Brain-reactive autoantibodies prevalent in human sera increase intraneuronal amyloid-β(1-42) deposition

Previous studies have reported immunoglobulin-positive neurons in Alzheimer's disease (AD) brains, an observation indicative of blood-brain barrier (BBB) breakdown. Recently, we demonstrated the nearly ubiquitous presence of brain-reactive autoantibodies in human sera. The significance of these observations to AD pathology is unknown. Here, we show that IgG-immunopositive neurons are abundant in brain regions exhibiting AD pathology, including intraneuronal amyloid-β(42) (Aβ(42)) and amyloid plaques, and confirm by western analysis that brain-reactive autoantibodies are nearly ubiquitous in human serum. To investigate a possible interrelationship between neuronal antibody binding and Aβ pathology, we tested the effects of human serum autoantibodies on the intraneuronal deposition of soluble Aβ(42) peptide in adult mouse neurons in vitro (organotypic brain slice cultures). Binding of human autoantibodies to mouse neurons dramatically increased the rate and extent of intraneuronal Aβ(42) accumulation in the mouse cerebral cortex and hippocampus. Additionally, individual sera exhibited variable potency related to their capacity to enhance intraneuronal Aβ(42) peptide accumulation and immunolabel neurons in AD brain sections. Replacement of human sera with antibodies targeting abundant neuronal surface proteins resulted in a comparable enhancement of Aβ(42) accumulation in mouse neurons. Overall, results suggest that brain-reactive autoantibodies are ubiquitous in the blood and that a defective BBB allows these antibodies to access the brain interstitium, bind to neuronal surfaces and enhance intraneuronal deposition of Aβ(42) in AD brains. Thus, in the context of BBB compromise, brain-reactive autoantibodies may be an important risk factor for the initiation and/or progression of AD as well as other neurodegenerative diseases.     

Immunoglobulin binding to brain in autoimmune mice

Brain-reactive autoantibodies (BRAA) are thought to play an important role in central nervous system (CNS) manifestations of systemic lupus erythematosus (SLE). Previous studies have shown the existence of BRAA in human and murine SLE. This study was undertaken to establish and characterize the presence of autoantibody binding to brain of autoimmune mice. Laser confocal microscopy was performed on frozen brain sections to detect the presence of immunoglobulin (Ig) in the brain of MRL/lpr and BXSB mice and compare that to control strains of MRL/mp and C57BL/6 mice. There was a dramatic increase in fluorescence in the brains of MRL/lpr and BXSB at 4 months of age. There was little or no Ig detected in the brains of control mice. This increase in presence of Ig in the autoimmune mouse brain was paralleled by an increase in the serum titers of BRAA and anti-DNA autoantibodies as determined by ELISA. These studies provide another link between the existence of brain-reactive autoantibodies and altered CNS functioning.     

Stiff-man syndrome: identification of 17 beta-hydroxysteroid dehydrogenase type 4 as a novel 80-kDa antineuronal antigen

Stiff-man syndrome (SMS) is a rare autoimmune disorder of the central nervous system associated with autoantibodies to glutamate decarboxylase (GAD). We isolated five brain-reactive human monoclonal antibodies, with reactivity distinct from GAD, from peripheral blood of a patient newly diagnosed with SMS. Two antibodies reacted with both Purkinje cells and ependymal cells, and precipitated an 80-kDa protein from rat neuronal primary cultures, which was also recognized by 12% (3/25) of SMS sera and 13% (2/15) of SMS cerebrospinal fluid (CSF) samples. The corresponding antigen was identified as 17 beta-hydroxysteroid dehydrogenase type 4 and may represent a possible novel target of autoimmunity in SMS.     

Autoreactive IgG to intracellular proteins in sera of MS patients.

IgG binding to multiple protein constituents in lysates of Jurkat cells was detected by Western blot in sera of patients with multiple sclerosis (MS) and systemic lupus erythematosus (SLE). The distribution patterns of bands with sera tested against protein lysates from normal Jurkat cells or from Jurkat cells exposed to apoptosis or oxidative stress inducing conditions were similar in most patients, but with inter-individual differences. The number of bands with sera of both patient populations far exceeded those (0 or 2 bands) detected with sera of healthy controls. Proteinase K, RNase and DNase pre-treatment of cell lysates suggested a protein nature for all of the antigens and a ribonucleoprotein (RNP) nature for some of the antigens recognized by serum IgG of MS and SLE patients. Only two MS patients had positive anti-nuclear antibody (ANA) titers, while all of them had positive Western blots. In addition to similarities, dissimilarities were also recognized between the humoral immune responses in MS and SLE. No IgG molecules were detected against phosphorylated proteins in the sera of MS patients, while multiple phosphoproteins were recognized by IgG molecules of SLE patients in immunoprecipitation experiments. These data suggest that in addition to ANA, the sera of MS patients contain autoantibodies directed against multiple intracellular proteins. The protein recognition patterns of immunoglobulins in MS share similarities, but also have distinct features when compared to those in SLE. The biological significance of these autoantibodies in MS remains to be understood.     

