Complement pores in erythrocyte membranes. Analysis of C8/C9 binding required for functional membrane damage

Structural biology of the membrane attack complex

The complement system is an intricate network of serum proteins that mediates humoral innate immunity through an amplification cascade that ultimately leads to recruitment of inflammatory cells or opsonisation or killing of pathogens. One effector arm of this network is the terminal pathway of complement, which leads to the formation of the membrane attack complex (MAC) composed of complement components C5b, C6, C7, C8 and C9. Upon formation of C5 convertases via the classical or alternative pathways of complement activation, C5b is generated from C5 by proteolytic cleavage, nucleating a series of association and polymerisation reactions of the MAC-constituting complement components that culminate in pore formation of pathogenic membranes. Recent structures of MAC components and homologous proteins significantly increased our understanding of oligomerisation, membrane association and integration, shedding light onto the molecular mechanism of this important branch of the innate immune system.

The structure and function of mammalian membrane-attack complex/perforin-like proteins

The membrane-attack complex (MAC) of complement pathway and perforin (PF) are important tools deployed by the immune system to target pathogens. Both perforin and the C9 component of the MAC contain a common 'MACPF' domain and form pores in the cell membrane as part of their function. The MAC targets gram-negative bacteria and certain pathogenic parasites, while perforin, released by natural killer cells or cytotoxic T lymphocytes (CTLs), targets virus-infected and transformed host cells (1). Remarkably, recent structural studies show that the MACPF domain is homologous to the pore-forming portion of bacterial cholesterol-dependent cytolysins; these data have provided important insight into the mechanism of pore-forming MACPF proteins. In addition to their role in immunity, MACPF family members have been identified as animal venoms, factors required for pathogen migration across host cell membranes and factors that govern developmental processes such as embryonic patterning and neuronal guidance (2). While most MACPF proteins characterized to date either form pores or span lipid membranes, some do not (e.g. the C6 component of the MAC). A current challenge is thus to understand the role, pore forming or otherwise, of MACPF proteins in developmental biology. This review discusses structural and functional diversity of the mammalian MACPF proteins.

Physiology and pathophysiology of complement: progress and trends

The complement system comprises a family of at least 20 plasma and membrane proteins that interact in a tightly regulated cascade system to destroy invading bacteria and prevent the deposition of immune complexes in the tissues. This brief review addresses the basic mechanisms of complement activation and control and describes the active fragments produced during complement activation. The biological importance of the complement system is amply illustrated in patients with complement deficiencies, who are susceptible to bacterial infections and immune complex diseases. The involvement of complement in other immunological diseases is an expanding area of clinical research, supported by the development of new assays for the identification of complement activation. This area is discussed here with particular reference to neurological diseases. A promising new prospect involves the use of complement inhibitory molecules in therapy of complement-mediated disease and this exciting area is also discussed. Novel physiological roles of complement also are being revealed and new evidence that complement and complement receptors play an important role in reproduction is summarized. It is hoped that this brief overview will convey some of the enthusiasm currently pervading research in this underappreciated area of immunology.

Effects of temperature, time, and toxin concentration on lesion formation by the Escherichia coli hemolysin

We performed osmotic protection experiments to test the hypothesis that the Escherichia coli hemolysin forms a discrete-size pore in erythrocyte membranes. The effects of toxin concentration, assay time, temperature, and protectant concentrations were examined. The results we present here raise doubts about the existing model of pore formation by hemolysin. We demonstrate that osmotic protection by various sugars of different sizes is a function of hemolysin concentration and assay time. The data indicate that under various conditions, lesion sizes with a diameter ranging from < 0.6 to > 1.2 nm can be inferred. Quantification of hemolysin permitted the estimation of the number of HlyA structural protein molecules required per erythrocyte for lysis in the presence of each protectant. It appears that hemolysin induces heterogeneous erythrocyte lesions which increase in size over time. Influx experiments utilizing radioactive sugar markers indicated that time-dependent osmotic protection patterns are independent of the diffusion rates of individual protectants. We demonstrate that the rate of the putative growth in the size of hemolysin-mediated lesions is temperature dependent. The erythrocyte membrane lesions formed at 37 degrees C can be stabilized in size when shifted to 4 degrees C. On the basis of these data, new models for the nature of the hemolysin-mediated erythrocyte membrane lesions are presented.

