Autoimmune responses are very similar to immune responses to non- self- antigens. Both are driven by antigen, involve the same adaptive immune cell types and produce tissue damage by the same effector mechanisms – both T cells and B cells. The development of autoimmunity, however, implies a failure of the normal regulatory mechanisms. These regulatory mechanisms are discussed first so that reasons for their breakdown can be examined.
Autoimmunity and self- tolerance
Strong protective mechanisms exist to prevent the development of autoimmune disease. As outlined in Chapter 1, the immune system has the ability to generate a vast diversity of different T- cell antigen receptors and immunoglobulin molecules by differential genetic recombination. This process produces many antigen- specific receptors capable of binding to self- molecules. To avoid autoimmune disease, the T and B cells bearing these self- reactive molecules must be either eliminated or downregulated. Because T cells (in particular CD4+ T cells) have a central role in controlling nearly all immune responses, the process of T- cell tolerance seems to be of greater importance in avoiding autoimmunity than B- cell tolerance, since self- reactive B cells require CD4+ T cells’ help to produce a fully fledged IgG antibody response. However, antigen receptors on B cells undergo somatic mutation as well as recombination events, resulting in additional opportunities for chance cross- reactivity with self- antigens. There are, therefore, several mechanisms to maintain tolerance. These are traditionally divided into central mechanisms, such as those that act early in development in the thymus, and peripheral mechanisms to eliminate those arising later in adaptive immune responses.
Thymic tolerance
The thymus plays a major role in eliminating T cells capable of recognizing peptides from self- proteins (Fig. 1). There are three processes. Developing T cells have antigen receptors (TCRs) that are produced randomly by recombination, but all have the ability to bind to major histocompatibility complex (MHC) molecules; only those that recognize the host MHC in the thymic cortex survive; >90% of the devel oping T cells die from ‘neglect’. T cells with TCRs that fail to bind the host MHC molecules in the thymic cortex die through apoptosis. T cells that survive this process move to the thymic medulla and bind, with a variety of affinities, to the MHC molecules and self- peptide complexes presented on medullary epithelial cells, from whom they also receive survival signals. Those T cells that bind with low affinity are allowed to survive and have the potential to bind to MHC plus a foreign peptide with high affinity, and so initiate protective immune responses at a later time. This is the process known as positive selection. In a further process known as negative selection, T cells that bind to MHC plus self- peptides with very high affinity in the thymus, and so have clear potential for self- recognition elsewhere in the body with consequent induction of autoimmunity, are eliminated on the basis of their high- affinity binding. This elimination of self- reactive cells is known as negative (deletional) selection (see Fig. 1).
Fig1. Representation of T- cell selection in the thymus by thymic epithelial cells (TEC). CMJ, corticomedullary junction. Courtesy of Georg Hollander.
This process of thymic education is only partially successful, as autoreactive T cells can be detected in healthy persons. The most important reason for failure of thymic tolerance is that many self- peptides are not expressed at a sufficient level in the thymus to induce negative selection. Many peptides found bound to MHC molecules in the thymus are from either ubiquitous intracellular or membrane- bound proteins or proteins present in the extracellular fluid, expression of which is controlled by a transcription factor known as AIRE. The negative selection of self- reactive thymocytes depends on the expression of tissue- specific antigens by medullary thymic epithelial cells. The AIRE protein plays an important role in presenting these antigens, and the absence of even one AIRE- induced tissue- specific antigen in the thymus can lead to autoimmunity in the antigen- expressing target organ.
The vast number of self- antigens, and the need to present them all to each developing T lymphocyte, renders thymic tolerance imperfect. Tolerance is induced to some, but not all, tissue- specific proteins. It is not surprising therefore that T cells responsive to tissue- specific proteins – for example, cartilage collagens or some central nervous system (CNS) antigens – can be detected in healthy people under certain laboratory conditions. A second level of control is needed to control these potentially autoreactive cells and this is known as peripheral tolerance.
Peripheral tolerance
There are several mechanisms by which peripheral tolerance is maintained. These are outlined here.
