these experimental Conformation-dependent MOG-specific autoantibodies mediate animals. However, other minor epitopes contribute to the poly- demyelination both in vitro and in vivo and are implicated in the clonal response to MOGex that may overlap with the immunodom- pathogenesis of MS. However, the epitope specificity of this re- inant region, or be structurally distinct. One of the latter is defined sponse is unknown and the clinical significance of MOG-specific by mAb Y11. This was the only Ab for which Ag recognition was autoantibodies in human disease remains controversial. We now independent of contributions from amino acids His103 and Ser104. report that Ag recognition by demyelinating MOGex-specific au- However, Y11 does bind to amino acids 76–100 (22) that encom- toantibodies in rodents is focused on the two amino acids His103 pass the EF loop (amino acids 86–93), suggesting that its epitope 104 and Ser that are located at the tip of the surface-exposed FG is located at the “bottom” membrane proximal surface of MOGex. loop of MOG. It should be noted that despite the high level of sequence iden- Analysis of the fine specificity of MOG-specific mAbs indicates tity between rat and human MOG, we observed differences in their that these two amino acids are part of a broader antigenic region ability to bind MOG-specific Abs. The surface-exposed amino ac- located at the “top” of MOGex centered on the FG loop. All the ids comprising the 8-18C5 epitope are fully conserved between mAbs used in this study bind to native, glycosylated MOGex as rodent and human MOG. This identity extends to the immediate expressed at the cell surface with six of them (8-18C5, Y8, Y9, Z2, vicinity of the 8-18C5 epitope with the exception of Ser42 that is Z4, Z8, and Z12) recognizing purely discontinuous epitopes (22, substituted by proline in the human protein. This single substitu- 37). However, although the binding of mAbs Y1 and Y10 to tion is sufficient to significantly influence binding of 8-18C5 as The Journal of Immunology 1261 Downloaded from

FIGURE 5. Concordance of MOG amino acids of the immunodominant epitope region with amino acids unique to MOG. The amino acid sequence

alignment contains sequences of human Ig-V domains related to MOG that were aligned to the MOG sequence. Green arrows indicate amino acids http://www.jimmunol.org/ characteristic for the Ig-V fold. Below the alignment the secondary structure elements of rat MOG are shown. Regions that form the 8-18C5 epitope on

MOG are enclosed by blue frames. On the left, the molecular surface of MOGex is displayed with amino acids contacting the Ab 8-18C5 colored from orange to green according to their position in the MOG sequence. In addition, the surface of MOGex is shown with the conservation of amino acid sequences of human BTN (BTN S1(A1)) and hERMAP compared with rat MOG, mapped onto it. Conserved amino acids are not colored, the order yellow–orange–red corresponds to decreasing conservation.

demonstrated using the “humanized” Ser42Pro mutant of rat MOG. erythroid membrane-associated protein (ERMAP) with amino acid Binding is further decreased when human MOG is used as an Ag. sequence identities in the range of 35–50% that can result in mo- by guest on September 28, 2021

