An official website of the United States government
NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.
Noebels JL, Avoli M, Rogawski MA, et al., editors. Jasper's Basic Mechanisms of the Epilepsies. 5th edition. New York: Oxford University Press; 2024. doi: 10.1093/med/9780197549469.003.0033
The last two decades have seen the discovery of a plethora of cell-surface autoantibodies causally implicated in clinically distinctive central nervous system (CNS) diseases. These autoantibody-mediated neurological disorders frequently present with multifocal neurological and psychiatric symptoms with both focal and generalized seizures as a core feature. This chapter describes the clinical spectrum of autoantibody-mediated CNS disorders, focusing on autoantibodies which target the extracellular domains of cell-surface antigens, including NMDAR and LGI1. In addition, the preclinical immune mechanisms underlying the genesis and propagation of these conditions are explored, and the chapter reviews how these autoantibodies directly modulate neuronal and circuit functions to explain the observed clinical features.
While the role of the immune system in causing neurological disorders has been proposed since at least the 1960s (Nutma et al., 2019), the last couple of decades have seen a rapid acceleration in our understanding of autoantibody-mediated neurological conditions.
In particular, the discovery of specific central nervous system (CNS)-expressed molecules which are targets of the immune response has allowed classification of distinct clinico-pathological syndromes, with far-reaching implications for the diagnosis and treatment of individual patients. These diseases affect multiple anatomical sites within the nervous system and demonstrate seizures as a frequent, and often prominent, feature.
When discussing the pathophysiology of potential autoantibody-mediated neurological syndromes, it is first important to distinguish whether the relevant antibodies target an intracellular or extracellular epitope. In the latter case, circulating antibodies are able to bind their targets on intact cell membranes, making it more likely that these autoantibodies themselves are involved in the initiating disease process. On the other hand, when epitopes are intracellular, the cell membrane has to be disrupted significantly (e.g., by cell death) before autoantibodies have access to their antigen. While these intracellular-targeting autoantibodies are frequently useful markers of an underlying disease process, they are likely to represent bystanders in the pathological cascade.
We will therefore focus this chapter on syndromes associated with autoantibodies directed against cell-surface exposed epitopes. So far, such autoantibody targets described in the literature are mainly synaptic receptors, including the N-Methyl-D-aspartate (NMDA), α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA), gamma aminobutyric acid A and B (GABAA and GABAB, respectively), glycine receptors, and the metabotropic glutamate 1 and 5 receptor (mGluR1 and mGluR5, respectively). A collection of other cell-surface proteins has also been identified as antigenic targets, including leucine-rich glioma in active 1 (LGI1), contactin-associated protein 2 like (CASPR2), dipeptidyl-peptidase-like protein 6 (DPPX), and IgLON5. In many of these cases, specific or characteristic clinical features are observed in the patients, frequently closely aligning autoantigens with an associated phenotype.
We will first describe the clinical spectrum of autoantibody-mediated conditions, and then proceed to discuss what is known about the underlying immune mechanisms, and how they lead—at the cell and circuit level—to the patient’s neurological symptoms, with a focus on seizures.
In general terms, autoantibody-mediated CNS disorders are characterized by a subacute course (i.e. with a progressive prodrome of days to weeks before reaching their nadir), and the clinical presentation affects multiple neurological domains. This generalization remains valuable in everyday practice, but, of course, there are a number of exceptions to be aware of, including the often insidious onset of LGI1-, CASPR2-, and Iglon5-antibody disorders.
One of the more frequent autoantibody-mediated clinical presentations is autoimmune encephalitis (AE), where patients experience memory and cognitive impairment, psychiatric symptoms, sleep disturbance, and seizures. Within this umbrella syndrome, there are a number of more nuanced clinical features which often permit the accurate prediction of the underlying single antigenic target. Table 33–1 lists the autoantibodies associated with CNS disease and seizures and their clinical features. We will describe each in further detail below.
Proteins Whose Extracellular Domains Are Targeted by Autoantibodies in Autoimmune Encephalitis, along with Their Clinical Features.