Ignorance
A restricted form of peripheral tolerance exists because the antigen is effectively invisible to the immune system. This is known as immunological ignorance (Fig. 2). Immunological ignorance can occur because the antigen is sequestered in an avascular organ such as the intact vitreous humour of the eye, although when limited amounts of antigen do escape from these sites due to inflammation this will break down. More importantly, immunological ignorance occurs because CD4+ T cells (which are required to initiate most immune responses) will only recognize antigens presented in association with MHC class II molecules. The limited distribution of MHC class II molecules in healthy tissues means that most organ- specific molecules are not expressed alongside MHC class II in health and so are not able to be presented at levels high enough to induce T- cell activation. However, in inflamed tissues MHC II is expressed as part of the inflammation, so additional processes are required to prevent or control self- reactivity.
Fig2. How peripheral T- cell tolerance is maintained. IL, interleukin; TCR, T- cell receptor; TGF, transforming growth factor.
Separation of autoreactive T cells and autoantigens
Self- antigens and lymphocytes are also kept separate by the restricted routes of lymphocyte circulation that limit naive lymphocytes to secondary lymphoid tissue and the blood (Fig. 3). Since antigens, including self- antigens, drain via afferent lymphatics to local lymph nodes, and the naive T cells in normal nodes do not express co- stimulatory molecules, no activation occurs in the absence of inflammation (see the next section), i.e. activation of dendritic cells via Toll- like receptors (Fig. 2).
Fig3. Diagrammatic representation of the different recirculation pathways taken by naive T cells (dotted line) and T cells that have been previously exposed to antigen (solid line).
To prevent large amounts of self- antigen from gaining access to dendritic cells, debris from self- tissue breakdown needs to be cleared rapidly and destroyed. This is achieved by cell death through apoptosis, preventing widespread spilling of cell contents, or autophagy to recycle cellular components. A variety of scavenger mechanisms help to clear cell debris, including the complement systems, certain acute- phase proteins (such as serum amyloid P and C- reactive protein) and a number of receptors found upon phagocytes. Defects of complement or phagocytes are associated with the development of autoimmunity against intracellular molecules (Table 1).
Table1. Single- gene defects that provide an insight into autoimmune diseases
Anergy, apoptosis and autophagy
The pattern of stimulation through the TCR may also be important in determining whether a cell becomes activated or anergic: continuous low- level stimulation of a small number of T- cell antigen receptors (as might occur with a self- antigen) tends to produce anergy, while strong and rap idly increasing stimulation (as in infection) favours activation.
More active mechanisms of peripheral tolerance also operate after antigen recognition (see Fig. 2). These involve either deletion of self- reactive cells by apoptosis or induction of a state of unresponsiveness, anergy. Naive CD4+ T cells need two signals to become activated and initiate an immune response: an antigen- specific signal through the T- cell antigen receptor and a second, non- specific co- stimulatory signal, usually signalled by CD28 (on the T cell) binding to one of the B7 family (CD80 or CD86) on the dendritic or specialized B cell. If the T cell receives both signals, then it will become activated and proliferate and produce cytokines. If no co- stimulatory molecules are engaged, then stimulation through the TCR alone leads to longstanding anergy or death of the T cell by apoptosis. The expression of these co- stimulatory molecules is tightly controlled. Constitutive expression is confined to specialized antigen- presenting cells such as dendritic cells; given their distribution and patterns of recirculation, interaction between CD4 cells and dendritic cells is likely to happen only in secondary lymphoid tissues such as lymph nodes. The restricted expression of co- stimulatory molecules means that, even if a T cell recognizes a tissue- specific peptide/MHC molecule complex (e.g. an antigen derived from a pancreatic islet cell), then anergy rather than activation is likely to follow, as no antigen- presenting cell (and therefore co- stimulatory signal) will be available in healthy tissue. Expression of co- stimulatory molecules can be induced upon other cells by a variety of stimuli, usually in association with inflammation or cell damage. However, because of the restricted pattern of lymphocyte recirculation, only previously activated cells will gain access to these peripheral sites and these peripheral co- stimulatory molecules are more likely to sustain than initiate an autoimmune response.
