Purification and characterization of the eighth and ninth components of carp complement.

Complement components corresponding to mammalian C8 and C9 were isolated from carp (Cyprinus carpio) serum. Carp C8 (M(r) 146,000) proved to be a gamma-globulin composed of three polypeptide chains (alpha-chain, M(r) 62,000; beta-chain, M(r) 62,000; gamma-chain, M(r) 22,000). The alpha-chain was disulfide-linked to the gamma-chain and the beta-chain was non-covalently associated with the alpha-gamma chain, in fair agreement with mammalian C8. However, the N-terminal amino acid sequences of the three subunits showed no homology with those of human C8. Carp C9 was an alpha-globulin composed of a single polypeptide (M(r) 91,000) and the N-terminus was blocked. Carp serum depleted of C8 did not hemolyse either carp antibody-sensitized sheep erythrocytes or non-sensitized rabbit erythrocytes, while C9-depleted carp serum did not hemolyse the former, but did hemolyse the latter target cells, as in the case of C9-depleted human serum.

Horse complement protein C9: primary structure and cytotoxic activity.

Lack of hemolytic activity of horse serum is an inherent property of horse C9. To understand the molecular reasons for this deficiency we have cloned C9 cDNA from a horse liver cDNA library and have sequenced the cDNA yielding the complete coding sequence for horse C9. Purification of C9 from horse plasma and microsequencing established the N-terminus of the mature protein and verified that the correct horse C9 cDNA clone had been isolated. The deduced amino acid sequence corresponds to a mature protein of 526 amino acids that is 77% identical to human C9. It has the same domain structure as human C9 and contains 22 cysteines and four invariant tryptophans. The few differences include the N-terminus, which is an unblocked glycine in horse C9 but pyroglutamine in human C9, and three potential N-glycosylation sites compared to two in human C9. The N-terminal difference is unimportant since microsequencing of bovine C9, which is strongly hemolytic, established that it also has an unblocked glycine identical to horse C9. There are no obvious structural differences apparent that could resolve the differences in hemolytic potency between the two molecules. Aside from a few conservative replacements, both C9 sequences are identical between positions 250 and 360. This region includes the membrane interaction domain in C9 and the postulated transmembrane segment that is thought to constitute the wall of a putative transmembrane pore and, therefore, should be required for cytotoxicity. In agreement with this prediction we have observed that, in contrast to the marked decrease in hemolytic activity, horse C9 is very efficient in killing a variety of Gram-negative bacteria. These results demonstrate that horse C9 is a structurally competent molecule with efficient cytotoxic activity. Its inability to lyse erythrocytes may be related to the action of control proteins on target cell membranes.

Isolation of the C9b fragment of human complement component C9 using urea in the absence of detergents.

The bactericidal activity of the C5b-9 complex of complement is dependent upon the terminal complement component C9. The precursor C5b-8 complex is not harmful to bacterial cells until C9 is added to complete the C5b-9 complex. The C9 molecule can be proteolytically cleaved by thrombin to yield an intact, nicked molecule that remains fully functional when added to either bacterial cells or erythrocytes bearing pre-formed C5b-8 complexes. In investigating the membranolytic function of C9 in the C5b-9 complex, the carboxyl-terminal portion of the nicked molecule (C9b) has been shown to be membranolytic when added to erythrocytes, liposomes, or bacterial inner membranes in the absence of any other complement components. The isolation of C9b from nicked C9 has been accomplished by preparative gel electrophoresis using detergents, however the study of the activity of C9b in membrane systems may be complicated by the possible presence of residual detergent. To address this concern, we have used 4 M urea in conjunction with hydroxyapatite chromatography and a phosphate elution procedure to separate the domains of nicked C9. The isolated C9b domain, free of detergents and in the absence of any other complement components, was found to be membranolytic. C9b isolated in this manner was capable of lysing erythrocytes and inhibiting the growth of bacterial spheroplasts.

Interaction between apolipoproteins A-I and A-II and the membrane attack complex of complement. Affinity of the apoproteins for polymeric C9.

