The membrane attack mechanism of complement: the three polypeptide chain structure of the eigth component (C8)

A selection signatures study among Middle Eastern and European sheep breeds

Selection, both natural and artificial, leaves patterns on the genome during domestication of animals and leads to changes in allele frequencies among populations. Detecting genomic regions influenced by selection in livestock may assist in understanding the processes involved in genome evolution and discovering genomic regions related to traits of economic and ecological interests. In the current study, genetic diversity analyses were conducted on 34,206 quality-filtered SNP positions from 450 individuals in 15 sheep breeds, including six indigenous breeds from the Middle East, namely Iranian Balouchi, Afshari, Moghani, Qezel, Karakas and Norduz, and nine breeds from Europe, namely East Friesian Sheep, Ile de France, Mourerous, Romane, Swiss Mirror, Spaelsau, Suffolk, Comisana and Engadine Red Sheep. The SNP genotype data generated by the Illumina OvineSNP50 Genotyping BeadChip array were used in this analysis. We applied two complementary statistical analyses, F_ST (fixation index) and xp-EHH (cross-population extended haplotype homozygosity), to detect selection signatures in Middle Eastern and European sheep populations. F_ST and xp-EHH detected 629 and 256 genes indicating signatures of selection, respectively. Genomic regions identified using F_ST and xp-EHH contained the CIDEA, HHATL, MGST1, FADS1, RTL1 and DGKG genes, which were reported earlier to influence a number of economic traits. Both F_ST and xp-EHH approaches identified 60 shared genes as the signatures of selection, including four candidate genes (NT5E, ADA2, C8A and C8B) that were enriched for two significant Gene Ontology (GO) terms associated with the adenosine metabolic procedure. Knowledge about the candidate genomic regions under selective pressure in sheep breeds may facilitate identification of the underlying genes and enhance our understanding on these genes role in local adaptation.

Molecular cloning of the alpha subunit of complement component C8 (CpC8α) of whitespotted bamboo shark (Chiloscyllium plagiosum)

Complement-mediated cytolysis is the important effect of immune response, which results from the assembly of terminal complement components (C5b-9). Among them, α subunit of C8 (C8α) is the first protein that traverses the lipid bilayer, and then initiates the recruitment of C9 molecules to form pore on target membranes. In this article, a full-length cDNA of C8α (CpC8α) is identified from the whitespotted bamboo shark (Chiloscyllium plagiosum) by RACE. The CpC8α cDNA is 2183 bp in length, encoding a protein of 591 amino acids. The deduced CpC8α exhibits 89%, 49% and 44% identity with nurse shark, frog and human orthologs, respectively. Sequence alignment indicates that the C8α is well conserved during the evolution process from sharks to mammals, with the same modular architecture as well as the identical cysteine composition in the mature protein. Phylogenetic analysis places CpC8α and nurse shark C8α in cartilaginous fish clade, in parallel with the teleost taxa, to form the C8α cluster with higher vertebrates. Hydrophobicity analysis also indicates a similar hydrophobicity of CpC8α to mammals. Finally, expression analysis revealed CpC8α transcripts were constitutively highly expressed in shark liver, with much less expression in other tissues. The well conserved structure and properties suggests an analogous function of CpC8α to mammalian C8α, though it remains to be confirmed by further study.

Glycyrrhizin inhibits the lytic pathway of complement--possible mechanism of its anti-inflammatory effect on liver cells in viral hepatitis

Glycyrrhizin, an aqueous extract of licorice root, has anti-inflammatory activity and has been used for the treatment of chronic viral hepatitis. In the present study we describe the mechanism by which glycyrrhizin inhibits complement. Glycyrrhizin inhibited the cytolytic activity of complement via the activation of both the classical and alternative pathways, while it had no effect on immune adherence, suggesting that it blocks C5 or a later stage of the complement cascade. Further analysis revealed that glycyrrhizin inhibits the lytic pathway in which the membrane attack complex (MAC) is formed. This mechanism suggests that glycyrrhizin may prevent tissue injury caused by MAC not only in chronic hepatitis but in many autoimmune and inflammatory diseases.

Genetic Basis of Human Complement C8α-γ Deficiency

Deficiency of the α-γ subunit of the eighth component of complement (C8α-γD) is frequently associated with recurrent neisserial infections, especially meningitis caused by Neisseria meningitidis. We here report the molecular basis of C8α-γD in two unrelated Japanese subjects. Screening all 11 exons of the C8α gene and all 7 exons of the C8γ gene and their boundaries by exon-specific PCR/single-strand conformation polymorphism demonstrated aberrant single-stranded DNA fragments in exon 2 of C8α gene in case 1 and in exons 2 and 9 of C8α gene in case 2. Nucleotide sequencing of the amplified DNA fragments in case 1 revealed a homozygous single-point mutation at the second exon-intron boundary, inactivating the universally conserved 5′ splice site consensus sequence of the second intron (IVS2+1G→T). Case 2 was a compound heterozygote for the splice junction mutation, IVS2+1G→T, and a nonsense mutation at Arg394 (R394X). R394X was caused by a C to T transition at nucleotide 1407, the first nucleotide of the codon CGA for Arg394, leading to a stop codon TGA. No mutations were detected in the C8γ gene by our method. Our results indicate that the pathogenesis of C8α-γD might be caused by heterogeneous molecular defects in the C8α gene.

Molecular, genetic, and functional analysis of homozygous C8 beta-chain deficiency in two siblings

C8 deficiency is associated with an increased susceptibility to neisserial infections. We present a case of an 11 year old boy who suffered from infection with Neisseria meningitidis. Medical history of the patient and his family (n = 5) did not indicate any previous immunodeficiency symptoms. Results from the analysis of phagocyte and lymphocyte functions were within the normal range. No hemolytic activities of the classical (CH50) and the alternative (APH50) pathways of complement were measurable, and SC5b-9 protein complexes could not be detected in the patient's plasma. Further analysis by highly sensitive ELISA and functional assays revealed a complete deficiency of C8. Upon the reconstitution with purified C8 total hemolytic activity could be restored. SDS-PAGE and Western blot analysis established a deficiency of the C8 beta chain. Genetic analysis at the genomic DNA level demonstrated the common C-T mutation in exon 9 of the C8B gene. Family analysis presented the older sister with non-detectable function of C8 in serum, both parents with about half-normal C8 titres, and the younger sister with normal C8 function. The parents and both sisters were asymptomatic, although the older of the sisters presented with the same complete C8 beta-chain deficiency as the patient described.

