Nature, nurture, and microbes: The development of multiple sclerosis.

This paper argues that multiple sclerosis (MS) is the result of an autoimmune attack against components of the central nervous system (CNS). The effector cells involved in the pathogenic process are CNS-autoreactive T cells present in the healthy immune system in a resting state. Upon activation, these cells cross the blood-brain barrier and attack the CNS target tissue. Recent evidence indicates that autoimmune activation may happen in the intestine, following an interaction of bacterial components of the gut flora with local CNS autoreactive T cells. The consequences of this concept are discussed.

[Multiple sclerosis and microbiota. From genome to metagenome?]. / Multiple Sklerose und Mikrobiota. Vom Genom zum Metagenom?

The individual risk of contracting multiple sclerosis (MS) is determined by genetic predisposition as well as environmental factors. In monozygotic twins the concordance rate for MS is approximately 30 % indicating that environmental factors are even more important than genetic factors. Observations in a T-cell receptor-transgenic, spontaneous mouse model strongly point to an important contribution of the individual gut microbiome (microbiota). Mice maintained in a germ-free environment are completely protected from experimental autoimmune encephalomyelitis (EAE) in this model, whereas mice that are kept under normal conditions spontaneously develop a relapsing-remitting central nervous system (CNS) disease which is astoundingly similar to human MS. It appears that the autoimmune reaction against CNS tissue is "remotely controlled" by the gut microbiota. This may be explained by the facts that the microbiota influences the gut-associated lymphoid tissue (GALT) and, vice versa, the GALT regulates systemic immunity. The precise role of the microbiota in MS remains to be clarified. New methods of DNA sequencing and bioinformatics allow the analysis of very complex bacterial metagenomes. If individual microbial risk profiles can be identified this would provide completely new perspectives for the prophylaxis and therapy of MS.

Genetic models for CNS inflammation.

The use of transgenic technology to over-express or prevent expression of genes encoding molecules related to inflammation has allowed direct examination of their role in experimental disease. This article reviews transgenic and knockout models of CNS demyelinating disease, focusing primarily on the autoimmune disease multiple sclerosis, as well as conditions in which an inflammatory response makes a secondary contribution to tissue injury or repair, such as neurodegeneration, ischemia and trauma.

Gene transfer through the blood-nerve barrier: NGF-engineered neuritogenic T lymphocytes attenuate experimental autoimmune neuritis.

Nerve-specific autoimmune T lymphocytes were used as vehicles to deliver therapeutically useful neurotrophic factors across the endothelial blood-nerve barrier. P2 protein-reactive T-lymphocyte lines from Lewis rats were transduced with a recombinant retrovirus containing the mouse nerve growth factor (NGF) gene. The engineered T cells released high amounts of NGF dependent on antigenic stimulation in vitro. After intravenous injection, the T cells infiltrated the rat peripheral nervous system and persisted there for at least two weeks. Local release of NGF from engineered T cells was demonstrable by immunocytochemistry and by an anti-inflammatory effect on infiltrating macrophages.

Immunological T-cell memory in the in vitro-induced experimental autoimmune orchitis: specificity of the reaction and tissue distribution of the autoantigens.

Immunological memory has been induced in vitro against testicular autoantigens by priming normal rat T lymphocytes against autologous testis cells, and by permitting the isolated blast cells to revert back to small secondary lymphocytes (secondary EAO cells) in the absence of the priming antigen. The secondary EAO cells vigorously respond in a secondary response when reconfronted with syngeneic testis or lymphoid cells. Their responsiveness to nonself stimulator cells is, however, reduced. Secondary cells derived from concanavalin A-stimulated blasts, do not show that pattern of specificity. The specificity of the secondary EAO cells is definite, and cannot be affected by further culture on allogeneic fibroblasts, which are antigenic for unprimed T lymphocytes. At least part of the autoantigens are determined by the major histocompatibility gene complex (MHC). Factors provided by the culture system do not appear to determine the specificity of this reaction. Only minor cell populations can restimulate secondary EAO cells. One of these populations is presumably phage-like cells within the lymphoid populations can elicit a secondary EAO response. Thus, the autoantigens relevant in the secondary EAO response are either MHC antigens restricted to these testicular and lymphoid subpopulations, or MHC antigens recognized in conjunction with organ-specific non-MHC determinants.

Regulation of autosensitization. The immune activation and specific inhibition of self-recognizing thymus-derived lymphocytes.