A method for in vitro synthesis of unglycosylated recombinant complement component C9

A method for in vitro synthesis of human complement component C9 has been established in order to generate unglycosylated normal and mutant proteins without the need to sub-clone. One or two step polymerase chain reaction (PCR) was used to add the T7 RNA polymerase promoter and introduce multiple mutations within the cDNA. The cDNA was then transcribed by T7 RNA polymerase and the mRNA translated in a rabbit reticulocyte lysate or wheat germ system. Successful synthesis was confirmed by: the correct size of PCR product DNA on agarose gel electrophoresis, incorporation of [alpha-32P]UTP into mRNA, and formation of [35S]methionine-labelled protein of the correct molecular mass for full length C9. The wheat germ extract generated up to 1.5 micrograms of recombinant C9. This unglycosylated C9 had at least 10% of the haemolytic activity of native C9. Unglycosylated C9 polymerised more readily than the native protein. This spontaneous polymerisation was increased by removal of the first 23 amino acids or mutating two cysteines at positions 33 and 36. This therefore provides a rapid method for screening the effect of multiple mutations on the biological activity and polymerisation of pore forming proteins.

Small angle neutron scattering studies of C8 and C9 and their interactions in solution

Small angle neutron scattering (SANS) results revealed that contrary to most reports C9 is not a globular protein. Its radius of gyration (Rg) at pH 8 and an ionic strength of 0.5 is 32.2 +/- 1.4 A increasing to 35 A at physiologic ionic strength. In contrast, C8, which has a 2.2-fold larger mass, has a similar Rg value [34.6 +/- 1.6 A]. Calibration plots of Rg vs. M(r) indicate that native C8 is a spherical protein whereas native C9 is elongated. From previous reports it was known that native C8 and C9 associate in solutions of low ionic strength. SANS results confirmed this observation but also demonstrated that C8-C9 heterodimers are already formed at physiologic ionic strength. The dimeric complex is globular [Rg = 40 +/- 0.8 A] indicating that the proteins associate side-by-side rather than end-to-end. In contrast, in presence of the drug Suramin, a potent inhibitor of the assembly of the C5b-9 complex, C9 forms a complex with twice the molecular mass that is still elongated (Rg = 48.8 +/- 0.8 A), suggesting that in this case the protein dimerizes end-to-end via a bridging Suramin molecule.

Complement membrane attack on nucleated cells: resistance, recovery and non-lethal effects

Images

Complement proteins C5b-9 induce transbilayer migration of membrane phospholipids

Transbilayer migration of membrane phospholipid arising from membrane insertion of the terminal human complement proteins has been investigated. Asymmetric vesicles containing pyrene-labeled phosphatidylcholine (pyrenePC) concentrated in the inner monolayer were prepared by outer monolayer exchange between pyrenePC-containing large unilamellar vesicles and excess (unlabeled) small unilamellar vesicles, using bovine liver phosphatidylcholine-specific exchange protein. After depletion of pyrenePC from the outer monolayer, the asymmetric large unilamellar vesicles were isolated by gel filtration and exposed to the purified C5b-9 proteins at 37 degrees C. Transbilayer exchange of phospholipid between inner and outer monolayers during C5b-9 assembly was monitored by changes in pyrene excimer and monomer fluorescence. Membrane deposition of the C5b67 complex (by incubation with C5b6 + C7) caused no change in pyrenePC fluorescence. Addition of C8 to the C5b67 vesicles resulted in a dose-dependent decrease in the excimer/monomer ratio. This change was observed both in the presence and absence of complement C9. No change in fluorescence was observed for control vesicles exposed to C8 (in the absence of membrane C5b67), or upon C5b-9 addition to vesicles containing pyrenePC symmetrically distributed between inner and outer monolayers. These data suggest that a transbilayer exchange of phospholipid between inner and outer monolayers is initiated upon C8 binding to C5b67. The fluorescence data were analyzed according to a "random walk" model for excimer formation developed for the case where pyrenePC is asymmetrically distributed between lipid bilayers. Based on this analysis, we estimate that a net transbilayer migration of approximately 1% of total membrane phospholipid is initiated upon C8 binding to C5b67. The potential significance of this transbilayer exchange of membrane phospholipid to the biological activity of the terminal complement proteins is considered.

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.

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.