IN CONCLUSION

the common C-T mutation in the C8B genes is the genetic basis of C8 beta-chain deficiency in two members of this Bosnian family.

Molecular cloning of Limulus alpha 2-macroglobulin

The American horseshoe crab Limulus polyphemus contains alpha 2-macroglobulin (alpha 2M) in the hemolymph plasma and hemocytes. alpha 2M from Limulus shows many of the typical characteristics of mammalian alpha 2M, including the presence of an internal thiol-ester, reactivity with a diversity of endopeptidases, a unique proteinase-trapping mechanism, and reactivity with the mammalian alpha 2M receptor. Additionally, Limulus alpha 2M has the unique property that it regulates the limulin-based hemolytic system of the plasma. A cDNA encoding Limulus alpha 2M has been obtained from a hemocyte cDNA library. The open reading frame encodes an N-terminal signal sequence of 25 amino acid residues and a mature protein of 1482 residues. The entire amino acid sequence is similar to those of the mammalian alpha 2Ms (28-29% identity) and contains common features found in mammalian alpha 2Ms. a bait region, an internal thiol-ester site, and a receptor-binding domain. However, the N-terminal portion (positions 24-105) has no sequence similarity with those of mammalian alpha 2Ms, and it is structurally related to that of the human complement factor C8 chain, consistent with a role for Limulus alpha 2M in host defense. The component sugar analysis of Limulus alpha 2M showed the existence of a complex type of oligosaccharide chain similar to those of mammalian alpha 2M. However, unlike mammalian alpha 2M, no sialic acid was detected in Limulus alpha 2M and it contained approximately 3 mol/mol N-acetylgalactosamine, suggesting the presence of O-linked sugar chains, which have not been found in mammalian alpha 2M. Expression of alpha 2M was detected in hemocytes, but not in hepatopancreas, heart, stomach, intestine, coxal gland, brain and skeletal muscle. Furthermore, immunoblotting of large and small granules of the hemocytes with antiserum against alpha 2M indicated the presence of the alpha 2M in large granules. Trypsin-treated Limulus alpha 2M, but not the native alpha 2M, displaced methylamine-treated human 125I-alpha 2M from the human alpha 2M receptor with a Kd of 30 nM, suggesting conservation of the proteinase-clearance mechanisms between mammalian and arthropod evolutionary lineages.

Human complement component C8. Molecular basis of the beta-chain polymorphism

The beta-chain of human complement component C8 exhibits a structural genetic polymorphism: using isoelectric focusing two major allotypes can be identified (C8B B ('basic') and C8B A ('acidic')). In the present report we describe a sequence polymorphism of the C8B gene (codon 63: AGA-->GGA) and demonstrate that the resulting amino acid substitution (Arg-->Gly) consistently differentiates between the two common charge variants of the C8 beta chain; the C8B B allotype is characterized by an Arg and the C8B A allotype by a Gly residue in position 63 of the C8 beta polypeptide chain.

The human complement component C8B gene: structure and phylogenetic relationship

The eighth component of human complement (C8) is a serum protein that consists of three chains (alpha, beta and gamma), encoded by three separate genes, viz., C8A, C8B, and C8G. In serum, the beta-subunit is non-covalently bound to the disulfide-linked alpha-gamma subunit. Using a full-length C8 beta cDNA probe, we isolated several clones from human genomic lambda DNA libraries. Four lambda clones covering the complete cDNA sequence were characterized by TaqI restriction mapping and were "shotgun" subcloned into M13. C8 beta-cDNA-positive clones were partially sequenced to characterize the 12 exons of the gene with sizes from 69 to 347 bp. All intron-exon junctions followed the GT-AG rule. By using polymerase chain reaction (PCR) primers located in the adjacent intron sequences, all 12 exons of the C8B gene could be amplified from genomic DNA. All fragments showed the expected sizes. The sizes of eight introns could be determined by using primer pairs that amplified two exons and the enclosed intron, and by restriction mapping. These analyses and the insert sizes of the genomic lambda clones indicate that the C8B gene has a total size of approximately 40 kb. The polymorphic TaqI site of the C8B gene localized in intron 11 could be demonstrated by direct restriction fragment analysis of a PCR fragment containing exons 11 and 12, and the enclosed intron 11. Homology comparison of the C8B gene with C8A and C9 on the basis of the exon structure confirmed the ancestral relationship known from the protein level.

Genetic deficiency of complement component C8 in the rabbit: evidence of a translational defect in expression of the alpha-gamma subunit

We examined the molecular basis for rabbit C8 alpha-gamma deficiency (C8D) using human C8 cDNA probes. Sequential probing of normal rabbit poly(A)+RNA revealed messages of 2.1, 1.9, and 0.8 kb for alpha, beta, and gamma, respectively. Corresponding analysis of C8D rabbit poly(A)+RNA identified messages for alpha and gamma of the same size and amounts as normal rabbits. Thus, C8 alpha-gamma deficiency in the rabbit appears to be the result of a translational rather than a transcriptional defect.

Gene responsible for deficient activity of the beta subunit of C8, the eighth component of complement, is located on mouse chromosome 4

Sera from 73 strains of mice were tested for hemolytic activity through the classical and the alternative pathways (CP and AP) in a single radial hemolysis assay. Sera from 16 out of 45 laboratory inbred strains had no lytic activity, and Ouchterlony analysis with anti-C5 serum showed them to be C5-deficient. Sera from 2 out of 28 strains derived from wild mice also had no lytic activity, but the C5 molecule was detectable in both. The hemolytic activity of sera from these strains can be restored by the addition of partially purified human C8 or human serum deficient in C8 alpha-gamma, leading us to conclude that strains M.MOL-MSM (MSM) and Mae are deficient in the beta subunit of C8. Typing of (DBA/2J X MSM)F1 hybrids and of progeny of a backcross to MSM showed that this C8 deficiency is controlled by a single recessive gene, designated C8b; the allele with hemolytic activity is C8b1; and the allele with no activity C8b0. Because of synteny homologies in mouse and human, we looked for and found close linkage between C8b and Pgm-2. Typing of recombinant mice for Mup-1 mapped the C8b locus 2.3 centimorgans (cM) telomeric to Pgm-2 on mouse chromosome 4.

