| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Published online before print May 12, 2006
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* Medical Microbiology and Immunology and the M.I.N.D. Institute and
Division of Rheumatology, Allergy and Clinical Immunology, University of California Davis, Sacramento
1 Correspondence: M.I.N.D. Institute, UC Davis, Wet Lab Building, 50th Street, Sacramento, CA 95817. E-mail: pashwood{at}ucdavis.edu
 |
ABSTRACT |
|---|
Key Words: autism spectrum disorder (ASD) • neurodevelopment • neurokine • immunity • inflammation • cytokines
|
INTRODUCTION |
|---|
There is emerging evidence and growing concern that a dysregulated or abnormal immune response may be involved in some forms of ASD. In general, the links between the immune and neurological systems are becoming increasingly well known. Cytokines and other products of immune activation have widespread effects on neuronal pathways and can alter behaviors such as mood and sleep. Aberrant immune activity during critical periods of brain and neuronal development could potentially play a role in neural dysfunction, typical of autism. Various hypotheses have attempted to link dysfunctional immune activity and autism, such as maternal immune abnormalities during early pregnancy, increased incidence of familial autoimmunity, and childhood vaccinations. Several lines of research have shown abnormalities in the nature, extent, and regulation of the immune response in autism, including a skewed generation of antibodies, cytokines, and immune cells. This is a review of current research linking immune dysfunction to autism.
|
NEUROLOGICAL ABNORMALITIES IN AUTISM |
|---|
30% of children with ASD develop epilepsy by adolescence [19
], and an additional, significant minority has subclinical epilepsy, as measured by epileptiform encephalogram, especially during sleep [20
, 21
]. These findings clearly indicate that there are neurological involvements in ASD that affect the development and differentiation of neurons in the brain. Immune dysregulation could result in the generation of localized or systemic inflammation and/or the release of immunomodulatory molecules that could influence, alter, or modify neurodevelopment and/or neuronal function, especially at critical times of development. |
IMMUNE SYSTEM ABNORMALITIES IN AUTISM |
|---|
(IFN-), and tumor necrosis factor 
(TNF-), which have pleiotropic effects in the CNS, including an emerging role in neurodevelopment [22
]. For example, the inflammatory cytokine IL-6 can induce sleep, and TNF- can induce anorexia [23
, 24
]. Indeed, products of immune activation including cytokines may be responsible for many common features of autism, such as mood and sleep disturbances. In turn, neuropeptides, derived from the central and peripheral nervous system, have profound effects on the immune system, including the chemotaxis and recruitment of innate immune cells [25
]. A number of findings in the immune system of patients with ASD point toward immune system dysregulation/dysfunction (Table 1
). Systemic, immunologic aberrations in autism have been linked with autoimmunity, leading to the generation of antibodies that are reactive to CNS proteins and have the potential for neuronal tissue destruction, and second, with dysfunctional immunity, such as abnormalities or deficits of function in immune cell subsets, leading to an inappropriate or ineffective immune response to pathogen challenge [56
].
|
|
CYTOKINES IN AUTISM |
|---|
were increased in autistic children compared with controls, but no changes were seen for IL-6, TNF-, and IFN- [40
], suggesting a potential TH1 shift. However, in a previous study, plasma IFN- was elevated in 10 autistic children when compared with four adult, control subjects [64
]. In another study, increased plasma IFN- levels were observed in 29 autistic children; it is interesting that these findings correlated positively with the generation of the intercellular CNS messenger and marker of oxidative stress, nitric oxide (NO) [35
]. In a separate study, the same authors observed that the macrophage product neopterin was present at high levels in serum samples from individuals with ASD compared with controls, which may reflect increased cell-mediated immune activation and IFN- production [36
]. These higher IFN- and neopterin levels correlated significantly with elevated, circulating numbers of monocytes observed in autistic children [36
]. In addition, neopterin and biopterin were shown to be increased in urine samples of ASD children [37
]. Recently, Zimmerman et al. [42
] found decreased levels of neopterin and quinolinic acid but increased levels of biopterin in the CSF from ASD patients compared with a control group of other neurological diseases. Increased soluble TNF receptor 1 was also observed in ASD serum samples, but cytokine levels were similar in ASD compared with neurological disease controls [42
]. Conversely, in cell culture experiments in which intracellular cytokine production was examined in 20 autistic patients compared with 20 aged-matched controls, intracellular production of IL-4 was increased, but IFN- and IL-2 were decreased in CD4+ and CD8+ lymphocytes following stimulation [41
