Molecular and cellular mechanisms underlying anti-neuronal antibody mediated disorders of the central nervous system

Autoimmunity Reviews 13 (2014) 299–312

Contents lists available at ScienceDirect

Autoimmunity Reviews journal homepage: www.elsevier.com/locate/autrev

Review

Molecular and cellular mechanisms underlying anti-neuronal antibody mediated disorders of the central nervous system M.H. van Coevorden-Hameete a,⁎, E. de Graaff a, M.J. Titulaer b, C.C. Hoogenraad a, P.A.E. Sillevis Smitt b a b

Department of Biology, Division of Cell Biology, Utrecht University, Padualaan 8, 3584 CH, Utrecht, The Netherlands Department of Neurology, Erasmus MC, 's-Gravendijkwal 230, 3015 CE Rotterdam, The Netherlands

a r t i c l e

i n f o

Article history: Received 19 October 2013 Accepted 30 October 2013 Available online 10 November 2013 Keywords: Central nervous system Autoimmune encephalitis Antibody Cell surface antigen Neurotransmitter receptor Ion channel

a b s t r a c t Over the last decade multiple autoantigens located on the plasma membrane of neurons have been identified. Neuronal surface antigens include molecules directly involved in neurotransmission and excitability. Binding of the antibody to the antigen may directly alter the target protein's function, resulting in neurological disorders. The often striking reversibility of symptoms following early aggressive immunotherapy supports a pathogenic role for autoantibodies to neuronal surface antigens. In order to better understand and treat these neurologic disorders it is important to gain insight in the underlying mechanisms of antibody pathogenicity. In this review we discuss the clinical, circumstantial, in vitro and in vivo evidence for neuronal surface antibody pathogenicity and the possible underlying cellular and molecular mechanisms. This review shows that antibodies to neuronal surface antigens are often directed at conformational epitopes located in the extracellular domain of the antigen. The conformation of the epitope can be affected by specific posttranslational modifications. This may explain the distinct clinical phenotypes that are seen in patients with antibodies to antigens that are expressed throughout the brain. Furthermore, it is likely that there is a heterogeneous antibody population, consisting of different IgG subtypes and directed at multiple epitopes located in an immunogenic region. Binding of these antibodies may result in different pathophysiological mechanisms occurring in the same patient, together contributing to the clinical syndrome. Unraveling the predominant mechanism in each distinct antigen could provide clues for therapeutic interventions. © 2013 Elsevier B.V. All rights reserved.

Contents 1. 2.

3.

Introduction . . . . . . . . . . . . . . . . . . . . Ligand gated ion channels . . . . . . . . . . . . . . 2.1. NMDA receptor . . . . . . . . . . . . . . . . 2.1.1. Antigen characteristics . . . . . . . . 2.1.2. Antibody characteristics . . . . . . . . 2.1.3. Evidence for antibody pathogenicity . . 2.1.4. Proposed pathophysiological mechanism 2.2. AMPA receptor . . . . . . . . . . . . . . . . 2.2.1. Antigen characteristics . . . . . . . . 2.2.2. Antibody characteristics . . . . . . . . 2.2.3. Evidence for antibody pathogenicity . . 2.2.4. Proposed pathophysiological mechanism 2.3. Glycine receptor . . . . . . . . . . . . . . . G-protein coupled receptors . . . . . . . . . . . . . 3.1. Metabotropic glutamate receptor . . . . . . . . 3.1.1. Antigen characteristics . . . . . . . . 3.1.2. Antibody characteristics . . . . . . . . 3.1.3. Evidence for antibody pathogenicity . . 3.1.4. Proposed pathophysiological mechanism

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

300 301 301 301 301 301 304 304 304 305 305 306 306 306 306 306 306 306 307

⁎ Corresponding author. Tel.: +31 30 2533184. E-mail addresses: [email protected] (M.H. van Coevorden-Hameete), [email protected] (E. de Graaff), [email protected] (M.J. Titulaer), [email protected] (C.C. Hoogenraad), [email protected] (P.A.E. Sillevis Smitt). 1568-9972/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.autrev.2013.10.016

300

M.H. van Coevorden-Hameete et al. / Autoimmunity Reviews 13 (2014) 299–312

3.2. GABAB receptor . . . . . . . . . . . . . . . . . 3.3. Dopamine receptor . . . . . . . . . . . . . . . 4. Potassium channel complex proteins . . . . . . . . . . . 4.1. LGI1 . . . . . . . . . . . . . . . . . . . . . . 4.1.1. Antigen characteristics . . . . . . . . . 4.1.2. Antibody characteristics . . . . . . . . . 4.1.3. Evidence for antibody pathogenicity . . . 4.1.4. Proposed pathophysiological mechanism . 4.2. Caspr2 . . . . . . . . . . . . . . . . . . . . . 4.3. DPP6 . . . . . . . . . . . . . . . . . . . . . . 5. Voltage gated ion channels . . . . . . . . . . . . . . . 5.1. Voltage gated calcium channels . . . . . . . . . . 5.1.1. Antigen characteristics . . . . . . . . . 5.1.2. Evidence for antibody pathogenicity in PCD 5.1.3. Proposed pathophysiological mechanism . 6. Other . . . . . . . . . . . . . . . . . . . . . . . . . 6.1. DNER (Tr) . . . . . . . . . . . . . . . . . . . 7. Conclusion . . . . . . . . . . . . . . . . . . . . . . 8. Search criteria . . . . . . . . . . . . . . . . . . . . . Take-home message . . . . . . . . . . . . . . . . . . . . . Disclosure statement . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

1. Introduction Immune responses affecting neurons of the central or peripheral nervous system can result in a broad spectrum of neurological syndromes ranging from encephalomyelitis to peripheral neuropathies. Sometimes these immune responses are parainfectious (e.g. Guillain Barré syndrome) and the neurological symptoms result from molecular mimicry. In other patients the disorder is paraneoplastic (e.g. anti-Yo paraneoplastic cerebellar degeneration (PCD) in ovarian cancer) in which ectopic expression of neuronal antigens by cancer cells induces immune activation. However, in many patients with suspected immune-mediated neurological syndromes the trigger of the immune response remains to be identified. When the central nervous system (CNS) is involved, these syndromes are generally called autoimmune encephalitis. Patients predominantly present with limbic encephalitis (LE), but other syndromes, including cerebellar ataxia (CA) and stiff persons' syndrome (SPS), have also been reported. Starting with HuD in 1991 [1], many paraneoplastic antigens were identified using cDNA expression libraries in Escherichia coli. Strikingly, all the antigens determined using this method are located intracellularly. Since 2000, autoantibodies against neuronal cell surface antigens have been identified in autoimmune encephalitis patients, with or without an underlying tumor. The first antigens (metabotropic glutamate receptor 1 (mGluR1) and N-methyl D-aspartate receptor (NMDAR) [2,3]) were identified by recognition of an antigen specific staining pattern in rat brain sections. Subsequently, antigens (e.g. α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptor (AMPAR), gamma-aminobutyric-acid-B receptor (GABABR) and delta/ notch-like epidermal growth factor-related receptor (DNER) [4–6]) were found by immunoprecipitation of the antigen using the patients' serum followed by mass spectrometry analysis. Autoantibodies against paraneoplastic intracellular antigens, such as HuD, are probably an epiphenomenon of a hypothesized T-cell mediated immune response. They do not appear to be directly pathogenic but can be very useful as a marker of disease. Because of cytotoxic neuronal damage, these patients often do not respond well to immunotherapy and their symptoms are mostly irreversible. Antibodies to neuronal cell surface molecules can be pathogenic by disrupting the function of the target protein. Often these are molecules involved in neurotransmission and binding of the antibodies directly leads to disrupted neuronal function. The neurological symptoms may be reversible and respond relatively well to immune suppressive therapy (for review see [7]).

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . .

307 307 308 308 308 308 308 308 308 309 309 309 309 309 309 309 309 310 310 310 310 310

Since 2007, the focus of autoimmune encephalitis research has mainly been on identification of new surface antigens and providing a description of the clinical features, diagnostic tests and therapeutic options in patients with antibodies to cell surface molecules (for review see [8]). However, it is important to strengthen the evidence for antibody pathogenicity and to deepen our understanding of the pathophysiological mechanisms involved in autoimmune encephalitis. Such improved understanding will not only provide cues for therapeutic interventions but can also teach us about the physiological function of the target proteins. In this review we summarize the evidence for pathogenicity of antibodies directed against neuronal cell surface antigens in the CNS (for overview see Table 1). Witebsky et al. drew up criteria to provide direct proof of the pathogenicity of autoantibodies, modeled after Koch's postulates: 1) antibodies have to be present in body fluids or bound to the site of pathology; 2) the antigenic target of the autoantibody should be known; 3) direct injection of patients' IgGs or immunization with a known antigen should clinically and pathologically reproduce the disorder in experimental animals [9]. One extra criterion was added by Drachman et al. in 1990; a reduction in antibody titer should co-occur with an improvement in clinical symptoms [10]. We classify the evidence of autoantibody pathogenicity into three distinct groups: The first group comprises clinical and circumstantial evidence, including symptom similarity following genetic or pharmacological disruption of the antigen, and response to immunotherapy. The second group includes studies in which functional effects of the antibodies have been demonstrated in vitro; the third group contains studies showing similar effects in vivo (Table 2). In addition, we address the possible cellular and molecular mechanisms by which antibody–antigen interaction could disrupt the function of the target protein. These mechanisms include agonistic or antagonistic effects on the receptor by binding of the antibody to the ligand binding site or allosteric binding site (Fig. 1A). An example is antagonistic autoantibodies acting on mGluR1 [2]. Furthermore, antibodies might block the pore of ion channels. Though this has not been shown for autoantibodies, antibodies experimentally generated against the extracellular domain of different types of voltage gated ion channels are able to block the pore and selectively reduce ion currents [11]. Furthermore, disruption of the interaction with neighboring molecules (auxiliary subunits, anchoring molecules, other receptors/cell surface proteins) could interfere with antigen localization as has been shown for the NMDAR and Ephrin-B2 receptor [12] (Fig. 1B). A different effect of

M.H. van Coevorden-Hameete et al. / Autoimmunity Reviews 13 (2014) 299–312

antibody binding is receptor crosslinking and subsequent internalization known as antigenic modulation [13] (Fig. 1C). This has been demonstrated for antibodies directed at the NMDAR [14,15]. Possible indirect pathogenic effects comprise complement dependent cytolysis (CDC) or antibody-dependent cell-mediated cytotoxicity (Fig. 1D). In a polyclonal immune response, antibodies can bind to multiple different epitopes and this heterogeneous antibody population can give rise to multiple pathophysiological mechanisms at the same time as demonstrated in myasthenia gravis (MG) (reviewed in [16,17]). 2. Ligand gated ion channels 2.1. NMDA receptor Antibodies directed to the NMDAR were initially recognized in 2007 in young women with encephalitis associated with ovarian teratoma [3]. Since then, hundreds of patients have been reported in literature, now including also men, children and elderly patients, frequently without an underlying tumor [18–20]. Anti-NMDAR encephalitis is often preceded by an aspecific, prodromal phase with flu-like symptoms. In adults, the disease frequently starts with psychiatric changes, later followed by seizures, orofacial dyskinesia, memory disturbance, and speech disorder. In children, the presenting symptoms can be psychiatric, but seizures and movement disorders occur as frequently. In both children and adults, these symptoms can eventually evolve into autonomic instability, central hypoventilation and coma [18,21]. Recovery often proceeds in the opposite direction, as the symptoms that occur last are the first to resolve [21]. However, the duration of recovery is usually much longer and can take well beyond 18 months [20]. 2.1.1. Antigen characteristics NMDARs are a subtype of ionotropic glutamate receptors that are widely expressed in the CNS. They play a major role in excitatory transmission, synaptic plasticity and excitotoxicity [22,23]. NMDARs are heterotetrameres that can be composed of seven different receptor subunits; NR1, NR2A-D and NR3A- and B. The NR2 subunits provide the binding site for glutamate while co-agonist glycine binds to a homologous site on NR1 and NR3. Most of the NMDARs are composed of two NR1 and two NR2 subunits [24]. The NR2 subunits are differentially expressed throughout development and in distinct regions of the brain, whereas the NR1 subunit is ubiquitously expressed [25]. The extracellular part of an NMDAR subunit is composed of the amino terminal domain (ATD) and the ligand binding domain (LBD). The ATD is subdivided into a top and a bottom lobe that form a clamshell-like structure. The ATD functions as a binding site for allosteric modulating molecules such as protons, polyamines and zinc ions [26]. Furthermore, the ATD is important in receptor assembly and interaction with other subunits [27]. Each subunit contains a transmembrane domain and an intracellular C-terminal domain that links the receptor to intracellular signaling molecules and scaffolds [28]. 2.1.2. Antibody characteristics Anti-NMDAR antibodies are primarily directed against the NR1 subunit and are mainly of the IgG1 and IgG3 subtype [14,29]. The epitope has been mapped to the ATD of the NR1 subunit. Antibody binding to NR1 is mostly independent of the other subunits in the receptor composition. Within the ATD a small region is crucial for antibody binding. This region harbors an essential glycosylation (N368) and deamination site (G369) and is located at the hinge between the two ATD lobes. These posttranslational modifications most likely affect the conformation of the clamshell-like structure of the ATD. Thereby, they affect the conformation or the availability of a distant epitope. Furthermore, upon deletion of the ATD top lobe antibody binding can be maintained, enhanced or abolished. This suggests the existence of multiple epitopes in an immunogenic region within the ATD, rather than a single epitope [30].

