Since Ehrlich first postulated the existence of mechanisms to prevent the generation of self-reactivity in 1900, ideas concerning the nature of this inhibition have developed in parallel with a progressive increase in understanding of the immune system. Burnet's clonal selection theory included the idea that interaction of lymphoid cells with their specific antigens during fetal or early postnatal life would lead to elimination of such “forbidden clones.” This idea became untenable, however, when it was shown that autoimmune diseases could be induced in experimental animals by simple immunization procedures, that autoantigen-binding cells could be demonstrated easily in the circulation of normal individuals, and that self-limited autoimmune phenomena frequently developed following tissue damage from infection or trauma. These observations indicated that clones of cells capable of responding to autoantigens were present in the repertoire of antigen-reactive cells in normal adults and suggested that mechanisms in addition to clonal deletion were responsible for preventing their activation.
Currently, three general processes are thought to be involved in the maintenance of selective unresponsiveness to autoantigens (Table 318-1): (1) sequestration of self-antigens, rendering them inaccessible to the immune system; (2) specific unresponsiveness (tolerance or anergy) of relevant T or B cells; and (3) limitation of potential reactivity by regulatory mechanisms.
Derangements of these normal processes may predispose to the development of autoimmunity (Table 318-2). In general, these abnormal responses require an exogenous trigger such as bacterial or viral infection or cigarette smoking and require the presence of endogenous abnormalities in the cells of the immune system. Microbial superantigens, such as staphylococcal protein A and staphylococcal enterotoxins, are substances that can stimulate a broad range of T and B cells based upon specific interactions with selected families of immune receptors, irrespective of their antigen specificity. If autoantigen-reactive T and/or B cells express these receptors, autoimmunity might develop. Alternatively, molecular mimicry or cross-reactivity between a microbial product and a self-antigen might lead to activation of autoreactive lymphocytes. One of the best examples of autoreactivity and autoimmune disease resulting from molecular mimicry is rheumatic fever, in which antibodies to the M protein of streptococci cross-react with myosin, laminin, and other matrix proteins as well as neuronal antigens. Deposition of these autoantibodies in the heart initiates an inflammatory response, whereas penetration of these antibodies into the brain can result in Sydenham's chorea. Molecular mimicry between microbial proteins and host tissues has been reported in type 1 diabetes mellitus, rheumatoid arthritis, and multiple sclerosis. It is presumed that infectious agents may be able to overcome self-tolerance because they possess molecules, such as bacterial endotoxin, RNA, or DNA, that have adjuvant-like effects on the immune system that increase the immunogenicity of the microbial antigens. The adjuvants activate dendritic cells through pattern recognition receptors and stimulate the activation of previously quiescent lymphocytes that recognize both microbial and self antigen.
Endogenous derangements of the immune system may also contribute to the loss of immunologic tolerance to self-antigens and the development of autoimmunity (Table 318-2). Some autoantigens reside in immunologically privileged sites, such as the brain or the anterior chamber of the eye. These sites are characterized by the inability of engrafted tissue to elicit immune responses. Immunologic privilege results from a number of events, including the limited entry of proteins from those sites into lymphatics, the local production of immunosuppressive cytokines such as transforming growth factor β, and the local expression of molecules such as Fas ligand that can induce apoptosis of activated T cells. Lymphoid cells remain in a state of immunologic ignorance (neither activated nor anergized) to proteins expressed uniquely in immunologically privileged sites. If the privileged site is damaged by trauma or inflammation, or if T cells are activated elsewhere, proteins expressed at this site can become the targets of immunologic assault. Such an event may occur in multiple sclerosis and sympathetic ophthalmia, in which antigens uniquely expressed in the brain and eye, respectively, become the target of activated T cells.