Structural and functional characterization of complement C8 gamma, a member of the lipocalin protein family

Human complement component C8 exhibits an unusual structure in that it contains three chains, two of which, alpha and beta, display high sequence homology to other complement and CTL pore-forming proteins. The third chain, C8 gamma, is covalently linked to C8 alpha by a disulfide linkage; it is demonstrated that Cys40 of C8 gamma is linked to Cys164 of C8 alpha, a unique cysteine located in a loop located between the cysteine-rich LDL-receptor class A module and the membrane-inserting region of C8 alpha. C8 gamma was recently identified as a member of the lipocalin protein family, in which all proteins were either shown to, or are believed to bind small hydrophobic ligands. The present results now demonstrate that C8 gamma incorporates retinol and retinoic acid in the presence of 2 M NaCl. Molecular modeling of C8 gamma, based on the crystal structure of the homologous beta-lactoglobulin, reveals a structure of eight antiparallel beta-strands, bearing a highly hydrophobic binding pocket. The residues participating in the pocket formation are highly conserved when compared with the structures of beta-lactoglobulin and retinol-binding protein, both of which are known to interact with retinol. It is therefore proposed that C8 gamma may act as a retinol transporting protein in plasma.

Prevalence of hereditary properdin, C7 and C8 deficiencies in patients with meningococcal infections

High incidence of hereditary complement (C) deficiencies was found among 101 patients who had a meningococcal disease. This study revealed 11 non-related patients with complete C deficiency: five deficient in C7, three in C8, two in properdin and one in C2. Additional C-deficient individuals, most of them with no history of severe bacterial infections, were detected in family studies. The C8-deficient patients were found to have a selective deficiency of the C8-beta subunit and a reduced expression of the alpha/gamma subunit. Only a few families with properdin deficiency have been described so far. However, it is likely that frequent analysis of the activity of the alternative C pathway in survivors of severe bacterial infections will disclose numerous properdin-deficient patients. All our C7-, C8- and properdin-deficient patients are Sephardic Jews whose families originated from Morocco, Yemen (C7 and C8 deficient) or Tunisia (properdin deficient). This and other findings indicate that the type of complement abnormality found in association with meningococcal infections varies with the ethnic origin of the patient.

Comparison of channels formed by poly C9, C5b-8 and the membrane attack complex of complement

The channels formed by poly C9, C5b-8 and C5b-9 were examined using the liposome swelling assay. By plotting the relative rate of swelling of C5b-8-containing liposomes vs the molecular weight of the sugar solute and by applying the Renkin equation, the size of the C5b-8 channel was estimated to be 1.5 mm radius. As increasing amounts of C9 were added during the formation of C5b-9, in C8:C9 ratios of 1:1, 1:2, 1:6 and 1:12, the size of the function channel increased. Poly C9 had a pore that was somewhat larger than C5b-9 at a C8:C9 ratio of 1:12. Using molecular sieving experiments with four different iodinated protein size markers, the channel diameter of poly C9 was estimated at between 90 and 100 A. Monoclonal antibodies to different complement proteins were added to the liposomes to see which might inhibit the channels. C5b-8 containing liposomes could be inhibited by antibodies to C8. Liposomes containing C5b-9 could be inhibited slightly by antibodies to C9 and most strongly by antibodies to the neoantigen of poly C9.

DNA polymorphism of the human complement C8 beta gene: formal genetics and intragenic localization

The eighth component of human complement consists of three subunits of different molecular mass, which are coded for by three separate genetic loci. Polymorphisms have been described at the protein level for the alpha and beta subunits by means of sodium dodecyl sulfate gel electrophoresis and isoelectric focusing. Using a full-length human C8 beta cDNA probe, we have studied more than 100 individuals by Southern blot analysis to detect DNA polymorphisms. We have found two restriction fragment length polymorphisms (RFLPs) with the enzymes Taq I and Bam HI. The Taq I polymorphism is defined by two alleles, i.e., a single 4.9 kb fragment or two 2.8/2.1 kb fragments. The allele frequencies are 0.68 and 0.32, respectively. The second RFLP with Bam HI is correlated with the Taq I variants: 3 kb Bam HI; 4.9 kb Taq I and 3.3 kb Bam HI; 2.8/2.1 kb Taq I. Both RFLPs could be mapped to the 3' portion of the C8 beta gene. Based on the size of genomic restriction fragments, the C8 beta gene can be estimated to have a size of 32-36 kb. Because of the even frequency distribution, the C8 beta DNA polymorphisms may be useful in gene mapping and disease association studies.

Biotinylation: a simple method for labelling complement component C8 with preservation of functional activity

Biotinylation of human C8 with the water-soluble biotin derivative biotinylamidohexanoic acid, N-hydroxysulfosuccinimide ester is an excellent method for labelling this terminal complement component with preservation of its functional activity. The biotinylated product can be detected both in native form and also following its incorporation into the terminal complement complexes. Detection assays include Western blotting, crossed immunoblotting, ELISA, and immunocytochemistry. Biotinylation is an attractive alternative method for labelling C8 and may be used for detecting and quantifying C8 and C5b-9 complexes in their soluble and membrane-bound forms.

