Interactions of complement proteins C1q and factor H with lipid A and Escherichia coli: further evidence that factor H regulates the classical complement pathway

Abstract

Proteins of the complement system are known to interact with many charged substances. We recently characterized binding of C1q and factor H to immobilized and liposomal anionic phospholipids. Factor H inhibited C1q binding to anionic phospholipids, suggesting a role for factor H in regulating activation of the complement classical pathway by anionic phospholipids. To extend this finding, we examined interactions of C1q and factor H with lipid A, a well-characterized activator of the classical pathway. We report that C1q and factor H both bind to immobilized lipid A, lipid A liposomes and intact Escherichia coli TG1. Factor H competes with C1q for binding to these targets. Furthermore, increasing the factor H: C1q molar ratio in serum diminished C4b fixation, indicating that factor H diminishes classical pathway activation. The recombinant forms of the C-terminal, globular heads of C1q A, B and C chains bound to lipid A and E. coli in a manner qualitatively similar to native C1q, confirming that C1q interacts with these targets via its globular head region. These observations reinforce our proposal that factor H has an additional complement regulatory role of down-regulating classical pathway activation in response to certain targets. This is distinct from its role as an alternative pathway downregulator. We suggest that under physiological conditions, factor H may serve as a downregulator of bacterially-driven inflammatory responses, thereby finetuning and balancing the inflammatory response in infections with Gram-negative bacteria.

Graphical abstract

Keywords

complement / lipid A / bacteria / factor H / C1q

INTRODUCTION

The human complement system plays a major role in resistance against microbial infections via antibody-dependent and antibody-independent mechanisms. Complement is known to recognize lipopolysaccharide (LPS), which is the main amphiphilic component of the outer layer of Gram-negative bacterial membranes. LPS is typically composed of three regions: lipid A, core polysaccharide and O-specific polysaccharide (Rietschel et al., 1994). Toxicity of LPS is associated with the lipid component (lipid A) while immunogenicity is associated with the polysaccharide components. The structure of lipid A is highly conserved among eubacteria. Lipid A is composed of a β-(1→6)-linked D-glucosamine disaccharide substituted at positions 4′ and 1 by phosphoester groups, with fatty acids linked to the remaining hydroxyl and amino groups (Rietschel et al., 1994).

LPS can activate complement via the classical, lectin and alternative pathways, resulting in the generation of pro-inflammatory anaphylatoxins (C3a, C4a and C5a) and subsequent recruitment of neutrophils and macrophages. In addition, C3b and iC3b opsonize both Gram-negative and Gram-positive bacteria (Joiner et al., 1986). The formation of the membrane attack complex and subsequent lysis of bacterial cells has been shown for a limited number of Gram-negative bacteria (Joiner et al., 1985). Opsonophagocytosis remains the major anti-bacterial defense mechanism.

LPS activates complement either in the presence of antibodies to its O-specific polysaccharide chain or by direct interaction with components of the complement system (Morrison and Kline, 1977). The lipid A region of LPS is responsible for classical pathway complement activation that involves a direct interaction between lipid A and C1q leading to the activation of C1 (Cooper and Morrison, 1978; Arvieux et al., 1984; Loos and Clas, 1987; Clas et al., 1989). The polysaccharide region, according to its composition, may stimulate alternative and lectin pathway activation through lipid A-independent mechanisms (Morrison and Kline, 1977; Jack and Turner, 2003). We have recently found that C1q and factor H bound to immobilized as well as liposomal anionic phospholipids, most notably to cardiolipin. Factor H strongly inhibited C1q binding to solid phase cardiolipin and cardiolipin liposomes, suggesting a role for factor H in regulating activation of the classical complement pathway by anionic phospholipids (Tan et al., 2010). In the work presented here, we investigated whether lipid A, a well-established classical pathway activator, also interacts with factor H as well as C1q, and followed this by tests of binding to a laboratory strain of E. coli as a source of lipid A.

Complement activation by the classical pathway occurs when the first component of complement, C1, binds via charge interactions to immune complexes containing IgG or IgM or directly to non-immunoglobulin target surfaces such as anionic phospholipids and Gram-negative bacteria (Sim and Malhotra, 1994; Gaboriaud et al., 2003; Kojouharova et al., 2004; Kishore et al., 2004a; Roumenina et al., 2006; Tan et al., 2010). C1 is made up of three glycoproteins, C1q, C1r and C1s. C1r and C1s are serine proteases that become activated when C1q binds to a target. C1q is a 460 kDa glycoprotein composed of 18 polypeptide chains (6A, 6B and 6C) and is present in human serum at a concentration of about 115 mg/L (Reid and Porter, 1976; McAleer and Sim, 1993; Dillon et al., 2009). The A, B and C chains, each about 25 kDa, combine to form six equivalent subunits each with a collagen-like “stalk” at the amino-terminus and a carboxyl-terminal heterotrimeric globular (gC1q) domain. The six subunits associate to form an umbel shape. The C1r and C1s subcomponents of C1 bind the collagen region of C1q. Binding of C1q to ligands, which occurs via the gC1q domains, brings about a conformational change in the collagen region that activates C1r and C1s (Dodds et al., 1978).

Factor H is a single polypeptide chain protein of 155 kDa, composed of 20 complement control protein (CCP) modules (Ripoche et al., 1988). Its concentration in human sera is reported to vary widely between individuals (Charlesworth et al., 1979; Maruvada et al., 2008; Edey et al., 2009; Ingram et al., 2010) up to over 800 mg/L, but probably on average 200 mg/L (Sim et al., 1993). It is well-known as a regulator of the complement alternative pathway. It prevents the assembly of and dissociates the alternative complement pathway C3 convertase, C3bBb, and acts as a cofactor for factor I-mediated cleavage of C3b (Weiler et al., 1976; Whaley and Ruddy, 1976; Fearon, 1978; Sim et al., 1993). The discrimination between activators and non-activators of the alternative pathway is determined by binding of factor H to surface-associated C3b (Meri and Pangburn, 1990). Factor H binds directly to many microorganisms and parasites where it is thought to downregulate complement alternative pathway activation (Díaz et al., 1997; Kraiczy and Würzner, 2006).

