Immunological Memory and Chronic Autoimmune Diseases


Though most autoimmune diseases become chronic, spontaneous remission occurs in a small, but considerable proportion of patients suffering from various autoimmune diseases [18, 19]. Complete remission is common among autoimmune diseases that had been induced by an exogenous stimulus in individuals bearing an otherwise intact immune system. This applies for most animal models, where autoreactivity is induced in animals that otherwise would not develop that autoimmune disorder such as experimental autoimmune encephalomyelitis (EAE) [20]. Similarly, drug-induced human SLE usually undergoes remission in patients no longer receiving this drug [21]. These observations suggest that the factors that induce acute autoreactivity and autoimmune tissue destruction are not necessarily sufficient to make immunopathology and inflammation chronic.

Clear evidence that induction and chronicity of an autoimmune disease could be driven by distinct mechanisms is provided by genetic approaches in animal models [5]. Genetic dissection shows that the gene loci involved in the onset of pristane-induced arthritis are different from those involved in the chronic phase of that disease [17, 22]. It seems likely that following the initial loss of immunological self-tolerance, additional factors are required to maintain inflammation. Such considerations are well in accordance with clinical observations and experimental results obtained in animal models. Some of these observations even suggest that those factors that had induced the disease initially are of minor importance at later stages of the disease. For example, in the K/BxN mouse model, the presence of both T cells and B cells is required to induce pathogenic autoantibody production and induction of arthritis [14,23]. Accordingly, depletion of T cells in vivo by therapeutic antibodies efficiently prevents the production of autoantibodies and the onset of inflammation, when applied early. However, when T cells are depleted once the disease has started, it inhibits neither the production of autoantibodies nor the progression of arthritis. When transferred into immunodeficient recipients, antibodies from arthritic K/BxN mice are sufficient to induce and maintain inflammation, in the complete absence of T cells [14]. Thus, in this model, T cells are of prime importance for the initial generation of au-toreactive antibodies, but they are of minor importance for the maintenance of inflammation. Despite their potential to induce autoreactivity, therapeutic targeting of T cells does not suffice to cure the established disease. For several human autoimmune diseases, similar principles are likely to apply. There is good evidence for a critical participation of T cells in the pathogenesis of RA, SLE, and many other autoimmune diseases [24-26]. Nevertheless, disease activity in RA patients usually remains unaltered upon T cell depletion [27]. Instead, recent clinical trials show that the therapeutic depletion of B cells expressing CD20 with rituximab results in remarkable clinical improvement in many patients with RA [28,29] or SLE [30]. This is in agreement with the speculation that T cells are involved in the initiation of rheumatic inflammation, but B cells are more relevant for maintenance of inflammation.

In understanding autoimmune inflammation, most attention so far has been directed to the mechanisms of breaking tolerance and developing tissue-specific autoreactivity, i.e., the initiation of inflammation. Predominantly, it is anticipated that inflammation is maintained by chronic autoimmune reactions. This view, already challenged by the apparent lack of efficacy of immunosuppressive therapies in many chronic inflammatory situations, has been challenged further by the recent development of new concepts of immunological memory. These concepts introduce the "memory plasma cell" as a new cellular entity [31, 32] and reveal a considerable epigenetic and transcriptional imprinting of distinct pro- and anti-inflammatory effector functions of memory T lymphocytes, including cytokine expression and expression of particular receptors for chemokines and adhesion molecules [3335]. In addition, it becomes clear that memory T and B lymphocytes, unlike their naïve precursors, can be reactivated independent of antigen, by proinflammatory cytokines and pathogen-associated molecular patterns [36,37]. In addition to these qualitative differences, quantitative differences between naïve and memory lymphocytes may contribute to the chronic course of most autoimmune diseases. These include the presence of increased numbers of

B and T lymphocytes specific for the recall antigen, higher affinities of the antibodies produced, a reduced threshold for activation of responding memory cells compared to naïve cells and an accelerated production of effector cytokines.

B Cell Memory

B cells that express autoantibodies with hypermutated antigen-binding sites are frequently found in patients suffering from RA, SLE, and other autoimmune diseases in which B cells or autoantibodies are likely to play key roles in the pathogenesis [38-40]. Memory B cells accumulate within inflamed salivary glands and within joints of patients suffering from Sjogren's syndrome (SS) or RA, respectively. This is due to immigration of memory B cells into inflamed tissues and their local clonal expansion there [41-43]. It had been assumed that this expansion might be autoantigen-driven and that the responding autoreactive cells might be directly involved in local pathogenesis [43]. Interestingly enough, in individuals suffering from autoimmune diseases, somatic mutation of the genes coding for the antibody binding sites can take place in lymphoid tissues outside of classical germinal centers [44-46] (Table 1). Such mutations also can occur in germinal center-like structures of inflamed synovial tissue of RA patients, and in B cells of the marginal sinus-bridging channels of MRL/lpr mice [47]. In MRL/lpr mice, the mutation rates of autoreactive B cells found in extrafollicular sites are similar to those of B cells of germinal centers, activated by an alloantigen [48].

