Autoantibodies occur in many different nervous system disorders, and are increasingly being found in disorders not traditionally associated with the immune system. Determining if the autoantibodies play a functional or pathogenic role is critical in selecting the most appropriate treatment options.
Association of autoantibodies with disorders of the nervous system
Some diseases affecting the nervous system have a long history of association with autoantibodies. For example, the presence of oligoclonal bands of immunoglobulin in the cerebrospinal fluid (CSF) has been used for nearly 50 years in the diagnosis of multiple sclerosis (MS) , and the presence of antibodies that bind to muscle in serum of patients with myasthenia gravis (MG) was described even earlier . More recently, it has become clear that autoantibodies are also found in various types of encephalitis, and in some patients with movement or psychiatric disorders or epilepsy [3–7]. Then there are autoantibodies associated with neoplasms (e.g., anti-Hu antibodies in small cell lung cancer with paraneoplastic sensory neuropathy ) or with other autoimmune diseases that may cross-react with neural components to cause CNS disease (e.g., cases of patients with type 1 diabetes developing stiff person syndrome, or vice versa, presumably due to immune attack against glutamic acid decarboxylase (GAD), which is present in both the pancreas and the nervous system ). What remains to be elucidated in many cases is whether these autoantibodies are merely a biomarker of the disease process, or whether they are pathogenic .
Levels of evidence in determining the pathogenicity of an autoantibody
Sixty years ago, Ernst Witebsky and colleagues set out some rules to determine if a disease was autoimmune in origin . These have been subsequently refined  to include direct evidence (i.e. transmissibility by antibody of the characteristic lesions of the disease from human to human or human to animal, or reproduction of the functional defects characteristic of the disease in vitro), indirect evidence (i.e. reproduction of the autoimmune disease in experimental animals or isolation of autoantibodies from the target organ), and circumstantial evidence (e.g. arising from clinical improvement in response to immunotherapy). There are many examples in the literature of the indirect and circumstantial levels of evidence for autoantibodies as the cause of neurological disease; however, only the direct evidence of the pathogenicity of autoantibodies will be discussed further.
Human to human transmission of a pathogenic autoantibody would be likely to occur only in two situations: either by accident (through transfusion from an affected donor), or by transplacental transmission of antibodies from mother to fetus. In approximately 10% of cases of babies born to mothers with MG  or the Lambert-Eaton myasthenic syndrome (LEMS) , the baby will have transient signs of disease. Disease transfer is dependent on the presence of antibodies of the IgG isotype and their ability to interact with the neonatal Fc receptor, FcRn, which facilitates transfer of IgG across the placenta (IgG1 binds better to FcRn than IgG4, which binds better than IgG3; IgG2 binds very poorly) . In MG and LEMS, the autoantibody targets (at the neuromuscular junction) could be easily accessed by the autoantibodies. However, it is of interest that there have been two recent reports of transplacental transmission of NMDA receptor antibodies, and development of encephalitis in at least one of these cases [16, 17]. In this situation, the antibody would have to navigate across not only the placenta, but also across the blood–CNS barriers. There are many conflicting reports regarding the permeability of the these barriers in neonates compared to adults ; but, in this patient at least, it would appear that the IgG has been able to cross into the CNS.
The first example of antibody-mediated transfer of a human nervous system disease to an experimental animal was reported in 1975, using immunoglobulin from a patient with MG . Somewhat surprisingly, there have been relatively few successful examples of human to experimental animal transmission since then. The main reason for the difficulty in inducing the characteristic lesions of the disease in an animal model is likely to be that the protein sequence of the target molecule differs somewhat at the antigen binding site between the two species, and even a single amino acid difference could potentially result in a loss of binding of an antibody. However, particularly with diseases of the CNS, the way in which the antibody is transferred will also affect the outcome of the transfer, as the adult blood–CNS barriers are largely impermeable to antibodies. Even if high levels of antibodies specific for CNS antigens are present in the serum of patients, unless the blood–CNS barriers have been compromised prior to transfer of the antibody, the antibody will usually need to be injected intracisternally or intrathecally into a recipient animal in order to maximise the likelihood that it reaches the target organ.