Overall, although to varying degrees, these conditions respond to immunotherapies which include corticosteroids, intravenous immunoglobulins (IVIG), plasma exchange, and drugs such as cyclophosphamide, rituximab, and bortezomib. Hence, these diseases are often considered “not-to-miss” clinical entities.
This entity was first described in 2007 (Dalmau et al., 2007) and is characterized by the presence of autoantibodies to the GluN1 subunit of the NMDA receptor (NMDAR) in the cerebrospinal fluid (CSF) (Dalmau et al., 2019). The disease is the commonest form of encephalitis in the young, affecting children or adolescents in about 60% of cases. It is rare in patients over 50 years of age. Affected females between the age of 18 and 35 have a high risk of an underlying ovarian teratoma with tumors rarely found in other groups. The “Immunopathogenesis” section below discusses the biological role of the teratoma.
In adults, and less so in children, the disease often presents with characteristic stages of illness:
Patients often first develop flu-like symptoms, fever, and/or headache over a few days.
Next, psychiatric features develop. These often occur early and appear transdiagnostic, crossing boundaries of traditionally distinct psychiatric categories with elements of psychosis, mood, behavioral, and sleep disturbances plus catatonia. Frequent features include agitation, aggression, hallucinations, delusions, mutism, irritability or mood instability, and depressed mood, with up to 20 psychiatric features noted per patient (Al-Diwani et al., 2019).
During this period, neurological features begin to manifest. These include the following:
cognitive dysfunction—amnesia, confusion, disorientation
focal and generalized seizures
movement disorders—in particular, characteristic orofacial dyskinesias. Akin to the psychiatric features, the movement disorders appear polymorphic, with movement disorder experts noting stereotypies, dystonia, and chorea often coexist with little tremor (Varley et al., 2019), which is a rare combination of features in other disorders.
autonomic dysfunction—including hypoventilation and cardiac dysrhythmias and can require intensive care unit (ITU) care and—in unusual instances—temporary pacemaking
depressed level of consciousness—often requiring airway protection on ITU
Neuroimaging is often normal, although patients can present with typical “limbic” encephalitis mesial temporal lobe (MTL) swelling visible on magnetic resonance imaging (MRI). The electroencephalogram (EEG) can show nonspecific focal and global slowing as well as a more characteristic pattern termed extreme delta-brush, limited to cases with more severe clinical presentations. CSF is often abnormal, with lymphocytic pleocytosis being the most common finding.
AE can also occur after herpes simplex virus encephalitis (HSVE—see section on “Immunopathogenesis”). Typically, the autoimmune entity presents 4 to 8 weeks following HSVE with neuropsychiatric symptoms, often including additional features such as seizures, a movement disorder, and dysautonomia. Frequently these patients have antibodies against the NMDA receptor alongside other neural antigens (Armangue et al., 2018).
Overall, about half of the patients with NMDAR-antibody encephalitis respond well to first-line immunotherapy (steroids, IVIG, plasma exchange) and—when relevant—removal of the teratoma. In those who don’t have a significant improvement within 2–4 weeks of treatment, second-line immunotherapies (such as cyclophosphamide, rituximab) are frequently used. Overall, about 81% of patients have a good outcome when treated (Titulaer et al., 2013) (defined as a Modified Rankin Scale [mRS] of 0 to 2 at 24 months), although the condition does have a mortality of ~10%, and relapses are observed in around 25% of cases, reducing to around 10% in those administered first- and second-line immunotherapies.
This condition frequently presents with prominent seizures and cognitive/behavioral changes and is the commonest form of AE in adults. It especially tends to occur in older adults (median age of onset in mid-60s) with a 2:1 male to female dominance (Irani et al., 2010, p. 1, 2011b). Seizures are reported to affect ~90% of patients, although in our experience, a history of seizures can be elicited from relatives of all patients with CNS disease. In fact, the seizure characteristics are the hallmark of this disease.
