Search

The molecular components of complement are a major part of the armoury of the mammalian immune system, being required for the lysis of antibody-targeted cells. Several of the complement proteins are known to be encoded by genes within the major histocompatibility complex (MHC). Molecular analysis of these genes is providing new information on the basis of complement action and the possible roles of this system in autoimmune disease.

Four cosmid clones, each with an average insert size of 40 kilobase pairs and containing the factor B gene, were isolated from a human genomic DNA library. The clones were identified by hybridization with a 515-base-pair cDNA probe isolated by using a unique 17-base synthetic oligonucleotide probe from a human liver cDNA library. The cosmid clones were characterized by restriction endonuclease digestion and Southern blotting, and a partial restriction map of the DNA represented in the cosmids was constructed. The Bb portion of the factor B gene is about 4 kb in length. DNA sequence analysis has resulted in the determination of 3.3 kb of sequence at the 3' end of the gene. This region codes for amino acids 87-505 of Bb and includes the whole of the serine proteinase domain of the protein. The three active site residues of histidine, aspartic acid, and serine found at positions 267, 317, and 440 of the Bb sequence, respectively, lie on separate exons. Other functional regions within the serine proteinase domain are separated also by intervening sequences.

Radioimmune assays were developed to assay the binding of complement components C1q, C1s and C4 to antibody aggregates and to cell-bound antibody. The binding of the components was compared with the haemolytic activity and with the capacity to form the C3 convertase activity in the presence of excess C2. The destruction of whole complement and of C4 activity is similar per 1,000 molecules of antibody in aggregates and cell-bound antibody, as is the binding of C1g and C1s, the latter being in a 1:2 molar ratio. The binding of C4 is about 12 times greater, per 1,000 molecules of antibody, on cells than in aggregates. However, the effective C4 molecules, as judged by the formation of C3 convertase activity, are much more similar on cells and aggregates. An assembly mechanism of the early components of complement on antibody-coated cells, which is compatible with these results, is suggested.

Human complement component C4 is coded by two genes situated between HLA-D and HLA-B. Both genes are highly polymorphic; C4-A gene products normally carry the blood group antigen Rodgers and C4-B proteins usually carry the Chido antigen. Using a monoclonal antibody which binds Rodgers-positive and Chido-positive proteins with different affinities, we have purified a number of less common C4 allotypes and compared their properties. All C4-B allotypes tested have similar specific hemolytic activities and binding efficiencies to small molecules. All C4-A proteins tested had similar binding to small molecules and hemolytic activities except for the C4-A6 proteins from two individuals with different extended haplotypes, both of which had identical hemolytic activities and much lower ones than other C4-A allotypes. Two allotypes, C4-A1, Rodgers-negative but Chido-positive, and C4-B5, Chido-negative but probably Rodgers-positive, were found to behave as typical C4-A and C4-B proteins, respectively, apart from the switch in their antigenic properties.

The heavy chain of a pathological human immunoglobulin IgG and also the Fd fragment have been isolated. No free alpha-amino group was present on either and the N-terminal sequence of both has been identified as pyrrolid-2-one-5-carbonylvalylthreonine. Splitting at the four methionine residues of the heavy chain with cyanogen bromide gave five fractions. The fraction from the C-terminal end of the chain was isolated in high yield and the amino acid sequence was: His-Glu-Ala-Leu-His-Asp(NH(2))-His-Tyr-Thr-Glu(NH(2))-Lys-Ser-Leu-Ser-Leu-Ser-Pro-Gly These results give strong support to the view that the heavy chain of immunoglobulin is a single peptide chain.

Fragment C-1, the N-terminal half of the heavy chain of rabbit immunoglobulin G, was prepared by cyanogen bromide cleavage from the heavy chain of immunoglobulin G obtained both from the pooled serum of normal rabbits and from specific anti-dinitrophenyl antibody. Tryptic digestion of fragment C-1 after the lysine residues had been allowed to react with S-ethyl trifluorothioacetate led to the isolation of six peptides from inert immunoglobulin G and specific antibody that appear to account for most of this section of the heavy chain. This approach should make possible comparative sequence studies of the Fd section of the heavy chain from different allotypes and from specific antibodies.

The partial amino acid sequence of two large peptides is described. These were prepared from the N-terminal half of the heavy chain of immunoglobulin G from pooled normal rabbit serum by tryptic digestion after the ∈-amino groups of the lysine residues had been blocked with S-ethyl trifluorothioacetate. These peptides are believed to account for about 145 residues of fragment C-1, the N-terminal section of rabbit immunoglobulin G heavy chain prepared by cyanogen bromide cleavage. The evidence from the present paper and the preceding paper (Cebra, Givol & Porter, 1968) suggests that it may be possible to deduce a predominant amino acid sequence for most, if not all, of this section of the molecule.

The complement system has two pathways of activity, both dependent on the sequential conversion of proteolytic zymogens to active proteases leading to a common lytic complex and both with control proteins which inhibit or inactivate different steps in the cascade. Three of the components--C2, factor B and C4--are coded by closely linked genes in the MHC of man and mouse and have been placed relative to each other. The genes are polymorphic, particularly C4, with variable numbers of loci as well as many mutant forms. Some alleles of C4 show strikingly different reactivities in their haemolytic activity and this may be relevant to the association of susceptibility to autoimmune diseases with particular haplotypes in this section of HLA.