Since both C1q and factor H have crucial roles in the recognition and removal of bacteria, we examined whether factor H might regulate C1q binding to lipid A, thus limiting the level of the activation of the classical pathway. The study was undertaken with lipid A coated on wells and incorporated into liposomes. The binding characteristics of both complement proteins to a Gram-negative bacterium, Escherichia coli strain TG1, were also investigated.

RESULTS

Interaction of ¹²⁵I-C1q, ¹²⁵I-factor H, ¹²⁵I-ghA, ¹²⁵I-ghB or ¹²⁵I-ghC with lipid A

Purified radioiodinated proteins were incubated in microtiter wells coated with lipid A. C1q bound strongly to lipid A-coated wells, with lower binding by factor H and ghC (Fig. 1a). Lowest binding to lipid A was seen with ghA and ghB but this was still above the background levels observed with negative control wells coated with phosphatidylcholine (PC). Cardiolipin (CL)-coated wells were used as a positive control (Tan et al., 2010).

Lipid A liposomes were also incubated with radioiodinated proteins and the binding was assessed (Fig. 1B). Factor H, ghA and ghC showed strongest binding to lipid A liposomes. The binding of C1q and ghB was less but well above background binding levels observed to control PC liposomes. CL liposomes were used as a positive control. These results show that C1q and factor H bind to lipid A and that the recombinant globular head regions of C1q bind differentially to lipid A. As noted previously (Tan et al., 2010), the oligomerization state of the recombinant gHA, gHB, gHC proteins varies, although they are mostly in trimer form. These materials may exhibit higher, but more usually lower, binding affinity than C1q, depending on polymerization state. They may possess potential binding sites not present in native C1q, because of different domain-domain interfaces.

Saturation binding of ¹²⁵I-C1q to lipid A

A fixed amount of ¹²⁵I-C1q was incubated with increasing amounts of unlabelled C1q on lipid A-coated wells or lipid A liposomes. The C1q binding to lipid A-coated wells (Fig. 2) and lipid A liposomes (Fig. 2) was dose dependent and saturable. ¹²⁵I-C1q binding to lipid A-coated wells and lipid A liposomes was saturated at a final concentration of 10 μg/mL C1q, which is well below the plasma concentration of ~115 μg/mL.

Inhibition of ¹²⁵I-C1q binding to lipid A by factor H

Since both C1q and factor H bound to lipid A, we examined whether they competed with each other. On lipid A-coated wells, a 6.75-fold molar excess of unlabelled factor H over ¹²⁵I-C1q was able to inhibit C1q binding by 40% (Fig. 3A). Higher amounts of factor H diminished C1q binding by > 80%. The molar ratio of factor H to C1q in serum may vary between about 5:1 and 30:1. Ovalbumin (OVA) did not influence the binding of ¹²⁵I-C1q to lipid A-coated wells.

When ¹²⁵I-C1q was incubated with 6.75-fold molar excess of factor H and mixed with lipid A liposomes, there was 30% decrease in C1q binding to lipid A liposomes (Fig. 3B) and again higher amounts of factor H diminished C1q binding by about 80%. OVA again did not interfere with the binding. These results indicate that C1q and factor H compete for binding to lipid A.

C1q binds to lipid A via its globular heads

The recombinant ghA, ghB and ghC modules were shown to bind to both lipid A-coated wells (Fig. 1A) and lipid A liposomes (Fig. 1B). To ascertain if the native C1q bound lipid A via its gC1q domain, ¹²⁵I-C1q was incubated with unlabelled ghA, ghB or ghC on lipid A-coated wells and lipid A liposomes.

C1q binding to lipid A-coated wells was inhibited most strongly by ghC (Fig. 3C) with weaker inhibition by ghA and ghB. The binding of C1q to lipid A liposomes was also strongly competed out by ghC and ghA with weaker inhibition by ghB (Fig. 3D), confirming that C1q binds to lipid A via its gC1q domain. The ghC module appeared to be the strongest lipid A binder (Fig. 1A and 1B), consistent with the inhibition assays where ghC was most effective in competing out C1q binding to lipid A-coated wells (Fig. 3C) and lipid A liposomes (Fig. 3D). As shown in Fig. 3C and 3D, the competition was incomplete since < 50% inhibition was achieved at the highest ghA, ghB and ghC concentrations used. This may be associated with the number of binding sites per molecule on C1q, compared with the ghA, ghB and ghC modules. C1q is likely to engage up to 18 binding sites, while ghA, ghB and ghC have, in principle, 3 sites each. The binding of C1q should therefore be of higher avidity, and the kinetics of binding may also be different. Alternatively, there may be a contribution of the C1q collagen region to the binding.

Binding of ¹²⁵I-C1q, ¹²⁵I-factor H, ¹²⁵I-ghA, ¹²⁵I-ghB or ¹²⁵I-ghC to E. coli TG1

Since all these proteins bind to lipid A isolated from E. coli, we tested the binding to the whole bacteria. Radioiodinated C1q, factor H, ghA, ghB and ghC did bind to the bacterial cells (Fig. 4). Factor H especially showed high binding to the bacterial cells, with approximately 20% of the protein incubated binding to the cells (Fig. 4). C1q binding was about 12% while the recombinant globular head regions showed 6%–8% binding to the bacterial cells. Relative binding by the different proteins is similar to that seen with lipid A liposomes (Fig. 1B), suggesting that binding to lipid A may be a major contributor to the total binding observed.

Saturation binding of ¹²⁵I-C1q and ¹²⁵I-factor H to E. coli

When ¹²⁵I-C1q or ¹²⁵I-factor H was incubated with their respective unlabelled counterparts together with the bacterial cells, we observed that C1q and factor H binding to E. coli was dose dependent. C1q binding was saturable at 550 ng of C1q (approximately 2 μg/mL final concentration) (Fig. 5A) while the saturation of factor H binding to E. coli was observed at an input of 800 ng (approximately 2.7 μg/mL) (Fig. 5B). The numbers of molecules per cell bound at saturation correspond to about 400 for C1q, and 1000 for factor H.