In this disorder, seizures are typically frequent, focal attacks often with some characteristic semiologies (Aurangzeb et al., 2017; Irani et al., 2011b; van Sonderen et al., 2017), including the following:
Faciobrachial dystonic seizures (FBDS). These seizures are characterized by facial grimacing and ipsilateral arm posturing, typically last 1–3 seconds, and show a prominent dystonic component. Since their description (Irani et al., 2011b), FBDS have become recognized as essentially specific for patients with LGI1-antibodies. Interestingly, these seizures are usually not accompanied by EEG changes (probably because they are generated subcortically)—and hence clinical recognition is paramount.
Dysautonomia. Piloerection is very typical of and modestly specific to LGI1-antibody encephalitis. Patients can also experience flushing and temperature alterations, perhaps implicating the amygdala or insula cortex in their genesis.
Paroxysmal dizziness spells (PDS). These attacks present with very brief stereotypical sensations of dizziness, without vertigo or EEG changes.
Sensory features. These typically include episodic tingling and vibratory sensations.
Motor features. These include automatisms, vocalizations, clonic jerks, dystonic posturing, and version; all are all recognized in this condition and do not aid differentiation from other causes of seizure disorders.
Importantly, all these attacks can occur up to around 300 times per day, making their frequency a characteristic feature of the illness. Many coexist in an individual, meaning that multifocal frequent focal seizures are a hallmark of LGI1-antibody encephalitis.
Cognitive dysfunction is another feature of LGI1-antibody encephalitis. It is seen in >90% of cases, not being discernible in a minority of patients who display seizures. The typical cognitive deficit in the acute phase is a severe anterograde amnesia with a retrograde gap, often extending several years. MRI of the brain is abnormal in about 50% of patients, displaying unilateral or bilateral mesial temporal lobe swelling with T2/FLAIR hyperintensities. In the chronic phase, LGI1-antibody encephalitis appears to most specifically damage the CA3 region of the hippocampus (Miller et al., 2020), in keeping with neuropsychological deficits above. CSF is usually normal. About 50% of patients have serum hyponatremia, which can be an early clue to considering this condition.
The seizures in LGI1-antibody encephalitis are very responsive to early immunotherapy, but antiseizure medications (ASMs) are far less efficacious, and therefore are not used aggressively in this condition (Irani et al., 2013; Thompson et al., 2018). In fact, the use of immunotherapies as antiseizure agents in this condition led us to observe that their early administration could prevent the onset of incipient cognitive impairment, which often appears a few weeks after the onset of seizures (Irani et al., 2013; Thompson et al., 2018). Hence, our findings have implications for short- and longer-term outcomes of these patients.
Indeed, the long-term disability as measured by mRS is comparable with NMDAR-antibody encephalitis, with ~80% of patients achieving an mRS less to or equal to 2: this is often considered as “good.” However, this crude measure obscures that at least two-thirds of patients have long-term cognitive deficits, affective dysfunction, and/or fatigue which limit function and their return to employment (Binks et al., 2021). This observation leads to the question of whether we are not treating these patients with sufficient immunotherapy—a testable hypothesis in forthcoming clinical trials.
In addition to NMDAR- and LGI1-antibody encephalitis, many other extracellular targets have been described, including synaptic receptors (such as AMPA, Glycine, GABAA, GABAB, mGluR1, and mGluR5) and other extracellular proteins (e.g., DPPX, CASPR2, and IgLon5), and it is anticipated that more antigens will be discovered in the next few years. Perhaps owing to their rarity, less is known about the clinical phenotypes and mechanisms underlying these conditions. Nevertheless, we will highlight a few of these conditions below.
GABAB receptor antibody encephalitis (van Coevorden-Hameete et al., 2019) is associated with small cell lung cancer in about 50% of cases. Seizures (focal or generalized) are a prominent feature (in about 90% of patients), with about 40% of patients developing status epilepticus. Hence, GABABR antibodies should be actively considered in elderly patients presenting with status epilepticus. Limbic encephalitis, movements disorders, dysautonomia, and opsoclonus-myoclonus have been described with this syndrome. CSF pleocytosis is common, and MRI frequently shows mesial temporal changes.