], suggesting a TH2 bias. Further in vitro studies of peripheral blood mononuclear cells stimulated with lipopolysaccharide (LPS) have shown an inappropriate innate immune response evinced by amplified production of proinflammatory cytokines TNF- and IL-1ß in ASD patients compared with controls [43
]. It is interesting that similar immune dysregulation of increased TNF- was also found in primary sibling family members of patients with ASD, indicating a possible similar genetic susceptibility in the patients studied. Taken together, these papers, while revealing differences in cytokine levels between ASD and typically developing or neurological disease controls, highlight the importance of carefully controlled, age-matched studies in the field. Many of these papers compare young ASD patients with adult controls or have a wide range of ages in both controls and case groups. Furthermore, drug treatments prescribed for the symptoms of ASD, including naltrexone, clozapine, risperidone, and tricyclic antidepressants, are all capable of affecting the immune response and cytokine production. Therefore, it cannot be discounted that medication status is a confounding factor in a number of these studies. Moreover, there is an increasing realization that ASD may be comprised of many different phenotypes, which share the same behavioral commonalities. The wide variety of disorders encompassed on the autism spectrum, the degree of severity, and variety of symptoms and co-morbidities often make it difficult to interpret the results of individual studies; however, it would appear that cytokine immune abnormalities are consistently observed in ASD subjects. Overall, these cytokine studies indicate first that a more complex pattern of cytokine production occurs in autism, which is not defined easily by the traditional TH1/TH2 paradigm, and second, that the differences between studies may be indicative of a possible patient selection bias and that particular cytokine profiles may potentially reflect different autism behavioral phenotypes. Cytokines can activate and exert trophic effects on glial cells, which can in turn produce cytokines and chemokines upon such activation. Cell culture studies have shown that neuropoietic cytokines such as IL-6 can have direct effects on neurons and glia, including changes in proliferation, survival, death, neurite outgrowth, and gene expression [65 , 66 ]. As the CNS is populated largely by astroglia and microglial cells, these cytokine-cell interactions are important for neuronal cell functioning and development. Recently, Vargas et al. [67 ] investigated the presence of immune activation in postmortem brain specimens and CSF from subjects with autism. The authors found active neuroinflammation in the cerebral cortex and cerebellum of brain tissue in autism. This inflammatory process was characterized by a marked cellular activation of microglial and astroglial cells and the presence of an altered cytokine pattern. In addition to activated microglia and astroglia, there was an accumulation of perivascular macrophages and monocytes but an absence of lymphocytes and antibody from the brain specimens, thus pointing toward an innate immune activation. In addition, an enhanced proinflammatory cytokine profile was observed in their CSF. These results suggest that abnormal immune responses in the neuroglia of autistic patients may influence neural function and neural development and that an aberrant immune response may contribute to the development of autism.
In general, the brain and CNS are considered to be protected and isolated from potentially harmful pathogens or agents within the blood, including inflammatory immune cells and proteins, by the blood brain barrier (BBB). Cytokines however, can gain entry into the brain through active transport mechanisms or at circumventricular regions, where the barrier is less controlling [68
]. Cytokines and inflammatory mediators may impair BBB function by binding to receptors on the endothelial cells directly. In addition, lymphocytes capable of mediating immune responses and production of cytokines can migrate into the brain from the blood via the CSF to the choroid plexus or from the blood to either the subarachnoid space or parenchymal perivascular space [69
]. It is important to note that peripheral cytokines could also directly affect afferent neurons and their functions [70
]. Autonomic innervation of immune organs such as the bone marrow, thymus, spleen, and lymph nodes plays a major role in immune system development; equally, the immune response is capable of changing expression and distribution of neural receptors in these organs [71
]. Cytokines can affect many behaviors including mood, sleep, appetite and nutritional uptake, exploratory behavior, and social interactions. For example, systemic cytokine administration at therapeutic doses of IFN-, IL-2, and TNF- has effects, including mood depression, sleep disorder, impaired cognitive function, decreased exploratory behavior, and changes in motivation [72
, 73
]. Systemic administration of cytokines can induce increased noradrenergic, dopaminergic, and serotonergic metabolism in the hypothalamus, hippocampus, and nucleus accumbens and modulate synaptic plasticity and thereby alter memory and learning [74
].
