Chapter 4: Antigen Recognition by B and T Cells PDF

PART II The recognition of antigen 4 5 6 Antigen Recognition by B-cell and T-cell Receptors The Generation of Lymphocyte Antigen Receptors Antigen Presentation to T Lymphocytes Antigen Recognition by B-cell and T-cell Receptors Innate immune responses are the body’s initial defense against infecti...

PART II The recognition of antigen 4 5 6 Antigen Recognition by B-cell and T-cell Receptors The Generation of Lymphocyte Antigen Receptors Antigen Presentation to T Lymphocytes Antigen Recognition by B-cell and T-cell Receptors Innate immune responses are the body’s initial defense against infection, but these work only to control pathogens that have certain molecular patterns or that induce interferons and other nonspecific defenses. To effectively fight the wide range of pathogens an individual will encounter, the lymphocytes of the adaptive immune system have evolved to recognize a great variety of different antigens from bacteria, viruses, and other disease-causing organisms. An antigen is any molecule or part of a molecule that is specifically recognized by the highly specialized recognition proteins of lymphocytes. On B cells these proteins are the immunoglobulins (Igs), which these cells produce in a vast range of antigen specificities, each B cell producing immunoglobulins of a single specificity (see Section 1-12). A membrane-bound form of immuno­ globulin on the B-cell surface serves as the cell’s receptor for antigen, and is known as the B-cell receptor (BCR). A secreted form of immunoglobulin of the same antigen specificity is the antibody produced by terminally differentiated B cells—plasmablasts and plasma cells. The secretion of antibodies, which bind pathogens or their toxic products in the extracellular spaces of the body (see Fig. 1.25), is the main effector function of B cells in adaptive immunity. Antibodies were the first proteins involved in specific immune recognition to be characterized, and are understood in great detail. The antibody molecule has two separate functions: one is to bind specifically to the pathogen or its products that have elicited the immune response; the other is to recruit other cells and molecules to destroy the pathogen once antibody has bound. For example, binding by antibodies can neutralize viruses and mark pathogens for destruction by phagocytes and complement, as described in Chapters 2 and 3. Recognition and effector functions are structurally separated in the antibody molecule, one part of which specifically binds to the antigen whereas the other engages the elimination mechanisms. The antigen-binding region varies extensively between antibody molecules and is known as the variable region or V region. The variability of antibody molecules allows each antibody to bind a different specific antigen, and the total repertoire of antibodies made by a single individual is large enough to ensure that virtually any structure can be recognized. The region of the antibody molecule that engages the effector functions of the immune system does not vary in the same way and ERRNVPHGLFRVRUJ 4 IN THIS CHAPTER The structure of a typical antibody molecule. The interaction of the antibody molecule with specific antigen. Antigen recognition by T cells. 140 Chapter 4: Antigen Recognition by B-cell and T-cell Receptors is known as the constant region or C region. It comes in five main forms, called isotypes, each of which is specialized for activating different effector mechanisms. The membrane-bound B-cell receptor does not have these effector functions, because the C region remains inserted in the membrane of the B cell. The function of the B-cell receptor is to recognize and bind antigen via the V regions exposed on the surface of the cell, thus transmitting a signal that activates the B cell, leading to clonal expansion and antibody production. To this end, the B-cell receptor is associated with a set of intracellular signaling proteins, which will be described in Chapter 7. Antibodies have become an important class of drug due to their highly specific activities, and we return to discuss their therapeutic uses in Chapter 16. The antigen-recognition molecules of T cells are made solely as membranebound proteins, which are associated with an intracellular signaling complex and function only to signal T cells for activation. These T-cell receptors (TCRs) are related to immunoglobulins both in their protein structure—having both V and C regions—and in the genetic mechanism that produces their great variability, which is discussed in Chapter 5. The T-cell receptor differs from the B-cell receptor in an important way, however: it does not recognize and bind antigen by itself, but instead recognizes short peptide fragments of protein antigens that are presented to them by proteins known as MHC molecules on the surface of host cells. The MHC molecules are transmembrane glycoproteins encoded in the large cluster of genes known as the major histocompatibility complex (MHC). The most striking structural feature of MHC molecules is a cleft in the extracellular face of the molecule in which peptides can be bound. MHC molecules are highly polymorphic—each type of MHC molecule occurs in many different versions—within the population. These are encoded by slightly different versions of individual genes called alleles. Most people are therefore hetero­ zygous for the MHC molecules: that is, they express two different alleles for each type of MHC molecule, thus increasing the range of pathogen-derived peptides and self-peptides that can be bound. T-cell receptors recognize features of both the peptide antigen and the MHC molecule to which it is bound. This introduces an extra dimension to antigen recognition by T cells, known as MHC restriction because any given T-cell receptor is specific for a particular peptide bound to a particular MHC molecule. In this chapter we focus on the structure and antigen-binding properties of immunoglobulins and T-cell receptors. Although B cells and T cells recognize foreign molecules in separate distinct fashions, the receptor molecules they use for this task are very similar in structure. We will see how this basic structure can accommodate great variability in antigen specificity, and how it enables immunoglobulins and T-cell receptors to perform their functions as the antigen-recognition molecules of the adaptive immune response. With this foundation, we will return to discuss the impact of MHC polymorphism on T-cell antigen recognition and T-cell development in Chapters 6 and 8, respectively. The structure of a typical antibody molecule. Antibodies are the secreted form of the B-cell receptor. Because they are soluble and secreted into the blood in large quantities, antibodies are easily obtained and easily studied. For this reason, most of what we know about the B-cell receptor comes from the study of antibodies. Antibody molecules are roughly Y-shaped, as represented in Fig. 4.1 using three different schematic styles. This part of the chapter will explain how this structure is formed and allows the antibody molecule to perform its dual tasks of binding to a wide variety of antigens while also binding to effector molecules ERRNVPHGLFRVRUJ The structure of a typical antibody molecule. Fig. 4.1 Structure of an antibody molecule. In panel a, the X-ray crystallographic structure of an IgG antibody is illustrated as a ribbon diagram