Alterations in antigen presentation may also contribute to autoimmunity. Peptide determinants (epitopes) of a self antigen that are not routinely presented to lymphocytes may be recognized as a result of altered proteolytic processing of the molecule and the ensuing presentation of novel peptides (cryptic epitopes). When B cells rather than dendritic cells present self antigen, they may also present cryptic epitopes that can activate autoreactive T cells. These cryptic epitopes will not have previously been available to effect the silencing of autoreactive lymphocytes. Furthermore, once there is immunologic recognition of one protein component of a multimolecular complex, reactivity may be induced to other components of the complex following internalization and presentation of all molecules within the complex (epitope spreading). Finally, inflammation, drug exposure, or normal senescence may cause a primary chemical alteration in proteins, resulting in the generation of immune responses that cross-react with normal self-proteins. For example, the induction and/or release of protein arginine deaminase enzymes results in the conversion of arginine residues to citrullines in a variety of proteins, thereby altering their capacity to induce immune responses. Production of anticitrullinated protein antibodies has been observed in rheumatoid arthritis, chronic lung disease, as well as normal smokers and may contribute to organ pathology. Alterations in the availability and presentation of autoantigens may be important components of immunoreactivity in certain models of organ-specific autoimmune diseases. In addition, these factors may be relevant in understanding the pathogenesis of various drug-induced autoimmune conditions. However, the diversity of autoreactivity manifest in non-organ-specific systemic autoimmune diseases suggests that these conditions might result from a more general activation of the immune system rather than from an alteration in individual self-antigens.
Many autoimmune diseases are characterized by the presence of antibodies that react with apoptotic material. Defects in the clearance of apoptotic material have been shown to elicit autoimmunity and autoimmune disease in a number of animal models. Moreover, defects in the clearance of apoptotic material have been found in subjects with systemic lupus erythematosus (SLE). Apoptotic debris not quickly cleared by the immune system can function as endogenous ligands for a number of pattern recognition receptors on dendritic cells. Under such circumstances, there is activation of dendritic cells, and an immune response to apoptotic debris can develop. In addition, the presence of extracellular apoptotic material within germinal centers of secondary lymphoid organs may facilitate the direct activation of autoimmune B cell clones or function to select autoimmune B cell clones during immune responses.
A number of experimental models have suggested that intense stimulation of T lymphocytes can produce nonspecific signals that bypass the need for antigen-specific helper T cells and lead to polyclonal B cell activation with the formation of multiple autoantibodies. For example, antinuclear, antierythrocyte, and antilymphocyte antibodies are produced during the chronic graft-versus-host reaction. In addition, true autoimmune diseases, including autoimmune hemolytic anemia and immune complex–mediated glomerulonephritis, can also be induced in this manner. While it is clear that such diffuse activation of helper T cell activity can cause autoimmunity, nonspecific stimulation of B lymphocytes can also lead to the production of autoantibodies. Thus, the administration of polyclonal B cell activators, such as bacterial endotoxin, to normal mice leads to the production of a number of autoantibodies, including those directed to DNA and IgG (rheumatoid factor). Moreover, excess BAFF can also cause T cell–independent B cell activation and heavy chain class switching and the development of autoimmunity. SLE, for example, can be induced in mice through exuberant dendritic cell activation, a redundancy of TLR7 on the y chromosome (BXSByaa mice) or through exposure to CpG, a ligand for TLR 9. The ensuing induction of inflammatory mediators can cause a switch from production of nonpathogenic IgM autoantibodies to pathogenic IgG autoantibodies in the absence of antigen-specific T cell help.
Aberrant selection of the B or T cell repertoire at the time of antigen receptor expression can also predispose to autoimmunity. For example, B cell immunodeficiency caused by an absence of the B cell receptor–associated kinase, Bruton's tyrosine kinase, leads to X-linked agammaglobulinemia. This syndrome is characterized by reduced B cell activation, but also by diminished negative selection of autoreactive B cells probably caused by high levels of BAFF, resulting in increased autoreactivity within a diminished B cell repertoire. Likewise, negative selection of autoreactive T cells in the thymus requires expression of the autoimmune regulator (AIRE) gene that enables the expression of tissue-specific proteins in thymic medullary epithelial cells. Peptides from these proteins are expressed in the context of major histocompatibility complex (MHC) molecules and mediate the elimination of autoreactive T cells. The absence of AIRE gene expression leads to a failure of negative selection of autoreactive cells, autoantibody production, and severe inflammatory destruction of multiple organs. Individuals deficient in AIRE gene expression develop autoimmune polyendocrinopathy-candidiasis-ectodermal dystrophy (APECED).