We have previously observed enhanced binding of HDL and apolipoproteins A-I and A-II to human endothelial cells exposed to activated complement. Induction of these binding sites required complement activation through C9, suggesting a specific role for the C9 component of the C5b-9 complex. We now report that specific and saturable binding sites for apoA-I and -A-II are expressed by C9 polymers (polyC9), whereas little binding was observed to native monomeric C9. These data suggested an interaction of the apoproteins with a site(s) which is exposed only upon C9 polymerization, and also suggested that binding of the apoproteins to this new site might interfere with assembly of C9 into the polyC9 tubule and insertion into the cell membrane. ApoA-I was found to inhibit zinc-catalyzed polymerization of C9 in a concentration-dependent fashion. Formation of SDS-resistant C9 polymers was completely inhibited at apoA-I or -A-II concentrations > or = 5 microM. ApoA-I also produced a concentration-dependent inhibition of C9 incorporation into C5b-9 complexes on endothelial cells, which was accompanied by a corresponding decrease in SDS-resistant C9 polymers associated with the cell membrane. In summary, the ability of the HDL apoproteins A-I and A-II to interact with an activation-dependent conformer(s) of the C9 component of the C5b-9 complex appears to explain the expression of HDL binding sites on endothelial cells exposed to complement. These apoproteins are also inhibitors of C9 polymerization, which may underlie the protective effect of HDL for blood cells exposed to activated complement.

[Purification of the ninth component of human complement (C9)].

A large amount of human C9 was purified from plasma by the following procedures: 1) Polyethylene glycol precipitation; 2) Depletion of plasminogen by passing over an L-lys-sepharose column; 3) DEAE-sephadex A-50 chromatography; and 4) Hydroxylapatite (HA) chromatography. The method of C9 purification was improved by altering the column-elution conditions and by the establishment of a novel method for preparing high-flow-rate HA. As a result, the rate of recovery of C9 was high (28.2%) and no impurities were detected either on gel electrophoretic or immunochemical examination. The hemolytic activity of purified C9 was retained.

Purification of the ninth component of the bovine complement cascade.

Conditions for purification of the ninth component of bovine complement (C9) were established. The conditions for binding and elution from diethylaminoethyl cellulose and hydroxylapatite were different than for human C9. Serum albumin, a frequent contaminant of bovine C9 preparations, was removed by chromatography on reactive-red agarose. The calculated molecular weight of bovine C9 was 66,000, and reduction with 2-mercaptoethanol affected its migration on polyacrylamide gel electrophoresis. Some preparations of bovine C9 migrated as 2 bands when partially reduced, but extensively reduced preparations had a single band.

Rapid isolation of human complement component C9 to verify the specificity of a haemolytic C9 microassay.

A sensitive, haemolytic microassay of human complement component C9 was developed. The assay is based on the principle of reactive (C5b6-initiated) haemolysis and uses commercially available C9-depleted serum as reagent for C9. The specificity of the assay was verified by rapid, activity-guided isolation of the haemolytic component from human serum using high-performance liquid chromatography (HPLC) on a system for fast protein liquid chromatography. This isolation yielded a single component with characteristics of C9. The results suggest that rapid, activity-guided isolation as a new application of HPLC can be a useful tool to demonstrate the specificity of a functional assay.

Molecular modeling of the domain structure of C9 of human complement by neutron and X-ray solution scattering.

