Interaction of poly(styrene sulfonic acid) with the alternative pathway of the serum complement system

Bioartificial pancreas, in which the islets of Langerhans are enclosed in artificial membrane to be protected from the host immune system, is expected to be a promising medical device to treat patients who suffer from insulin-dependent diabetes. Our strategy for preparation of a bioartificial pancreas involves utilizing a membrane including polymeric materials that can inhibit the complement reaction. In this study, we examined the effects of poly(styrene sulfonic acid) (PSSa) on the alternative pathway of the serum complement system to identify the mechanism(s) involved. PSSa was dissolved in pooled normal human serum (NHS), and the mixtures were incubated at 37°C for 30 min. Complement activities in sera were determined by hemolytic assays. Amounts of complement activation products released were determined by ELISA. Interactions of PSSa with complement components and fragments were examined with electrophoresis and immunoblotting. From these examinations, it appeared that the manner of PSSa effects on the alternative pathway (AP) highly depends on its concentration. PSSa seemingly acted as an activator when its concentration was 0.005 g/dl to 0.05 g/dl, while it acted as an inhibitor when its concentration was more than 0.1 g/dl. In terms of activation or inhibition of the AP, forming complex of PSSa with factor H induced activation, and that with factor D induced inhibition.     

Interaction of poly(styrene sulfonic acid) with the classical pathway of the serum complement system

Bioartificial pancreas, in which the islets of Langerhans (islets) are enclosed in artificial membrane to be protected from the host immune system, is expected to be a promising medical device to treat patients who suffer from insulin-dependent diabetes. Our strategy for the preparation of a bioartificial pancreas involves utilizing a membrane including polymeric materials that can inhibit the complement. When we examined a membrane containing poly(styrene sulfonic acid), long survival of islets enclosed in the membrane was observed in recipients carrying antibodies against islet cells. This fact stimulated us to start examinations of effects of PSSa on the complement system. In this study, we examined effects of PSSa on the classical pathway (CP) of the serum complement system to identify the mechanism(s) involved. The electric static interaction between cationic C1q (pI 9.3) and anionic PSSa induces PSSa-C1q complex formation. The dissociation of C1q(r2s2) complex by PSSa results inactivation of the CP activity. Those results indicate that PSSa was not an activator of the CP, but an inhibitor of CP activation. This study clarifies the mechanism by which PSSa protects islets in a microcapsule from the humoral immunity of the recipient carrying anti-islet antibodies. A microcapsule containing PSSa seems to effectively protect the islet from attacks of the host immune system after transplantation carrying antibodies against islet cells.     

Interaction of poly(styrene sulfonic acid) with the classical pathway of the serum complement system

Bioartificial pancreas, in which the islets of Langerhans (islets) are enclosed in artificial membrane to be protected from the host immune system, is expected to be a promising medical device to treat patients who suffer from insulin-dependent diabetes. Our strategy for the preparation of a bioartificial pancreas involves utilizing a membrane including polymeric materials that can inhibit the complement. When we examined a membrane containing poly(styrene sulfonic acid), long survival of islets enclosed in the membrane was observed in recipients carrying antibodies against islet cells. This fact stimulated us to start examinations of effects of PSSa on the complement system. In this study, we examined effects of PSSa on the classical pathway (CP) of the serum complement system to identify the mechanism(s) involved. The electric static interaction between cationic C1q (pI 9.3) and anionic PSSa induces PSSa-C1q complex formation. The dissociation of C1q(r₂s₂) complex by PSSa results inactivation of the CP activity. Those results indicate that PSSa was not an activator of the CP, but an inhibitor of CP activation. This study clarifies the mechanism by which PSSa protects islets in a microcapsule from the humoral immunity of the recipient carrying anti-islet antibodies. A microcapsule containing PSSa seems to effectively protect the islet from attacks of the host immune system after transplantation carrying antibodies against islet cells.     

Alternative pathway amplification and infections

The alternative pathway (AP) is the phylogenetically oldest arm of the complement system and may have evolved to mark pathogens for elimination by phagocytes. Studies using purified AP proteins or AP-specific serum showed that C3b amplification on bacteria commenced following a lag phase of about 5 min and was highly dependent on the concentration of complement. Most pathogens have evolved several elegant mechanisms to evade complement, including expressing proteases that degrade AP proteins and secreting proteins that block function of C3 convertases. In an example of convergent evolution, many microbes recruit the AP inhibitor factor H (FH) using molecular mechanisms that mimic FH interactions with host cells. In most instances, the AP serves to amplify C3b deposited on microbes by the classical pathway (CP). The role of properdin on microbes appears to be restricted to stabilization of C3 convertases; scant evidence exists for its role as an initiator of the AP on pathogens in the context of serum. Therapeutic complement inhibition carries with it an increased risk of infection. Antibody (Ab)-dependent AP activation may be critical for complement activation by vaccine-elicited Ab when the CP is blocked, and its molecular mechanism is discussed.     