Specific binding of ¹²⁵I-C1q and ¹²⁵I-factor H to E. coli

Since factor H and C1q binding to E. coli was saturable, we examined if the binding of both proteins to E. coli was specific. Inhibition of ¹²⁵I-C1q with unlabelled C1q, using OVA as a control, was first studied. At a molar ratio of unlabelled C1q to labelled C1q of 4:1, ¹²⁵I-C1q binding was inhibited by about 50% (Fig. 5C). Thus, the radioiodinated C1q bound slightly better than unlabelled C1q, suggesting limited modification of binding properties during iodination (for example, by oxidation). Alternatively, specific binding may be superimposed on a significant non-specific component. OVA did not inhibit ¹²⁵I-C1q binding to E. coli.

When the specificity of ¹²⁵I-factor H binding to bacterial cells was examined, a 5:1 molar ratio of unlabelled factor H to labelled factor H inhibited ¹²⁵I-factor H binding by about 50% (Fig. 5D). Again this may indicate some damage to the protein on radioiodination or a non-specific component to the binding, as for C1q. OVA did not interfere with ¹²⁵I-factor H binding. These results show that the binding of C1q and factor H to E. coli have specific and saturable components, but there may be a significant background of non-specific binding.

Saturation binding of ¹²⁵I-ghA, ¹²⁵I-ghB and ¹²⁵I-ghC to E. coli

¹²⁵I-ghA, ghB and ghC were incubated with various quantities of unlabelled ghA, ghB and ghC respectively and it was shown that all three proteins bound to E. coli in a dose-dependent manner. Saturation was observed at an input of 1.55 μg for ghA (Fig. 6A), 2.05 μg for ghB (Fig. 6B) and 3.05 μg for ghC (Fig. 6C). Saturation required a higher quantity of these proteins than for C1q (0.55 μg), consistent with probable difference in avidity as discussed above.

Specific binding of ¹²⁵I-ghA, ¹²⁵I-ghB and ¹²⁵I-ghC to E. coli

Labelled ghA, ghB and ghC were incubated with their respective unlabelled counterpart and the binding of ¹²⁵I-ghA, ¹²⁵I-ghB and ¹²⁵I-ghC to E. coli was assessed. A 5:1 molar ratio of ghA to labelled ghA yielded about 50% decrease in ¹²⁵I-ghA binding (Fig. 6D) and similar results were obtained for ghB and ghC. OVA did not inhibit the binding of all three labelled proteins to E. coli. These results demonstrate that ghA, ghB and ghC binding to E. coli have a specific and saturable component.

C1q and factor H compete for binding to E. coli

Since both C1q and factor H bind to E. coli in a specific and saturable manner, we extended our study to whole bacterial cells in order to compare our results with immobilized and liposome lipid A (Fig. 3A and 3B). When ¹²⁵I-C1q was mixed with unlabelled factor H, there was partial inhibition of C1q binding to E. coli. When a 9:1 molar ratio of unlabelled factor H to C1q was incubated (1.66 μg of factor H), there was about 20% inhibition of ¹²⁵I-C1q binding (9:1 is within the range of molar ratios found in plasma) (Fig. 7A). At the highest quantity of factor H used (approx 40:1 molar ratio to C1q) inhibition approached 50%.

Unlabelled C1q also competed with ¹²⁵I-factor H for binding (Fig. 7B) and about 25% inhibition was seen with a 1:1 C1q to factor H molar ratio (2.4 μg C1q). Inhibition approached 50% at the highest C1q concentration used (15 μg C1q, molar ratio C1q : factor H = 6.3:1). Since the relative number of binding sites and their affinities differ for C1q and factor H, it is possible that different incubation times would give a different binding profile. The OVA control did not interfere with ¹²⁵I-C1q or ¹²⁵I-factor H binding to bacterial cells. These results demonstrate that C1q and factor H do compete for binding to E. coli but only partial inhibition was achieved. While factor H and C1q share some binding sites on bacterial cell surfaces, there are likely to be other sites on E. coli that are uniquely bound by either factor H or C1q.

Assessment of C4b deposition on lipid A-coated wells

Since factor H inhibited C1q binding to lipid A-coated wells and liposomes, it was of interest to evaluate if this was of significance in the context of regulation of classical complement pathway activation. We used factor H- and C1q-depleted serum repleted with various C1q and factor H ratios, incubated on lipid A-coated wells. The extent of activation of the classical pathway was assessed by measuring C4b bound to the wells. We observed a trend that an increase in the factor H concentration decreases the amount of C4b deposition (Fig. 8). These results are consistent with a regulatory role for factor H in down-regulating classical complement pathway activation.

DISCUSSION

LPS is the major component of the outer membrane of Gram-negative bacteria and a well known activator of the humoral and cellular components of the host defense system (Rietschel et al., 1994). Activation of host immune mechanisms, including the complement system, is essential to fight infection with Gram-negative bacteria, although excessive stimulation can also result in the serious-life threatening symptoms of septic shock (Bone 1991). LPS activates complement either in the presence of antibodies to its O-specific polysaccharide chain or by direct interaction with components of the complement system (Morrison and Kline, 1977). They first demonstrated that the lipid A region of LPS is responsible for classical pathway complement activation while the polysaccharide region is responsible for alternative pathway activation through a lipid A-independent mechanism. More recent data indicate that the polysaccharide regions may also activate the lectin pathway (Jack and Turner, 2003; Shang et al., 2005). Direct interaction of isolated lipid A with C1q leading to the activation of C1 was reported by Cooper and Morrison (1978).

We have studied interactions of C1q and factor H with lipid A purified from E. coli and also intact E. coli cells. C1q was shown to bind to lipid A coated on wells (Fig. 1A and 1B) and in liposomes. The binding of C1q is predominantly via its globular heads (Fig. 1A, 1B, 3A and 3D). The observation of factor H binding to both forms of lipid A is novel (Fig. 1A and 1B). Since C1q and factor H both bind to lipid A, we examined if factor H competes with C1q for anionic binding sites like those presented by lipid A. As illustrated in Fig. 3A and 3C, unlabelled factor H does effectively compete with ¹²⁵I-C1q binding to lipid A. This novel observation suggests that factor H has an additional regulatory role of down-regulating classical pathway activation. This was confirmed (Fig. 8) by showing that increasing the factor H: C1q molar ratio in serum diminishes C4b fixation.