We studied the mechanisms underlying the natural tolerance of thymus-derived (T) lymphocytes for self-antigens. Lymphocytes from the thymus or lymph nodes of inbred rats were autosensitized in vitro against monolayers of autochthonous thymus reticulum cells or syngeneic fibroblasts. Receptors for self-antigens were detected by the specific adherence of normal lymphocytes to syngeneic cells. The achievement of active cell-mediated autosensitization was assayed by measuring the immunospecific lysis of syngeneic target cells in vitro, or graft-versus-host (GvH) reactions in vivo. The following observations were made using these systems. (a) A fraction of normal lymphocytes was found to have specific surface receptors that are able to recognize self-antigens which seem to be accessible in vivo. These potentially self-reactive lymphocytes were activated by incubation with syngeneic or autochthonous cells in vitro. Hence, the elimination of potentially self-reactive lymphocytes cannot be the only basis for natural self-tolerance. Therefore, the maintenance of self-tolerance in vivo appears to involve suppression of the immune reactivity of such self-tolerant lymphocytes. (b) We found that control of autosensitization depends upon the inhibition of the recognition of self-antigens. A GvH reaction in vivo could not be suppressed once recognition of self-antigens had occurred in vitro. Moreover, studies of the kinetics of antigen recognition indicated that several hours of incubation in vitro were needed for the inactivation of factors specifically inhibiting self-recognition. (c) We found that factors which inhibit self-recognition are present in fresh autologous serum. Treatment of the lymphocytes, but not syngeneic adsorbing cells, with autologous serum prevented recognition of syngeneic antigens. Allogeneic serum did not prevent self-recognition, and autologous serum did not inhibit the recognition of foreign antigens. These findings indicate that natural tolerance of T lymphocytes to self-antigens can be regulated by serum factors which act on the lymphocytes. The immunospecificity of the inhibitory effect suggests that these factors may be soluble self-antigens in a tolerogenic form.

Thymic nurse cells. Lymphoepithelial cell complexes in murine thymuses: morphological and serological characterization.

We describe a new cellular component of normal mouse thymuses, which is isolated by fractionated trypsin dissociation of minced thymus tissue followed by repeated unit gravity sedimentation. These cells are of unusually large size, with diameters of 30 mum and more. They represent cellular complexes of single large cells filled with high numbers of lymphoid cells. The majority of the engulfed lymphoid cells is not only fully intact, as judged by morphological criteria, but, moreover, includes a high proportion of mitotic figures. Electron microscopic investigations reveal the epithelial character of the large thymic nurse cells (TNC). The peripherally situated cytoplasmic tonofilament streams, and characteristic vacuoles filled with coarse, unidentified material, closely resemble cytoplasmic organelles found in the cortical reticuloepithelial cells described in situ. The internalized lymphocytes are located within caveolae lined by plasma membranes. These TNC caveolae are completely sequestered, and have lost any communication with the extracellular space, as demonstrated by the inability of an electrondense marker, cationized ferritin, to diffuse into the perilymphocytic clefts. The structural interactions between the membranes of the engulfed thymocytes with the surrounding TNC caveolar membranes were investigated both in ultrathin sections and in freeze-etch preparates. Two distinct contact types between both membranes were discerned: (a) complete, close contact along the entire lymphocyte circumference, and (b) more frequently, contact restricted to discrete, localized areas. Judging from their size and distribution, the localized contacts could correspond particle aggregates of freeze-etch preparates, which morphologically resemble certain stages of gap junction. Furthermore, we regularly found square arrays of particles of uniform size, which so far have been thought to be typical for cell membranes actively engaged in ion exchange. Tight junction-like particle arrays, which were present on TNC outer membranes, and probably represented disrupted contacts between adjacent TNC in the intact tissue, could not be found on caveolar or lymphocyte membranes. Finally, one of the most conspicuous specializations of the TNC caveolar membrane were membrane invaginations, which were arranged mainly in groups, and which probably reflect endo- or exocytotoxic events. We investigated the surface antigen phenotype of TNC by indirect immunofluorescence, with monoclonal antibodies against determinants of H-2- complex subregions as well as against lymphocyte differentiation markers. Semiquantification was reached with flow cytofluorimetry, followed by morphological control by fluorescence microscopy. The surface antigen formula of TNC is: Ig(-), Thy-l(-), H-2K(++), I-A (++), I-E/C(+), H-D(++), Ly-1(-), Ly-2(-), Qat-4(-), Qat-5(-), and peanut agglutinin (PNA)(-). Thymic macrophages, which were identified by double fluorescence, with rhodamine- coupled zymosan as a phagocytosis marker, were serologically identical with TNC. Free thymocytes, in contrast, had the following antigen formula: Ig(-), Thy-1(++), H-2K(+/-), I-A(-), I-E/C(-), H-2D(+/-), Ly-1(+/-), Ly-2(+), Qat- 4(-), Qat-5(-), and PNA(+). The unprecedented finding of high numbers of dividing thymocytes sojourning within thymic epithelial cells, and the particular specializations of the TNC caveolar membranes surrounding these engulfed thymocytes is the basis of a hypothesis that postulates that an intraepithelial differentiation cycle is one essential step in, intrathymic T lymphocyte generation.