301

Besides in the classical anti-NMDAR encephalitis, antibodies to the NMDAR have been reported in other neurological or psychiatric disorders. However, a clear distinction needs to be made between IgG antibodies directed at a conformational epitope on NR1, and anti-NR1 antibodies that are of different Ig subtypes or directed at linear epitopes. Since none of the latter antibodies has been linked to a specific syndrome, they are up to now of less clinical significance. One paper reports IgA-NMDAR antibodies in serum of patients with slow cognitive decline of unknown cause. IgA-NMDAR antibodies did have morphological and electrophysiological effects on cultured hippocampal neurons, however they affect multiple proteins at the synapse and the antibodies are not detected in cerebrospinal fluid (CSF) [31]. A large study including over 2800 sera of patients with neuropsychiatric disorders and healthy individuals shows no difference in seroprevalence of IgG, IgA and IgM anti-NR1 antibodies between these groups (0.6%, 5.9% and 5.7%, respectively) [32]. This study, and another cohort study [33] indicate that IgG anti-NR1 antibodies are very rare in schizophrenia. Antibodies to the NR2A and -B subunit have been described in patients with neuropsychiatric lupus. Most likely these antibodies are anti-dsDNA antibodies that cross react with a pentapeptide of the NR2 subunit. The results of human studies relating anti-NR2 antibodies to a distinct neuropsychiatric phenotype in systemic lupus erythematosus (SLE) are contradictory and the exact role of anti-NR2 antibodies in disease pathogenesis remains to be elucidated [34]. 2.1.3. Evidence for antibody pathogenicity 2.1.3.1. Clinical and circumstantial. Genetically modified mice with reduced expression of the NR1 subunit display behavioral abnormalities resembling schizophrenia [35]. Furthermore, exposure of humans to non-competitive NMDAR antagonists, such as ketamine or phencyclidine, results in symptoms similar to anti-NMDAR encephalitis [36]. In all cases tested the ovarian tumors from anti-NMDAR encephalitis patients expressed NMDAR [3,21,29]. Tumor removal and/or first line immunotherapy mostly results in substantial neurological improvement, co-occurring with a drop in serum antibody titers. However, in spite of effective reduction of the antibody load in the serum, some patients remain symptomatic and have persistently high CSF antibody titers. Overall, CSF antibody titers seem to correlate better with symptom severity than serum titers [14,21,37–40]. With the increased clinical recognition of autoimmune encephalitis, pathological studies on brain biopsy and autopsy tissues are less often performed. The available immunopathological studies show antibody depositions and microgliosis, as is consistent with the antibodies' pathophysiological role [3,14,15,29,41]. Although the autoantibodies are predominantly of the complement activating subtype IgG1 and are able to fix complement in vitro, there is no detectable complement apposition in the patients' brain [42]. Cytotoxic T-cells are sparse and B-cells and plasma cells are observed to different extents [29,41–43]. A total of 6 patients with anti-NMDAR encephalitis during pregnancy have been reported. In all cases the disease occurred relatively early during pregnancy, when cross-placental IgG transfer is still limited. Immunosuppressive therapy was started promptly in all patients. All babies were born healthy, however long-term cognitive follow up results are not yet available. The results of extensive antibody testing in one baby were negative [44–47]. 2.1.3.2. In vitro. In vitro functional work using patients' IgGs points out multiple possible mechanisms of action for anti-NR1 antibodies. Several studies indicate that patients' antibodies cause antigenic modulation of NMDARs. Incubation of primary hippocampal neurons with patient's CSF or purified IgGs, for hours up to a week, specifically decreases the amount of NMDAR clusters on dendrites in a titer dependent manner [14,15,48]. This effect is reversible upon removal of the patient's CSF and cannot be induced when only (non-crosslinking) fragment

302

Table 1 Overview of clinical and circumstantial and in vivo and in vitro evidence for antibody pathogenicity. Evidence for pathogenicity Epitope

Antibody subclass

Pharmacologic or genetic antigen disruption

Treatment effect

Neuropathology

Tumor pathology

In vitro

In vivo

Ovarium Prodromal flu-like teratoma symptoms, followed by psychiatric chang- (38%) es, movement disorders and seizures. Eventually resulting in autonomic instability and coma.

N N 500 [3,18,20,21]

Multiple conformational epitopes on the ATD of the NR1 subunit [30]

IgG1, IgG3

- Reduced NR1 expression in mice: schizophrenia related behavior [35]. - NMDAR antagonists: symptoms similar to anti-NMDAR encephalitis [36].

Relatively good response to tumor removal and immunotherapy [20,21].

Antibody depositions and microgliosis. No complement depositions and sparse CD8+ T-cell infiltration. Occasional B-cells and plasma cells [15,29,41–43].

NMDAR expression on all examined ovarian tumors [3,21,29].

- Reduced surface expression [14,15,44]. - Disrupted interaction with Ephrin B2 receptor [12]. - Fast effect on receptor kinetics [30].

AMPAR

Limbic encephalitis

Lung, breast and thymus tumors (66%)

N = 15 [4,54–56]

GluR1 and GluR2 subunit [70]. No further epitope mapping.

IgG

Relatively good response to tumor removal and immunotherapy [54].

Not performed.

SPS and PERM

Thymoma, Hodgkin lymphoma, breast, lung (19%)

N = 21 [70–75,77–79]

Extracellular epitope on the GlyRα1 subunit [70]. No further epitope mapping.

IgG

Relatively good response to immunotherapy [70,71,76].

Not performed.

GluR1/2 expression on all examined tumors (breast, SCLC, thymus) [4,54]. Not performed.

Reduced surface expression [54].

GlyR

mGluR1

Cerebellar ataxia

Hodgkin lymphoma, prostate carcinoma (60%)

N = 5 [2,92–94]

Extracellular epitope on mGluR1 [2]. No further epitope mapping.

IgG

GluR1 knockout mice: symptoms resembling schizophrenia [65,66]. GluR2 knockout mice: learning impairment [67,68]. - Mutations in GlyRα1: hyperekplexia in human and rodents [89,90]. - GlyR antagonist: muscles cramps and hyperreflexia [91]. mGluR1 knockout mice: ataxia and intention tremor [104].

- Injection of IgGs in rats: overactivity of glutamatergic pathway [48,49]. - Reduced clusters of NMDA receptors [15]. - No neonatal disorder reported (clinical) [44–47]. Not performed.

Mixed response to immunotherapy [2].

Purkinje cell loss. No CD8+ T-cells [108].

mGluR5

Ophelia syndrome

Hodgkin lymphoma (100%)

N = 3 [92,95]

Extracellular epitope on mGluR5 [92].

IgG

Relatively good response to oncological and immunotherapy [92,95]

Not performed.

Mixed response to immunotherapy [137,138]

Not performed.

Predominant clinical syndrome

NMDAR

GABABR Limbic encephalitis with prominent seizures

Associated tumors (% of patients)

Predominantly N = 62 [5,109–111,182,183] small cell lung carcinoma (60%)

Extracellular conformational epitope on the GABAB1 subunit,

Both genetic and pharmacologic disruption of mGluR5 results in impaired memory and learning [181]. Predominantly GABAB1 knockout IgG1, IgG2 and mice: epilepsy and -3 also occur impaired memory [117].

Not performed.

Not performed.

mGluR1 expression on prostate carcinoma, not detected in Hodgkin Lymphoma [2,93,94]. Not performed.

Activation block [2]. Reduced LTD at parallel fiber synapse [108].

Subarachnoid infusion of patient IgGs: severe ataxia within 30 min that lasted 24 h [2]

Not performed.

Not performed.

Not performed.

Not performed.

Not performed.

M.H. van Coevorden-Hameete et al. / Autoimmunity Reviews 13 (2014) 299–312

Number of patients reported

Antigen

Basal ganglia encephalitis

LGI1

Limbic encephalitis

Caspr2

Morvans' syndrome or neuromyotonia

DPP6

Encephalitis, preceded by severe diarrhea

VGCC

Cerebellar ataxiaa

DNER

Cerebellar ataxia

a

-DRD2 knockout mice: uncoordinated movements [128]. -DR agonist: induce hallucinations, DR antagonist: induce extrapyramidal locomotor symptoms [127]. Predominantly - LGI knockout mice: N = 188 Conformational Rarely IgG4, IgG1 also lethal phenotype paraneoplastic [137,139,147,150,152,184–186] epitope on LGI1 with seizures [148]. occurs [139]. No further (8%) -LGI1 mutations: TLE epitope mapping. [149]. IgG1 Caspr2 knockout Extracellular Predominantly N = 90 mice: disrupted K+ [137,138,147,158,185,187,188] epitope on Caspr2 thymoma [137,138]. No (22%) channel clustering, further epitope no neurological mapping. symptoms [155]. Frameshift mutation: epilepsy and mental retardation [157]. DPP6 knockout mice: No tumors N = 4 [160] Both the intra- and IgG increased dendritic reported extracellular doexcitability [163]. main of DPP6 [160]. No further epitope mapping. Not performed. IgG Mutations in P/QSmall cell lung N = 40 [165–168,189] type channels: epicarcinoma sodic cerebellar atax(74%) ia and spinocerebellar ataxia [172]. No tumors reported

Hodgkin lymphoma (92%)

N = 26 [122]

N = 65 (anti-Tr patients, not all Main epitope between EGF retested for DNER) peat 2 and 3, an [6,180,190–192] additional epitope situated in repeat 3-10 [6].

IgG

IgG1, IgG3

Mixed response to immunotherapy [122].

Not performed.

Not performed.

Not performed.

Not performed.

Relatively good response to immunotherapy [137,138].

CD8+ T-cell infiltration. Mild neuronal loss [43,152,153]

Not performed.

Not performed.

Not performed.

Relatively good response to immunotherapy [137,138]

Not performed.

Not performed.

Not performed.

Not performed.

Relatively good response to immunotherapy [160]

Not performed.

Not performed.

Not performed.

Not performed.

No response to immunotherapy [174].

Not Purkinje cell loss. performed. Cortical gliosis. Reduced P/Q-type channels in molecular layer of the cerebellum [173].

Experimentally generated antibody against S5-6 loop: impaired synaptic transmission [175].

-Cerebellar infusion of experimentally generated antibodies against S5-6 in mice: CA [175]. -Intrathecal injection of IgG from PCD/ LEMS in mice: CA. IgG from LEMS alone: no CA [177]. Not performed.

DNER knockout mice: Rarely response Not performed. to disturbed motor immunotherapy coordination [179]. [6,180,190–192].

Not performed.

No morphological changes upon application of patients' IgGs to cerebellar slices or dissociated hippocampal neurons [6].

M.H. van Coevorden-Hameete et al. / Autoimmunity Reviews 13 (2014) 299–312

DRD2

and rarely the GABAB2 subunit [109,111]. No further epitope mapping. Extracellular epitope on DRD2 [122]. No further epitope mapping.

LEMS not discussed here.

