Overview

The evolutionary pressures that have molded the immune response and promoted a highly diverse repertoire clearly derive from infectious agents. Two different strategies exist. The more primitive innate immune response (see Chapter 18) uses a limited repertoire of nonpolymorphic receptors that recognize structural motifs that are common to many microorganisms. These include small glycolipids and lipopeptides. The alternative strategy of the evolutionarily newer adaptive immune response (see Chapter 19) relies on generating myriad different receptors that can recognize a wide array of foreign compounds from infectious agents. Whereas the innate immune response allows a rapid focused response, adaptive immunity permits a broader, albeit slower, response but critically offers the additional benefit to the host of immune memory.

T lymphocyte development constantly confronts the dilemma of combating infection without provoking a response to the host. The price for generating an increasingly varied population of antigen receptors needed to recognize a wide spectrum of pathogens is the progressive risk of producing self-reactive lymphocytes that can provoke an autoimmune diathesis. To minimize the possibility of self-reactive cells, T lymphocytes are subjected to a rigorous selection process during development in the thymus. In addition, premature activation of mature T cells is prevented by requiring two signals for activation. Finally, the tremendous expansion of T cells that occurs during the response to an infection is resolved by the active induction of cell death. The consequences of inefficient lymphocyte removal at any one of these junctures can be devastating to the health of the organism. This is vividly displayed in both humans and mice where naturally arising mutations in death receptors such as Fas result in massive accumulation of lymphocytes and autoimmune sequelae. These are discussed in more detail in Chapter 27 on cell survival and death.

The activation of T lymphocytes yields a variety of effector functions that are pivotal to combating infections. Cytolytic T cells can kill infected cells through the expression of perforin, which induces holes in cell membranes, or ligands for death receptors such as Fas or tumor necrosis factor (TNF) receptor. Production of T cell cytokines such as interferon-γ (IFN-γ) can inhibit viral replication, whereas other cytokines such as interleukin (IL)-4, IL-5, and IL-21 are critical for optimal B cell growth and immunoglobulin production.¹ However, this same armamentarium, if not tightly regulated, can also precipitate damage to host tissues and provoke autoimmune responses. This is particularly apparent in situations where T cell infiltration can be observed histologically such as in the synovium of inflammatory arthritides, pancreatic islets in type 1 diabetes, and the central nervous system in multiple sclerosis. Damage in these cases need not be the direct result of recognition of target tissues by the T cells. T cells may be activated elsewhere and then migrate to the tissue and damage innocent bystander cells. T cells may also promote autoimmunity through the augmentation of B cell responses.

T Cell Development

T cells must traverse two stringent hurdles during their development. First, they must successfully rearrange the genes encoding the two chains of the T cell antigen receptor (TCR). Second, T cells must survive thymic selection during which T cells that interact strongly with self-peptides are eliminated. This minimizes the chances of autoreactive T cells escaping to the periphery.

The TCR is an 80- to 90-kD disulfide-linked heterodimer composed of a 48- to 54-kD α chain and a 37- to 42-kD β chain. An alternate TCR composed of γ and δ chains is expressed on 2% to 3% of peripheral blood T cells and is discussed later. The TCR has an extracellular ligand binding pocket and a short cytoplasmic tail that by itself cannot signal. Consequently it is noncovalently associated with as many as five invariant chains of the CD3 complex that relay information to the intracellular signaling machinery via immunoreceptor tyrosine activation motifs (ITAMs) (see later). Not surprisingly, the structure of the TCR gene is similar to what was first described for immunoglobulin genes in B cells (see details in Chapter 14). Each overcame the problem of how to encode approximately 10 million different T or B cell specificities within the human genome, which contains only 30,000 genes. To economically package this diversity, the process of gene rearrangement and splicing evolved using machinery similar to that which already existed to promote gene translocations. The β and δ chain genes of the TCR contain four segments known as the V (variable), D (diversity), J (joining), and C (constant) regions. The α and γ chains are similar but lack the J component. Each of the segments has several family members (≈ 50 to 100 V, 15 D, 6 to 60 J, and 1 to 2 C members). An orderly process occurs during TCR gene rearrangement in which a D segment is spliced adjacent to a J segment, which is subsequently spliced to a V segment.

