1. Induction of autoimmunity by immunization ('immunization models')
This approach, the first to be introduced, has been of primary importance in establishing the validity of the autoimmune concept. Furthermore, since it has been possible by this means to raise autoimmune responses to an extensive range of self components (Table 1), it is evident that potentially autoreactive cells are not only a normal component of the immune repertoire of every individual but also that they are quite capable of being activated under appropriate conditions. As a corollary, it also follows that, under normal circumstances, these cells are effectively subjugated by one or more regulatory mechanisms. In consequence, it is generally necessary to inject native autoantigens in combination with powerful nonspecific immunostimulants such as Freund's water-in-oil adjuvant in order to override these inhibitory influences. As an alternative procedure, the immunogen can be altered self components or material derived from other species which may provide a more effective autoimmune stimulation by way of 'modified self' or cross-reactive antigenicity. Weigle, for example, used rabbit thyroglo-bulin to induce thyroiditis in mice and the injection of rat erythrocytes into mice also induces antibodies to mouse erythrocytes and transient anemia. Presumably, under these circumstances, the inclusion of new antigenic determinants can provide a bypass mechanism permitting new T cell help for the generation of autoimmune responses.
Immunization models have been extensively used
Table 1 Induction of experimental autoimmunity by immunization with autoantigens
Target organAissue Thyroid follicles
Pancreas (islet cell)
Central nervous system neurons
Eye (uvea/retina) Kidney (tubules/glomeruli)
Joint (articular surface) Liver (hepatocytes) Testes (spermatozoa) Heart (myocardium) Thrombocytes Erythrocytes Neuromuscular junction
Rabbit, mouse, guinea pig, rat, monkey, chicken, dog Guinea pig, rabbit, rat (Lewis, BN), mouse (CF1, BSVS), monkey, dog Rat, rabbit
Rat, dog, chicken, mouse
Rat, mouse, rabbit
Rat (Lewis), guinea pig (strain B), mouse, sheep, monkey Guinea pig (Hartley and NIH) Rat (BN) Rat
Mouse (DBA), rat, rabbit Mouse
Mouse (C57 black) Guinea pig, mouse Rat (BN) Mouse
Rat, guinea pig
Analogous human disease Hashimoto's thyroiditis
Autoimmune adrenal failure (Addison's disease) Lymphocytic adenohypophys'rtis Autoimmune parathyroiditis Type I diabetes Multiple sclerosis
Autoimmune uveitis Autoimmune tubulointerstitial nephritis Autoimmune glomerulonephritis Rheumatoid arthritis Chronic active hepatitis Autoimmune orchitis Autoimmune myocarditis Idiopathic thrombocytopenia Autoimmune hemolytic anemia Myasthenia gravis to study genetic influences on autoimmune susceptibility. As might be expected, strong major histocompatibility complex (MHC) influences have been observed in some models as, for example, in mouse thyroiditis and experimental autoimmune encephalomyelitis (EAE) in rats. However, this does not seem to apply to the mouse antierythrocyte/anemia model mentioned above, although non-H2 genetic associations are evident.
One major disadvantage of this procedure is that the pathological effects induced are often transient in nature, presumably due to the imposition of autoimmunity on the normal animal with intact immuno-regulatory feedbacks. As a consequence, overt clinical disease is not observed in many immunization models, even though short-term tissue infiltration is readily apparent histologically. However, EAE is a notable exception in that severe nervous signs can be readily induced by immunization of susceptible animals with myelin basic protein or its peptide derivatives. Clearly, the nervous system is more sensitive to the effects of immunological attack and these are more apparent clinically than, for example, those involving the endocrine tissues.
Although the relapsing form of EAE can be induced in guinea pigs by this method, in most experimental systems autoimmunity is rarely reactivated even by further immunization. There is evidence in a number of models that this effect may be due to the generation of antigen specific 'suppressor' cells, as demonstrated by adoptive cell transfer stud ies. Evidence for such activity has been found in mouse thyroiditis, mouse antierythrocyte autoimmunity and EAE in rats.
Although the injection of autoantigens in adjuvant can clearly have little relevance to the natural triggering events leading to human autoimmune disease development, immunization models continue to provide a wealth of information on many other aspects of autoimmunity. Currently, with the rapid accumulation of information at the molecular level concerning the subcellular components involved in autoimmunity, this procedure is likely to be utilized as a requisite step in the evaluation of purified recombinant polypeptides derived from such structures as potential B or T autoepitopes.