SC5b-7, SC5b-8 and SC5b-9 complexes of complement: ultrastructure and localization of the S-protein (vitronectin) within the macromolecules

Purified terminal components of the complement system were used together with purified S-protein, the inhibitor of the membrane attack complex, to generate the soluble complexes SC5b-7, SC5b-8 and SC5b-9. These complexes were purified by ultracentrifugation in sucrose density gradients with 50-70% yield, exhibiting sedimentation coefficients of 20 S, 21 S and 23 S, respectively. In Ouchterlony double-diffusion analysis, the purified complexes gave a line of identity against all antisera of the precursor components indicating that complex formation had occurred. The identity of the complexes was also revealed by the appearance of all subunit components after polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate. Since the inhibitor function of S-protein in the terminal complement cascade should also be manifested in the morphology of the macromolecules generated, the ultrastructures of the three complexes were analyzed by electron microscopy. In contrast to aggregated (C5b-7)n and (C5b-8)n, negatively stained SC5b-7 and SC5b-8 imaged mostly as monomeric irregularly shaped cylindrical structures, whereas SC5b-9 less than 27 S) appeared as wedge-shaped structure lacking the tubular polymerized C9. (All three complexes were also generated in the presence of biotinyl-S-protein and labeled with avidin-gold conjugates as electron-dense marker). Analysis of the modified complexes in electron micrographs demonstrated that the complexes were marked exclusively at one site of their ultrastructures, suggesting this region to be the location of S-protein and the critical site for membrane binding of C5b-7 or C5b-8 and for initiation of C9 polymerization. These results support recent findings in which the function of S-protein as complement inhibitor was dependent on conformational changes of the protein molecule with concomitant exposure of the heparin-binding domain.

Induction of expression of cell-surface homologous restriction factor upon anti-CD3 stimulation of human peripheral lymphocytes

Homologous restriction factor (HRF) is a 65-kDa membrane protein that inhibits transmembrane channel formation by the membrane-attack complex of complement and by the complement component C9-related cytolytic lymphocyte protein. Stimulation of resting peripheral human lymphocytes with the anti-CD3 monoclonal antibody OKT3 has been shown to induce cytotoxicity in the CD8+ subpopulation. As demonstrated here, OKT3 stimulation also induces expression of cell-surface HRF by CD4+ and CD8+ T lymphocytes. The small proportion of Leu 19+ natural killer lymphocytes present in peripheral blood mononuclear cells was found to express HRF prior to stimulation. Whereas unstimulated peripheral blood mononuclear cells were susceptible to lysis by the membrane-attack complex or by the C9-related protein, OKT3-stimulated peripheral blood mononuclear cells were relatively resistant to both the membrane-attack complex and C9-related protein. This acquired resistance was abrogated by blocking surface HRF with F(ab')2 anti-HRF, suggesting that resistance was due to lymphocyte-membrane HRF. By using solid-phase anti-HRF, a 65-kDa protein was isolated from the activated peripheral blood mononuclear cells and shown to be capable of conferring upon sheep erythrocytes the characteristic activity of human HRF.

Structural homology of human complement component C8 gamma and plasma protein HC: identity of the cysteine bond pattern

Anti-C8 alpha-gamma specific antibodies were used to isolate cDNA clones from a human liver expression library. Antibodies affinity-purified on the expressed hybrid protein of one clone bound exclusively to the gamma-chain of reduced C8 alpha-gamma. This clone, as well as a second full length cDNA clone obtained by hybridization screening, were sequenced and the complete primary structure for C8 gamma was established. Cyanogen bromide cleavage of C8 alpha-gamma released a 12 kDa carboxy-terminal C8 gamma fragment under both reducing and nonreducing conditions which was identified by fragment-specific, affinity-purified antibodies. Our data clearly show that C8 gamma has one internal disulfide bridge between cys-76 and cys-168 within the carboxy-terminal 12 kDa fragment, whereas the remaining cysteine residue 40 forms the disulfide bridge with C8 alpha. The overall sequence homology to plasma protein HC (23% amino acid identities) and the conservation of one internal cysteine bond and one free, surface-located cysteine residue suggests a highly conserved three-dimensional structure of C8 gamma and protein HC and also a possible functional relationship between these proteins.

Regulation of the alternative pathway of human complement by C1q

The interaction of C1q with C3b and its effect on C3b activities in the alternative pathway of complement (APC) have been studied. Purified C1q markedly inhibited C3b deposition on and lysis of rabbit erythrocytes by the isolated cytolytic APC. It also blocked formation of the C3 convertase, C3b, Bb as well as binding of Factors B and H to sheep erythrocytes (E) bearing C3b. The direct and specific binding of C1q to C3b was clearly demonstrated using the hemagglutination technique at low ionic strength (0.1 M NaCl). C1q concns of 2 micrograms/ml and higher agglutinated, in a dose-dependent fashion, EC3b but not E, EC3bi or EC3d. Addition of C1r and C1s to C1q and formation of C1 did not affect its capacity to agglutinate EC3b. The C1q-mediated agglutination of EC3b was inhibited by EDTA, MgEGTA, C3b and Factor B but not by native C3 or collagen. Heating C1q (56 degrees C) markedly potentiated its agglutinating activity whereas collagenase-treated C1q lost most of its activity. Taken together, these results suggest that C1q binds through its "heads" and in the presence of calcium ions to a site on C3b that is adjacent to the Factor B and Factor H binding sites. This interaction may down-regulate the activity of the alternative pathway of complement on surfaces which activate both the classical and alternative pathways of complement.

The eighth component of human complement: evidence that it is an oligomeric serum protein assembled from products of three different genes

The eighth component of human complement (C8) consists of three nonidentical subunits arranged asymmetrically as a disulfide-linked alpha-gamma dimer and a noncovalently associated beta chain. Genetic studies of C8 polymorphisms established that alpha-gamma and beta are encoded at different loci. Implicit in this finding was the existence of two different genes and the likelihood that alpha-gamma would be synthesized in single-chain precursor form. However, recent characterization of cDNA clones revealed separate mRNAs for human alpha and beta but no evidence of a single-chain precursor for alpha-gamma. A cDNA clone containing the entire coding region for human gamma has now been characterized, and its sequence supports the existence of a separate gamma mRNA. Included are a consensus translation initiation sequence, an apparent initiation methionine, and a signal peptide. By use of cDNA probes specific for human alpha, beta, or gamma, analysis of poly(A) RNA from normal baboon liver revealed separate mRNAs of 2.5, 2.6, and 1.0 kilobases (kb), respectively. Parallel analysis of poly(A) RNA from rat liver identified mRNAs of 3.4, 2.3, and 0.9 kb. These results argue against the possibility that C8 is assembled from products of two different genes (alpha-gamma and beta) and suggest it is comprised of three different gene products (alpha, beta, and gamma) that undergo both covalent and noncovalent association to yield the mature protein.

