Chapter 37. Humoral Immunity and Complement

During evolution, jawed vertebrates developed the capacity to respond with exquisite specificity to foreign organisms.1 Specific immunity is characterized by an enormous diversity of possible responses and by refinement in the immune response with successive exposures to the organism.2 The cells that can discriminate with fine specificity through their vast repertoire of receptors are lymphocytes. Specific immunity, also called adaptive immunity because it develops as an adaptation to infection, can be segregated into humoral immunity, mediated by antibodies produced by B lymphocytes, and cellular immunity, mediated by T lymphocytes. These two forms of specific immunity developed to serve different functions. Humoral immunity is directed primarily toward extracellular antigens such as circulating bacteria and toxins. Cellular immunity is directed primarily toward antigens that infect or inhabit cells (see Chapter 10). To combat extracellular pathogens, the defending agent needs to be abundant and widely distributed in the body, particularly at its interfaces with the environment. Antibodies fulfill these characteristics by being capable of being secreted in great quantity from the cells that produce them and by being distributed in blood, mucosa, and interstitial fluid. In addition, antibodies can attach through Fc receptors (FcRs) to the surface of certain other cells of the immune system, such as mast cells, conferring antigen specificity to cells that do not have their own endogenously produced antigen-specific receptors. In addition to their major function in humoral immunity as antibody producers, B lymphocytes have a role in antigen presentation, regulation of T-cell subsets and dendritic cells, organization of lymphoid tissues, and cytokine and chemokine production.3,4

Antibody Structure

Antibodies, or immunoglobulins (Ig), are a family of glycoproteins that share a common structure.2,5,6 The antibody molecule has a symmetric Y-shape consisting of two identical light chains, each about 24 kDa, that are covalently linked to two identical heavy chains, each about 55 or 70 kDa, that are covalently linked to one another (Fig. 37-1). Within the light and heavy chains are variable and constant regions. The major function of the variable region is to recognize antigen, whereas the constant region mediates effector functions. The light and heavy chains contain a series of repeating, homologous units of about 110 amino acids that assume a globular structure and are called Ig domains. The Ig domain motif is found not only in antibody molecules but also in a variety of other molecules of the Ig “superfamily,” including the T-cell receptor, the major histocompatibility complex (MHC), CD4, CD8, intercellular adhesion molecule 1, among other molecules. The light chain has two major domains, (1) a variable (VL) and (2) a constant (CL) domain. The heavy chains have four or five domains, a variable (VH) and three (in IgA, IgD, and IgG) or four (in IgM and IgE) constant (CH1–4) domains. In IgA, IgD, and IgG, there is a hinge region between CH1 and CH2 that confers additional flexibility to the molecule. The variable domains are at the N-terminus. At the C-terminus are the constant domains and, in the heavy chains of membrane-bound antibodies, the transmembrane and cytoplasmic domains.

Within the variable regions of the light and heavy chains are three areas of intense variability called hypervariable regions. These three regions, which are in proximity to one another in the three-dimensional structure of the antibody, are the areas most responsible for binding antigen. Because the hypervariable regions form a shape complementary to that of the antigen, the hypervariable regions are also called the complementarity-determining regions. The unique areas formed by the hypervariable regions are present in too low an amount in the individual to generate self-tolerance. Thus, the immune system may not distinguish the unique portion of the antibody as self and may produce antibodies to that region of the antibody. The area of the antibody capable of generating an immune response is called an idiotope, and antibody responses to idiotopes result in a network of idiotypic–anti-idiotypic interactions that may help regulate the humoral immune response.7

There are two types of light chains, κ and λ, each encoded on different chromosomes. Each antibody molecule has either two κ or two λ chains, never one of each. The functional differences, if any, between κ and λ are not known. There are five types of heavy chains, (1) α, (2) δ, (3) ϵ, (4) γ, and (5) μ, corresponding to the antibody classes IgA, IgD, IgE, IgG, and IgM, respectively. The different heavy chain classes have significantly different functions, as discussed in Section “Antibody Classes”. The IgA and IgG classes contain closely related subclasses, consisting of IgA1 and IgA2, and IgG1, IgG2, IgG3, and IgG4 (Table 37-1).

Enzymatic digestion of IgG molecules by papain results in three cleavage products, two identical Fab fragments consisting of a light chain bound to the V–CH1 region of the heavy chain and an Fc portion consisting of two CH2–CH3 heavy chains bound to each other. Fab was so named for its property of antigen binding, and Fc was so named for its property of crystallizing. When IgG is digested by pepsin, the C-terminal region is digested into small fragments. The remaining product consists of the Fab region along with the hinge region. Fab fragments containing the hinge region are termed Fab′. When the two Fab′ fragments in an antibody molecule remain associated, the fragment is called F(ab′)2.

(See Table 37-1)

Immunoglobulin M

IgM is evolutionarily the most ancient antibody class and is the first Ig molecule to be expressed during B-cell development.1 Its secretory form exists mainly as a pentamer consisting of five IgM molecules joined at their C-termini by tail pieces and stabilized by a molecule called a joining (J) chain. The engagement of membrane-bound IgM by antigen results in the activation of naive B cells. Secreted IgM recognizes antigen, usually through low-affinity interactions, and it can activate complement. IgM is the major effector of the primary antibody response. Although IgM interactions are typically low affinity, IgM can be very effective in responding to a polyvalent antigen (such as a polysaccharide with repeating epitopes) because its pentameric structure allows for multiple low-affinity interactions, resulting in a high-avidity interaction. (Avidity refers to the overall strength of attachment, whereas affinity refers to the strength of attachment at a single antigen-binding site.)