The GABAA receptor can also be a target for autoantibodies (Petit-Pedrol et al., 2014; Pettingill et al., 2015). In this encephalitis, refractory focal and generalized status epilepticus are common features. The imaging features of GABAA receptor encephalitis appear highly characteristic, with patients showing multifocal fluffy cortical and subcortical FLAIR hyperintensities. We have recognized several of these patients simply from their imaging, and most do not have an associated cancer.
Dipeptidyl-peptidase-like protein 6 (DPPX) is a regulatory protein associated with Kv4.2. Patients with antibodies to DPPX present with weight loss and gastrointestinal symptoms, including diarrhea (Hara et al., 2017) in addition to CNS features of cognitive and memory dysfunction and—more specifically—features of hyperexcitability, including tremor, myoclonus, seizures, and hyperekplexia. Lymphoma is a recognized association, but most patients do not have an associated tumor.
While seizures are often a prominent feature of these autoantibody-mediated conditions, recent evidence suggests that the seizures subside with immunosuppression and that long-term post-encephalitis epilepsy is rare (de Bruijn et al., 2019). This is in contrast to classical presumed immune-mediated syndromes such as Rasmussen’s encephalitis, where seizures remain a prominent feature throughout the life of the patient and often surgical resection is required for seizure freedom (Irani et al., 2011a), and conflicts with the observation of residual medial temporal lobe atrophy in these diseases (Miller et al., 2020).
Which immunological mechanisms underlie the emergence and propagation of autoantibodies in these disorders? To understand this fundamental etiological question, we first need to review the biology of B-cells, which are the precursors to antibody-secreting cells (ASCs) (Sun et al., 2020).
B-cells are derived from common lymphoid progenitor cells in the bone marrow. As they mature, they exit the bone marrow to enter the peripheral blood as new emigrant and naïve B-cells. The B-cell receptor (BCR) expressed on the B-cell surface arises from recombination of heavy- and light-chain gene segments, resulting in a large and diverse repertoire of BCRs. Thereafter, typically in secondary lymphoid organ germinal centers, B-cell reactivity can further be modified by somatic hypermutation (SHM), a process by which point mutations are introduced in the variable region of the BCR, creating further diversity and—traditionally—improves the quality (i.e. affinity) of the immune response toward antigenic targets. The resultant diversity allows recognition of many antigens which belong to pathogens such as bacteria and viruses. Unfortunately, consequently, a large proportion of BCRs also recognize self-antigens creating a potential for autoimmunity.
In order to facilitate B-cell tolerance toward self-antigens, the immune system has several checkpoints in the bone marrow, the periphery, and the germinal centers, which aim to inactivate autoreactive B-cells. Mechanisms include apoptosis of autoreactive B-cells, induction of anergy (a state where B-cells persist in a quiescent state, with limited capacity to respond to antigen), and B-cell receptor editing whereby autoreactive cells have an opportunity to delete their autoreactive specificity. Current evidence, mainly derived from studies in patients with neuromyelitis optica, suggests that autoantibody-mediated CNS disorders arise secondary to points at which there is a breakdown in these tolerance mechanisms, both before and within germinal centers. This is evidenced by the presence of both aquaporin-4 reactive naïve and memory B-cells in patients (Wilson et al., 2018). This paradigm now requires clarification and extension to autoantibody-mediated seizure disorders.
Downstream of most B-cell tolerance mechanisms, the pathogenic autoantibodies may be secreted by two main populations of ASCs. The first are the long-lived plasma cells (LLPCs), which are terminally differentiated B-cells which reside in the bone marrow. The second are the short-lived ASCs, which are continuously being generated from B-cells in germinal centers. The relative contributions of these two populations are key to understanding both the origins of the disorders and their ideal therapeutic regimes. As LLPCs do not divide or express CD20, autoantibodies secreted by these autonomous cells are likely refractory to antiproliferative agents (e.g., azathioprine) and to anti-CD20 medications such as rituximab. By contrast, germinal center-derived ASCs should be more sensitive to rituximab and inhibition of mitosis.