With recent advances in molecular biology, the ability to reproduce and to measure functional defects characteristic of the disease in vitro has improved markedly, and such systems are now providing the best direct evidence of potential pathogenic effects of autoantibodies . The use of such methods is only limited by our current understanding of how autoantibodies could act . Traditionally, we tend to think of antibodies exerting functional/pathogenic effects only if they bind to surface antigens. If the target of the antibody is not a receptor, the potential pathogenic mechanisms considered usually include cell lysis (either via antibody-mediated activation of complement, which appears to be one mechanism by which myelin might be destroyed in MS , or antibody-dependent cell-mediated cytotoxicity mediated via Fc receptors on NK cells, macrophages or polymorphonuclear cells) or coating (opsonization) of the cell to send an “eat me” message to macrophages (this is thought to underlie the abundance of myelin-loaded macrophages in MS). If the target is a receptor, then cell lysis is still an option (e.g., antibody and complement mediated destruction of muscle morphology at the postsynaptic membrane by antibodies targeting the skeletal muscle acetylcholine receptor in MG ), but other mechanisms could include blocking or modulation of receptor signalling (e.g., binding of anti-LGI1 antibodies appears to result in closing of the Kv1 channel and a reduction of K+ levels in the synapse in limbic encephalitis ), internalization of receptors (e.g., internalization and break down of NMDA receptors following binding of anti-NMDAR antibody ), or downstream effects such as modulation of cell architecture (e.g., binding of antibodies specific for MOG from patients with acute disseminated encephalomyelitis leads to changes in microtubule arrangement in oligodendrocytes ). Other methods of action could be envisaged, e.g., effects on cell trafficking or on repair mechanisms, but these remain to be elucidated.
When considering the antibody-mediated CNS disorders, a question that frequently arises is the site of synthesis of the autoantibodies. Autoantibodies can often be found both in the CSF and the serum (e.g., in NMDAR encephalitis ), or even sometimes in the serum but not the CSF (e.g., anti-dopamine receptor 2 (DR2) and anti-voltage-gated potassium channel-complex (VGKC) antibodies in non-malignant related encephalitis ). Activated B cells and plasma cells expressing the adhesion molecules ICAM-1, VLA-4 and ALCAM can cross freely into the CNS [29, 30], and would have the potential to synthesize antibody both within and outside of the CNS. There have also been reports of the formation of ectopic lymphoid follicles containing plasma cells in the meningeal space in MS , and antibodies formed within the CNS could drain via the interstitial fluid and CSF into the bloodstream . Brain-specific antibodies produced in the CNS/CSF also potentially have a large absorptive surface on which to bind (thereby decreasing the levels of free antibody detectable within the CSF) , whereas, in the serum, it is less likely they would encounter their specific antigen, and would therefore be present at higher levels. It has been argued that if antibodies produced within the serum were the sole source of the pathogenic antibodies, then it might be expected that there would be much higher levels of neurological disease in patients with diseases such as systemic lupus erythematosus (SLE) or type 1 diabetes, who often have high serum levels of autoantibodies against molecules that are present in the CNS. However, it could also be that, in such cases, the autoantibodies target slightly different epitopes of the antigen which are not as available for antibody binding within the CNS, or which do not induce the same functional effects upon antibody binding, as suggested by studies of anti-GAD antibodies in type 1 diabetes and stiff person syndrome [9, 34].