This pattern of relative binding was also observed for five other lecular mimicry with MOGex. Intriguingly, multiple sequence mAbs as well as for the polyclonal Ab response to MOG. These alignment of MOGex and related structures reveals that those observations suggest that substitutions affecting internal amino ac- amino acids that form the 8-18C5 epitope as defined by x-ray 96 111 ids such as Tyr Phe and Val Met that are located in proximity crystallography are among the least conserved between MOGex to Ser/Pro42 lead to slight but significant changes of the surface and its homologs expressed outside the CNS (Fig. 5). This corre- structure of MOG—a hypothesis that can only be definitely proven lation becomes even more obvious in three-dimensional space by a three-dimensional structure of human MOG. when the degree of conservation between MOGex and its ho- In general, any surface patch of a protein can be targeted by Abs mologs is mapped onto the molecular surface of MOGex (Fig. 5). (44, 45). Yet, in many cases, the autoantibody repertoire is focused The largest surface patch of nonconserved amino acids corre- on restricted parts of the corresponding autoantigens in vivo (46– sponds strikingly to the 8-18C5 epitope centered about the FG 49). Some of these immunodominant regions of nonsequestered loop. In addition, it contains the tip of the DE loop recognized by autoantigens were identified to be cryptotopes that are normally mAbs Y1 and Y10. Though it has to be kept in mind that in general not exposed to the immune response but become accessible, e.g., single amino acid exchanges can be sufficient for rendering a pro- after disassembly or proteolytic cleavage of the autoantigen, others tein region antigenic, the demonstrated correspondence between were found to be epitopes newly generated by posttranslational antigenicity and lack of conservation is striking and may represent modifications or oxidative damage during apoptosis (50–54). The a simple explanation for the observed epitope specificity of con- restricted epitope specificity of the Ab response to MOGex is sur- formation-dependent MOGex-specific Abs. It should be stressed prising as this autoantigen is sequestered within the CNS where it that this argument applies to the development of the Ab response is unable to influence the selection of the immune repertoire (55– to the intact folded protein, not to responses to short linear epitopes 57). In this case, naive B cells with a wide range of potential MOG that may be generated during protein degradation. Whether this specificities should evade Ag-driven selection and become integral correlation between “molecular foreignness” and antigenicity also components of the repertoire. This leads to the obvious question of applies to other sequestered autoantigens remains to be why then is the epitope specificity of conformation-dependent investigated.

MOGex-specific Abs focused on one main immunogenic region of Although recent studies emphasize the importance of conforma- the protein? Several observations suggest that this may be due to tion-dependent MOGex-specific Abs in MS, consistent data regard- the deletion of certain epitope specificities from the Ab repertoire ing the clinical significance of MOG-specific Abs in MS do not by components of self proteins that mimic the corresponding MOG exist. Apart from methodological difficulties concerning, for ex- epitopes (58, 59). ample, the preparation of the Ag MOG or the selection of patients, The butyrophilin (BTN) gene family provides a large number of assays are often complicated by low sensitivity and specificity. structural homologs of MOGex, like BTN, BTN-like proteins, or Recently, Zhou et al. (25) reported that the mAb 8-18C5 competes 1262 SPECIFICITY OF CONFORMATION-DEPENDENT ANTI-MOG Abs with the binding of sera from MS patients to native MOG ex- 20. Reindl, M., C. Linington, U. Brehm, R. Egg, E. Dilitz, F. Deisenhammer, pressed on the cell surface which suggests that the 8-18C5 epitope W. Poewe, and T. Berger. 