C9 is the most abundant component of the membrane attack complex of the complement system of immune defense. This is a typical mosaic protein with thrombospondin (TSR) and low density lipoprotein receptor (LDLr) domains at its N-terminus and an epidermal growth factor-like (EGF) domain at its C-terminus. Between these lies a perforin-like sequence. In order to define the arrangement in solution of these four moieties in C9, high-flux neutron and synchrotron X-ray solution scattering studies were carried out. The neutron radius of gyration RG at infinite contrast is 3.33 nm, and its cross-sectional RG (RXS) is 1.66 nm. Similar values were obtained by synchrotron X-ray scattering after allowance for radiation effects. Stuhrmann analyses showed that the neutron radial inhomogeneity of scattering density alpha is 35 X 10(-5) from the RG data and 16 X 10(-5) from the RXS data. These values are typical for soluble glycoproteins and show no evidence for the existence of any large hydrophobic surface patches on free C9 that might form contacts with lipids. Indirect transformation of the neutron and X-ray scattering curves into real space showed that C9 had a maximum dimension estimated at 12 +/- 2 nm, and this suggests that the lengths of 7-8 nm deduced from previous electron microscopy studies in vacuo are underestimated. Molecular modeling of the C9 scattering curves utilized small spheres in the Debye equation, in which the analyses were constrained by the known volumes of the four moieties of C9 and the known sizes of the TSR and EGF-like domains. The most likely models for C9 suggest that these four regions of C9 are arranged in a V-shaped structure, with an angle of 10 degrees between the two arms, each of length 11.1 nm. This structure has a more hydrophobic character between the two arms. The scattering model is fully consistent with hydrodynamic sedimentation data on C9. Similar V-shaped hydrodynamic models could be developed for C6, C7, C8, and C9 of complement. Such a compact structure is atypical of other multidomain complement proteins so far studied by solution scattering and is fully compatible with mechanisms in which C9 is postulated, on activation, to undergo a drastic unfolding of its domain structure and to expose a more hydrophobic surface which can be embedded into lipid bilayers.

Thermal unfolding and aggregation of human complement protein C9: a differential scanning calorimetry study.

The thermotropic behavior of purified human complement protein C9 was investigated by high-sensitivity differential scanning calorimetry. When dissolved in physiological buffers (pH 7.2, 150 mM NaCl), C9 underwent three endothermic transitions with transition temperatures (Tm) centered at about 32, 48, and 53 degrees C, respectively, and one exothermic transition above 64 degrees C that correlated with protein aggregation. The associated calorimetric enthalpies of the three endothermic transitions were about 45, 60, and 161 kcal/mol with cooperative ratios (delta Hcal/delta HvH) close to unity. The total calorimetric enthalphy for the unfolding process was in the range of 260-280 kcal/mol under all conditions. The exothermic aggregation temperature was strongly pH dependent, changing from 60 degrees C at pH 6.6 to 81.4 degrees C at pH 8.0, whereas none of the three endothermic transitions was significantly affected by pH changes. They were, however, sensitive to addition of calcium ions; most affected was Tm1 which shifted from 32 to 35.8 degrees C in the presence of 3 mM calcium, i.e., the normal blood concentration. Kosmotropic ions stabilized the protein by shifting the endothermic transitions to slightly higher temperatures whereas inclusion of chaotropic ions (such as choline), removal of bound calcium by addition of EDTA, or proteolysis with thrombin lowered the transition temperatures. Previous studies had indicated the formation of at least three different forms of C9 during membrane insertion or during heat polymerization, and it is suggested that the three endothermic transitions reflect the formation of such C9 conformers. Choline, which is present at high concentrations on the surface of biological membranes, and calcium ions have the ability to shift the transition temperatures of the first two transitions to be either close to or below body temperature. Thus, it is very likely that C9 is present in vivo in a partially unfolded state when bound to a membrane surface, and we propose that this facilitates membrane insertion and refolding of the protein into an amphiphilic conformation.

Evasion of Yersinia enterocolitica serotype 03 from complement-mediated killing.

Using isogenic strains of Y. enterocolitica serotype 03 (isolate 75) differing in LPS side chains (S and R), plasmid content (p+ and p-) or selective failure of YOP1 expression (YOP1-) we observed comparable C9 consumption via the alternative pathway by all strains. Differences became apparent in the bactericidal assays in which the 75S p+ strain was resistant whereas the plasmid-negative S and R strains were killed. Increased but submaximal resistance was observed with the 75R p+ and 75S YOP1- strains indicating that LPS side chains and plasmid-encoded factors other than YOP1 also contribute to serum resistance. Deposition of C9 and terminal complement complex (TCC) formation were greatly reduced on the resistant 75S p+ strain compared to the sensitive 75S p- strain. The 75S YOP1- variant behaved like the 75S p- strain with respect to C9 deposition and TCC formation suggesting that YOP1 prevents terminal C activation. A comparison of TCC deposited on resistant and sensitive Y. enterocolitica revealed no differences with respect to their salt- and protease-resistance and their size. We conclude from our experiments that serum resistance of 75S p+ depends on several surface components YOP1 being the most important. The YOP1 protein clearly interferes with C9 deposition and TCC formation and thereby contributes to serum resistance of plasmid-positive, YOP1 expressing Y. enterocolitica.