Together with aberrant cytokine profiles, several studies have shown abnormal levels of blood lymphocytes in autism. Significantly decreased CD4+ T cells have been observed in ASD [30
31
32
33
]. It is interesting that in mouse models, systemic T cell deficiency leads to learning and memory impairment, which can be ameliorated by T cell replacement [75
]. In early studies, decreased responsiveness to T cell mitogens such as phytohemagglutinin has been shown in ASD [26
, 27
]. Furthermore, an incomplete or partial activation of T cells following stimulation, with an increased expression of HLA-DR+ but not the IL-2R chain (CD25), was observed in ASD [28
, 29
]. NK cells are an important cytotoxic cell subset of the innate immune system and an important cytokine contributor, in particular, IFN-. In children with autism and in children with the related disorder Rett syndrome, lower levels of circulating numbers of NK cells are noted compared with controls [38
, 39
]. Furthermore, decreased NK cell activity demonstrated by target cell lysis has been shown in ASD children [39
]. Reduced numbers and activity of NK cells could impair the ability to eradicate or prevent viral infections in these children, which could potentially be damaging to neural tissues during critical windows of CNS development.
Abnormal concentrations of plasma Ig classes have been observed in some ASD children [30 , 47 , 48 ]. Moreover, a skewing or imbalance in Ig isotype has been observed, with increased IgG2 and IgG4 present in ASD [47 ]. Trajkovski et al. [48 ] showed a similar skewing with skewed IgG1 and IgG4 as well as IgM and IgG levels. These imbalances in Ig levels may be indicative of an underlying autoimmune disorder and/or an aberrant susceptibility to infections.
|
"NEUROKINES" IN AUTISM |
|---|
-aminobutryic acid (GABA), have been studied in ASD [76
]. For example, in postmortem brain specimens obtained from patients with ASD, there was a 48–61% decrease in glutamic acid decarboxylase, an enzyme that converts glutamate into GABA, in the parietal and cerebellar regions of the brain compared with controls [77
]. In ASD, this may cause suppression of the GABA-ergic system, resulting in heightened stimulation of the glutamate system, which has been associated with seizures. Second, excitotoxic damage of neurons, possibly resulting from glutamate hyperactivity, may result in abnormal, structural development of the brain [78
].
The neurotransmitter serotonin has a wide range of affects on normal physiological functions including circadian rhythyms, appetite, mood, sleep, anxiety, motor activity, and cognition. Serotonin is detected, not only in neurons of the nervous system but also in platelets and lymphocytes of the immune system, where it can exert dose-dependent, suppressive, or proliferative effects. In normal development, serotonin levels are high in the brain up until the age of five and then decrease dramatically [79
]. Serotonin levels increase in the hypothalamus, hippocampus, and cortex in response to various cytokines, such as IL-1ß, IFN-, and TNF- [74
, 80
]. Moreover, enzymes that control the conversion of tryptophan into serotonin are under the influence of IFN- and IL-1 [81
]. Increased serotonin levels in peripheral blood platelets have been described in approximately one-third of patients with autism [82
]. It is interesting that selective serotonin (5-HT) reuptake inhibitors have been shown to be beneficial in treating obsessional and repetitive behaviors in some ASD patients [83
]. The reason for the difference in serotonin levels is unknown; potentially, it may be a result of the presence of inflammatory cytokines or more likely, to alterations in the platelets themselves, which could modify serotonin uptake [84
]. In addition, proinflammatory cytokines IL-1ß, IFN-, and TNF- are capable of affecting the activity of the serotonin transporter gene, a potential susceptibility gene in ASD [85
, 86
].
Cytokines and chemokines play a major role in many stages of development of the CNS and are known to induce the secretion of many neurotransmitters and neuropeptides [22 , 87 ]. In turn, neuropeptides play an important role in all phases of immune system development, often acting as trophic factors, which has led to the hypothesis that neurotrophins (NTs) should be considered as neurokines, as they act in a cytokine-like manner, influencing the development and function of the immune system [88 ]. Several NTs with potent immunomodulatory actions, including neuropeptide Y, substance P, calcitonin gene-related peptide (CGRP), vasoactive intestinal peptide (VIP), BDNF, and NT-4/5, which have multiple affects on neurodevelopment and neuron maintenance, have been implicated in ASD. Analysis of neonatal blood spots by recycling immunoaffinity chromatography found that BDNF, VIP, CGRP, and NT-4/5 were elevated in ASD compared with typically developing control children but could not be distinguished from those with mental retardation [89 ]. Brain-derived neurotrophic factor is a major player in neurodevelopment known to regulate neuronal cell survival, growth, plasticity and differentiation, and is now considered to be a growth factor with a wide spectrum of functions outside the nervous system, including modulation and regulation of immune function [90 , 91 ].