Following transcription, the VDJ sequence is spliced to a C segment to produce a mature TCR messenger RNA. Arithmetically, this random rearrangement of a single chain of the TCR locus can give rise to a minimum of 50V × 15D × 6J × 2C, or about 9000 possible combinations. At each of the splice sites, which must occur in-frame to be functional, additional nucleotides not encoded by the genome (so-called N-region nucleotides) can be incorporated, adding further diversity to the rearranging gene. The combinations from the two TCR chains, plus N-region diversity, yield at least 10⁸ possible combinations. Cutting, rearranging, and splicing are directed by specific enzymes. Mutations in the genes mediating these processes can result in arrest in lymphocyte development. For example, mutation in the gene encoding a DNA-dependent protein kinase required for receptor gene recombination results in a severe combined immunodeficiency (SCID).

Because the developing T cell has two copies of each chromosome, there are two chances to successfully rearrange each of the two TCR chains. As soon as successful rearrangement occurs, further β-chain rearrangements on either the same or the other chromosome are suppressed, a process known as allelic exclusion. This limits the chance of dual TCR expression by an individual T cell. The high percentage of T cells that contain rearrangements of both β-chain genes attests to the inefficiency of this complex event. Rearrangement of the α chain occurs later in thymocyte development in a similar fashion, although without apparent allelic exclusion. This can result in dual TCR expression by a single T cell.

Development of T cells occurs within a microenvironment provided by the thymic epithelial stroma. The thymic anlage is formed from embryonic ectoderm and endoderm and is then colonized by hematopoietic cells, which give rise to dendritic cells, macrophages, and developing T cells.² The hematopoietic and epithelial components combine to form two histologically defined compartments: the cortex, which contains immature thymocytes, and the medulla, which contains mature thymocytes (Figure 13-1A). As few as 50 to 100 bone marrow–derived stem cells enter the thymus daily.

Figure 13-1 Sequence of thymocyte development. A, The earliest thymocyte precursors lack expression of CD4 and CD8 (CD4⁻CD8⁻). These can be further divided into four subpopulations based on sequential expression of CD44 and CD25. It is at the CD44⁻CD25⁺ stage that the TCR-β-chain rearranges. The SCID mutation or deficiencies of the rearrangement enzymes Rag-1 and Rag-2 result in inability to rearrange the β-chain and maturational arrest at this stage. Those thymocytes that successfully rearrange the β-chain express it associated with a surrogate α-chain known as pre-Tα. Concomitant with a proliferative burst, development can then progress to the CD4⁺CD8⁺ stage in the cortex where the TCR-α-chain rearranges and pairs with the β-chain to express a mature TCR complex. These cells then undergo thymic positive and negative selection (as diagrammed in Figure 13-3B). Successful completion of this rigorous selection process results in mature CD4⁺ or CD8⁺ T cells in the medulla, which eventually emigrate to peripheral lymphoid sites. B, Schematic two-color flow cytometry showing subpopulations of thymocytes defined by CD4 and CD8 expression in their relative proportions.

The stages of thymocyte development can be defined by the status of rearrangement and expression of the two genes that encode the α and β chains of the TCR and the expression of CD4 and CD8, proceeding in an orderly fashion from CD4⁻8⁻ → CD4⁺8⁺ → CD4⁺8⁻ or CD4⁻8⁺ (Figure 13-1B). CD4 and CD8 define, respectively, the helper and cytolytic subsets of mature T cells.

CD4⁻8⁻ thymocytes can be further subdivided based on their expression of CD25 (the high-affinity IL-2 receptor α chain) and CD44 (the hyaluronate receptor).³ Development proceeds in this order: CD25⁻CD44⁺ → CD25⁺CD44⁺ → CD25⁺CD44⁻ → CD25⁻CD44⁻. These subpopulations correspond to discrete stages of thymocyte differentiation. CD25⁻44⁺ cells express low levels of CD4, and their TCR genes are in germline configuration. These cells downregulate CD4 and upregulate CD25 to give rise to CD25⁺CD44⁺ thymocytes, which now express surface CD2 and low levels of CD3ε. At the next stage (CD25⁺CD44⁻), there is a brief burst of proliferation followed by upregulation of the recombination-activating enzymes, RAG-1 and RAG-2, and the concomitant rearrangement of the genes of the TCR β chain. A small subpopulation of T cells rearranges and expresses a second pair of TCR genes known as γ and δ. Productive TCR β-chain rearrangement results in downregulation of RAG and a second proliferative burst. Loss of CD25 then yields CD25⁻CD44⁻ thymocytes.