Killing of human melanoma cells by the membrane attack complex of human complement as a function of its molecular composition

The efficiency of the membrane attack complex (MAC) in killing M21 melanoma cells was determined varying the molar ratio of cell-bound C9:C8. It was found that C5b-8 produced functional channels as evidenced by 86Rb release and propidium iodide uptake; cell killing occurred in the absence of C9 with greater than 5 X 10(5) C5b-8/cell; the maximal molar ratio of C9:C8 was 6.6:1; using nonlytic numbers of C5b-8 (4.7 X 10(5)/cell), greater than 90% killing ensued at a C9:C8 molar ratio of 2.8:1 at which approximately 9,000 poly C9/cell were formed, and 50% killing at a ratio of 1:1; (e) when the MAC was assembled on cells at 0 degree C, consisting of C5b-8(1)9(1), and unbound C9 was removed before incubation at 37 degrees C, killing was similar to that observed when poly C9 formation was allowed to occur. Thus, MAC lytic efficiency toward M21 cells may be enhanced by but does not depend on poly C9 formation.

The pore-forming protein (perforin) of cytolytic T lymphocytes is immunologically related to the components of membrane attack complex of complement through cysteine-rich domains

Structural, functional and immunological similarities between the ninth component of complement (C9) and the lymphocyte pore-forming protein (PFP, perforin) have recently been described (8-10). PFP is shown here to be immunologically related to all other components of the membrane attack complex (MAC) of human complement, namely, C5b-6, C7, C8, and C9. Polyclonal antibodies raised against purified human C5b-6, C7, C8, or C9 react with other components of the MAC and with mouse lymphocyte PFP. The antigenic epitopes shared by human complement proteins and mouse lymphocyte PFP are limited to cysteine-rich domains. Only complement proteins that have been reduced and alkylated elicit the production of crossreactive antibodies when used as immunogens. The nonreduced forms of complement components or lymphocyte PFP neither react with these antibodies nor give rise to crossreactive antibodies. The homologous domains of complement proteins and lymphocyte PFP may play related functions in their attachment to lipid membranes and assembly of membrane lesions.

Dysfunctional C8 beta chain in patients with C8 deficiency

Two sera from unrelated individuals, each lacking C8 activity, were examined by Western blot analysis. Using antisera raised against whole C8, the two sera are shown to lack the C8 beta chain, indicating a C8 beta deficiency, which is frequently observed in cases of dysfunctional C8. In contrast, by means of a specific anti-C8-beta antiserum, a C8 beta-like polypeptide chain of apparently identical molecular weight compared to normal C8 beta was detected. Digestion of normal and dysfunctional C8 beta with Staphylococcus aureus V8 protease revealed distinct differences in the enzymatic digestion pattern. We conclude that the dysfunction in the C8 protein in these two patients resides in the dysfunctional C8 beta chain, and that this form of C8 deficiency is distinct from C8 deficiencies previously reported, in which one or both C8 subunits are lacking.

Isolation of a human erythrocyte membrane protein capable of inhibiting expression of homologous complement transmembrane channels

Erythrocytes are poorly lysed by homologous complement, whereas they are readily lysed by heterologous complement. This phenomenon had been attributed to an interference by the cell surface with the action of complement components C8 and C9. To isolate the responsible membrane constituent, detergent-solubilized human erythrocyte (EH) membranes were subjected to affinity chromatography by using human C9-Sepharose. The isolated protein had a mass of 38 kDa and, incorporated into liposomes, was highly effective in inhibiting complement-mediated channel expression, including the C5b-8, membrane attack complex, and tubular polymer of C9 channels. Antibody produced to the 38-kDa protein caused a 20-fold increase in reactive lysis of EH by isolated C5b6, C7, C8, and C9. The antibody did not enhance C5b-7 uptake, but it affected C9 binding to the target cell membrane. Antibody to human decay-accelerating factor, used as a control, had no effect on reactive lysis of EH. Anti-38-kDa protein did not enhance the action on EH of C8 and C9 from other species, indicating that the action of this regulatory protein is species specific. It was therefore termed homologous restriction factor (HRF). Blood cells other than erythrocytes, such as polymorphonuclear leukocytes, also exhibited cell-surface HRF activity. In immunoblots of freshly isolated EH membranes, anti-38-kDa HRF detected primarily a 65-kDa protein, suggesting that the 38-kDa protein constitutes an active fragment of membrane HRF. Because of the specific binding reaction observed between HRF and C8 or C9, HRF was tested with anti-human C8 and anti-human C9. A limited immunochemical relationship of HRF to C8 and C9 could be established and solid-phase anti-C9 proved an efficient tool for the isolation of HRF from solubilized EH membranes.

The terminal complement complex in sera deficient in the eighth component of complement (C8)

The terminal complement complex (TCC) was quantified in sera from patients with a genetic deficiency of C8 alpha-gamma or C8 beta. The individual sera contained only trace amounts of TCC compared with a normal serum pool. The content of TCC increased after mixing the two sera, which was consistent with reconstitution of C8 activity. Only a moderate increase in TCC was obtained after zymosan activation of the individual sera, whereas activation of the mixture resulted in high amounts of TCC. C8 was demonstrated in the TCC of both deficient sera. These results may indicate that functional C8 is present in trace amounts despite the genetic deficiency, and that the terminal pathway may function to some extent although not enough to be detectable in less sensitive assays.