1999. against the myelin oligodendrocyte glycoprotein and the myelin basic protein in multiple sclerosis and other neuro- on MOG is also recognized by human MOG-specific Abs. In case logical diseases: a comparative study. Brain 122: 2047–2056. 21. Markovic, M., V. Trajkovic, J. Drulovic, S. Mesaros, N. Stojsavljevic, the specificity of conformation-dependent MOGex-specific Abs in I. Dujmovic, and M. M. Stojkovic. 2003. Antibodies against myelin oligoden- humans is restricted to the same immunodominant epitope region drocyte glycoprotein in the cerebrospinal fluid of multiple sclerosis patients. that we report for rodents, testing the differential Ab reactivity of J. Neurol. Sci. 211: 67–73. sera to mutant vs wild-type MOG may represent an alternative way 22. Brehm, U., S. J. Piddlesden, M. V. Gardinier, and C. Linington. 1999. Epitope specificity of demyelinating monoclonal autoantibodies directed against the hu- to analyze MOG-specific Abs in MS. man myelin oligodendrocyte glycoprotein (MOG). J. Neuroimmunol. 97: 9–15. 23. von Budingen, H. C., S. L. Hauser, A. Fuhrmann, C. B. Nabavi, J. I. Lee, and Acknowledgments C. P. Genain. 2002. Molecular characterization of specificities against Proc. Natl. 42 myelin/oligodendrocyte glycoprotein in autoimmune demyelination. We thank Nancy Ruddle for providing the plasmids of hMOG and Ser Pro Acad. Sci. USA 99: 8207–8212. rat MOG. We are grateful to Tad Holak and Loyola d’Silva for recording 24. von Budingen, H. C., S. L. Hauser, J. C. Ouallet, N. Tanuma, T. Menge, and the one-dimensional proton NMR spectra and to Piotr Neumann for data C. P. Genain. 2004. Frontline: epitope recognition on the myelin/oligodendrocyte collection of DM2 mutant diffraction data. glycoprotein differentially influences disease phenotype and antibody effector functions in autoimmune demyelination. Eur. J. Immunol. 34: 2072–2083. 25. Zhou, D., R. Srivastava, S. Nessler, V. Grummel, N. Sommer, W. Bruck, Disclosures H. P. Hartung, C. Stadelmann, and B. Hemmer. 2006. Identification of a patho- The authors have no financial conflict of interest. genic antibody response to native myelin oligodendrocyte glycoprotein in mul- tiple sclerosis. Proc. Natl. Acad. Sci. USA 103: 19057–19062. 26. Breithaupt, C., A. Schubart, H. Zander, A. Skerra, R. Huber, C. Linington, and

References U. Jacob. 2003. Structural insights into the antigenicity of myelin oligodendro- Downloaded from 1. Gardinier, M. V., P. Amiguet, C. Linington, and J. M. Matthieu. 1992. Myelin/ cyte glycoprotein. Proc. Natl. Acad. Sci. USA 100: 9446–9451. oligodendrocyte glycoprotein is a unique member of the immunoglobulin super- 27. Collaborative Computational Project. 1994. The CCP4 suite–programs for pro- family. J. Neurosci. Res. 33: 177–187. tein crystallography. Acta Crystallogr. D 50: 760–763. 2. Johns, T. G., and C. C. A. Bernard. 1999. The structure and function of myelin 28. Jones, T. A., J. Y. Zou, S. W. Cowan, and M. Kjeldgaard. 1991. Improved oligodendrocyte glycoprotein. J. Neurochem. 72: 1–9. methods for building protein models in electron-density maps and the location of 3. Lebar, R., C. Lubetzki, C. Vincent, P. Lombrail, and J. M. Boutry. 