303

304

M.H. van Coevorden-Hameete et al. / Autoimmunity Reviews 13 (2014) 299–312

Table 2 Types of evidence for antibody pathogenicity. Type of evidence

Experimental method

Possible outcome fitting with antibody pathogenicity

Clinical and circumstantial

Genetic disruption of the antigen Pharmacological disruption of the antigen Patients' response to immunotherapy Correlation of serum/CSF antibody titers with severity of clinical symptoms Neuropathological studies on brain autopsy or biopsy tissue

Induces a similar behavioral syndrome Induces a similar behavioral syndrome (Relatively) good response to immunotherapy High/low titers correlating with a more severe/mild clinical syndrome

Pathological studies on tumor tissue Disease episode during pregnancy In vitro

In vivo

Dissociated neuronal cultures (mainly immunofluorescence) Biochemistry Organotypic slices (mainly electrophysiology) Passive immunization: infusion of affinity purified patients' IgGs into experimental animals Active immunization with antigen

antigen-binding (Fab) fragments are used [14,15]. Also, hippocampus from autopsied anti-NMDAR positive patients and rats infused with CSF of anti-NMDAR encephalitis patients show diminished NR1 staining [15]. In addition, patients' IgGs prevent surface interaction between the Ephrin B2 receptor and the NR1 extracellular domain, leading to the diffusion of NR2A/NR1 NMDARs to perisynaptic membrane areas. In the extra synaptic membrane there is increased clustering and internalization of NR2B/NR1 receptors [12]. Electrophysiological studies have been performed to indicate the functional effects of incubation with patients' IgGs. Single channel recordings after only 7 min of antibody application show a prolonged opening duration of the NMDAR in the presence of glutamate and glycine, suggesting an initial agonistic effect of anti-NR1 antibodies [30]. Multiple other studies demonstrate a selective decrease in NMDAmediated currents after different durations of IgG incubation, probably reflecting the reduction in the number of NMDAR present in the synaptic membrane; In one study using tetanic stimulation to induce LTP, a quick suppressive effect on NMDAR currents is visible already after 5 min of antibody incubation [49]. Also Janzen et al. demonstrate a reduction in spike rates within 15 min after application [50]. Incubation of cultured cerebellar granular cells with patient's CSF for 1 h decreases the calcium influx [51]. Whole patch-clamp recordings in cultured hippocampal neurons demonstrate a selective decline in NMDARmediated currents after incubation with patients' IgGs for one day [15]. These results indicate that the effect of decreased NMDAR surface expression might be already detectable by electrophysiology long before these changes can be visualized using light microscopy. 2.1.3.3. In vivo. In vivo studies on anti-NMDAR antibodies are sparse. Injection of anti-NR1 IgGs in the CA1 hippocampal area increases the amount of glutamate in the extracellular fluid compartment [52]. In addition, injection of patients' IgGs in the prefrontal motor cortex of rats enhances the excitability of the motor cortex [53]. Both these studies suggest overactivity of glutamatergic pathways.

IgG (and possibly complement) depositions, B-cell and plasma cell infiltration, limited cytotoxic T-cell apposition Expression of the neuronal antigen in tumor tissue Occurrence of a similar neonatal syndrome (caused by a transfer of IgGs across the placenta) Reduced receptor surface expression, altered receptor localization, altered neuronal morphology Altered levels of second messengers Altered receptor kinetics and/or currents in the presence of antibodies Induces a similar behavioral syndrome Induces a similar behavioral syndrome

Paradoxically, the results from in vivo work show a hyperexcitable state, and in vitro work predominantly indicates reduced NMDAR currents upon incubation with anti-NR1 antibodies. However, as NMDARs mainly affect inhibitory pathways, a hypofunctional state of the NMDAR leads to dysinhibition. This results in symptoms reflecting a hyperexcitable state, such as psychosis and seizures that are seen in anti-NMDAR encephalitis patients. In addition to the in vivo and in vitro studies, results from brain autopsy studies and the reversibility of the disorder upon immunosuppression suggest a direct pathogenic role of anti-NMDAR antibodies. However, it should be taken into account that autopsy material is often obtained from patients who have been treated with immunosuppressant therapy. Immunopathological studies on hippocampal biopsies in the active phase of the disease are rare. In addition, as hippocampal atrophy has been reported in patients with anti-NMDAR encephalitis [3,29] a role for T-cell mediated or antibody induced cytotoxic mechanisms cannot be excluded and might especially be important in patients who have a protracted disease course. The epitope of the anti-NMDAR antibodies has been mapped to an antigenic region in the ATD of NR1. While the NR1 subunit is expressed ubiquitously in the brain, patients' sera predominantly react with the hippocampus, whereas other regions such as the cerebellum show little immunoreactivity [3]. Also, although not uncommon in young children, adults rarely display cerebellar symptoms [14,20]. These findings suggest that there are differences in NR1 conformation and epitope availability between different brain regions. This could for example be the result of different types of posttranslational modifications in different areas of the brain. Despite the body of work that has been performed in cultured neurons and slices, behavioral correlates of anti-NMDAR antibody binding in rodents are still lacking. Induction of symptoms similar to NMDAR encephalitis by active or passive immunization would provide the ultimate prove of pathogenicity. 2.2. AMPA receptor

2.1.4. Proposed pathophysiological mechanism Similar to autoantibodies in MG, anti-NMDAR antibodies most likely have multiple effects on a cellular level: an effect on receptor surface expression via antigenic modulation [14,15,48], an effect on receptor localization via disrupting the interaction with neighboring molecules on the plasma membrane [12] and possibly a fast agonistic effect on receptor kinetics [30]. The latter could result from interference of antibody binding with ligand binding, allosteric modulators or by stabilizing an open receptor conformation. To further clarify whether the suppressive effects seen after minutes of IgG application should be attributed to early NMDAR internalization or to a fast effect on receptor kinetics one could make use of Fabs.

The AMPAR was identified as an auto antigen in 2009, in 10 patients out of a series of 109 patients with LE [4,54]. In 2010, a clinical description of four more patients was published. This study included two patients that did not present with LE, but with atypical psychosis [55]. Recently, a case report of anti-AMPAR encephalitis in a young pregnant woman showed a rapid progressive disease course with more extensive CNS involvement [56]. 2.2.1. Antigen characteristics The AMPAR is an ionotropic glutamate receptor that mediates fast excitatory synaptic transmission and synaptic plasticity. A decrease in

M.H. van Coevorden-Hameete et al. / Autoimmunity Reviews 13 (2014) 299–312

A. Agonistic or antagonistic effects

305

B. Block of interaction

+ + +

C. Antigenic modulation C

complement factor

+

ion

D. Cytotoxicity

adhesion molecule natural killer cell allosteric molecule

C C

antibody receptor subunit with ligand binding site receptor subunit with allosteric binding site neurotransmitter Fig. 1. Molecular and cellular effects of antibody–antigen interaction. As an example ligand gated ion channels are depicted. A. Agonistic or antagonistic effects. Possible binding sites for antibodies are ligand or allosteric binding domains at which they can exhibit an agonistic or antagonistic effect. In addition antibodies could have a pore blocking effect or stabilize open or closed conformation of the receptor. B. Blocking of interaction. Antibody binding can prevent binding of the receptor to neighboring membrane molecules such as other receptors or adhesion molecules. This might alter the localization of the receptor. C. Antigenic modulation. Antibody binding links neighboring receptor molecules causing receptor clustering and internalization, resulting in diminished surface expression of receptors. D. Cytotoxicity. Antibody binding can lead to complement deposition and activation of natural killer cells resulting in cell death.

number of AMPARs at the postsynaptic membrane leads to long-term depression (LTD), whereas an increase leads to LTP, processes which are thought to be the underlying cellular mechanisms for memory and learning [57]. AMPARs are composed of 4 different subunits, namely GluR1-4, that combine to form heterotetramers. The different subunits show region specific expression patterns, with the GluR1/GluR2 composition being predominant in the hippocampus [58]. However, AMPAR subunit composition is remodeled continuously in response to neuronal activity [59]. The extracellular part of all subunits is composed of a long N-terminal domain and a LBD. The extracellular and transmembrane domains of the different subunits are highly homologous, but their intracellular C-terminal domains are divergent and are involved in binding to auxiliary proteins, scaffolds and signaling molecules [57,60]. Interacting molecules can also bind the AMPAR at its extracellular domain; examples are cell adhesion molecules such as cadherins and neuroligins [61]. 2.2.2. Antibody characteristics Anti-AMPAR antibodies are directed against the GluR1 and/or GluR2 subunit, with GluR2 antibodies occurring more frequently. Epitope

mapping has not been performed, CSF samples containing antibodies can however compete with each other in a cell based assay (CBA), suggesting a common epitope [54]. Antibodies against the GluR3 subunit have been described in patients with Rasmussen's encephalitis [62] and other severe forms of epilepsy [63]. However, the presence of anti-GluR3 antibodies in Rasmussen's encephalitis could not be reproduced by other research groups [64]; this hypothesis is therefore no longer pursued. 2.2.3. Evidence for antibody pathogenicity 2.2.3.1. Clinical and circumstantial. GluR1 knockout mice show symptoms resembling schizophrenia [65,66], while knockout of GluR2 has been linked to multiple types of learning impairment [67,68]. AMPAR antagonists have mainly been studied for their neuroprotective effect and little is known about their behavioral effects [69]. All tumors examined expressed either GluR1 or GluR2 or both subunits [4,54]. AMPAR encephalitis patients respond well to the administered immunotherapy and tumor treatment and nearly all return to baseline levels after the first episode. However, patients with anti-AMPAR associated LE have a high tendency to relapse. These second episodes are usually much less

306

M.H. van Coevorden-Hameete et al. / Autoimmunity Reviews 13 (2014) 299–312

responsive to therapy [54]. Up till now correlations between serum antibody levels and clinical symptoms have not been reported, neither have immunopathological studies on brain biopsy or autopsy material.

GlyR antibody titers, suggesting that GABA levels might increase to compensate for a lack of glycine responsiveness [72].

2.2.3.2. In vitro. Incubation of cultured rat hippocampal neurons with patient's antibodies for 6 days, but not yet after three days, results in a selective decrease of the number of AMPAR clusters at synaptic sites, and to a lesser extent to a reduction of the total number of receptor clusters along the dendrites. This effect is reversible after withdrawal of the patient's serum [54].

3.1. Metabotropic glutamate receptor

2.2.3.3. In vivo. No work has been reported in animal models. 2.2.4. Proposed pathophysiological mechanism The reduction of AMPAR clusters predominantly at the synapse suggests a mechanism by which patients' antibodies disturb the synaptic localization of the receptors. This could be via interfering with its trafficking or turnover, relocating the receptor to either the extra synaptic membrane or to the intracellular pool. Alike the NMDAR, this might be due to enhanced crosslinking and internalization [54]. However, in general the effects of antigenic modulation are visible within hours instead of days after antibody application. Therefore, specific loss of clusters at the synapse could be a consequence of a disrupted interaction of the extracellular domain of the AMPAR with anchoring or clustering molecules. 2.3. Glycine receptor Antibodies against the glycine receptor (GlyR) were first reported in 2008, in a patient with progressive encephalomyelitis with rigidity and myoclonus (PERM) [70]. In the following years 21 patients have been reported, all presenting with muscle stiffness, hyperactive startle responses (hyperekplexia) and limb spasms [71–80]. Three different syndromes can be distinguished with increasing severity of symptoms: stiff-leg syndrome, SPS and PERM. In addition, anti-GlyR antibodies were found in one individual suffering from progressive vision loss [76]. Glycine is an inhibitory neurotransmitter in the adult mammalian brain. It acts on GlyRs in inhibitory synapses and as a co-agonist on NMDAR [81]. GlyRs are heteropentameric receptors composed of 2 GlyRα subunits and 3 GlyRβ subunits. There are 4 different GlyRα subunits (α1–α4) and one single type of GlyRβ subunit. The GlyRα subunits are responsible for ligand binding, whereas GlyRβ is essential for receptor trafficking and synaptic clustering via binding to the scaffolding molecule gephyrin [82]. GlyRs are predominantly expressed in the brainstem, spinal cord and retina [83,84]. Patient antibodies target an extracellular epitope on the GlyRα1 subunit [70]. In this respect they differ significantly from other autoantibodies associated with spinal hyperexcitability disorders directed at glutamic acid decarboxylase 65 (GAD65), amphiphysin and gephyrin [85–87]. These antigens are located intracellular and the antibody pathogenicity is still under debate [88]. No in vivo or in vitro studies on the direct or indirect pathogenic effect of anti-GlyR antibodies have been performed. However, the clinical and circumstantial evidence directing at their pathogenicity is quite robust. Mutations in the GlyRα subunit have been found in patients with a hereditary hyperekplexia syndrome and result in a similar phenotype in rodents [89,90]. The toxin strychnine, a specific antagonist of GlyRs, among other things causes severe muscle cramps and hyperreflexia [91]. None of the tumors found in patients with anti-GlyR α1 antibodies were tested for GlyR expression and patients generally respond well to immunosuppressive therapy [70,71,76]. Moreover, anti-GlyR antibody patients respond to therapy more frequently than anti-GAD65 patients and the severity of their symptoms correlates with antibody titers [70,72,74,76]. In one patient increased GABA levels were found in the CSF that normalized upon decline of