Immunoglobulin D

The IgD molecule exists primarily in a membrane-bound form and is the second antibody class to be expressed during B-cell development. Its function is not completely understood, but in its membrane-bound form it can serve as an antigen receptor for naive B cells.8 Secreted IgD has been found on the surface of basophils, where it induces production of antimicrobial, opsonizing, inflammatory, and B-cell–stimulating factors.9

Immunoglobulin A

IgA is the most abundant Ig in the body, being present in large quantity at mucosal sites. It is responsible for mucosal immunity and is secreted in breast milk, thus contributing to neonatal immunity. In its secreted form, it exists as a monomer, dimer, or trimer, with the multimers being formed by interactions between tail pieces and stabilized by the J chain. For transport across epithelial surfaces, IgA dimers attach to a type of FcR called the polymeric Ig receptor.10 Once the transport process is complete, the IgA dimers remain attached to the extracellular portion of the receptor, called the secretory component, which protects the IgA from proteolysis. Cells of the immune system that have receptors for IgA include neutrophils, eosinophils, and monocytes.

Immunoglobulin G

IgG is the most abundant Ig in the circulation. Its secreted form is a monomer. IgG plays an important role in secondary antibody responses, and its interactions with antigen tend to be high affinity, particularly as the immune response matures. A number of cells have FcRs for IgG, including monocytes, neutrophils, eosinophils, natural killer (NK) cells, and B cells. IgG opsonizes (coats) antigen, allowing phagocytosis of the antigen, and activates complement. An exception is IgG4, which does not activate complement. IgG is important in neonatal immunity, as it is the only Ig class to cross the placenta, and it is secreted in breast milk. The interaction of IgG with the MHC class I-related receptor FcRn is involved in the delivery of IgG across the placenta as well as in prolonging its level in the circulation.11 The serum half-life of IgG is 23 days, considerably longer than that of the other Ig classes.

Immunoglobulin E

IgE is found in very small amounts in the circulation. High-affinity receptors for the Fc portion of IgE are present on mast cells, basophils, and eosinophils, and low-affinity receptors are present on B cells and Langerhans cells. In mast cells and basophils, IgE engagement with antigen activates the cells. IgE mediates immediate hypersensitivity, but its principal protective role may be to combat parasites.

The information encoded by an individual's DNA is limited by the need for the DNA to fit into a package of the size of a cell. This space is far too small for sufficient DNA to encode billions of different lymphocyte receptors if the genes were encoded separately. Lymphocytes have adapted to this limitation by special mechanisms that increase by orders of magnitude the number of different possible antigen receptors.12 Each clone of B cells produces identical antigen receptors (i.e., antibodies) with unique specificity. It is estimated that an individual has approximately 107 different B-cell clones, resulting in 107 distinct antibodies. A major mechanism for generating this enormous diversity is gene rearrangement, whereby segments of DNA within a lymphocyte undergo somatic recombinations.13 Light chain genes contain three regions, (1) V (variable), (2) J (joining), and (3) C (constant), and heavy chain genes contain four regions, (1) V, (2) D (diversity), (3) J, and (4) C. Within each region are many gene segments from which to select for the final antibody product, which is comprised of one gene segment randomly selected from each region. The initial event in antibody formation is the joining of one D and one J segment from a heavy chain gene, with subsequent deletion of the DNA between the two segments. Next, a V segment is selected to join to the DJ segment, and any remaining D segments are deleted. The VDJ complex has attached 3′ to it any remaining J segments plus the C region. The unused J segments are removed during RNA processing. A similar process occurs in light chain loci; because there are no D segments in light chain loci, a VJ rather than a VDJ complex is formed. (Particularly in the k locus, VJ recombination may occur through a somewhat different mechanism involving inversion of the DNA without deletion of intervening sequences, but the functional result is the same.)

The ability to select one segment each from the many segments available in the V, D, and J regions leads to a vast increase in the repertoire of possible antibodies. Additional diversity is generated by the juxtaposition of a rearranged light chain to a rearranged heavy chain; by the addition, deletion, or transposition of nucleotides at the junctions between V and D, D and J, and V and J segments, a phenomenon called junctional diversity; and by somatic hypermutation after antigen stimulation (see below).

Cells destined to become mature B cells undergo an orderly progression of events during development, resulting in the formation sequentially of heavy chains, light chains, and whole antibody molecules, with checkpoints to select against cells making unproductive gene rearrangements or autoreactive antibodies, and survival signals to select for cells making potentially useful antibodies. The process of B-cell development occurs in distinct stages, characterized by specific events and identifiable by specific cell surface markers and Ig gene expression.

Bone marrow and fetal liver stem cells that give rise to B cells are initially pluripotent.2,14 Stem cells developing in the lymphocytic pathway initially become common lymphoid progenitors, which can give rise to B, T, or NK cells. B cells originating from fetal liver are mainly B1 cells (see Section “B-Cell Activation and Antibody Function”), whereas B cells originating in the bone marrow are primarily follicular B cells. Cells and extracellular molecules in the stromal microenvironment provide signals required for differentiation of lymphocytes. Induction of the transcriptional regulators EBF, E2A, and Pax-5 leads to the expression of proteins critical to B-cell development. Posttranscriptional regulation of mRNA by RNA-binding proteins and microRNAs provides further control over the process of B-cell differentiation.15

The earliest cell committed to the B-cell lineage is called a pro-B cell. At the pro-B cell stage, the cell expresses recombination activating gene (RAG) and terminal deoxyribonucleotidyl transferase (TdT) proteins, which will be needed subsequently for somatic recombination and nucleoside transfers involved in junctional diversity, respectively. At the pro-B-cell stage, limited somatic recombination has taken place, and Ig is not yet expressed.