of the backbones of the polypeptide chains. The two heavy chains are shown in yellow and purple. The two light chains are both shown in red. Three globular regions form an approximate Y shape. The two antigen-binding sites are at the tips of the arms, which are tethered at their other end to the trunk of the Y by a flexible hinge region. The light-chain variable (VL) and constant region (CL) are indicated. The heavy-chain variable region (VH) and VL together form the antigenbinding site of the antibody. In panel b, a schematic representation of the same structure denotes each immunoglobulin domain as a separate rectangle. The hinge that tethers each heavy chain’s first constant domain (CH1) to its second (CH2) is illustrated by a thin purple or yellow line, respectively. The antibody-binding sites are indicated by concave regions in VL and VH. Positions of carbohydrate modifications and disulfide linkages are indicated. In panel c, a more simplified schematic is shown that will be used throughout this book with the variable region in red and the constant region in blue. C terminus, carboxy terminus; N terminus, amino terminus. Structure courtesy of R.L. Stanfield and I.A. Wilson. antigen-binding sites VH VH CH1 VL VL CL CH2 CH3 a antigen-binding sites and to cells that destroy the antigen. Each of these tasks is performed by different parts of the molecule. The ends of the two arms of the Y—the V regions—are involved in antigen binding, and they vary in their detailed structure between different antibody molecules. The stem of the Y—the C region—is far less variable and is the part that interacts with effector molecules and cells. There are five different classes of immunoglobulins, distinguished in their being constructed from C regions that have different structures and properties. These are known as immunoglobulin M (IgM), immunoglobulin D (IgD), immunoglobulin G (IgG), immunoglobulin A (IgA), and immunoglobulin E (IgE). All antibodies are constructed in the same way from paired heavy and light polypeptide chains, and the generic term immunoglobulin is used for all such proteins. More subtle differences confined to the V region account for the specificity of antigen binding. We will use the IgG antibody molecule as an example to describe the general structural features of immunoglobulins. 4-1 IgG antibodies consist of four polypeptide chains. IgG antibodies are large molecules with a molecular weight of approximately 150 kDa and are composed of two different kinds of polypeptide chains. One, of approximately 50 kDa, is called the heavy or H chain, and the other, of 25 kDa, is the light or L chain (Fig. 4.2). Each IgG molecule consists of two heavy chains and two light chains. The two heavy chains are linked to each other by disulfide bonds, and each heavy chain is linked to a light chain by a disulfide bond. In any given immunoglobulin molecule, the two heavy chains and the two light chains are identical, giving an antibody molecule two identical antigen-binding sites. This gives the antibody the ability to bind simultaneously to two identical antigens on a surface, thereby increasing the total strength of the interaction, which is called its avidity. The strength of the interaction between a single antigen-binding site and its antigen is called its affinity. Two types of light chains, lambda (λ) and kappa (κ), are found in antibodies. A given immunoglobulin has either κ chains or λ chains, never one of each. No functional difference has been found between antibodies having λ or κ light chains, and either type of light chain can be found in antibodies of any of the five major classes. The ratio of the two types of light chains varies from species to species. In mice, the average κ to λ ratio is 20:1, whereas in humans it is 2:1 and in cattle it is 1:20. The reason for this variation is unknown. Distortions of this ratio can sometimes be used to detect the abnormal proliferation of a Fig. 4.2 Immunoglobulin molecules are composed of two types of protein chains: heavy chains and light chains. Each immunoglobulin molecule is made up of two hinged heavy chains (green) and two light chains (yellow) joined by disulfide bonds so that each heavy chain is linked to a light chain and the two heavy chains are linked together. ERRNVPHGLFRVRUJ VL CL VH hinge CH1 disulfide bonds CH2 carbohydrate CH3 b N terminus variable region VH CH1 VL CL disulfide bonds CH2 constant region CH3 c C terminus Immunobiology | chapter 1 | 04_001 Murphy et al | Ninth edition © Garland Science design by blink studio limited light chain heavy chain Immunobiology | chapter 4 | 04_002 Murphy et al | Ninth edition © Garland Science design by blink studio limited disulfide bonds 141 142 Chapter 4: Antigen Recognition by B-cell and T-cell Receptors B-cell clone, since all progeny of a particular B cell will express an identical light chain. For example, an abnormally high level of λ light chains in a person might indicate the presence of a B-cell tumor that is producing λ chains. The class, and thus the effector function, of an antibody is defined by the structure of its heavy chain. There are five main heavy-chain classes, or isotypes, some of which have several subtypes, and these determine the functional activity of an antibody molecule. The five major classes of immunoglobulin are IgM, IgD, IgG, IgA, and IgE, and their heavy chains are denoted by the lowercase Greek letters μ, δ, γ, α, and ε, respectively. For example, the constant region of IgM is denoted by Cμ. IgG is by far the most abundant immunoglobulin in serum and has several subclasses (IgG1, 2, 3, and 4 in humans). The distinctive functional properties of the different classes and subclasses of antibodies are conferred by the carboxy-terminal part of the heavy chain, where this chain is not associated with the light chain. The general structural features of all the isotypes are similar, particularly with respect to antigen binding. Here we will consider IgG as a typical antibody molecule, and we will return to discuss the structural and functional properties of the different heavy-chain isotypes in Chapter 5. The structure of a B-cell receptor is identical to that of its corresponding antibody except for a small portion of the carboxy terminus of the heavy-chain C region. In the B-cell receptor, the carboxy terminus is a hydrophobic amino acid sequence that anchors the molecule in the membrane, and in the antibody it is a hydrophilic sequence that allows secretion. 