Bacterial killing by complement. C9-mediated killing in the absence of C5b-8

The ability of serum complement to kill Gram-negative bacteria requires assembly of the membrane attack complex (MAC) on the cell surface. The molecular events that lead to cell killing after MAC assembly are unknown. We have investigated the effect of C9 on bacterial survival in the presence and absence of its receptor, the C5b-8 complex, on the outer membrane. A fluorescence assay of the membrane potential across the inner bacterial membrane revealed that addition of C9 to cells bearing the performed C5b-8 complex caused a rapid and complete dissipation of the membrane potential. No fluorescence change was observed in serum-resistant strains of Escherichia coli. Addition of trypsin, after C9 was bound to C5b-8, did not rescue the cells from the lethal effects of C9. Furthermore, assays of cell killing kinetics and C9 binding indicate that formation of tubular poly(C9) is not required for killing. When C9 was introduced into the periplasmic space in the absence of its receptor by means of an osmotic shock procedure, cell killing occurred. Other proteins, such as C8 or serum albumin, were not toxic, and C9 was ineffective against two resistant strains. The results presented here and previously [Dankert & Esser (1986) Biochemistry 25, 1094-1100], when considered together, indicate that the 'lethal unit' in complement killing of some Gram-negative bacteria is a C9-derived product that acts by dissipation of cellular energy.

Formation of transmural complement pores in serum-sensitive Escherichia coli

The binding of C9 at 0 and 37 degrees C to viable Escherichia coli K-12 cells carrying C5b-8 complexes was quantified. At low temperature, limited average binding of only 1 to 1.4 molecules of C9 per C8 molecule occurred, whereas 6 to 8 C9 molecules were bound per C8 molecule at 37 degrees C. Despite incorporation of C9 into C5b-9 complexes at 0 degrees C, these terminal complexes caused no loss of bacterial viability even when present in very large numbers (1,000 to 1,500 per CFU) on the bacterial cells. In contrast, generation of 50 to 100 C5b-9 complexes carrying multiple C9 molecules per CFU caused loss of viability. The failure of C5b-81C91 complexes to generate transmural pores was confirmed by measurements of o-nitrophenyl-beta-D-galactoside influx into the cells. Whereas treatment of C5b-8-laden cells with C9 at 32 degrees C caused virtually instantaneous influx of the marker, almost no influx was registered in cells receiving C9 at 0 degrees C. When cells carrying C5b-7 were brought into the stationary phase and given C8 and C9 at 32 degrees C, a C9-dependent disruption of the outer membrane permeability barrier immediately occurred as demonstrated by cleavage of a chromogenic substrate by periplasmic beta-lactamase. In sharp contrast, o-nitrophenyl-beta-D-galactoside influx was markedly retarded over a prolonged period, with abrupt permeability increases of the inner membrane toward this molecule being noted just before bacterial cell division occurred. We conclude that killing of E. coli requires binding of C5b-9 complexes containing C9 oligomers to the outer membrane and suggest that formation of pores in the inner membrane occurs when these complexes are "hit" by transiently forming zones of bioadhesion. Formation of the latter may be a dynamic process that is accentuated during cell division and quiescent during the stationary phase.

Single-channel analysis of the conductance fluctuations induced in lipid bilayer membranes by complement proteins C5b-9

Single-channel analysis of electrical fluctuations induced in planar bilayer membranes by the purified human complement proteins C5b6, C7, C8, and C9 have been analyzed. Reconstitution experiments with lipid bilayer membranes showed that the C5b-9 proteins formed pores only if all proteins were present at one side of the membrane. The complement pores had an average single-channel conductance of 3.1 nS at 0.15 M KCl. The histogram of the complement pores suggested a substantial variation of the size of the single channel. The linear relationship between single-channel conductance at fixed ionic strength and the aqueous mobility of the ions in the bulk aqueous phase indicated that the ions move inside the complement pore in a manner similar to the way they move in the aqueous phase. The minimum diameter of the pores as judged from the conductance data is approximately 3 nm. The complement channels showed no apparent voltage control or regulation up to transmembrane potentials of 100 mV. At neutral pH the pore is three to four times more permeable for alkali ions than for chloride, which may be explained by the existence of fixed negatively charged groups in or near the pore. The significance of these observations to current molecular models of the membrane lesion formed by these cytolytic serum proteins is considered.

