4-15. The immunoglobulin heavy-chain isotypes are distinguished by the structure of their constant regions

The five main isotypes of immunoglobulin are IgM, IgD, IgG, IgE, and IgA. In humans, IgG antibodies can be further subdivided into four subclasses (IgG1, IgG2, IgG3, and IgG4), whereas IgA antibodies are found as two subclasses (IgA1 and IgA2). The IgG isotypes in humans are named in order of their abundance in serum, with IgG1 being the most abundant. The heavy chains that define these isotypes are designated by the lower-case Greek letters μ, δ, γ, ε, and α, as shown in Fig. 4.16, which also lists the major physical properties of the different human isotypes. IgM forms pentamers in serum, which accounts for its high molecular weight. Secreted IgA can occur as either a monomer or as a dimer.

Figure 4.16

The properties of the human immunoglobulin isotypes. IgM is so called because of its size: although monomeric IgM is only 190 kDa, it normally forms pentamers, known as macroglobulin (hence the M), of very large molecular weight (see Fig. 4.23). IgA dimerizes (more...)

Sequence differences between immunoglobulin heavy chains cause the various isotypes to differ in several characteristic respects. These include the number and location of interchain disulfide bonds, the number of attached oligosaccharide moieties, the number of C domains, and the length of the hinge region (Fig. 4.17). IgM and IgE heavy chains contain an extra C domain that replaces the hinge region found in γ, δ, and α chains. The absence of the hinge region does not imply that IgM and IgE molecules lack flexibility; electron micrographs of IgM molecules binding to ligands show that the Fab arms can bend relative to the Fc portion. However, such a difference in structure may have functional consequences that are not yet characterized. Different isotypes and subtypes also differ in their ability to engage various effector functions, as will be described in Section 4-18.

Figure 4.17

The structural organization of the main human immunoglobulin isotype monomers. Both IgM and IgE lack a hinge region but each contains an extra heavy-chain domain. Note the differences in the numbers and locations of the disulfide bonds (black lines) linking (more...)

4-16. The same V_H exon can associate with different C_H genes in the course of an immune response

The V-region exons expressed by any given B cell are determined during its early differentiation in the bone marrow and, although they may subsequently be modified by somatic hypermutation, no further V(D)J recombination occurs. All the progeny of that B cell will therefore express the same assembled V genes. By contrast, several different C-region genes can be expressed in the B cell's progeny as the cells mature and proliferate in the course of an immune response. Every B cell begins by expressing IgM as its B-cell receptor, and the first antibody produced in an immune response is always IgM. Later in the immune response, however, the same assembled V region may be expressed in IgG, IgA, or IgE antibodies. This change is known as isotype switching. It is stimulated in the course of an immune response by external signals such as cytokines released by T cells or mitogenic signals delivered by pathogens, as we will discuss further in Chapter 9. Here we are concerned with the molecular basis of the isotype switch.

The immunoglobulin C_H genes form a large cluster spanning about 200 kb to the 3′ side of the J_H gene segments (Fig. 4.18). Each C_H gene is split into several exons (not shown in the figure), each corresponding to an individual immunoglobulin domain in the folded C region. The gene encoding the μ C region lies closest to the J_H gene segments, and therefore closest to the assembled V-region exon after DNA rearrangement. A complete μ heavy-chain transcript is produced from the newly rearranged gene. Any J_H gene segments remaining between the assembled V gene and the C_μ gene are removed during RNA processing to generate the mature mRNA. μ heavy chains are therefore the first to be expressed and IgM is the first immunoglobulin isotype to be expressed during B-cell development.

Figure 4.18

The organization of the immunoglobulin heavy-chain C-region genes in mice and humans (not to scale). In humans, the cluster shows evidence of evolutionary duplication of a unit consisting of two γ genes, an ε gene and an α gene. (more...)

Immediately 3′ to the μ gene lies the δ gene, which encodes the C region of the IgD heavy chain. IgD is coexpressed with IgM on the surface of almost all mature B cells, although this isotype is secreted in only small amounts and its function is unknown. Indeed, mice lacking the C_δ exons seem to have essentially normal immune systems. B cells expressing IgM and IgD have not undergone isotype switching, which, as we will see, entails an irreversible change in the DNA. Instead, these cells produce a long primary transcript that is differentially cleaved and spliced to yield one of two distinct mRNA molecules. In one of these, the VDJ exon is linked to the C_μ exons to encode a μ heavy chain, and in the other the VDJ exon is linked to the C_δ exons to encode a δ heavy chain (Fig. 4.19). The differential processing of the long mRNA transcript is developmentally regulated, with immature B cells making mostly the μ transcript and mature B cells making mostly the δ form along with some of the μ transcript.