4-2 Immunoglobulin heavy and light chains are composed of constant and variable regions. The amino acid sequences of many immunoglobulin heavy and light chains have been determined and reveal two important features of antibody molecules. First, each chain consists of a series of similar, although not identical, sequences, each about 110 amino acids long. Each of these repeats corresponds to a discrete, compactly folded region of protein known as an immuno­globulin domain, or Ig domain. The light chain consists of two Ig domains, whereas the heavy chain of the IgG antibody contains four Ig domains (see Fig. 4.2). This suggests that the immunoglobulin chains have evolved by repeated duplication of ancestral gene segments corresponding to a single Ig domain. The second important feature is that the amino-terminal amino acid sequences of the heavy and light chains vary greatly between different antibodies. The variability is limited to approximately the first 110 amino acids, corresponding to the first Ig domain, whereas the remaining domains are constant between immunoglobulin chains of the same isotype. The amino-terminal variable Ig domains (V domains) of the heavy and light chains (VH and VL, respectively) together make up the V region of the antibody and determine its antigenbinding specificity, whereas the constant Ig domains (C domains) of the heavy and light chains (CH and CL, respectively) make up the C region (see Fig. 4.1). The multiple heavy-chain C domains are numbered from the aminoterminal end to the carboxy terminus, for example, CH1, CH2, and so on. 4-3 The domains of an immunoglobulin molecule have similar structures. Immunoglobulin heavy and light chains are composed of a series of Ig domains that have a similar overall structure. Within this basic structure, there are distinct differences between V and C domains that are illustrated for the light chain in Fig. 4.3. Each V or C domain is constructed from two β sheets. A β sheet is ERRNVPHGLFRVRUJ The structure of a typical antibody molecule. Light-chain V domain Light-chain C domain amino terminus carboxy terminus β strands β strands disulfide bond Arrangement of β strands disulfide bond D E disulfide bond B A G F C D E B A G F C C´ C´´ Immunobiology | chapter 4 | 04_005 Murphy et al | Ninth edition built several strands, which are regions of protein where several conScience design by blink β studio limited © Garlandfrom secutive polypeptides have their peptide backbone bonds arranged in an extended, or flat, conformation. β strands in proteins are sometimes shown as ‘ribbons with an arrow’ to indicate the direction of the polypeptide backbone (see Fig. 4.3). β strands can pack together in a side-by-side manner, being stabilized laterally by two or three backbone hydrogen bonds between adjacent strands. This arrangement is called a β sheet. The Ig domain has two β sheets that are folded onto each other, like two pieces of bread, into a structure called a β sandwich, and are covalently linked by a disulfide bond between cysteine residues from each β sheet. This distinctive structure is known as the immuno­ globulin fold. The similarities and differences between V and C domains can be seen in the bottom panels of Fig. 4.3. Here, the Ig domains have been opened out to show how their respective polypeptide chains fold to create each of the β sheets and how each polypeptide chain forms flexible loops between adjacent β strands as it turns to change direction. The main difference between the V and C domains is that the V domain is larger and contains extra β strands, called Cʹ and Cʹʹ. In the V domain, the flexible loops formed between some of the β strands contribute to the antigen-binding site of the immunoglobulin molecule. Many of the amino acids that are common to the C and V domains are present in the core of the immunoglobulin fold and are essential for its stability. Other proteins with sequences similar to those of immunoglobulins have been ERRNVPHGLFRVRUJ Fig. 4.3 The structure of immunoglobulin constant and variable domains. The upper panels show schematically the folding pattern of the constant (C) and variable (V) domains of an immunoglobulin light chain. Each domain is a barrel-shaped structure in which strands of polypeptide chain (β strands) running in opposite directions (antiparallel) pack together to form two β sheets (shown in yellow and green for the C domain and red and blue for the V domain), which are held together by a disulfide bond. The way in which the polypeptide chain folds to give the final structure can be seen more clearly when the sheets are opened out, as shown in the lower panels. The β strands are lettered sequentially with respect to the order of their occurrence in the amino acid sequence of the domains; the order in each β sheet is characteristic of immunoglobulin domains. The β strands Cʹ and Cʹʹ that are found in the V domains but not in the C domains are indicated by a blue-shaded background. The characteristic four-strand plus three-strand (C-region type domain) or four-strand plus five-strand (V-region type domain) arrangements are typical immunoglobulin superfamily domain building blocks, found in a whole range of other proteins as well as antibodies and T-cell receptors. 143 144 Chapter 4: Antigen Recognition by B-cell and T-cell Receptors Proteolytic cleavage by papain found to have domains with a similar structure, called immunoglobulin-like domains (Ig-like domains). These domains are present in many proteins of the immune system, such as the KIRs expressed by NK cells described in Chapter 3. They are also frequently involved in cell–cell recognition and adhesion, and together with the immunoglobulins and the T-cell receptors, these proteins make up the extensive immunoglobulin superfamily. amino terminus 4-4 carboxy terminus Fab Fab Fc Proteolytic cleavage by pepsin The antibody molecule can readily be cleaved into functionally distinct fragments. When fully assembled, an antibody molecule comprises three equal-sized globular portions, with its two arms joined to its trunk by a flexible stretch of polypeptide chain known as the hinge region (see Fig. 4.1b). Each arm of this Y-shaped structure is formed by the association of a light chain with the aminoterminal half of a heavy chain; the VH domain is paired with the VL domain, and the CH1 domain is paired with the CL domain. The two antigen-binding sites are formed by the paired VH and VL domains at the ends of the two arms of the Y (see Fig. 4.1b). The trunk of the Y is formed by the pairing of the carboxyterminal halves of the two heavy chains. The CH3 domains pair with each other but the CH2 domains do not interact. Carbohydrate side chains attached to the CH2 domains lie between the two heavy chains. Proteolytic enzymes (proteases) were an important tool in early studies of antibody structure, and it is valuable to review the terminology they generated. Limited digestion with the protease papain cleaves antibody molecules into three fragments (Fig. 4.4). Papain cuts the antibody molecule on the aminoterminal side of the disulfide bonds that link the two heavy chains, releasing the two arms of the antibody molecule as two identical fragments that contain the antigen-binding activity. These are called the Fab fragments, for fragment antigen binding. The other fragment contains no antigen-binding activity, but because it crystallized readily, it was named the Fc fragment (fragment crystallizable). It corresponds to the paired CH2 and CH3 domains. The Fc fragment is the part of the antibody molecule that does not interact with antigen, but rather interacts with effector molecules and cells, and it differs between heavy-chain isotypes. Another protease, pepsin, cuts on the carboxy-terminal side of the disulfide bonds (see Fig. 4.4). This produces a fragment, the F(abʹ)2 fragment, in which the two antigen-binding arms of the antibody molecule remain linked. Pepsin cuts the remaining part of the heavy chain into several small fragments. The F(abʹ)2 fragment has exactly the same antigen-binding characteristics as the original antibody but is unable to interact with any effector molecule, such as C1q or Fc receptors, and can be used experimentally to separate the antigen-binding functions from the antibody’s other effector functions. Many antibody-related molecules can be constructed using genetic engineering techniques, and