A major contribution of germinal center-based mechanisms to autoantibody production are supported by several emerging observations. First, patients appear to often respond to rituximab. While this observation requires formal validation through randomized controlled trials, it supports a role for CD20+ cells. Also, the teratomas associated with NMDAR-antibody encephalitis display germinal center-like structures with rich inflammatory infiltrates, and B-cells cultured from these tumors can secrete NMDAR-reactive antibodies (Makuch et al., 2018). The fact that prompt tumor removal is associated with a significant outcome benefit provides evidence for the crucial role of these peripheral B-cells in the pathogenesis of NMDAR-antibody encephalitis. It has also been demonstrated that autoantigen-specific IgMs can be detected in some of these conditions. IgM is usually the first antibody class to be generated in a geminal center reaction and has a half-life of 5 days. Hence, it is likely produced from continual germinal center activity. Finally, a key role for specific T- and B-cell interactions is provided by genetic observations. LGI1-antibody encephalitis is strongly associated with HLA-DRB1*07:01, while CASPR2-antibody encephalitis is linked with HLA-DRB1*11:01 (Binks et al., 2018). These alleles encode HLA class II molecules which can present peptides to T-cell receptors, hence tuning the T-cell repertoire for autoantigen reactivity.
In almost all these disorders, the serum levels of autoantibodies are higher than levels in the CSF (Sun et al., 2020). Hence, a key question remains whether soluble serum antibodies can access the CNS to cause disease or whether B-cells infiltrate the CSF and locally induce disease. As in NMDAR- and LGI1-antibody encephalitis (Kornau et al., 2020; Kreye et al., 2016), autoantigen-specific antibodies have been cloned from intrathecal memory B-cells and ASCs, often at high frequencies. Hence, it appears that intrathecal synthesis of autoantibodies against self-autoantigens is a major mechanism in delivering the autoantibodies to the CNS parenchyma. Furthermore, the robust clinical observation that NMDAR-antibodies can frequently occur after herpes simplex virus encephalitis provides a model for how neuronal autoimmunity can be generated. In this phenomenon, NMDAR-antibodies (and other reactivities) are—again—found at higher concentrations in serum than CSF. One plausible explanation for this observation, after a primary CNS insult, is that HSVE infection causes the apoptosis of neurons, which leads to the release of neural antigens into the CSF. These antigens may drain via the (recently rediscovered) cerebral lymphatic drainage system into the deep cervical lymph nodes, where a B-cell response can be mounted. This can lead to the differentiation of self-directed ASCs, which migrate from the periphery into the CNS. This model is presented in Figure 33–1.
Hypothesised immunopathogenesis of autoantibody-mediated CNS disorders. Autoantibody access to CNS targets. Tissue destruction, either as a result of infection or a pro-inflammatory milieu, can release soluble neuronal autoantigens into the cerebrospinal (more...)
In order for the immune processes—notably the autoantibodies in Table 33–1—to produce neurological symptoms, including seizures, they must affect the electrophysiological properties of single neurons and/or their firing within neuronal networks. Here, we will review these processes, first for antibodies directed against synaptic receptors and then for the other extracellular proteins.
Much of our knowledge about the mechanisms of synaptic receptor autoantibodies comes from NMDAR-antibody encephalitis, where the direct pathogenic role of the autoantibodies is strongly supported by passive transfer experiments (Fig. 33–2). In these experiments, human NMDAR autoantibodies are infused into rodent CSF, usually by comparison to IgG from healthy patients as a control. Rodents infused with IgGs from patients with autoantibodies against the NMDAR display neurobehavioral deficits (Malviya et al., 2017; Planagumà et al., 2015, 2016), recapitulating some of the clinical features seen in NMDAR-antibody encephalitis. One study focused on the epileptogenic potential of NMDAR-antibodies (Wright et al., 2015). The antibodies were injected in the left lateral ventricle of C57BL/6 mice and EEG signals recorded using a miniature telemetry system with manual video analysis and automated EEG spike detection. This strain of mice is resistant to seizures, and therefore needed to be challenged with a subthreshold concentration of the chemo-convulsant pentylenetetrazol (PTZ). Mice treated with NMDAR-antibodies had more frequent and more severe seizures than mice treated with antibodies from healthy controls. This provides direct evidence that NMDAR-antibodies alone are sufficient to lower the seizure threshold. More recently, it has been shown that C57BL/6 mice develop seizures after continuous intracerebroventricular infusion of patient CSF-derived anti-NMDAR IgG, without evidence of behavioral deficits or histological changes such as neuronal cell loss or astrocytes proliferation (Taraschenko et al., 2019). Moreover, anakinra, the IL-1β receptor antagonist, attenuated daily seizure and improved novel object recognition, a measure of nonspatial memory, in these mice (Taraschenko et al., 2021).