Borna disease, a progressive meningoencephalomyelitis as a model for CD4+ T cell-mediated immunopathology in the brain.

A homogeneous T cell line NM1 with Borna disease (BD) virus reactivity could be established. The NM1 cells have been characterized as CD4+ T cells. Adoptive transfer revealed that this MHC class II-restricted immune cell is responsible for the immunopathological effect leading to BD, a progressive meningoencephalomyelitis.

Interferon gamma gene expression in sensory neurons: evidence for autocrine gene regulation.

We explored expression and possible function of interferon-gamma (IFN-gamma) in cultured fetal (E15) rat dorsal root ganglion neurons combining whole cell patch-clamp electrophysiology with single cell reverse transcriptase polymerase chain reaction and confocal laser immunocytochemistry. Morphologically, we located IFN-gamma protein in the cytoplasm of the neurons in culture as well as in situ during peri- and postnatal development. Transcripts for classic IFN-gamma and for its receptor were determined in probes of cytoplasm sampled from individual cultured neurons, which had been identified by patch clamp electrophysiology. In addition, the cultured neurons expressed both chains of the IFN-gamma receptor. Locally produced IFN-gamma acts back on its cellular source. Phosphorylation and nuclear translocation of the IFN-inducible transcriptional factor STAT1 as well as IFN-gamma-dependent expression of major histocompatibility complex class I molecules on the neuronal membrane were noted in untreated cultures. However, both processes were substantially blocked in the presence of antibodies neutralizing IFN-gamma. Our findings indicate a role of IFN-gamma in autocrine regulation of sensory neurons.

Major histocompatibility complex (MHC) class I gene expression in single neurons of the central nervous system: differential regulation by interferon (IFN)-gamma and tumor necrosis factor (TNF)-alpha.

This study examined the effect of the pro-inflammatory cytokines interferon-gamma (IFN-gamma) and tumor necrosis factor-alpha (TNF-alpha) on the induction of MHC class I-related genes in functionally mature brain neurons derived from cultures of dissociated rat hippocampal tissue. Patch clamp electrophysiology combined with single cell RT-PCR demonstrated that approximately 50% of the untreated neurons contained mRNA for MHC class I heavy chains, while, with few exceptions, the cells failed to transcribe beta2-microglobulin and TAP1/TAP2 gene transcripts. No constitutive expression of MHC class I protein was detectable by confocal laser microscopy on the surface of neurons. All neurons transcribed the alpha-chain of the interferon-type II receptor (binding IFN-gamma) along with the p55 receptor for TNF-alpha. Sustained exposure to IFN-gamma resulted in transcription of beta2-microglobulin and TAP1/TAP2 genes and MHC class I surface expression in a minor part of the neurons, but did not alter their electrophysiological activities as assessed by whole cell electrophysiology. Suppression of neuronal electric activity by the sodium channel blocker tetrodotoxin drastically increased to almost 100% IFN-gamma-mediated induction of MHC class I chains, of both TAP transporters, and of membrane expression of MHC class I protein. The effect of tetrodotoxin is at least partly reverted by the neurotransmitter glutamate. In contrast to IFN-gamma, treatment with TNF-alpha did neither upregulate TAP1/TAP2 nor beta2-microglobulin gene expression, but induced MHC class I heavy chain gene transcription in all neurons. Consequently, no MHC class I molecules were detectable on the membranes of TNF-alpha-treated neurons.

B lymphocytes producing demyelinating autoantibodies: development and function in gene-targeted transgenic mice.