We extended our study to E. coli TG1 cells and showed that C1q (Fig. 4 and 5C), the recombinant forms of the globular head modules of human C1q (Fig. 4 and 6D) and factor H (Fig. 4 and 5D) bind specifically to E. coli. In addition, factor H was able partially to inhibit C1q binding to E. coli cells (Fig. 7A) and vice versa (Fig. 7B). When comparing the extent of inhibition by factor H of C1q binding to purified lipid A (Fig. 3A and 3B) and whole E. coli cells (Fig. 7A), it is clear that factor H competed out C1q binding more effectively using purified lipid A compared with whole bacterial cells. These results suggest that, on E. coli, C1q and factor H share overlapping binding sites, but that C1q and factor H have other distinct, unshared binding sites. C1q or the C1 complex has been reported to bind to several candidate molecules on Gram-negative bacteria. These include porins or outer membrane proteins from Aeromonas, Klebsiella, Salmonella and Brucella species (Latsch et al., 1992; Albertí et al., 1996; Eisenschenk et al., 1999; Merino et al., 2005) as well as lipid A and LPS components such as polyribosyl-ribitolphosphate from Haemophilus influenzae (Bunse and Heinz, 1993). Other reports indicate other modes of C1q binding via its collagen region, in a manner which does not activate the classical pathway. This has been reported for a 51/57 kDa protein from E. coli (van den Berg et al., 1996) and also the LPS component 2-keto-3-deoxyoctonic acid (KDO) from E. coli (Zohair et al., 1989). For the interaction of factor H with Gram-negative bacteria, there are many reports of specific factor H-binding bacterial molecules (Kraiczy and Würzner, 2006; Zipfel et al., 2007). Factor H binding to Neisseria gonorrhoeae and N. meningitidis has been characterized. Factor H is reported to bind to sialylated lipooligosaccharides (LOS) in gonococci, but not in meningococci (Gulati et al., 2005; Schneider et al., 2006). The protein por1A of gonococci and a 33 kDA protein in meningococci bind factor H, as does the protein YadA of Yersinia (Kraiczy and Würzner, 2006; Gulati et al., 2005; Schneider et al., 2006). There are no previous reports of direct binding of isolated factor H to E. coli, although binding of factor H during complement activation in serum has been observed (Kubens et al., 1989; Maruvada et al., 2008). These reports do not distinguish direct binding from binding via surface C3b.

The deposition of C4b on lipid A-coated wells by serum depleted of factor H and C1q and repleted with various ratios of both proteins was also examined (Fig. 8). With an increase in the molar ratio of factor H to C1q within the physiological ratio (about 5:1 up to 30:1), there was a steady decrease in the amount of C4b deposited on the lipid A coated wells. Similarly, C4b deposition increased when the ratio of factor H to C1q was decreased below the physiological ratio. These results are consistent with C1q (in the form of C1) and factor H in serum competing for binding to lipid A, resulting in either an increase or decrease in the activation of the classical pathway. Under physiological conditions, factor H may serve as a downregulator of bacterially-driven inflammatory responses, thereby finetuning and balancing the inflammatory response in infections with Gram-negative bacteria.

Serum amyloid P component (SAP), which is constitutively expressed in plasma at a level of about 30–50 μg/mL, has been shown to prevent LPS-mediated activation of the classical pathway (de Haas et al., 2000). SAP binding to the lipid A component of LPS prevented the deposition of C1q and hence interfered with the antibody-independent activation of the classical pathway. SAP binding, however, did not interfere with antibody-mediated C1q activation by LPS/antibody complexes nor alternative pathway activation. In contrast, C1q has been suggested to bind SAP and activate the classical pathway (Ying et al., 1993), but SAP bound to lipid A might present itself in a conformation which is unsuitable for C1q binding and/or classical pathway activation. The inhibition of C1q binding to Gram-negative bacteria by SAP, and in our study, by factor H, might therefore influence the pathophysiology of infection with such bacteria. It is also likely, although this is outside the scope of this study, that factor H may modulate lipid A-mediated macrophage activation that is enhanced by C1q. Thus, factor H may have a scavenging role as an anti-inflammatory mediator where the inflammatory reactions involve interaction between C1q and bacteria or LPS.

MATERIALS AND METHODS

Buffers

The following buffers were used: Veronal buffers (VB, 5 mmol/L sodium barbital, 142 mmol/L NaCl, pH 7.4; VB²⁺, 5 mmol/L sodium barbital, 142 mmol/L NaCl, 0.15 mmol/L CaCl₂, 0.5 mmol/L MgCl₂, pH 7.4) and SDS-PAGE sample buffer (0.2 mol/L Tris-HCl, 8 mol/L urea, 2% w/v SDS, pH 8.0).

Purified proteins and antibodies

C1q was purified from pooled human serum using affinity chromatography on IgG-Sepharose (Tan et al., 2010). Human factor H, purified by monoclonal antibody affinity chromatography (Sim et al., 1993) was supplied by BE Moffatt, MRC Immunochemistry Unit, Oxford. The globular head regions of human C1q A (ghA), B (ghB) and C (ghC) chains were expressed in E. coli as fusions to maltose binding protein (MBP) and purified as described previously (Kishore et al., 2003).

Radioiodination of proteins

Factor H was radioiodinated with ¹²⁵I as previously described (Tan et al., 2010). Typical specific activity obtained was 1.1 × 10⁷ cpm/μg (counting efficiency of 70%). C1q, ghA, ghB and ghC were radioiodinated by a less oxidizing procedure (Tan et al., 2010). Typical specific activities were (2.8–9.2) × 10⁵ cpm/μg for C1q and (1.5–1.9) × 10⁶ cpm/μg for ghA, ghB and ghC. For some experiments, radioiodinated components were diluted with unlabelled material to reduce the specific activity.

Preparation of normal human serum and serum depleted of both C1q and factor H

Plasma pooled from at least 20 donors (HD Supplies, Aylesbury, UK) was recalcified to obtain serum by adding 1 mol/L CaCl₂ to a final concentration of 16 mmol/L and the plasma was left to clot overnight at 4°C. The serum was then collected by filtering through muslin. Serum depleted of C1q and factor H was made as described by Tan et al. (2010).