many antibodies and antibody-related molecules are being used therapeutically to treat a variety of diseases. We will return to this topic in Chapter 16, where we discuss the various therapeutic uses of anti­ bodies that have been developed over the last two decades. F(ab´ ) 2 pFc´ Immunobiology | chapter 4 | 04_003 Murphy et al | Ninth edition © Garland Science design by blink studio limited Fig. 4.4 The Y-shaped immunoglobulin molecule can be dissected by partial digestion with proteases. Upper panels: papain cleaves the immunoglobulin molecule into three pieces, two Fab fragments and one Fc fragment. The Fab fragment contains the V regions and binds antigen. The Fc fragment is crystallizable and contains C regions. Lower panels: pepsin cleaves immunoglobulin to yield one F(abʹ)2 fragment and many small pieces of the Fc fragment, the largest of which is called the pFcʹ fragment. F(abʹ)2 is written with a prime because it contains a few more amino acids than Fab, including the cysteines that form the disulfide bonds. ERRNVPHGLFRVRUJ The structure of a typical antibody molecule. (Micrograph ×300,000) Angle between arms is 60o Immunobiology | chapter 4 | 04_004 Murphy | Ninthhinge edition region 4-5 et alThe of the immunoglobulin molecule allows flexibility in binding to multiple antigens. © Garland Science design by blink studio limited The hinge region between the Fc and Fab portions of the IgG molecule allows for some degree of independent movement of the two Fab arms. For example, in the antibody molecule shown in Fig. 4.1a, not only are the two hinge regions clearly bent differently, but the angle between the V and C domains in each of the two Fab arms is also different. This range of motion has led to the junction between the V and C domains being referred to as a ‘molecular ball-and-socket joint.’ This flexibility can be revealed by studies of antibodies bound to small antigens known as haptens. These are molecules of various types that are typically about the size of a tyrosine side chain. Although haptens are specifically recognized by antibody, they can stimulate the production of anti-hapten antibodies only when linked to a protein (see Appendix I, Section A-1). Two identical hapten molecules joined by a short flexible region can link two or more anti-hapten antibodies, forming dimers, trimers, tetra­ mers, and so on, which can be seen by electron microscopy (Fig. 4.5). The shapes formed by these complexes show that antibody molecules are flexible at the hinge region. Some flexibility is also found at the junction between the V and C domains, allowing bending and rotation of the V domain relative to the C domain. Flexibility at both the hinge and the V–C junction enables the two arms of an antibody molecule to bind to sites some distance apart, such as the repeating sites on bacterial cell-wall polysaccharides. Flexibility at the hinge also enables antibodies to interact with the antibody-binding proteins that mediate immune effector mechanisms. Summary. The IgG antibody molecule is made up of four polypeptide chains, comprising two identical light chains and two identical heavy chains, and can be thought of as forming a flexible Y-shaped structure. Each of the four chains has a variable (V) region at its amino terminus, which contributes to the antigen-binding site, and a constant (C) region. The light chains are bound to the heavy chains by many noncovalent interactions and by disulfide bonds, and the V regions of the heavy and light chains pair in each arm of the Y to generate two identical antigen-binding sites, which lie at the tips of the arms of the Y. The possession of two antigen-binding sites allows antibody molecules to cross-link antigens and to bind them much more stably and with higher avidity. The trunk of the Y, also called the Fc fragment, is composed of the carboxy-terminal domains of the heavy chains, and it is these domains that determine the antibody’s isotype. Joining the arms of the Y to the trunk are the flexible hinge regions. The Fc fragment and hinge regions differ in antibodies of different isotypes. ERRNVPHGLFRVRUJ Angle between arms is 90o Fig. 4.5 Antibody arms are joined by a flexible hinge. An antigen consisting of two hapten molecules (red balls in diagrams) that can cross-link two antigen-binding sites is used to create antigen:antibody complexes, which can be seen in the electron micrograph. Linear, triangular, and square forms are seen, with short projections or spikes. Limited pepsin digestion removes these spikes (not shown in the figure), which therefore correspond to the Fc portion of the antibody; the F(abʹ)2 pieces remain cross-linked by antigen. The interpretation of some of the complexes is shown in the diagrams. The angle between the arms of the antibody molecules varies. In the triangular forms, this angle is 60°, whereas it is 90° in the square forms, showing that the connections between the arms are flexible. Photograph courtesy of N.M. Green. 145 Chapter 4: Antigen Recognition by B-cell and T-cell Receptors Different isotypes have different properties and therefore differ in their interactions with effector molecules and cell types. However, the overall organization of the domains is similar in all isotypes. The interaction of the antibody molecule with specific antigen. In this part of the chapter we look at the antigen-binding site of an immunoglobulin molecule in more detail. We discuss the different ways in which antigens can bind to antibody, and address the question of how variation in the sequences of the antibody V domains determines the specificity for antigen. 4-6 Localized regions of hypervariable sequence form the antigenbinding site. The V regions of any given antibody molecule differ from those of every other. Sequence variability is not, however, distributed evenly throughout the V region but is concentrated in certain segments, as is clearly seen in a variability plot (Fig. 4.6), in which the amino acid sequences of many different antibody V regions are compared. Three particularly variable segments can be identified in both the VH and VL domains. They are designated hypervariable regions and are denoted HV1, HV2, and HV3. In the heavy chains they are located at residues 30 to 36, 49 to 65, and 95 to 103, respectively, while in the light chains they are located at residues 28 to 35, 49 to 59, and 92 to 103, respectively. The most variable part of the domain is in the HV3 region. The regions between the hypervariable regions comprise the rest of the V domain; they show less variability and are termed the framework regions. There are four such regions in each V domain, designated FR1, FR2, FR3, and FR4. The framework regions form the β sheets that provide the structural framework of the immunoglobulin domain. The hypervariable sequences correspond to three loops and are positioned near one another in the folded domain at the outer edge of the β sandwich (Fig. 4.7). Thus, not only is diversity concentrated in particular parts of the V domain sequence, but it is also localized to a particular Heavy-chain V region Light-chain V region Variability Fig. 4.6 There are discrete regions of hypervariability in V domains. The hypervariability regions of both the heavy and the light chain contribute to antigen binding of an antibody molecule. A variability plot derived from comparison of the amino acid sequences of several dozen heavy-chain and light-chain V domains is