Neuropathogenesis: neural mechanisms in autoimmune encephalitis. A. In NMDAR-antibody encephalitis, circulating autoantibodies specifically bind synaptic NMDA receptors (left panel). This causes internalisation of the synaptic receptor, thereby specifically (more...)
At the cellular level, using primary hippocampal cultures, NMDAR-reactive antibodies have been shown to cross-link and internalize the NMDA receptor with selective reduction in the NMDA current in mEPSCs (Hughes et al., 2010). While NMDAR-antibodies reduce surface NMDAR, they do not affect cell morphology, and they do not appear to reduce the other components of the synapse, including AMPA and GABA receptors. This is analogous to myasthenia gravis (MG), where it has been shown that autoantibodies to the skeletal muscle acetylcholine receptor lead to the reduction in miniature end-plate potentials (mEPPs) at the neuromuscular junction (NMJ) (Hughes et al., 2004) by causing the receptor to be internalized.
NMDAR-autoantibodies have also been shown to affect NMDA receptor trafficking, leading to displacement of NMDA receptors outside of the synapse, and blocking NMDA-dependent plasticity (Mikasova et al., 2012). Activation of the NMDAR-associated tyrosine kinase Ephrin B2 was shown to this reverse these antibody-mediated effects (Mikasova et al., 2012; Planagumà et al., 2016).
This receptor-specific effect of patient autoantibodies has led to the concept of NMDAR-encephalitis being a syndrome of NMDA hypofunction, which explains the clinical commonalities it shares with other such disorders such as phencyclidine (PCP)/ketamine abuse. Interestingly, the synaptic changes induced in the cells by incubation with NMDA antibodies appear reversible after a washout period. This could explain why patients can make a very good recovery with prompt immunosuppression.
A study using an allosteric modulator of the NMDA receptor, SGE-301 (Mannara et al., 2020), found that treatment with the modulator could inhibit the reduction of surface NMDA receptors, the disruption of synaptic plasticity, and memory deficits caused by subsequently treating mice with NMDAR-antibodies. It is yet to be understood whether this effect will persist if mice are made ill with antibodies, prior to rescue with the drug—akin to the human disease. Nevertheless, not only does this provide further evidence of the NMDA receptor specificity of the antibody-mediated pathology, it also provides a new potential line of therapeutic attack for these conditions exploiting direct receptor modulation rather than immunosuppression.
While most of the attention has understandably been focused on neurons, other cell types can express NMDA receptors. A recent study (Matute et al., 2020) found that NMDAR-autoantibodies affect oligodendrocyte function, too: incubation of oligodendrocyte cultures with NMDAR-antibodies blocked NMDAR-dependent calcium transients and blocked NMDAR-dependent increases in GLUT1 expression. Since oligodendrocytes are responsible for axonal myelin, this may explain why NMDAR-antibody encephalitis has been found to affect both grey and white matter.
One of the remarkable features of NMDAR-antibody encephalitis is the disparity between the severity of the patients’ symptoms and the lack of extensive changes on clinical MRI imaging. Not surprisingly, however, functional MRI studies of NMDAR-antibody encephalitis patients show brain-wide disruption in functional connectivity which correlate with disease activity, including diffusion tensor imaging with widespread changes in white matter tracts, measured by fractional anisotropy, mean diffusivity, and radial diffusivity (Finke et al., 2013; Peer et al., 2017). Consistent with the effect of NMDAR-antibodies on oligodendrocytes, the authors interpreted these findings to be more consistent with demyelination rather than axonal damage.