Cyanine dye fluorescence used to measure membrane potential changes due to the assembly of complement proteins C5b-9

The fluorescent potentiometric indicator diS-C3-(5) has been used to investigate changes in membrane potential due to assembly of the C5b-9 membrane attack complex of the complement system. EAC1-7 human red blood cells and resealed erythrocyte ghosts--bearing membrane-assembled C5b67 complexes--were generated by immune activation in C8-deficient human serum. Studies performed with these cellular intermediates revealed that the membrane potential of EAC1-7 red cells and ghosts is unchanged from control red cells (-7 mV) and ghosts (O mV), respectively. Addition of complement proteins C8 and C9 to EAC1-7 red cells results in a dose-dependent depolarization of membrane potential which precedes hemolysis. This prelytic depolarization of membrane potential--and the consequent onset of hemolysis--is accelerated by raising external [K+], suggesting that the diffusional equilibration of transmembrane cation gradients is rate limiting to the cytolytic event. In the case of EAC1-7 resealed ghosts suspended at either high external [K+] or [Na+], no change in membrane potential (from O mV) could be detected after C8/C9 additions. When the membrane potential of the EAC1-7 ghost was displaced from O mV by selectively increasing the K+ conductance with valinomycin, a dose-dependent depolarization of the membrane was observed upon addition of C8 and C9. In these experiments, lytic breakdown of the ghost membranes was less than 5%. Conclusions derived from this study include: (i) measured prelytic depolarization of the red cell Donnan potential directly confirms the colloid-osmotic theory of immune cytolysis. (ii) The diffusional transmembrane equilibration of Na+ and K+ through the C5b-9 pore results in a dose-dependent depolarization of the membrane potential (Em) which appears to be rate-limiting to cytolytic rupture of the target erythrocyte. (iii) Enhanced immune hemolysis observed in high K+ media cannot be attributed to cation-selective conductance across the C5b-9 pore, and is probably related to the near-equilibrium condition of potassium-containing red cells when suspended at high external K+. These experiments demonstrate that carbocyanine dye fluorescent indicators can be used to monitor electrochemical changes arising from immune damage to the plasma membrane under both cytolytic and noncytolytic conditions. Potential application of this method to the detection of sublytic pathophysiological changes in the plasma membrane of complement-damaged cells are discussed.

Characterization of the complement sensitivity of paroxysmal nocturnal hemoglobinuria erythrocytes

The affected erythrocytes of paroxysmal nocturnal hemoglobinuria (PNH II and PNH III cells) are abnormally sensitive to complement-mediated lysis. Normal human erythrocytes chemically modified by treatment with 2-amino-ethylisothiouronium bromide (AET) have been used as models for PNH cells inasmuch as they also exhibit an enhanced susceptibility to complement. To investigate the bases for the greater sensitivity of these abnormal cells to complement-mediated lysis, we compared binding of C3 and constituents of the membrane attack complex to normal, PNH II, PNH III, and AET-treated cells after classical pathway activation by antibody and fluid-phase activation by cobra venom factor complexes. When whole serum complement was activated by antibody, there was increased binding of C3 and C9 to PNH II, PNH III, and AET-treated cells, although the binding of these complement components to PNH II and PNH III cells was considerably greater than their binding to the AET-treated cells. In addition, all of the abnormal cell types showed a greater degree of lysis per C9 bound than did the normal erythrocytes. PNH III and AET-treated cells were readily lysed by fluid-phase activation of complement, whereas normal and PNH II erythrocytes were not susceptible to bystander lysis. The greater hemolysis of PNH III and AET-treated cells in this reactive lysis system was due to a quantitative increase in binding of constituents of the membrane attack complex. This more efficient binding of the terminal components after fluid-phase activation of whole serum complement was not mediated by cell-bound C3 fragments. These investigations demonstrate that the molecular events that characterize the enhanced susceptibility of PNH II, PNH III, and AET-treated erythrocytes to complement-mediated lysis are heterogeneous.

Ninth component of complement: self-aggregation and interaction with lipids

We have investigated environmental conditions that might be of importance for the polymerization of the ninth component (C9) of human complement. In disagreement with earlier reports summarized by Tschopp et al. [Tschopp, J., Müller-Eberhard, H. J., & Podack, E. R. (1982) Nature (London) 298, 534-538] we find no evidence for significant aggregation or loss of hemolytic activity of C9 when incubated at 37 degrees C even after 12 days of incubation. Higher temperatures cause denaturation of the protein and formation of stringlike aggregates. In contrast, short-term proteolysis with 1% (w/w) trypsin at room temperature causes rapid polymerization of part of the C9 into tubular structures (poly-C9), and the remainder of the monomeric C9 is digested. This polymerization reaction is inhibitable by trypsin inhibitor; alpha-thrombin and proteinase K are ineffective in creating polymers. A second discrepancy to the earlier reports is our finding that monomeric C9 immediately interacts with small unilamellar lipid vesicles (SUV) without a required heating step. As a result of this interaction about half of the C9 aggregates to form strings and tubules, and these aggregates cause agglutination of vesicles. The other half of the C9 associates with a second population of SUV without causing a change in Stokes' radius of these vesicles, and no proteinaceous structures are detectable on the vesicle surface by electron microscopy. When these two vesicle populations are tested for their membrane integrity, no release of an encapsulated fluorescent marker can be detected, nor is there leakage of potassium ions across the bilayer membrane since a membrane diffusion potential can be developed.(ABSTRACT TRUNCATED AT 250 WORDS)