The generation of an autoreactive B cell memory could contribute in several ways to autoimmune inflammation. Memory B cells are increased in numbers and show a reduced threshold for reactivation. Plasma cells derived from memory B cells will secrete autoantibodies of increased affinities and switched isotypes, which may switch them from harmless to pathogenic. Probably most importantly, memory B cells can be reactivated independent of antigen-receptor signaling, by pathogen-associated molecular patterns and cytokines. Such signals will drive memory B cells into differentiation to antibody-secreting cells [36]. Reactivation of memory B cells can occur also independent of T cell help [49]. Thus even in a situation when T cell tolerance is (re-)established, autoreactive memory B cells, when reactivated by pathogenic insult, may start an inflammatory autoreactive flare, or entertain chronic autoimmune inflammation, by differentiating into potent autoantigen-presenting cells [50] and into cells secreting potentially pathogenic autoantibodies.

Table 1 B cell and humoral memory origin and persistence

Protective immunity

Formation of memory B cells in germinal centers within secondary lymphoid tissues [44,45] Persistence of long-lived plasma cells providing humoral antibody-memory mainly in thebone marrow [61, 62], and to a lower extent in spleen [64]

Autoimmune diseases

Spontaneous formation of splenic germinal centers in murine models [66, 94]

Persistence of long-lived plasma cells providing humoral autoreactive antibody memory in the spleen [65] andmost likely the inflamed kidneys of NZB/W mice [66] Somatic hypermutation of autoreactive B cells (indicating memory B cell formation) in extrafollicular areas within secondary lymphoid tissues demonstrated in Fas-deficient MRL/lpr mice [47] Ectopic germinal center-like structures (indicating memory B cell formation) within chronically inflamed joints of RA patients [95-97]

T Cell Memory

Like memory B cells, memory T cells, compared to naive T cells, have a reduced threshold for reactivation by antigen and can also be reactivated independent of antigen, e.g., by ligands for TLR2 [37] and cytokines, such as type I interferons [51]. Autoreactive T cells have been found in patients suffering from various autoimmune diseases. Upon restimulation in vitro, many of these cells express effector cytokines and show the phenotype of memory cells [52,53]. It is controversial whether the frequencies of autoreactive T cells are higher in autoimmune patients than in normal controls, with some reports finding no differences [54, 55], while others do [56]. In an attempt to determine the frequencies of autoreactive T cells with high functional avidity to myelin in patients and controls, Bilekova and colleagues have used low amounts of the autoantigen myelin for the ex vivo restimulation of T cells from patients with multiple sclerosis (MS) and found the frequencies of autoreactive T cells about four times as high in patients compared to controls [57]. This increase, though small, is in accordance with the idea that an autoreactive T cell memory is formed in MS patients. Many autoreactive T cells isolated from patients also show a cytokine memory [33], e.g., for IFN-y [58].

The contribution of autoreactive T cell memory to the pathology of murine models for autoimmune diseases is poorly understood. T cells are considered to be of prime relevance for the development of MS and its murine model EAE. However, the presence of autoreactive memory T cells might not be required for the chronic phase of the disease [59], but rather for the development of chronicity [60]. A single immunization with self-antigen leads to acute EAE inflammation, followed by stable remission. Induction of autoreactive immunological memory to neuronal antigens leads to a chronic relapsing-remitting form of EAE, more closely resembling MS [60]. While the detailed mechanisms behind these observations are not completely understood, they suggest that the formation of autoreactive immunological memory might be a prerequisite for chronicity in EAE. Whether or not T cell memory is crucial in this process remains speculative. Autoreactive T cells have a great potential to contribute to autoreactive pathogenicity and definitively can induce autoimmune diseases in animal models, but experiments and clinical observations demonstrating a crucial role for T cell memory in causing chronicity of autoimmune diseases are still lacking.