1986. The M2 errors in these models. Acta Crystallogr. A 47: 110–119. autoantigen of central nervous system myelin, a glycoprotein present in oligo- 29. Nicholls, A., K. A. Sharp, and B. Honig. 1991. Protein folding and association– dendrocyte membrane. Clin. Exp. Immunol. 66: 423–434. insights from the interfacial and thermodynamic properties of hydrocarbons. Pro- http://www.jimmunol.org/ 4. Linnington, C., M. Webb, and P. L. Woodhams. 1984. A novel myelin-associated teins 11: 281–296. glycoprotein defined by a mouse . J. Neuroimmunol. 6: 30. Weiner, S. J., P. A. Kollman, D. A. Case, U. C. Singh, C. Ghio, G. Alagona, 387–396. S. Profeta, and P. Weiner. 1984. A new force-field for molecular mechanical 5. Adelmann, M., J. Wood, I. Benzel, P. Fiori, H. Lassmann, J. M. Matthieu, simulation of nucleic-acids and proteins. J. Am. Chem. Soc. 106: 765–784. M. V. Gardinier, and K. Dornmair. 1995. The N-terminal domain of the myelin 31. Oliver, A. R., G. M. Lyon, and N. H. Ruddle. 2003. Rat and human myelin oligodendrocyte glycoprotein (MOG) induces acute demyelinating experimental oligodendrocyte glycoproteins induce experimental autoimmune encephalomy- autoimmune encephalomyelitis in the Lewis rat. J. Neuroimmunol. 63: 17–27. elitis by different mechanisms in C57BL/6 mice. J. Immunol. 171: 462–468. 6. Genain, C. P., M. H. Nguyen, N. L. Letvin, R. Pearl, R. L. Davis, M. Adelman, 32. Kabsch, W. 1993. Automatic processing of rotation diffraction data from crystals M. B. Lees, C. Linington, and S. L. Hauser. 1995. Antibody facilitation of mul- of initially unknown symmetry and cell constants. J. Appl. Crystallogr. 26: tiple sclerosis-like lesions in a nonhuman primate. J. Clin. Invest. 96: 2966–2974. 795–800. 7. Iglesias, A., J. Bauer, T. Litzenburger, A. Schubart, and C. Linington. 2001. T- 33. McCoy, A. J., R. W. Grosse-Kunstleve, P. D. Adams, M. D. Winn, L. C. Storoni,

and B-cell responses to myelin oligodendrocyte glycoprotein in experimental and R. J. Read. 2007. Phaser crystallographic software. J. Appl. Crystallogr. 40: by guest on September 28, 2021 autoimmune encephalomyelitis and multiple sclerosis. Glia 36: 220–234. 658–674. 8. Cross, A. H., J. L. Trotter, and J. A. Lyons. 2001. B cells and antibodies in CNS 34. Brunger, A. T., P. D. Adams, G. M. Clore, W. L. DeLano, P. Gros, demyelinating disease. J. Neuroimmunol. 112: 1–14. R. W. Grosse-Kunstleve, J. S. Jiang, J. Kuszewski, M. Nilges, N. S. Pannu, et al. 9. Genain, C. P., B. Cannella, S. L. Hauser, and C. S. Raine. 1999. Identification of 1998. Crystallography and NMR system: a new software suite for macromolec- autoantibodies associated with myelin damage in multiple sclerosis. Nat. Med. 5: ular structure determination. Acta Crystallogr. D 54: 905–921. 170–175. 35. Engh, R. A., and R. Huber. 1991. Accurate bond and angle parameters for x-ray 10. O’Connor, K. C., H. Appel, L. Bregoli, M. E. Call, I. Catz, J. A. Chan, protein-structure refinement. Acta Crystallogr. A 47: 392–400. N. H. Moore, K. G. Warren, S. J. Wong, D. A. Hafler, and K. W. Wucherpfennig. 36. Bourquin, C., A. Iglesias, T. Berger, H. Wekerle, and C. Linington. 2000. Myelin 2005. Antibodies from inflamed central nervous system tissue recognize myelin oligodendrocyte glycoprotein-DNA vaccination induces antibody-mediated au- oligodendrocyte glycoprotein. J. Immunol. 175: 1974–1982. toaggression in experimental autoimmune encephalomyelitis. Eur. J. Immunol. 11. De March, A. K., M. De Bouwerie, M. N. Kolopp-Sarda, G. C. Faure, 30: 3663–3671. M. C. Bene, and C. C. A. Bernard. 