The next stage of B-cell maturation is represented by the pre-B cell and is marked by the synthesis of a cytoplasmic μ heavy chain. Because light chains are not yet expressed at this stage, surface Ig is not present. Some of the μ heavy chains associate with invariant molecules called surrogate light chains and with the signal transducing proteins Ig α and Ig β to form complexes called pre-B cell receptors. Cells that have synthesized heavy chains that are capable of forming part of a pre-B cell receptor are selected for at this stage, as pre-B cell receptors provide important signals for survival, proliferation, and maturation.

The formation of light chains marks the next stage in B-cell maturation, the immature B-cell stage. When light chains join with the μ heavy chains, an IgM molecule results and can be expressed on the cell surface in association with Ig α and Ig β. Although the presence of a B-cell receptor complex confers the ability to recognize specific antigens, at this stage such recognition does not result in proliferation or differentiation. Rather, the cells may undergo negative selection when antigen is encountered. Immature B cells recognizing self-antigen may be negatively selected through deletion,16 anergy, or receptor editing, a process of secondary gene rearrangement by which a new, nonself specificity is acquired.17

The exit of immature B cells from the bone marrow to the spleen marks the beginning of the next stage, the transitional B-cell stage.18 Transitional cells gradually acquire surface IgD, CD21, and CD23 expression and become more immune competent. Alternative splicing of RNA allows the simultaneous expression of IgM and IgD. At the beginning of the stage, cross-linking of the B-cell receptor leads to negative selection. With further maturation, transitional cells become responsive to T-cell help and lose sensitivity to negative selection.

The mature B cell expresses IgM and IgD and is competent to respond to antigen. The cell is considered naive because it has not been activated by antigen. The majority of mature B cells circulate through peripheral lymphoid tissues (spleen, lymph nodes, mucosal lymphoid tissue) and are called follicular B cells, or recirculating B cells. B cells are recruited to the follicle by the chemokine CXCL13, secreted by follicular dendritic cells, and survive in the follicle with the assistance of a cytokine called BAFF (B-cell activating factor), also known as BLyS (B lymphocyte stimulator). A small percentage of mature B cells home to the marginal zone of the spleen and remain resident there.

The encounter of antigen by mature naive B cells leads to B-cell activation, proliferation, and differentiation (see Section “B-Cell Activation and Antibody Function”). A subset of B cells become memory B cells, which can persist for long periods apparently without stimulation by antigen, and which respond rapidly if the antigen is encountered subsequently.19 Another subset of B cells differentiates into cells that make progressively less membrane-bound Ig and more secreted Ig. The terminally differentiated B cells committed to the production of secreted Ig are plasma cells and have abundant rough endoplasmic reticulum, consistent with the function of the cells as antibody factories.20

B cells recognize a variety of macromolecules, including proteins, lipids, carbohydrates, and nucleic acids. The portion of the molecule recognized by the antibody is called an epitope or determinant. B cells recognize both linear epitopes (epitopes formed by several adjacent amino acids) and, quite commonly, conformational epitopes (epitopes present as a result of folding of the macromolecule).21 In contrast to B cells, T-cell responses are almost entirely restricted to linear epitopes of peptides.

Macromolecules, particularly large proteins, may contain several different epitopes, and a humoral response to a macromolecule typically is comprised of multiple different antibodies. Although each different antibody is specific for a given epitopic configuration, similarities in epitopes may exist such that an antibody to a given epitope on a given macromolecule also may be able to bind a different epitope on a different macromolecule. This phenomenon is called cross-reactivity, and may be important in the genesis of autoimmune antibody responses.

Macromolecules that have multiple identical epitopes are classified as being polyvalent or multivalent. Antibodies to these macromolecules or aggregates of macromolecules may form complexes called immune complexes with the antigen. At a particular concentration of antibody and antigen, called the zone of equivalence, a large network of linked antigens and antibodies forms. At lower or higher concentrations of antibody or antigen, the complexes are much smaller. Immune complexes, formed in the circulation or in tissue, may be responsible for disease through the initiation of an inflammatory response.

On cross-linking of the mature B-cell receptor by antigen, clustering of receptors initiates signaling transduced by Igα and Igβ. The complex signaling cascade involving the phosphorylation of tyrosine kinases, including Lyn, Fyn, Btk, and Syk, eventuates in the expression of genes involved in B-cell activation.22 B-cell activation is facilitated by second signals, one of which is provided by the complement protein C3d.23 Complement fragment C3d is formed as a result of complement activation through any of the complement activation pathways (see Section “Complement”). The B-cell surface contains a coreceptor complex consisting of complement receptor 2 (CR2), CD19, and CD81 (also called TAPA-1, or target for antiproliferative antigen-1). Simultaneous binding of antigen by antibody on the B-cell surface and of C3d by CR2 leads to markedly increased B-cell activation. B-cell activation may also occur through Toll-like receptors that recognize specific microbial products.24

