key: cord-017954-vobslprh
authors: Croxford, J. Ludovic; Miyake, Sachiko
title: Animal Models for the Study of Neuroimmunological Disease
date: 2016-03-26
journal: Neuroimmunological Diseases
DOI: 10.1007/978-4-431-55594-0_3
sha: 
doc_id: 17954
cord_uid: vobslprh

The development and use of numerous animal models of human autoimmune diseases have provided important advances in our understanding of pathogenic mechanisms of disease and provided robust and reliable models to test novel therapeutic strategies. However, few preclinical studies of therapeutic treatments have demonstrated efficacy in the clinic, possibly because of the biological differences between humans and other animals. Although animal models of human disease are imperfect, it is important to understand the differences between the human disease and its animal models and to design experimental studies using animal models appropriately for the questions being asked. This review provides an overview of the currently used animal models of three human neuroimmunological diseases, multiple sclerosis, Guillain-Barré syndrome, and myasthenia gravis, as well as the advantages and disadvantages of each model and how they correlate or differ from their human counterpart.

Following decades of research, scientists have developed a large number of drugs and therapeutic agents that can be used to reduce the symptoms and severity of a number of neuroimmunological diseases such as multiple sclerosis (MS), neuritis, and myasthenia gravis. However, none of these treatment methodologies is curative, and many have a limited life span, such as interferon (IFN)-β that induces neutralizing antibodies in some MS patients [1] , or cause serious side effects (progressive multifocal leucoencephalopathy) such as monoclonal anti-very late antigen (VLA)-4 antibodies (natalizumab, Tysabri®) [2] and thus cause patients to drop out of clinical trials or stop taking the treatment. In addition, although many studies have dissected the intricate pathways involved in the aetiology, development, and pathogenesis of diseases such as MS, we still do not have a definitive understanding of how these diseases manifest and, indeed in the case of MS, whether it is a single disease or rather a spectrum of disorders with similar characteristics. Therefore, researchers have used animal models to aid our understanding of disease pathways, immune cell functions, and pathology. An example of the importance of animal models was the demonstration that CD4 T cells specific for a myelin epitope injected into naïve animals were sufficient to induce a central nervous system (CNS) demyelinating disease with CNS lesions similar to those observed in MS [3, 4] . In addition, these models are a useful preclinical aid to developing novel therapeutic agents. Many different animal models of neuroimmunological disorders have been developed over the last few decades with some success. However, it is important to understand the advantages and limitations of each model to ensure that the correct model is being used for the purpose of the study.

In this review, we provide an overview of the currently available neuroimmunological disease animal models, describing their advantages and disadvantages as well as their relationship and correlation to the human disease they are attempting to model. Although no one animal model is completely identical with human disease, they are nevertheless very useful and important tools with which to increase our understanding of neuroimmunological disease when used correctly.

MS is a human, inflammatory, demyelinating disease of the CNS, with characteristic demyelinating lesions containing immune cell infiltrate and activated CNS-resident cells located in the white matter of the brain and spinal cord; therefore, studies to investigate its pathology are difficult and often rely on autopsy tissues. In these cases, patients might have had the disease for decades and therefore the tissues are representative of disease at the end stage. Furthermore, epidemiological studies have suggested that the initiating factors for MS might be infections that occur during childhood [5] . Therefore, animal models of MS are critical for the study of susceptibility, disease initiation, and pathology during the early stages of disease. Experimental autoimmune encephalomyelitis (EAE) is an autoimmune T cell-mediated disease of the CNS mediated primarily by CD4þ T cells and a commonly used animal model for the study of MS. Disease is characterized by perivascular lesions containing inflammatory infiltrates and nerve conduction block that causes reversible hind-limb paralysis. In addition, late-stage EAE animals also develop axonal demyelination and loss, a pathological hallmark of MS, which leads to severe permanent disability. Characterization of the immune cell infiltrate (T cells, B cells, activated macrophages, and microglia) and the pathology of perivascular demyelinating lesions have shown similarities with human MS lesions [6, 7] although in EAE the spinal cord is the target organ, whereas in MS the brain is more often targeted, especially the white matter. However, EAE is a useful model to study the inflammatory stage of MS and shares some characteristics of MS such as optic neuritis, increased susceptibility of females, perivascular lesions, axonal demyelination, and partial remyelination, as well as eventual axonal loss and flaccid-limb paralysis.

There are two commonly used models of EAE. Active EAE is induced by subcutaneously immunizing genetically susceptible animals, usually rodents, with myelin antigen in Freund's complete adjuvant-containing mineral oil and Mycobacterium tuberculosis strain H37RA. In some EAE low-responder mouse strains, such as C57BL/6, additional injections of pertussis toxin are required. During the induction phase, myelin-specific T cells are activated by antigen-presenting cells (APC) presenting myelin peptide fragments in the draining lymph nodes to produce T helper 1 (Th1) type cytokines, such as interferon (IFN)-γ and tumour necrosis factor (TNF)-α, which allows them to escape the lymph nodes and traffic to the CNS. The effector phase of disease involves the extravasation of activated myelinspecific T cells through the blood-brain barrier and into perivascular spaces in the spinal cord. Here, the encephalitogenic T cells encounter CNS-resident cells mediating further stimulation of pro-inflammatory cytokines and chemokines that mediate a secondary influx of other peripheral inflammatory cells including B cells and mononuclear phagocytes. A study in EAE and a viral model of MS demonstrated that APCs in the CNS restimulated myelin-specific T cells to initiate epitope spreading to other myelin antigens and perpetuate disease [8] . The consequence of this pro-inflammatory cytokine milieu is the demyelination of CNS axons, thought to be mediated in part by numerous mechanisms including phagocytosis by activated mononuclear cells, destructive effects of anti-myelin antibodies, the production of free radicals, and the direct cytotoxic effects of pro-inflammatory cytokines secreted by activated CD4þ T cells and monocytes.