shown. At each amino acid position, the degree of variability is the ratio of the number of different amino acids seen in all of the sequences together to the frequency of the most common amino acid. Three hypervariable regions (HV1, HV2, and HV3) are indicated in red. They are flanked by less variable framework regions (FR1, FR2, FR3, and FR4, shown in blue or yellow). Variability 146 100 80 60 50 40 30 40 20 20 10 0 0 20 FR1 40 HV1 FR2 60 HV2 100 80 FR3 Immunobiology | chapter 4 | 04_006 Murphy et al | Ninth edition HV3 ERRNVPHGLFRVRUJ © Garland Science design by blink studio limited 120 Residue FR4 0 0 20 FR1 40 HV1 FR2 60 HV2 80 FR3 100 120 Residue HV3 FR4 The interaction of the antibody molecule with specific antigen. Fig. 4.7 The hypervariable regions lie in discrete loops of the folded structure. First panel: the hypervariable regions (red) are positioned on the structure of a map of the coding region of the V domain. Second panel: when shown as a flattened ribbon diagram, hypervariable regions are seen to occur in loops (red) that join particular β strands. Third panel: in the folded structure of the V domain, these loops (red) are brought together to form antigen-binding regions. Fourth panel: in a complete antibody molecule, the pairing of a heavy chain and a light chain brings together the hypervariable loops from each chain to create a single hypervariable surface, which forms the antigen-binding site at the tip of each arm. Because they are complementary to the antigen surface, the hypervariable regions are also commonly known as the complementarity-determining regions (CDRs). C, carboxy terminus; N, amino terminus. Light-chain V region Variability 50 40 30 20 10 region on the surface of the molecule. When the VH and VL immunoglobulin domains are paired in the antibody molecule, the three hypervariable loops from each domain are brought together, creating a single hypervariable site at the tip of each arm of the molecule. This is the antigen-binding site, or antibody-combining site, which determines the antigen specificity of the antibody. These six hypervariable loops are more commonly termed the complementarity-determining regions, or CDRs, because the surface they form is complementary to that of the antigen they bind. There are three CDRs from each of the heavy and light chains, namely, CDR1, CDR2, and CDR3. In most cases, CDRs from both VH and VL domains contribute to the antigenbinding site; thus it is the combination of the heavy and the light chain that usually determines the final antigen specificity (see Fig. 4.6). However, there are some Fab crystal structures that show antigen interaction with just the heavy chain; for example, in one influenza Fab, antigen interaction involves binding mostly to the VH CDR3, and only minor contacts with other CDRs. Thus, one way in which the immune system is able to generate antibodies of different specificities is by generating different combinations of heavy-chain and light-chain V regions. This is known as combinatorial diversity; we will encounter a second form of combinatorial diversity in Chapter 5 when we consider how the genes encoding the heavy-chain and light-chain V regions are created from smaller segments of DNA during the development of B cells in the bone marrow. 4-7 Antibodies bind antigens via contacts in CDRs that are complementary to the size and shape of the antigen. In early investigations of antigen binding to antibodies, the only available sources of large quantities of a single type of antibody molecule were tumors of antibody-secreting cells. The antigen specificities of these antibodies were unknown, and therefore many compounds had to be screened to identify ligands that could be used to study antigen binding. In general, the substances found to bind to these antibodies were haptens (see Section 4-5) such as phosphocholine or vitamin K1. Structural analysis of complexes of antibodies with their hapten ligands provided the first direct evidence that the hypervariable regions form the antigen-binding site, and demonstrated the structural basis of specificity for the hapten. Subsequently, with the discovery of methods of generating monoclonal antibodies (see Appendix I, Section A-7), it became possible to make large amounts of pure antibody specific for a given antigen. This has provided a more general picture of how antibodies interact with their antigens, confirming and extending the view of antibody–antigen interactions derived from the study of haptens. The surface of the antibody molecule formed by the juxtaposition of the CDRs of the heavy and light chains is the site to which an antigen binds. The amino acid sequences of the CDRs are different in different antibodies, and so too are the shapes and properties of the surfaces created by these CDRs. As a general principle, antibodies bind ligands whose surfaces are complementary to that of the antigen-binding site. A small antigen, such as a hapten or a short ERRNVPHGLFRVRUJ 0 0 20 FR1 40 60 HV1 FR2 HV2 80 FR3 100 Residue HV3 FR4 N C N HV3 (CDR3) HV1 (CDR1) C HV2 (CDR2) antigenbinding site Immunobiology | chapter 4 | 04_007 Murphy et al | Ninth edition © Garland Science design by blink studio limited 147 148 Chapter 4: Antigen Recognition by B-cell and T-cell Receptors Fig. 4.8 Antigens can bind in pockets, or grooves, or on extended surfaces in the binding sites of antibodies. The panels in the top row show schematic representations of the different types of binding sites in a Fab fragment of an antibody: first panel, pocket; second panel, groove; third panel, extended surface; and fourth panel, protruding surface. Below are examples of each type. Panel a: the top image shows the molecular surface of the interaction of a small hapten with the complementarity-determining regions (CDRs) of a Fab fragment as viewed looking into the antigen-binding site. The ferrocene hapten, shown in red, is bound into the antigen-binding pocket (yellow). In the bottom image (and in those of panels b, c, and d), the molecule has been rotated by about 90° to give a side-on view of the binding site. Panel b: in a complex of an antibody with a peptide from the human immunodeficiency virus (HIV), the peptide (red) binds along a groove (yellow) formed between the heavy-chain and light-chain V domains. Panel c: shown is a complex between hen egg-white lysozyme and the Fab fragment of its corresponding antibody (HyHel5). The surface on the antibody that comes into contact with the lysozyme is colored yellow. All six CDRs of the antigenbinding site are involved in the binding. Panel d: an antibody molecule against the HIV gp120 antigen has an elongated CDR3 loop (arrow) that protrudes into a recess on the side of the antigen. The structure of the complex between this antibody and gp120 has been solved, and in this case only the heavy chain interacts with gp120. Structures courtesy of R.L. Stanfield and I.A. Wilson. VH a VL b c d Immunobiology | chapter 4 | 04_008 Murphy et al | Ninth edition peptide, generally binds design by blink studio limitedin a pocket or groove lying between the heavy-chain © Garland Science and light-chain V domains (Fig. 4.8a and b). Some antigens, such as proteins, can be the same size as, or larger than, the antibody itself. In these cases, the interface between antigen and antibody is often an extended surface that involves all the CDRs and, in some cases, part of the framework region as well (see Fig. 4.8c). This surface need not be concave but can be flat, undulating, or even convex. In some cases, antibody molecules with elongated CDR3 loops can protrude a ‘finger’ into recesses in the surface of the antigen, as shown in Fig. 4.8d, where an antibody binding to the HIV gp120 antigen projects a long loop into its target. 