We studied the cellular basis of self tolerance of B cells specific for brain autoantigens using transgenic mice engineered to produce high titers of autoantibodies against the myelin oligodendrocyte glycoprotein (MOG), a surface component of central nervous system myelin. We generated "knock-in" mice by replacing the germline JH locus with the rearranged immunoglobulin (Ig) H chain variable (V) gene of a pathogenic MOG-specific monoclonal antibody. In the transgenic mice, conventional B cells reach normal numbers in bone marrow and periphery and express exclusively transgenic H chains, resulting in high titers of MOG-specific serum Igs. Additionally, about one third of transgenic B cells bind MOG, thus demonstrating the absence of active tolerization. Furthermore, peritoneal B-1 lymphocytes are strongly depleted. Upon immunization with MOG, the mature transgenic B cell population undergoes normal differentiation to plasma cells secreting MOG-specific IgG antibodies, during which both Ig isotype switching and somatic mutation occur. In naive transgenic mice, the presence of this substantial autoreactive B cell population is benign, and the mice fail to develop either spontaneous neurological disease or pathological evidence of demyelination. However, the presence of the transgene both accelerates and exacerbates experimental autoimmune encephalitis, irrespective of the identity of the initial autoimmune insult.

T cell receptor repertoire in polymyositis: clonal expansion of autoaggressive CD8+ T cells.

In polymyositis (PM), CD8+ T cell receptor (TCR) alpha/beta + cells invade and destroy major histocompatibility complex class I-positive muscle fibers. We combined polymerase chain reaction (PCR) and double-fluorescence immunocytochemistry to analyze the T cell receptor (TCR) repertoire expressed in muscle of PM patients. In patient 1, inverse PCR revealed a preferential usage of TCR V alpha 33.1, V beta 13.1, and V beta 5.1. Six of six TCR V alpha 33.1+ clones and five of seven V beta 13.1+ clones had identical nucleotide sequences. In contrast, the V beta 5.1+ TCRs were more heterogeneous. Similar results were obtained with an independent PCR method using primers specific for TCR V alpha 33, V beta 13, or V beta 5. No TCR sequences could be amplified from noninflammatory control muscle. Furthermore, none of the TCR sequences found in PM muscle could be detected in blood from the same patient or from a normal control subject. Immunohistochemistry confirmed that V beta 5.1 and V beta 13.1 were overrepresented in the muscle lesions of this patient. 32% of all CD8+ T cells were V beta 13.1+, and 16% were V beta 5.1+. However, approximately 60% of the CD8+ T cells that invaded muscle fibers were V beta 13.1+, whereas 10% were V beta 5.1+. In patient 2, 50% of the T cells were V beta 5.1+, and as in patient 1, these T cells were mainly located in interstitial areas. In patient 3, > 75% of the autoinvasive T cells stained with an anti-V beta 3 mAb. Sequence analysis of 15 PCR clones amplified with a V beta 3-specific primer showed that 9 (60%) sequences were identical. The results suggest that (a) a strikingly limited TCR repertoire is expressed in PM muscle; (b) there is a dissociation between the TCR usage of autoinvasive and interstitial T cells; and (c) the autoinvasive T cells are clonally expanded.

Experimental autoimmune panencephalitis and uveoretinitis transferred to the Lewis rat by T lymphocytes specific for the S100 beta molecule, a calcium binding protein of astroglia.

The pathogenic potential of autoimmune T cell responses to nonmyelin autoantigens was investigated in the Lewis rat using the astrocyte-derived calcium binding protein S100 beta, as a model nonmyelin autoantigen. The Lewis rat mounts a vigorous RT1B1 (major histocompatibility complex class II) restricted autoimmune response to an immunodominant S100 beta epitope (amino acid residues 76-91). The adoptive transfer of S100 beta-specific T cell lines induced a severe inflammatory response in the nervous system, but only minimal neurological dysfunction in naive syngeneic recipients. The inability of S100 beta-specific T cell transfer to induce severe disease was associated with a decreased recruitment of ED1+ macrophages into the central nervous system (CNS) in comparison with that seen in severe experimental autoimmune encephalomyelitis (EAE) induced by the adoptive transfer of myelin basic protein (MBP)-specific T line cells. Moreover, unlike encephalitogenic MBP-specific T cell lines, S100 beta-specific T cell lines exhibited no cytotoxic activity in vitro. Histopathological analysis also revealed striking differences in the distribution of inflammatory lesions in MBP- and S100 beta-specific T cell-mediated disease. In contrast to the MBP paradigm, S100 beta-specific T cell transfer induces intense inflammation not only in the spinal cord, but throughout the entire CNS and also in the uvea and retina of the eye. In view of the distribution of lesions throughout the grey and white matter of the CNS we propose to term this new model experimental autoimmune panencephalomyelitis (EAP) to differentiate it from EAE. These experiments demonstrate for the first time that nonmyelin CNS autoantigens can initiate a pathogenic autoimmune T cell response, although the nature of the target autoantigen profoundly influences the clinical and histopathological characteristics of the resulting autoimmune disease. This is not simply a consequence of the distribution of the autoantigen, as both MBP and S100 beta are coexpressed in many areas of the CNS, but reflects differences in the capacity of different regions of the CNS to process and present specific autoantigens. This new model of T cell-mediated autoimmune CNS disease exhibits a number of similarities to multiple sclerosis (MS), such as its mild clinical course and the involvement of areas of the brain and eye, which are absent in myelin-mediated models of EAE. Nonmyelin autoantigens may therefore play an unexpectedly important role in the immunopathogenesis of inflammatory diseases of the CNS.