Depletion of C1q was tested by hemolytic assay as previously described (Tan et al., 2010). To measure the extent of factor H depletion, the activity of factor H as a cofactor for factor I-mediated cleavage of C3b/inactive C3 was tested. ¹²⁵I-labelled-C3 (NH₃) was used as the substrate (Sim and Sim, 1983). This was supplied by Dr. Stefanos Tsiftsoglou, MRC Immunochemistry Unit, Oxford. Serial 2-fold dilutions of control and depleted sera were diluted with PBS-EDTA containing 5 μg/mL soybean trypsin inhibitor (Type IS-SBTI: Sigma) and incubated with a fixed amount of ¹²⁵I-C3(NH₃) for 12 h at 37°C. The reaction was stopped with SDS-PAGE sample buffer containing 20 mmol/L DTT. The samples were run on a 10% SDS-PAGE gel (Laemmli, 1970), dried down and exposed to X-ray film (Fuji RX, Fuji Photo Film UK Ltd., London, UK) in autoradiography cassettes with intensifying screen for 8 h at −70°C. The extent of C3(NH₃) cleavage by factor I in serum is proportional to the amount of its co-factor, factor H, present in serum. The amount of cleavage was judged by the loss of the C3(NH₃) α chain visualized on the X-ray film. The results showed that factor H was depleted by 75% relative to the control serum.

Binding of ¹²⁵I-C1q, ¹²⁵I-factor H, ¹²⁵I-ghA, B, C to solid-phase lipid A

Diphosphoryl lipid A from E. coli (Sigma L-5399) was used throughout. Lipid A (5 μg/well) in methanol: chloroform (2:1 v/v) solution was added to Maxisorp microtitre plate wells (Nunc, Kamstrup, Denmark) and solvents were allowed to evaporate at 4°C overnight. The wells were then blocked with 10 mmol/L potassium phosphate buffer pH 7.0 + 500 μg/mL ovalbumin for 2 h at room temperature. ¹²⁵I-C1q, ¹²⁵I-factor H and ¹²⁵I-ghA, ¹²⁵I-ghB and ¹²⁵I-ghC were separately added to lipid A coated wells (10,000 cpm/well of ¹²⁵I-C1q or ¹²⁵I-factor H, 150,000 cpm/well of ¹²⁵I-ghA, ¹²⁵I-ghB and ¹²⁵I-ghC) and incubated for 1 h at room temperature in a final volume of 100 μL of VB²⁺. The wells were then washed three times with 200 μL of VB²⁺. Two hundred microliters of 0.1 mol/L NaOH was added to each well and allowed to incubate for 5 min to dissociate bound proteins and 180 μL of the supernatant was collected and counted in a γ-counter.

Binding of ¹²⁵I-C1q, ¹²⁵I-factor H, ¹²⁵I-ghA, ¹²⁵I-ghB or ¹²⁵I-ghC to lipid A liposomes

Multilamellar lipid A liposomes were prepared according to established methods (New, 1990). The liposome preparations were composed of 25 molar percent dipalmitoylphosphatidylcholine (PC), 45 molar percent cholesterol (CHOL) with 30 molar percent of lipid A. Control PC liposomes contained only 55 molar percent PC and 45 molar percent CHOL. Quantitative determination of total phospholipids was performed by using the Stewart assay (Stewart, 1980). One hundred microliters of multilamellar phospholipid vesicles (100 μg/mL) were mixed with 200 μL of VB²⁺ + 750 μg/mL ovalbumin containing either 90,000 cpm ¹²⁵I-C1q, or 100,000 cpm ¹²⁵I-factor H or 150,000 cpm ¹²⁵I-ghA or ¹²⁵I-ghB or ¹²⁵I-ghC in 1.7 mL conical ultracentrifuge tubes which had been blocked overnight with 500 μL of VB²⁺ + 1 mg/mL ovalbumin and incubated for 1 h at room temperature. The tubes were then spun at 10,000 g in a microfuge for 10 min and the supernatant was removed. The pellet was washed once with 300 μL VB²⁺, spun and the pellet was counted in a γ-counter.

Saturation binding of ¹²⁵I-C1q to lipid A-coated wells and liposomes

A fixed amount of ¹²⁵I-C1q (250 ng/well or reaction) was premixed with various quantities of unlabelled C1q (0–3.75 μg) to lower the specific activity, so that saturation could be achieved without using high levels of radioactivity. The tubes were then incubated on ice for 1 h after which they were incubated with lipid A coated wells or lipid A liposomes, as described above, and the amount of C1q binding was calculated from the radioactivity bound and the known specific activity.

Inhibition of ¹²⁵I-C1q binding to lipid A-coated wells and lipid A liposomes by ghA, ghB, ghC or factor H

A fixed amount of ¹²⁵I-C1q in VB²⁺ (3 μg of protein/well or reaction) was incubated with various quantities of unlabelled ghA, ghB, ghC or factor H (0–54 μg for factor H, 0–18 μg for ghA, ghB, ghC) for 1 h on ice. The mixtures were then placed on lipid A-coated wells, or in contact with lipid A liposomes in 1.7 mL conical ultracentrifuge tubes (blocked overnight with 500 μL of VB²⁺ + 1 mg/mL ovalbumin) containing 100 μL of lipid A liposomes (100 μg/mL) for 1 h at room temperature. The lipid A coated wells or lipids A liposomes were washed and the amount of ¹²⁵I-C1q bound was determined as above.

Binding of ¹²⁵I-C1q, ¹²⁵I-factor H, ¹²⁵I-ghA, ¹²⁵I-ghB and ¹²⁵I-ghC to E. coli

E. coli TG1 was grown in Luria-Bertani broth (50 mL) at 37°C with vigorous agitation on a shaker to a concentration of 1 × 10⁹ cells/mL. The bacterial cells were harvested by centrifuging at 5000 g for 10 min at room temperature. The pellet was washed twice and resuspended in VB²⁺ + 0.5 mg/mL ovalbumin. One hundred microliters of VB²⁺ + 0.5 mg/mL ovalbumin containing 125 ng of ¹²⁵I-C1q, ¹²⁵I-factor H, ¹²⁵I-ghA, ¹²⁵I-ghB or ¹²⁵I-ghC was incubated with 200 μL of E. coli (1 × 10⁹ cells/mL in VB²⁺-0.5 mg/mL ovalbumin) in 1.7 mL conical ultracentrifuge tubes (blocked overnight with 500 μL of VB²⁺ + 1 mg/mL ovalbumin) at 37°C for 1 h with mixing every 15 min. Triplicates of the reaction mix (50 μL aliquots) were layered on 200 μL of 15% sucrose (d = 1.05 g/mL) in 0.4 mL microcentrifuge tubes. The tubes were spun for 1 min at 10,000 g and the tubes were snapfrozen in liquid nitrogen. The tips were cut off and the amount of radioactivity bound to the bacterial pellet was counted. A correction for non-specific entrapment of ¹²⁵I-C1q, ¹²⁵I-factor H, ¹²⁵I-ghA, ¹²⁵I-ghB and ¹²⁵I-ghC was done by taking an aliquot of the cell mix at the start of the experiment (within 30 s of mixing), spinning it down and quantifying the amount bound as described above.