2. Induction of autoimmunity by impairment of immunological function ('immune depletion models')
It is now accepted that partial impairment of normal immunological function can lead to the onset of autoimmune reactivity, particularly of the organ-specific type. The earliest evidence in this context was the observation that thymectomy of certain mouse strains leads to the appearance of antinuclear antibodies. However, this relationship was firmly established in 1973 by two groups who independently reported the induction of organ-specific autoimmunity by procedures causing partial immuno-depletion. Penhale and coworkers found that autoimmune thyroiditis could be induced by thymec-
tomy of 3-week-old Wistar rats followed by repeated low-dose irradiation. In a similar vein, Nishizuka and colleagues made the observation that thyroiditis spontaneously developed in mice following thymectomy alone, provided that this was carried out in the neonatal period. The autoimmune basis of these conditions was evident from the presence of infiltrating lymphocytes in the target tissue, circulating tissue-specific autoantibodies and the ability to transfer the disease to naive recipients with lymphocytes from affected animals. From 1975 onwards a number of other manipulations and treatments involving the immune system were shown to induce organ-specific autoimmune disease in rodents. A feature of all these procedures is either a deficiency of circulating T lymphocytes or their functional impairment as, for example, by a reduction in the range of the T cell receptor (TCR) repertoire. The spectrum of autoimmune diseases induced by this means is now impressive and includes diabetes, thyroiditis, gastritis, orchitis, oophoritis, prostatitis and coagulating adenitis. These conditions are rarely, if ever, seen in normal unmanipulated rodents of the same strains. The various depletion methods employed to induce these diseases are shown in Table 2.
A central question in understanding the process of autoimmunity is how the aberrant immune response is initiated. The development of immunodepletion models has been conceptually important in demonstrating the critical requirement for regulatory elements in the control of potential autoimmune reactivity in the normal individual. They also strongly suggest that neither aberrant antigen modification nor presentation are likely to be necessary antecedents of the autoimmune process. In consequence, they provide a powerful tool for studying the basis for the loss of self tolerance to peripheral self antigens and the development of destructive autoimmune lesions. In this regard they have proved particularly suitable for studying the immunological regulation of autoimmunity by reconstitution studies
Table 2 Manipulations of the immune system leading to the development of organ-specific autoimmunity
Combined sublethal irradiation/thymectomy of adult rodents Neonatal thymectomy of rodents Neonatal administration of cyclosporine in mice Combined cyclophosphamide/thymectomy of adult rodents High dose (42.5 Gy), fractionated total lymphoid irradiation of adult mice Single-chain TCR transgenic mice T cell transfer to syngeneic, T cell-deficient mice using syngeneic lymphocytes derived from normal animals.
Immunodepletion models do not require immunization with putative autoantigens in adjuvant, as in the previous model above, but presumably involve the appropriate targeting of natural autoantigens rather than surrogate model antigens. In addition, they are also similar to their clinical counterparts in humans in that, once initiated, the disease is sustained and chronic. Furthermore, an association between various immunodeficient states and clinical autoimmunity in humans has been recognized. Because of these similarities in many pathological and immunological features they are providing invaluable insights into the pathogenesis of their human disease counterparts.
3. Induction of autoimmunity by infectious agents ('infection models')
The natural history of autoimmune disease in some individual patients strongly suggests the involvement of environmental or infectious agents in the inductive process. For example, postmeasles encephalomyelitis may result from an autoimmune response to myelin basic protein, and infection with Coxsackie virus group B has been associated with myocarditis in which there are autoantibodies to heart antigens. Despite this circumstantial evidence in isolated instances, direct evidence for the involvement of infectious agents in human autoimmune disease is unavailable. In contrast, experimental studies in animals provide some compelling evidence of the involvement of infectious agents, particularly viruses, in the triggering of certain forms of autoimmunity. Several viruses have been shown to induce diabeteslike conditions in inbred strains of mice with evidence of associated autoimmunity. Polyendocrine syndromes and diseases of nonendocrine organs have also been observed in certain infections. Certain bacteria or their products may also cause experimental autoimmune effects. For example, the injection of mycobacterial proteoglycan may induce arthritis, several bacterial products may similarly cause myasthenia (see below) and the injection of live Listeria directly into the testes can provoke an autoimmune orchitis.
The phenomenology of infection models require further investigation before a detailed understanding of their mechanisms can be gained and an autoimmune basis for their pathogenesis is unequivocally established. However, a number of immunological processes may underlie their pathogenicity, including:
Apart from the expression of virus gene products in persistently infected cells which may target them for host immune responses, virus may also have subtle effects on cell genes and the resulting enhanced or altered expression of host cell proteins could induce an autoimmune response. A further possibility is interaction of microorganisms with immunocompetent cells, leading to their polyclonal activation or reduced susceptibility to regulation. They may also cause a loss of function of regulatory cells with autoimmune consequences.
2. Molecular mimicry. Partial sequence homology between microbial antigens and self determinants may lead to the generation of host-specific responses by T helper cell bypass.
3. Idiotype-anti-idiotype interactions (see below).
4. Tissue injury induced by infection may lead to the breaking of tolerance to self antigens and subsequent perpetuation of injury by autoimmune mechanisms. This mechanism could account for the induction of autoimmune disease by an acute, nonpersisting infection.