3. G-protein coupled receptors

In 2000 mGluR1 was identified as a neuronal auto antigen in two patients with CA and Hodgkin's disease in remission [2]. Three additional patients with mGluR1 antibodies and CA have been reported since, one with a prostate adenocarcinoma and two without an underlying tumor [92–94]. In addition, three patients with anti-mGluR5 antibodies were described, which all presented with Ophelia syndrome, a rare type of LE that occurs in the context of Hodgkin's lymphoma [92,95]. Although mGluR1 antibodies have been described in only 5 patients, mGluR1 is one of the best studied neuronal surface antigens and will therefore be the focus of the following section. 3.1.1. Antigen characteristics There are 8 mammalian mGluRs which can be subdivided into three different classes; class I (mGluR1 and mGluR5), class II (mGluR2 and mGluR3) and class III (mGluR4, mGluR6, mGluR7 and mGluR8) [96]. mGluRs are homodimeric G-protein coupled receptors that contain seven transmembrane domains, an extracellular LBD and intracellular C-terminal domain that interacts with the G-protein. mGluRs mainly act via activation of the second messenger inositol-1,4,5-triphosphate (IP3) and subsequent release of calcium ions from the intracellular storage compartments. Both mGluR1 and mGluR5 are located at the perisynaptic membrane of the postsynaptic terminal [97]. mGluR1 is abundantly expressed in cerebellar Purkinje cells and the olfactory bulb. It is located at Purkinje cell dendritic spines, that form excitatory synapses with both parallel and climbing fibers [98]. It plays a critical role in LTD at the parallel fiber synapse [99–101]. In addition, mGluR1 mediates slow excitatory post synaptic potentials at synapses made between parallel fibers and Purkinje cells [102]. 3.1.2. Antibody characteristics CBAs show that the patients' antibodies react to the N-terminal extracellular domain of mGluR1. No further epitope mapping or competition experiments have been performed. Despite the homology between mGluR1 and mGluR5 the antibodies do not cross react [2,92]. Interestingly, also antibodies to Homer3, an intracellular regulator of mGluR1 and mGluR5, were described in a patient with cerebellitis [103]. 3.1.3. Evidence for antibody pathogenicity 3.1.3.1. Clinical and circumstantial. mGluR1 knockout mice show gait ataxia and intention tremor [104]. Furthermore, multiple spontaneous and induced mutations in the mGluR1 gene in mice have been reported, all resulting in ataxia and intention tremor [105,106]. Class I mGluR antagonists have been studied extensively for their analgesic and antiaddictive effects, nevertheless none of these report symptoms of ataxia [107]. Expression of mGluR1 could not be detected in tumor tissue of the Hodgkin lymphoma patients [2,93], whereas the prostate carcinoma did express mGluR1 [94]. If patients were treated with immunotherapy early in the disease course, serum and CSF titers dropped and their symptoms significantly improved [2,92,93]. Post-mortem analysis of the cerebellum of one anti-mGluR1 patient showed a normal sized cerebellum, although the density of Purkinje cells was decreased and there were changes in dendritic arborization. However, Purkinje cell loss was much less prominent than in other forms of PCD (e.g. anti-Yo associated PCD). There were neither signs of an on-going inflammatory response nor the presence of CD8+ T-cells, suggesting that the Purkinje cell degeneration is caused by other mechanisms than T-cell mediated

M.H. van Coevorden-Hameete et al. / Autoimmunity Reviews 13 (2014) 299–312

cytotoxicity [108]. Alternatively, the lack of on-going inflammation was due to the long interval between onset of symptoms and autopsy. 3.1.3.2. In vitro. Patient's antibodies specifically block the activation of mGluR1 receptors by glutamate in a dose dependent way, as was measured by the formation of IP3 in mGluR1 expressing CHO cells [2]. Purified patients' IgGs reduce Purkinje cell excitability and spontaneous firing rate in cultured mouse cerebellum slices [108]. In addition to these acute effects on cerebellar Purkinje cell functioning, mGluR1 antibodies affect synaptic plasticity; the extracellular presence of patients' mGluR1 antibodies strongly attenuates LTD of the Purkinje cell–parallel fiber synapse in cultured embryonic Purkinje cells [108]. 3.1.3.3. In vivo. Injection of purified patient's IgGs into the subarachnoid space of mice induces severe ataxia within 30 min after injection that subsides after 24 h. This effect can be abolished by specifically depleting the anti-mGluR1 IgGs while affinity purified mGluR1-IgGs have a stronger effect [2]. Furthermore, mice infused in the flocculus with purified patients IgGs show severe but reversible disruption of their compensatory eye movements [108]. 3.1.4. Proposed pathophysiological mechanism The major findings indicate that mGluR1 antibodies can affect Purkinje cell functioning at three different levels: excitability, plasticity and survival. On the short term, the antibodies reduce excitability and firing rate of Purkinje cells, which is reflected in the deficits in cerebellar motor performance. In addition mGluR1 antibodies block LTD induction, corresponding to impaired adaptation of eye movements in patients [108]. Lastly, upon prolonged exposure to the antibodies, patients have a reduced number of Purkinje cells, but this effect is less severe than in patients with other forms of PCD. Also the absence of CD8+ T-cells in autopsy material suggests a different mechanism then a cytotoxic T-cell response. Moreover, the effect of antibodies on receptor surface expression remains to be studied. 3.2. GABAB receptor In 2010 antibodies to the GABABR were described in 15 patients with LE and prominent seizures [5]. An additional series of 10 patients was reported in 2011 [109]. Anti-GABABR patients frequently have an underlying small cell lung carcinoma (SCLC). Two recent publications of 17 [110] and 20 patients [111] showed that besides with LE, anti-GABABR patients can in rare cases also present with CA and opsoclonus-myoclonus syndrome. The main inhibitory neurotransmitter GABA interacts with two types of receptors, the ionotropic GABAAR and GABACR, and the metabotropic GABABR. The GABABR is an obligatory heterodimer composed of the GABAB1 and GABAB2 subunits. It is distributed throughout the CNS and PNS, with the highest density in the cerebral cortex, thalamus, cerebellum, amygdala and hippocampus [112]. The GABABR is a G-protein coupled receptor that modulates calcium and potassium ion-channels and elicits both inhibition of presynaptic neurotransmitter release as well as postsynaptic hyperpolarization [113]. The GABAB1 subunit contains a large extracellular domain which is responsible for ligand binding, whereas the GABAB2 subunit couples the ligand-bound receptor to the effector G-protein and is essential for receptor surface trafficking [113]. The two subunits interact via their C-terminal coiledcoil domains, which also provide a binding site for auxiliary subunits [114]. There are two isoforms of the GABAB1 subunit: GABAB1a and GABAB1b, which have a different synaptic distribution. GABAB1a is located in glutamatergic terminals, where it detects GABA spillover from neighboring synapses and inhibits glutamate release. Both subtypes are expressed in the GABAergic terminals and post-synaptic membrane where GABA binding induces hyperpolarization [115,116]. The patients' antibodies target both synaptic and extra synaptic GABABRs. They are mainly directed at the GABAB1 subunit, with one

307

patient showing additional reactivity to the GABAB2 subunit [5]. A small subset of patients did not show any reactivity to the GABABR on Western blot, suggesting a conformational epitope [111]. No in vivo or in vitro pathogenicity studies using sera from antiGABABR LE patients have been reported, therefore the only indications for antibody pathogenicity can be found in clinical and circumstantial evidence. Genetic disruption of the GABABR results in symptoms similar to LE: spontaneous seizures, increased anxiety, and memory impairment [117,118]. In humans, polymorphisms in the GABABR may contribute to the risk of mesial temporal lobe epilepsy (TLE), in accordance with the prominent seizures seen in patients with GABABR antibodies [119,120]. The majority of the patients substantially improve with immunotherapy and/or oncological therapy [5,109–111]. In none of the studies the patients' SCLC tissues have been tested for GABABR expression, SCLC of patients without a neurological syndrome can however express GABABR [5]. One autopsy study of a patient with additional antiamphiphysin antibodies shows extensive T-lymphocytic infiltration, mainly in the limbic areas of the brain [111]. In contrast with AMPAR and NMDAR, it was suggested from unpublished data that antibody mediated internalization is not the primary mechanism for GABABR malfunctioning [121]. As the antibodies are mainly directed at the GABAB1 subunit, which is responsible for ligand binding, it is possible that the patients' antibodies are interfering with ligand binding and/or the subsequent conformational change that induces the downstream signaling pathway. These effects could be studied by looking at activation of GABABRs' second messengers as has been performed for mGluR1. 3.3. Dopamine receptor In 2012 antibodies to the dopamine receptor D2 (DRD2) were found in 26 pediatric patients with basal ganglia disorders, including suspected autoimmune basal ganglia encephalitis, Sydenham's chorea (SC) and Tourette's syndrome. Anti-DRD2 patients have prominent movement disorders, including chorea, dystonia, tics and Parkinsonism, combined with psychiatric symptoms [122]. Other publications reported antibodies to DRD2 and DRD1 in patients with SC and Pediatric Autoimmune Disorder Associated with Streptococcal infections (PANDAS). However, the assays that were used did not discriminate between intra- and extracellular epitopes [123,124]. The modulatory neurotransmitter dopamine binds to 5 different but closely related G-protein coupled receptors, named DRD1-5. DRD2 is localized at the postsynapse and is expressed highest in the striatum, nucleus accumbens and olfactory tubercle [125]. Dopamine is involved in a variety of different brain functions and many neurologic and psychiatric disorders have been related to dopaminergic dysfunction. A hypofunctional dopaminergic state underlies Parkinson's disease, whereas dopamine hyperactivity is thought to play an important role in schizophrenia [126]. Many well-known antipsychotic drugs, such as haloperidol, are DRD2-antagonist and cause locomotor side effects. Vice versa, dopamine agonist reduce the extrapyramidal symptoms in Parkinson's disease, but can cause hallucinations in higher dosage [127]. Moreover DRD2 knockout mice show reduced and uncoordinated movements [128]. The patients' antibodies react to the extracellular domain of the DRD2 [122]. No results have been published on a possible pathogenic mechanism for anti-DRD2 antibodies. Also, it must be noted that so far no antibodies have been detected in CSF and that the titers of antibodies binding to DRD2 expressed on the surface are low [122]. Due to diagnostic uncertainty, immunotherapy was administered empirically, but motor, cognitive and psychiatric symptoms often did not improve. However, more recent patients suspected of autoimmune basal ganglia encephalitis were treated early and aggressively and have made a complete recovery [122]. The reported patients show a large variety of symptoms of dopamine dysfunction, ranging from hyper-dopaminergic

308

M.H. van Coevorden-Hameete et al. / Autoimmunity Reviews 13 (2014) 299–312

(psychosis) to hypo-dopaminergic (Parkinsonism). Therefore no statements can be made about a possible agonistic or antagonistic effect of DRD2 antibodies and the data need to be reproduced. 4. Potassium channel complex proteins Several clinical syndromes such as neuromyotonia (NMT), LE, Morvan's syndrome and some cases of adult onset epilepsy were for a long time thought to be associated with voltage gated potassium channel (VGKC) autoantibodies [129–132]. Already in the early 90s, the evidence for anti-VGKC antibody pathogenicity was substantial. Patients with NMT and anti-VGKC antibodies responded well to plasma exchange. In addition, infusion of IgGs of some of the NMT patients into mice induced increased nerve terminal excitability in a phrenic nerve– diaphragm preparation [133,134]. These results were similar to the effects of VGKC blockers, such as aminopyridines and α-dendrotoxin, which delay nerve repolarization [135,136]. It was hypothesized that anti-VGKC antibodies cause a reduction in functional VGKCs, resulting in a prolonged action potential and increased acetylcholine release. In contrast with antibodies directed at other ion channels (described in Section 2), anti-VGKC antibodies were almost always directed at multiple VGKC subtypes. This was attributed to antibody heterogeneity or a possible common epitope on the different VGKC subtypes [131]. By now, it is known that many of the antibodies are directed at three proteins that are in complex with VGKCs: leucin-rich glioma inactivated 1 (LGI1), contactin-associated protein 2 (Caspr2), and contactin-2 [137–139]. Contactin-2 antibodies mainly co-occur with LGI1 and Caspr2 antibodies and are therefore thought to be of lesser clinical significance [137]. Roughly, it can be stated that patients with antiLGI1 antibodies mainly present with LE, while Caspr2 antibodies are predominantly associated with NMT and Morvan's syndrome. NMT is a peripheral nerve hyperexcitability syndrome with symptoms of muscle cramps and stiffness [131]. In Morvan's syndrome, NMT co-occurs with dysautonomia and cognitive and sleep disorders. Still, a portion of patients positive for VGKC antibodies in a radioimmunoassay does not react with any of the three identified VGKC-complex proteins. It is unclear what the nature and clinical significance of these antibodies is.