Based on animal studies, two structurally related neuropeptides, oxytocin and vasopressin, are believed to play a critical role in the formation of social bonding and recognition and in the processing of social cues [92 ]. Prairie voles are highly social animals, which form long-lasting pair bonds; in contrast, montane voles are asocial or solitary and do not form pair bonds [93 ]. Central infusion of oxytocin in female or vasopressin in male prairie voles helps establish partner-bonding; this phenomenon can be blocked using specific antagonists [94 , 95 ]. Furthermore, oxytocin knockout mice have normal, cognitive abilities but diminished social recognition, suggesting a key role of oxytocin in social interactions [96 ]. In ASD patients, Modahl et al. [97 ] found significantly lower levels of plasma oxytocin when compared with age-matched, normal subjects. Moreover, this decrease in oxytocin levels may be a result of a reduction in the processing of oxytocin, as increased levels of the pro-hormone form of oxytocin were found in autism patients [98 ]. In prairie voles, oxytocin and vasopressin receptors are located in the ventral forebrain, whereas the pattern of expression of oxytocin receptors differs in montane voles [92 ]. It would seem that not only the concentration of neuropeptides but also the pattern of receptor distribution may be important in the establishment of socially rewarding interactions. So far, signature patterns of neuropeptides and neurotransmitters and their respective receptors have yet to be established in ASD. Further studies that address this issue in ASD may provide clues into the development of impaired social interactions that are present in ASD.
It is interesting that Dunzendorfer et al. [25
] have suggested a novel role for neuropeptides in the regulation of dendritic cell (DC) migration. They investigated locomotion of mononuclear cell-derived DCs at different maturation stages toward gradients of sensory neuropeptides in vitro. Calcitonin gene-related peptide, VIP, secretin, and secretoneurin induced immature DC chemotaxis comparable with the potency of the chemokine regulated on activation, normal T expressed and secreted (RANTES), whereas substance P and macrophage-inflammatory protein-3ß (MIP-3ß) stimulated immature cell migration only slightly [25
]. Moreover, the neuropeptide VIP synergized with cytokines such as TNF- in the induction of DC maturation [99
]. In the CNS, DCs have been found in normal meninges, the choroid plexus, and CSF and are actively recruited during inflammation, where they may play equal roles in the defense against infections and contribute to the break-down of tolerance to CNS autoantigens [100
]. These findings suggest a central role for DC- and neuropeptide-mediated chemotaxis in the control of CNS inflammation and the generation of T cell reactivity against CNS antigens, and present an intriguing concept in the context of autism.
|
AUTOIMMUNITY AND AUTISM |
|---|
Various anti-brain antibodies have been found in autistic patients, including autoantibodies to serotonin receptor [20
], myelin basic protein (MBP), neuron axon filament protein, cerebellar neurofilaments, nerve growth factor, -2-adrenergic-binding sites, anti-brain endothelial cell proteins, and antibodies directed against an as-yet unknown brain protein [49
50
51
52
53
, 106
]. The pathophysiological significance of these antibodies reported in children with autism is uncertain. For example, increased autoantibodies would suggest that there is increased neuronal damage, as is the case in multiple sclerosis, where following demyelination, MBP is unmasked, and there is a subsequent generation of antibodies. However, evidence of demyelination in autism has remained elusive [107
]. Glial fibrillary acidic protein (GFAP), measured in the CSF of 47 autistic childen, was elevated significantly compared with 10 age-matched control children, suggesting that gliosis and unspecific brain damage may occur in autism [108
]. However, as GFAP correlates strongly with age, most likely as a result of age-dependent expansion of fibrillary astrocytes, caution must be shown in interpreting these data [99
]. Needless to say, in the absence of neuronal damage, the presence of serum antibodies to brain tissues may be abnormal and may be detrimental to appropriate development and function of the CNS. In a study by Singh and Rivas [109
], antibodies directed to the rat caudate nucleus (the portion of the brain responsible for assembly of peripheral information) were found in 49% of the autism patients evaluated and in none of the control cases.
