image

3 - The biology of immunoglobulins

Chapter 3

Summary:

  • Immunoglobulin light and heavy chains contain variable and constant domains.
  • The structural diversity of immunoglobulin antigen-binding sites arises from genetic recombination and somatic hypermutation of genes corresponding to the light and heavy chain variable region domains.
  • Further diversity of κ and λ light chain constant domains arises from isotypic and allotypic variation.
  • FLCs are secreted by plasma cells.
  • sFLCs have a half-life of a few hours due to rapid renal clearance.
  • Serum IgG has a prolonged and variable half-lie due to FcRn recycling.

3.1. Immunoglobulin structure

Antibody (immunoglobulin) molecules are composed of two identical heavy chains and two identical light chains. Heavy chains are each paired with a single light chain via a disulphide bridge and non-covalent interactions to form a heavy-light chain pair (or half-molecule). Two heavy-light chain pairs are linked by disulphide bonds in the so-called ‘hinge region’ to form a Y-shaped structure that is arranged symmetrically about a two-fold axis (Figure 3.1).

Immunoglobulin heavy and light chains each have constant and variable regions. A pair of heavy and light chain variable regions together forms the antigen-binding site. The variable regions exhibit enormous structural diversity, particularly of antigen-binding contacts, allowing the recognition of a huge variety of antigens.

Antibody heavy and light chains are composed of homologous structural units known as 'immunoglobulin domains'. Each domain is approximately 110 amino acids long and is constructed from a series of antiparallel β-strands connected to form two β-pleated sheets. The sheets are covalently linked by an intrachain disulphide bridge and each domain adopts a roughly barrel-shaped structure characteristic of an immunoglobulin fold [88][89].

The light chain tertiary structure consists of two immunoglobulin domains joined by a loop to form a single variable region and single constant region (Figure 3.2). In humans, light chains are encoded by two different gene loci, resulting in the serologically distinguishable light chain types, κ and λ. Immunoglobulin molecules are assembled in plasma cells with exclusively κ or λ light chain types, never both.

Similar to light chains, the heavy chain contains one variable domain corresponding to a single variable region. By contrast, the number of heavy chain constant domains (comprising the constant region) varies between immunoglobulin classes, of which there are five: IgG, IgA, IgM, IgD and IgE. Human IgG and IgA can be further divided into closely related subclasses IgG1, 2, 3 and 4, IgA1 and 2. These classes and subclasses are encoded by separate heavy chain constant genes (γ1-4, α1-2, μ, δ and ε, respectively). The constant regions of the heavy chain mediate most of the biological functions of antibodies by interacting with other effector molecules and immune cells. The majority of secreted antibodies are monomeric, although several immunoglobulin subtypes form oligomers, such as IgA and IgM (Figure 3.3).

In B-cells, the heavy chains, κ light chains and λ light chains are each encoded by independent chromosomal loci containing multiple copies of analogous gene segments. The gene segments within each locus are rearranged stochastically by somatic recombination and RNA processing mechanisms, ultimately resulting in the expression of functional immunoglobulin proteins. The light chain variable domain is constructed from variable (V) and joining (J) gene segments, whilst the constant domain is encoded by a separate constant (C) gene segment (Figure 3.4). The heavy chain variable domain is constructed from three gene segments: V, D (diversity) and J.

3.2. Immunoglobulin diversity

In humans it is calculated that there are at least 1011 unique antibody structural variants possible which allows for the recognition of a vast number of different antigens [90]. Such diversity is generated in four main ways:

Firstly, different combinations of gene segments are used in the rearrangement of heavy and light chain genes during early B-cell development. κ light chains are constructed from one of approximately 40 functional Vκ gene segments, one of 5 Jκ gene segments and a single Cκ gene. λ light chains are constructed from one of approximately 30 Vλ gene segments, and one of four (or more) pairs of functional Jλ gene segments and Cλ genes [90] (Figure 3.5) . The heavy chain variable region is formed from one of around 60 VH, one of 30 DH, and one of six JH gene segments [90]. This combinational diversity accounts for a substantial amount of variable region diversity. Secondly, diversity arises from the addition or removal of nucleotides at the junctions between V (D) and J gene segments during recombination. A third source of diversity arises from the many different combinations of heavy and light chains, and finally, somatic hypermutation introduces point mutations in the variable region genes of light and heavy chains in mature activated B-cells [90].

In light chains, variations are also found in a region of the variable domain corresponding to the first 23 amino acids of the first framework region (a region not associated with antigen binding). Using monoclonal antibodies, four κ (Vκ I - Vκ IV) and six λ subgroups (Vλ I - Vλ VI) have been identified [91] (Figure 3.6A). The specific subgroup structures influence the potential of free light chains (FLCs) to polymerise. For example, AL amyloidosis is associated with Vλ VI, and light chain deposition disease (LCDD) with Vκ I and Vκ IV.