also non-secreted cytoplasmic LGI1. This suggests that patients' antibodies recognize a conformational epitope, which is strengthened by the fact that patients' antibodies do not recognize LGI1 on Western blot [139]. Anti-LGI1 antibodies are of the IgG1 and -4 subtype, with IgG4 being the predominant subclass [147]. Further characterization with epitope mapping has not been reported. 4.1.3. Evidence for antibody pathogenicity No in vivo or in vitro studies have been performed on anti-LGI1 antibodies. LGI1 deficient mice show a lethal phenotype that consists of several types of seizures [148]. Autosomal dominant TLE can be caused by numerous mutations throughout LGI1 protein coding sequence [149]. These epileptic findings in genetic disruption of LGI1 are consistent with the faciobrachial dystonic seizures (also referred to as tonic seizures) that often precede anti-LGI1 LE [150,151]. Many patients achieve substantial clinical improvement after treatment with immunosuppressive therapy [137,139]. Results from an autopsy study of a patient with anti-LGI1 antibodies demonstrate infiltration of cytotoxic T-cells in the hippocampus with mild neuronal loss [152]. This is in agreement with previous pathological examination in patients with anti-VGKC-complex antibodies [153]. A different immunopathological study on one patient with proven LGI antibodies shows loss of neurons with deposition of immunoglobulin and complement on neurons. Also, in this study brain atrophy on MRI was observed in one anti-LGI1 patient [43]. 4.1.4. Proposed pathophysiological mechanism Immunopathological studies on brain specimens show inconclusive results. Interestingly a recent study in one anti-LGI1 patient [43] suggests a role for CDC leading to neuronal cell loss. However this finding is not consistent with the fact that anti-LGI1 antibodies are mainly of the IgG4 subtype, which is not capable of complement activation. Since no in vitro or in vivo studies have been performed we can only speculate about a direct pathogenic effect of anti-LGI1 antibodies. As LGI1 is secreted there could be pathophysiological mechanism different from the receptor antigens described earlier on. 4.2. Caspr2

4.1. LGI1 4.1.1. Antigen characteristics The LGI protein family consists of 4 highly conserved members (LGI1-4) which are differentially expressed throughout the nervous system. LGI1 is mainly expressed in the dentate gyrus, the mossy fiber layer of the CA3 region of the hippocampus, lateral temporal cortex and cerebellar pinceau [140]. It is a secreted glycoprotein that binds to the cell surface via a disintegrin and metalloproteinase (ADAM) transmembrane proteins. LGI1 thereby connects ADAM22 to ADAM23 [141,142]. The N-terminal part of LGI1 contains three leucin-rich repeats that function as protein binding domain. The C-terminal part consists of 7 epitempin repeats that mediate LGI1 binding to ADAM22 and ADAM23 [143]. ADAM22 and LGI1 form complexes with voltagegated potassium channels (Kv1) in the juxtaparanodal regions, axon initial segment and synaptic terminals [144,145]. When co-assembled with Kv1.4 and Kvβ1 subunits LGI1 acts as an antagonist of Kvβ1 mediated channel inactivation, thereby decreasing the probability of neurotransmitter release [145]. The domain involved in this regulation has not yet been identified. Furthermore, ADAM22 and LGI1 interact with postsynaptic-density 95 (PSD95), which acts as a scaffold protein at excitatory synapses. PSD95 interacts with many postsynaptic proteins, such as AMPA- and NMDA receptors, signaling- and adhesion molecules [146]. 4.1.2. Antibody characteristics In CBAs patients' antibodies react to LGI1 most prominently when secreted and co-expressed with ADAM22 or -23 and to a lesser extent

Caspr2 is a member of the neurexin superfamily that consists of transmembrane proteins involved in cell–cell interactions within the nervous system. It is expressed in neurons throughout the PNS and CNS, including the CA3 hippocampal region and cerebellar molecularand granular layer [154]. Caspr2 serves as a scaffold for Kv1.1/Kv1.2 channels at the juxtaparanodal region in both PNS and CNS to ensure the channel clustering necessary for saltatory conduction [155]. Caspr2 has a large extracellular region containing multiple domains involved in protein–protein interactions. The smaller intracellular C-terminal part harbors a binding site for signaling molecules and cytoskeletal proteins [154,156]. A frameshift mutation in the CNTNAP2 gene, encoding for Caspr2, results in childhood onset refractory epilepsy, developmental regression, mental retardation and neuropsychiatric disturbances [157]. Caspr2 null mice show a marked reduction of K+ channels at the juxtaparanodal region, however this does not result in increased excitability, nor do these mice display any neurological abnormalities [155]. Antibodies against Caspr2 were found in patients positive or negative in the radioimmunoassay, indicating that VGKC dependent and independent epitopes might exist [158]. Also the fact that anti-Caspr2 antibodies might occur in patients with CA suggests the existence of multiple epitopes [159]. Anti-Caspr2 antibodies are predominantly of the IgG1 subtype [147], implying that there might be a role for CDC in anti-Caspr2 encephalitis. However, brain pathological studies specifically addressing patients with anti-Caspr2 antibodies are not available. Also, thymoma tissues from anti-Caspr2 patients have not been tested for Caspr2 expression.

M.H. van Coevorden-Hameete et al. / Autoimmunity Reviews 13 (2014) 299–312

Patients usually respond well to immunosuppressive therapy [137,138]. Alike the mutant Caspr2, antibody binding to Caspr2 could lead to its failure to concentrate Kv1.1/Kv1.2 channels at the juxtaparanodal region. This could result in failure to repolarize and subsequent hyperexcitability. 4.3. DPP6 Dipeptyl-peptidase-like protein-6 (DPP6), also named DPPX was identified as an auto-antigen in 2012 in 4 patients with subacute psychiatric symptoms, myoclonus and seizures. In 3 patients the neuropsychiatric disorder was preceded by severe diarrhea of unclear etiology [160]. DPP6 is a cell surface auxiliary subunit of Kv4.2 potassium channels, of which the K+ current is critical for dendritic integration, neuronal firing rate and synaptic plasticity. The Kv4.2 channel is most prominent in hippocampal neurons. It associates with two types of auxiliary proteins, Kv-channel interacting proteins and dipeptyl-peptidase-like proteins (DPP6 and -10), of which DPP6 is most abundant in the hippocampus [161]. DPP6 facilitates the trafficking of Kv4 channels to the plasma membrane, thereby modulating neuronal excitability [162]. DPP6 knockout mice have not been characterized behaviorally but display increased dendritic excitability [163]. Patient antibodies target both the intracellular and the extracellular domains of DPP6 [160]. Furthermore, the expression of DPP6 in the mesenteric plexus might fit with the patients' unexplained diarrhea. The 3 patients of which detailed follow up information was available all had substantial recovery after immunotherapy [160]. Since functional work on DPP6 antibodies has not been performed, we can only speculate about their possible mechanism of action. Possibly, antibody binding results in a loss of interaction with the Kv4 channels which alters its surface expression and results in hyperexcitability. 5. Voltage gated ion channels 5.1. Voltage gated calcium channels The role of antibodies to P/Q type voltage gated calcium channels (VGCCs) in Lambert–Eaton myasthenic syndrome (LEMS) is well established. In this disorder anti-VGCC antibodies bind to the presynaptic calcium channels at the neuromuscular junction (NMJ) leading to muscle weakness (for review see [164]). However, besides their role in LEMS, anti-VGCC antibodies have also been associated with PCD, often in the context of a SCLC [165–168]. In contrast with the role of anti-VGCC antibodies in LEMS, the contribution of anti-VGCC antibodies to PCD is still under debate. 5.1.1. Antigen characteristics Both antibodies to the P/Q-type, as well as the N-type VGCC occur in patients with PCD [167,168]. The P/Q- and N channels are located at the presynaptic membrane where they couple neuronal excitation to neurotransmitter release [169]. P/Q type channels are expressed throughout the brain with a high concentration in the cerebellar Purkinje cells [170]. The ion-conduction pore of VGCCs is formed by the α1 subunit that is associated with accessory subunits termed β and α2δ. These auxiliary subunits modulate channel kinetics and assist in trafficking and anchoring of the α1 subunit. The α1 subunit consists of 4 homologous domains (I–IV), with each domain containing 6 transmembrane segments (S1–S6). The channel's voltage sensor is localized in the S4 segment, whereas the S5 and S6 segments provide the Ca2+ specificity. The N- and C-termini of the α1 subunit are involved in receptor anchoring and trafficking [171]. 5.1.2. Evidence for antibody pathogenicity in PCD 5.1.2.1. Clinical and circumstantial. Mutations in the P/Q-type VGCCs are associated with hereditary episodic cerebellar ataxia and spinocerebellar

309

ataxia [172]. Analysis on autopsy brain material of 6 PCD-LEMS patients showed significant reduction of P/Q-type channels in the molecular layer, but not the granular layer of the cerebellum. In addition there was Purkinje cell loss and cerebellar cortex gliosis. This suggests that the immunogenic target might selectively be the P/Q-channel in the molecular layer [173]. In contrast with LEMS, patients with anti-VGCC antibodies and CA do not benefit from immunotherapy [174]. 5.1.2.2. In vitro. A peptide antibody was generated against the extracellular S5–S6 loop of both N and P/Q-type VGCCs that are essential in channel functioning. This antibody specifically inhibited N and P/Q type VGCCs in cerebellar granule neurons and impaired synaptic transmission in cerebellar slices [175]. It is however unknown whether the anti-P/Q type IgGs of PCD patients recognize this epitope. 5.1.2.3. In vivo. In one in vivo study antibodies generated against the S5–S6 loop were infused over the mice cerebellum in order to bypass the blood–brain barrier and avoid confounding effect of the antibodies on the NMJ. Infused animals displayed CA [175]. However, immunization of rats using the same S5–S6 peptide, induced muscle weakness but failed to induce cerebellar dysfunction [176]. Notably, CA symptoms were not reported in LEMS mouse model studies [164]. A recent in vivo study shows that intrathecal injection of IgGs from a PCD/LEMS patient, but not from LEMS patients without PCD, induces ataxia in mice in a similar manner as anti-mGluR1 IgGs [177]. 5.1.3. Proposed pathophysiological mechanism Most likely anti-VGCC antibodies with a distinct epitope repertoire exists in patients with PCD/LEMS compared to patients with LEMS [177]. Since antibodies to both N type and the P/Q-type channels occur in the CSF of patients with CA [167], antibodies to common or unique epitopes on N and P/Q type channels may both contribute to the CNS disorder. It remains uncertain why anti-VGCC PCD patients show limited response to immunotherapy, what causes Purkinje cell loss and if cellular and molecular disease mechanisms are comparable between LEMS and CA. 6. Other 6.1. DNER (Tr) DNER was found to be the antigen of anti-Tr antibodies that occur in patients with Hodgkin disease and PCD [6]. DNER functions as a ligand for Notch and is important for the neuron–glia interaction that is essential for the maturation of Bergmann glia cells. It contains a large extracellular domain with 10 epidermal growth factor (EGF) repeats, of which the second and third serve as a Notch binding site. DNER is expressed throughout the adult and developing brain with highest levels in the somatodendritic compartment of Purkinje Cells [178]. DNER knockout mice show disturbed motor coordination [179]. Anti-Tr antibodies are of the IgG1 and -3 subclass and occur in both serum and CSF [180]. The main epitope of the anti-Tr antibodies was mapped to a region of 176 amino acids between EGF repeats 2 and 3 of the extracellular part of DNER. However, a minority of the patients have antibodies targeting an additional epitope situated in EGF repeats 3–10. Mutation of 4N-glycosylation sites in the main epitope strongly reduced the recognition of the epitope by the anti-Tr sera, suggesting that the epitope might be conformational [6]. When applied to organotypic cerebellar slices or dissociated neuronal cultures anti-DNER IgGs did not induce any morphological changes in neurons [6]. Therefore, it remains undetermined if and how antiDNER antibodies could disturb cerebellar function in the adult cerebellum. As most patients with anti-DNER associated PCD respond poorly to treatment and severe Purkinje cell loss was observed in one autopsied patient, cytotoxic mechanisms might play a role as well [6,180].