It is also important to note that for each antibody tested, the number of autistic children showing positivity is far from 100%. The observations of elevated anti-CNS antibodies in autism are at best unconfirmed and in some cases, such as serotonin receptors and MBP, markedly conflicting. Furthermore, it is difficult to determine whether the autoantibodies present in the plasma of patients with autism contribute to the development of the disorder or if they are a consequence of the disease. However, the collective findings of autoimmunity in families and the plethora of anti-brain antibodies suggest that in some patients, autoantibodies that target the CNS may be a pathological or an exacerbating factor in neuronal development in children with ASD. It can be inferred that increased autoimmunity may be confined to a subset of autistic patients. Indeed, large cohort studies with thoroughly defined and specifically phenotyped autistic patient groups and well-matched age and sex controls need to be performed to confirm the potential role of autoantibodies in the pathology of all or subsets of autistic patients. Moreover, the development of an animal model will be crucial to determine the role of autoantibodies in the pathology of autism.
|
MUCOSAL IMMUNE RESPONSE AND AUSTISM |
|---|
20% of young children previously diagnosed with autism [112
]. In contrast, prospective reports from pediatric gastroenterology and general autism clinics have described GI symptoms in 46–84% of patients with ASD [110
]. However, prevalence estimates from population-based epidemiologic studies are largely lacking. Reported GI abnormalities include low activities of disaccharidase enzymes, defective sulfation of ingested phenolic amines (tylenol), bacterial overgrowth with greater diversity and number of clostridial species, more numerous Paneth cells, increased intestinal permeability, and positive effects on behavioral cognition following dietary intervention [110
, 113
114
115
116
]. Clinical and pathological studies have described an apparently characteristic GI immunopathology in this subset of children with ASD [117
, 118
], in which chronic, ileo-colonic lymphoid nodular hyperplasia (LNH) and entero-colitis are key features [118
]. The mucosal lesion consists of a pan-enteric lymphocytic infiltrate with a variable degree of acute inflammation and eosinophil infiltration [118
119
120
]. Flow cytometric and immunohistochemical analyses of mucosal lymphocyte populations in this ASD subgroup have demonstrated qualitatively consistent abnormalities at different anatomical sites including stomach, duodenum, ileum, and colon [120
121
122
], which indicate a relatively homogenous, mucosal lymphocyte infiltrate. Although LNH is not an uncommon finding in children with allergies or immunodeficiency, there is increased frequency and severity in patients with autism [123
]. Indeed, these findings, in particular, LNH and associated colitis, have been described by others as "not normal" [124
, 125
]. In addition, other findings included a focal deposition of serum IgG from ASD children with GI symptoms, which colocalize with complement C1q on the basolateral enterocyte membrane; these changes were not seen in histologically normal and inflamed mucosa of developmentally normal children or children with cerebral palsy [121
]. This focused immune response direct to the epithelia may firstly be indicative of an autoimmune process directed against self-antigen contained within epithelial cells and secondly, is suggestive of an inflammatory process that may perturb the intestinal barrier function in this ASD subgroup. Increased basement membrane thickness and abnormal patterns of epithelial glycosaminoglycans in this ASD subgroup are a further indication of inflammatory degradation, which could contribute to disruption of the intestinal barrier function [120
]. Although there is a great deal of speculation, the exact mechanism of how mucosal changes may influence autistic development or behavior is still not clear. It is plausible that these data reflect a primary intestinal immune activation and immunopathology, which leads to heightened, systemic immune activation and results in neuroinflammation. Indeed, in a recent study, the generation of chronic colitis in an animal model through the rectal administration of trinitrobenzene sulfonic acid not only caused GI inflammation but also led to the activation of brain areas that are abnormal in autism, as measured by c-Fos expression [126 ]. In celiac disease, it is recognized that primary mucosal immunopathology, as a result of gliadin intolerance, can produce secondary neurological disease including cerebral inflammation, dementia, cerebellar ataxia, epilepsy, and heterotopic cerebral calcification [127 ]. Further investigation of gut-brain interactions in this cohort of children with ASD and GI symptoms is necessary to clarify the potential links with the intestinal pathology and the effect on behaviors.