The subsequent response to an antigen often involves a complex interaction between B cells and T cells, leading to a fine-tuning of the immune response.25 Recognition of antigen by both B cells and T cells leads to increased expression of cell surface proteins and cytokines that render these cells increasingly capable of migrating toward and productively interacting with each other. T cells recognizing peptide-class II MHC complexes on dendritic cells receive a primary signal from the complex and a secondary signal from costimulatory interactions involving the binding of B7–1 and B7–2 on dendritic cells to CD28 on T cells.26 These activated T cells express CXCR5, the ligand for CXCL13, which results in T-cell migration toward the follicle and therefore increasingly toward B cells. In response to a protein antigen, B cells take up the antigen, process it, and present processed antigen on the cell surface in complex with class II MHC. Activated B cells express less CXCR5, which allows them to migrate from the follicle toward the T-cell zone. At the boundary between follicles and T-cell zones, activated T cells interact with B cells and provide signals to the B cells through the binding of CD40 on B cells to CD40 ligand (CD154) on T cells and through the action of cytokines, notably interleukin 2 (IL-2), IL-4, IL-21, BAFF, and APRIL (a proliferation-inducing ligand).2 These signals will be necessary for subsequent class (heavy chain isotype) switching, affinity maturation, and memory B-cell generation. The overall effects on B cells are stimulation of proliferation and differentiation.

At this phase, some of the activated B cells become short-lived plasma cells, which provide a prompt initial response to an antigen, while others migrate back from the periphery of the follicle to proliferate rapidly and form germinal centers. It is primarily in the germinal centers that class switching, affinity maturation, and generation of memory B cells occur. Class switching from IgM to IgA, IgE, or IgG occurs as a result of T cell–B cell interactions.27,28 The determination of the antibody class selected is based on the site where the antigen is encountered and the cytokine milieu. For example, B-cell responses to antigens encountered on mucosal surfaces characteristically result in class switching to IgA, and transforming growth factor-β is an important contributing cytokine. IL-4 is an important signal for class switching to IgE.

T-cell interaction with B cells also results in affinity maturation, whereby the affinity of antibodies for the antigen progressively increases. During affinity maturation, somatic hypermutations in antibody genes result in antibodies with both greater and lesser affinity for the antigen.29,30 Those antibodies with greater affinity confer a survival advantage on the B cells that produce them. Progressively, the population of B cells evolves in favor of those producing higher affinity antibodies for the antigen. Both class switching and affinity maturation require the expression of an enzyme called activation-induced cytosine deaminase (AID).31

The culmination of germinal center activity is the formation of memory B cells and long-lived plasma cells.32 A number of transcriptional regulators are involved in late B-cell development, including BLIMP1 (B-lymphocyte maturation protein 1), IRF4 (interferon-regulatory factor 4), and XBP1 (X-box-binding protein 1). Plasma cells may arise from and be replenished by memory B cells or may arise from an intermediate cell, the plasmablast. Long-lived plasma cells migrate to and have a survival niche in the bone marrow, where they can persist indefinitely. These bone marrow long-lived plasma cells are the major source of antigen-specific antibody in the circulation.

As noted in Section “Antigens Bound by B Cells,” T-cell responses are limited almost entirely to peptides. Thus, B-cell responses to nonprotein antigens may not result in T-cell help through the mechanisms described earlier.33 In selected cases, T-cell independent nonprotein antigens can induce class switching, but in general, T-cell independent responses are characterized by IgM antibodies of lower affinity. One type of T-cell-independent B-cell response produces so-called natural antibodies—IgM antibodies that are largely anticarbohydrate antibodies produced without apparent antigen exposure.34 These natural antibodies are characterized by a limited repertoire and are produced primarily by B1 peritoneal cells either spontaneously or in response to bacteria that colonize the gut. Marginal zone B cells, located near the marginal sinus in the spleen, may also produce natural antibodies.

Antigen occupation of antibody-binding sites on B cells leads to functional results, called effector functions. With the exception of direct neutralization of antigen by antibody binding, effector functions are typically mediated through the binding of Ig to FcRs.35,36 FcRs can be categorized as those that trigger cell activation and those that do not. Those that can trigger activation contain one or more motifs called immunoreceptor tyrosine-based activation motifs. Of those that do not trigger activation, some can inhibit cell activation and contain a motif called immunoreceptor tyrosine-based inhibition motif. FcRs that neither activate nor inhibit cell activation are involved in the transport of Ig through epithelia and the prolongation of the half-life of IgG.

The effector functions of antibodies serve to eliminate the antigen that initiated the immune response and also to downregulate the immune response when activation is not required. Effector functions of antibodies include neutralization of antigen, complement activation, cell activation (of monocytes, neutrophils, eosinophils, and B cells), phagocytosis (by monocytes and neutrophils), and antibody-dependent cell-mediated cytotoxicity (mediated by NK cells and eosinophils). In addition, engagement of IgG by antigen provides a negative signal to B cells, mediated through the binding of the antigen-antibody complex to an immunoreceptor tyrosine-based inhibition motif-containing Fcγ receptor, FcγIIB, on the B cell.37

Disorders of B cells or antibodies cause or contribute to many diseases of dermatologic relevance. Immunodeficiency diseases may result from abnormalities of B-cell development or activation, or from abnormalities in effector function pathways.38 B-cell lymphomas may result from failure to regulate proliferation, differentiation, or programed cell death.39 Ectopic lymphoid aggregates can arise as a result of aberrant chemokine-mediated lymphocyte homing.40 Antibodies may initiate an inflammatory response that results in injury, as in IgE-mediated allergic reactions or immune-complex diseases.41,42 In some cases, the antigen may not be obviously harmful, but the response to the antigen is. In other cases, the antigen may be pathogenic, but the character or magnitude of the immune response is inappropriate or inadequately controlled. The regulatory systems that protect an organism from attack by its own immune system occasionally go awry.43,44 Failure to eliminate autoreactive cells may be a major underlying abnormality in many patients with autoimmunity. In some cases, autoimmunity may be initiated as a result of an immune response to a pathogen.45 The pathogen may act as a nonspecific activator of the immune system, or may activate the immune response specifically (e.g., by containing an epitope or epitopes that are cross-reactive with an autoepitope). These responses may be particularly difficult to control because the major stimulus for the immune response, the antigen, is a normal component of “self” and cannot be eliminated. As mentioned earlier, B cells have important roles beyond antibody production. Abnormalities or imbalance of immune regulatory functions by B cells may lead to autoimmunity. Thus, through many mechanisms, the normal protective B-cell response, which developed as an elegant means to discriminate very finely among various potential pathogens, can be subverted to result in harm to the organism.