310

M.H. van Coevorden-Hameete et al. / Autoimmunity Reviews 13 (2014) 299–312

7. Conclusion Autoimmune encephalitis includes a range of neurologic disorders that can result in severe disability or death if not recognized and treated early. Since the identification of the NMDAR as an antigenic target, research in this field has intensified leading to the identification of multiple neuronal surface antigens. However, sound evidence for antibody pathogenicity and understanding of the underlying molecular mechanisms are still very limited for many of these surface antigens. Antibodies targeting the NMDAR have been studied most thoroughly. Most likely patients possess a heterogeneous antibody population targeting an immunogenic region, rather than a single common epitope, similar to antibodies to the acetylcholine receptor in MG. Commonly the epitopes of cell surface antigens are conformational, non-linear. Probably antibody heterogeneity with respect to epitope and IgG subtype results in multiple pathophysiological effects of antibody–antigen binding occurring at the same time. Together with brain region specific conformational changes in the antigens this could contribute to differences in clinical phenotype. First line therapies currently applied in patients with autoimmune encephalitis include steroids, plasmapheresis and intravenous IgGs. Often patients require additional therapy and cyclophosphamide and rituximab are used most frequently. Determining the predominant pathogenic mechanism for each individual antigen can provide clues for the applicability of immunotherapy targeting different components of the immune system. 8. Search criteria Literature for this review was obtained by performing Pubmed searches for each specific published neuronal surface antigen in the CNS (NMDA receptor, AMPA receptor, glycine receptor, metabotropic glutamate receptor 1/5, GABAB receptor, dopamine receptor, LGI1, Caspr2, DPP6/DPPX, voltage gated calcium channels, DNER/Tr) combined with ‘antibodies’, ‘autoimmune’, ‘autoimmunity’ or the predominant clinical syndrome such as ‘limbic encephalitis’ or ‘neuromyotonia’ starting from the date of the first publication on antibodies to this specific antigen till 16th of October 2013. Take-home message • Antibodies to neuronal surface antigens are often directed at conformational epitopes in an immunogenic region. • A heterogeneous antibody population that gives rise to multiple pathogenic effects and brain region specific posttranslational modifications may explain large clinical variations. • For many neuronal surface antigens the evidence for antibody pathogenicity is still limited.

Disclosure statement EG and PSS received a research grant from Euroimmun for a patent for the use of DNER as an autoantibody test. The work of MT is supported by grants from the Netherlands Organisation for Scientific Research (NWO, Veni-incentive), the Dutch Epilepsy Foundations (NEF, project 14–19), an ErasmusMC fellowship and a clinical research fellowship by the Dutch Cancer Society (KWF, number 2009-4451). MT received a travel grant for lecturing in India from Sun Pharma, India. CH and MC have nothing to disclose. References [1] Szabo A, et al. HuD, a paraneoplastic encephalomyelitis antigen, contains RNA-binding domains and is homologous to Elav and sex-lethal. Cell 1991;67(2):325–33. [2] Sillevis Smitt P, et al. Paraneoplastic cerebellar ataxia due to autoantibodies against a glutamate receptor. N Engl J Med 2000;342(1):21–7.

[3] Dalmau J, et al. Paraneoplastic anti-N-methyl-D-aspartate receptor encephalitis associated with ovarian teratoma. Ann Neurol 2007;61(1):25–36. [4] Bataller L, et al. Reversible paraneoplastic limbic encephalitis associated with antibodies to the AMPA receptor. Neurology 2010;74(3):265–7. [5] Lancaster E, et al. Antibodies to the GABA(B) receptor in limbic encephalitis with seizures: case series and characterisation of the antigen. Lancet Neurol 2010;9(1):67–76. [6] de Graaff E, et al. Identification of delta/notch-like epidermal growth factor-related receptor as the Tr antigen in paraneoplastic cerebellar degeneration. Ann Neurol 2012;71(6):515–24. [7] Iorio R, Lennon VA. Neural antigen-specific autoimmune disorders. Immunol Rev 2012;248(1):104–21. [8] Rosenfeld MR, Dalmau J. Diagnosis and management of paraneoplastic neurologic disorders. Curr Treat Options Oncol 2013;14(4):528–38. [9] Witebsky E, et al. Chronic thyroiditis and autoimmunization. J Am Med Assoc 1957;164(13):1439–47. [10] Drachman DB. How to recognize an antibody-mediated autoimmune disease: criteria. Res Publ Assoc Res Nerv Ment Dis 1990;68:183–6. [11] Xu SZ, et al. Generation of functional ion-channel tools by E3 targeting. Nat Biotechnol 2005;23(10):1289–93. [12] Mikasova L, et al. Disrupted surface cross-talk between NMDA and Ephrin-B2 receptors in anti-NMDA encephalitis. Brain 2012;135(Pt 5):1606–21. [13] Boyse EA, Stockert E, Old LJ. Modification of the antigenic structure of the cell membrane by thymus-leukemia (TL) antibody. Proc Natl Acad Sci U S A 1967;58(3):954–7. [14] Dalmau J, et al. Anti-NMDA-receptor encephalitis: case series and analysis of the effects of antibodies. Lancet Neurol 2008;7(12):1091–8. [15] Hughes EG, et al. Cellular and synaptic mechanisms of anti-NMDA receptor encephalitis. J Neurosci 2010;30(17):5866–75. [16] Conti-Fine BM, Milani M, Kaminski HJ. Myasthenia gravis: past, present, and future. J Clin Invest 2006;116(11):2843–54. [17] Diamond B, et al. Losing your nerves? Maybe it's the antibodies. Nat Rev Immunol 2009;9(6):449–56. [18] Irani SR, Vincent A. NMDA receptor antibody encephalitis. Curr Neurol Neurosci Rep 2011;11(3):298–304. [19] Peery HE, et al. Anti-NMDA receptor encephalitis. The disorder, the diagnosis and the immunobiology. Autoimmun Rev 2012;11(12):863–72. [20] Titulaer MJ, et al. Treatment and prognostic factors for long-term outcome in patients with anti-NMDA receptor encephalitis: an observational cohort study. Lancet Neurol 2013;12(2):157–65. [21] Dalmau J, et al. Clinical experience and laboratory investigations in patients with anti-NMDAR encephalitis. Lancet Neurol 2011;10(1):63–74. [22] Hunt DL, Castillo PE. Synaptic plasticity of NMDA receptors: mechanisms and functional implications. Curr Opin Neurobiol 2012;22(3):496–508. [23] Waxman EA, Lynch DR. N-methyl-D-aspartate receptor subtypes: multiple roles in excitotoxicity and neurological disease. Neuroscientist 2005;11(1):37–49. [24] Monaghan DT, et al. Pharmacological modulation of NMDA receptor activity and the advent of negative and positive allosteric modulators. Neurochem Int 2012;61(4):581–92. [25] Monyer H, et al. Developmental and regional expression in the rat brain and functional properties of four NMDA receptors. Neuron 1994;12(3):529–40. [26] Huggins DJ, Grant GH. The function of the amino terminal domain in NMDA receptor modulation. J Mol Graph Model 2005;23(4):381–8. [27] Paoletti P. Molecular basis of NMDA receptor functional diversity. Eur J Neurosci 2011;33(8):1351–65. [28] Karakas E, Simorowski N, Furukawa H. Structure of the zinc-bound amino-terminal domain of the NMDA receptor NR2B subunit. EMBO J 2009;28(24):3910–20. [29] Tuzun E, et al. Evidence for antibody-mediated pathogenesis in anti-NMDAR encephalitis associated with ovarian teratoma. Acta Neuropathol 2009;118(6):737–43. [30] Gleichman AJ, et al. Anti-NMDA receptor encephalitis antibody binding is dependent on amino acid identity of a small region within the GluN1 amino terminal domain. J Neurosci 2012;32(32):11082–94. [31] Pruss H, et al. IgA NMDA receptor antibodies are markers of synaptic immunity in slow cognitive impairment. Neurology 2012;78(22):1743–53. [32] Hammer C, et al. Neuropsychiatric disease relevance of circulating anti-NMDA receptor autoantibodies depends on blood–brain barrier integrity. Mol Psychiatry 2013 [E-pub ahead of print]. [33] Steiner J, et al. Increased prevalence of diverse N-methyl-D-aspartate glutamate receptor antibodies in patients with an initial diagnosis of schizophrenia: specific relevance of IgG NR1a antibodies for distinction from N-methyl-D-aspartate glutamate receptor encephalitis. JAMA Psychiatry 2013;70(3):271–8. [34] Lauvsnes MB, Omdal R. Systemic lupus erythematosus, the brain, and anti-NR2 antibodies. J Neurol 2012;259(4):622–9. [35] Mohn AR, et al. Mice with reduced NMDA receptor expression display behaviors related to schizophrenia. Cell 1999;98(4):427–36. [36] Abi-Saab WM, et al. The NMDA antagonist model for schizophrenia: promise and pitfalls. Pharmacopsychiatry 1998;31(Suppl. 2):104–9. [37] Iizuka T, et al. Anti-NMDA receptor encephalitis in Japan: long-term outcome without tumor removal. Neurology 2008;70(7):504–11. [38] Ishiura H, et al. Response of anti-NMDA receptor encephalitis without tumor to immunotherapy including rituximab. Neurology 2008;71(23):1921–3. [39] Seki M, et al. Neurological response to early removal of ovarian teratoma in antiNMDAR encephalitis. J Neurol Neurosurg Psychiatry 2008;79(3):324–6. [40] Frechette ES, et al. Prolonged follow-up and CSF antibody titers in a patient with anti-NMDA receptor encephalitis. Neurology 2011;76(7 Suppl. 2):S64–6. [41] Camdessanche JP, et al. Brain immunohistopathological study in a patient with anti-NMDAR encephalitis. Eur J Neurol 2011;18(6):929–31.