​3.3. Isotypic and allotypic variation of light chain constant domains

In addition to genetic recombination and somatic hypermutation of the variable domains, further heterogeneity of light chains arises from isotypic and allotypic variation of the constant domains (Figure 3.6B and C). The human genome contains a variable number of λ constant genes, giving rise to multiple λ chain isotypes, which can be distinguished serologically by the expression of Mcg, Kern and Oz markers. Additionally, the genome contains a single κ constant gene for which three serologically-defined allotypes have been identified, designated Km1, Km2 and Km3 [92]. These allotypes define three Km alleles, which differ in two amino acids, as presented in Table 3.1. There is no evidence that these constant region variants affect FLC measurements using Freelite® assays, which are based on polyclonal antisera (Chapter 5).

Km allele Amino acids
Km1 Val(153) Leu(191)
Km1,2 Ala(153) Leu(191)
Km3 Ala(153) Val(191)

Table 3.1. Amino acid substitutions in κ light chain constant domains.

3.4. Immunoglobulin and FLC production

Immunoglobulins and FLCs are produced by B-cells. During their development the earliest immunoglobulin polypeptide to be produced is the μ heavy chain, in the pre-B-cell. Immature and mature B-cells produce either κ or λ light chains, which associate with μ heavy chains to form membrane-bound IgM. Upon activation, mature B-cells differentiate into plasma cells, which secrete immunoglobulin into the serum. Activation may also stimulate plasma cells to switch the heavy chain constant domain, and hence the class of antibody produced, for example, IgM to IgG. Approximately 40% more light chains than heavy chains are synthesised [93], and this excess of FLCs is thought to favour accurate assembly of intact immunoglobulin molecules. Light chains which remain unbound from their heavy chain partner are secreted into the blood as FLCs. Secretion of FLC is highest from plasma cells, with twice as many producing κ-chains than λ-chains. κ FLCs are normally monomeric, while λ FLCs tend to be dimeric, joined by disulphide bonds; however, higher polymeric forms of both FLCs may occur (Figure 3.7). Tumours associated with the different stages of B-cell maturation may secrete monoclonal FLCs and/or monoclonal intact immunoglobulins into the serum (Figure 3.8).

In normal individuals, approximately 500 mg of FLCs are produced each day from bone marrow and lymph node cells [91][93]. The molecules enter the blood and are rapidly partitioned between the intravascular and extravascular compartments. The normal plasma cell content of the bone marrow is about 1%, whereas in multiple myeloma (MM) this can rise to over 90%. In chronic infections and autoimmmune diseases the bone marrow may contain 5 - 10% plasma cells, and may be associated with hypergammaglobulinaemia and corresponding increases in polyclonal serum FLC (sFLC) concentrations. Identification of monoclonal plasma cells in the bone marrow by histology or flow cytometry is an essential part of MM diagnosis, and is frequently based on identifying intracellular κ and λ light chains by direct immunofluorescence techniques (Figure 3.9).

3.5. Clearance and metabolism

Serum concentrations of FLCs and intact immunoglobulins reflect the balance between their production and clearance rates. An understanding of immunoglobulin clearance mechanisms in both normal and pathological conditions is important when considering the utility of sFLCs and intact immunoglobulins as tumour markers in monoclonal gammopathies.

3.5.1. Half-life of sFLCs

sFLCs are rapidly cleared and metabolised by the kidneys. At around 25 kDa in size, monomeric FLCs, characteristically κ, are cleared in 2 - 4 hours at 40% of the glomerular filtration rate. Dimeric FLCs of around 50 kDa, typically λ, are cleared in 3 - 6 hours at 20% of the glomerular filtration rate (Figure 3.10), while larger polymers are cleared more slowly [93][94]. In contrast, IgG has a half-life of approximately 21 days with minimal renal clearance (Section 5.3). Although κ FLC production rates are estimated to be twice that of λ, their faster removal ensures that actual serum concentrations are approximately 50% lower (Chapter 5). The half-life of FLCs is dependent upon kidney function, so that FLC removal may be prolonged to 2 - 3 days in MM patients with complete renal failure [93][94][91][95]. In patients with chronic kidney disease (CKD), κ and λ sFLC concentrations increase due to reduced renal clearance [96]. When renal clearance is reduced, a greater proportion of sFLC are removed through pinocytosis by cells of the reticuloendothelial system [97]. This mechanism removes κ and λ sFLC at the same rate so the relative FLC concentrations change to reflect more closely the higher rate of κ production and there are minor increases in the κ/λ sFLC ratio [96].