Activated human T cells, B cells, and monocytes produce brain-derived neurotrophic factor in vitro and in inflammatory brain lesions: a neuroprotective role of inflammation?

Brain-derived neurotrophic factor (BDNF) has potent effects on neuronal survival and plasticity during development and after injury. In the nervous system, neurons are considered the major cellular source of BDNF. We demonstrate here that in addition, activated human T cells, B cells, and monocytes secrete bioactive BDNF in vitro. Notably, in T helper (Th)1- and Th2-type CD4(+) T cell lines specific for myelin autoantigens such as myelin basic protein or myelin oligodendrocyte glycoprotein, BDNF production is increased upon antigen stimulation. The BDNF secreted by immune cells is bioactive, as it supports neuronal survival in vitro. Using anti-BDNF monoclonal antibody and polyclonal antiserum, BDNF immunoreactivity is demonstrable in inflammatory infiltrates in the brain of patients with acute disseminated encephalitis and multiple sclerosis. The results raise the possibility that in the nervous system, inflammatory infiltrates have a neuroprotective effect, which may limit the success of nonselective immunotherapies.

Blood-borne soluble protein antigen intensifies T cell activation in autoimmune CNS lesions and exacerbates clinical disease.

We explored the effect of i.v. soluble antigen on autoaggressive, myelin basic protein-specific effector T cells within their target organ during acute experimental autoimmune encephalomyelitis (EAE). Intravital two-photon imaging revealed that i.v. autoantigen reached the CNS and was taken up and processed by antigen-presenting cells within 30 min after injection. The exogenous autoantigen dramatically changed the motility and function of autoreactive effector T cells within the EAE lesions: T cells that had been cruising through the tissue slowed down and became tethered to local antigen-presenting cells within 1 h. One hour later, the effector T cells massively produced proinflammatory cytokines and up-regulated membranous activation markers. This strong activation of the T cells boosted CNS inflammation and aggravated clinical disease. Postactivated effector and resting memory T cells specific for a non-CNS antigen (ovalbumin) were recruited to EAE lesions and moved there without contacting antigen-presenting cells. These cells were similarly arrested and activated after i.v. infusion of ovalbumin, and they also exacerbated clinical disease. Our data are relevant for autoantigen-based therapies of autoimmune disorders. Further, the study indicates how brain unrelated antigens (microbial components) leaking into the chronically inflamed CNS through the bloodstream might trigger relapses in multiple sclerosis.

Immune function of the blood-brain barrier: incomplete presentation of protein (auto-)antigens by rat brain microvascular endothelium in vitro.