2003. Anti-myelin oligodendrocyte glycop- 37. Piddlesden, S. J., H. Lassmann, F. Zimprich, B. P. Morgan, and C. Linington. rotein B-cell responses in multiple sclerosis. J. Neuroimmunol. 135: 117–125. 1993. The demyelinating potential of antibodies to myelin oligodendrocyte gly- 12. Gaertner, S., K. de Graaf, B. Greve, and R. Weissert. 2004. Antibodies against coprotein is related to their ability to fix complement. Am. J. Pathol. 143: glycosylated native MOG are elevated in patients with multiple sclerosis. Neu- 555–564. rology 63: 2381–2383. 38. Kleywegt, G. J., and T. A. Jones. 1997. Detecting folding motifs and similarities 13. Sun, J., H. Link, T. Olsson, B. G. Xiao, G. Andersson, H. P. Ekre, C. Linington, in protein structures. Methods Enzymol. 277: 525–545. and P. Diener. 1991. T and responses to myelin-oligodendrocyte glycop- 39. Kraulis, P. J. 1991. Molscript–a program to produce both detailed and schematic rotein in multiple sclerosis. J. Immunol. 146: 1490–1495. plots of protein structures. J. Appl. Crystallogr. 24: 946–950. 14. Berger, T., P. Rubner, F. Schautzer, R. Egg, H. Ulmer, I. Mayringer, E. Dilitz, 40. Merritt, E. A., and M. E. P. Murphy. 1994. Raster3d version-2.0–a program for F. Deisenhammer, and M. Reindl. 2003. Antimyelin antibodies as a predictor of photorealistic molecular graphics. Acta Crystallogr. D 50: 869–873. clinically definite multiple sclerosis after a first demyelinating event. N. Engl. J. Med. 349: 139–145. 41. Barton, G. J. 1993. ALSCRIPT: a tool to format multiple sequence alignments. 15. Vojdani, A., E. Vojdani, and E. Cooper. 2003. Antibodies to myelin basic protein, Protein Eng. 6: 37–40. myelin oligodendrocytes peptides, ␣-␤-crystallin, activation and cy- 42. Clements, C. S., H. H. Reid, T. Beddoe, F. E. Tynan, M. A. Perugini, T. G. Johns, tokine production in patients with multiple sclerosis. J. Intern. Med. 254: C. C. A. Bernard, and J. Rossjohn. 2003. The crystal structure of myelin oligo- 363–374. dendrocyte glycoprotein, a key autoantigen in multiple sclerosis. Proc. Natl. 16. Lalive, P. H., T. Menge, C. Delarasse, B. Della Gaspera, D. Pham-Dinh, Acad. Sci. USA 100: 11059–11064. P. Villoslada, H. C. von Budingen, and C. P. Genain. 2006. Antibodies to native 43. Rehm, T., R. Huber, and T. A. Holak. 2002. Application of NMR in structural myelin oligodendrocyte glycoprotein are serologic markers of early inflammation proteomics: screening for proteins amenable to structural analysis. Structure 10: in multiple sclerosis. Proc. Natl. Acad. Sci. USA 103: 2280–2285. 1613–1618. 17. Haase, C. G., J. Guggenmos, U. Brehm, M. Andersson, T. Olsson, M. Reindl, 44. Davies, D. R., and G. H. Cohen. 1996. Interactions of protein with J. M. Schneidewind, U. K. Zettl, F. Heidenreich, T. Berger, et al. 2001. The fine antibodies. Proc. Natl. Acad. Sci. USA 93: 7–12. specificity of the myelin oligodendrocyte glycoprotein autoantibody response in 45. Newman, M. A., C. R. Mainhart, C. P. Mallett, T. B. Lavoie, and S. J. Smith-Gill. patients with multiple sclerosis and normal healthy controls. J. Neuroimmunol. 1992. Patterns of antibody specificity during the BALB/c immune response to 114: 220–225. hen egg white lysozyme. J. Immunol. 149: 3260–3272. 18. Lampasona, V., D. Franciotta, R. Furlan, S. Zanaboni, R. Fazio, E. Bonifacio, 46. Henriksson, E. W., M. Wahren-Herlenius, I. Lundberg, E. Mellquist, and G. Comi, and G. Martino. 2004. Similar low frequency of anti-MOG IgG and I. Pettersson. 