Autoantibodies against pituitary peptides in sera from patients with multiple sclerosis

Autoantibodies against pituitary peptides were demonstrated in sera from multiple sclerosis (MS) patients. Ten patients with lupus erythematosus disseminatus (SLE) and 97 healthy blood donors served as controls. The sera were used as primary antibodies in the indirect immuno-enzyme cytochemical (IEC) method, with fixed, paraffin-embedded rat brains and rat and hog pituitaries as antigen substrates. Eleven of the 33 MS sera reacted with peptides in the neural lobe/hypothalamic nuclei or distal lobe. The MS had a significantly higher incidence of peptide antibodies than sera from controls (11/33 vs 9/97). The mean antibody titers were significantly different (1577 vs 333). Comparison with rabbit reference antibodies specific to each of the 6 distal lobe hormones showed that the 9 distal lobe-positive MS sera reacted with cells harboring peptides of the somatotropin family. The presence of peptide autoantibodies was not related to clinical status or medical treatment. No antibodies against pituitary peptides were found in the SLE sera. One of the 11 positive MS sera showed antibodies against gastric parietal cells. None of the 11 sera showed antibodies against muscle, mitochondria, thyroid, adrenal, or parotid antigens. We propose that in a proportion of patients with MS, these autoantibodies might be involved in the demyelinization process by interfering with the peptide/receptor interplay, thus placing MS as a disease in analogy with myasthenia gravis. Alternatively, these autoantibodies might be involved in the altered immunoregulation of MS or be secondary to the disease.     

Characterization of cell surface antigens on the adrenergic neuroblastoma clone A2(1)

Neuroblastoma cell lines have been extensively used to identify the presence of brain reactive autoantibodies in the sera of patients with systemic lupus erythematosus (SLE) who have neuropsychiatric involvement and in the animal models (murine) of this disorder. In this study, a characterization of murine neuroblastoma cell surface antigens, from the adrenergic A2(1) cell line, have indicated both similarities and differences with the cell surface antigens of normal mouse brain. It has also shown that some of these antigens are nervous system specific, whereas others are not. These data indicate that a more precise definition of the antigens on the surface of neuroblastoma cells, with which anti-brain autoantibodies react, is necessary for an understanding of the neuropsychiatric manifestations associated with autoimmune diseases such as SLE.     

Autoantibodies to neuroblastoma cell surface antigens in neuropsychiatric lupus

A human neuroblastoma cell line, LA-N-1 was used as a target cell in a I131 radiolabeled staphylococcal protein-A (I131-SpA) binding assay, to characterize the pattern of antineuronal activity of human sera in fifty-four cases of systemic lupus erythematosus (SLE) including twenty-six patients with neuropsychiatric manifestations of SLE (LE-CNS), out of which ten were pediatric patients aged eleven to eighteen years, thirty-six normal donors and sixteen rheumatoid arthritis patients. The IgG binding activity of normal control sera with LA-N-1 neuroblastoma cells was determined to be 998 +/- 490 cpm I131-SpA, per 5 micrograms LA-N-1 protein (mean +/- SD), 2936 +/- 2607 cpm I131-SpA per 5 micrograms LA-N-1 protein for rheumatoid arthritis patients, 5109 +/- 3304 cpm I131-SpA per 5 micrograms LA-N-1 protein for SLE patients. The binding activity of for LE-CNS patients sera was: 10 565 +/- 2993 and 15 346 +/- 2993 cpm I131-SpA per 5 micrograms LA-N-1 protein, for the pediatric and adult group of patients respectively. Absorption assays disclosed that the antineuronal IgG autoantibody detected in the LE-CNS group of patients is cross reacting with human adult brain, while the anti-neuronal activity of rheumatoid arthritis and SLE patients could be removed by sequential absorption with homogenates of human lung, liver, and kidney, fetal calf serum, human muscle and human lymphocytes. We conclude that detection of autoantibodies binding to LA-N-1 human neuroblastoma cells may be helpful in the diagnostic workup of pediatric and adult LE-CNS patients.     