The structure of immunoglobulin raises several obvious questions about the structural genes coding for their peptide chains.

An alternative method of isolation from human plasma is described for C3b inactivator, C3bINA, the proteinase that in conjunction with either beta 1H or C4b-binding protein will hydrolyse respectively C3b or C4b, the activation products of the third, C3 and fourth, C4, components of complement. The purification is by chromatography of plasma on columns of QAE-Sephadex, wheat-germ agglutinin-Sepharose, hydroxyapatite and Sephacryl S-200. The yield of C3bINA (6 mg from 500ml of plasma) is severalfold higher than in previously described methods. The sensitivity of the assay for C3bINA has been increased by including optimal amounts of beta 1H, and it was observed that beta 1H was essential for hydrolysis by C3bINA of C3b, whether the C3b was in solution or bound to a cell surface. Native C3 is not hydrolysed by C3bINA + beta 1H, but the haemolytically inactive form that appears on prolonged storage at 4 degrees C or on freezing and thawing is hydrolysed and gives fragments of the alpha-chain of 75000 and 43000 apparent mol.wt. As the alpha'-chain of C3b, which has lost an N-terminal peptide C3a, gives fragments of 67000 and 43000 apparent mol.wt. when incubated with C3bINA + beta 1H, this suggests that the larger fragment is N-terminal and the smaller one C-terminal. The pH optimum of C3bINA with soluble substrates is 6.0, but no clear classification of the type of proteinase to which this enzyme belongs has been obtained.

Polymorphism of complement components, recognized by differences in either their antigenic specificity or their electrophoretic mobility, together with studies of inherited deficiencies, has enabled many of their structural genes to be mapped. In humans, three genes (for C2, C4, and factor B) have been placed between HLA-D and HLA-B on chromosome 6 and in mice, C4 between H2-I and H2-D, chromosome 17. Structural studies show that these components have exceptional features. C2 and factor B which contain the proteolytic active site of the C3 and C5 convertases are of the classical and alternative pathway respectively and are similar in structure and function. Both are novel types of serine proteases. C4 (as C3) contains an intrachain thioester bond essential for hemolytic activity. Molecular genetic investigations are determining the relative positions of these genes, and their precise structure, and should clarify their relation to the inherited diseases which are associated with defects in this section of the human genome.

A remarkable difference has been observed between the reactivity of the two forms of human complement component C4. C4B binds twice as effectively as C4A to antibody-coated red cells, but the reverse occurs with protein-antigen complexes. C4B reacts much more effectively with hydroxyl groups than C4A and this is reversed for reaction with amino groups in spite of the very small difference in amino acid sequence between the two forms of C4. No other differences in stability, activation or inactivation were observed. These findings emphasise the biological advantage of the duplication of the C4 gene in its reaction with a wide range of antigenic structures. The correlation of the presence of different forms of C4 with susceptibility to autoimmune diseases may be explicable by these big differences in binding reactivity.

The heavy chain of rabbit immunoglobulin G exists in three major allotypic patterns, Aa1–Aa3. A comparison of the amino acid compositions of the heavy chains isolated from immunoglobulin IgG homozygous for each allotypic determinant revealed the presence of an additional methionine residue per chain in the Aa3 allotype relative to the Aa1 and Aa2 allotypes. The position of the additional methionine residue was determined by cyanogen bromide cleavage and by tryptic digestion of the γ-chains; it coincided with the inter-Fd–Fc area of the chain. Isolation and characterization of the corresponding tryptic peptides of 31 amino acid residues from each of the allotypes showed the presence of a methionine-for-threonine replacement in the Aa3 allotype, but only in about 70–80% of the molecules. No other allotypic variations were seen in this tryptic peptide. Allotypically related variations in composition were also detected in the N-terminal cyanogen bromide-cleavage peptide.

THE method developed by Mishell and Dutton1 for the in vitro stimulation of an apparently primary response to red cell antigens offers advantages over in vivo experiments in analysing the cellular basis of the immune response. The possible uses of the in vitro system would be much increased were it possible to get reproducible responses to a well defined hapten.

Although an antibody molecule possesses many biological properties essential to its role in immune reactions, the one on which all others depend is its capacity to bind specifically with the antigen, the injection of which has stimulated its synthesis. Increasing knowledge of the structure of immunoglobulins—the group of proteins to which all antibodies belong—has made it seem even more remarkable that molecules of identical basic structure should be able to combine specifically with such a wide range of different substances.

Hapten-antibody complexes prepared at equivalence with the bivalent hapten bis-DNP-octamethylene-diamine and purified rabbit anti-DNP antibody were fractionated by Sepharose gel-filtration and the fractions examined by electron microscopy. Individual fractions were tested for whole-complement fixation and C1 fixation. Dimer forms did not show this type of biological activity, while fractions containing tetramers and larger polymers exhibited both C and C1 fixation, which could be inhibited by prior exposure of the complexes to the univalent hapten epsilon-DNP-caproic acid. The dose-response result indicated that the C-fixation observed was not due to interpolymeric cooperative effects. It was concluded that in the generation of biological activity by soluble antigen-antibody complexes made with complement-fixing antibody, quaternary structural changes following specific combination with antigen may be as important as any tertiary structural alterations that occur in the individual immunoglobulin molecule.