Saturation binding of ¹²⁵I-C1q and ¹²⁵I-factor H to E. coli

¹²⁵I-C1q and ¹²⁵I-factor H (50 ng/reaction) were mixed with varying quantities of unlabelled C1q and factor H in 100 μL of VB²⁺ + 0.5 mg/mL ovalbumin respectively for 1 h on ice. The mix was then incubated with 200 μL of E. coli (1×10⁹ cells/mL in VB²⁺ + 0.5 mg/mL ovalbumin) in 1.7 mL conical ultracentrifuge tubes (blocked overnight with 500 μL of VB²⁺ + 1 mg/mL ovalbumin) and incubated at 37°C for 1 h with mixing every 15 min. The quantity of ¹²⁵I-C1q and ¹²⁵I-factor H bound was quantified as described above.

Specific binding of ¹²⁵I-C1q and ¹²⁵I-factor H to E. coli

To examine if the binding of ¹²⁵I-C1q or ¹²⁵I-factor H to E. coli was specific, 0.55 μg of ¹²⁵I-C1q or 0.80 μg of ¹²⁵I-factor H was incubated with increasing quantities of unlabelled C1q (0–3 μg) or factor H (0–10 μg), respectively, in 100 μL of VB²⁺ + 0.5 mg/mL ovalbumin and incubated on ice for 1 h. The mix was added to 200 μL of E. coli (1 × 10⁹ cells/mL in VB²⁺ + 0.5 mg/mL ovalbumin) in 1.7 mL conical ultracentrifuge tubes (blocked overnight with 500 μL of VB²⁺ + 1 mg/mL ovalbumin) and incubated at 37°C for 1 h with mixing every 15 min. The quantity of ¹²⁵I-C1q or ¹²⁵I-factor H bound was quantified as above.

Saturation binding of ¹²⁵I-ghA, ¹²⁵I-ghB or ¹²⁵I-ghC to E. coli

¹²⁵I-ghA, ¹²⁵I-ghB or ¹²⁵I-ghC (50 ng/reaction) was mixed with varying quantities of corresponding unlabelled recombinant globular head proteins in 100 μL of VB²⁺ + 0.5 mg/mL ovalbumin respectively for 1 h on ice. The mix was then incubated with 200 μL of E. coli (1 × 10⁹ cells/mL in VB²⁺ + 0.5 mg/mL ovalbumin) at 37°C for 1 h with mixing every 15 min. The quantity of ¹²⁵I-ghA, ¹²⁵I-ghB or ¹²⁵I-ghC bound was quantified as above.

Specific binding of ¹²⁵I-ghA, ¹²⁵I-ghB or ¹²⁵I-ghC to E. coli

To examine if the binding of ¹²⁵I-ghA, ¹²⁵I-ghB or ¹²⁵I-ghC to E. coli was specific, 1.55 μg of ¹²⁵I-ghA, 2.05 μg of ¹²⁵I-ghB or 3.05 μg of ¹²⁵I-ghC was incubated with increasing quantities of the corresponding unlabelled ghA, ghB and ghC in 100 μL of VB²⁺ 0.5 mg/mL ovalbumin and incubated on ice for 1 h. The mix was added to 200 μL of E. coli (1 × 10⁹ cells/mL in VB²⁺ + 0.5 mg/mL ovalbumin) in 1.7 mL conical ultracentrifuge tubes (blocked overnight with 500 μL of VB²⁺-1 mg/mL ovalbumin) and incubated at 37°C for 1 h with mixing every 15 min. The radioactivity bound was quantified as above.

Cross-inhibition of ¹²⁵I-C1q and ¹²⁵I-factor H binding to E. coli by unlabelled factor H and C1q, respectively

¹²⁵I-C1q (0.55 μg) or ¹²⁵I-factor H (0.80 μg) was incubated with increasing quantities of unlabelled factor H or C1q respectively for 1 h on ice in 100 μL of VB²⁺ + 0.5 mg/mL ovalbumin. The mix was added to 200 μL of E. coli (1 × 10⁹ cells/mL in VB²⁺ + 0.5 mg/mL ovalbumin) in 1.7 mL conical ultracentrifuge tubes (blocked overnight with 500 μL of VB²⁺ + 1 mg/mL ovalbumin) and incubated at 37°C for 1 h with mixing every 15 min. The quantity of ¹²⁵I-C1q or ¹²⁵I-factor H bound was quantified as above.

Assessment of C4b deposition, on lipid A coated-wells, from serum containing different C1q to factor H ratios

Serum depleted of C1q and factor H was diluted with 1 volume of VB²⁺ and repleted at various molar ratios of factor H to a fixed amount of C1q by adding back the purified proteins in VB²⁺. The various serum samples were incubated on ice for 1 h before incubation on lipid A coated wells (blocked for 2 h with VB²⁺ + 1 mg/mL ovalbumin) for 1 h at 37°C. The wells were then washed thrice with VB²⁺ before adding to the wells 100 μL of 1 in 500 dilution of affinity purified chicken polyclonal anti-human C4-alkaline phosphatase conjugate (Immune Systems Limited, Devon, UK) in VB²⁺ and incubating at room temperature for 1 h. Following further washes with 200 μL of VB²⁺, color was developed at 37°C with 100 μL buffered p-nitrophenyl phosphate (pNPP) substrate (Sigma, N2770). The absorbances were read at 405 nm. The positive control used was control serum that was not depleted of C1q or factor H but was diluted to the same extent. The negative control was the C1q and factor H depleted serum with no added C1q or factor H.