4-8 Antibodies bind to conformational shapes on the surfaces of antigens using a variety of noncovalent forces. The biological function of antibodies is to bind to pathogens and their products, and to facilitate their removal from the body. An antibody generally recognizes only a small region on the surface of a large molecule such as a polysaccharide or protein. The structure recognized by an antibody is called an antigenic determinant or epitope. Some of the most important pathogens have polysaccharide coats, and antibodies that recognize epitopes formed by the sugar subunits of these molecules are essential in providing immune protection against such pathogens. In many cases, however, the antigens that provoke an immune response are proteins. For example, many protective antibodies against viruses recognize viral coat proteins. In all such cases, the structures recognized by the antibody are located on the surface of the protein. Such sites are likely to be composed of amino acids from different parts of the polypeptide chain that have been brought together by protein folding. Antigenic determinants of this kind are known as conformational or discontinuous epitopes because the structure recognized is composed of segments of the protein that are discontinuous in the amino acid sequence of the antigen but are brought together in the three-dimensional structure. In contrast, an epitope composed of a single segment of polypeptide chain is termed a continuous or linear epitope. Although most antibodies raised against intact, fully folded proteins recognize discontinuous epitopes, some will bind to peptide fragments of the protein. Conversely, antibodies raised against peptides of a protein or against synthetic peptides corresponding to part of its sequence ERRNVPHGLFRVRUJ The interaction of the antibody molecule with specific antigen. are occasionally found to bind to the natural folded protein. This makes it possible, in some cases, to use synthetic peptides in vaccines that aim to raise antibodies against a pathogen’s protein. The interaction between an antibody and its antigen can be disrupted by high salt concentrations, by extremes of pH, by detergents, and sometimes by competition with high concentrations of the pure epitope itself. The binding is therefore a reversible noncovalent interaction. The forces, or bonds, involved in these noncovalent interactions are outlined in Fig. 4.9. Electrostatic interactions occur between charged amino acid side chains, as in salt bridges. Most antibody–antigen interactions involve at least one electrostatic interaction. Interactions also occur between electric dipoles, as in hydrogen bonds, or can involve short-range van der Waals forces. High salt concentrations and extremes of pH disrupt antigen–antibody binding by weakening electrostatic interactions and/or hydrogen bonds. This principle is employed in the purification of antigens by using affinity columns of immobilized antibodies (or in the purification of antibody by using antigens in a like manner) (see Appendix I, Section A-3). Hydrophobic interactions occur when two hydrophobic surfaces come together to exclude water. The strength of a hydrophobic interaction is proportional to the surface area that is hidden from water, and for some antigens, hydrophobic interactions probably account for most of the binding energy. In some cases, water molecules are trapped in pockets in the interface between antigen and antibody. These trapped water molecules, especially those between polar amino acid residues, may also contribute to binding and hence to the specificity of the antibody. The contribution of each of these forces to the overall interaction depends on the particular antibody and antigen involved. A striking difference between antibody interactions with protein antigens and most other natural protein– protein interactions is that antibodies often have many aromatic amino acids Noncovalent forces Origin Electrostatic forces Attraction between opposite charges Hydrogen bonds Hydrogen shared between electronegative atoms (N, O) Van der Waals forces Fluctuations in electron clouds around molecules polarize neighboring atoms oppositely δ+ δ– δ– δ+ H O δ+ H Hydrophobic forces Hydrophobic groups interact unfavorably with water and tend to pack together to exclude water molecules. The attraction also involves van der Waals forces NH 3 N δ– OOC O C δ– H δ+ H H H δ– O H δ+ δ– O O H H Na+ Cation-pi interaction Non-covalent interaction between a cation and an electron cloud of a nearby aromatic group δ– H δ– H δ– H Immunobiology | chapter 4 | 04_009 Murphy et al | Ninth edition © Garland Science design by blink studio limited δ– δ–H δ– H H ERRNVPHGLFRVRUJ Fig. 4.9 The noncovalent forces that hold together the antigen:antibody complex. Partial charges found in electric dipoles are shown as δ+ or δ–. Electrostatic forces diminish as the inverse square of the distance separating the charges, whereas van der Waals forces, which are more numerous in most antigen–antibody contacts, fall off as the sixth power of the separation and therefore operate only over very short ranges. Covalent bonds never occur between antigens and naturally produced antibodies. 149 150 Chapter 4: Antigen Recognition by B-cell and T-cell Receptors in their antigen-binding sites. These amino acids participate mainly in van der Waals and hydrophobic interactions, and sometimes in hydrogen bonds and pi-cation interactions. Tyrosine, for example, can take part in both hydrogen bonding and hydrophobic interactions; it is therefore particularly suitable for providing diversity in antigen recognition and is overrepresented in antigen-binding sites. In general, the hydrophobic and van der Waals forces operate over very short ranges and serve to pull together two surfaces that are complementary in shape: hills on one surface must fit into valleys on the other for good binding to occur. In contrast, electrostatic interactions between charged side chains, and hydrogen bonds bridging oxygen and/or nitrogen atoms, accommodate more specific chemical interactions while strengthening the interaction overall. The side chains of aromatic amino acids such as tyrosine can interact noncovalently through their pi-electron system with nearby cations, including nitrogen-containing side chains that may be in a protonated cationic state. 