The complement system was discovered through its ability to contribute to, or “complement,” bacterial cell lysis by antibodies.46,47 At physiologic temperatures, serum containing antibacterial antibodies lysed bacteria effectively, whereas serum heated to 56°C (133°F) lost its ability to lyse bacteria. Because antibodies are quite stable at 56°C (133°F), it was postulated that the loss of lytic ability was due to the degradation of heat-labile nonantibody molecules.

Although the complement system was first identified through its role in humoral immunity, a primitive complement system emerged more than 1.3 billion years ago as a component of innate immunity (see Chapter 10).48 The complement system is constituted in part by a set of plasma proteins that are normally inactive or minimally active. The initial steps of activation involve the cleavage of an inactive protein into a smaller and a larger fragment. The larger fragment, then, is itself able to cleave other proteins in the cascade. Because an activated molecule in one step is capable of generating many activated molecules in the following step, the sequential cleavage and activation of complement proteins result in amplification of the cascade.

The early stages of the activation of complement ultimately result in cleavage of the complement protein C3. This is followed by the cleavage of C5 and initiation of the final steps of complement activation. There are three distinct pathways that lead to the cleavage of C3, the classical, the alternative, and the lectin pathways. Although the classical pathway was the first to be described, the alternative pathway is evolutionarily older.49

Alternative Pathway

(Fig. 37-2)

The first step in the activation of the alternative pathway is the binding of C3b to a cell surface such as a bacterial cell surface. Intact C3 is an inactive molecule, but there is a low-level spontaneous cleavage of C3, called tickover, which results in the continuous availability of the C3b fragment. C3b can bind stably to a cell surface through an interaction between a thioester group of C3b and a hydroxyl group of the cell surface. In intact C3, the thioester domain is covered by hydrophobic residues that prevent hydrolysis of the thioester bond. The anaphylatoxin (ANA) domain of C3 stabilizes this inactive conformation. When the ANA domain is cleaved to release C3a, the thioester group is exposed, and conformational change results in its ability to bond to the cell surface.50,51 If this chemical bonding does not occur, the thioester group is hydrolyzed and C3b is inactivated.

Once stable attachment of C3b to the cell surface takes place, a plasma protein called Factor B binds to C3b. Factor B is in turn cleaved by factor D, generating Bb and Ba. The complex of C3b and Bb, stabilized by the plasma protein properidin, is the alternative pathway C3 convertase (i.e., it cleaves C3 into C3a and C3b). The result of the activity of the C3 convertase is an amplification of the pathway by two or three orders of magnitude. The addition of further C3b to the complex results in C3bBbC3b, which constitutes the alternative pathway to C5 convertase. The late steps of complement activation, after the cleavage of C5, are common to the three pathways and are described in Final Steps of Complement Activation.

Thus, the low level of C3b in the plasma acts as a sentinel for microbes. Once C3b is bound to the cell surface, subsequent molecular interactions result in a substantial amplification of the alternative pathway and cleavage of C5. The requirement for binding of C3b to a structural element serves to limit the effect of complement activation to the area where complement activation is needed. The alternative pathway of complement activation does not require finely specific recognition of antigen and so is considered a component of innate immunity. It follows that if specific recognition is not required, C3b can bind to human cells as well as microbes. However, activation on human cells is generally prevented by the intervention of regulatory proteins present on the surface of human cells, protecting these cells from inappropriate and harmful attack.52

Classical Pathway

(Fig. 37-3)

The initial step in the activation of the classical pathway is characteristically the binding of the portion of the C1 complex called C1q to IgG or IgM antibodies.46 The C1q molecule consists of six identical arms attached to a central trunk. The globular ends of the arms attach to the complement-binding regions of the heavy chains of certain Ig classes. In order for C1q to be activated, it must bind simultaneously to at least two Ig heavy chains. This means, in effect, that the Ig must have bound antigen. In the case of IgG, binding of multiple epitopes by antibodies results in close proximity of the antibodies and thus the proper configuration for C1q activation. (As mentioned in Section “Immunoglobulin G,” IgG4 is an exception, in that it does not bind C1q or activate complement.) Because IgM exists as a pentamer, theoretically IgM without bound antigen could activate complement. However, when IgM is not bound by antigen, the C1q binding site is not accessible. When antigen is bound, a conformational change results in exposure of the C1q binding site.