The TCR β chain cannot be stably expressed without an α chain. Because the TCR α chain has not yet rearranged, a surrogate invariant TCR pre-α chain is disulfide linked to the β chain.⁴ When associated with components of the CD3 complex, this allows a low-level surface expression of a pre-TCR and progression to the next developmental stage. Failure to successfully rearrange the TCR β chain results in a developmental arrest at the transition from CD25⁺CD44⁻ to CD25⁻CD44⁻. This occurs in RAG-deficient mice, as well as in mice and humans with SCID.⁵

A number of signaling molecules are required for early T cell development (Figure 13-2). The Ikaros gene encodes a family of transcription factors required for the development of cells of lymphoid origin. Notch-1, a molecule known to regulate cell fate decisions, is also required at the earliest stage of T cell lineage development.⁶ Cytokines including IL-7 promote the survival and expansion of the earliest thymocytes. In mice deficient for IL-7, its receptor components IL-7Rα or γ_c, or the cytokine receptor–associated signaling molecule JAK-3, thymocyte development is inhibited at the CD25⁻CD44⁺ stage. In humans, mutations in γ_c or JAK-3 result in the most frequent form of SCID.⁷ Pre-TCR signaling is required for the CD25⁺CD44⁻ → CD25⁻CD44⁻ transition. Thus loss of signaling components including Lck, SLP-76, and LAT-1 results in a block at this stage of T cell development. TCR signals are also required for differentiation of CD4⁺CD8⁺ to CD4⁺ or CD8⁺ cells. Humans deficient in ZAP-70 (see later) have CD4⁺ but not CD8⁺ T cells in the thymus and periphery.⁸

Figure 13-2 Sequence of αβ T cell development in the thymus. The earliest thymocyte precursors lack expression of CD4 and CD8 (CD4⁻CD8⁻). These can be further divided into four subpopulations based on the sequential expression of CD25 and CD44. At the CD25⁺CD44⁻ stage, the TCR-β-chain rearranges and associates a surrogate α chain known as pre-Tα. Concomitant with a proliferative burst, thymocytes progress to the CD4⁺CD8⁺ stage, rearrange the TCR-α-chain, and express a mature TCR complex. These cells then undergo thymic positive and negative selection. Those thymocytes that survive this rigorous selection process differentiate into mature CD4⁺ or CD8⁺ T cells. Shown also are the various signaling molecules that are involved at specific stages of thymic development.

CD25⁻CD44⁻ cells upregulate expression of CD4 and CD8 to become CD4⁺8⁺. It is as a CD4⁺CD8⁺ thymocyte that the α chain of the TCR rearranges. Unlike the β chain, allelic exclusion of the α chain is not apparent. Rearrangement of the α chain can occur simultaneously on both chromosomes, and if one attempt is unsuccessful, repeat rearrangements to other Vα segments are possible. Reports exist of dual TCR expression by up to 30% of mature T cells in which the same T cell expresses different α chains paired with the same β chain.⁹ However, in most cases of dual TCR α chains, one is downregulated during positive selection by Lck and Cbl, through ubiquitination, endocytosis, and degradation.

Although the structure of immunoglobulin and TCR are quite similar, they recognize fundamentally different antigens. Immunoglobulins recognize intact antigens in isolation, either soluble or membrane bound, and are often sensitive to the tertiary structure. The TCRαβ recognizes linear stretches of antigen peptide fragments bound within the grooves of either major histocompatibility complex (MHC) class I or class II molecules (Figure 13-3A). Thymic selection molds the repertoire of emerging TCR so that they recognize peptides within the groove of self-MHC molecules, ensuring self-MHC restriction of T cell responses. The MHC structure is described in detail in Chapters 19 through 21. Pockets within the MHC groove bind particular residues along the peptide sequence of 7 to 9 amino acids for MHC class I and 9 to 15 amino acids for MHC class II molecules. As a result, depending on the particular MHC molecule, certain amino acids will make strong contact with the MHC groove while others will contact the TCR.