A recent systematic review of the EEG changes in NMDAR-antibody encephalitis (Gillinder et al., 2019) found that the most commonly reported EEG abnormalities were nonspecific encephalopathic changes. When present, epileptiform changes included sharp waves, periodic lateralized epileptiform discharges (PLEDs), and generalized periodic epileptiform discharges (GPEDs). Interestingly many clinically suspected seizures are not associated with EEG changes, which suggests they may be generated deep in the brain or have highly localized sources. The extreme delta-brush pattern (Schmitt et al., 2012), which is considered specific to severe NMDAR encephalitis, is characterized by slow delta activity on which are superimposed bursts of rhythmic beta oscillations. It is unclear how this rhythm arises, but a plausible hypothesis (Gillinder et al., 2019) is that the delta waves are driven by thalamocortical oscillations caused by deafferentation, and that the beta waves arise from cortical networks as a result of excitatory-inhibitory imbalances due to NMDAR hypofunction.
Much less is known about the mechanisms targeting other synaptic receptors, but they seem to follow a similar template to NMDAR-antibody encephalitis. In GABAA-receptor encephalitis (Petit-Pedrol et al., 2014), patient CSF containing GABAA-antibodies causes a specific reduction in synaptic clusters of GABAA-receptors while sparing the other components of the synapse. Interestingly, the authors found that the antibodies did not cause a reduction in extra-synaptic GABAA-receptors. Similarly, in AMPAR-antibody encephalitis, application of autoantibodies to cultured neurons causes a reduction in synaptic GluR2 containing AMPAR clusters (Lai et al., 2009). Consistent with this, antibodies derived from patients with AMPAR-antibody encephalitis have been found to reduce both mEPSC amplitude and frequency when applied to cultured neurons (Gleichman et al., 2014).
In the previous section, we reviewed evidence that synaptic receptor autoantibodies tend to cause their effects by reducing the density of their target receptor within the postsynaptic membrane. For antibodies targeting non-receptor proteins, the mechanisms are likely to be more indirect and perhaps more varied.
The most studied such case is LGI1-antibody encephalitis (Fig. 33–2). LGI1 is a secreted protein that consists of a N-terminal leucine-rich repeat (LRR) domain plus a C-terminal epitempin (EPTP) repeat domain. The EPTP domain interacts with the presynaptic disintegrin and metalloproteinase domain-containing protein (ADAM23) and the postsynaptic ADAM22 protein (Schulte et al., 2006). The presynaptic ADAM23 interacts with Kv1 channels, whereas the ADAM22 protein interacts with postsynaptic AMPA receptors, with interactions involving postsynaptic density protein 95 (PSD95). LGI1-ADAM22 and LGI1-ADAM23 complexes can trans-synaptically heterodimerize via their LRR domains, thereby functionally coupling presynaptic Kv1 channels to postsynaptic AMPA receptors (van Sonderen et al., 2017; Zekeridou and Pittock, 2020).
Patients with LGI1-antibody encephalitis have polyclonal antibodies, mainly of the IgG4 subclass, directed against both the EPTP and the LRR domains of LGI1. A recent study (Ramberger et al., 2020) dissected out the contribution of individual patient-derived monoclonal antibodies against LGI1, and found that those directed against the EPTP domain prevented the docking of LGI1 with ADAM22/23. On the other hand, antibodies against the leucine-rich repeat (LRR) domain bound the predocked LGI1-ADAM22 and LGI1-ADAM23 complexes and led to their internalization. Since most patients with LGI1-antibody encephalitis have both EPTP and LRR targeting antibodies, it is likely that a combination of (at least) these two mechanisms are at play in vivo in humans. Therefore, in contrast to NMDAR-antibody encephalitis where the effect of autoantibodies seems to be mainly postsynaptic, LGI1-antibodies have both pre- and postsynaptic mechanisms.
When applied to rat hippocampal neurons, patient serum antibodies against LGI1 cause a reduction in postsynaptic AMPAR receptors (Ohkawa et al., 2013). Acute slices derived from mice who were treated with 14 days of CSF infusion of LGI1-antibodies show increased presynaptic excitability and potentiated synaptic release, proposed to be through a reduction in Kv1.1 (Petit-Pedrol et al., 2018), as well impaired hippocampal LTP. This is mirrored at the behavioral level by defects in object recognition in LGI1-antibody treated rodents.