The complement component C1 is a complex of C1q, C1r, and C1s. Binding of two or more of the globular heads of C1q results in activation of C1r. Activated C1r is a protease that cleaves and activates C1s, and activated C1s, in turn, cleaves C4 into C4a and C4b. (The numbering of the complement proteins differs from their positions in the activation sequence, as components were discovered before the elucidation of their positions in the pathway.) C4b, like C3b, contains a thioester group that can form stable bonds with hydroxyl groups on a particular structure. Bound C4b is then bound by C2, which is cleaved into C2a and C2b. The C4bC2b complex is the classical pathway C3 convertase. (Note: Typically, suffix “a” denotes the smaller and “b” the larger complement fragment. Historically, the exception was C2, where C2a represented the larger and C2b the smaller fragment. Some recent textbooks now identify the smaller fragment as C2a and the larger as C2b, to maintain consistency with the nomenclature of the other complement proteins. However, many recent publications continue to adhere to the historic nomenclature.)

The cleavage of C3 results in C3a and C3b. The C3b fragment may then go on to activate complement by the alternative pathway, or act in concert with C4bC2b to form C4bC2bC3b, the classical pathway C5 convertase.

The major characteristics of the classical pathway are much the same as those of the alternative pathway. Activation of the pathway requires attachment of a complement protein to a structure such as a cell surface or immune complex, so that the effects of complement activation are spatially limited. Initial activation steps result in the formation of a C3 convertase that cleaves C3. Cleavage of C3 leads to the formation of a C5 convertase that cleaves C5 into C5a and C5b. A major difference is the initiation of the classical pathway by specific recognition of antigen by antibodies, in contrast to the less specific binding that occurs to initiate the alternative pathway.

Lectin Pathway

The lectin pathway is very similar to the classical pathway, with the exception of the initiating steps. The first step is the binding of a plasma lectin, mannose-binding protein (MBP), to polysaccharides on microbial cell surfaces.53 MBP, a member of the collectin (collagenous lectin) family, is structurally similar to C1q and can associate with C1r and C1s. MBP attachment to a microbe can begin the cascade through the activation of C1r and C1s. MBP also interacts with MBP-associated serine proteases (MASPs), which are analogous to C1r and C1s. Binding of MBP to microbial cell surfaces results in cleavage of MBP-associated serine protease and thence cleavage of C4.54 In either event, the effect is the formation of C4b and its stable binding to a cell surface. As is the case for the alternative pathway, the lectin pathway is a component of innate immunity.

Final Steps of Complement Activation

(Fig. 37-4)

The alternative, classical, and lectin pathways converge at the cleavage of C5. The C5b fragment remains surface bound. The next steps do not involve enzymatic cleavage, but rather the sequential binding of C6, C7, and C8 to C5b. The C5b-8 complex stably bound to a cell membrane becomes an active membrane attack complex (MAC) through the addition of C9. C9 polymerizes around the complex and forms pores in cell membranes. These pores may result in cell death through osmotic rupture, particularly in nonnucleated erythrocytes.

Nucleated cells are more resistant to lysis, but may still exhibit effects attributable to MAC binding.55 It is quite possible that the nonlytic changes induced by MAC are of more functional and pathologic significance overall than is MAC-induced cell lysis.56 These nonlytic effects may differ depending on cell type and milieu, and similar effects may lead to different outcomes. For example, MAC insertion into phagocyte cell membranes can lead to the production of inflammatory mediators such as reactive oxygen species and prostaglandins, resulting in phagocyte activation.57 Glomerular epithelial cells may also exhibit inflammatory mediator production, but, in that setting, the inflammatory mediators may lead to tissue injury.58 MAC has also been reported to cause proliferation of certain cells, and MAC has been reported to have both apoptotic and antiapoptotic properties.55

Additional Initiators of Complement Activation

In addition to the characteristic initiators of the three pathways, certain additional structures can trigger complement activation.59 These include, among others, the following: in the alternative pathway, IgA immune complexes and endotoxin; in the classical pathway, C-reactive protein, apoptotic bodies, and serum amyloid P; and in the lectin pathway, serum ficolins (lectins which bind N-acetylglucosamine).53,60

As mentioned earlier, the earliest function of the complement system to be discovered was the lysis of bacteria. Killing of microbes through direct lysis is mediated by the MAC, C5b-9. Microbes may also be destroyed through coating, or opsonization, by complement and phagocytosis of the opsonized particles by phagocytic cells. The processes of opsonization and phagocytosis are also mechanisms for another important function of complement—the clearance of immune complexes and apoptotic debris from the circulation.

An ancillary function of complement activation is the induction of inflammation. Inflammation, characterized by vascular changes and ingress and activation of leukocytes and inflammatory proteins, serves to augment the localized immune response in tissue. Three mediators of inflammation initiated by complement activation are the complement fragments C3a, C4a, and C5a. These are called anaphylatoxins because of their ability to induce degranulation of mast cells.61 The most potent of these ANAs is C5a. Receptors for C5a are expressed on endothelial cells, mast cells, eosinophils, basophils, monocytes, neutrophils, smooth muscle cells, and epithelial cells. Binding of C5a to endothelial cells results in increased vascular permeability and expression of P-selectin, both of which promote leukocyte accumulation in tissue. Binding to neutrophils results in increased neutrophil motility, adhesion to endothelial cells, and production of reactive oxygen species. The overall result is the accumulation of inflammatory cells at local sites in tissue where they can phagocytose and efficiently kill microbes.

The activation of complement also results in stimulation of the humoral immune system through the generation of C3d. B cells whose cell surface antibodies recognize complement-bound antigen are upregulated by the concurrent binding of C3d to CR2 on the B-cell surface. Opsonization by complement also facilitates antigen presentation to B cells by follicular dendritic cells.

There are several proteins that interact with complement and serve to mediate or regulate its functions. The C5a receptor, which is a member of the seven-transmembrane α-helical G-protein-coupled receptor family, was mentioned in the previous section. Some of the best described of the complement receptors are CR1–CR4.