The issues of how an autoantibody acts and where it is produced will determine not only the choice of therapy, but also the likelihood of success. Autoantibodies targeting surface antigens have been reported to be more susceptible to therapeutic intervention than autoantibodies targeting intracellular antigens . Furthermore, disorders of the peripheral nervous system or neuromuscular junction, or paraneoplastic syndromes where the tumor is located peripherally, are typically more amenable to treatment than disease contained within the CNS. Many of the immunomodulatory and immunosuppressive treatments currently available target both B cell/antibody and T cell arms of the immune response, but not always in the same way. For example, the use of interferon-b, which is frequently used successfully in relapsing-remitting MS (primarily T cell driven), induced increased levels of the B cell activating factor (BAFF) within the CNS , and appeared to increase the relapse rate in patients with the antibody-mediated neuromyelitis optica (NMO) . Intravenous immunoglobulin (IVIg) has been used successfully in MG, chronic inflammatory demyelinating polyneuropathy (CIDP), and multifocal motor neuropathy (MMN), and it is thought that its beneficial effects may be mediated through upregulation of expression of the inhibitory Fc receptor FcγRIIB on macrophages, thereby shortening the half-life of autoreactive antibodies or inhibiting their functional effects . Similarly, plasmapheresis also appears to be most effective in disorders where the target of the antibodies lies outside of the CNS, e.g., in CIDP, Guillain-Barré syndrome (GBS), LEMS and MG [39, 40], or in cases of autoimmune encephalitis associated with the presence of a tumor .
The use of monoclonal antibody therapies is becoming increasingly popular, as their specificity enhances the potential to target specific molecules or mechanisms. For example, rituximab, which depletes CD20+ B cells, has been used successfully in the treatment of a wide range of neurological disorders, including MS, NMO, CIDP, MMN, IgM-MAG neuropathy, MG and paraneoplastic opsoclonus-myoclonus , and has been reported to be particularly effective in disorders associated with IgG4 antibodies . Other monoclonal antibodies currently in development or in clinical trials, or currently approved only for other (often autoimmune) disorders include belimumab to deplete BAFF  and eculizumab to inhibit complement activation and the formation of the membrane attack complex . At present, as with most of the other treatment options available, targeting pathogenic antibody production within the CNS remains a challenge, unless the blood–CNS barriers are disrupted. Other types of drugs that inhibit the same target molecules are also under development, but to improve the efficacy of therapeutic strategies for CNS disorders, it will be necessary to improve drug delivery across the blood–CNS barriers.
Autoantibodies occur in many different diseases of the nervous system, and have the potential to be of pathogenic relevance. Current therapeutic strategies are more successful in disorders where the antibodies are produced and/or act in the periphery. There is a need to develop more effective ways of targeting autoantibodies that are produced and/or act within the CNS.
The views expressed in this article are those of the author and do not necessarily represent those of Neurology Central or Future Science Group.
- Cohen S, Bannister R. Immunoglobulin synthesis within the central nervous system in disseminated sclerosis. The Lancet 289(7486), 366–367 (1967).
- Strauss AJL, Seegal BC, Hsu KC, Burkholder PM, Nastuk WL, Osserman KE. Immunofluorescence demonstration of a muscle binding, complement-fixing serum globulin fraction in myasthenia gravis. Proc. Soc. Exp. Biol. Med. 105 184–191 (1960).
- Leypoldt F, Armangue T, Dalmau J. Autoimmune encephalopathies. Ann. N. Y. Acad. Sci. 1338 94–114 (2015).
- Iorio R, Lennon VA. Neural antigen-specific autoimmune disorders. Immunol. Rev. 248(1), 104–121 (2012).
- Kruse JL, Lapid MI, Lennon VA et al. Psychiatric Autoimmunity: N-Methyl-D-Aspartate Receptor IgG and Beyond. Psychosomatics 56(3), 227–241 (2015).
- Sinmaz N, Amatoury M, Merheb V, Ramanathan S, Dale RC, Brilot F. Autoantibodies in movement and psychiatric disorders: updated concepts in detection methods, pathogenicity, and CNS entry. Ann. N. Y. Acad. Sci. 1351, 22–38 (2015).
- Davis R, Dalmau J. Autoimmunity, seizures, and status epilepticus. Epilepsia 54(SUPPL. 6), 46-49 (2013).
- Graus F, Cordon-Cardo C, Posner JB. Neuronal antinuclear antibody in sensory neuronopathy from lung cancer. Neurology 35(4), 538–543 (1985).