References

[1] Albertí, S., Marqués, G., Hernández-Allés, S., Rubires, X., Tomás, J.M., Vivanco, F., and Benedí, V.J. (1996). Interaction between complement subcomponent C1q and the Klebsiella pneumoniae porin OmpK36. Infect Immun 64, 4719–4725.

[2] Arvieux, J., Reboul, A., Bensa, J.C., and Colomb, M.G. (1984). Characterization of the C1q receptor on a human macrophage cell line, U937. Biochem J 218, 547–555.

[3] Bone, R.C. (1991). The pathogenesis of sepsis. Ann Intern Med 115, 457–469.

[4] Bunse, R., and Heinz, H.P. (1993). Interaction of the capsular polysaccharide of Haemophilus influenzae type B with C1q. Behring Inst Mitt 93, 148–164.

[5] Clas, F., Euteneuer, B., Stemmer, F., and Loos, M. (1989). Interaction of fluid phase C1/C1q and macrophage membrane-associated C1q with gram-negative bacteria. Behring Inst Mitt 84, 236–254.

[6] Charlesworth, J.A., Scott, D.M., Pussell, B.A., and Peters, D.K. (1979). Metabolism of human beta 1H: studies in man and experimental animals. Clin Exp Immunol 38, 397–404.

[7] Cooper, N.R., and Morrison, D.C. (1978). Binding and activation of the first component of human complement by the lipid A region of lipopolysaccharides. J Immunol 120, 1862–1868.

[8] Díaz, A., Ferreira, A., and Sim, R.B. (1997). Complement evasion by Echinococcus granulosus: sequestration of host factor H in the hydatid cyst wall. J Immunol 158, 3779–3786.

[9] Dillon, S.P., D’Souza, A., Kurien, B.T., and Scofield, R.H. (2009). Systemic lupus erythematosus and C1q: A quantitative ELISA for determining C1q levels in serum. Biotechnol J 4, 1210–1214.

[10] Dodds, A.W., Sim, R.B., Porter, R.R., and Kerr, M.A. (1978). Activation of the first component of human complement (C1) by antibody-antigen aggregates. Biochem J 175, 383–390.

[11] Edey, M., Strain, L., Ward, R., Ahmed, S., Thomas, T., and Goodship. T.H. (2009). Is complement factor H a susceptibility factor for IgA nephropathy? Mol Immunol 46, 1405–1408.

[12] Eisenschenk, F.C., Houle, J.J., and Hoffmann, E.M. (1999). Mechanism of serum resistance among Brucella abortus isolates. Vet Microbiol 68, 235–244.

[13] Fearon, D.T. (1978). Regulation by membrane sialic acid of 1H-dependent decay-dissociation of amplification C3 convertase of the alternative complement pathway. Proc Natl Acad Sci U S A 75, 1971–1975.

[14] Gaboriaud, C., Juanhuix, J., Gruez, A., Lacroix, M., Darnault, C., Pignol, D., Verger, D., Fontecilla-Camps, J.C., and Arlaud, G.J. (2003). The crystal structure of the globular head of complement protein C1q provides a basis for its versatile recognition properties. J Biol Chem 278, 46974–46982.

[15] Gulati, S., Cox, A., Lewis, L.A., Michael, F.S., Li, J., Boden, R., Ram, S., and Rice, P.A. (2005). Enhanced factor H binding to sialylated Gonococci is restricted to the sialylated lacto-N-neotetraose lipooligosaccharide species: implications for serum resistance and evidence for a bifunctional lipooligosaccharide sialyltransferase in Gonococci. Infect Immun 73, 7390–7397.

[16] de Haas, C.J., van Leeuwen, E.M., van Bommel, T., Verhoef, J., van Kessel, K.P., and van Strijp, J.A. (2000). Serum amyloid P component bound to gram-negative bacteria prevents lipopolysaccharide-mediated classical pathway complement activation. Infect Immun 68, 1753–1759.

[17] Ingram, G., Hakobyan, S., Hirst, C.L., Harris, C.L., Pickersgill, T.P., Cossburn, M.D., Loveless, S., Robertson, N.P., and Morgan, B.P. (2010). Complement regulator factor H as a serum biomarker of multiple sclerosis disease state. Brain 133, 1602–1611.

[18] Jack, D.L., and Turner, M.W. (2003). Anti-microbial activities of mannose-binding lectin. Biochem Soc Trans 31, 753–757.

[19] Joiner, K.A., Schmetz, M.A., Sanders, M.E., Murray, T.G., Hammer, C.H., Dourmashkin, R., and Frank, M.M. (1985). Multimeric complement component C9 is necessary for killing of Escherichia coli J5 by terminal attack complex C5b-9. Proc Natl Acad Sci U S A 82, 4808–4812.

[20] Joiner, K.A., Grossman, N., Schmetz, M., and Leive, L. (1986). C3 binds preferentially to long-chain lipopolysaccharide during alternative pathway activation by Salmonella montevideo. J Immunol 136, 710–715.

[21] Kishore, U., Gupta, S.K., Perdikoulis, M.V., Kojouharova, M.S., Urban, B.C., and Reid, K.B.M. (2003). Modular organization of the carboxyl-terminal, globular head region of human C1q A, B, and C chains. J Immunol 171, 812–820.

[22] Kishore, U., Ghai, R., Greenhough, T.J., Shrive, A.K., Bonifati, D.M., Gadjeva, M.G., Waters, P., Kojouharova, M.S., Chakraborty, T., and Agrawal, A. (2004a). Structural and functional anatomy of the globular domain of complement protein C1q. Immunol Lett 95, 113–128.

[23] Kishore, U., Gaboriaud, C., Waters, P., Shrive, A.K., Greenhough, T.J., Reid, K.B.M., Sim, R.B., and Arlaud, G.J. (2004b). C1q and tumor necrosis factor superfamily: modularity and versatility. Trends Immunol 25, 551–561.

[24] Kojouharova, M.S., Gadjeva, M.G., Tsacheva, I.G., Zlatarova, A., Roumenina, L.T., Tchorbadjieva, M.I., Atanasov, B.P., Waters, P., Urban, B.C., Sim, R.B., et al. (2004). Mutational analyses of the recombinant globular regions of human C1q A, B, and C chains suggest an essential role for arginine and histidine residues in the C1q-IgG interaction. J Immunol 172, 4351–4358.