Initially guinea pigs and rats were the animals of choice for EAE studies. However, with the advent of transgenic and gene knockout technology, mice have become the more commonly used animal for EAE studies. Numerous mouse models of EAE have been developed, and these are differentiated by the strain of mouse used and the immunodominant myelin peptide for that particular strain. Commonly used EAE-susceptible mouse strains include SJL, B10.PL, C57BL/6, C3H, SWR, and Biozzi ABH. Depending upon the immunizing myelin antigen and mouse strain used, different forms of EAE can be induced. The most commonly used models are SJL mice immunized with proteolipid protein (PLP) 139-151 that induces a relapsing-remitting form of disease, the most common form of MS, and C57BL/6 mice immunized with myelin oligodendrocyte protein (MOG) 35-55 that initiates a chronic-progressive type of EAE. Other strains, such as Biozzi ABH mice, can be immunized with a suspension of whole Biozzi ABH spinal cord to develop a very reproducible relapsing-remitting disease [9] . The clinical symptoms of EAE usually develop at 7-15 days post-immunization and include weight loss, loss of tail tone and gait, and eventually paralysis in either or both hind limbs. Weight loss precedes the onset of disease symptoms and thus can be used as a marker for disease onset. Daily observation of EAE symptoms and body weight is noninvasive and is therefore a great advantage to researchers, allowing the disease process to be followed, especially when studying the effects of drug treatments.

Of note, some mouse strains are resistant to EAE (A/J, C3H/HeJ, AKR, NZW, and DBA/2). In addition to the importance of environmental factors, it is thought that multiple predisposing genetic elements might be involved in susceptibility to MS (reviewed by Ebers 1994) [10] . Therefore, the backcrossing of EAE-resistant mice with EAE-susceptible strains has been useful for genetic susceptibility studies.

The second model of EAE is passive and involves the in vitro restimulation of encephalitogenic T cells with the myelin peptide used to immunize the original T-cell donor animals [11] . The successful culture of these cells usually requires the addition of Th1 cytokines such as IL-12 [12] . Once cells have been sufficiently activated, they are intravenously administered to naïve animals, where they traffic to the CNS and induce disease. The CNS pathology and disease course are similar to that for active EAE. The ability of MBP-specific T cells to induce EAE in rats and mice was some of the earliest evidence to suggest that MS has an autoimmune aetiology [3, 4] . Another evidence for an autoimmune inflammatory pathogenesis includes the presence of MHC class II-restricted CD4þ T cells that recognize myelin antigens such as myelin basic protein (MBP), PLP, and MOG in MS patients as well as healthy individuals and that MHC class II genes are associated with MS susceptibility [13] [14] [15] . Because T cells are already activated when administered into naïve animals, passive EAE is a suitable model for investigating the effector phase of disease. It has some advantages over passive EAE as an immunization step is not required; thus, there is no antigen reserve that continually activates naïve T cells, which likely boosts the immune system nonspecifically. Furthermore, the direct injection of effector T cells into naïve mice allows the definitive starting point of disease induction to be known, which might be useful for treatment studies, and finally encephalitogenic T cells can be directly tracked in vivo to study methods of extravasation into the CNS and for isolation of antigen-specific T cells.

Early studies demonstrated that activated myelin-specific Th1 CD4þ T cells secreting IFN-γ, TNF-α, and IL-2 were encephalitogenic and sufficient to induce EAE when transferred into naïve mice of a susceptible genetic background [16] [17] [18] . However, the knockout of genes encoding IFN-γ or TNF-α [19, 20] exacerbated EAE and therefore the role of Th1-induced EAE was not straightforward. IL-17 production by CNS-infiltrating T cells is important for blood-brain barrier dysfunction and lesion formation in MS patients [21, 22] , and the genetic inhibition of IL-17A in mice was sufficient to partly ameliorate EAE in mice [23] . This indicated that Th17 cells, an alternative effector T-cell subset to Th1, might be the critical effector cells in EAE. Indeed, Th17 cells cultured in the presence of IL-23 and other cytokines become highly pathogenic and can induce EAE when transferred into naïve mice [24] . Th17 cells show an enhanced efficiency at inducing EAE compared with Th1 cells [24] and therefore might require less cell manipulation in vitro for adoptive transfer studies.

Although there is a predominance of EAE papers investigating the role of myelinspecific CD4þ T cells in EAE, MS lesions have been reported to contain greater numbers of CD8þ T cells compared with CD4þ T cells [6, 25, 26] . However, the function of CD8 T cells in MS lesions is unclear, and therefore, the study of CD8þ T cells in EAE is important. A number of EAE models induced by MHC class I-specific CD8þ T cells have been reported. MBP79-87-specific CD8þ T-cell clones isolated from MBP-immunized C3H wild-type mice induced EAE with numerous neurological deficits (ataxia, spastic reflexes, spinning) as well as hindlimb paralysis when intravenously injected into C3H wild-type mice [27] . This form of EAE was very severe and all mice were moribund by day 14. Interestingly this model showed different clinical symptoms to the CD4þ T cell-mediated type of EAE, but importantly had some similarities to MS, in that perivascular lesions were predominant in the brain, compared with CD4þ EAE where perivascular lesions are located in the spinal cord.

Another study reported that MOG35-55-specific CD8þ T cell lines could also induce a severe, chronic form of EAE when adoptively transferred to naïve C57BL/ 6 mice [28] . In contrast to the MBP-induced CD8 T-cell model of EAE, lesions were present in both the spinal cord and brain. Differences in genetic background, availability of myelin antigens in the CNS, or induction procedure might explain the differences observed between the two models. In contrast to these early CD8 studies, more recent investigations have indicated inhibitory roles for CD8 T cells in EAE: neuroantigen-specific autoregulatory CD8þ T cells inhibited autoimmune demyelination by modulating dendritic cell functions [29] and IL-15-dependent CD8þ CD122þ T cells ameliorated EAE by reducing IL-17 production by CD4þ T cells [30] .