The endothelial blood-brain barrier (BBB) has a critical role in controlling lymphocyte traffic into the central nervous system (CNS), both in physiological immunosurveillance, and in its pathological aberrations. The intercellular signals that possibly could induce lymphocytes to cross the BBB include immunogenic presentation of protein (auto-)antigens by BBB endothelia to circulating T lymphocytes. This concept has raised much, though controversial, attention. We approached this problem by analyzing in vitro immunospecific interactions between clonal rat T lymphocyte lines with syngeneic, stringently purified endothelial monolayer cultures from adult brain micro-vessels. The rat brain endothelia (RBE) were established from rat brain capillaries using double collagenase digestion, density gradient fractionation and selective cytolysis of contaminating pericytes by anti-Thy 1.1 antibodies and complement. Incubation with interferon-gamma in most of the brain-derived endothelial cells induced Ia-antigens in the cytoplasm and on the cell surface in some of the cells. Before the treatment, the cells were completely Ia-negative. Pericytes were unresponsive to IFN-gamma treatment. When confronted with syngeneic T cell lines specific for protein (auto-)antigens (e.g., ovalbumin and myelin basic protein, MBP), RBE were completely unable to induce antigen-specific proliferation of syngeneic T lymphocytes irrespective of pretreatment with IFN-gamma and of cell density. RBE were inert towards the T cells, and did not suppress T cell activation induced by other "professional" antigen presenting cells (APC) such as thymus-derived dendritic cells or macrophages. IFN-gamma-treated RBE were, however, susceptible to immunospecific T cell killing. They were lysed by MBP-specific T cells in the presence of the specific antigen or Con A. Antigen dependent lysis was restricted by the appropriate (MHC) class II product. We conclude that the interaction of brain endothelial cells with encephalitogenic T lymphocytes may involve recognition of antigen in the molecular context of relevant MHC products, but that this interaction per se is insufficient to initiate the full T cell activation program.

Autosensitization of lymphocytes against thymus reticulum cells.

Lymph nodes of normal rats contain lymphocytes that can be induced in vitro to mediate a specific cellular immune reaction against reticulum cells derived from syngeneic adult thymus glands. It is likely that these lymphocytes had access to reticulum cell antigens during the period in which they developed immunocompetence within the thymus. This suggests that contact with self-antigens during development may not eliminate self-reactive lymphocytes. These findings are in conflict with the theory that natural tolerance to self-antigens depends upon elimination of lymphocyte clones.

Induction of MHC class I genes in neurons.

Whether neurons express major histocompatibility complex (MHC) class I genes has not been firmly established. The techniques of confocal laser microscopy, patch clamp electrophysiology, and reverse transcriptase-polymerase chain reaction were combined here to directly examine the inducibility of MHC class I genes in individual cultured rat hippocampal neurons. Transcription of MHC class I genes was very rare in neurons with spontaneous action potentials. In electrically silent neurons, transcription was noted, with expression of beta 2-microglobulin under tighter control than in class I heavy chain molecules. Surface expression of class I molecules occurred only in electrically silent neurons treated with interferon gamma. Immunosurveillance by cytotoxic T cells may be focused on functionally impaired neurons.

Lessons from multiple sclerosis: models, concepts, observations.

Experimental autoimmune encephalomyelitis (EAE) is often termed "the" model of human multiple sclerosis (MS). This is, however, an oversimplification. MS is a multifaceted disorder, with no single experimental model representing the entire complexity of the human disease. On the other hand, EAE comes in numerous, distinct variants, which may reflect individual aspects of MS. This presentation reviews EAE variants and their usability as models for human MS. New transgenic models representing mechanisms determining spontaneous initiation, the course of central nervous system (CNS) autoimmunity, the distribution of lesions within the CNS and the cellular composition of the inflammatory infiltrate are discussed. Aspects of the early, inflammatory phase of MS plaque generation, in particular concerning the dynamics of immune cell invasion into the CNS, are also reviewed. Finally, the usability of EAE models for discovery and validation of MS drugs is discussed.

Transplantation of thymic autoimmune microenvironment to severe combined immunodeficiency mice. A new model of myasthenia gravis.

To study the role of the thymus in the cellular pathogenesis of myasthenia gravis (MG) we transplanted thymus tissue fragments from MG thymuses beneath the kidney capsule of severe combined immunodeficiency (SCID) mice. Immunocytochemical studies documented that the human thymus tissues are accepted as long-term grafts in the host SCID mice, with human lymphocytes, thymic stroma, and thymic myoid cells demonstrable in transplanted thymus for at least 15 weeks after transplantation. Human anti-acetylcholine receptor antibodies became detectable 1 to 2 weeks after transplantation, and in most chimeras the titers increased over at least 11 weeks to reach levels typically found in severe human MG. Human Ig deposits were detected at skeletal muscle end-plates, demonstrating that the human (auto)antibodies bound to murine acetylcholine receptor. In contrast, transfers of dissociated thymus cells only lead to a transient increase of anti-acetylcholine receptor antibodies. Our data prove that myasthenia gravis thymus is able to induce and maintain autoantibody production in immunodeprived host animals, and that this tissue contains all cellular components required for autoantibody production. Transplantation of solid thymus tissue seems to transfer an autoimmune microenvironment, which will allow direct studies of the mechanism of autosensitization inside the thymus.