Primary alterations in the activity of T and/or B cells, cytokine imbalances, or defective immunoregulatory circuits may also contribute to the emergence of autoimmunity. Diminished production of tumor necrosis factor (TNF) and interleukin (IL) 10 has been reported to be associated with the development of autoimmunity. Overproduction of type 1 interferon has also been associated with autoimmunity. Overexpression of co-stimulatory molecules on T cells similarly can lead to autoantibody production.
Autoimmunity may also result from an abnormality of immunoregulatory mechanisms. Observations made in both human autoimmune disease and animal models suggest that defects in the generation and expression of regulatory T cell activity may allow for the production of autoimmunity. It has recently been appreciated that the IPEX (immunodysregulation, polyendocrinopathy, enteropathy X-linked) syndrome results from the failure to express the FOXP3 gene, which encodes a molecule critical in the differentiation of regulatory T cells. Administration of normal regulatory T cells or factors derived from them can prevent the development of autoimmune disease in rodent models of autoimmunity. Abnormalities in the function of regulatory T cells have been noted in a number of human autoimmune diseases, although it remains uncertain whether these are causative or are secondary abnormalities owing to inflammation. Finally, recent data indicate that B cells may also exert regulatory function, largely through the production of the cytokine IL-10. Deficiency of IL-10-producing regulatory B cells can prolong the course of an animal model of multiple sclerosis.
It should be apparent that no single mechanism can explain all the varied manifestations of autoimmunity. Furthermore, genetic evaluation has shown that a number of abnormalities often need to converge to induce an autoimmune disease. Additional factors that appear to be important determinants in the induction of autoimmunity include age, sex (many autoimmune diseases are far more common in women), genetic background, exposure to infectious agents, and environmental contacts. How all of these disparate factors affect the capacity to develop self-reactivity is currently being investigated intensively.
Evidence in humans that there are susceptibility genes for autoimmunity comes from family studies and especially from studies of twins. Studies in type 1 diabetes mellitus, rheumatoid arthritis, multiple sclerosis, and SLE have shown that approximately 15–30% of pairs of monozygotic twins show disease concordance, compared with <5% of dizygotic twins. The occurrence of different autoimmune diseases within the same family has suggested that certain susceptibility genes may predispose to a variety of autoimmune diseases. Genetic mapping has begun to identify chromosomal regions that predispose to specific autoimmune diseases. It is notable that some genes are associated with multiple autoimmune diseases, whereas others are more specifically associated with only one autoimmune condition. The gene encoding PTPN22 is associated with multiple autoimmune diseases. Its product is a phosphatase expressed by a variety of hematopoietic cells that downregulates antigen receptor–mediated stimulation of T and B cells. A gain-of-function polymorphism of this gene is associated with type 1 diabetes mellitus, rheumatoid arthritis, and SLE in some populations. The explanation of the association of this polymorphism with autoimmune disease is uncertain, but it is likely that it diminishes antigen receptor signaling during lymphocyte development permitting escape of autoreactive clones or decreased activation-induced apoptosis of autoantigen-reactive lymphocytes in the periphery. In recent years, genomewide association studies have demonstrated a variety of other genes that are involved in human autoimmune diseases. Most genes individually confer a relatively low risk for autoimmune diseases and are found in normal individuals. No gene has been identified that is essential for autoimmune diseases. In addition to this evidence from humans, certain inbred mouse strains reproducibly develop specific spontaneous or experimentally induced autoimmune diseases, whereas others do not. These findings have led to an extensive search for genes that determine susceptibility to autoimmune disease.
The strongest consistent association for susceptibility to autoimmune disease has been found with particular alleles of the MHC. It has been suggested that the association of MHC genotype with autoimmune disease relates to differences in the ability of different allelic variations of MHC molecules to present autoantigenic peptides to autoreactive T cells. An alternative hypothesis involves the role of MHC alleles in shaping the T cell receptor repertoire during T cell ontogeny in the thymus. Additionally, specific MHC gene products may themselves be the source of peptides that can be recognized by T cells. Cross-reactivity between such MHC peptides and peptides derived from proteins produced by common microbes may trigger autoimmunity by molecular mimicry. However, MHC genotype alone does not determine the development of autoimmunity. Identical twins are far more likely to develop the same autoimmune disease than MHC-identical nontwin siblings, suggesting that genetic factors other than the MHC also affect disease susceptibility. Recent studies of the genetics of type 1 diabetes mellitus, SLE, rheumatoid arthritis, and multiple sclerosis in humans and mice have shown that there are several independently segregating disease susceptibility loci in addition to the MHC. Genes that encode molecules of the innate immune response are also involved in autoimmunity. In humans, inherited homozygous deficiency of the early proteins of the classic pathway of complement (C1q, C4, or C2) as well as genes involved in the type 1 interferon pathway are very strongly associated with the development of SLE.