Several epitopes on native human complement C9 are involved in interaction with the C5b-8 complex and other C9 molecules.

Ten monoclonal antibodies (mAb) against native human C9 exhibiting various inhibitory effects on the hemolytic activity of C9 (Bausback, J., Kontermann, R. and Rauterberg, E. W., Immunobiology 1988. 178: 58) were further analyzed regarding their reactivities with monomeric C9 (mC9), polymerized C9 (pC9), and the non-lytic SC5b-9 complex in enzyme-linked immunosorbent assay and with the membrane attack complex (MAC) generated on rabbit erythrocytes analyzed by flow cytometry. In addition, the inhibitory effects of mAb on zinc-induced C9 polymerization were investigated. One epitope of the C-terminal half of C9b exposed on the surface of pC9 and the MAC seems not to participate directly in lytic function or polymerization since no inhibitory effect of the respective mAb was observed. The nine other mAb directed against epitopes of the C9a part exhibit various inhibitory potentials. The mAb inhibit either hemolysis or polymerization, or both processes. Due to the reactivity with the tested antigens the mAb can be divided into two groups. mAb of the first group bind with nearly the same affinity to all four antigens, whereas mAb of the second group react preferentially with mC9 while their affinity to pC9, SC5b-9 and the MAC is reduced. Comparison of reaction patterns and inhibitory effects strongly suggest that different epitopes on the surface of native C9 are involved in interaction of C9 with C5b-8 and/or in C9-C9 interaction. The finding that mAb inhibiting polymerization of C9 in vitro have no inhibitory effect on hemolysis confirms that C9 polymers are no prerequisite for lysis.

Purification of C8 and C9 from rat serum.

A procedure based on modifications of published methods for human proteins for the isolation of rat C8 and C9 from one batch of serum is described. The procedure allows the rapid, large-scale isolation of pure and haemolytically active proteins. Rat C9 had a slightly higher molecular weight than human C9 on SDS-PAGE and similar isoelectric point. Rat C8 differed from human C8 in the molecular weight of the gamma chain (23,000 and 21,000 kD respectively), and on isoelectric focusing (pI rat C8: 6.5-6.9; pI human C8: 7.4-7.9).

Isolation and characterization of a membrane-attack-complex-inhibiting protein present in human serum and other biological fluids.

We have previously reported the isolation of a membrane-attack-complex-inhibiting protein (MIP) from human erythrocyte membranes [Watts, Patel & Morgan (1987) Complement 4, 236] and the production of polyclonal antibodies to this protein. Here we report the identification in plasma, urine, saliva and cerebrospinal fluid of a protein immunochemically identical with the membrane-derived MIP. The protein has been isolated from plasma by immunoaffinity chromatography on an anti-(erythrocyte MIP)-Sepharose column and shown by SDS/polyacrylamide-gel electrophoresis to be of similar molecular mass to the erythrocyte protein (55 kDa non-reduced and 65 kDa under reducing conditions). Monoclonal antibodies have been raised against plasma MIP and used to establish a two-site enzyme-linked immunoadsorbent assay, enabling quantification of MIP in plasma, urine and cerebrospinal fluid. Plasma MIP, though not able to incorporate spontaneously into membranes, was deposited on heterologous and homologous erythrocyte membranes during complement activation in a C8-dependent manner. Depletion of MIP from plasma resulted in enhancement of the lytic capacity of the plasma on heterologous erythrocytes.

Multimeric C9 within C5b-9 deposits in unique locations in the cell wall of Salmonella typhimurium.