Intrathecal antibodies and brain damage in autoimmune MRL mice

Neuropsychiatric (NP) manifestations and brain pathology are poorly understood and potentially fatal concomitants of systemic lupus erythematosus (SLE). For many years, autoantibodies to brain tissue (i.e., brain-reactive antibodies, BRA) were proposed as a key factor in pathogenesis of CNS manifestations. Recent evidence suggests that intrathecal BRA, rather than serum autoantibodies, are a better predictor of disturbed brain morphology and function. We presently test this hypothesis by examining the relationship among BRA in cerebrospinal fluid (CSF), behavioral deficits, and brain pathology in a well-established animal model of CNS lupus. We showed earlier that significant diversity in disease manifestations within genetically homogenous MRL-lpr mice allows for constructive and informative correlational analysis. Therefore, levels of CSF antibodies were presently correlated with behavioral, neuropathological and immune measures in a cohort of diseased MRL-lpr males (N = 40). ELISA, Western Blotting, standardized behavioral battery, digital planimetry, HE staining, and immunohistochemistry were employed in overall data collection. The IgG antibodies from CSF were binding to different regions of brain parenchyma, with dentate gyrus, amygdale, and subventricular zones showing enhanced immunoreactivity. High levels of CSF antibodies correlated with increased immobility in the forced-swim test and density of HE⁺ cells in the paraventricular nucleus. Peripheral measures of autoimmunity were associated with other deficits in behavior and neuropathology. This correlation pattern suggests that etiology of brain damage in lupus-prone mice is multifactorial. Intrathecal BRA may be important in altering motivated responses and activity of major neuroendocrine axes at the onset of SLE-like disease.     

Autoantibodies to glutamate receptors can damage the brain in epilepsy, systemic lupus erythematosus and encephalitis

Glutamate is the major excitatory CNS neurotransmitter. Glutamate receptor autoantibodies have now been called to our attention, as they are found in many patients with epilepsy, systemic lupus erythematosus (SLE) and encephalitis, and can unquestionably cause brain damage. AMPA GluR3 autoantibodies have been found thus far in 27% of patients with different epilepsies, while NMDA NR2A or NR2B autoantibodies, some of which cross-react with double-stranded DNA, have been detected in 30% of SLE patients, with or without neuropsychiatric impairments. NR2 autoantibodies were also found in patients with epilepsy (33%), encephalitis and stroke. NR2 and GluR3 autoantibodies do not cross-react in patients with epilepsy. Human and animal studies show that both types of glutamate receptor autoantibodies can certainly damage the brain. GluR3 autoantibodies bind to neurons, possess a unique ability to activate their glutamate-receptor antigen, and cause neuronal death (either by excitotoxicity or by complement fixation independent of receptor activation), multiple brain damage and neurobehavioral/cognitive impairments. In animal models (mice, rats or rabbits) GluR3 autoantibodies may cause seizures, augment their severity or modulate their threshold. NR2/dsDNA autoantibodies, once present in the CNS, can bind and subsequently kill hippocampal and cortical neurons by an excitotoxic complement-independent mechanism. Herein, we discuss epilepsy, autoimmune epilepsy, SLE and neuropsychiatric SLE in general; summarize the up-to-date in vivo and in vitro evidence concerning the presence of glutamate receptor autoantibodies in human diseases; discuss the activity and pathogenicity of different glutamate receptor autoantibodies; and end with our conclusions, recommendations and suggested future directions.     

Synergistic effect of low doses of tumor necrosis factor and sera from patients with systemic lupus erythematosus on the expression of procoagulant activity by cultured endothelial cells

The effect of 15 antiphospholipid antibody (APA) positive SLE sera, 13 APA negative SLE sera, 10 APA negative sera from patients with other connective tissue diseases (CTD) and 15 control sera on the expression of endothelial procoagulant activity (PCA) was studied. Endothelial cells (EC) were stimulated with tumor necrosis factor (TNF) and 20% serum for 4 hr and the surface PCA expression was measured. Without TNF, none of the sera stimulated PCA expression. With suboptimal TNF stimulation, 14/15 APA positive SLE sera, 7/13 APA negative SLE sera, 2/10 CTD sera and 2/15 control sera enhanced PCA expression. This stimulating effect resided in the IgG fraction and was associated with the presence of APA, but not with a history of thrombosis. Purified APA had no PCA stimulating activity. PCA expression was inhibited by cycloheximide and heat treatment (30 min, 56 degrees C) of serum. In conclusion, 21/28 (75%) SLE sera increase the TNF-induced endothelial PCA expression. Although this effect predominantly occurs with APA positive serum, a causative role of APA was not demonstrated.     

