An abnormal but functionally active complement component C9 protein found in an Irish family with subtotal C9 deficiency

Abstract

Two independently segregating C9 genetic defects have previously been reported in two siblings in an Irish family with subtotal C9 deficiency. One defect would lead to an abnormal C9 protein, with replacement of a cysteine by a glycine (C98G). The second defect is a premature stop codon at amino acid 406 which would lead to a truncated C9. However, at least one of two abnormal proteins was present in the circulation of the proband at 0.2% of normal C9 concentration. In this study, the abnormal protein was shown to have a molecular weight approximately equal to that of normal C9, and to carry the binding site for monoclonal antibody (mAb) Mc42 which is known to react with an epitope at amino acid positions 412-426, distal to 406. Therefore, the subtotal C9 protein carries the C98G defect. The protein was incorporated into the terminal complement complex, and was active in haemolytic, bactericidal and lipopolysaccharide release assays. A quantitative haemolytic assay indicated even slightly greater haemolytic efficiency than normal C9. Epitope mapping with six antihuman C9 mAbs showed the abnormal protein to react to these antibodies in the same way as normal C9. However, none of these mAbs have epitopes within the lipoprotein receptor A module, where the C98G defect is located. The role of this region in C9 functionality is still unclear. In conclusion, we have shown that the lack of a cysteine led to the production of a protein present in the circulation at very much reduced levels, but which was fully functionally active.

Figure 1

Molecular composition of C9. Amino acid numbers, organization of modules, location of the two defects (indicated by arrows) and schematic representation of the two arrangements of disulphide bonding patterns within the LDL-A module of C9 (Cys98 may form a disulphide bridge with Cys 113 or Cys 104) are shown (modified from Lengweiler et al.11). The modules of C9 are: TSP, thrombospondin repeat; LDL-A, low density lipoprotein class A module; MACPF, membrane attack/perforin segment; EGF, epidermal growth factor domain; … represents the intramolecular disulphide bridges. Note that one-quarter of the 12 intramolecular disulphide bridges are located within the LDL-A domain representing approximately 7% of the coding sequence.

Figure 2

Results of SDS-PAGE and Western blotting of purified preparations of normal C9 and C9SD probed with two of the anti-C9 mAbs. In all gels: track 1, C9SD, track 3, normal C9, and tracks 2 and 4 empty. Gels (a) and (b), unreduced and reduced samples, respectively, probed with mAb 11.60. Gels (c) and (d), unreduced and reduced samples, respectively, probed with mAb Mc47. mAb 11.60 reacts with monomeric and polymerized/aggregated C9, whereas Mc47 reacts almost exclusively with monomeric C9. The positions of molecular weight markers and normal C9 are indicated.

Figure 3

Functional haemolytic activity of purified C9 and C9SD preparations. Haemolytic activity was measured by the optical density (OD) of the test samples at 415 nm.

Figure 4

Release of [³H]-labelled LPS from E. coli J5 (2877) by NHS and C9SD serum over 1 hr. Release of LPS by HIS and C9D serum lacking all C9 functional activity, included for comparison.