Humanized mouse models have also been developed to study the effect of MS-related myelin epitopes presented to CD8 T cells by human HLA molecules and showed that MOG181-189 presented by HLA-A*201 to MOG-specific CD8 T cells exacerbated CD4 T cell-induced EAE [31] .

Overall, these models of CD8 T cell-induced EAE are important for determining the function of CD8 T cells during CNS autoimmune disease, i.e. pathological versus inhibitory/regulatory effects, and have an advantage over CD4 T cellinduced EAE in that the target organ is predominantly the brain, similar to that in MS.

A number of spontaneous mouse models of EAE have been developed by the transgenic expression of T-cell receptors (TCR) specific for myelin antigens (PLP, MOG, or MOG) in T cells that overcome the negative selection of autoreactive T cells in the thymus (reviewed by Croxford 2011) [32] . Therefore, these models represent a more "natural" type of disease and are useful for examining the role of autoimmune T-cell activation by environmental stimuli. However, depending on the type of study involved, for example, drug efficacy testing, the use of spontaneous EAE models is not recommended because of the wide variance in disease onset and severity.

The advantages of EAE are its robust and reproducible disease course that is useful for studying the early stages of disease initiation; the activation of immune cells, especially antigen-specific T cells; as well as mechanisms in the effector phase such as the role of resident CNS cells, regulatory T cells, and other regulatory mechanisms. In addition, pathological studies at all stages of disease can provide important information regarding sites and mechanisms of demyelination, remyelination, nerve conduction block, and axonal loss. Furthermore, depending upon the models used, the different mechanisms involved in relapsing vs chronic forms of disease can be studied. The reader is encouraged to read review articles describing the specific induction protocols for the numerous EAE models that are available [11, 12] .

The most significant disadvantage of EAE is the location of pathology in spinal cord, unlike MS, where many lesions occur in the brain. Furthermore, although EAE is clearly an autoimmune disease, MS does not have all the hallmarks of an autoimmune disease, and evidence for a specific immunodominant myelin antigen is lacking. Although CD4þ T cells are critical for disease induction in most EAE models, their role in MS is less clear, where many other types of immune cells such as CD8þ T cells, B cells, and monocytes probably also have important roles. For example, B cells have little or no role in many EAE models, at least at the early time points usually studied; however, their importance in MS is suggested by the beneficial effects of anti-CD20 therapy in some MS patients [33] . EAE is a very useful tool to test potential new immunomodulatory therapies. Although four approved MS treatments were studied in EAE (natalizumab, mitoxantrone, glatiramer acetate, and fingolimod), the large majority of new treatments fail when they enter MS clinical trials. An explanation for this might be that treatment studies using the EAE model do not mimic the patient disease course in the clinic. For example, the administration of drugs before the initiation of EAE disease is only useful for indicating an effect on the activation of T cells and has no real clinical significance for MS treatment, where patients have often experienced symptoms for some time before treatments are started. In addition, treatments administered before the onset of symptoms that prevent paralysis are often said to prevent demyelination. However, if these treatments are immunomodulatory, then it is difficult to differentiate between their anti-demyelination and immunosuppressive effects. Therefore, potential treatments should be administered after the onset of EAE symptoms for clinically meaningful results. In summary, although EAE is an imperfect model of MS, its correct usage by investigators is critical when studying novel therapeutic compounds.

A number of etiological studies have indicated the potential role of viruses in the susceptibility, onset, and exacerbation of MS. Therefore, in addition to immunization models that are useful to investigate the early immunological pathways involved in pathogenesis, a number of virus-induced demyelinating models have been developed that allow the study of the potential viral aetiology of MS. Using these models a number of hypotheses of the mechanism of disease onset have been developed including molecular mimicry, epitope spreading, direct bystander activation, and release of cryptic epitopes.

Theiler's murine encephalomyelitis (TMEV) is a single-strand RNA virus that belongs to the cardiovirus group of the Picornaviridae and is a natural mouse pathogen. TMEV-induced demyelinating disease (TMEV-IDD) develops in susceptible strains of mice (SJL/J) upon intracranial injection of the BeAn 8386 strain of TMEV, which persistently infect microglial cell populations in the CNS [34, 35] . The persistent infection of TMEV is thought to cause demyelination mediated by macrophage bystander destruction activated by TMEV-specific CD4þ T-cell responses that target the persistent viral infection, leading to the release of myelin antigens that can activate autoreactive PLP156-171-specific T cells that have escaped negative selection [36] . Epitope spreading to other myelin epitopes propagates the disease, which shows similar inflammatory and demyelinating pathological hallmarks as those seen in MS patients [37, 38] . The onset of a chronic progressive demyelinating disease with no recovery or remission periods occurs around day 30-35 post-infection and continues for over 100 days. TMEV-IDD is characterized by the development of a spastic hind-limb paralysis and perivascular and parenchymal lesions in the spinal cord with demyelination of white matter tracts containing mononuclear cell infiltrates.

Although epitope spreading is thought to be involved in TMEV-IDD, molecular mimicry, the mistaken recognition of a pathogenic epitope for a "self" epitope due to shared amino acid sequences, is another potential mechanism for the induction of autoimmunity. Early studies demonstrated that the immunization of viral peptides could stimulate myelin peptide-specific T-cell responses in vivo [39, 40] . However, the need for epitope processing and the role of a "live" viral infection to stimulate the host innate immune system are not addressed by short-length peptide immunization. Therefore, recombinant TMEV strains engineered to incorporate 30-mer myelin or bacterial/viral myelin mimic epitopes were used to study the potential viral induction of autoimmunity by molecular mimicry [41, 42] . Infection of SJL mice with TMEV engineered to express a PLP mimic peptide derived from Haemophilus influenzae, a natural mouse pathogen, with 6/13 homologous amino acids to PLP139-151, including primary TCR and MHC class II contact residues, induced a mild but rapid onset CNS disease [42] . Interestingly, the immunization of mice with the Haemophilus influenzae peptide in complete Freund's adjuvant failed to induce overt disease, indicating the importance of pathogen-delivered innate immune signals for the induction of disease by molecular mimicry. Of note, viral peptide sequences in TMEV do not share any sequence homology with PLP, the immunodominant myelin peptide in TMEV-IDD in SJL mice, therefore indicating the role of the engineered Haemophilus influenzae peptide sequence in disease induction. These studies were expanded to include TMEV expressing 30-mer from murine hepatitis virus (MHV), which only shares 3/13 amino acids with PLP139-151, and demonstrated the importance of a proline residue at the secondary MHC class II contact point [43] .