Immunopathogenic Mechanisms in Autoimmune Diseases
The mechanisms of tissue injury in autoimmune diseases can be divided into antibody-mediated and cell-mediated processes. Representative examples are listed in Table 318-3.
Table 318-3 Mechanisms of Tissue Damage in Autoimmune Disease |Favorite Table|Download (.pdf)
Table 318-3 Mechanisms of Tissue Damage in Autoimmune Disease
Blocking or inactivation
α Chain of the nicotinic acetylcholine receptor
Phospholipid–β2-glycoprotein 1 complex
Insulin-resistant diabetes mellitus
TSH receptor (LATS)
Granulomatosis with polyangiitis (Wegener's)
α3 Chain of collagen IV
Systemic lupus erythematosus
|Opsonization||Platelet GpIIb:IIIa||Autoimmune thrombocytopenic purpura|
Rh antigens, I antigen
Autoimmune hemolytic anemia
Antibody-dependent cellular cytotoxicity
Thyroid peroxidase, thyroglobulin
Rheumatoid arthritis, multiple sclerosis, type 1 diabetes mellitus
Type 1 diabetes mellitus
The pathogenicity of autoantibodies can be mediated through several mechanisms, including opsonization of soluble factors or cells, activation of an inflammatory cascade via the complement system, and interference with the physiologic function of soluble molecules or cells.
In autoimmune thrombocytopenic purpura, opsonization of platelets targets them for elimination by phagocytes. Likewise, in autoimmune hemolytic anemia, binding of immunoglobulin to red cell membranes leads to phagocytosis and lysis of the opsonized cell. Goodpasture's syndrome, a disease characterized by lung hemorrhage and severe glomerulonephritis, represents an example of antibody binding leading to local activation of complement and neutrophil accumulation and activation. The autoantibody in this disease binds to the α3 chain of type IV collagen in the basement membrane. In SLE, activation of the complement cascade at sites of immunoglobulin deposition in renal glomeruli is considered to be a major mechanism of renal damage. Moreover, the DNA- and RNA-containing immune complexes in SLE activate TLR 9 and 7, respectively, in dendritic cells and promote a proinflammatory, immunogenic milieu conducive to amplifying the autoimmune response.
Autoantibodies can also interfere with normal physiologic functions of cells or soluble factors. Autoantibodies against hormone receptors can lead to stimulation of cells or to inhibition of cell function through interference with receptor signaling. For example, long-acting thyroid stimulators, which are autoantibodies that bind to the receptor for thyroid-stimulating hormone (TSH), are present in Graves' disease and function as agonists, causing the thyroid to respond as if there were an excess of TSH. Alternatively, antibodies to the insulin receptor can cause insulin-resistant diabetes mellitus through receptor blockade. In myasthenia gravis, autoantibodies to the acetylcholine receptor can be detected in 85–90% of patients and are responsible for muscle weakness. The exact location of the antigenic epitope, the valence and affinity of the antibody, and perhaps other characteristics determine whether activation or blockade results from antibody binding.
Antiphospholipid antibodies are associated with thromboembolic events in primary and secondary antiphospholipid syndrome and have also been associated with fetal wastage. The major antibody is directed to the phospholipid–β2-glycoprotein I complex and appears to exert a procoagulant effect. In pemphigus vulgaris, autoantibodies bind to a component of the epidermal cell desmosome, desmoglein 3, and play a role in the induction of the disease. They exert their pathologic effect by disrupting cell-cell junctions through stimulation of the production of epithelial proteases, leading to blister formation. Cytoplasmic antineutrophil cytoplasmic antibody (c-ANCA), found in granulomatosis with polyangiitis (Wegener's), is an antibody to an intracellular antigen, the 29-kDa serine protease (proteinase-3). In vitro experiments have shown that IgG anti-c-ANCA causes cellular activation and degranulation of primed neutrophils.