Humoral Memory

Immunological memory is characterized by increased numbers of clonally expanded B and T cells specific for an antigen previously encountered, along with an increase in the affinity of the antibody response. Long-lived plasma cells provide humoral memory in terms of specific antibodies [61-64]. These plasma cells have alifespan comparable to that of memory B cells and maintain antibody secretion for months and years, without need for restimulation. They are a major source of long-term persisting serum antibody titers in protective immunity. Recently, it has been demonstrated that long-lived plasma cells also contribute to the production of autoantibodies in lupus-prone NZB/W mice [65]. In these mice, total plasma cell numbers are significantly increased in the spleen, as compared to healthy mice [66]. The splenic plasma cell population of NZB/W mice consists of about 60% short-lived cells, and about 40% nondividing, long-lived plasma cells with half-lives of several months, at least. Cells secreting autoantibodies binding to DNA are present in the shortlived and in the long-lived plasma cell population [65]. Treatment with the immunosuppressive drug cyclophosphamide readily depletes the short-lived, but not the long-lived plasma cells in NZB/W mice. Whether autoreactive long-lived plasma cells are also found in other tissues of NZB/W mice remains to be established. Chronically inflamed tissues are potential candidates to harbor long-lived plasma cells. Plasma cells accumulate at these sites in large numbers, and many factors supporting plasma cell survival belong to the family of inflammatory cytokines [67]. Interestingly, the bone marrow plasma cell compartment that contains the largest numbers of long-lived plasma cells in nonautoimmune individuals is not increased in NZB/W mice.

In humans, the lifetime of individual plasma cells cannot be measured directly, for ethical reasons. Indirect evidence suggests, however, that also in humans long-lived plasma cells contribute to the pathogenesis of autoimmune diseases. Long-lived plasma cells will produce antibodies in the absence of an acute reactive immune flare, i.e., during disease remission, and will survive conventional immunosuppressive therapy [32]. This is exactly the observation for many patients with autoimmune diseases, which maintain expression of autoantibodies during clinical quiescent phases [32, 68-70] and despite immunosuppressive or B cell- or T cell-depleting therapy [28, 71-73]. The autoantibodies produced in the remitting phases, although obviously not leading to acute clinical symptoms, may well contribute essentially to the maintenance of chronic inflammation and set the stage for the next relapse.

Autoantibody-Mediated Pathological Mechanisms

In mice, and probably also in humans, long-lived plasma cells can maintain a relevant level of autoantibody titers during clinically quiescent states, even in face of immunosuppressive therapy. Though some autoantibodies may have no pathogenetic potential, others fuel chronic inflammation and accelerate relapses of acute pathology by a variety of mechanisms (Table 2). Autoan-tibodies directed to targets on cell surfaces can damage or destroy these cells by classical immune effector mechanisms, such as complement lysis or antibody-dependent cell-mediated immunity (ADCC), e.g., autoantibodies binding to erythrocytes and thrombocytes can cause autoimmune hemolytic anemia and thrombocytopenia, respectively. Autoantibodies reacting with cell surface receptors can modify cell activity by modulation, blockage, or stimulation [74]. Autoantibody-induced cross-linking of cell surface receptors may result in aggregation and redistribution of the receptors and their internalization, mimicking the respective signals or making the cell refractory to them. By this mechanism, anti-acetylcholine receptor antibodies impair neuromuscular function in myasthenia gravis [75]. Type 1 anti-intrinsic factor antibodies block the cobalamin binding site of the gastric protein intrinsic factor, required for the uptake of cobalamin, causing pernicious anemia [76]. In Graves' disease, thyroid-stimulating autoantibodies mimic the action of the thyroid-stimulating hormone. The resulting stimulation of thyroid cells leads to hyperthyroidism and overproduction of thyroid hormone [77].

Table 2 Antibody mediated pathological mechanisms





Autoimmune anemia,


to cell surface


and matrix antigens

Blockade or stimulation

Myasthenia gravis,

[74, 76, 77]

of proteins/receptors

pernicious anemia;

Graves' disease

Immune complex-mediated

Nephritis, vasculitis, etc.


Antibodies to intracellular autoantigens can also be pathogenetic. Several studies have demonstrated binding of such autoantibodies to the outside of the cell membrane. The reason for this extracellular binding is less clear. Membrane proteins may show cross-reactive epitopes to intracellular proteins, and/or under certain conditions intracellular antigens maybe exposed at the cell surface, e.g., in cells undergoing apoptosis [78]. It has also been reported that autoantibodies specific for intracellular antigens penetrate into living cells [79-81], although this idea is controversially discussed [81].

DNA-anti-DNA antibody complexes directly induce production of type I interferon by plasmacytoid dendritic cells [82, 83]. Deposits of immune complexes can mediate vasculitis, cryoglobulinemia, nephritis, and other syndromes [84-87]. The significant contribution of secreted autoantibodies to pathogenesis of autoimmune diseases and the resistance of long-lived plasma cells to conventional immunosuppression underlines the necessity of developing novel therapeutic approaches for the elimination of long-lived plasma cells.

0 0

Post a comment