Figure 4.19

Co-expression of IgD and IgM is regulated by RNA processing. In mature B cells, transcription initiated at the V_H promoter extends through both C_μ and C_δ exons. This long primary transcript is then processed by cleavage and polyadenylation (more...)

Switching to other isotypes occurs only after B cells have been stimulated by antigen. It occurs through a specialized nonhomologous DNA recombination mechanism guided by stretches of repetitive DNA known as switch regions. Switch regions lie in the intron between the J_H gene segments and the C_μ gene, and at equivalent sites upstream of the genes for each of the other heavy-chain isotypes, with the exception of the δ gene (Fig. 4.20, top panel). The μ switch region (S_μ) consists of about 150 repeats of the sequence [(GAGCT)_n (GGGGGT)], where n is usually 3 but can be as many as 7. The sequences of the other switch regions (S_γ, S_α, and S_ε) differ in detail but all contain repeats of the GAGCT and GGGGGT sequences. Exactly how these repetitive sequences promote switch recombination is unclear because the enzyme(s) that promote switch recombination have not been identified; however, it is thought that the repetitive sequences might promote short stretches of homologous alignment that in turn promote DNA recombination.

Figure 4.20

Isotype switching involves recombination between specific switch signals. Repetitive DNA sequences that guide isotype switching are found upstream of each of the immunoglobulin C-region genes, with the exception of the δ gene. The figure illustrates (more...)

When a B cell switches from coexpression of IgM and IgD to expression of an IgG subtype, DNA recombination occurs between S_μ and the S_γ of that IgG subtype. The C_μ and C_δ coding regions are deleted, and γ heavy-chain transcripts are made from the recombined gene. Figure 4.20 (left panels) illustrates switching to γ3 in the mouse. Some of the progeny of this IgG-producing cell may subsequently undergo a further switching event to produce a different isotype, for example IgA, as shown in the bottom panel of Fig. 4.20. Alternatively, as shown in the right panels of Fig. 4.20, the switch recombination may occur between S_μ and one of the switch regions downstream of the C_γ genes so that the cell switches from IgM to IgA or IgE (illustrated for IgA only). All switch recombination events produce genes that can encode a functional protein because the switch sequences lie in introns and therefore cannot cause frame shift mutations.

The enzymes that carry out isotype switching have not been clearly defined. However, we do know that DNA repair enzymes are involved since switching is markedly reduced in Ku protein knockouts; Ku proteins are also essential for the rejoining of DNA during V(D)J joining (see Section 4-5). Recently, it was discovered that deficiency in Activation Induced Cytidine Deaminase completely blocks isotype switching. As mentioned in Section 4-9, this deficiency also blocks somatic hypermutation. Activation Induced Cytidine Deaminise is thought to be an RNA editing enzyme and how it works to enable both hypermutation and switching is unknown at present. A deficiency in this enzyme in humans has now been associated with a form of immuno-deficiency known as Hyper IgM type 2 syndrome, which is characterized by an absence of immunoglobulins other than IgM. A failure of T cells to activate isotype switching leads to a similar syndrome now classified as Hyper IgM type 1 (see Section 11-9). (Hyper IgM Immunodeficiency, in Case Studies in Immunology, see Preface for details)

Isotype switch recombination is unlike V(D)J recombination in several ways. First, all isotype switch recombination is productive; second, it uses different recombination signal sequences and enzymes; third, it happens after antigen stimulation and not during B-cell development in the bone marrow; and fourth, the switching process is not random but is regulated by external signals such as those provided by T cells, as will be discussed in Chapter 9.

4-17. Transmembrane and secreted forms of immunoglobulin are generated from alternative heavy-chain transcripts

Immunoglobulins of all heavy-chain isotypes can be produced either in secreted form or as a membrane-bound receptor. All B cells initially express the transmembrane form of IgM; after antigen stimulation, some of their progeny differentiate into plasma cells producing the secreted form of IgM, whereas others undergo isotype switching to express transmembrane immunoglobulins of a different isotype before switching to the production of secreted antibody of the new isotype.