Prevalence of anti-cardiolipin, anti-beta2 glycoprotein I, and anti-prothrombin antibodies in young patients with epilepsy

PURPOSE: To measure anti-cardiolipin (aCL), anti-beta2 glycoprotein I (anti-beta2GPI), and anti-prothrombin (aPT) antibodies in young patients with epilepsy, and to correlate their presence with demographic data, clinical diagnoses, laboratory and neuroradiologic findings, and antiepileptic drugs (AEDs). METHODS: Sera from one hundred forty-two consecutive patients with epilepsy with a median age of 10 years were tested for aCL and anti-beta2GPI autoantibodies by solid-phase assays. aPT antibodies also were assayed in sera from 90 patients. Positive results were confirmed after a minimum of 6 weeks. Antinuclear antibodies (ANAs) and antibodies against extractable nuclear antigens (ENAs) also were tested. RESULTS: An overall positivity of 41 (28.8%) of 142 sera was found. Fifteen patients were positive for aCL, 25 for anti-beta2GPI, and 18 for aPT antibodies. Several patients (12%) displayed more than one specificity in their serum. Only one of these patients had a concurrent positivity for ANAs and ENAs. A predominance of younger patients was found in the antibody-positive group. All types of epilepsy were represented in the positive group. No relation between antibody positivity and AEDs was found. Diffuse ischemic lesions at computed tomography (CT)/magnetic resonance imaging (MRI) scans were present in higher percentages in patients who were antibody positive. No positive patient had a history of previous thrombosis or other features related to systemic lupus erythematosus (SLE), and no patient was born of a mother with SLE. CONCLUSIONS: Our study suggests a relation between epilepsy and aPL in young patients. A pathogenetic role for these autoantibodies cannot be excluded, and their determination might prove useful even from a therapeutic point of view.     

Thrombosis associated with antiphospholipid antibodies cannot be explained by effects on endothelial and platelet prostanoid synthesis

The effect of 23 antiphospholipid antibody positive SLE sera, 4 antiphospholipid antibody negative SLE sera and 17 control sera on endothelial prostacyclin and platelet thromboxane A2 production was studied. Endothelial cells and platelets were stimulated with different agonists. Depending on the stimulus used, 4-19% of the SLE sera inhibited the prostacyclin release, whereas 4-28% enhanced prostacyclin production. Our data suggest that the pathophysiological mechanisms underlying decreased prostacyclin production are heterogeneous. Follow-up of two patients showed that prostacyclin inhibitory activity was variable in time. Platelet thromboxane production was normal or increased, but never decreased in the presence of the SLE sera. An imbalance in thromboxane A2/prostacyclin ratio was present in some patients, but did not correlate with a history of thrombosis. We conclude that, in general, interference of antiphospholipid antibodies with endothelial or platelet prostanoid synthesis does not explain the occurrence of thromboembolic manifestations in antiphospholipid antibody positive SLE patients.     

Anti-ribosomal P antibody in schizophrenia

Background: A series of epidemiological, clinical and laboratory findings suggest an autoimmune process in schizophrenia and include, among others, high titers of various autoantibodies in the sera of patients. Antiribosomal P antibody is known to exist in systemic lupus erythematosus (SLE) patients with a psychiatric presentation, including psychosis, rationalizing the examination of its existence in patients with schizophrenia. Methods: Sera of 59 patients, 48 diagnosed with schizophrenia and 11 diagnosed with a schizoaffective disorder, were examined for the presence of antiribosomal P antibody titers using ELISA. The control group consisted of 94 healthy subjects with similar age and gender distribution. Results: Anti-ribosomal P antibody titers were below cut-off level in 58 patients and borderline in one patient, similar to the low titers of the control group. Conclusions: Previous investigations have demonstrated high specificity for anti-ribosomal P antibody in SLE patients with psychosis. In view of the results of this study, however, anti-ribosomal P antibody is not a biological marker for schizophrenia.     

