General Structural Features of Immunoglobulin Molecules; Myeloma Proteins

One purpose of this text is to review the literature leading to our current knowledge of the structure of immunoglobulins. In subsequent chapters, references are provided to much of the pertinent research. In this introductory section we will outline some of the basic structural characteristics of immunoglobulins without citing the references on which the information is based. The discussion will be brief and, at times, sketchy. Its purpose is to facilitate reading of the remainder of the book, and perhaps to permit the reading of other chapters without strict regard to their sequence.

Immunoglobulins are present throughout the vertebrate kingdom but have not been identified in invertebrates. A basic feature, maintained during evolution, is the four-chain structure, comprising two light (L) and two heavy (H) chains. Both the H and L chain contribute amino acid residues which form the antigen-binding site. Each H-L pair produces one binding site, and the four-chain structure thus yields two sites. An individual molecule normally contains one species of H chain and one species of L chain, so that the combining sites of a molecule are identical. This property has its origin in the expression of only one gene for an L chain (variable region) and of one gene for the H chain (variable region) in a lymphoid cell that is synthesizing antibody.

An immunoglobulin molecule may have more than four chains. Such molecules are polymers of the four-chain unit, which are held together by disulfide bonds. For example, molecules of the IgM class may have from 16 to 24 chains (4–6 four-chain units, or 8–12 potential binding sites), depending on the animal species. Serum IgA may also exist as a polymer, and secretory IgA, found in colostrum, tears, saliva, and other secretory fluids, generally occurs as a dimer of the four-chain unit.

Thus the minimum “valence” of an immunoglobulin is 2. The presence of more than one combining site on the antibody molecule is of biological importance because it permits aggregation of macromolecular or particulate antigens. Bivalence is essential for enhancement of phagocytosis by antibody and for fixation of complement in the presence of antigen. Also, bivalent or multivalent attachment of a single antibody molecule to a particle increases the energy of interaction or “avidity”; this is significant, for example, in the neutralization of viruses, particularly when the affinity of a single combining site is low, as is often the case early in the process of immunization.

It was demonstrated by R.R. Porter in 1958 that cleavage of rabbit IgG by papain at neutral pH gives rise to large fragments with distinctive properties. The nature of this cleavage, as it applies to rabbit IgG or human IgG1, is illustrated in Figs. 1.1 and 1.2. Papain selectively attacks each H chain at a site just N-terminal to the interheavy chain disulfide bonds. This liberates three large fragments. Two are called Fab fragments (ab for antigen-binding) and the other is fragment Fc (c for crystallizable). Each Fab fragment (mol. wt. 45,000) has one antigen-binding site, i.e., is univalent. Fab fragments therefore cannot precipitate macromolecular antigens or agglutinate particulate antigens, but they can, when present in excess, block the precipitation or agglutination that would otherwise occur upon addition of untreated, bivalent antibody. The fragments do this by occupying antigenic determinants, thus denying access to bivalent antibodies. Each Fab fragment comprises a complete L chain and the N-terminal half of an H chain, designated fragment Fd. The L chain and fragment Fd are held together by noncovalent bonds and by a disulfide bond.

The third fragment, Fc (mol. wt. 50,000), crystallizes spontaneously in cold neutral buffer from a digest of rabbit IgG, which is heterogeneous. Crystallizability is generally associated with homogeneity and this was the first evidence for the presence of an invariant segment in the heterogeneous IgG molecules. We know now that fragment Fc consists of the C-terminal halves of two H chains, which are held together by noncovalent interactions.

Figures 1.1 and 1.2 also show the site of attack of rabbit or human IgG by pepsin at pH 4 to 4.5. Owing to the fact that its preferential site of cleavage is on the C-terminal, rather than the N-terminal side of the interheavy chain disulfide bond(s), a large bivalent fragment is liberated. Only the interheavy chain disulfide bond(s) join the two univalent fragments after peptic digestion; these bonds are very easily reduced, liberating two univalent fragments, designated Fab′, which are slightly larger than fragment Fab. [The initial product of digestion is F(ab′)2.] The Fab′ segments can be reunited, to form fragment F(ab′)2, by reoxidation of the disulfide bond; this cannot, for obvious reasons, be accomplished with fragment Fab. Pepsin partially degrades fragment Fc, possibly owing to the low pH required for peptic activity.

Because of the susceptibility of a particular region in the middle of the H chain to attack by papain, pepsin, and other enzymes, it is generally thought that this region must be loosely folded. It has been designated the hinge region because the two univalent (Fab) fragments can move in space in relation to one another, and it is believed that the “swivel” is in this region of the chain.

This flexibility permits bivalent attachment of a single antibody to a particulate antigen (horseshoe configuration of the antibody) or, alternatively, the linking of two antigen particles with the antibody stretched out to its maximum length (about 140 Å for IgG). The combining sites are situated at the ends of the bivalent IgG molecule. A single, decavalent IgM molecule can attach itself to a particulate antigen through several combining sites. (Information of this kind is obtained by electron microscopy; the location of the combining site has been confirmed by X-ray crystallography, which also provides information as to the dimensions of the site.)

As indicated in Fig. 1.1, each H chain is joined to an L chain through a disulfide bond. In addition, strong noncovalent bonds serve to link the two chains. The latter can be disrupted by reagents such as 6 M guanidine hydrochloride, 8 M urea, or 1 M propionic acid. If the interchain disulfide bond joining the H to the L chain is first reduced, the chains can be separated on a preparative scale by gel filtration in the solvents mentioned. The chains reassociate as H-L pairs, in quite good yield, if the dissociating solvent is replaced by dialysis against neutral buffer. In the native molecule the H chains are joined to one another through disulfide and noncovalent bonds in the hinge and Fc regions. Under appropriate conditions it is possible to produce half-molecules, each consisting of a complete H chain and a complete L chain.

Besides L and H chains, secretory IgA contains a polypeptide called secretory component, or SC, which has a molecular weight of 65,000 – 70,000 and is bound to the H chains of the molecule. Another component, J chain (J for joining; mol. wt. 15,000), is found in the IgM of many primitive and advanced species, and in polymeric, but not monomeric (four-chain) IgA. J chain is linked to H chains through disulfide bonds. The amino acid sequences and antigenic properties of SC and J chain do not appear to be homologous to those of the immunoglobulin polypeptides; i.e., they do not seem to have a common evolutionary origin.

In humans and in many other species there...