- Lohmann T, Hawa M, Leslie RD, Lane R, Picard J, Londei M. Immune reactivity to glutamic acid decarboxylase 65 in stiffman syndrome and type 1 diabetes mellitus. Lancet 356(9223), 31–35 (2000).
- Diamond B, Huerta PT, Mina-Osorio P, Kowal C, Volpe BT. Losing your nerves? Maybe it’s the antibodies. Nature reviews. Immunology 9(6), 449–456 (2009).
- Witebsky E, Rose NR, Terplan K, Paine JR, Egan RW. Chronic thyroiditis and autoimmunization. J. Am. Med. Assoc. 164(13), 1439–1447 (1957).
- Rose NR, Bona C. Defining criteria for autoimmune diseases (Witebsky’s postulates revisited). Immunol. Today 14(9), 426–430 (1993).
- Plauche WC. Myasthenia gravis in mothers and their newborns. Clin. Obstet. Gynecol. 34(1), 82–99 (1991).
- Reuner U, Kamin G, Ramantani G, Reichmann H, Dinger J. Transient neonatal Lambert-Eaton syndrome. J. Neurol. 255(11), 1827–1828 (2008).
- Palmeira P, Quinello C, Silveira-Lessa AL et al. IgG Placental Transfer in Healthy and Pathological Pregnancies. Clin. Dev. Immunol. 2012 13 (2012).
- Hilderink M, Titulaer MJ, Schreurs MWJ, Keizer K, Bunt JEH. Transient anti-NMDAR encephalitis in a newborn infant due to transplacental transmission. Neurology® Neuroimmunology & Neuroinflammation 2(4), e126 (2015).
- Jagota P, Vincent A, Bhidayasiri R. Transplacental transfer of NMDA receptor antibodies in an infant with cortical dysplasia. Neurology 82(18), 1662–1663 (2014).
- Saunders NR, Liddelow SA, Dziegielewska KM. Barrier Mechanisms in the Developing Brain. Front. Pharmacol. 3 46 (2012).
- Toyka KV, Brachman DB, Pestronk A, Kao I. Myasthenia gravis: passive transfer from man to mouse. Science 190(4212), 397–399 (1975).
- Ohkawa T, Fukata Y, Yamasaki M et al. Autoantibodies to epilepsy-related LGI1 in limbic encephalitis neutralize LGI1-ADAM22 interaction and reduce synaptic AMPA receptors. J. Neurosci. 33(46), 18161–18174 (2013).
- Beasley SJ, Greer JM. Autoantibodies and their potential roles in diseases of the nervous system. Clinical and Experimental Neuroimmunology 6(4), 370–386 (2015).
- Storch MK, Piddlesden S, Haltia M, Iivanainen M, Morgan P, Lassmann H. Multiple sclerosis: in situ evidence for antibody- and complement-mediated demyelination. Ann. Neurol. 43(4), 465–471 (1998).
- Kusner LL, Kaminski HJ, Soltys J. Effect of complement and its regulation on myasthenia gravis pathogenesis. Expert Rev. Clin. Immunol. 4(1), 43–52 (2008).
- Schulte U, Thumfart JO, Klocker N et al. The epilepsy-linked Lgi1 protein assembles into presynaptic Kv1 channels and inhibits inactivation by Kvbeta1. Neuron 49(5), 697–706 (2006).
- Moscato EH, Peng X, Jain A, Parsons TD, Dalmau J, Balice-Gordon RJ. Acute mechanisms underlying antibody effects in anti-N-methyl-D-aspartate receptor encephalitis. Ann. Neurol. 76(1), 108–119 (2014).
- Dale RC, Tantsis EM, Merheb V et al. Antibodies to MOG have a demyelination phenotype and affect oligodendrocyte cytoskeleton. Neurol Neuroimmunol Neuroinflamm 1(1), e12 (2014).