Effect of the nephrotic syndrome on the concentration of serum complement components

The concentration of 12 component and four control proteins of the complement system was measured in serum from 43 children with a nephrotic syndrome, which subsequently proved to be steroid-responsive, and from 13 children with focal glomerulosclerosis (FGS) and was compared with values from 197 normal subjects. Of classical pathway complement components, 40% of patients had low C1q levels and 20%, low C2 levels. Mean serum levels of C1s, C4, C1INH, and C4bp were elevated. Of alternative pathway components, factors B and I were low in one third, while levels of C3 and H were commonly elevated. Of the terminal components, only C8 and C9 were low. In five patients with FGS with hypoalbuminemia without edema, all component levels were normal. With the exception of C1q, C1s, and C8, high molecular weight (mol wt) components were in high concentration and low mol wt components in low concentration. The three exceptions may be explained by the subunit structure of C1 and C8. From a practical standpoint, the study indicates that edematous patients with a nephrotic syndrome may have low serum levels of C1q and C2, simulating classical pathway complement activation such as commonly occurs in glomerulonephritis. However, low levels of C4, and possibly C1s, can be used as indicators of classical pathway activation since their levels are not reduced by a nephrotic syndrome.

Antigenic crossreactivity of the alpha subunit of complement component C8 with the cysteine-rich domain shared by complement component C9 and low density lipoprotein receptor

Complement component C9 contains two distinct cysteine-rich domains exhibiting high sequence resemblance to a domain present in the low density lipoprotein (LDL) receptor and epidermal growth factor precursor, respectively. Antibodies were raised against a peptide corresponding to the most conserved region of the LDL receptor/C9-homology segment. The antibodies were shown by immunoblotting to bind specifically to C9 but also to crossreact with C8 alpha, the alpha subunit of complement component C8. Moreover, a monoclonal antibody to a neoantigen present in polymerized C9 bound to an epitope exposed on C8 within the C5b-8 complex but buried in monomeric C8, suggesting that C8 and C9 undergo similar conformational changes during membrane-attack-complex assembly. Isolated C8 alpha-gamma exhibited the propensity to polymerize in the presence of Zn2+ and urea, as already demonstrated for C9. These data indicate that C8 alpha is closely related, both structurally and functionally, to C9.

Human C81 (alpha-gamma) polymorphism: detection in the alpha-gamma subunit on SDS-PAGE, formal genetics and linkage relationship

The molecular basis of human C81 (alpha-gamma) polymorphism could be elucidated by immunoprecipitation of human C81 allotypes and separation of the alpha-gamma and beta subunits on sodium dodecylsulfate polyacrylamide gel electrophoresis (SDS-PAGE) under nonreducing conditions. If the C8 molecules were completely reduced, C81 polymorphism was no longer detectable on SDS-PAGE. It is concluded that C81 variation depends on charge rather than molecular weight differences. Four C81 allotypes, the common A and B and two rare allotypes provisionally named A2 and B1, could be distinguished. The rare allotype A1 as detected by isoelectric focusing with subsequent C8 (alpha-gamma)-dependent functional overlay could no longer be visualized on SDS-PAGE. This allotype may therefore be elicited only in the intact C8 molecule. The beta-chain polymorphism named C82, probably also reflecting charge variation of the C8 molecule, could not be detected yet on SDS-PAGE. The distributions of C81 phenotypes and their respective allele frequencies were in good agreement with previously reported data. In the study of 30 families with 100 offspring, no deviation from the rule of at least four codominant alleles at one genetic locus was found. Linkage between C81 gene(s) and PGM1a encoded on chromosome 1 could be confirmed. The following estimates were obtained: (formula; see text) with S theta being the standard error of the maximum likelihood estimate theta. The new technique for allotyping human C81 at the subunit may provide a new tool for the differentiation of qualitative and quantitative variation of the eighth component of human complement.

Genetic polymorphism of human complement component C81 in the Japanese population

Genetic polymorphism of human C81 has been investigated using polyacrylamide gel isoelectric focusing (PAGIEF) in the presence of 3.1 M urea followed by electroblotting with enzyme immunoassay. In 448 individuals phenotypes of C81 were classified into three common and four rare patterns, and these were considered to be controlled by two common alleles, C81 A and C81 B, and three rare alleles which were tentatively designated C81 A1J and C81 A2J for acidic variants and C81 B1J for the basic variant. The alleles of C81 A2J and C81 B1J are new rare alleles, but C81 A1J might correspond to C81 A1 in the former studies. Family data were in accordance with the hereditary rules. The gene frequencies were estimated as C81 A is 0.6228, C81 B is 0.3672, C81 A1J is 0.0078, C81 A2J is 0.0011, and C81 B1J is 0.0011, respectively. The gene frequencies of the two common alleles agreed approximately with other ethnic groups. PAGIEF of neuraminidase-treated plasma samples followed by electroblotting with enzyme immunoassay is applicable to the study of heterogeneity of C81.

Complement C8 deficiency with recurrent meningococcemia: examination of meningococcal opsonization

A 10 year old girl presented with recurrent febrile episodes over 2 months. A non-typable strain of Neisseria meningitidis was grown from blood cultures on three occasions. She was found to lack functional C8 activity in serum but material with C8 antigenic activity was present. The opsonic activity of the C8 deficient serum for N. meningitidis was equivalent to that of normal controls.

Genetic polymorphism of complement component C8

Extensive genetic polymorphism of complement component C8 was demonstrated by isoelectric focusing of serum or plasma samples followed by immunoblotting procedures. Using these methods, we could detect both alpha-gamma (C81) and beta (C82) chain polymorphisms in the same gel. Two-dimensional (2D) electrophoresis of C8 immunoprecipitates was used to obtain further information of the C8 patterns. Evidence was obtained that the C81 polymorphism resides in the structural gene of the C8 alpha chain. Both C8 systems show autosomal, chiefly codominant inheritance, and the distribution of phenotypes agrees with the Hardy-Weinberg equilibrium. Our findings suggest at least five different alleles in the C81 system; the gene frequencies of the two most common ones, C81*A and C81*B being 0.59 and 0.39, respectively. In C82 we found evidence for at least three codominant alleles, the gene frequencies for the two most common ones, C82*B and C82*A being 0.94 and 0.05, respectively. In addition, family studies disclosed the existence of a null allele, C82*Q0.