It is interesting that mucosal lymphocytes isolated from the duodenum, ileum, and colon as well as peripheral lymphocytes of ASD patients with GI symptoms showed increased, spontaneous production of proinflammatory, intracellular cytokines, most notably TNF-, when compared with aged-matched controls, including those with similar symptoms of constipation [44
, 45
]. These data support the hypothesis that there is mucosal immune dysregulation with a proinflammatory lymphocyte cytokine profile in ASD children. These findings have since been confirmed in peripheral blood, where proinflammatory cytokines were increased upon stimulation with dietary proteins in similarly affected autistic children compared with controls [46
]. In some studies, circulating antibodies to food substances, namely the casein and gliadin, have been found [128
, 129
]. However, these antibodies are also found with similar frequency to that in the general population. Furthermore, antibodies to neuronal-specific antigens in the sera of children with autism could cross-react with dietary peptides, including milk butyrophilin, Streptococcus M protein, and Chlamydia pneumoniae [129
], suggesting that bacterial infections and milk antigens may modulate an autoimmune process in autism.
It may be that antigens in the diet can cross into the mucosa more easily via a disrupted intestinal barrier, where they cause local, inflammatory reactions generating proinflammatory cytokine signals, which interact with afferent neurons. Another possibility is that the failure to detoxify neuroactive antigens from the gut may lead to cognitive impairments. In an open-label trial administration of vancomycin, an antibiotic, which is poorly absorbed, resulted in objective, cognitive improvements in autistic children [130 ], presumably as a result of treatment of intestinal dysbiosis. However, once administration of vancomycin was ceased, the patients’ cognitive functions regressed, suggesting that the initial improvement was a result of beneficial effects on the intestinal pathology [130 ]. The exacerbation of GI and behavioral symptoms in autism induced by certain foods, particularly those containing gluten and casein, has been shown through dietary intervention and their removal from the diet [131 ]. Autistic children on gluten and casein-free diets also showed significantly lower eosinophil infiltrate in intestinal biopsies compared with those on a conventional diet [30 ]. The significance of this finding is still unclear. However, it has been proposed recently that immune responses associated with allergy may contribute to the pathogenesis of autoimmune diseases of the CNS in humans and in animal models [132 ].
The increased passage of exorphins and/or opioids from the diet such as gliadomorphin and casomorphin into the body, where they may interact with the CNS, could play a role in inducing the behavioral features of autism. Opioid peptides and opioid receptors are important modulators of neural development, influencing migration, proliferation, and differentiation within the CNS [133
]. Peripherally, opioid peptides are contained and/or produced by the gut, lung, placenta, testis, lymphoid tissue, and immune cells, but also another important source of opioids is from the diet. The endogenous opiates ß, , and endorphins can directly influence the immune response, enhancing generation of cytotoxic T cells and NK cells, and antibody synthesis and act as chemoattractants for monocytes and neutrophils [134
]. The precise mechanisms that underlie the immunosuppressive effects of opioids remain unknown; however, they may operate as cytokines, acting through receptors on peripheral blood and/or glial cells [135
]. It has been hypothesized that an excess of opioid peptides will have detrimental effects on brain development and behavior, and that autism may result from abnormal levels or activity of opioid peptides. ß-Casemorphine-7, an opioid exclusively of dietary origin, has been shown to be present in patients with psychoses including autism [136
]. Indeed, the beneficial effects on autistic behavior following dietary exclusion therapy are thought, in part, to be a result of reduced opioid intake [136
, 137
]. Furthermore, therapeutic trials using the oral opioid antagonist naltrexone in some patients with ASD have shown improvements in behavioral characteristics such as repetitive stereotypes, hyperactivity, social contact, and self-injurious behavior [138
].
|
ANIMAL MODELS OF AUTISM/AUTISTIC BEHAVIOR |
|---|
Several lines of research have suggested that direct viral infection of the fetal CNS during a critical time in development could lead to autism (reviewed in ref. [141 ]). Congenital cytomegalovirus (CMV), BDV, and congenital rubella and measles virus are examples of such infections that have been linked to autism. A recent retrospective case study of a child with autism found evidence of a congenital CMV infection through dried blood spots taken at birth [142 ]. In animal models, several studies have shown that neonatal infection with BDV, a persistent, neurotropic RNA virus, leads to ASD-like symptoms, including neuroanatomical, neurochemical, neuroimmune, and behavioral changes [143 ]. A 2004 study demonstrated that BDV infection could interfere directly with proper neuronal connectivity and function in vitro using a rat model [144 ]. It is interesting that genetic background effects were seen in the BDV rat model of Pletnikov et al. [145 ]. Inbred Lewis and Fisher rats, and outbred Sprague-Daley rats were shown to exhibit differential patterns of irregularity dependent on their genetic background, which highlights the importance of the complex interaction of genes on the development of ASD. Congenital rubella has also been related to ASD in a similar manner to CMV and BDV. Fetal infection with rubella has been known to cause miscarriage, stillbirths, and severe birth defects. In the 1970s, several studies demonstrated a correlation between congenital rubella and the development of autism-like behavioral symptoms [146 , 147 ].