4-9 VH D1.3 VL Gln121 HEL VL Phe91 VL Tyr32 VH Tyr101 Gln121 VL Trp92 VL Ser93 HEL Asp119 HEL Arg125 Antibody interaction with intact antigens is influenced by steric constraints. An example of an antibody–antigen interaction involving a specific amino acid in the antigen can be seen in the complex of hen egg-white lysozyme with the antibody D1.3 (Fig. 4.10). In this structure, strong hydrogen bonds are formed between the antibody and a particular glutamine in the lysozyme mole­cule that protrudes between the VH and VL domains. Lysozymes from partridge and turkey have another amino acid in place of the glutamine and do not bind to this antibody. In the high-affinity complex of hen egg-white lysozyme with another antibody, HyHel5 (see Fig. 4.8c), two salt bridges between two basic arginines on the surface of the lysozyme interact with two glutamic acids, one each from the VH CDR1 and CDR2 loops. Lysozymes that lack one of the two arginine residues show a 1000-fold decrease in affinity for HyHel5. Overall surface complementarity must have an important role in antigen–antibody interactions, but in most antibodies that have been studied at this level of detail, only a few residues make a major contribution to the binding energy and hence to the final specificity of the antibody. Although many antibodies naturally bind their ligands with high affinity, in the nanomolar range, genetic engineering by site-directed mutagenesis can tailor an antibody to bind even more strongly to its epitope. Even when antibodies have high affinity for antigens on a larger structure, such as an intact viral particle, antibody binding may be prevented by their particular arrangement. For example, the intact West Nile virion is built from an icosahedral scaffold that has 90 homodimers of a membrane-anchored envelope glycoprotein, E, which has three domains, DI, DII, and DIII. The DIII domain has four polypeptide loops that protrude outward from the viral particle. A neutralizing antibody against West Nile virus, E16, recognizes these loops of DIII, as shown in Fig. 4.11. In theory, there should be 180 possible antigen-binding sites for the E16 antibody on the West Nile viral particle. However, a combination of crystallographic and electron micrographic studies show that even with an excess of the E16 Fab fragment, only about 120 of the total 180 DIII domains of E are able to bind E16 Fab fragment (see Fig. 4.11). Fig. 4.10 The complex of lysozyme with the antibody D1.3. Top panel: The interaction of the Fab fragment of D1.3 with hen egg-white lysozyme (HEL) is shown. HEL is depicted in yellow, the heavy chain (VH) in turquoise, and the light chain (VL) in green. Bottom panel: A glutamine residue (Gln121) that protrudes from HEL (yellow) extends its side chain (shown in red) between the VL (green) and VH (turquoise) domains of the antigen-binding site and makes hydrogen bonds with the hydroxyl group (red dots) of the indicated amino acids of both domains. These hydrogen bonds are important to the antigen–antibody binding. Courtesy of R. Mariuzza and R.J. Poljak. Immunobiology | chapter 4 | 04_100 Murphy et al | Ninth edition © Garland Science design by blink studio limited ERRNVPHGLFRVRUJ The interaction of the antibody molecule with specific antigen. Fig. 4.11 Steric hindrance occludes the binding of antibody to native antigen in the intact West Nile viral particle. Top panel: the monoclonal antibody E16 recognizes DIII, one of the three structural domains in the West Nile virus glycoprotein E. Shown is a crystal structure of the E16 Fab bound to the DIII epitope. Bottom left panel: a computer model was used to dock E16 Fab to the mature West Nile virion. E16 Fabs were able to bind 120 of the 180 DIII epitopes. Sixty of the five-fold clustered DIII epitopes are sterically hindered by the binding of Fab to four nearby DIII epitopes. An example of an occluded epitope is the blue area indicated by the arrow. Bottom right panel: cryogenic electron microscopic reconstruction of saturating E16 Fab bound to West Nile virion confirmed the predicted steric hindrance. The vertices of the triangle shown in the figure indicate the icosahedral symmetry axes. This appears to result from steric hindrance, with the presence of one Fab blocking the ability of another Fab to bind to some nearby E protein sites. Presumably, such steric hindrance would become more severe with intact antibody than is evident with the smaller Fab fragment. This study also showed that the Fab bound to the DIII region using only one of its antigen-binding arms, indicating that antibodies may not always bind to antigens with both antigen-binding sites, depending on the orientation of the antigens being recognized. These constraints will impact the ability of antibodies to neutralize their targets. E16 Fab binds four outward-facing loops of WNV DIII envelope protein CL CH E16 Fab VH VL WNV envelope DIII 4-10 Some species generate antibodies with alternative structures. Our focus in this chapter has been on the structure of antibodies produced by humans, which is generally similar in most mammalian species, including mice, an important model system for immunology research. However, some mammals have the ability to produce an alternative form of antibody that is based on the ability of a single VH domain to interact with antigen in the absence of a VL domain (Fig. 4.12). It has been known for some time that the serum of camels contained abundant immunoglobulin-like material composed of heavy-chain dimers that lack associated light chains but retain the capacity to bind antigens. These antibodies are called heavy-chain-only IgGs (hcIgGs). This property is shared by other camelids, including llamas and alpacas. These species have retained the genes for the immunoglobulin light chains, and some IgG-like material in their sera remains associated with light chains, and so it is unclear what led to this particular adaptation during their evolution. In camelids, the ability to produce hcIgGs arises from mutations that allow the alternative splicing of the heavy-chain mRNA, with loss of the CH1 exon and thus the joining of the VH directly to the CH2 domain in the protein. Other mutations stabilize this structure in the absence of VL domains. Molecular model of E16 Fab bound to mature WNV particle Cryo-EM reconstruction of E16 Fab bound to mature WNV particle Immunobiology | chapter 4 | 04_101 Murphy et al | Ninth edition © Garland Science design by blink studio limited Cartilaginous fish, in particular the shark, also have an antibody molecule that differs substantially from human or murine antibodies (see Fig. 4.12). Like camelids, the shark also has genes encoding both immunoglobulin heavy and light chains, and does produce immunoglobulins containing both Human IgG Camelid IgG VH CH1 VL CL Shark IgNAR CH1 VH CH2 CH3 CH2 CH2 CH4 CH3 C H3 CH5 Immunobiology | chapter 4 | 04_102 Murphy et al | Ninth edition © Garland Science design by blink studio limited ERRNVPHGLFRVRUJ Fig. 4.12 Camelid and shark antibody can consist of heavy chain only. In the camelid heavy-chain-only antibody, a splicing event of the mature heavy chain can delete the exon encoding the CH1 region and thereby create an in‑frame hinge region linking the VH1 to the CH2 region. In the shark, the heavy-chainonly Ig molecule retains the CH1 region, suggesting that this form of antibody may predate the evolution of light chains. For both, the repertoire of antigen-binding sites involves extensive variations in long CDR3 regions of the VH domain relative to other types of antibody. 