The type 1 complement receptor, CR1 (CD35), is a member of a family of proteins called regulators of complement activation (RCA), which share a common structure consisting of multiple short consensus repeats, also known as complement control protein repeats.52 CR1 binds C3b or C4b and is expressed on peripheral blood cells, including monocytes, B and T lymphocytes, neutrophils, eosinophils, and erythrocytes; on follicular dendritic cells; and on keratinocytes.62 On phagocytic cells, binding of CR1 to C3b or C4b results in phagocytosis of particles opsonized by complement fragments, as well as activation of microbicidal mechanisms in the phagocytic cells. On erythrocytes, binding of CR1 to C3b- or C4b-coated immune complexes results in transport of the complexes to the spleen and liver, where they are cleared from the circulation by phagocytes. Thus, CR1 serves as an important mediator of complement function. It can also serve as a downregulator of complement activation, as it is involved in the dissociation of C3 convertase complexes.

The type 2 complement receptor, CR2 (CD21), is also a member of the RCA family.59 CR2 binds C3 fragments iC3b (“i” stands for inactive), C3dg, and C3d, as well as Epstein–Barr virus, interferon-α, and the immunoregulatory protein CD23. CR2 is expressed on subsets of B and T lymphocytes, basophils, mast cells, follicular dendritic cells, and some epithelial cells, including keratinocytes. On B cells, CR2 serves as a coreceptor for B-cell activation. When CR2 is bound by C3d, the level of B-cell activation is increased by orders of magnitude.63 On dendritic cells, CR2 engagement results in a trapping of immune complexes in germinal centers. CR2 also appears to play a role in antigen presentation to T cells.

The type 3 complement receptor, CR3 (CD11b/CD18, Mac-1), is an integrin cell surface molecule expressed on monocytes, neutrophils, NK cells, and mast cells.64 It functions to promote phagocytosis of microbes through binding to iC3b and through direct binding to microbes. It interacts with intercellular adhesion molecule 1 expressed endogenously on endothelial cells to stabilize the adhesion of leukocytes to endothelium, facilitating the recruitment of leukocytes from the circulation into tissue.

The type 4 complement receptor, CR4 (CD11c/CD18), is also an integrin cell surface molecule. It is expressed on monocytes, neutrophils, NK cells, and dendritic cells, and probably functions similarly to CR3.

Among the recently described complement receptors are SIGN-R1,65 which binds C1q and is expressed on splenic marginal zone macrophages, and CRIg (complement receptor of the immunoglobulin family),66 which binds C3b and iC3b and is expressed on a subset of tissue-resident macrophages.

Molecules involved in the regulation of complement activation serve to downregulate the immune response once an immune response is no longer needed and to limit the immune response to the sites required, specifically protecting self from complement attack.

The C1 inhibitor (C1 INH) is a protease inhibitor that inhibits certain plasma serine proteases, including C1, kallikrein, and factor XII.46 C1 exists as a complex of C1q and a tetramer of two C1r and two C1s fragments. When C1q binds antibody, C1 INH can act to limit complement activation by binding to the C1rC1s tetramer, dissociating it from C1q and preventing downstream activation of the pathway.

Another major point of interaction of regulatory proteins is with bound C3b or C4b.52 As mentioned earlier, the thioester group of unbound C3b or C4b is rapidly hydrolyzed, rendering these molecules inactive. For surface-bound C3b or C4b, inactivation occurs through the displacement of components of the alternative or classical pathway C3 convertase from C3b or C4b or through proteolysis of C3b or C4b. In the alternative pathway, inactivation of the C3 convertase complex, C3bBb, can take place through the displacement of Bb from C3b by the plasma protein factor H or the cell surface proteins decay accelerating factor (DAF, CD55), membrane cofactor protein (MCP, CD46), and CR1. These three cell surface proteins, all members of the RCA family, are expressed on human cells but not microbes, thereby allowing complement activation to proceed on microbes while protecting human cells from injury. Factor H preferentially binds cell surfaces with high levels of sialic acid, and the relative abundance of sialic acid on human cells but not microbes further focuses the downregulation of complement activation on human cells. In the classical and lectin pathways, the C3 convertase is C4bC2b. DAF, MCP, and CR1 can displace C2b from C4b, as can the plasma protein C4-binding protein (C4BP). Thus, the cell surface proteins DAF, MCP, and CR1 can dissociate the C3 convertases of both the alternative and the classical/lectin pathways, whereas the plasma proteins factor H and C4BP are specific for alternative or classical/lectin, respectively. Proteolysis of C3b or C4b is mediated by factor I, a plasma protein that requires cofactors for its activity. MCP, CR1, factor H, and C4BP can all serve as cofactors for factor I.