Rasmussen encephalitis and non-herpetic acute limbic encephalitis]

Rasmussen syndrome (RS) and non-herpetic acute limbic encephalitis (NHALE) have pathophysiological background related with autoimmunity to glutamate receptors (GluRs) after infections. RS and NHALE were reviewed, depending mainly on our recent studies. RS is the prototype of autoimmune-mediated epilepsy. In patients with RS, several kinds of autoantibodies against neuronal molecules, for example, GluR3, GluRepsilon2 (NMDA-R2B), etc., are reported. These autoantibodies are not specific for RS. About autoantibodies against GluR3, significance and stimulating effects to GluR3 are controversial. Autoantibodies against GluRepsilon2 were detected in all patients within six months from epilepsy onset, and in some patients at chronic stage. These data suggest that autoantibodies against GluRepsilon2 may be involved in the pathological mechanisms in the early stage, but we could not confirm the effect of the autoantibodies from RS patients on excitatory postsynaptic NMDA current using patch clump methods. However, anti-double-stranded DNA antibodies in patients with SLE are reported to cross-react with n-terminal of GluRepsilon2, and cause neuronal apoptosis in rat hippocampus, ensuing memory impairment, and emotional behavior impairment in mice. Therefore, autoantibodies against GluRepsilon2 may contribute to the cognitive and behavioral changes in RS. Concerning about cellular immunity in RS, lymphocytes stimulating tests revealed peripheral lymphocytes sensitized by antigens containing GluRepsilon2. Cytotoxic T cells (CTLs) excreting Granzyme B were reported in resected brain tissue, and we confirmed the elevated levels of Granzyme B, not in sera, but in CSF. These data suggest that CTLs activated by infection invade into CNS, and recognize neural antigens, and excrete Granzyme B. The incidence of NHALE is 4.1/1 million/year in Japanese adults. Our study in 91 adult patients with NHALE revealed the following characteristics. Mean onset age was 35.2 +/- 16.9 years old, and preceding infections existed in 68.7% of patients, and predominant symptoms at the onset were psychiatric symptoms (33.3%) and convulsions (25.0%). CSF showed slightly elevated cell counts (55.5 +/- 139.9), protein levels (48.1 +/- 36.0 mg/dl), and IgG levels (4.5 +/- 3.9 mg/dl). MRI lesions with high intensity were found in 40.8% (DWI) and 54.2% (FLAIR) of patients in various stages after onsets. Autoantibodies against GluRepsilon2 in sera were detected in approximately 60% of NHALE patients from acute to chronic stages, and the autoantibodies in CSF were detected in 51.8% (acute stage), 41.4% (recovery stage), 28.6% (chronic stage) of patients and included epitopes to n-terminal of GluRepsilon2 (NT1). These data suggest that autoantibodies against GluRepsilon2 produced in sera after infection infiltrate into CNS through damaged BBB in acute stages, and affect n-terminal of GluRepsilon2. In chronic stage, recovery of function of BBB reduces levels of the autoantibodies in CSF. Because BBB in hippocampi and amygdala are vulnerable, autoantibodies against GluRepsilon2 including epitopes to n-terminal may contribute to the limbic symptoms around onset. Among several autoantibodies related with NHALE, autoantibodies against GluRepsilon2 were found in patients around 15-34 years old, autoantibodies against VGKC were around 50.4 years old, autoantibodies against NAE were around 59 years old, autoantibodies against Hu were around 61.5 years old. These data suggest that autoantibodies related with NHALE have age-dependent heterogeneity.     

Extracellular domains of the beta, gamma and epsilon subunits of the human acetylcholine receptor as immunoadsorbents for myasthenic autoantibodies: a combination of immunoadsorbents results in increased efficiency