Semliki Forest virus (SFV) strain A774 is a neurotropic, single-stranded RNA alphavirus of the Togaviridae family that induces a demyelinating disease upon intraperitoneal injection in SJL/J mice. Disease is characterized by virus-induced demyelination of the CNS. Despite the clearance of the virus by the immune system, maximal demyelinating lesions with the expression of IFN-γ and TNF-α are observed at 14 days post-infection up to 1 month in BALB/c mice, whereas both demyelination and pro-inflammatory cytokine expression can be detected in SJL/J mice up to 1 year [44] . SFV infection induces an MBP-specific T-cell response [45] and demyelination is mediated by CD8 T cells [46, 47] .

MHV, a group II positive-strand RNA coronavirus, is a natural pathogen of mice that upon intracranial injection causes an acute encephalomyelitis that develops into a chronic CNS immune-mediated demyelinating disease with some clinical and pathological similarities with MS [48] . MHV can induce either acute or chronic forms of disease. The acute form is characterized by the production of pro-inflammatory cytokines and chemokines that eventually reduce MHV viral load in the CNS. However, persistent infection of spinal cord white matter tracts propagates and promotes antiviral responses that cause demyelination leading to symptoms of limp tails and partial to complete hind-limb paralysis, similar to that in EAE. In contrast to EAE and other viral models of CNS demyelinating disease, myelin-specific T cells and epitope spreading are not thought to be required for demyelination; rather it is the persistent effect of antiviral immune responses including macrophages and CD4þ and CD8þ T cells that cause chronic demyelination (reviewed by Lane 2010) [49] .

In summary, TMEV-IDD is a demyelinating model that is associated with persistent infection of the CNS, whereas SFV infection of SJL mice might represent a model of MS where immune-mediated demyelination is triggered by a virus infection of the CNS that is cleared efficiently by the host immune system. Although some studies have indicated MS might have a viral aetiology, the mechanisms involved are unknown but might involve an infection in early life that primes autoreactive T cells by molecular mimicry or a persistent infection of the CNS that causes demyelination via epitope spreading and/or molecular mimicry [50] . Thus, both mouse models allow the study of how viruses might induce CNS demeylinating autoimmune disease. In contrast, MHV provides a different scope for study, the underlying mechanisms that mediate host defence against an acute viral infection that later becomes chronic and which is associated with CNS demyelination and neurological symptoms in the absence of myelin-specific T-cell responses.

Guillain-Barré syndrome (GBS) is a rapid autoimmune-mediated acute disease of the peripheral nervous system, which is often caused by infection with Campylobacter jejuni. Clinical disease characteristics usually occur rapidly following onset, usually within 4 weeks, and include neuromuscular paralysis, progressive limb weakness, and deficits of the sensory and autonomic systems as well as cranial nerve involvement. The generation of antibodies to cell-surface gangliosides highly expressed on peripheral nerves by cross-reactivity to Campylobacter jejuni outer membrane components (lipo-oligosaccharides) mediates disease [51] and is thought to be caused by postinfectious molecular mimicry (reviewed by Shahrizaila 2011) [52] .

Two forms of GBS have been characterized: (i) a primary demyelinating form (acute inflammatory demyelinating polyradiculoneuropathy [AIDP]) and (ii) those with axonal involvement (acute motor axonal neuropathy [AMAN], acute motorsensory axonal neuropathy [AMSAN]), with either the presence or absence of antiganglioside antibodies [53, 54] . The pathogenic mechanisms of the AIDP form of GBS include the demyelination of the peripheral nerve myelin sheath by inflammation including upregulated adhesion molecules, pro-inflammatory cytokines, and the infiltration of CD4þ T cells and plasma cells secreting anti-ganglioside antibodies and macrophages [53] [54] [55] [56] , whereas the axonal forms of GBS (AMAN, AMSAN) usually do not include lymphocyte or monocyte involvement, rather Wallerian degeneration as well as complement and anti-ganglioside antibodies that directly mediate myelin destruction [57] .

Disease is often acute and the recovery of motor function is common in approximately 60 % of GBS patients. However, in other cases, residual sensory deficits remain, and in extreme cases, severe alterations to the autonomic nervous system can cause death by respiratory failure, embolism, or cardiac arrest. Current treatment of GBS is generally nonspecific and typically includes the administration of intravenous immunoglobulins or plasmapheresis, which reduce the duration to recovery. However, neither of these treatments is curative, and therefore, the use of animal models of GBS to identify pathogenic mechanisms and to test novel treatments is important.