Complement-activating ability of leucocytes from patients with complement factor I deficiency.

Previous studies from this laboratory have shown that normal peripheral blood B cells are capable of activating complement via the alternative pathway (AP), that the activation is associated with complement receptor type 2 (CR2) expression, and that erythrocytes at normal blood levels partially inhibit the activation. The purpose of the present study was to investigate whether factor I (FI) deficiency, which leads to continued formation of the AP convertase (C3bBb) resulting in the consumption of factor B and C3 and large scale generation of C3b fragments, affects the phenotype and/or function of the patients' B cells. Using flow cytometry, peripheral blood leucocytes (PBL) from two FI-deficient patients were investigated for expression of complement receptors and complement regulatory proteins, in vivo-deposited C3 fragments and in vitro complement-activating ability. CR1 levels on B cells were significantly lower in FI-deficient patients than in normal individuals, whereas CR2 levels were found to be reduced, although not to a significant extent. CR1 levels on monocytes and polymorphonuclear leucocytes (PMN) were found to be normal or slightly raised. All leucocyte subpopulations were found to be covered in vivo with C3b fragments. AP activation on B cells from FI-deficient patients in homologous serum was significantly reduced compared with that for normal individuals, whereas no in vitro activation was seen in autologous serum. In addition, the in vivo-bound C3b fragments were degraded to C3d,g when the patients' PBL were incubated in homologous serum containing EDTA. Finally, the patients, erythrocytes failed to exert any inhibition on AP activation in homologous serum.     

New insight into the effects of heparinoids on complement inhibition by C1‐inhibitor

Complement activation is of major importance in numerous pathological conditions. Therefore, targeted complement inhibition is a promising therapeutic strategy. C1‐esterase inhibitor (C1‐INH) controls activation of the classical pathway (CP) and the lectin pathway (LP). However, conflicting data exist on inhibition of the alternative pathway (AP) by C1‐INH. The inhibitory capacity of C1‐INH for the CP is potentiated by heparin and other glycosaminoglycans, but no data exist for the LP and AP. The current study investigates the effects of C1‐INH in the presence or absence of different clinically used heparinoids on the CP, LP and AP. Furthermore, the combined effects of heparinoids and C1‐INH on coagulation were investigated. C1‐INH, heparinoids or combinations were analysed in a dose‐dependent fashion in the presence of pooled serum. Functional complement activities were measured simultaneously using the Wielisa^®‐kit. The activated partial thrombin time was determined using an automated coagulation analyser. The results showed that all three complement pathways were inhibited significantly by C1‐INH or heparinoids. Next to their individual effects on complement activation, heparinoids also enhanced the inhibitory capacity of C1‐INH significantly on the CP and LP. For the AP, significant potentiation of C1‐INH by heparinoids was found; however, this was restricted to certain concentration ranges. At low concentrations the effect on blood coagulation by combining heparinoids with C1‐INH was minimal. In conclusion, our study shows significant potentiating effects of heparinoids on the inhibition of all complement pathways by C1‐INH. Therefore, their combined use is a promising and a potentially cost‐effective treatment option for complement‐mediated diseases.     

Design of a complement mannose-binding lectin pathway-specific activation system applicable at low serum dilutions