Regulation of complement at the late steps is mediated in part by CD59, a cell surface protein expressed on human cells but not microbes. It binds the C5b-8 complex and inhibits addition of C9, blocking formation of the MAC. Plasma S protein binds the C5b-7 complex and blocks its insertion into the cell membrane and also inhibits C9 polymerization.67 Intact MACs may be removed from cells through shedding on membrane vesicles or by internalization and degradation.55

Carboxypeptidase N can remove the terminal arginine of C3a, C4a, and C5a and has been referred to as ANA inactivator.68 Carboxypeptidase R has also been shown to remove the terminal arginine of C3a and C5a.69

Genetic Abnormalities of the Complement System

Deficiencies of complement cascade proteins, complement receptors, or complement regulatory proteins can lead to a variety of diseases.70 Genetic deficiencies of complement have been associated primarily with increased risk for infection or autoimmunity. As examples, deficiencies of many complement components, particularly the early complement components C1–C4, have been associated with early-onset systemic lupus erythematosus (SLE), C3 deficiency has been associated with life-threatening pyogenic infections, and C5–C9 deficiencies have been associated with Neisserial infections (reviewed in Chapter 143). Genetic deficiency of mannose-binding lectin is relatively common, with significantly low levels occurring in about 10% of Caucasians, and is associated with increased risk for infection and autoimmunity, including SLE.71,72

Altered expression of CR3 (CD11b/CD18) and CR4 (CD11c/CD18) occurs in leukocyte adhesion deficiency-1, a congenital disorder resulting from mutations in the gene encoding CD18. Mutation in CD18 also affects expression of CD11a/CD18 (leukocyte function-associated antigen-1). Patients with leukocyte adhesion deficiency-1 exhibit significant abnormalities of leukocyte adhesion and have recurrent infections (see Chapter 143).

Deficiency of the complement regulatory protein C1 INH causes angioneurotic edema as a result both of poorly regulated classical pathway activation and of excess bradykinin due to the actions of kallikrein and factor XII. Angioedema may also occur if one has markedly reduced levels of the ANA inactivator, carboxypeptidase N73 (see Chapter 38).

Deficiency of a protein required for the proper expression of DAF and CD59 on the cell surface is associated with paroxysmal nocturnal hemoglobinuria, a disease characterized by complement-mediated erythrocyte lysis.74 Hemolytic-uremic syndrome has been associated with mutations in MCP, factor H, and factor I.75 Factor H deficiency has also been associated with membranoproliferative glomerulonephritis.

The risk for developing age-related macular degeneration is affected significantly by the presence of certain polymorphisms in genes of the complement system, particularly complement factor H but also factor B and C2.76

Complement, Systemic Lupus Erythematosus, and Autoantibodies

Complement has been closely associated with SLE (see also Chapter 155), albeit in seemingly paradoxical ways.77,78 Genetic deficiencies of complement components are associated with SLE, but some of the tissue injury seen in SLE appears to be mediated in part by complement activation. Thus, complement seems to be simultaneously protective and deleterious. These observations underscore the protean roles of complement in the immune system. Complement activation has the potential for causing tissue injury, but complement components may be important in clearance of immune complexes and apoptotic debris. Apoptotic bodies that are not cleared effectively may be able to trigger autoimmunity through presentation of normally sequestered autoantigens to the immune system. It has also been suggested that complement participates in eliminating self-reactive immature B cells, a further mechanism for a protective effect of complement in SLE.78

Altered expression of both CR1 and CR2 has been observed in patients with SLE.79 In murine models of lupus, knockout mice lacking expression of CR1 and CR2 (located on the same gene in mice and produced through alternative splicing) have accelerated autoimmunity if the mice otherwise have the optimal genetic background. These findings indicate that the interaction of CR2 and C3d, important in B-cell response to antigen, is another factor that determines susceptibility to SLE.

Autoantibodies to complement components can result in or exacerbate disease.80 Autoantibodies to C1q are relatively common in SLE and have been associated with more severe renal disease, possibly through an adverse affect on the clearance of immune complexes or apoptotic bodies.81 C3 nephritic factor is an autoantibody to the C3 convertase, C3bBb, which acts to stabilize the complex. Its clinical significance is its association with membranoproliferative glomerulonephritis type II and partial lipodystrophy.

Complement and Vascular Disease

Complement activation has been implicated in the pathogenesis of atherosclerosis, reperfusion injury after myocardial ischemia, and cerebral infarct in ischemic stroke.82,83 In the microvascular proliferative disease associated with diabetes, glycation, and thereby inactivation of CD59, may result in cellular proliferation due to nonlytic proliferative effects of MAC.84 Complement activation has also been implicated in hyperacute rejection of xenotransplants due to the presence of natural antibodies to components of the endothelial cells of the transplanted organ, with resultant complement activation, endothelial cell injury, and intravascular coagulation.

Evasion or Subversion of Complement by Microbes

The observation that individuals with deficiencies of components of the late stages of complement activation are at increased risk for only a limited set of infections illustrates that infectious agents have evolved means of evading destruction by complement.85 Gram-positive bacteria have thick cell walls that are difficult to penetrate by MAC.46 Group A streptococcus M protein binds factor H, which downregulates complement activation, and many pathogens have evolved mechanisms to attract factor H through the expression of sialic acids on their surfaces. Staphylococcus aureus expresses several proteins that inhibit C3 activation. A protein of vaccinia virus (VCP-1, vaccinia virus complement control protein-1) acts as a cofactor for factor I, leading to proteolysis of C3b and C4b. In human immunodeficiency virus (HIV) infection, the inclusion of downregulatory molecules of the complement system into the viral or host cell membrane allows HIV to evade complement-mediated destruction.86 DAF and CD59 can be subsumed into the HIV virus membrane upon budding from infected human cells, and factor H can bind to HIV surface glycoproteins on infected human cells.

The complement system has been subverted by certain infectious agents for entry into the cell. Epstein–Barr virus penetrates B cells via binding to CR2 on the B-cell surface.46 Measles virus binds to cells via MCP. Mycobacteria make C4-like molecules that bind C2b and then cleave C3. The deposition of C3b on the mycobacterial cell membrane leads to its uptake into macrophages, where it exists as an intracellular parasite. Knowledge of these evasive and subversive strategies of pathogens may be useful in designing vaccines and targeted therapies.85