Extreme controversy has erupted recently over the possible role of mercury-containing vaccines in the development of ASD. Two different mechanisms have been proposed for the potential role of vaccines in autism. The first suggests that antibodies stimulated by the viral components within the vaccines could potentially cross-react with host tissues and induce autoimmunity. Alternatively, the mercury-containing preservative thimerosal (ethylmercurithiosalicylate), previously found in several vaccines, has been proposed to induce a host response that leads to autism. Organic mercury, such as methyl mercury and ethyl mercury, are potent immunosuppressors, more so than inorganic mercury. Thimerosal is metabolized rapidily to ethyl mercury, and many of its potential effects on the immune system and general health and its toxicology in general are largely unknown and in many quarters, hotly debated. Information about thimerosal has generally been derived as a comparison with methyl mercury, which has been more widely examined as a result of its presence as a common environmental contaminant, primarily through fish consumption. Many hypothesis are based on the assumption that the effects of ethyl mercury and ergo thimerosal on the immune system are similar to those of methyl mercury. One study demonstrated that injection of thimerosal into young SJL mice, a strain that is highly susceptible to the development of autoimmune disorders, caused ASD-like behavioral symptoms, whereas injection into mice, that are less susceptible to autoimmunity (C57BL/6J and BALB/cJ strains) did not cause symptoms [148 ]. The thimerosal dosing and timing regime were designed to mimic the potential exposure present in the pediatric immunization schedule. These findings suggest that genetic factors, such as a predisposition to autoimmune diseases, may play a role in thimerosal-related neurotoxicity. A population study by Geier and Geier [149 ] demonstrated evidence for an association between increasing concentrations of thimerosal in vaccines and the occurrence of neurodevelopmental disorders. However, the overwhelming majority of epidemiological population studies indicates there is no established correlation between vaccinations and autism.
Other potential animal models of ASD have been developed that utilize biochemical anomalies, transgenic models, and neural structural defects. For example, the administration of valproic acid to rats prior to the closure of the neural tube results in behavioral abnormalities similar to those seen in autism [150 ]. In humans, such exposure could arise as a result of the administration of antiseizure medication to pregnant mothers. Experiments where lesions are generated in specific regions of the brain have been performed in order to study the behavioral abnormalities that result from their injury at different time-points during neurodevelopment. Amygdala lesions in primates resulted in social and emotional deficits that were similar to those found in ASD [151 ]. Similar behavioral abnormalities were observed in rats that had undergone the same procedure [152 ]. Other examples of ASD models include the Brattleboro rats, which are unable to secrete vasopressin; the disheveled 1 null mouse, which is homozygous for a targeted deletion of the disheveled 1 gene; and oxytocin knockout mice [153 154 155 ], which show deficits in social memory and social interaction (i.e., pup separation-distress calling, huddling behaviors, and nest-building). The guinea pigs of Caston and colleagues [156 ] were thought to be one of the closest animal models to ASD. Naturally occurring cerebellar defects were seen 3 weeks postnatally, including neuronal dropout and decreases in dendritic arborization and slight cell body shrinkage. The neuroanatomical changes were similar to those found in some patients with ASD. In addition, social interaction and performance of motor learning tasks were deficient compared with controls. However, as a result of "reproduction difficulties", this model is no longer available. The complexity of ASD increases the difficulty of obtaining a multilateral animal model that is complete for all the neuroanatomical, genetic, neurochemical, immunological, and behavioral irregularities, which are present in ASD. However, the development of such a model would greatly enhance research into various aspects of the ASD phenotype and may highlight possible clues to the etiology and pathogenesis of ASD.