Figure 13-3 T cell antigen receptor (TCR) interaction with the major histocompatibility complex (MHC)–peptide complex. A, Polymorphic residues within the variable region of the α and β chains of the TCR make contact with determinants on the MHC molecule on an antigen-presenting cell (APC), as well as with the peptide fragment that sits in the MHC binding groove. B, Schematic diagram illustrating that during thymocyte development, those TCR conferring either a very low signal intensity (null selection) or high intensity (negative selection) each lead to apoptosis. Only those thymocytes whose TCR can engage MHC peptides and confer moderate intensity survive by positive selection.

The contact between the TCR and MHC/peptide has been revealed by crystal structure to be remarkably flat, rather than a deep lock and key structure one might imagine.¹⁰ The TCR axis is tipped about 30 degrees to the long axis of the MHC class I molecule and is slightly more skewed to MHC class II. The affinity of the TCR for antigen/MHC is in the micromolar range. This is less than many antibody-antigen affinities and is several logs less than many enzyme-substrate affinities. This has led to the notion that TCR interactions with antigen/MHC are brief and that successful activation of the T cell requires multiple interactions, resulting in a cumulative signal.

Once the T cell has successfully rearranged and expressed a TCR in association with the CD3 complex, it encounters the second major hurdle in T cell development, thymic selection. Selection has two phases, positive and negative, and the outcome is based largely on the intensity of TCR signaling in response to interactions with MHC/self-peptides expressed on thymic epithelium and dendritic cells. TCR signals that are either too weak (death by neglect) or too intense (negative selection) result in elimination by apoptosis, whereas those with intermediate signaling intensity survive positive selection (Figure 13-3B). Successful positive selection at the CD4⁺8⁺ stage is coincident with upregulation of surface TCR, the activation markers CD5 and CD69, and the survival factor Bcl-2.¹¹ T cells bearing a TCR that recognizes MHC class I maintain CD8 expression, downregulate CD4, and become CD4⁻8⁺. T cells expressing a TCR that recognizes MHC class II become CD4⁺8⁻.

An enigma for thymic selection has been how to present the myriad self-proteins to developing thymocytes so that self-reactive thymocytes are effectively eliminated by negative selection. This includes particularly those antigens with tissue or developmentally restricted expression. A solution was found with the discovery of the autoimmune regulator (AIRE) gene. AIRE is a transcription factor expressed by the medullary epithelium of the thymus that induces transcription of a wide array of organ-specific genes, such as insulin, that might otherwise be sequestered from developing thymocytes.¹² This effectively creates a self-transcriptome within the thymus against which autoreactive T cells can be deleted. Gene knockout mice of AIRE and humans bearing AIRE mutations manifest various autoimmune sequelae in a syndrome known as autoimmune polyendocrinopathy-candidiasis-ectodermal dystrophy (APECED).¹³

Not surprisingly, a variety of signaling molecules activated by TCR engagement are important to thymic selection. Lck, the RasПRaf-1ПMEK1ПERK kinase cascade, the kinase ZAP-70, and the phosphatases CD45 and calcineurin are involved with positive selection. Among these the Ras ПERK pathway is particularly important because dominant negative variants of these molecules can disrupt positive selection. Conversely, an activator of Ras known as GRP1 assists the positive selection of thymocytes expressing weakly selecting signals. These molecules are discussed in more detail in the section on TCR signaling. By contrast, although a number of molecules may promote negative selection, among them the MAP kinases JNK and p38, there appears to be sufficient redundancy so that only rarely does elimination of any one of these molecules affect deletion of thymocytes. The few exceptions include CD40, CD40L, CD30, or the pro-apoptotic Bcl-2 family member, Bim, where preservation of at least some thymocytes bearing self-reactive TCR could be observed in mice deficient in these molecules.^14–16