Two types of external signals can initiate apoptosis (programmed cell death). Tumour necrosis factor alpha (TNFɑ), the inflammatory cytokine secreted mainly by macrophages, is a major process since almost all cells in the human body have receptors for TNF. The other major external apoptotic pathway involves the expression of FAS ligand (FASL) on T cells following activation. This protein then binds to FAS on adjacent cells and results in activation- induced cell death (AICD). This process is needed to prevent excessive immune responses and to eliminate autoreactive T cells. There is a third, intrinsic pathway and possibly more, since these are important regulatory processes. There are also inhibitors of apoptosis, the absence of which can result in excessive cell death such as in inflammatory diseases like in the gut.
Autophagy is a basic mechanism conserved in evolution, which is essential for maintaining cellular components in the face of cellular stress (such as nutrient starvation). Damaged proteins and organelles are recognized and recycled, preventing initiation of an immune response. The role of defective autophagy is gradually being appreciated in many inflammatory and neurodegenerative conditions.
Regulation and suppression
Suppression of unwanted immune responses is so important that there are several other mechanisms for peripheral tolerance, including active suppression of self- reactive T cells by regulatory populations of T cells (Tregs). There are both antigen- specific induced Tregs, which recognize the same antigen as T- effector cells, and natural Tregs, which are non- antigen specific. Both are formed in the thymus and are generally defined by the markers CD4, CD25 and forkhead box protein P3 (FoxP3). CD4+CD25+FoxP3+ cells exert their regulatory effects either through secretion of immunosuppressive cytokines such as IL- 10 and TGF- β or through cell contact- dependent mechanisms, such as CTLA- 4 expression.
Activated T cells can also express cell- surface molecules similar in structure to co- stimulatory molecules, but which exert a negative effect upon T- cell activation; in particular, CTLA- 4 has a similar structure to CD28 and binds to the same ligands to prevent or limit activation. Binding of CTLA- 4 to CD80 or CD86 induces anergy or death by apoptosis, as a negative counterpart to co- stimulation that may be important in terminating an immune response. The importance of apoptotic death of autoreactive lymphocytes in preventing autoimmune disease is emphasized by the development of autoimmunity in patients with immune- genetic defects in the control of apoptosis (see Table 1). This has also been exploited in the development of a therapeutic monoclonal antibody that blocks the CTLA4 regulatory signal, which allows cytotoxic T cells to destroy malignant melanoma cells. However, removal of regulatory T- cell inhibition is associated with the induction of iatrogenic autoimmunity in a range of organs (see Table 1 and Fig. 4). There is also evidence in mice of a variety of specific CD8+ T- regulatory cells (known as CD8+ Ti) that contribute to peripheral tolerance and such cells have particular features (interferon and apoptosis gene activation) to suggest a protective phenotype. However, their role in humans has yet to be defined.
Fig4. Schematic representation of sites of action of checkpoint inhibitors.
B- cell tolerance
B- cell tolerance is less complete than T- cell tolerance since there is no organ equivalent of the thymus in mammalian B- cell development. So B- cell tolerance operates at a peripheral rather than central level, and multiple mechanisms exist to limit B- cell autoreactivity: clonal deletion, clonal anergy, receptor editing and maturation arrest.
The production of self- reactive antibodies is limited mainly by the lack of T- cell help for self- antigens. New B cells are being produced continuously from bone marrow precursors and many of these are autoreactive. The process of somatic hypermutation of immunoglobulin genes in mature B cells in the germinal centres of lymph nodes also has the potential for the generation of autoantibodies. However, B cells are able to undergo receptor editing in the bone marrow – a process that allows new light- chain rearrangement if an autoreactive B cell is produced, to enable new non- self- specificities. In addition, if a newly developed or recently hypermutated B cell binds to the self- antigen, then in the absence of specific antigen T- cell help, that B cell will undergo apoptosis or anergy. Thus, there are some similarities between T- and B- cell activation and tolerance, in that two signals are required for activation and the presence of an antigen- specific signal alone leads to death or anergy. As with T cells, low- level chronic stimulation via the antigen receptor favours anergy, and rapidly increasing amounts of antigen favour activation.