Experimental autoimmune neuritis (EAN) is a commonly used, robust, and highly predictable animal model of GBS. It was originally developed in rabbits [58] , but has since been induced in a wide variety of animals including rats, mice, rabbits, guinea pigs, and monkeys. Disease is induced by the active immunization of susceptible animals with purified whole peripheral nerve myelin or specific myelin protein components (myelin protein 2 [P2], myelin protein zero [p0]) in complete Freund's adjuvant [58] [59] [60] . EAN in rats and mice is a monophasic acute demyelinating inflammatory disease of the peripheral nervous system [58] with many similarities to GBS including clinical, immunological, and morphological characteristics. In rats, onset of disease is observed at 12-13 post-immunization with a peak of disease severity at day 16. Clinical disease is characterized by tail and limb weakness, and histopathological analyses have shown the presence of nerve oedema and demyelinated peripheral nerves accompanied by the infiltration of inflammatory cells, features which are present in GBS. Symptoms are likely to be caused by a combination of nerve conduction block and the effects of demyelination. During the effector phase of EAN disease, typical pro-inflammatory components including chemokines (MIP-1α and MIP-1β, MCP-1, RANTES, and IP-10) (reviewed by Fujioka 1999) [61, 62] , cytokines (reviewed by Zhu 1998) [63] , and adhesion molecules (VCAM-1) [64] have been shown to be upregulated. Th1-type cytokines might mediate the demyelination of peripheral nerves as evidence shows that the addition of IFN-γ can enhance EAN, whereas blockade of IFN-γ with neutralizing antibodies can ameliorate EAN symptoms and disease course [65] .

EAN mouse models with severe clinical symptoms and pathological features similar to rat EAN have also been reported. An early study reported the induction of EAN SJL/J mice, which showed subclinical damage to peripheral nerve myelin but without clinical symptoms, in contrast to EAN Lewis rats, that developed typical hind-limb weakness and histopathology of the peripheral nervous system [66] . A subsequent study demonstrated that the addition of pertussis toxin to SJL/J mice immunized with bovine peripheral nerve myelin in complete Freund's adjuvant enhanced the mild disease seen in the absence of pertussis toxin [67] . When immunized mice with pertussis toxin were treated with recombinant mouse IL-12, the disease course duration was prolonged and recovery was delayed. Histological analysis demonstrated severe demyelination of the caudae equinae and sciatic nerves during the recovery stage as well as mononuclear cell infiltration. Of note, although C57BL/6 mice were initially thought to be resistant to EAN induction, immunization of male C57BL/6 mice with a synthetic P0 180-199 peptide induced the clinical and pathological characteristics of acute monophasic EAN [68] . As previously shown, the addition of intravenously administrated pertussis toxin increased the incidence of EAN and enhanced inflammation and demyelination of peripheral nerves.

Similar to that seen for EAE, in addition to immunization models of EAN, adoptive transfer models of EAN have been reported. P2-and P0-specific CD4þ T cell lines have been shown to transfer histopathologically similar EAN to naïve syngeneic Lewis rat recipients when injected intravenously [69] [70] [71] . However, the onset of disease was earlier (day 7) in a P2-adoptive transfer EAN model compared with active immunization EAN models (day 12-13).

Although EAN is a robust, reproducible model of neuritis, sharing many pathological and clinical features with GBS, one critical disadvantage is that Campylobacter jejuni infection, which is thought to be involved in a high percentage of human cases of GBS, does not induce disease in rats. Furthermore, immunization of rats with various gangliosides, the main immunological target of the immune system in Campylobacter jejuni-induced GBS, also does not induce disease [72] . The addition of gangliosides to immunization protocols failed to have an enhancing effect of EAN severity or disease onset. Although other species of animals have shown some promise in terms of developing conduction block (rabbits immunized with GM1 demonstrated sciatic nerve conduction block [73] , 33/100 chickens administered Campylobacter jejuni isolated from a Chinese patient showed sciatic nerve Wallerian degeneration with minor demyelination) [74] , these models have not been studied extensively. Therefore, the existing EAN model is useful for studies related to the effector phase immune-mediated mechanisms of peripheral nerve demyelination; it is less useful for studies to determine the pathogenic mechanisms involved following Campylobacter jejuni infection.

Myasthenia gravis (MG) is a rare neuromuscular disease that causes excessive fatigue and generalized muscle weakness that can fluctuate and is characterized by clinical symptoms including ptosis and diplopia. Muscle weakness often worsens upon use but usually improves with rest. As the disease course of MG progresses, bulbar and respiratory muscle weakness worsens, which can become life-threatening, often requiring the use of mechanical ventilation by intubation. MG is thought to be a T cell-dependent antibody-mediated autoimmune disease because approximately 80 % of MG patients have autoantibodies against acetylcholine receptors (AChR) [75, 76] , possibly as a consequence of the loss of "self"tolerance in the thymus [77] . AChR antibodies bind to AChR expressed at neuromuscular junctions (NMJ) and block neuromuscular transmission. Interestingly, the first experimental evidence to suggest AChR antibodies might be pathogenic was demonstrated using an experimental rabbit model where immunization with purified acetylcholine receptor in complete Freund's adjuvant induced the production of antibodies to acetylcholine receptor, which mediated neuromuscular blockade that caused flaccid paralysis and MG-like symptoms [78] .

Despite the early identification of AChR antibodies as potential mediators of disease, the mechanisms involved that precede the production of AChR antibodies are less clear, although involvement of the thymus has been indicated. The symptomatic treatment of MG using acetylcholinesterase inhibitors and/or immunosuppressive drugs can improve muscle function although these treatments are limited by either a reduction in efficacy over time or severe side effects. Furthermore, plasma exchange and surgery in patients that develop thymoma are invasive procedures. Currently, no treatments directly target the autoimmune component of disease, and therefore, there is a still a need for the further elucidation of MG pathogenesis and the development of novel drugs, which highlights the importance of using animal models of MG.

Experimental autoimmune myasthenia gravis (EAMG) was originally induced in rabbits by immunization of highly purified AChR isolated from the electric organ of Electrophorus electricus emulsified in complete Freund's adjuvant [78] . This induced the production of antibodies that specifically recognize AChR and bind to these receptors at NMJs, subsequently blocking neurotransmission, causing muscle fatigue and weakness, which mimic the symptoms observed in human MG. Since the first description of EAMG, it has been induced in a wide variety of animals including rabbits [78, 79] rats [80] , mice [81, 82] , guinea pigs [80] , goats [83] , monkeys [84] , and frogs [85] .