Recently we showed that alternative pathway (AP) amplification was responsible for more than 80% of specific classical pathway-induced terminal pathway activation under physiological conditions. The present study aimed to design a system for specific lectin pathway (LP) activation applicable at low serum dilutions with a fully functional AP. Comparison between activation of normal human serum (NHS), a mannose-binding lectin (MBL) homozygous D/D-deficient serum, and sera deficient in C1q and C2, all diluted 1 : 2, was essential to document optimal conditions for LP specificity. Mannan on the solid phase of enzyme-linked immunosorbent assay (ELISA) plates was used for activation, showing 0.5 microg mannan/well to give optimal conditions because at this concentration a good signal was preserved for C4 and TCC deposition in NHS, whereas the C3 deposition observed in C2-deficient serum at higher mannan concentrations reached nadir at 0.5 microg/well, indicating a lack of direct AP activation under these conditions. Pooled NHS and C1q-deficient serum gave the same degree of C4 and terminal complement complex (TCC) deposition, whereas deposition of these products was not obtained with MBL-deficient serum. Reconstitution with purified MBL, however, restored the depositions. A blocking anti-MBL monoclonal antibody (mAb) completely abolished the complement deposition, in contrast to a non-inhibiting anti-MBL mAb. Activation of C2-deficient serum induced C4 deposition similar to NHS, but negligible deposition of C3 and TCC, confirming the lack of direct activation of AP. Thus, this assay is unique in being LP-specific at low serum dilution and thus particularly suitable to study LP activation mechanisms and the role of AP amplification under physiological conditions.     

Complement resistance of pathogenic Entamoeba histolytica mediated by trypsin-sensitive surface component(s).

Pathogenic forms of the protozoan parasite Entamoeba histolytica were reported previously to resist the cytolytic effect of the alternative complement pathway (AP) only temporarily during exposure to complement. In contrast, nonpathogenic forms of E. histolytica had been found to show AP resistance as a stable property. We studied the mechanisms of AP resistance of the two forms. Upon exposure to AP activity, resistant pathogenic or nonpathogenic forms bound significantly less C3 products than complement-sensitive pathogenic amebae, indicating that the two resistant forms both inhibited AP amplification. Various enzymatic treatments and inhibition of membrane mobility by cytochalasin B and glutaraldehyde fixation showed that the mechanisms of AP inhibition differed between pathogenic and nonpathogenic forms; in contrast to nonpathogenic forms, pathogenic amebae required intact membrane mobility and a trypsin-sensitive surface component(s) to inhibit AP activation.     

Inhibition of complement activation on a model biomaterial surface by streptococcal M protein-derived peptides

The aim of this study was to evaluate a new approach to inhibit complement activation triggered by biomaterial surfaces in contact with blood. In order to inhibit complement activation initiated by the classical pathway (CP), we used streptococcal M protein-derived peptides that specifically bind human C4BP, an inhibitor of the CP. The peptides were used to coat polystyrene microtiter wells which served as a model biomaterial. The ability of coated peptides to bind C4BP and to attenuate complement activation via the CP (monitored as generation of fluid-phase C3a and binding of fragments of C3 and C4 to the surface) was investigated using diluted normal human serum, where complement activation by the AP is minimal, as well as serum from a patient lacking alternative pathway activation. Complement activation (all parameters) was significantly decreased in serum incubated in well surfaces coated with peptides. Total inhibition of complement activation was obtained at peptide coating concentrations as low as 1–5 μg/mL. Successful use of Streptococcus-derived peptides shows that it is feasible to control complement activation at a model biomaterial surface by capturing autologous complement regulatory molecules from plasma.     

Interactions of chrysotile asbestos fibres with the complement system.

Type A chrysotile fibres (white asbestos) were tested in vitro for activation of the complement system. Fibres were incubated in normal human serum (NHS), factor B-depleted human serum, and normal and C4-deficient guinea-pig sera; the supernates were assayed for the remaining complement activity. Activation of the alternative pathway (AP) was shown in three ways. First, quantitative measurement of factor B; second, kinetic analysis of rabbit red blood cell lysis in whole alternative pathway (AP) and factor B lytic assays; third, qualitative measurement of C3 and factor B conversion by crossed immunoelectrophoresis. No C3 convertase activity could be demonstrated on the fibres but other possible mechanisms of AP activation are discussed. Magnesium itself is not responsible for this activation because acid-treated fibres retain this property. The early classical pathway is not involved as shown by normal whole complement activity of a factor B-depleted human serum and the absence of decrease of C4 functional activity. Knowing that complement proteins are present in pulmonary alveoli, mainly provided by cell synthesis, we suggest that complement activation in vivo may be relevant to the genesis of the chronic inflammation and fibrosis in the lung.     