Nonhuman primates may represent the best animal model of autism in which several complex and sensitive social behaviors can be assessed in particular animals or in mother-infant pairings. However, these experiments require a high level of expertise and are highly expensive and time-consuming, especially when considering the study of neurodevelopment, which may be protracted to last several years. Murine models hold several advantages in the study of neurodevelopment as a result of their significantly increased rate of development and that in general, neurodevelopmental timelines for specific brain regions are comparable with those in humans. However, in murine models, most neurodevelopment is postnatal compared with humans and can make the interpretation of data regarding exposures difficult. The differences in mass/volume of relative structures may also complicate findings (for example, the olfactory system). Moreover, the greatest problem for any animal model is the identification and consistent measurement of measures of key autistic-like phenomena, such as social behavior, repetitive behavior, and restricted interests. At present, the test of a good animal model strongly depends on the further development of standard behavioral measures of social interaction along with measures of repetitive and stereotyped patterns of behavior, which are essential in a valid animal model.
|
POTENTIAL IMMUNE COMMONALITIES WITH OTHER NEURODEVELOPMENTAL DISORDERS |
|---|
|
were found in patients with TS [178
]. In addition, an increased frequency of specific antibodies for neural proteins have recently been observed in a few patients with TS and in their first-degree family members compared with control groups, pointing toward a potential genetic susceptibility [179
]. In one proposed model of pathogenesis, molecular mimicry may play a role in TS, exhibiting similarities with Sydenham’s chorea [180
]. Sydenham’s chorea is characterized by involuntary movements, obsessive-compulsive and neuropsychiatric symptoms [181
]. Group A streptococcal infection is thought to be responsible for the development of Sydenham’s chorea and rheumatic fever, although the pathogenesis is unknown. In genetically susceptible individuals who have group A ß-hemolytic streptococci infection, antibodies directed against the streptococci are thought to cross-react with CNS structures, particularly gangliosides, that induce abnormal signal transduction leading to CaM kinase II activation, which could result in neurotransmitter imbalance [182
]. Pediatric autoimmune neuropsychiatric disorders associated with striptoeoccal infections (PANDAS) is the term given to the subgroup of patients on the TS and OCD spectrum, who acquire OCD and/or tic disorders following streptococcal infection but do not meet the criteria for Sydenham’s chorea [183 ]. Unlike OCD alone, PANDAS has a relapsing-remitting disease course [184 ]. Rates of tic disorders and OCD in first degree relatives of patients with PANDAS are increased compared with individuals in the general population [185 ]. Effectiveness of immunotherapy, such as plasmapheresis, has been shown to alleviate symptom severity in some patients [186 ]. Enlargement of the basal ganglia, caudate, putamen, and globus pallidus has been reported in functional MRI study of PANDAS patients, as have serum autoantibodies that react with the basal ganglia [187 ], and may pertain to the presence of inflammation in these regions. The effect of antibiotic prophylaxis as a means of preventing the exacerbations of obsessive-compulsive traits and tic disorders in PANDAS patients was found to be effective in preventing post-streptococcal neuropsychiatric episodes [188 ]. PANDAS is still a controversial concept, in part, as a result of the lack of strictly defined classification criteria of symptoms and in part, as a result of the temporal connection between streptococcal infection and neuropsychiatric symptoms, which can be as long as six months [189 ].
It is estimated that as many as 75% of SLE patients have neurological complications [190 ]. A study by Kowal and colleagues [191 ] showed that neuropsychiatric symptoms of lupus are caused by anti-DNA antibodies, which are able to cross-react with N-methyl-D-aspartate (glutamate) receptors, resulting in excitotoxic death of neurons. These autoantibodies have recently been shown to cause neuronal death and affect cognitive processes in an animal model [191 ]. This study showed that systemic immune dysregulation can lead to brain effects in the presence of a BBB made permeable with LPS, which suggests that BBB abrogation by infection, stress, or disease-related factors may also be pivotal in SLE. It is notable that an increase in circulating immune complexes, capable of abrogating the BBB, has also been demonstrated in SLE [192 ].
In Alzheimer’s disease, ß-amyloid is believed to inflict vascular damage to the BBB, facilitating the entry of autoantibodies, which cause antibody-mediated neuronal death and memory impairment [193 ]. In normal controls, the entry into the CNS of these naturally occurring antibodies is prevented by presence of an intact BBB. Vascular changes related with aging (artherosclerosis, hypertension) are risk factors for AD [194 , 195 ]. In addition to amyloid plaques that can be detected in the cerebral cortex, deposits are found in capillaries, meninges, and choroid plexus and are often associated with endothelial damage and basement membrane defects, all of which may affect BBB permeability [196 , 197 ]. Reactive microglia associated with plaques are found in AD brains [198 ]. The dysfunction of the BBB does not always lead to AD; the avidity and affinity of the antibody must be taken into account, as should the cytokine environment, which can affect the permeability and adhesion molecule properties of the BBB [199 ].
|
CONCLUSION |
|---|
TOP
ABSTRACT
INTRODUCTION