Currently, rat and mouse models are the animal models of choice, as gene knockout and transgenic technology has allowed a greater in-depth investigation into the specific molecules involved in disease pathogenesis. Susceptible rat strains include Lewis, Fischer, and Wistar-Munich rats, and nonresponder strains include Wistar Furth and Copenhagen rats. It is important to note that some mouse strains have different susceptibilities to EAMG; H-2 b, s haplotype strains (C57BL/6 and SJL/J) are high responders, whereas H-2 k, p haplotypes are nonresponders; therefore, it is important to determine the strain used before initiating studies [81, 86] . Interestingly, rats appear to be more susceptible to EAMG than mice, as usually only one immunization with AChR in complete Freund's adjuvant is required to induce autoantibodies to AChR. In contrast, susceptible mouse strains usually require two or three immunizations with AChR in complete Freund's adjuvant. Furthermore, disease severity in mouse EAMG is reduced compared with the rat model, and therefore, this is an important consideration if the efficacy of novel therapeutic agents is to be tested. Of note, susceptibility to both MG and EAMG is linked to the HLA/MHC region [87] . Thus, EAMG has many clinical and pathological similarities to human MG, especially the presence of autoantibodies that recognize and bind to AChR.

Following immunization, T cells mediate the production of AChR-specific antibodies in murine EAMG as well as human MG [88] [89] [90] , and treatment with anti-CD4 or anti-Ia antibodies can block the induction or induce remission of EAMG [89, 91] . Further evidence for a role of T cells in EAMG pathogenesis was shown by the use of lymphocyte immunosuppressive agents and oral tolerance to AChR that inhibited disease onset [92] [93] [94] . Similar to that observed in EAE, the role of pro-inflammatory cytokines including IL-1, IL-12, IFN-γ, and TNF-α is also important in the onset of EAMG, with functions related to T-cell development, proliferation, and differentiation. Studies in EAMG demonstrated that treatment of EAMG rats or mice with anti-TNF-α treatments reduced EAMG development [95] and significantly improved established disease [96] . Confirmation of a role for TNF in MG was demonstrated by a trial investigating the use of a TNF inhibitor (etanercept) in MG patients that reduced muscle weakness [97] .

Once the autoreactive T cells become activated, they stimulate B cells to produce and secrete anti-AChR antibodies that induce the observed clinical symptoms. EAMG studies have indicated that following the binding of AChR antibodies to NMJs, complement activation including C3 and C9 deposition and membrane attack complex might mediate the destruction of NMJ plasma membranes [98] [99] [100] . Confirmation of a role for complement in EAMG was shown in studies reporting that the blockade of the complement system by complement inhibitors protected rats against the induction of EAMG [101] [102] [103] . Importantly, the role of complement was indicated in human MG [99, 100, 104] , indicating the validity of EAMG.

Clinical signs of active EAMG (tremor, hunched posture, muscle weakness, and fatigue) usually occur 3-10 days after the second immunization, and mice are observed for signs of muscle weakness by the paw-grip test at least once a week. In addition, a number of other tests have been developed to assess the extent of disease including the quantitative measurement of muscle weakness by electromyography, the evaluation of EAMG induction by the quantification of muscle AChR loss, and serum anti-AChR antibody levels by radioimmunoassay or ELISA. Therefore, active EAMG is a useful model for myasthenia gravis and the extent of disease can be measured using fairly noninvasive techniques.

Another method for the induction of EAMG is the passive transfer model, where autoantibodies specific for AChR from donor AChR-immunized animals are injected daily into naïve animals, and this was first shown in a rat model [83, 105] . Another method for the passive induction of EAMG is the injection of AChR-specific antibodies isolated from the serum of MG patients [106] .

Interestingly, the transfer of EAMG by autoreactive lymphocytes is less robust than that by autoantibodies [107] and indicates that autoreactive antibodies are probably the major mechanism involved in the onset of NMJ destruction.

In summary, active EAMG is useful for investigating the induction phase of disease (T-and B-cell activation and autoantibody production) including loss of selftolerance, mechanisms of antigen-specific immune response induction, and their modulation by therapeutic agents (to induce tolerance or immunosuppression), whereas passive EAMG is useful for the study of the effector phase of disease including IgG deposition in NMJs, complement molecules, and investigating treatments using regulatory proteins to prevent the degradation of NMJs. Clinical disease in the passive model of EAMG is similar to that observed in active EAMG and therefore serves as a useful model of MG when full activation of the immune system is not required, as is seen following active immunization protocols.

Although most MG patients develop autoreactive antibodies to AChR, approximately 20 % of MG patients are AChR antibody negative; however, 30-40 % of these MG patients have antibodies that recognize muscle-specific kinase (MuSK) [108] . MuSK is a tyrosine kinase receptor that is involved in the development of postsynaptic membranes in the NMJ. However, whether MuSK antibodies are involved in the pathogenicity of MG is unclear and it is unknown whether they contribute to muscle weakness. Therefore, the use of animal models is useful to help dissect the potential pathogenic function of MuSK antibodies. A recent study demonstrated that immunization of rabbits or mice with a MuSK ectodomain induced muscle weakness as measured by electromyographic analysis and flaccid paralysis, which was similar to that in human MG [109, 110] . However, the appearance of disease-related symptoms is longer than that for AChR-active and AChR-passive models of EAMG, and the passive transfer of MuSK antibodies is less effective than the equivalent AChR-passive model. Therefore, MuSK antibody-related MG might represent a subtype of MG or reflect differences in environmental or genetic susceptibility factors.