Determination of alternative pathway of complement activity in mouse serum using rabbit erythrocytes

Rabbit, mouse and sheep erythrocytes expressing different concentrations of membrane sialic acid were used to study possible modes of activation of the alternative complement (C) pathway in mouse, human and guinea pig serum. Mouse erythrocytes activated only human serum, whereas rabbit erythrocytes activated the sera of all three species. Based on the observation that rabbit erythrocytes activate the murine alternative C pathway a method for estimation of alternative C pathway activity (AP50 value) in mouse serum was devised analogous to that used for human AP50 determination. The method is not very sensitive to ageing or to batch variation of the indicator cells. The AP50 value of mouse serum measured by this method is of the same order as for human and guinea pig serum. Mouse serum AP50 activity is partly determined by natural antirabbit erythrocyte antibodies and is sensitive to heating (15′ at 48°C and 4′ at 56°C), and to the actions of cobra venom factor, zymosan and cysteine. Strain and sex differences with respect to AP50 activities of mouse sera were observed.     

Comparison of immunological adjuvants

In this study, several innate immunological adjuvants and related compounds were compared with respect to complement activation in serum and induction of cytokine release in whole blood samples using immunoassays. As found, simple lipids had no effect on the complement system or on cytokine release, whereas lipopolysaccharides induced prominent release of pro‐inflammatory cytokines (IL1β, TNF and IFNγ) without affecting the complement system, except for one, which activated the lectin pathway (LP). Moreover, saponin induced IL1β and MCP1 release and did not affect the complement system. The polysaccharide inulin exhausted the alternative pathway (AP) completely without affecting the LP and the classical pathway (CP), whereas zymosan exhausted the AP and had a major effect on the LP and CP as well. Peptidoglycans mainly affected the LP. Inulin, agarose and cellulose induced IL1β and MCP1 release, while dextran had no effect on cytokine secretion. Zymosan mainly induced IL1β release. The inorganic compound aluminium hydroxide, Al(OH)₃, activated the complement system very efficiently (all three pathways) but only induced MCP1 release. Other compounds tested had minor/individual effects. Collectively, well‐known adjuvants, such as aluminum hydroxide, activated the complement system and/or induced pro‐inflammatory cytokine release. Since complement activation generates anaphylactic peptides, a simple definition of an (innate) immunological adjuvant can be inferred: it activates the (innate) immune system by complement activation and/or release of cytokines so as to attract cells of the adaptive immune system.     

Trichosanthin, an initiator of the alternative complement activation pathway.

Trichosanthin (TCS), a protein purified from the plant Trichosanthes Kirilowii Maxim, activates normal human serum complement via the alternative pathway, as shown by TCS-induced C3 conversion in normal serum and its prevention by depletion of factor B, but not with the addition of EGTA. Injection of TCS to BALB/c mice consumed the complement alternative pathway (APC) activity in serum, implying in vivo activation of the alternative pathway by TCS. Elevation of peripheral blood leucocyte count as well as protein exudation and neutrophil accumulation in the peritoneal cavities could be induced by peritoneal injection of TCS. The main effect of complement activation by TCS was demonstrated to be induction of neutrophil accumulation.     

Complement activation by malignant B cells from patients with chronic lymphocytic leukaemia (CLL).

It has previously been reported that the expression of the complement receptors CR1 (CD35) and CR2 (CD21) on malignant B cells in CLL is reduced compared with the expression on normal B cells, while deposition of complement C3 fragments, as a consequence of alternative pathway (AP) activation of complement, is observed on mononuclear cells from patients with B CLL. Following our demonstration that normal B cells are capable of activating the AP of complement in a CR2-dependent fashion, we have chosen to re-examine the complement-activating ability of B CLL cells in relation to their altered phenotype with respect to CR2 and the complement regulatory membrane proteins, CR1, decay accelerating factor (DAF) (CD55) and membrane cofactor protein (MCP) (CD46). Flow cytometry was used to measure expression of complement receptors and regulatory proteins on CD5+ B cells from CLL patients, as well as the deposition of C3 fragments occurring both in vivo and after in vitro AP activation. We have confirmed the reduced expression of CR1 and CR2 on CLL cells and have shown that AP activation in the presence of homologous, normal serum was reduced on B CLL cells compared with normal B cells. The degree of AP activation correlated directly with CR2 expression. In addition, we observed that CLL cells bear in vivo-deposited C3d,g, although at a significantly lower level than normal B cells.     

Control of immune complex and zymosan-mediated anaphylatoxin generation by proteins B and H of the alternative complement pathway.