M.H. van Coevorden-Hameete et al. / Autoimmunity Reviews 13 (2014) 299–312 [42] Martinez-Hernandez E, et al. Analysis of complement and plasma cells in the brain of patients with anti-NMDAR encephalitis. Neurology 2011;77(6):589–93. [43] Bien CG, et al. Immunopathology of autoantibody-associated encephalitides: clues for pathogenesis. Brain 2012;135(Pt 5):1622–38. [44] Kumar MA, et al. Anti-N-methyl-D-aspartate receptor encephalitis during pregnancy. Arch Neurol 2010;67(7):884–7. [45] McCarthy A, et al. Anti-NMDA receptor encephalitis with associated catatonia during pregnancy. J Neurol 2012;259(12):2632–5. [46] Ito Y, et al. Anti-NMDA receptor encephalitis during pregnancy. Rinsho Shinkeigaku 2010;50(2):103–7. [47] Magley J, et al. Pregnancy outcome in anti-N-methyl-D-aspartate receptor encephalitis. Obstet Gynecol 2012;120(2 Pt 2):480–3. [48] Pruss H, et al. Retrospective analysis of NMDA receptor antibodies in encephalitis of unknown origin. Neurology 2010;75(19):1735–9. [49] Zhang Q, et al. Suppression of synaptic plasticity by cerebrospinal fluid from antiNMDA receptor encephalitis patients. Neurobiol Dis 2012;45(1):610–5. [50] Jantzen SU, et al. In vitro neuronal network activity in NMDA receptor encephalitis. BMC Neurosci 2013;14:17. [51] Rubio-Agusti I, et al. Isolated hemidystonia associated with NMDA receptor antibodies. Mov Disord 2011;26(2):351–2. [52] Manto M, et al. In vivo effects of antibodies from patients with anti-NMDA receptor encephalitis: further evidence of synaptic glutamatergic dysfunction. Orphanet J Rare Dis 2010;5:31. [53] Manto M, et al. Afferent facilitation of corticomotor responses is increased by IgGs of patients with NMDA-receptor antibodies. J Neurol 2011;258(1):27–33. [54] Lai M, et al. AMPA receptor antibodies in limbic encephalitis alter synaptic receptor location. Ann Neurol 2009;65(4):424–34. [55] Graus F, et al. The expanding clinical profile of anti-AMPA receptor encephalitis. Neurology 2010;74(10):857–9. [56] Wei YC, et al. Rapid progression and brain atrophy in anti-AMPA receptor encephalitis. J Neuroimmunol 2013;261(1–2):129–33. [57] Anggono V, Huganir RL. Regulation of AMPA receptor trafficking and synaptic plasticity. Curr Opin Neurobiol 2012;22(3):461–9. [58] Lu W, et al. Subunit composition of synaptic AMPA receptors revealed by a singlecell genetic approach. Neuron 2009;62(2):254–68. [59] Liu SJ, Zukin RS. Ca2+-permeable AMPA receptors in synaptic plasticity and neuronal death. Trends Neurosci 2007;30(3):126–34. [60] Palmer CL, Cotton L, Henley JM. The molecular pharmacology and cell biology of alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptors. Pharmacol Rev 2005;57(2):253–77. [61] Czondor K, Thoumine O. Biophysical mechanisms regulating AMPA receptor accumulation at synapses. Brain Res Bull 2013;93:57–68. [62] Rogers SW, et al. Autoantibodies to glutamate receptor GluR3 in Rasmussen's encephalitis. Science 1994;265(5172):648–51. [63] Mantegazza R, et al. Antibodies against GluR3 peptides are not specific for Rasmussen's encephalitis but are also present in epilepsy patients with severe, early onset disease and intractable seizures. J Neuroimmunol 2002;131(1–2):179–85. [64] Watson R, et al. Absence of antibodies to glutamate receptor type 3 (GluR3) in Rasmussen encephalitis. Neurology 2004;63(1):43–50. [65] Wiedholz LM, et al. Mice lacking the AMPA GluR1 receptor exhibit striatal hyperdopaminergia and ‘schizophrenia-related’ behaviors. Mol Psychiatry 2008;13(6):631–40. [66] Barkus C, et al. Do GluA1 knockout mice exhibit behavioral abnormalities relevant to the negative or cognitive symptoms of schizophrenia and schizoaffective disorder? Neuropharmacology 2012;62(3):1263–72. [67] Mead AN, et al. AMPA receptor GluR2, but not GluR1, subunit deletion impairs emotional response conditioning in mice. Behav Neurosci 2006;120(2):241–8. [68] Mead AN, Stephens DN. Involvement of AMPA receptor GluR2 subunits in stimulus-reward learning: evidence from glutamate receptor gria2 knock-out mice. J Neurosci 2003;23(29):9500–7. [69] Stone TW, Addae JI. The pharmacological manipulation of glutamate receptors and neuroprotection. Eur J Pharmacol 2002;447(2–3):285–96. [70] Hutchinson M, et al. Progressive encephalomyelitis, rigidity, and myoclonus: a novel glycine receptor antibody. Neurology 2008;71(16):1291–2. [71] Mas N, et al. Antiglycine-receptor encephalomyelitis with rigidity. J Neurol Neurosurg Psychiatry 2011;82(12):1399–401. [72] Piotrowicz A, et al. A case of glycine-receptor antibody-associated encephalomyelitis with rigidity and myoclonus (PERM): clinical course, treatment and CSF findings. J Neurol 2011;258(12):2268–70. [73] Turner MR, et al. Progressive encephalomyelitis with rigidity and myoclonus: glycine and NMDA receptor antibodies. Neurology 2011;77(5):439–43. [74] Iizuka T, et al. Glycine receptor antibodies are detected in progressive encephalomyelitis with rigidity and myoclonus (PERM) but not in saccadic oscillations. J Neurol 2012;259(8):1566–73. [75] Clerinx K, et al. Progressive encephalomyelitis with rigidity and myoclonus: resolution after thymectomy. Neurology 2011;76(3):303–4. [76] McKeon A, et al. Glycine receptor autoimmune spectrum with stiff-man syndrome phenotype. JAMA Neurol 2013;70(1):44–50. [77] Damasio J, et al. Progressive encephalomyelitis with rigidity and myoclonus: the first pediatric case with glycine receptor antibodies. JAMA Neurol 2013:1–4. [78] De Blauwe SN, Santens P, Vanopdenbosch LJ. Anti-glycine receptor antibody mediated progressive encephalomyelitis with rigidity and myoclonus associated with breast cancer. Case Rep Neurol Med 2013;2013:589154. [79] Kyskan R, et al. Antiglycine receptor antibody and encephalomyelitis with rigidity and myoclonus (PERM) related to small cell lung cancer. BMJ Case Rep 2013;2013.

311

[80] Peeters E, et al. Supranuclear gaze palsy in glycine receptor antibodypositive progressive encephalomyelitis with rigidity and myoclonus. Mov Disord 2012;27(14):1830–2. [81] Dutertre S, Becker CM, Betz H. Inhibitory glycine receptors: an update. J Biol Chem 2012;287(48):40216–23. [82] Dresbach T, et al. Molecular architecture of glycinergic synapses. Histochem Cell Biol 2008;130(4):617–33. [83] Baer K, et al. Localization of glycine receptors in the human forebrain, brainstem, and cervical spinal cord: an immunohistochemical review. Front Mol Neurosci 2009;2:25. [84] Lynch JW. Molecular structure and function of the glycine receptor chloride channel. Physiol Rev 2004;84(4):1051–95. [85] Butler MH, et al. Identification of a dominant epitope of glutamic acid decarboxylase (GAD-65) recognized by autoantibodies in stiff-man syndrome. J Exp Med 1993;178(6):2097–106. [86] Butler MH, et al. Autoimmunity to gephyrin in stiff-man syndrome. Neuron 2000;26(2):307–12. [87] De Camilli P, et al. The synaptic vesicle-associated protein amphiphysin is the 128-kD autoantigen of stiff-man syndrome with breast cancer. J Exp Med 1993;178(6):2219–23. [88] Holmoy T, Geis C. The immunological basis for treatment of stiff person syndrome. J Neuroimmunol 2011;231(1–2):55–60. [89] Bakker MJ, et al. Startle syndromes. Lancet Neurol 2006;5(6):513–24. [90] Schaefer N, Vogel N, Villmann C. Glycine receptor mutants of the mouse: what are possible routes of inhibitory compensation? Front Mol Neurosci 2012;5:98. [91] Parker AJ, et al. Strychnine poisoning: gone but not forgotten. Emerg Med J 2011;28(1):84. [92] Lancaster E, et al. Antibodies to metabotropic glutamate receptor 5 in the Ophelia syndrome. Neurology 2011;77(18):1698–701. [93] Marignier R, et al. Metabotropic glutamate receptor type 1 autoantibody-associated cerebellitis: a primary autoimmune disease? Arch Neurol 2010;67(5):627–30. [94] Iorio R, et al. Cerebellar degeneration associated with mGluR1 autoantibodies as a paraneoplastic manifestation of prostate adenocarcinoma. J Neuroimmunol 2013;263(1–2):155–8. [95] Mat A, et al. Ophelia syndrome with metabotropic glutamate receptor 5 antibodies in CSF. Neurology 2013;80(14):1349–50. [96] Conn PJ, Pin JP. Pharmacology and functions of metabotropic glutamate receptors. Annu Rev Pharmacol Toxicol 1997;37:205–37. [97] Enz R. Structure of metabotropic glutamate receptor C-terminal domains in contact with interacting proteins. Front Mol Neurosci 2012;5:52. [98] Nusser Z, et al. Subsynaptic segregation of metabotropic and ionotropic glutamate receptors as revealed by immunogold localization. Neuroscience 1994;61(3):421–7. [99] Kano M, Hashimoto K, Tabata T. Type-1 metabotropic glutamate receptor in cerebellar Purkinje cells: a key molecule responsible for long-term depression, endocannabinoid signalling and synapse elimination. Philos Trans R Soc Lond B Biol Sci 2008;363(1500):2173–86. [100] Jin Y, et al. Long-term depression of mGluR1 signaling. Neuron 2007;55(2):277–87. [101] Ichise T, et al. mGluR1 in cerebellar Purkinje cells essential for long-term depression, synapse elimination, and motor coordination. Science 2000;288(5472):1832–5. [102] Batchelor AM, et al. Pharmacological characterization of synaptic transmission through mGluRs in rat cerebellar slices. Neuropharmacology 1997;36(3):401–3. [103] Hoftberger R, et al. Patient with homer-3 antibodies and cerebellitis. JAMA Neurol 2013;70(4):506–9. [104] Aiba A, et al. Deficient cerebellar long-term depression and impaired motor learning in mGluR1 mutant mice. Cell 1994;79(2):377–88. [105] Conti V, et al. crv4, a mouse model for human ataxia associated with kyphoscoliosis caused by an mRNA splicing mutation of the metabotropic glutamate receptor 1 (Grm1). Int J Mol Med 2006;18(4):593–600. [106] Sachs AJ, et al. The mouse mutants recoil wobbler and nmf373 represent a series of Grm1 mutations. Mamm Genome 2007;18(11):749–56. [107] Pietraszek M, et al. mGluR5, but not mGluR1, antagonist modifies MK-801induced locomotor activity and deficit of prepulse inhibition. Neuropharmacology 2005;49(1):73–85. [108] Coesmans M, et al. Mechanisms underlying cerebellar motor deficits due to mGluR1-autoantibodies. Ann Neurol 2003;53(3):325–36. [109] Boronat A, et al. GABA(B) receptor antibodies in limbic encephalitis and anti-GADassociated neurologic disorders. Neurology 2011;76(9):795–800. [110] Jeffery OJ, et al. GABAB receptor autoantibody frequency in service serologic evaluation. Neurology 2013;81(10):882–7. [111] Hoftberger R, et al. Encephalitis and GABAB receptor antibodies: novel findings in a new case series of 20 patients. Neurology 2013;81(17):1500–6. [112] Bettler B, et al. Molecular structure and physiological functions of GABA(B) receptors. Physiol Rev 2004;84(3):835–67. [113] Benarroch EE. GABAB receptors: structure, functions, and clinical implications. Neurology 2012;78(8):578–84. [114] Pinard A, Seddik R, Bettler B. GABAB receptors: physiological functions and mechanisms of diversity. Adv Pharmacol 2010;58:231–55. [115] Benke D, Zemoura K, Maier PJ. Modulation of cell surface GABA(B) receptors by desensitization, trafficking and regulated degradation. World J Biol Chem 2012;3(4):61–72. [116] Vigot R, et al. Differential compartmentalization and distinct functions of GABAB receptor variants. Neuron 2006;50(4):589–601. [117] Schuler V, et al. Epilepsy, hyperalgesia, impaired memory, and loss of pre- and postsynaptic GABA(B) responses in mice lacking GABA(B(1)). Neuron 2001;31(1):47–58. [118] Mombereau C, et al. Genetic and pharmacological evidence of a role for GABA(B) receptors in the modulation of anxiety- and antidepressant-like behavior. Neuropsychopharmacology 2004;29(6):1050–62.