EAMG is a very useful animal model for the study of the pathways involved in disease pathogenesis as well as investigating novel therapeutic strategies. Importantly, EAMG shares similar symptoms with MG, especially muscle weakness and fatigue. Furthermore, they share such immunopathological features as AChRspecific antibodies in the serum, muscle AChR loss, presence of complement factors such as C3 and C9 at NMJs, and a supportive role for T and B cells as regards autoantibody production. One of the major differences between EAMG and MG is the involvement of the thymus (loss of self-tolerance) in human MG, which is absent in EAMG, where tolerance must be "broken" by immunization of the autoantigen in complete Freund's adjuvant. Although many therapeutic strategies have been demonstrated to be beneficial in EAMG, few have translated to the clinic for human MG. This might be due to the human form of disease having a more complex aetiology than EAMG. However, EAMG still has an important role to play in preclinical studies, whether to investigate pathogenic mechanisms or potential therapeutic strategies.

Despite the numerous beneficial advances of researchers in each of these human neuroimmune diseases, the underlying mechanisms of many of these disease processes are still unclear and this has hindered the search for curative therapies.

Although the use of human tissues and samples is critical for these types of studies, often they are invasive and provide an indication of a single time point within a disease process that might have been ongoing for years or decades. Therefore, it is vital to use animal models to fill in the "gaps" that cannot be analysed by research using human tissues. However, whilst animal models of disease can provide useful information, it is important to note that no single model is a perfect model of its human counterpart and each has their advantages and disadvantages. Therefore, it is important to use the correct model for the study involved, i.e. immune cell activation, mechanisms of demyelination, immune cell trafficking into the target organ, induction vs effector phases of disease or routine drug testing. In addition, it is important to remember that each mouse strain used is genetically representative of a single human individual; therefore, multiple strains should ideally be used when developing novel treatments or investigating pathogenic mechanisms, to reduce the possibility of strain irregularities confounding the results.

Incidence and significance of neutralizing antibodies to interferon beta-1a in multiple sclerosis. Multiple Sclerosis Collaborative Research Group (MSCRG)

Progressive multifocal leukoencephalopathy complicating treatment with natalizumab and interferon beta-1a for multiple sclerosis

The rapid isolation of clonable antigen-specific T lymphocyte lines capable of mediating autoimmune encephalomyelitis

Adoptive transfer of experimental allergic encephalomyelitis in SJL/J mice after in vitro activation of lymph node cells by myelin basic protein: requirement for Lyt 1þ 2-T lymphocytes

Epidemiologic evidence for multiple sclerosis as an infection

Multiple sclerosis: distribution of T cell subsets within active chronic lesions

Immunohistochemical analysis of the cellular infiltrate in multiple sclerosis lesions

Epitope spreading initiates in the CNS in two mouse models of multiple sclerosis

Induction of chronic relapsing experimental allergic encephalomyelitis in Biozzi mice

The role of genetic factors in multiple sclerosis susceptibility

Experimental autoimmune encephalomyelitis in the mouse

Mouse models of multiple sclerosis: experimental autoimmune encephalomyelitis and Theiler's virus-induced demyelinating disease

A genome screen in multiple sclerosis reveals susceptibility loci on chromosome 6p21 and 17q22

A full genome search in multiple sclerosis

A complete genomic screen for multiple sclerosis underscores a role for the major histocompatibility complex. The Multiple Sclerosis Genetics Group

Encephalitogenic T cells in the B10.PL model of experimental allergic encephalomyelitis (EAE) are of the Th-1 lymphokine subtype

Surface expression of alpha 4 integrin by CD4 T cells is required for their entry into brain parenchyma

Lymphotoxin and tumor necrosis factor-alpha production by myelin basic protein-specific T cell clones correlates with encephalitogenicity

Fathman CG Mice with a disrupted IFN-gamma gene are susceptible to the induction of experimental autoimmune encephalomyelitis (EAE)

TNF is a potent anti-inflammatory cytokine in autoimmune-mediated demyelination

Gene-microarray analysis of multiple sclerosis lesions yields new targets validated in autoimmune encephalomyelitis

Human TH17 lymphocytes promote blood-brain barrier disruption and central nervous system inflammation

IL-17 plays an important role in the development of experimental autoimmune encephalomyelitis

IL-23 drives a pathogenic T cell population that induces autoimmune inflammation

Immunohistological analysis of T lymphocyte subsets in the central nervous system in chronic progressive multiple sclerosis

The immunopathology of multiple sclerosis: an overview

A pathogenic role for myelin-specific CD8(þ) T cells in a model for multiple sclerosis

Myelin antigenspecific CD8þ T cells are encephalitogenic and produce severe disease in C57BL/6 mice

Neuroantigen-specific autoregulatory CD8þ T cells inhibit autoimmune demyelination through modulation of dendritic cell function

IL-15-dependent CD8þ CD122þ T cells ameliorate experimental autoimmune encephalomyelitis by modulating IL-17 production by CD4þ T cells

CD8 T cell responses to myelin oligodendrocyte glycoprotein-derived peptides in humanized HLA-A*0201-transgenic mice

Mouse models for multiple sclerosis: historical facts and future implications

B-cell depletion with rituximab in relapsing-remitting multiple sclerosis

Monocytes/macrophages isolated from the mouse central nervous system contain infectious Theiler's murine encephalomyelitis virus (TMEV)

Split tolerance of Th1 and Th2 cells in tolerance to Theiler's murine encephalomyelitis virus

Class II-restricted T cell responses in Theiler's murine encephalomyelitis virus (TMEV)-induced demyelinating disease. III. Failure of neuroantigen-specific immune tolerance to affect the clinical course of demyelination

Pathology of the early lesion in multiple sclerosis

Primary demyelination in Theiler's virus infection. An ultrastructural study

Expansion by self antigen is necessary for the induction of experimental autoimmune encephalomyelitis by T cells primed with a cross-reactive environmental antigen

Molecular mimicry in T cell-mediated autoimmunity: viral peptides activate human T cell clones specific for myelin basic protein

A virus-induced molecular mimicry model of multiple sclerosis

Viral delivery of an epitope from Haemophilus influenzae induces central nervous system autoimmune disease by molecular mimicry