The generation of histamine releasing activity (HRA) from human basophils in fresh serum by tetanus toxoid (Te)/anti-Te complexes or by zymosan can be modulated through introduction of incremental amounts of proteins B and H of the alternative complement pathway. Serum treated at 50 degrees in order to abolish alternative pathway-mediated haemolytic activity, lost 90% of its capacity to generate HRA upon addition of Te/aTe; such loss could be reversed through additions of purified B. Amounts of B sufficient to restore normal alternative pathway haemolytic activity also restored HRA induced by Te/aTe; as little as a 33% increase above the normal serum concentration of B increased the capacity to support Te/aTe induced HRA by a factor of 1.4. In contrast, additions of incremental doses of purified H to fresh serum reduced generation of HRA by both Te/aTe and zymosan. Total inhibition was achieved by increasing the serum H concentration by 12.5-30%; further increases of H up to 200% again permitted HRA generation induced by immune complexed aTe. H also inhibited Te/aTe induced HRA in a serum heated at 50 degrees but only 30% inhibition of HRA could be achieved over a range of H inputs up to 187% above normal serum concentration. Additions of H also inhibited HRA generation in fresh serum when induced with plain or C3b-coated zymosan (Z) particles. By increasing the serum concentration of H from 12.5 to 125%, dose-dependent inhibition of HRA generation was observed; the H input necessary to suppress 48% of HRA generation was ten times higher when HRA was generated by Z-C3b than by plain zymosan. Thus, the complement-dependent generation of HRA from fresh serum strongly depends on modest variations in the concentrations of the two regulatory proteins B and H of the alternative complement pathway, suggesting their direct effect on generation of anaphylatoxins C3a and C5a.     

The effect of the antithyroid drug propylthiouracil on the alternative pathway of complement in rats

The effect of propylthiouracil (PTU) on the lytic activity of complement in rat serum was investigated in vivo. Rats (180+/-10 g) were treated daily by gavage with PTU doses of 1-50 mg/200 g body weight for time intervals ranging from 1 to 30 days. Serum classical pathway (CP) and alternative pathway (AP) activities were determined 24 h after the last dose. A single dose of 50 mg/200 g body weight was administered to additional groups and the animals were sacrificed after periods of 1-48 h.The results showed a relatively small reduction ( approximately 30%) in CP activity, evident only in animals treated with 50 mg of PTU for three weeks. However, a clear and opposite effect of PTU, an increase in lytic activity reaching values up to 180% of controls, was observed on AP activity. This effect was seen at all PTU doses used, and occurred within 4 days of treatment with the highest dose. Maximum activity was observed at intermediate intervals, depending on the PTU dose, with a return to control levels occurring after the longer periods of treatment. The lytic activity of serum from animals treated with a single PTU dose of 50 mg/200 g body weight and sacrificed 1-48 h after dosing did not differ from controls. Serum levels of thyroid hormone (triiodo L-thyronine, T3, and thyroxine, T4) were determined in representative groups of treated animals (injected with 5 mg of PTU/200 g body weight/day). These were either undetectable or considerably lower than those of controls. The serum PTU levels of these rats increased for up to 22 days, reaching values of 2-4 microg/ml.PTU is described in the literature as a modulator of both cellular immune responses and antibody production. Upon complement activation fragments of complement components bind to immune complexes and to specific receptors on cells of the immune system. Thus, alteration in AP activity caused by PTU treatment suggests a possible mechanism by which the drug exerts its modulatory effect. Increased complement AP activity might affect events as antigen presentation and hence the onset and course of the immune response.     

Effect of supraphysiologic levels of C1-inhibitor on the classical, lectin and alternative pathways of complement

C1-inhibitor is increasingly used experimentally and clinically in inflammatory conditions like septicemia and ischemia-reperfusion injury. Several mechanisms may account for the anti-inflammatory effects of C1-inhibitor, including inhibition of complement. The aim of the present study was to investigate and compare the supraphysiologic effect of C1-inhibitor on the three complement pathways. Novel assays for specific evaluation of the classical, lectin and alternative pathways were employed using normal human serum supplemented with increasing concentrations of C1-inhibitor. Solid-phase classical- and lectin pathway activation was dose-dependently and significantly reduced up to 85% in the range of 2-28 times physiologic C1-inhibitor concentration. The lectin pathway was more potently inhibited than the classical at low doses. A functional lectin pathway assay demonstrated a significant reduction of C4 deposition up to 86% even at low concentration of C1-inhibitor and documented the effect to be at the level of MBL/MASPs. In contrast, C1-inhibitor had no effect on solid-phase alternative pathway activation, but significantly reduced cobra venom factor-induced fluid-phase activation up to 88%. The negative controls albumin and IgG had no effect on complement activation. The positive inhibitory controls compstatin (C3 inhibition), EDTA- or MBL-deficient sera reduced complement activation by 82-100%. We conclude that C1-inhibitor in high physiologic doses differentially inhibits all three-complement pathways. The inhibition pattern was strikingly different in the classical and lectin pathway, compared to the alternative. Previous studies interpreting the effects of C1-inhibitor as only due to classical pathway inhibition needs reconsideration. The data has implications for the therapeutic use of C1-inhibitor.     