The membrane forms of all isotypes are monomers comprised of two heavy and two light chains: IgM and IgA polymerize only when they are secreted. In its membrane-bound form the immunoglobulin heavy chain has a hydrophobic transmembrane domain of about 25 amino acid residues at the carboxy terminus, which anchors it to the surface of the B lymphocyte. This transmembrane domain is absent from the secreted form, whose carboxy terminus is a hydrophilic secretory tail. The two different carboxy termini of the transmembrane and secreted forms of immunoglobulin heavy chains are encoded in separate exons and production of the two forms is achieved by alternative RNA processing (Fig. 4.21). The last two exons of each C_H gene contain the sequences encoding the secreted and the transmembrane regions respectively; if the primary transcript is cleaved and polyadenylated at a site downstream of these exons, the sequence encoding the carboxy terminus of the secreted form is removed by splicing and the cell-surface form of immunoglobulin is produced. Alternatively, if the primary transcript is cleaved at the polyadenylation site located before the last two exons, only the secreted molecule can be produced. This differential RNA processing is illustrated for C_μ in Fig. 4.21, but occurs in the same way for all isotypes.

Figure 4.21

Transmembrane and secreted forms of immunoglobulins are derived from the same heavy-chain sequence by alternative RNA processing. Each heavy-chain C gene has two exons (membrane-coding (MC) yellow) that encode the transmembrane region and cytoplasmic (more...)

Although the production of membrane-bound and secreted versions of the heavy chain is achieved by similar mechanisms to those that allow the co-expression of surface IgM and IgD (see Fig. 4.19), these two instances of alternative RNA processing act at different stages in the life of the B cell, and on different primary transcripts. B cells make a long heavy-chain transcript that can be processed to give either transmembrane IgM or IgD before they are stimulated by antigen. A B cell that is activated ceases to coexpress IgD with IgM, either because μ and δ sequences have been removed as a consequence of an isotype switch or, in IgM-secreting plasma cells, because transcription from the V_H promoter no longer extends through the C_δ exons. In activated B cells that differentiate to become antibody-secreting plasma cells, much of the transcript is spliced to the secreted rather than transmembrane form of whichever isotype the B cell happens to express.

4-18. Antibody C regions confer functional specialization

The secreted antibodies protect the body in a variety of ways, as we briefly outline here and discuss further in Chapter 9. In some cases it is enough for the antibody simply to bind antigen. For instance, by binding tightly to a toxin or virus, an antibody can prevent it from recognizing its receptor on a host cell. The V regions on their own are sufficient for this. The C region is essential, however, for recruiting the help of other cells and molecules to destroy and dispose of pathogens to which the antibody has bound, and it confers functionally distinct properties on each of the various isotypes.

The C regions of antibodies have three main effector functions. First, the Fc portions of different isotypes are recognized by specialized receptors expressed by immune effector cells. The Fc portions of IgG1 and IgG3 antibodies are recognized by Fc receptors present on the surface of phagocytic cells such as macrophages and neutrophils, which can thereby bind and engulf pathogens coated with antibodies of these isotypes. The Fc portion of IgE binds to a high-affinity Fcε receptor on mast cells, basophils, and activated eosinophils, enabling these cells to respond to the binding of specific antigen by releasing inflammatory mediators. Second, the Fc portions of antigen: antibody complexes can bind to complement (see Fig. 1.24) and initiate the complement cascade, which helps to recruit and activate phagocytes, can aid the engulfment of microbes by phagocytes, and can also directly destroy pathogens. Third, the Fc portion can deliver antibodies to places they would not reach without active transport. These include the mucus secretions, tears, and milk (IgA), and the fetal blood circulation by transfer from the pregnant mother (IgG). In both cases, the Fc portion engages a specific receptor that leads to the active transport of the immunoglobulin through cells to reach different body compartments.

The role of the Fc portion in these effector functions can be demonstrated by studying enzymatically treated immunoglobulins that have had one or other domain of the Fc cleaved off (see Section 3-3) or, more recently, by genetic engineering, which permits detailed mapping of the exact amino acid residues within the Fc that are needed for particular functions. Many kinds of microorganism seem to have responded to the destructive potential of the Fc portion by manufacturing proteins that either bind to it or proteolytically cleave it, and so prevent the Fc region from working. Examples of these are Protein A and Protein G made by Staphylococcus species (Fig. 4.22), and Protein D of Haemophilus species. Researchers can exploit these proteins to help to map the Fc and as immunological reagents (see Appendix I, Section A-10). Not all immunoglobulin isotypes or subtypes have the same capacity to engage each of the effector functions. The differential capabilities of each isotype are summarized in Fig. 4.16. For example, IgG1 and IgG3 have higher affinity for the most common type of Fc receptor.

Figure 4.22

Protein A of Staphylococcus aureus bound to a fragment of the Fc region of IgG. A fragment of the Fc portion of a single IgG heavy chain is complexed with a fragment of the immunoglobulin-binding Protein A from Staphylococcus aureus. The Fc fragment has (more...)