It is important to note that autoantibodies of a given specificity may cause disease only in genetically susceptible hosts, as has been shown in experimental models of myasthenia gravis, SLE, rheumatic fever and rheumatoid arthritis. It is also important to be aware that once organ damage is initiated, new inflammatory cascades are initiated that can sustain and amplify the autoimmune process. Finally, some autoantibodies seem to be markers for disease but have as yet no known pathogenic potential.
Manifestations of autoimmunity are found in a large number of pathologic conditions. However, their presence does not necessarily imply that the pathologic process is an autoimmune disease. A number of attempts to establish formal criteria for the diagnosis of autoimmune diseases have been made, but none is universally accepted. One set of criteria is shown in Table 318-4; however, this should be viewed merely as a guide in consideration of the problem.
Table 318-4 Human Autoimmune Disease: Presumptive Evidence for an Immunologic Pathogenesis |Favorite Table|Download (.pdf)
Table 318-4 Human Autoimmune Disease: Presumptive Evidence for an Immunologic Pathogenesis
|1. Presence of autoantibodies or evidence of cellular reactivity to self|
|2. Documentation of relevant autoantibody or lymphocytic infiltrate in the pathologic lesion|
|3. Demonstration that relevant autoantibody or T cells can cause tissue pathology|
|a. Transplacental transmission|
|b. Adaptive transfer into animals|
|c. In vitro impact on cellular function Supportive Evidence|
|1. Reasonable animal model|
|2. Beneficial effect from immunosuppressive agents|
|3. Association with other evidence of autoimmunity|
|4. No evidence of infection or other obvious cause|
To classify a disease as autoimmune, it is necessary to demonstrate that the immune response to a self-antigen causes the observed pathology. Initially, the demonstration that antibodies against the affected tissue could be detected in the serum of patients suffering from various diseases was taken as evidence that these diseases had an autoimmune basis. However, such autoantibodies are also found when tissue damage is caused by trauma or infection, and the autoantibody is secondary to tissue damage. Thus, it is necessary to show that autoimmunity is pathogenic before classifying a disease as autoimmune.
If the autoantibodies are pathogenic, it may be possible to transfer disease to experimental animals by the administration of autoantibodies, with the subsequent development of pathology in the recipient similar to that seen in the patient from whom the antibodies were obtained. This has been shown, for example, in Graves' disease. Some autoimmune diseases can be transferred from mother to fetus and are observed in the newborn babies of diseased mothers. The symptoms of the disease in the newborn usually disappear as the levels of the maternal antibody decrease. An exception, however, is congenital heart block, in which damage to the developing conducting system of the heart follows in utero transfer of anti-Ro antibody from the mother to the fetus. This can result in a permanent developmental defect in the heart.
In most situations, the critical factors that determine when the development of autoimmunity results in autoimmune disease have not been delineated. The relationship of autoimmunity to the development of autoimmune disease may relate to the fine specificity of the antibodies or T cells or their specific effector capabilities. In many circumstances, a mechanistic understanding of the pathogenic potential of autoantibodies has not been established. In some autoimmune diseases, biased production of cytokines by helper T (TH) cells may play a role in pathogenesis. In this regard, T cells can differentiate into specialized effector cells that predominantly produce interferon γ (TH1), IL-4 (TH2), IL-17 (TH17) or provide help to B cells (T follicular helper, TFH) (Chap. 314). TH1 cells facilitate macrophage activation and classic cell-mediated immunity, whereas TH2 cells are thought to have regulatory functions and are involved in the resolution of normal immune responses and also the development of responses to a variety of parasites; TH17 cells produce a number of inflammatory cytokines, including IL-17 and IL-22, and TFH cells help B cells by constitutively producing IL-21. In a number of autoimmune diseases, such as rheumatoid arthritis, multiple sclerosis, type 1 diabetes mellitus, and Crohn's disease, there appears to be biased differentiation of TH1 cells, with resultant organ damage. More recently, studies suggest accentuated differentiation of TH17 cells associated with animal models of inflammatory arthritis and also rheumatoid arthritis, whereas increased differentiation of TFH cells has been associated with animal models of SLE.