Sequential onset of permeability changes in mouse ascites cells induced by Sendai virus

The addition of haemolytic Sendai virus to cells induces membrane changes in the following sequence: (i) Increased permeability to ions, (ii) increased permeability to low molecular weight metabolites, (iii) increased permeability to proteins. The consequences of an increased permeability to ions are: (a) alteration of membrane potential, (b) net changes in intracellular cations and (c) cell swelling, in that order. Depending on virus: cell ratio, Ca2+ concentration and temperature, it is possible to observe ion leakage without metabolite or protein leakage, and ion and metabolite leakage without protein leakage. A model for the induction of permeability changes is presented.

Proteolytic modification of human complement protein C9: loss of poly(C9) and circular lesion formation without impairment of function

We have compared the ability of thrombin-cleaved C9 (C9n) with that of native C9 to produce tubular or ring-like poly(C9) and to express the classical complement lesion on target membranes. Three procedures were used to produce poly(C9): (i) limited proteolysis with trypsin, (ii) interaction with small unilamellar lipid vesicles, and (iii) incubation with a 2- to 4-fold molar excess of ZnCl2. In contrast to C9, which could be converted to tubular poly(C9), C9n was converted to smaller peptides by the first procedure and was aggregated into string-like poly(C9) by the other two methods. C9-depleted human serum (R-9 serum) was reconstituted with either C9 or C9n and these sera were then used to lyse sensitized sheep erythrocytes. Numerous classical complement lesions could be detected on ghost membranes obtained from cells lysed by C9-reconstituted R-9 serum but only a few on ghost membranes produced by C9n-reconstituted R-9 serum. C9n was shown to be hemolytically as active as C9 even when tested under "single-hit" conditions and it was about twice as efficient when compared with C9 in releasing sucrose and inulin from resealed ghosts. These results are interpreted to indicate that formation of the classical complement lesion is only incidental to lysis and not an obligatory event and that enlargement of the "functional pore size" of the complement lesion is not linked to formation of a circular membrane attack complex.

The effect of Ca2+ on virus-cell fusion and permeability changes

Sendai virus-mediated permeability changes in Lettre cells or red blood cells are affected by extracellular Ca2+ in the following way: the lag period to onset of permeability changes is lengthened and the subsequent extent of leakage is reduced. Ca2+ neither stimulates nor inhibits fusion of the viral envelope to the plasma membrane of Lettre cells or red blood cells. It is concluded that Ca2+ protects cells against virally-induced permeability changes in a manner not involving membrane fusion.

The influence of electrochemical gradients of Na+ and K+ upon the membrane binding and pore forming activity of the terminal complement proteins

The hemolytic activity of the terminal complement proteins (C5b-9) towards erythrocytes containing high potassium concentration has been reported to be dramatically increased when extracellular Na+ is substituted isotonically by K+ (Dalmasso, A.P., et al., 1975, J. Immunol. 115:63-68). This phenomenon was now further investigated using resealed human erythrocyte ghosts (ghosts), which can be maintained at a nonlytic osmotic steady state subsequent to C5b-9 binding: (1) The functional state of C5b-9-treated ghosts was studied from their ability to retain trapped [14C]-sucrose or [3H]-inulin when suspended either in the presence of Na+ or K+. A dramatic increase in the permeability of the ghost membrane to both nonelectrolytes - in the absence of significant hemoglobin release - was observed for C5b-9 assembly in the presence of external K+. (2) The physical binding of the individual 125I-labeled terminal complement proteins to ghost membranes was directly measured as a function of intra- and extracellular K+ and Na+. The uptake of 125I-C7, 125I-C8, and 125I-C9 into membrane C5b-9 was unaltered by substitution of Na+ by K+. (3) The binding of the terminal complement proteins to ghosts subjected to a transient membrane potential generated by the K+-ionophore valinomycin (in the presence of K+ concentration gradients) was measured. No significant change in membrane binding of any of the C5b-9 proteins was detected under the influence of both depolarizing and hyperpolarizing membrane potentials.(ABSTRACT TRUNCATED AT 250 WORDS)