151 152 Chapter 4: Antigen Recognition by B-cell and T-cell Receptors heavy and light chains. But sharks also produce an immunoglobulin new antigen receptor (IgNAR), with heavy-chain-only antibody in which the VH is spliced to the CH1 exon, rather than the CH1 exon being spliced out as in camelids. These differences suggest that hcIgG production by camelids and sharks represents an event of convergent evolution. The ability of camelid VH domains to interact efficiently with antigens is the basis for producing so-called single-chain antibody. The simplification of using only a single domain for antigen recognition has prompted recent interest in single-chain monoclonal antibodies as an alternative to standard monoclonal antibodies, which we will discuss more in Chapter 16. Summary. X-ray crystallographic analyses of antigen:antibody complexes have shown that the hypervariable loops (complementarity-determining regions, CDRs) of immunoglobulin V regions determine the binding specificity of an antibody. Contact between an antibody molecule and a protein antigen usually occurs over a broad area of the antibody surface that is complementary to the surface recognized on the antigen. Electrostatic interactions, hydrogen bonds, van der Waals forces, and hydrophobic and pi-cation interactions can all contribute to binding. Depending on the size of the antigen, amino acid side chains in most or all of the CDRs make contact with antigen and determine both the specificity and the affinity of the interaction. Other parts of the V region normally play little part in the direct contact with the antigen, but they provide a stable structural framework for the CDRs and help to determine their position and conformation. Antibodies raised against intact proteins usually bind to the surface of the protein and make contact with residues that are discontinuous in the primary structure of the molecule; they may, however, occasionally bind peptide fragments of the protein, and antibodies raised against peptides derived from a protein can sometimes be used to detect the native protein molecule. Peptides binding to antibodies usually bind in a cleft or pocket between the V regions of the heavy and light chains, where they make specific contact with some, but not necessarily all, of the CDRs. This is also the usual mode of binding for carbohydrate antigens and small molecules such as haptens. Antigen recognition by T cells. In contrast to the immunoglobulins, which interact with pathogens and their toxic products in the extracellular spaces of the body, T cells recognize foreign antigens only when they are displayed on the surface of the body’s own cells. These antigens can derive from pathogens such as viruses or intracellular bacteria, which replicate within cells, or from pathogens or their products that have been internalized by endocytosis from the extracellular fluid. T cells detect the presence of an intracellular pathogen because the infected cells display peptide fragments of the pathogen’s proteins on their surface. These foreign peptides are delivered to the cell surface by specialized hostcell glycoproteins—the MHC molecules. These are encoded in a large cluster of genes that were first identified by their powerful effects on the immune response to transplanted tissues. For that reason, the gene complex was called the major histocompatibility complex (MHC), and the peptide-binding glycoproteins are known as MHC molecules. The recognition of antigen as a small peptide fragment bound to an MHC molecule and displayed at the cell surface is one of the most distinctive features of T cells, and will be the focus of this part of the chapter. How the peptide fragments of antigen are generated and become associated with MHC molecules will be described in Chapter 6. ERRNVPHGLFRVRUJ Antigen recognition by T cells. We describe here the structure and properties of the T-cell receptor (TCR). As might be expected from the T-cell receptors’ function as highly variable antigen-recognition structures, the genes for TCRs are closely related to those for immunoglobulins. There are, however, important differences between T-cell receptors and immunoglobulins, and these differences reflect the special features of antigen recognition by T cells. 4-11 The TCRα:β heterodimer is very similar to a Fab fragment of immunoglobulin. T-cell receptors were first identified by using monoclonal antibodies that bound to a single cloned T-cell line: such antibodies either specifically inhibit antigen recognition by the clone or specifically activate it by mimicking the antigen (see Appendix I, Section A-20). These clonotypic antibodies were then used to show that each T cell bears about 30,000 identical antigen receptors on its surface, each receptor consisting of two different polypeptide chains, termed the T-cell receptor α (TCRα) and β (TCRβ) chains. Each chain of the α:β heterodimer is composed of two Ig domains, and the two chains are linked by a disulfide bond, similar to the structure of the Fab fragment of an immunoglobulin molecule (Fig. 4.13). α:β heterodimers account for antigen recognition by most T cells. A minority of T cells bear an alternative, but structurally similar, receptor made up of a different pair of polypeptide chains designated γ and δ. The γ:δ T-cell receptors seem to have different antigenrecognition properties from the α:β T-cell receptors, and the functions of γ:δ T cells in immune responses are still being clarified as the various ligands they recognize are identified (see Section 6-20). In the rest of this chapter and elsewhere in the book we use the term T-cell receptor to mean the α:β receptor, except where specified otherwise. Both types of T-cell receptors differ from the membrane-bound immunoglobulin that serves as the B-cell receptor in two main ways. A T-cell receptor has only one antigen-binding site, whereas a B-cell receptor has two, and T-cell receptors are never secreted, whereas immunoglobulins can be secreted as antibodies. Further insights into the structure and function of the α:β T-cell receptor came from studies of cloned cDNA encoding the receptor chains. The amino acid sequences predicted from the cDNA showed that both chains of the T-cell receptor have an amino-terminal variable (V) region with sequence homology to an immunoglobulin V domain, a constant (C) region with homology to an immunoglobulin C domain, and a short stalk segment containing a cysteine residue that forms the interchain disulfide bond (Fig. 4.14). Each chain spans the lipid bilayer by a hydrophobic transmembrane domain, and ends in a short cytoplasmic tail. These close similarities of T-cell receptor chains to the heavy and light immunoglobulin chains first enabled prediction of the structural resemblance of the T-cell receptor heterodimer to a Fab fragment of immunoglobulin. antigen-binding site VL antibody VH Fab CL CH Fc antigen-binding site Vα Vβ Cα Cβ T cell Immunobiology | chapter | 04_011 Fig. 4.13 The T-cell4 receptor resembles Murphy et al | Ninth edition Fab fragment. The a membrane-bound © Garland Science design by blink studio limited Fab fragment of an antibody molecule is a disulfide-linked heterodimer, each chain of which contains one immunoglobulin C domain and one V domain; the juxtaposition of the V domains forms the antigen-binding site (see Section 4-6). The T-cell receptor is also a disulfide-linked heterodimer, with each chain containing an immunoglobulin C-like domain and an immunoglobulin V-like domain. As in the Fab fragment, the juxtaposition of the V domains forms the site for antigen recognition. carbohydrate α chain β chain The three-dimensional structure of the T-cell receptor determined by X-ray crystallography in Fig. 4.15a shows that T-cell receptor chains fold in much the same way as the regions comprising the Fab fragment in Fig. 4.1a. Fig. 4.14 Structure of the T-cell receptor. The T-cell receptor heterodimer is composed of two transmembrane glycoprotein chains, α and β. The extracellular portion of each chain consists of two domains, resembl