4-19. IgM and IgA can form polymers

Although all immunoglobulin molecules are constructed from a basic unit of two heavy and two light chains, both IgM and IgA can form multimers (Fig. 4.23). IgM and IgA C regions contain a ‘tailpiece’ of 18 amino acids that contains a cysteine residue essential for polymerization. An additional separate 15 kDa polypeptide chain called the J chain promotes polymerization by linking to the cysteines of the tailpiece, which is found only in the secreted forms of the μ and α chains. (This J chain should not be confused with the J gene segment; see Section 4-2.) In the case of IgA, polymerization is required for transport through epithelia, as we discuss in Chapter 9. IgM molecules are found as pentamers, and occasionally hexamers (without J chain), in plasma, whereas IgA in mucous secretions, but not in plasma, is mainly found as a dimer (see Fig. 4.23).

Figure 4.23

The IgM and IgA molecules can form multimers. IgM and IgA are usually synthesized as multimers in association with an additional polypeptide chain, the J chain. In pentameric IgM, the monomers are cross-linked by disulfide bonds to each other and to the (more...)

The polymerization of immunoglobulin molecules is thought to be important in the binding of antibody to repetitive epitopes. The dissociation rate of an individual epitope from an individual antibody binding site influences the strength of binding, or affinity, of that site: the lower the dissociation rate, the higher the affinity (see Appendix I, Section A-9). An antibody molecule has two or more identical antigen-binding sites, and if it attaches to two or more repeating epitopes on a single target antigen, it will only dissociate when all sites dissociate. The dissociation rate of the whole antibody from the whole antigen will therefore be much slower than the rate for the individual binding sites, giving a greater effective total binding strength, or avidity. This consideration is particularly relevant for pentameric IgM, which has ten antigen-binding sites. IgM antibodies frequently recognize repetitive epitopes such as those expressed by bacterial cell-wall polysaccharides, but the binding of individual sites is often of low affinity because IgM is made early in immune responses, before somatic hypermutation and affinity maturation. Multisite binding makes up for this, dramatically improving the overall functional binding strength.

4-20. Various differences between immunoglobulins can be detected by antibodies

When an immunoglobulin is used as an antigen, it will be treated like any other foreign protein and will elicit an antibody response. Anti-immunoglobulin antibodies can be made that recognize the amino acids that characterize the isotype of the injected antibody. Such anti-isotypic antibodies recognize all immunoglobulins of the same isotype in all members of the species from which the injected antibody came.

It is also possible to raise antibodies that recognize differences in immuno-globulins from members of the same species that are due to the presence of multiple alleles of the individual C genes in the population (genetic polymorphism). Such allelic variants are called allotypes. In contrast to anti-isotypic antibodies, anti-allotypic antibodies will recognize immuno-globulin of a particular isotype only in some members of a species. Finally, as individual antibodies differ in their V regions, one can raise antibodies against unique sequence variants, which are called idiotypes.

A schematic picture of the differences between idiotypes, allotypes, and isotypes is given in Fig. 4.24. Historically, the main features of immunoglobulins were defined by using isotypic and allotypic genetic markers identified by antisera raised in different species or in genetically distinct members of the same species (see Appendix I, Section A-10). The independent segregation of allotypic and isotypic markers revealed the existence of separate heavy-chain, κ, and λ genes.

Figure 4.24

Different types of variation between immunoglobulins. Differences between constant regions due to usage of different C-region genes are called isotypes; differences due to different alleles of the same C gene are called allotypes; differences due to particular (more...)

Summary

The isotypes of immunoglobulins are defined by their heavy-chain C regions, each isotype being encoded by a separate C-region gene. The heavy-chain C-region genes lie in a cluster 3′ to the V-region gene segments. A productively rearranged V-region exon is initially expressed in association with μ and δ C_H regions, but the same V-region exon can subsequently be associated with any one of the other isotypes by the process of isotype switching, in which the DNA is rearranged to place the V region 5′ to a different C-region gene. Unlike V(D)J recombination, isotype switching is always productive and occurs only in B cells activated by antigen. The immunological functions of the various isotypes differ; thus, isotype switching varies the response to the same antigen at different times or under different conditions. Immunoglobulin RNA can be processed in two different ways to produce either membrane-bound immunoglobulin, which acts as the B-cell receptor for antigen, or secreted antibody. In this way, the B-cell antigen receptor has the same specificity as the antibody that the B cell secretes upon activation.