1999. Key residues revealed in a major conformational epitope of IgM in MS patients and healthy subjects. Neurology 62: 2092–2094. the U1–70K protein. Proc. Natl. Acad. Sci. USA 96: 14487–14492. 19. Lim, E. T., T. Berger, M. Reindl, C. M. Dalton, K. Fernando, G. Keir, 47. Hudson, B. G., K. Tryggvason, M. Sundaramoorthy, and E. G. Neilson. 2003. E. J. Thompson, D. H. Miller, and G. Giovannoni. 2005. Anti-myelin antibodies Alport’s syndrome, Goodpasture’s syndrome, and type IV collagen. N. Engl. do not allow earlier diagnosis of multiple sclerosis. Mult. Scler. 11: 492–494. J. Med. 348: 2543–2556. The Journal of Immunology 1263

48. Mahler, M., M. Bluthner, and K. M. Pollard. 2003. Advances in B-cell epitope 55. Litzenburger, T., R. Fassler, J. Bauer, H. Lassmann, C. Linington, H. Wekerle, analysis of autoantigens in connective tissue diseases. Clin. Immunol. 107: 65–79. and A. Iglesias. 1998. B producing demyelinating autoantibodies– 49. Zillikens, D., P. A. Rose, S. D. Balding, Z. Liu, M. Olaguemarchan, L. A. Diaz, development and function in gene-targeted transgenic mice. J. Exp. Med. 188: and G. J. Giudice. 1997. Tight clustering of extracellular Bp180 epitopes recog- 169–180. nized by bullous pemphigoid autoantibodies. J. Invest. Dermatol. 109: 573–579. 56. Fazilleau, N., C. Delarasse, C. H. Sweenie, S. M. Anderton, S. Fillatreau, 50. Plotz, P. H. 2003. The autoantibody repertoire: searching for order. Nat. Rev. F. A. Lemonnier, D. Pham-Dinh, and J. M. Kanellopoulos. 2006. Persistence of Immunol. 3: 73–78. 51. Cloos, P. A. C., and S. Christgau. 2004. Post-translational modifications of pro- autoreactive myelin oligodendrocyte glycoprotein (MOG)-specific T cell reper- teins: implications for aging, recognition, and . Biogeron- toires in MOG-expressing mice. Eur. J. Immunol. 36: 533–543. tology 5: 139–158. 57. Delarasse, C., P. Daubas, L. T. Mars, C. Vizler, T. Litzenburger, A. Iglesias, 52. Schellekens, G. A., H. Visser, B. A. de Jong, F. H. van den Hoogen, J. M. Hazes, J. Bauer, B. Della Gaspera, A. Schubart, L. Decker, et al. 2003. Myelin/oligo- F. C. Breedveld, and W. J. van Venrooij. 2000. The diagnostic properties of dendrocyte glycoprotein-deficient (MOG-deficient) mice reveal lack of immune rheumatoid arthritis antibodies recognizing a cyclic citrullinated peptide. Arthritis tolerance to MOG in wild-type mice. J. Clin. Invest. 112: 544–553. Rheum. 43: 155–163. 58. Litzenburger, T., H. Bluthmann, P. Morales, D. Pham-Dinh, A. Dautigny, 53. Schellekens, G. A., B. A. de Jong, F. H. van den Hoogen, L. B. van de Putte, and H. Wekerle, and A. Iglesias. 2000. Development of myelin oligodendrocyte gly- W. J. van Venrooij. 1998. Citrulline is an essential constituent of antigenic de- coprotein autoreactive transgenic B lymphocytes: receptor editing in vivo after terminants recognized by rheumatoid arthritis-specific autoantibodies. J. Clin. encounter of a self-antigen distinct from myelin oligodendrocyte glycoprotein. Invest. 101: 273–281. 54. Schachna, L., F. M. Wigley, S. Morris, A. C. Gelber, A. Rosen, and J. Immunol. 165: 5360–5366. L. Casciola-Rosen. 2002. Recognition of granzyme B-generated autoantigen 59. Krishnamoorthy, G., H. Lassmann, H. Wekerle, and A. Holz. 2006. Spontaneous fragments in scleroderma patients with ischemic digital loss. Arthritis Rheum. 46: opticospinal encephalomyelitis in a double-transgenic mouse model of autoim- 1873–1884. mune T cell B cell cooperation. J. Clin. Invest. 116: 2385–2392. Downloaded from http://www.jimmunol.org/ by guest on September 28, 2021