Inherited C8 beta subunit deficiency in a patient with recurrent meningococcal infections: in vivo functional kinetic analysis of C8

A 16 year old with recurrent meningococcal infections is reported. Absence of haemolytic activity in both the classical and alternative pathways resulted from an absence of functional C8. Addition of functional C8 restored hemolytic activity. Antigenically deficient C8 was present in the serum and isoelectric focusing of serum confirmed the absence of the C8 beta chain. Following the infusion of fresh frozen plasma, we followed the decay in C8 functional activity as well as total haemolytic activity. C8 activity peaked at about 3 h with a half-life survival estimated to be 28 h. The kinetics of total haemolytic activity showed a slower decay with an exponential decline over 72 h and a half-life of 55 h. Fresh frozen plasma may be of value in the treatment of patients with C8 deficiency and acute Neisserial infections.

Population and formal genetics of the human C81(alpha-gamma) polymorphism

One hundred and ninety-six unrelated healthy individuals and 30 families with 75 offspring have been studied for the C81(alpha-gamma) polymorphism. The following allele frequencies were calculated: C81*A = 0.5536; C81*B = 0.4286; C81*A1 = 0.0178. Observed and expected phenotype frequencies were in a good agreement according to the Hardy Weinberg law. No exceptions from the mode of inheritance were found. In family W the segregation of the rare allele C81*A1 could be followed. Comparing the results of this study with previous data from Boston and Oslo, a combined technology including C8-dependent lysis and C8 structural variation is suggested for future investigations.

Membrane attack by complement

Membrane attack by complement involves the self-assembly on membranes of five hydrophilic proteins (C5b, C6, C7, C8 and C9) to an amphiphilic tubular complex comprising approximately 20 subunits. The hydrophilic-amphiphilic transition of the precursor proteins is achieved by restricted unfolding and exposure of previously hidden hydrophobic domains. Restricted unfolding, in turn, is driven by high-affinity protein-protein interactions resulting in the formation of amphilic complexes. Circular polymerization of C9 to a tubular complex (poly C9) constitutes the molecular mechanism for transmembrane channel assembly and formation of ultrastructural membrane lesions.

Genetic polymorphism in C8 beta-chains. Evidence for two unlinked genetic loci for the eighth component of human complement (C8)

Genetic polymorphism in the beta-subunit of the eighth component of human complement, C8, was defined by isoelectric focusing of serum in polyacrylamide gel in the presence of urea and development of specific patterns of hemolysis in an overlay gel containing antibody-sensitized erythrocytes and C8 beta-chain-deficient serum. Bands of hemolysis induced by serum from unrelated Caucasians suggested autosomal codominant inheritance of three structural alleles at a single locus, C82: C82 degrees A (acidic), C82 degrees B (basic), and C82 degrees A1 (very acidic) with frequencies of 0.952, 0.044, and 0.004, as well as the probable null allele C82 degrees Q0. The distribution of phenotypes agreed with the Hardy-Weinberg equilibrium. The previously described genetic polymorphism in human C8 defined with the use of "complete" C8 (C8 alpha-gamma-chain)-deficient serum was distinct from and independent of the inherited structural variation at C82. Therefore, the locus for C8 alpha-gamma-chains has been redesignated C81, and has the alleles C81 degrees A, C81 degrees A1, and C81 Q0. Linkage studies failed to show close linkage between the two loci for C8, C81, and C82, and between C82 and the major histocompatibility complex or C6.

Interaction of the eighth component of guinea pig complement (C8) with the membrane-bound C5b-7 complex. The binding site of C8 to C5b-7 is formed on association of the alpha-gamma subunit with the beta-chain

The eighth component (c8) of guinea pig complement consists of three polypeptide chains, the alpha-, beta-, and gamma-chains with M.W. of 60,000, 60,000, and 24,000, respectively. The alpha- and gamma-chains are bound by a disulfide bond(s) forming an alpha-gamma subunit, which is linked noncovalently to the beta-chain. The alpha-gamma subunit and the beta-chain were separated and purified from C8 by treatment with sodium dodecyl sulfate (SDS) and gel chromatography on Sephacryl S-300 in the presence of SDS. After removal of SDS, neither alpha-gamma nor beta showed the hemolytic activity of C8 when assayed independently, but showed significant activity in combination, indicating reconstitution of active C8. The recovery of hemolytic activity was 3.48%. When alpha-gamma and beta were incubated successively with EAC-7 with intervening reconstitution of active C8 on the cells was insignificant, irrespective of the order of the reactions. alpha-gamma and beta did not bind to EAC-7 when added separately, but after recombination 7% of alpha-gamma and 9% of beta bound to EAC-7 when EAC-7 was in excess. These results indicate that the binding site of guinea pig C8 to the membrane-bound C5b-7 complex does not exist on either alpha-gamma or beta only but stretches over both or is formed on one subunit after recombination of the subunits.

The membrane attack complex of complement and its precursor proteins lack phospholipase activity

The membrane attack complex of human complement and its highly purified precursor proteins have been analyzed for phospholipase activity. Using three different sensitive assays, phospholipase A1, A2, C or D activity could not be detected. Based on the sensitivity of the assays employed, these results indicate the complement-mediated membrane damage is not enhanced by covalent breakdown of membrane phospholipids, but is entirely caused by physical action of the membrane attack complex. The results also imply that the putative serine esterase sites of C6 and C7 are not acting on phospholipids.

Human monocyte spreading induced by activated factor B of the complement alternative pathway: differential effects of Fab' and F(ab')2 antibody fragments directed to C5, C6, and C7

Human peripheral blood mononuclear phagocytes are induced by activated Factor B (Bb) of the complement alternative pathway to undergo morphological shape changes in vitro which have been described as "spreading." The spreading reaction induced by Bb has previously been shown to depend upon the enzymatic activity of Bb and to be inhibited by Fab' antibody fragments directed to C5 (but not anti-C3 Fab'). The possibility that Bb may exert its effect on monocytes by initiating assembly of terminal complement complexes comprised of C5b, 6, 7, C5b-8, or C5b-9 was addressed in the present study. The effects were tested of Fab' and F(ab')2 antibody fragments directed to C5, C6, C7, and C8 and to neoantigens expressed in the assembling terminal complement complexes on the monocyte spreading reaction induced by Bb. Differential effects of monovalent Fab' and divalent F(ab')2 antibody fragments were observed. Anti-C5, C6, and C7 Fab' were found to inhibit the spreading reaction induced by Bb in an immunologically specific manner. Divalent F(ab')2 fragments directed to these same proteins (but not to C3, C4, C8, or C9) induced monocyte spreading in the complete absence of Bb or other recognized inducing agents. Monocyte spreading induced by hybridoma immunoglobulin (Ig) directed to C5 and C7 was found to be correlated with the binding of 10(6) molecules Ig per cell. These findings support the notion that C5, C6, and C7 (or an analogous system of cellular proteins) are associated with the surface of human peripheral blood monocytes and that these proteins may play a role in certain reactions by which mononuclear phagocytes are induced to altered states of cellular physiology.