5. Enhanced expression of MHC antigen by microbial influences directly on the target cell or indirectly via lymphocytes.
4. Induction of autoimmunity by chemical agents ('chemically induced models')
A number of chemical substances appear to give rise to autoimmune effects in patients after prolonged use as therapeutic agents; for example, methyl dopa may cause a Coornbs'-positive hemolytic anemia. Similarly, in the experimental context, several chemical agents have been shown to induce autoimmune-like syndromes, such as streptozotocin-induced diabetes in mice and mercuric chloride-induced glomerulonephritis in brown Norway rats. There are several possible mechanisms whereby chemicals may be involved in generating autoimmune manifestations, including: 1) modification of tissue components, particularly cell membrane structures; 2) breaking of self tolerance following liberation of cellular constituents by direct tissue damage; and 3) modification of immune function, particularly affecting regulatory cells. In general, withdrawal of the chemical agent results in recovery and diseases induced by this means have few clear analogies with naturally occurring human disorders. Low dose streptozotocin-induced diabetes mellitus in mice is a possible exception to the above statement, although the actual pathogenesis, whether direct toxicity or true autoimmunity, is still in doubt.
5. Induction of autoimmunity by anti-idiotype (Id) responses ('anti-idiotype models')
Theoretically, several mechanisms of induction of autoimmunity through Id interactions appear to be possible, although few examples of such are currently documented and none established as regularly utilized models for inducing autoimmunity. First, an anti-Id antibody may resemble an autoantigen sufficiently to be cross-reactive. This is exemplified in the insulin system, where it has been found that an antibody to the Id of anti-insulin antibody may interact with and actually trigger insulin receptors. Another, more dramatic example of the same mechanism appears to be the myasthenia gravis-like syndrome observed when rabbits are injected with a synthetic agonist (Bis Q) of the acetylcholine receptor. A second way in which the Id network can stimulate autoimmunity is through the development of 'parallel sets'. These may arise when immunologically distinct antibodies share a 'public' idiotype. As an example, extensive idiotypy has been discovered between the antibodies to the acetylcholine receptor and the DEX antigen (1,3-dextran) present on the surface of many bacteria. Thus, the injection of such bacteria may induce autoantibodies to the receptor and hence myasthenia gravis. It may be no coincidence that many examples of autoimmune anti-Id effects involve receptor molecules.
6. Induction of autoimmunity by genetic manipulation ('transgenic models')
Transgenic technologies have provided new opportunities for developing animal models of autoimmune diseases. To date their development has been largely directed at obtaining new insights into the fundamental questions of autoimmunity, such as the nature of self tolerance, rather than the provision of relevant models of human disease. Expression of well-defined molecules can be induced in specific cell types by this technology and three broad approaches have been utilized to study aspects of autoimmunity:
1. The genes selected may code for a normally foreign antigen which will be synthesized at a particular development stage and behave essentially like a self antigen. The gene may be linked to a tissue-specific promoter which ensures its expression in a particular tissue while simultaneously avoiding thymic expression. Thus it is possible to generate cells expressing a genuine self molecule without concurrent thymic deletion of the complementary lymphocytes.
2. The genes selected may be rearranged T cell receptor or immunoglobulin genes coding for recognition molecules with specificity for a self component. This procedure has the potential to generate antiself lymphocytes at high frequency, allowing the fate of such cells to be followed during and after ontogeny. Alternatively, the TCR repertoire can be greatly restricted by introducing the gene coding for an irrelevant nonself specificity. Thus the T cell compartment can be functionally depleted, leading to autoimmunity (see method 2 above). 3. The transgene may code for MHC molecules in order to address the question of the significance of their upregulation on target tissues during autoimmunity. This has led to the surprising finding that diabetes of nonimmune origin can be induced by overexpression of these molecules on the islet ¡3 cells. This is presumably due to derangement of cell metabolism as a consequence of high class II expression.
These strategies can also be combined to produce multiple transgenic animals in which, for example, an antigenic molecule can be directed to a specific cell type and the complementary TCR introduced into lymphocytes simultaneously.
The ¡3 cell of the pancreatic islet currently appears to be the most popular location for transgenic modification. Apart from the importance of type I diabetes as a human autoimmune syndrome, this cell is particularly convenient in that the transgene can be linked to the specific promoter for insulin production. Studies with transgenically modified 0 cells have been revealing in a number of respects. For example, experiments with lymphocytic choriomeningitis virus (LCMV) glycoprotein antigen expression on these cells point to immunological ignorance rather than tolerance as the mechanism responsible for the absence of reactivity to islet cell antigens, as diabetes does not occur spontaneously in this model and can only be induced by subsequent infection with LCMV. Furthermore, a number of additional lines of evidence indicate strongly that the virus is not breaking immunological tolerance in this experiment. New generations of these transgenic models are likely to revolutionize our understanding of this central phenomenon of autoimmunity.
Was this article helpful?