We have shown previously that multimeric C9 within C5b-9 (C9:C5b-8 greater than 3:1) is needed for killing of a rough strain of Escherichia coli. We now extend these studies using serum sensitive, rough (R) and serum resistant, wild type (WT) strains of Salmonella typhimurium as well as a mutant S. typhimurium strain (TS) with a temperature sensitive mutation in synthesis of keto-deoxy-octulosonate, a constituent within the deep core structure of Salmonella LPS. Both R and TS required multimeric C9 within C5b-9 to be killed. Addition at 37 degrees C of increasing inputs of C9 to TS or R bearing C5b-9 led to a dose-related increase in C9 binding and killing. In contrast, addition of high inputs of C9 to the same strains at 4 degrees C, a procedure that limits the C9:C5b-8 ratio to 1:1, resulted in low C9 binding and minimal killing. Bactericidal C5b-9 formed at 37 degrees C on R and TS with high inputs of C9 co-sedimented with the bacterial outer membrane on sucrose density gradient analysis. Non-bactericidal C5b-9 on R, WT, and TS co-sedimented near the inner membrane, despite the presumed lack of association between these constituents. Whereas 125I C9 within the non-bactericidal pools immunoprecipitate with anti-C5, 125I C9 within bactericidal pools did not immunoprecipitate with anti-C5, anti-C7, or anti-C9. These findings suggest that bactericidal C5b-9 may be deposited in a unique location or configuration within the bacterial cell wall.

A biotin-avidin sandwich ELISA for quantification of intact complement component C9. The sera from hereditary C9 deficient individuals completely lack C9.

A two-site sandwich ELISA method was developed for quantitating intact C9 protein using MoAb P40 (anti-C9b antibody). This antibody reacted with monomeric C9 but not with polymerized C9. MoAb P40 was used as a capture antibody and MoAb X195 (anti-C9a antibody) as a detection antibody. This method is highly sensitive and can detect approximately 0.5 ng/ml of native C9. No cross-reactivities of either C6, C7, or C8 were observed even at concentrations of 10 micrograms/ml per component. In addition, this method allows for measurement of only intact C9 molecules, eliminating the interference of polymerized C9 or inactivated C9. Using this assay, no C9 at all was detected in sera from inherited C9 deficient individuals, including both healthy blood donors and patients with meningococcal meningitis; although by hemolytic assay, C9 levels were reported to be less than 0.2% those of NHS. Therefore, this two-site sandwich ELISA method can replace the hemolytic assay, and is especially useful for measuring small amounts of C9 in serum.

The cytolytic protein of human lymphocytes related to the ninth component (C9) of human complement: isolation from anti-CD3-activated peripheral blood mononuclear cells.

A 70-kDa channel-forming protein has recently been isolated from human large granular lymphocytes maintained in interleukin-2-dependent culture. The protein was shown to be immunochemically related to the ninth component of complement (C9) and was therefore designated C9-related protein (C9RP). Using the procedure that was developed for the isolation of C9RP from large granular lymphocytes--i.e., affinity chromatography employing anti-human C9 linked to Sepharose, a cytolytic protein has now been isolated from OKT3-activated human peripheral blood mononuclear cells. Nineteen to 40 micrograms of active protein was obtained from 1 X 10(9) human peripheral blood mononuclear cells after the cells were cultured for 3 days with OKT3 (monoclonal antibody to cell surface antigen T3). During this period, a marked increment occurred in the amount of the cytotoxic protein contained per cell, indicating that OKT3 induced de novo synthesis of the protein. By NaDodSO4/PAGE the molecular mass was determined to be 70 kDa. By ELISA the isolated protein and C9RP of large granular lymphocytes reacted to the same extent with anti-C9RP. Using K-562 or M21 human melanoma cells as targets, the cytotoxic activity of the isolated protein, in the presence of 5 mM Ca2+, was comparable to that of C9RP. The same cytolytic protein was isolated from peripheral blood mononuclear cells that were depleted of CD16+ cells prior to OKT3 activation and that consisted primarily of CD4+ and CD8+ T lymphocytes. These results suggest that the cytolytic protein of OKT3-activated cytotoxic T lymphocytes is identical with C9RP of interleukin-2-stimulated large granular lymphocytes.

Restriction in the lytic efficiency of complement of different erythrocyte targets: a re-examination of the activities of horse C8 and C9.