Two types of dysfunctional eighth component of complement (C8) molecules in C8 deficiency in man. Reconstitution of normal C8 from the mixture of two abnormal C8 molecules

Restoration of hemolytic activity was examined in sera from seven unrelated eighth component of complement (C8)-deficient subjects. The sera fell into two groups, depending on whether hemolytic activity was restored by the addition of the beta-chain (group 1) or the alpha-gamma-subunit (group 2) purified from normal human C8. Antigenic analysis of these sera by double-immunodiffusion using anti-human C8 confirmed previous findings of a dysfunctional C8 in the four sera of group 1 and established the presence of a different dysfunctional C8 in one of the sera of group 2 when tested at a high concentration. Further characterization of the dysfunctional C8 molecules in the two sera by sodium dodecyl sulfate-polyacrylamide gel electrophoresis demonstrated that group 1 sera were missing the beta-subunit and group 2 sera were missing the alpha-gamma-subunit of the C8 molecule. Sera from either of these two groups alone did not produce hemolysis in hemolytic plates containing sheep erythrocytes coated with antibody and complement components up to C7 (EAC1-7) and C9. When sera from the two groups were added to adjacent wells in the hemolytic plates, a zone of hemolysis developed between the wells. The contribution of C8 alpha-gamma from the sera of group 1 and of C8 beta from those of group 2 to the lysis of EAC1-7 in the presence of C9 was confirmed by the inhibitory effect of specific antibodies against the two C8 subunits. In experiments in which hemolytic activity was reconstituted by mixing sera from group 1 with sera from group 2, the serum source of C8 beta (group 2) was the limiting reagent. The dysfunctional C8 molecule in this serum was able to bind to EAC1-7. Chromatographic analysis demonstrated that the generation of hemolytic activity in the mixture of the two sera resulted from the reconstitution of the C8 molecule rather than the sequential action of the two C8 subunits.

Beta chain deficiency in three patients with dysfunctional C8 molecules

Structural and functional studies were performed on a dysfunctional C8 molecule present in the serum of two siblings and an unrelated individual. The C8 in these three sera exhibited a pattern of partial immunologic identity with C8 in normal serum but was devoid of functional activity. The C8 was immunoprecipitated from the three sera and from a control serum with an antihuman C8 antiserum and analyzed by SDS-PAGE using highly purified human C8 as a reference. A selective absence of a band of 62,000 mol. wt was observed in the immunoprecipitates from the sera containing dysfunctional C8. Experiments performed with the purified alpha-gamma and beta subunits showed that the hemolytic activity of the C8 deficient sera could be reconstituted by the addition of the beta chain but not the alpha-gamma dimer. Binding of the dysfunctional C8 to C567 was excluded by the following observations: (1) EAC1-7 treated with the C8 deficient sera and then washed could not be lysed after the addition of the beta subunit and C9; and (2) the abnormal molecules did not interfere with the consumption of normal C8 by the soluble complex SC5b-7.

Inherited deficiency of C8 in a patient with recurrent meningococcal infections: further evidence for a dysfunctional C8 molecule and nonlinkage to the HLA system

An adult male with recurrent meningococcal infections is reported whose serum lacked functional C8 activity but possessed antigenic C8. The addition of 1500 U of purified C8/ml of serum restored hemolytic activity to normal. Four to five times more C8 was required to restore bactericidal activity than to restore hemolytic activity. Bactericidal activity could also be restored by mixing the patient's serum with a second C8-deficient serum that lacked detectable antigenic or functional C8. The patient's serum contained bactericidal antibody for groups A, B, C, and Y meningococci and specific antibody to group Y capsular polysaccharide. There was two to three times more bactericidal antibody activity in the serum than in a pool of normal sera for the infecting strain. Family studies disclosed a sibling who was HLA identical to the patient but whose serum contained normal amounts of total hemolytic and C8 functional activity.

Ultrastructure of the membrane attack complex of complement: detection of the tetramolecular C9-polymerizing complex C5b-8

The ultrastructure of the membrane attack complex (MAC) of complement had been described as representing a hollow cylinder of defined dimensions that is composed of the proteins C5b, C6, C7, C8, and C9. After the characteristic cylindrical structure was identified as polymerized C9 [poly(C9)], the question arose as to the ultrastructural identity and topology of the C9-polymerizing complex C5b-8. An electron microscopic analysis of isolated MAC revealed an asymmetry of individual complexes with respect to their length. Whereas the length of one boundary (+/- SEM) was always 16 +/- 1 nm, the length of the other varied between 16 and 32 nm. In contrast, poly(C9), formed spontaneously from isolated C9, had a uniform tubule length (+/- SEM) of 16 +/- 1 nm. On examination of MAC-phospholipid vesicle complexes, an elongated structure was detected that was closely associated with the poly(C9) tubule and that extended 16-18 nm beyond the torus of the tubule and 28-30 nm above the membrane surface. The width of this structure varied depending on its two-dimensional projection in the electron microscope. By using biotinyl C5b-6 in the formation of the MAC and avidin-coated colloidal gold particles for the ultrastructural analysis, this heretofore unrecognized subunit of the MAC could be identified as the tetramolecular C5b-8 complex. Identification also was achieved by using anti-C5 Fab-coated colloidal gold particles. A similar elongated structure of 25 nm length (above the surface of the membrane) was observed on single C5b-8-vesicle complexes. It is concluded that the C5b-8 complex, which catalyzes poly(C9) formation, constitutes a structure of discrete morphology that remains as such identifiable in the fully assembled MAC, in which it is closely associated with the poly(C9) tubule.