Much less is known about the mechanisms of antibodies against other extracellular targets. Patients with antibodies against CASPR2 can have both peripheral and central nervous system manifestations, including encephalitis and neuromyotonia. CASPR2 is a cell adhesion molecule located in both central and peripheral axonal juxtaparanodes, where it is reported to interact with contactin-2 expressed on myelinating cells (van Sonderen et al., 2017). This CASPR2-contactin-2 complex is required for the clustering of axonal Kv1 potassium channels. The CASPR2-autoantibodies are mainly IgG4, and while—in contrast to NMDAR antibodies—they do not seem to cause internalization of CASPR2, they have been shown to inhibit the contactin-2/CASPR2 interaction (Patterson et al., 2018, p. 2). Although it remains to be demonstrated, presumably these can affect the Kv1 channels downstream.
DPPX is a regulatory protein associated with a different potassium channel, Kv4.2, an A-type potassium channel which regulates somatodendritic excitability in hippocampal pyramidal neurons (Kim et al., 2008, p. 4). While DPPX autoantibodies do not directly bind Kv4.2, incubation of neurons with DPPX-antibodies causes a reduction in membrane DPPX/Kv4.2 clusters (Hara et al., 2017). While the downstream effect on excitability has not been shown, it is plausible to hypothesize that the antibodies would downregulate the A-type current and therefore lead to hyperexcitability and seizures (Fransén and Tigerholm, 2010). Interestingly, it was also found that as well as the hippocampus, DPPX is highly expressed in the myenteric plexus (Boronat et al., 2013), which may explain why patients with this condition have prominent gastrointestinal symptoms.
In this chapter, we have reviewed the clinical features and immunological and neural pathogenesis of autoantibody-mediated CNS disorders. We have described how disruption in both central and peripheral immune tolerance leads to the generation of autoreactivity in B-lineage cells and shown how the presence of autoantibodies against cell-surface proteins can alter single neuron and brain network physiology, leading to the specific clinical syndromes.
The understanding of these conditions is still in its infancy. However, we anticipate that further studies will lead to disease-specific treatments which improve patient outcomes in these disorders.
SRI is a coapplicant and receives royalties on patent application WO/2010/046716 (U.K. patent no., PCT/GB2009/051441) entitled “Neurological Autoimmune Disorders.” The patent has been licensed commercially for the development of assays for LGI1 and other VGKC-complex antibodies. SRI is an applicant on a patent application entitled “Diagnostic Strategy to Improve Specificity of CASPR2 Antibody Detection” (PCT/GB2019/051257, publication number WO/2019/211633 and UK1807410.4). SRI has received honoraria from UCB, MedImmun, ADC therapeutics, and Medlink Neurology and research support from CSL Behring, UCB, and ONO Pharma.
AM is supported by the Wellcome Trust (220708/Z/20/Z). SRI is supported by the Wellcome Trust (104079/Z/14/Z), BMA Research Grants—Vera Down grant (2013), Margaret Temple (2017), Epilepsy Research UK (P1201), the Fulbright UK-US commission (MS Society research award), and by the NIHR Oxford Biomedical Research Centre. This research was funded in whole, or in part, by the Wellcome Trust (Grant number 104079/Z/14/Z). For the purpose of Open Access, the author has applied a CC BY public copyright license to any Author Accepted Manuscript version arising from this submission. The views expressed are those of the author(s) and not necessarily those of the NHS, the NIHR, the Department of Health, UBC, or Vancouver Coastal Health. The funders had no role in the preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication. We would like to thank our patients for teaching us so much about their diseases.
This is an open access publication, available online and distributed under the terms of a Creative Commons Attribution-Non Commercial-No Derivatives 4.0 International licence (CC BY-NC-ND 4.0), a copy of which is available at https://creativecommons.org/licenses/by-nc-nd/4.0/. Subject to this license, all rights are reserved.
Your browsing activity is empty.
Activity recording is turned off.
See more...