312

M.H. van Coevorden-Hameete et al. / Autoimmunity Reviews 13 (2014) 299–312

[119] Gambardella A, et al. GABA(B) receptor 1 polymorphism (G1465A) is associated with temporal lobe epilepsy. Neurology 2003;60(4):560–3. [120] Wang X, et al. Association between the gamma-aminobutyric acid type B receptor 1 and 2 gene polymorphisms and mesial temporal lobe epilepsy in a Han Chinese population. Epilepsy Res 2008;81(2–3):198–203. [121] Lancaster E, Dalmau J. Neuronal autoantigens-pathogenesis, associated disorders and antibody testing. Nat Rev Neurol 2012;8(7):380–90. [122] Dale RC, et al. Antibodies to surface dopamine-2 receptor in autoimmune movement and psychiatric disorders. Brain 2012;135(Pt 11):3453–68. [123] Brimberg L, et al. Behavioral, pharmacological, and immunological abnormalities after streptococcal exposure: a novel rat model of Sydenham chorea and related neuropsychiatric disorders. Neuropsychopharmacology 2012;37(9):2076–87. [124] Ben-Pazi H, Stoner JA, Cunningham MW. Dopamine receptor autoantibodies correlate with symptoms in Sydenham's chorea. PLoS One 2013;8(9):e73516. [125] Seeman P, et al. Dopamine D2 and D3 receptors in human putamen, caudate nucleus, and globus pallidus. Synapse 2006;60(3):205–11. [126] Iversen SD, Iversen LL. Dopamine: 50 years in perspective. Trends Neurosci 2007;30(5):188–93. [127] Beaulieu JM, Gainetdinov RR. The physiology, signaling, and pharmacology of dopamine receptors. Pharmacol Rev 2011;63(1):182–217. [128] Baik JH, et al. Parkinsonian-like locomotor impairment in mice lacking dopamine D2 receptors. Nature 1995;377(6548):424–8. [129] Vincent A. Autoimmune channelopathies: John Newsom-Davis's work and legacy. A summary of the Newsom-Davis memorial lecture 2008. J Neuroimmunol 2008;201–202:245–9. [130] Tan KM, et al. Clinical spectrum of voltage-gated potassium channel autoimmunity. Neurology 2008;70(20):1883–90. [131] Hart IK, et al. Autoantibodies detected to expressed K+ channels are implicated in neuromyotonia. Ann Neurol 1997;41(2):238–46. [132] Kleopa KA, et al. Neuromyotonia and limbic encephalitis sera target mature Shaker-type K+ channels: subunit specificity correlates with clinical manifestations. Brain 2006;129(Pt 6):1570–84. [133] Sinha S, et al. Autoimmune aetiology for acquired neuromyotonia (Isaacs' syndrome). Lancet 1991;338(8759):75–7. [134] Shillito P, et al. Acquired neuromyotonia: evidence for autoantibodies directed against K+ channels of peripheral nerves. Ann Neurol 1995;38(5):714–22. [135] Fu WM, Lin RH, Lin-Shiau SY. Study on the neuromuscular action of 4-aminopyridine in the mouse diaphragm. Arch Int Pharmacodyn Ther 1987;289(2):198–211. [136] Roper J, Schwarz JR. Heterogeneous distribution of fast and slow potassium channels in myelinated rat nerve fibres. J Physiol 1989;416:93–110. [137] Irani SR, et al. Antibodies to Kv1 potassium channel-complex proteins leucine-rich, glioma inactivated 1 protein and contactin-associated protein-2 in limbic encephalitis, Morvan's syndrome and acquired neuromyotonia. Brain 2010;133(9):2734–48. [138] Lancaster E, et al. Investigations of caspr2, an autoantigen of encephalitis and neuromyotonia. Ann Neurol 2011;69(2):303–11. [139] Lai M, et al. Investigation of LGI1 as the antigen in limbic encephalitis previously attributed to potassium channels: a case series. Lancet Neurol 2010;9(8):776–85. [140] Herranz-Perez V, et al. Regional distribution of the leucine-rich glioma inactivated (LGI) gene family transcripts in the adult mouse brain. Brain Res 2010;1307:177–94. [141] Leonardi E, et al. A computational model of the LGI1 protein suggests a common binding site for ADAM proteins. PLoS One 2011;6(3):e18142. [142] Sirerol-Piquer MS, et al. The epilepsy gene LGI1 encodes a secreted glycoprotein that binds to the cell surface. Hum Mol Genet 2006;15(23):3436–45. [143] Morante-Redolat JM, et al. Mutations in the LGI1/Epitempin gene on 10q24 cause autosomal dominant lateral temporal epilepsy. Hum Mol Genet 2002;11(9):1119–28. [144] Ogawa Y, et al. ADAM22, a Kv1 channel-interacting protein, recruits membraneassociated guanylate kinases to juxtaparanodes of myelinated axons. J Neurosci 2010;30(3):1038–48. [145] Schulte U, et al. The epilepsy-linked Lgi1 protein assembles into presynaptic Kv1 channels and inhibits inactivation by Kvbeta1. Neuron 2006;49(5):697–706. [146] Han K, Kim E. Synaptic adhesion molecules and PSD-95. Prog Neurobiol 2008;84(3):263–83. [147] Irani SR, et al. Morvan syndrome: clinical and serological observations in 29 cases. Ann Neurol 2012;72(2):241–55. [148] Chabrol E, et al. Electroclinical characterization of epileptic seizures in leucine-rich, glioma-inactivated 1-deficient mice. Brain 2010;133(9):2749–62. [149] Nobile C, et al. LGI1 mutations in autosomal dominant and sporadic lateral temporal epilepsy. Hum Mutat 2009;30(4):530–6. [150] Irani SR, et al. Faciobrachial dystonic seizures precede Lgi1 antibody limbic encephalitis. Ann Neurol 2011;69(5):892–900. [151] Andrade DM, et al. Tonic seizures: a diagnostic clue of anti-LGI1 encephalitis? Neurology 2011;76(15):1355–7. [152] Schultze-Amberger J, Pehl D, Stenzel W. LGI-1-positive limbic encephalitis: a clinicopathological study. J Neurol 2012;259(11):2478–80. [153] Park DC, et al. An autopsy case of limbic encephalitis with voltage-gated potassium channel antibodies. Eur J Neurol 2007;14(10):e5–6. [154] Poliak S, et al. Caspr2, a new member of the neurexin superfamily, is localized at the juxtaparanodes of myelinated axons and associates with K+ channels. Neuron 1999;24(4):1037–47. [155] Poliak S, et al. Juxtaparanodal clustering of Shaker-like K+ channels in myelinated axons depends on Caspr2 and TAG-1. J Cell Biol 2003;162(6):1149–60.

[156] Horresh I, et al. Organization of myelinated axons by Caspr and Caspr2 requires the cytoskeletal adapter protein 4.1B. J Neurosci 2010;30(7):2480–9. [157] Strauss KA, et al. Recessive symptomatic focal epilepsy and mutant contactinassociated protein-like 2. N Engl J Med 2006;354(13):1370–7. [158] Becker EB, et al. Contactin-associated protein-2 antibodies in non-paraneoplastic cerebellar ataxia. J Neurol Neurosurg Psychiatry 2012;83(4):437–40. [159] Becker EB, et al. Candidate screening of the TRPC3 gene in cerebellar ataxia. Cerebellum 2011;10(2):296–9. [160] Boronat A, et al. Encephalitis and antibodies to dipeptidyl-peptidase-like protein-6, a subunit of Kv4.2 potassium channels. Ann Neurol 2013;73(1):120–8. [161] Kim J, et al. Kv4 accessory protein DPPX (DPP6) is a critical regulator of membrane excitability in hippocampal CA1 pyramidal neurons. J Neurophysiol 2008;100(4):1835–47. [162] Nadal MS, et al. The CD26-related dipeptidyl aminopeptidase-like protein DPPX is a critical component of neuronal A-type K+ channels. Neuron 2003;37(3):449–61. [163] Sun W, et al. DPP6 establishes the A-type K(+) current gradient critical for the regulation of dendritic excitability in CA1 hippocampal neurons. Neuron 2011;71(6):1102–15. [164] Titulaer MJ, Lang B, Verschuuren JJ. Lambert–Eaton myasthenic syndrome: from clinical characteristics to therapeutic strategies. Lancet Neurol 2011;10(12):1098–107. [165] Burk K, et al. Antineuronal antibodies in sporadic late-onset cerebellar ataxia. J Neurol 2010;257(1):59–62. [166] Lorenzoni PJ, et al. Cerebellar ataxia in non-paraneoplastic Lambert–Eaton myasthenic syndrome. J Neurol Sci 2008;270(1–2):194–6. [167] Graus F, et al. P/Q type calcium-channel antibodies in paraneoplastic cerebellar degeneration with lung cancer. Neurology 2002;59(5):764–6. [168] Mason WP, et al. Small-cell lung cancer, paraneoplastic cerebellar degeneration and the Lambert–Eaton myasthenic syndrome. Brain 1997;120(Pt 8):1279–300. [169] Ishikawa T, et al. Presynaptic N-type and P/Q-type Ca2+ channels mediating synaptic transmission at the calyx of Held of mice. J Physiol 2005;568(Pt 1):199–209. [170] Westenbroek RE, et al. Immunochemical identification and subcellular distribution of the alpha 1A subunits of brain calcium channels. J Neurosci 1995;15(10):6403–18. [171] Catterall WA. Voltage-gated calcium channels. Cold Spring Harb Perspect Biol 2011;3(8):a003947. [172] Rajakulendran S, Kaski D, Hanna MG. Neuronal P/Q-type calcium channel dysfunction in inherited disorders of the CNS. Nat Rev Neurol 2012;8(2):86–96. [173] Fukuda T, et al. Reduction of P/Q-type calcium channels in the postmortem cerebellum of paraneoplastic cerebellar degeneration with Lambert–Eaton myasthenic syndrome. Ann Neurol 2003;53(1):21–8. [174] Blumenfeld AM, et al. Coexistence of Lambert–Eaton myasthenic syndrome and subacute cerebellar degeneration: differential effects of treatment. Neurology 1991;41(10):1682–5. [175] Liao YJ, et al. Anti-Ca2+ channel antibody attenuates Ca2+ currents and mimics cerebellar ataxia in vivo. Proc Natl Acad Sci U S A 2008;105(7):2705–10. [176] Komai K, Iwasa K, Takamori M. Calcium channel peptide can cause an autoimmune-mediated model of Lambert–Eaton myasthenic syndrome in rats. J Neurol Sci 1999;166(2):126–30. [177] Martin-Garcia E, et al. Intrathecal injection of P/Q type voltage-gated calcium channel antibodies from paraneoplastic cerebellar degeneration cause ataxia in mice. J Neuroimmunol 2013;261(1–2):53–9. [178] Eiraku M, et al. Delta/notch-like epidermal growth factor (EGF)-related receptor, a novel EGF-like repeat-containing protein targeted to dendrites of developing and adult central nervous system neurons. J Biol Chem 2002;277(28):25400–7. [179] Tohgo A, et al. Impaired cerebellar functions in mutant mice lacking DNER. Mol Cell Neurosci 2006;31(2):326–33. [180] Bernal F, et al. Anti-Tr antibodies as markers of paraneoplastic cerebellar degeneration and Hodgkin's disease. Neurology 2003;60(2):230–4. [181] Simonyi A, Schachtman TR, Christoffersen GR. Metabotropic glutamate receptor subtype 5 antagonism in learning and memory. Eur J Pharmacol 2010;639(1–3):17–25. [182] Jarius S, et al. GABAB receptor antibodies in paraneoplastic cerebellar ataxia. J Neuroimmunol 2013;256(1–2):94–6. [183] Wysota B, et al. Autoimmune gabab antibody encephalitis associated with nonmalignant lung lesion. J Neurol Neurosurg Psychiatry 2013;84(11):e2. [184] Tofaris GK, et al. Immunotherapy-responsive chorea as the presenting feature of LGI1-antibody encephalitis. Neurology 2012;79(2):195–6. [185] Klein CJ, et al. Insights from LGI1 and CASPR2 potassium channel complex autoantibody subtyping. JAMA Neurol 2013;70(2):229–34. [186] Krastinova E, et al. Treatment of limbic encephalitis with anti-glioma-inactivated 1 (LGI1) antibodies. J Clin Neurosci 2012;19(11):1580–2. [187] Balint B, et al. Caspr2 antibodies in limbic encephalitis with cerebellar ataxia, dyskinesias and myoclonus. J Neurol Sci 2013;327(1–2):73–4. [188] Vincent A, Irani SR. Caspr2 antibodies in patients with thymomas. J Thorac Oncol 2010;5(10 Suppl. 4):S277–80. [189] Ogawa E, et al. VGCC antibody-positive paraneoplastic cerebellar degeneration presenting with positioning vertigo. Neurol Sci 2011;32(6):1209–12. [190] Shams'ili S, et al. Paraneoplastic cerebellar degeneration associated with antineuronal antibodies: analysis of 50 patients. Brain 2003;126(Pt 6):1409–18. [191] Briani C, et al. Spectrum of paraneoplastic disease associated with lymphoma. Neurology 2011;76(8):705–10. [192] Graus F, et al. Immunological characterization of a neuronal antibody (anti-Tr) associated with paraneoplastic cerebellar degeneration and Hodgkin's disease. J Neuroimmunol 1997;74(1–2):55–61.