Structural requirements for initiation of cross-reactivity and CNS autoimmunity with a PLP139-151 mimic peptide derived from murine hepatitis virus

Long-term effects of Semliki Forest virus infection in the mouse central nervous system

Predisposition to EAE induction in resistant mice by prior infection with Semliki Forest virus

The pathogenicity of the M9 mutant of Semliki Forest virus in immune-compromised mice

Semliki Forest virus induced, immune mediated demyelination: the effect of irradiation

Dissociation of demyelination and viral clearance in congenitally immunodeficient mice infected with murine coronavirus JHM

The pathogenesis of murine coronavirus infection of the central nervous system

Epitope spreading and molecular mimicry as triggers of autoimmunity in the Theiler's virus-induced demyelinating disease model of multiple sclerosis

Serologic evidence of previous Campylobacter jejuni infection in patients with the Guillain-Barre' syndrome

Guillain-barré syndrome animal model: the first proof of molecular mimicry in human autoimmune disorder

Pathogenesis of Guillain-Barré syndrome

The role of cytokines in Guillain-Barré syndrome

Acute immunoinflammatory neuropathy: update on Guillain-Barré syndrome

From bench to bedside-experimental rationale for immune-specific therapies in the inflamed peripheral nerve

Pathology of the motor-sensory axonal Guillain-Barre' syndrome

Allergic neuritis: an experimental disease of rabbits induced by the injection of peripheral nervous tissue and adjuvants

Identification of the neuritogen for experimental allergic neuritis

P0 myelin protein produces experimental allergic neuritis in Lewis rats

Sequential expression of chemokines in experimental autoimmune neuritis

Chemokines and peripheral nerve demyelination

Cytokine production and the pathogenesis of experimental autoimmune neuritis and Guillain-Barré syndrome

The role of the very late antigen-4 and its counterligand vascular cell adhesion molecule-1 in the pathogenesis of experimental autoimmune neuritis of the Lewis rat

The role of interferon-gamma in the pathogenesis of experimental autoimmune disease of the peripheral nervous system

Experimental allergic neuritis in the SJL/J mouse: dysfunction of peripheral nerve without clinical signs

Experimental allergic neuritis in the SJL/J mouse: induction of severe and reproducible disease with bovine peripheral nerve myelin and pertussis toxin with or without interleukin-12

P0 protein peptide 180-199 together with pertussis toxin induces experimental autoimmune neuritis in resistant C57BL/6 mice

Transfer of experimental allergic neuritis with P2-reactive T-cell lines

Cell adhesion molecules of the immunoglobulin supergene family as tissue-specific autoantigens: induction of experimental allergic neuritis (EAN) by P0 protein-specific T cell lines

Induction of experimental allergic neuritis in the BN rat: P2 protein-specific T cells overcome resistance to actively induced disease

Effects of gangliosides on the expression of autoimmune demyelination in the peripheral nervous system

Experimental sensory neuropathy induced by sensitization with ganglioside GD1b

Experimental Campylobacter jejuni infection in the chicken: an animal model of axonal Guillain-Barre' syndrome

Myasthenia gravis: past, present, and future

Autoimmune channelopathies and related neurological disorders

Thymus and myasthenia gravis: antigen processing in the human thymus and the consequences for the generation of autoreactive T cells

Autoimmune response to acetylcholine receptor

Experimental myasthenia in rabbits: immunological, electrophysiological and morphological aspects

Experimental autoimmune myasthenia: a model of myasthenia gravis in rats and guinea pigs

Strain differences in the autoimmune response of mice to acetylcholine receptors

Genetic control of experimental autoimmune myasthenia gravis in mice

Immunological studies of acetylcholine receptors

Experimental autoimmune myasthenia induced in monkeys by purified acetylcholine receptor

Myasthenia in frogs immunized against cholinergic receptor protein

Immunogenetics of experimental autoimmune myasthenia gravis

A molecular basis for MHC class Il-associated autoimmunity

Autoimmune human T lymphocytes specific for acetylcholine receptor

Immuno-therapy for myasthenia gravis: a murine model

Animal models of myasthenia gravis

In vivo therapy with monoclonal anti-IA antibody suppresses immune response to acetylcholine receptor

Daunomycin treatment prevents clinical expression of experimental autoimmune myasthenia gravis

Oral administration of acetylcholine receptors: effects on experimental myasthenia gravis

IFN-α treatment suppresses the development of experimental autoimmune myasthenia gravis

Anti-TNF-alpha antibodies suppress the development of experimental autoimmune myasthenia gravis

Treatment of experimental autoimmune myasthenia gravis with recombinant human tumor necrosis factor receptor Fc protein

Etanercept treatment in corticosteroid dependent myasthenia gravis

Ultrastructural localization of immune complexes (IgG and C3) at the end-plate in experimental autoimmune myasthenia gravis

Ultrastructural localization of the terminal and lytic ninth complement component (C9) at the motor end-plate in myasthenia gravis

Myasthenia gravis: quantitative immunocytochemical analysis of inflammatory cells and detection of complement membrane attack complex at the end-plate in 30 patients

Inhibition of acute passive transfer experimental autoimmune myasthenia gravis with Fab antibody to complement C6

Novel complement inhibitor limits severity of experimentally myasthenia gravis

Anti-C5 antibody treatment ameliorates weakness in experimentally acquired myasthenia gravis

Immune complexes (IgG and C3) at the motor end-plate in myasthenia gravis: ultrastructural and light microscopic localization and electrophysiologic correlations

Passively transferred experimental autoimmune myasthenia gravis

Myasthenia gravis: passive transfer from man to mouse

Experimental autoimmune myasthenia: cellular and humoral immune responses

Auto-antibodies to the receptor tyrosine kinase MuSK in patients with myasthenia gravis without acetylcholine receptor antibodies

Induction of myasthenia by immunization against muscle-specific kinase

Myasthenia gravis experimentally induced with muscle-specific kinase