Organ-Specific versus Systemic Autoimmune Diseases
Autoimmune diseases form a spectrum, from those specifically affecting a single organ to systemic disorders with involvement of many organs (Table 318-5). Hashimoto's autoimmune thyroiditis is an example of an organ-specific autoimmune disease (Chap. 341). In this disorder, there is a specific lesion in the thyroid associated with infiltration of mononuclear cells and damage to follicular cells. Antibody to thyroid constituents can be demonstrated in nearly all cases. Other organ- or tissue-specific autoimmune disorders include pemphigus vulgaris, autoimmune hemolytic anemia, idiopathic thrombocytopenic purpura, Goodpasture's syndrome, myasthenia gravis, and sympathetic ophthalmia. One important feature of some organ-specific autoimmune diseases is the tendency for overlap, such that an individual with one specific syndrome is more likely to develop a second syndrome. For example, there is a high incidence of pernicious anemia in individuals with autoimmune thyroiditis. More striking is the tendency for individuals with an organ-specific autoimmune disease to develop multiple other manifestations of autoimmunity without the development of associated organ pathology. Thus, as many as 50% of individuals with pernicious anemia have non-cross-reacting antibodies to thyroid constituents, whereas patients with myasthenia gravis may develop antinuclear antibodies, antithyroid antibodies, rheumatoid factor, antilymphocyte antibodies, and polyclonal hypergammaglobulinemia. Part of the explanation for this may relate to the genetic elements shared by individuals with these different diseases.
Table 318-5 Some Autoimmune Diseases |Favorite Table|Download (.pdf)
Table 318-5 Some Autoimmune Diseases
Autoimmune hemolytic anemia
Autoimmune polyglandular syndrome
Autoimmune thrombocytopenic purpura
Type 1 diabetes mellitus
Insulin-resistant diabetes mellitus
Autoimmune Addison's disease
Acute rheumatic fever
Organ Nonspecific (Systemic)
Systemic lupus erythematosus
Granulomatosis with polyangiitis (Wegener's)
Systemic necrotizing vasculitis
Systemic autoimmune diseases differ from organ-specific diseases in that pathologic lesions are found in multiple diverse organs and tissues. The hallmark of these conditions is the demonstration of associated relevant autoimmune manifestations that are likely to be etiologic in the organ pathology. SLE represents the prototype of these disorders because of its abundance of autoimmune manifestations.
SLE is a disease of protean manifestations that characteristically involves the kidneys, joints, skin, serosal surfaces, blood vessels, and central nervous system (Chap. 319). The disease is associated with a vast array of autoantibodies whose production appears to be a part of a generalized hyperreactivity of the humoral immune system. Other features of SLE include generalized B cell hyperresponsiveness and polyclonal hypergammaglobulinemia. Current evidence suggests that both hypo- and hyperresponsiveness to antigen can lead to survival and activation of autoreactive B cells in SLE.
Treatment: Autoimmune Diseases
Treatment of autoimmune diseases can focus on either suppressing the induction of autoimmunity, restoring normal regulatory mechanisms, or inhibiting the effector mechanisms. To eliminate autoreactive cells, immunosuppressive or ablative therapies are most commonly used. In recent years, cytokine blockade has been demonstrated to be effective in preventing immune activation in some diseases. New therapies have also been developed to target lymphoid cells more specifically, either by blocking a co-stimulatory signal needed for T or B cell activation, by blocking the migratory capacity of lymphocytes, or by eliminating the effector T cells or B cells. The efficacy of these therapies is not yet demonstrated. Newer trials are testing the possibility of using autoantigen itself to induce tolerance. One major advance in inhibiting effector mechanisms has been the introduction of cytokine blockade, targeting TNF or IL-1, that appears to limit organ damage in some diseases. Biologicals that interface with T cell activation (CTLA-4Ig) or delete B cells (anti-CD20 antibody) have also recently been approved for the treatment of rheumatoid arthritis. Therapies that prevent target organ damage or support target organ function remain an important therapeutic approach to autoimmune disease.