Myasthenia gravis (MG) is usually caused by autoantibodies against the human muscle acetylcholine receptor (AChR). Plasmapheresis offers a therapeutic option, but, as well as removing the pathogenic anti-AChR autoantibodies, it non-specifically removes indispensable immunoglobulins. An attractive alternative to plasmapheresis would be the extracorporeal specific removal of the autoantibodies using AChR-based immunoadsorbents. Previously, we used the N-terminal extracellular domain (ECD) of the AChR alpha subunit to immunoadsorb anti-alpha subunit autoantibodies from MG sera. In this study, we immobilised the beta -, gamma- and epsilon-AChR ECDs on Sepharose and tested them as immunoadsorbents on 50 MG sera. A given ECD removed a different percentage of autoantibodies from different sera and different ECDs removed different percentages from the same serum; on average, the beta-, gamma- and epsilon-ECDs removed 22%, 20% and 15.5% of the autoantibodies, respectively. Immunoadsorption was completed in 3 min, 1 mug of ECD removed approximately 2 pmol of autoantibodies, and the immunoadsorbent could be recycled approximately 4 times. The combined use of two (alpha+gamma) or four (alpha+beta+gamma+epsilon) ECDs in a single immunoadsorbent resulted in much higher (often additive) immunoadsorption. These results show that MG sera have autoantibodies against several AChR subunits, and suggest that the combined use of all AChR ECDs could provide the basis for a novel, antigen-specific therapy for MG.     

Serum antibodies to GM1 and GM3-gangliosides in systemic lupus erythematosus with chronic inflammatory demyelinating polyradiculoneuropathy

Acute symmetric demyelinating polyneuropathy of the Guillain-Barré type is known in systemic lupus erythematosus (SLE). Chronic idiopathic demyelinating polyneuropathy (CIDP) has been reported rarely with SLE. A case is reported of CIDP accompaning SLE with autoantibodies against GM1-and GM3-gangliosides. There was no historical evidence to suggest SLE, and CIDP was the first manifestation of SLE. The 38-year-old patient had elevated CSF-protein, slow nerve conduction velocities, sural nerve biopsy revealed mixed axon loss with demyelination and CIDP white matter lesions were observed in magnetic resonance imaging. the GM1-and GM3-autoantibodies may play a role in the pathogenesis of CIDP in SLE.     

Clinical and magnetic resonance imaging features of multiple sclerosis with autoreactive antibodies in Ishikawa Prefecture, Japan

Previous reports of multiple sclerosis (MS) with autoantibodies might include neuromyelitis optica (NMO). We investigated the frequency of autoreactive antibodies (AR) in both MS and NMO. Systemic lupus erythematosus (SLE)-associated autoantibodies such as anti-Sm antibodies, anti-single stranded DNA antibodies and lupus anticoagulants were only identified in MS, whereas SLE itself is more commonly associated with NMO. Moreover, when magnetic resonance imaging features between autoreactive antibody-positive (AR(+))MS and -negative (AR(-))MS were compared, AR(+)MS cases showed significantly fewer than 3 periventricular lesions compared to AR(-)MS cases. These results may indicate different pathogenetic mechanisms underlying AR(+)MS and AR(-)MS.     

Endothelial prostacyclin release in systemic lupus erythematosus

The ability of sera from patients with SLE to stimulate endothelial cell prostacyclin production was studied using a standardized assay system for testing the effects of serum on cultured human endothelial cell monolayers. The effects of 20 normal and 32 SLE sera on endothelial prostacyclin production were measured. No differences between the rates of prostacyclin production were seen between the two groups, either basally or when prostacyclin release was stimulated with thrombin or bradykinin. In the SLE samples there was no correlation between anticardiolipin IgG or IgM titres and their ability to stimulate basal or agonist-induced prostacyclin release. These results suggest that the elevated risk of thrombosis in SLE patients is not associated with inhibition of endothelial cell prostacyclin synthesis.     

Plasmacytoid dendritic cells and interferon-alpha in Aicardi-Goutières syndrome

In Aicardi-Goutières syndrome (AGS), as in systemic lupus erythematosus (SLE) and Sj?gren's syndrome, an increased level of interferon alpha (IFN-alpha) is involved in the pathogenesis of the disease. In SLE and Sj?gren's syndrome, cytokine production originates in plasmacytoid dendritic cells (pDCs) under the influence of immune complexes formed by DNA and RNA from improperly removed apoptotic or necrotic cells, together with IgG autoantibodies. We studied the role of soluble factors in the serum or cerebrospinal fluid (CSF) of AGS patients and their capacity to stimulate pDCs to produce IFN-alpha. Our findings show that, in contrast to SLE, there is no decrease in the number of circulating pDCs in AGS patients. Secondly, unlike the autoantibodies in the serum of patients with SLE or Sj?gren's syndrome, there is no increased frequency of antinuclear antibodies (ANA) or other soluble factors inducing IFN-alpha from pDCs. These data indicate that the origin of IFN-alpha in AGS is different from that in the autoimmune diseases tested.