Differences in the lytic efficiency of different complement sources have frequently been observed. This effect has been shown to be related to both the species of the target erythrocyte and the species composition of terminal complement components within the 5b-9 membrane attack complex. The majority of studies have indicated that the source of C9 is critical in controlling the range of erythrocyte species that can be lysed efficiently. One exception to this general finding was the report by Lachmann et al., 1973 (Immunology 24, 135-145), using horse serum as a complement source. In that study, horse C8 rather than C9 was implicated as the critical component. In this study, we have re-examined this observation and find that the restricted hemolytic potential of horse complement correlates absolutely with the presence of horse C9. The reason for the differences between our findings and those of the earlier study are discussed.

Complement-mediated killing of Escherichia coli: dissipation of membrane potential by a C9-derived peptide.

The molecular mechanism of complement-mediated killing of Gram-negative bacteria has yet to be resolved, but it is generally accepted that assembly of the membrane attack complex (MAC) of complement on the outer bacterial membrane is a required step. We have now investigated the effect of the MAC and its precursor complex, C5b-8, on the membrane potential (delta Em) across the inner bacterial membrane. Delta Em of whole cells was measured directly by using a lipophilic cation (tetraphenylphosphonium) that equilibrates with the potential or indirectly by measuring transport of solutes (proline and galactoside), which is dependent on delta Em. Our results indicate that the C5b-8 complex caused a transient collapse of delta Em in the absence of cell killing. Addition of C9 to allow formation of the MAC dissipated delta Em irreversibly, and the cells were killed. Since delta Em is generated across the inner membrane in Gram-negative bacteria, inner membrane vesicles were prepared and membrane potentials were generated either by adding D-lactate to energize the electron-transport chain or by creating a K+ diffusion potential with valinomycin. C9 added in the absence of earlier acting complement proteins had no effect on delta Em of isolated, actively respiring vesicles or on K+ diffusion potentials. In contrast, its C-terminal thrombin fragment (C9b), which has been shown earlier to contain the membrane-active domain of C9, efficiently collapsed delta Em in such vesicles. C9b did not require a specific receptor since it was effective on "right-side-out" and "inside-out" vesicles. These results are interpreted to indicate that a C9-derived fragment deenergizes cells and may be the causative agent for cell death.

The architecture of complement component C9 and poly(C9).

The architecture of human complement component C9 and poly(C9) was investigated by transmission electron microscopy. Monomeric native C9 (Mr = 66,000) exhibits an ellipsoid appearance (70 X 50 A) with a crevice visible on one face. C9 polymerizes spontaneously to form hollow tubular structures consisting of 12-16 monomeric subunits. Poly(C9) is a cylinder (150 A-outer diameters and 90 A-inner diameter) rimmed by a torus (46-A thick) on one end. Electron micrographs of poly(C9) indicate that the torus is formed by radial strands of polypeptide. Each subunit of poly(C9) is apparently tilted relative to the central axis of the cylindrical structure. C9 can be cleaved by alpha-thrombin into two single-chain polypeptide fragments: C9a (Mr = 28,000) and C9b (Mr = 38,000), which are the amino- and carboxyl-terminal segments of the protein, respectively. The cleaved form of the protein, C9a,b, can be induced to polymerize under suitable conditions to form sodium dodecyl sulfate-resistant poly(C9), indicating that the resistance of poly(C9) to denaturation is a collective feature of both C9a and C9b. The C9a and C9b polypeptide regions have been mapped on poly(C9) by immunoelectron microscopy. Determinants for the C9a region were observed about the torus, base, and on the midsection of the poly(C9) cylinder. C9b epitopes are concentrated predominantly about the torus and base, but were rarely observed on the midsection of poly(C9). Thus, the C9a and C9b segments of the C9 polypeptide are not clearly segregated in poly(C9). The locations of oligosaccharide units on poly(C9) were visualized by electron microscopy after labeling of the complex with concanavalin A bound to colloidal gold. The oligosaccharide positions were found on the periphery of the torus and base. In summary, C9 appears to be a single-domain protein. Polymerization involves a major rearrangement. To form a subunit of poly(C9) the polypeptide chain must form at least one major fold parallel to the central axis of the tubule.