- Dalmau J, Gleichman AJ, Hughes EG et al. Anti-NMDA-receptor encephalitis: case series and analysis of the effects of antibodies. Lancet Neurol. 7(12), 1091–1098 (2008).
- Irani SR, Gelfand JM, Al-Diwani A, Vincent A. Cell-surface central nervous system autoantibodies: clinical relevance and emerging paradigms. Ann. Neurol. 76(2), 168–184 (2014).
- Alter A, Duddy M, Hebert S et al. Determinants of human B cell migration across brain endothelial cells. J. Immunol. 170(9), 4497–4505 (2003).
- Cayrol R, Wosik K, Berard JL et al. Activated leukocyte cell adhesion molecule promotes leukocyte trafficking into the central nervous system. Nat. Immunol. 9(2), 137–145 (2008).
- Howell OW, Reeves CA, Nicholas R et al. Meningeal inflammation is widespread and linked to cortical pathology in multiple sclerosis. Brain 134(Pt 9), 2755–2771 (2011).
- Louveau A, Harris TH, Kipnis J. Revisiting the Mechanisms of CNS Immune Privilege. Trends Immunol. 36(10), 569–577 (2015).
- Chang T, Alexopoulos H, Pettingill P et al. Immunization against GAD induces antibody binding to GAD-independent antigens and brainstem GABAergic neuronal loss. PLoS One 8(9), e72921 (2013).
- Sinmaz N, Nguyen T, Tea F, Dale RC, Brilot F. Mapping autoantigen epitopes: molecular insights into autoantibody-associated disorders of the nervous system. J. Neuroinflammation 13(1), 219 (2016).
- Dalmau J, Lancaster E, Martinez-Hernandez E, Rosenfeld MR, Balice-Gordon R. Clinical experience and laboratory investigations in patients with anti-NMDAR encephalitis. Lancet Neurol. 10(1), 63–74 (2011).
- Vaknin-Dembinsky A, Brill L, Orpaz N, Abramsky O, Karussis D. Preferential increase of B-cell activating factor in the cerebrospinal fluid of neuromyelitis optica in a white population. Mult. Scler. 16(12), 1453–1457 (2010).
- Kim SH, Kim W, Li XF, Jung IJ, Kim HJ. Does interferon beta treatment exacerbate neuromyelitis optica spectrum disorder? Mult. Scler. 18(10), 1480–1483 (2012).
- Nimmerjahn F, Ravetch JV. The antiinflammatory activity of IgG: the intravenous IgG paradox. J. Exp. Med. 204(1), 11–15 (2007).
- Cortese I, Chaudhry V, So YT, Cantor F, Cornblath DR, Rae-Grant A. Evidence-based guideline update: Plasmapheresis in neurologic disorders: report of the Therapeutics and Technology Assessment Subcommittee of the American Academy of Neurology. Neurology 76(3), 294–300 (2011).
- Gwathmey K, Balogun RA, Burns T. Neurologic indications for therapeutic plasma exchange: 2013 update. J. Clin. Apher. 29(4), 211–219 (2014).
- Rypulak E, Borys M, Piwowarczyk P et al. Successful treatment of anti-NMDA receptor encephalitis with a prompt ovarian tumour removal and prolonged course of plasmapheresis: A case report. Molecular and clinical oncology 5(6), 845–849 (2016).
- Dalakas MC. B cells as therapeutic targets in autoimmune neurological disorders. Nat. Clin. Pract. Neurol. 4(10), 557–567 (2008).
- Khosroshahi A, Carruthers MN, Deshpande V, Unizony S, Bloch DB, Stone JH. Rituximab for the treatment of IgG4-related disease: lessons from 10 consecutive patients. Medicine (Baltimore) 91(1), 57–66 (2012).
- Stohl W, Hilbert DM. The discovery and development of belimumab: the anti-BLyS-lupus connection. Nat. Biotechnol. 30(1), 69–77 (2012).
- Legendre CM, Licht C, Loirat C. Eculizumab in atypical hemolytic-uremic syndrome. N. Engl. J. Med. 369(14), 1379–1380 (2013).