[25] Kraiczy, P., and Würzner, R. (2006). Complement escape of human pathogenic bacteria by acquisition of complement regulators. Mol Immunol 43, 31–44.

[26] Kubens, B.S., Nikolai, S., and Opferkuch, W. (1989). A third mechanism of serum resistance in Escherichia coli. Zentralbl Bakteriol 271, 222–230.

[27] Laemmli, U.K. (1970). Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680–685.

[28] Latsch, M., Stemmer, F., and Loos, M. (1992). Purification and characterization of LPS-free porins isolated from Salmonella minnesota. FEMS Microbiol Lett 69, 275–281.

[29] Loos, M., and Clas, F. (1987). Antibody-independent killing of gram-negative bacteria via the classical pathway of complement. Immunol Lett 14, 203–208.

[30] Maruvada, R., Blom, A.M., and Prasadarao, N.V. (2008). Effects of complement regulators bound to Escherichia coli K1 and Group B Streptococcus on the interaction with host cells. Immunology 124, 265–276.

[31] McAleer, M.A., and Sim, R.B. (1993). The complement system. In: Activators and Inhibitors of Complement. Sim, R.B. ed. Dordrecht, Netherlands: Kluwer.1–15.

[32] Meri, S., and Pangburn, M.K. (1990). Discrimination between activators and nonactivators of the alternative pathway of complement: regulation via a sialic acid/polyanion binding site on factor H. Proc Natl Acad Sci U S A 87, 3982–3986.

[33] Merino, S., Vilches, S., Canals, R., Ramirez, S., and Tomás, J.M. (2005). A C1q-binding 40 kDa porin from Aeromonas salmonicida: cloning, sequencing, role in serum susceptibility and fish immunoprotection. Microb Pathog 38, 227–237.

[34] Morrison, D.C., and Kline, L.F. (1977). Activation of the classical and properdin pathways of complement by bacterial lipopolysaccharides (LPS). J Immunol 118, 362–368.

[35] New, R.C.C. (1990). Preparation of liposomes, In: Liposomes - a practical approach. New, R. C. C. (ed.). Oxford, U.K. IRl Press, 3–104.

[36] Reid, K.B., and Porter, R.R. (1976). Subunit composition and structure of subcomponent C1q of the first component of human complement. Biochem J 155, 19–23.

[37] Rietschel, E.T., Kirikae, T., Schade, F.U., Mamat, U., Schmidt, G., Loppnow, H., Ulmer, A.J., Zähringer, U., Seydel, U., Di Padova, F., et al. (1994). Bacterial endotoxin: molecular relationships of structure to activity and function. FASEB J 8, 217–225.

[38] Ripoche, J., Day, A.J., Harris, T.J., and Sim, R.B. (1988). The complete amino acid sequence of human complement factor H. Biochem J 249, 593–602.

[39] Roumenina, L.T., Rouseva, M., Zlatarova, A., Kolev, M., Ghai, R., Olova, N., Gadjeva, M., Agrawal, A., Mantovani, A., Bottazzi, B., et al. (2006). Interaction of C1q with IgG1, C-reactive protein and pentraxin 3: mutational analyses using recombinant globular regions of C1q A, B and C chains. Biochemistry 45, 4093–4104.

[40] Schneider, M.C., Exley, R.M., Chan, H., Feavers, I., Kang, Y.H., Sim, R.B., and Tang, C.M. (2006). Functional significance of factor H binding to Neisseria meningitidis. J Immunol 176, 7566–7575.

[41] Shang, S.Q., Chen, G.X., Shen, J., Yu, X.H., and Wang, K.Y. (2005). The binding of MBL to common bacteria in infectious diseases of children. J Zhejiang Univ Sci B 6, 53–56.

[42] Sim, R.B., and Malhotra, R. (1994). Interactions of carbohydrates and lectins with complement. Biochem Soc Trans 22, 106–111.

[43] Sim, E., and Sim, R.B. (1983). Enzymic assay of C3b receptor on intact cells and solubilized cells. Biochem J 210, 567–576.

[44] Sim, R.B., Day, A.J., Moffatt, B.E., and Fontaine, M. (1993). Complement factor I and cofactors in control of complement system convertase enzymes. Methods Enzymol 223, 13–35.

[45] Stewart, J.C. (1980). Colorimetric determination of phospholipids with ammonium ferrothiocyanate. Anal Biochem 104, 10–14.

[46] Tan, L.A., Yu, B., Sim, F.C., Kishore, U., and Sim, R.B. (2010). Complement activation by phospholipids: the interplay of factor H and C1q. Protein Cell 1, 1033–1049.

[47] van den Berg, R.H., Faber-Krol, M.C., van de Klundert, J.A., van Es, L.A., and Daha, M.R. (1996). Inhibition of the hemolytic activity of the first component of complement C1 by an Escherichia coli C1q binding protein. J Immunol 156, 4466–4473.

[48] Weiler, J.M., Daha, M.R., Austen, K.F., and Fearon, D.T. (1976). Control of the amplification convertase of complement by the plasma protein β1H. Proc Natl Acad Sci U S A 73, 3268–3272.

[49] Whaley, K., and Ruddy, S. (1976). Modulation of the alternative complement pathways by β1H globulin. J Exp Med 144, 1147–1163.

[50] Ying, S.C., Gewurz, A.T., Jiang, H., and Gewurz, H. (1993). Human serum amyloid P component oligomers bind and activate the classical complement pathway via residues 14-26 and 76-92 of the A chain collagen-like region of C1q. J Immunol 150, 169–176.

[51] Zipfel, P.F., Würzner, R., and Skerka, C. (2007). Complement evasion of pathogens: common strategies are shared by diverse organisms. Mol Immunol 44, 3850–3857.

[52] Zohair, A., Chesne, S., Wade, R.H., and Colomb, M.G. (1989). Interaction between complement subcomponent C1q and bacterial lipopolysaccharides. Biochem J 257, 865–873.

RIGHTS & PERMISSIONS

Higher Education Press and Springer-Verlag Berlin Heidelberg 2011