3.5.2. Renal clearance of FLCs

Figure 3.11 shows the glomerular filtration and metabolism of FLCs within a kidney nephron. Each nephron contains a glomerulus with basement membrane fenestrations, which allow filtration of serum molecules into the proximal tubules. Pore sizes are variable, with restricted filtration of molecules that are greater than 20 kDa in size, and a molecular weight cut-off of around 60 kDa. Protein molecules that pass through the glomerular pores are bound by the multi-ligand megalin and cubulin receptors on proximal tubule epithelium; these are then absorbed unchanged, degraded in the proximal tubular cells into their constituent amino acids, or excreted as fragments [99]. This megalin/cubulin absorption pathway is designed to prevent loss of large amounts of proteins and peptides into urine. It is very efficient and can process between 10 and 30 g of small molecular weight proteins daily. Therefore, the 500 mg of FLCs produced each day by the normal lymphoid system are filtered by the glomeruli and completely processed in the proximal tubules [95][100][101].

In normal individuals, between 1 and 10 mg of FLCs are excreted per day into the urine. Their exact origin is unclear, but they probably enter the urine via the mucosal surfaces of the distal part of the nephrons and the urethra, alongside secretory IgA. This secretion is part of the mucosal defence system that prevents infectious agents entering the body.

Because of the huge metabolic capacity of the proximal tubule, the amount of FLCs in urine (even when production is considerably increased in a patient with MM), is more dependent upon renal function than synthesis by the tumour. As a consequence, serum and urine FLC concentrations may differ during the evolution of light chain MM (LCMM) (Figure 3.12). From low initial starting concentrations, sFLCs increase steadily with growing tumour mass, while concentrations in the urine show little change until the proximal tubular metabolism is exceeded and overflow proteinuria develops. Hence, early disease and oligo-secretory disease are not identified from urine tests. Subsequently, urine FLCs rise rapidly as overflow occurs, to reach a maximum. Concentrations then decrease as renal impairment occurs, and are low in complete renal failure. By contrast, sFLC levels increase as renal impairment develops due to the lengthening half-life of FLCs that are no longer cleared by the kidneys. Because of the biphasic urine curve, decreasing concentrations may indicate response to treatment or deterioration of renal function. Urine measurements are therefore unreliable during disease monitoring. Serum levels, however, rise or fall in correct relationship to worsening or improving disease status. The merits of serum over urine testing are further discussed in Chapter 24.

3.5.3. Half-life of IgG, IgA and IgM

Under normal circumstances, most serum proteins that are too large for renal filtration (> around 60 kDa) are removed by pinocytosis, a process that occurs in all nucleated cells as they obtain their essential nutrients from plasma. This accounts for the half-life of IgA and IgM, which is constant at around 5 - 6 days. By contrast, IgG has a concentration-dependent half-life of approximately 21 days due to recycling by FcRn receptors [103][104][105][102]. These receptors have a structure similar to Class I MHC molecules with a heavy chain of three domains and a single domain light chain comprising β2-microglobulin (Figure 3.13). FcRn receptors are functional in most nucleated cells, including renal podocytes, which may account for the presence of IgG in the urine at high serum concentrations (Chapter 24) [105][106][107][108][109][110][111]. These are the same receptors that transport IgG from the pregnant mother to the developing foetus in the last trimester of pregnancy.

The heterodimeric FcRn molecules protect both IgG and albumin from acid digestion in lysosomes, recycling them back to the cell surface (Figure 3.14). Interestingly, IgG and albumin molecules do not compete for the same sites on the receptor, although the exact mechanism and sites of binding are unknown. In the absence of functioning FcRn receptors, as in patients with familial hypercatabolic hypoproteinaemia (a disease associated with a genetic deficiency of β2-microglobulin), the half-lives of IgG and albumin are only 3 days. Such patients have hypogammaglobulinaemia, not from failure of production, but simply from excessive catabolism.

At high IgG concentrations, the FcRn recycling system can reach saturation and the half-life of IgG falls as there are insufficient FcRn receptors to protect all IgG molecules (Figure 3.15). Hence, a patient presenting with, for example, a monoclonal IgG of 90 g/L is producing far more than 3 times the amount of IgG than a patient presenting with 30 g/L of IgG. In contrast, at low IgG concentrations, when FcRn receptor protection is maximal, the IgG half-life extends to several months. Serum IgG concentrations may therefore be an unreliable indicator of tumour production rates in patients with IgG MM (Chapter 18).

Questions

  1. What accounts for immunoglobulin light chain heterogeneity?
  2. What are the normal serum half-lives of IgG and FLCs?
  3. Do urine FLC concentrations always increase alongside rising sFLC concentrations?
  4. Why does the IgG half-life vary with concentration?

Answers

  1. Light chain heterogeneity arises from genetic recombination, isotypic, allotypic and idiotypic variation and somatic hypermutation of the variable regions after antigen exposure (Sections 3.1 and 3.2)
  2. IgG is approximately 21 days and FLCs 2 - 6 hours (Section 3.4).
  3. No. If there is significant renal impairment, urine FLC excretion falls (Section 3.4).
  4. FcRn receptors saturate at high IgG concentrations so the half-life shortens (Section 3.4). At low IgG concentrations, the half-life lengthens because FcRn receptor recycling is maximal (Section 3.4).
VIEW ANSWERS

Figures

References