Inhibition of human and guinea pig complement by heparin fractions differing in affinity for antithrombin III or in average molecular weight

Heparin comprises a mixture of structurally related molecules. Affinity for antithrombin III (AT III) is a prerequisite for its anticoagulant activity, which also is dependent on its molecular weight. In this study heparin fractions prepared by affinity chromatography on immobilized AT III and by gel filtration chromatography were compared for their ability to inhibit complement mediated haemolysis and both classical and alternative pathway C3 activation as measured by crossed immunoelectrophoresis. In normal human serum, inhibition of haemolysis and of heat-aggregated IgG (HAGG) as well as zymosan induced C3 activation by heparin was found to be independent of its AT III affinity and of its molecular weight (range 4800-17,000), on a weight basis. In guinea pig serum similar results were obtained for inhibition of HAGG induced C3 activation, but inhibition of haemolysis showed a marked molecular weight dependency and was also reduced for the fraction with low affinity for AT III. This may reflect a species difference in the haemolytic action of human and guinea pig complement. It is concluded that inhibition of human complement by heparin is independent of its anticoagulant activity and of its size and it is suggested that a fraction of heparin with reduced risk for bleeding and platelet aggregation is a potential anti-inflammatory agent.     

Activation of the alternative complement pathway by IgM antibody reacted on paragloboside incorporated into liposome membrane

Liposomes containing paragloboside (PG) on their membrane were readily lysed by C4 deficient guinea pig serum (C4D-GPS). On the other hand, guinea pig serum from specific pathogen free Hartley guinea pigs (SPF-GPS) did not lyse PG-liposomes in Mg-EGTA-GVB (gelatin veronal buffered saline containing MgCl2 and ethyleneglycol-bis(beta-aminoethyl ether)N,N'-tetraacetate) which permits complement activation via the alternative pathway but not via the classical pathway. However, the SPF-GPS could lyse the liposomes in Mg-EGTA-GVB when heated (56 degrees C, 30 min) C4D-GPS or other guinea pig serum (GPS) was added. Gel filtration of Hartley (Htl)-GPS through a Sephadex G-200 column revealed that IgM antibody to PG in Htl-GPS was responsible for sensitization of the liposomes to lysis by complement of SPF-GPS via the alternative pathway. This result indicated that guinea pig IgM antibody can initiate the activation of the alternative pathway of homologous complement on liposome membrane.     

Activation of the alternative pathway of human complement by autologous cells expressing transmembrane recombinant properdin

Properdin (P) is a serum glycoprotein that stabilizes the labile C3 convertase (C3bBb) of the alternative pathway of the complement system (AP). Thanks to its oligomeric nature, P specifically upregulates AP on surfaces without activating AP in the fluid-phase. We investigated whether human cells, displaying P at their membrane, could activate autologous AP. The cDNAs encoding human P and the transmembrane domain of human platelet derived growth factor receptor were fused together and expressed in human embryo kidney cells (HEK-293). Selected cells displayed P at their surface as shown by FACS. In contact with human serum at 37 degrees C, they triggered AP-mediated C3 deposition. SDS-PAGE analysis showed C3 covalently bound to various membrane proteins, but not to P itself. However, displayed P affinity could bind to serum or purified C3i at 4 degrees C. C3 binding was restricted to the cells displaying P, was inhibited by an anti-P mAb, and did not require serum P. Bound C3 allowed further C5, C7 and C9 deposition as well as cell lysis after blocking CD59 function. In contrast, wild-type cells, cells displaying factor D or truncated P (deleted from its 6th thrombospondin-like repeat) did not activate AP. We hypothesize that displayed P activates AP by stabilizing bystander C3b and/or by capturing serum C3iBb convertase. Finally, we suggest that P could be used for retargeting autologous complement to AP-resistant pathogens and tumor cells.