The survivors of these two stringent processes of TCR gene rearrangement and thymic selection represent less than 3% of total immature thymocytes. This is reflected in the presence of a high rate of cell death in developing thymocytes. This can be visualized by the measurement of DNA degradation, a hallmark of apoptosis, as shown in Figure 13-4. The survivors become either CD4⁺ helper or CD8⁺ cytolytic T cells and reside in the thymic medulla for 12 to 14 days before emigrating to the periphery. The decision to become a CD4⁺ versus CD8⁺ T cell involves further developmental signals including once again Notch-1. Notch-1 signaling is required for progression to CD8⁺ but not CD4⁺ thymocytes. This parallels the observation that long TCR interactions are required for CD4 progression, whereas shorter TCR engagement is required for CD8 progression.¹⁷

Figure 13-4 T cell antigen receptor (TCR) signal pathways. Schema showing the principal signal pathways resulting from TCR activation and how they impinge on the regulatory region of the interleukin-2 (IL-2) gene. See text for details.

Peripheral Migration of T Cells

The migration of naïve T cells to peripheral lymphoid structures or their infiltration into other tissues requires the coordinate regulation of an array of cell adhesion molecules. T cell recirculation is essential for host surveillance and is carefully regulated by a specific array of homing receptors. Entry from the circulation to tissues occurs via two main anatomic sites: the flat endothelium of the blood vessels and specialized postcapillary venules known as high endothelial venules (HEV). A three-step model has been proposed for lymphocyte migration: rolling, adhesion, and migration.²⁰ L-selectin expressed by naïve T cells binds via lectin domains to carbohydrate moieties of GlyCAM-1 and CD34 (collectively known as peripheral node addressin) that are expressed on endothelial cells, particularly HEV. The weak binding of CD62L to its ligand mediates a weak adhesion to the vessel wall, which, combined with the force of blood flow, results in rolling of the T cell along the endothelium. The increased cell contact assists the interaction of a second adhesion molecule on lymphocytes, the integrin LFA-1 (CD11a/CD18) with its ligands, ICAM-1 (CD54), and ICAM-2 (CD102). This results in the arrest of rolling and firm attachment. Migration into the extracellular matrix of tissues may involve additional lymphocyte cell surface molecules such as the hyaluronate receptor (CD44) or the integrin α4β7 (CD49d/β7) that binds the mucosal addressin cell adhesion molecule-1 (MAdCAM-1) on endothelium of Peyer’s patches and other endothelial cells.

Other cytokines known as chemokines may contribute to lymphocyte homing. Chemokines are structurally and functionally related to proteins bearing an affinity for heparan sulfate proteoglycan and promote migration of various cell types.²¹ The chemokines RANTES, MIP-1α, MIP-1β, MCP-1, and IL-8 are produced by a number of cell types including endothelium, activated T cells, and monocytes and are present at inflammatory sites such as rheumatoid synovium (see Chapter 70).

Once mature T cells have reached the peripheral lymphoid tissues of lymph node and spleen, they undergo a low level of turnover known as homeostatic proliferation in response to self-peptide/MHC complexes, IL-7, and IL-15.²² This serves to maintain the number of peripheral T cells, which can become accelerated in states of lymphopenia such as following chemotherapy or irradiation.²³ Because homeostatic proliferation is driven by self-MHC/peptides, its acceleration could precipitate an autoimmune syndrome. In this regard, it is of interest that one of the standard models of autoimmunity is day 3 thymectomy, which results in lymphopenia,²⁴ NOD diabetic mice have chronic lymphopenia that contributes to their diabetes,²⁵ and evidence of augmented homeostatic proliferation has been suggested to occur in rheumatoid arthritis.²⁶

Activation of T Cells

T cell activation initiates intracellular signaling cascades that ultimately result in proliferation, effector function, or death, depending on the developmental stage of the cell. To guard against premature or excessive activation, T cells have a requirement of two independent signals for full activation. Signal 1 is an antigen-specific signal provided by the binding of the TCR to antigenic peptide complexed with MHC. Signal 2 is mediated by either cytokines or the engagement of co-stimulatory molecules such as B7.1 (CD80) and B7.2 (CD86) on the antigen-presenting cell (APC). Receiving only signal 1 without co-stimulation results in T cell unresponsiveness or anergy.