OSCS may have direct or indirect effects on complement. OSCS inducedFXII-dependent generation of C3a and C5a in the plasma with the resultant anaphylactoid and chemotactic functions. The generation of C3a and C5a would also consume the complement components C3 and C5, and the depletion of these components may impact complement activation pathways. A more direct interaction of OSCS with complement components has been demonstrated using surface plasmon resonance [4] and this binding may impact complement activation. Another indirect
effect of OSCS could be mediated through an interaction with regulators of the complement system. For example, previous studies have shown complement activation can be regulated by heparin and related GAGs through the complement regulator, C1inh. Heparin has been shown to potentiate the inhibition of C1s by C1 inhibitor by 15- to 35-fold, leading to decreased formation of C3 convertase in assays performed either with purified complement proteins or in whole serum [15]. However, a study using surface plasmon resonance did not reveal a difference in heparin and OSCS binding to complement components [4]. It is important to further investigate the comparative effects of heparin and OSCS on the complement pathways. In the present study we investigated the interaction of OSCS with the complement classical pathway using a biologically relevant functional model as well as surface plasmon resonance. Although OSCS-contaminated heparin is unlikely to appear in the future due to current regulatory expectations, a related product, polysulfated glycosaminoglycan (PSGAG), is an approved veterinary medicine. The effect of PSGAG on complement activity in animal plasma was also investigated in this report.
Pre-treatment of Complement Plasma with OSCS Dramatically Enhanced Inhibition of Complement Fixation on Bacteria
Pre-incubation of the complement plasma with OSCS at 37uC for 5 minutes significantly decreased C3 fixation on bacteria in comparison to later addition of OSCS (Figure 2B). This indicates that OSCS interacted with some factor or factors in the plasma that inhibited the activity of complement. The decrease in the percentage of C3 fixation with a fixed dose (20 mg/ml) of OSCS was more notable at greater dilutions of complement (Figure 2B). The larger impact of a fixed dose of OSCS on diluted complement plasma may indicate that the ratio of complement components or other plasma factors to OSCS is critical for the inhibition effect.Consumption of C3 is not an Explanation of the Complement Inhibition by OSCS
The activation of FXII into FXIIa by OSCS generates kallikrein and further activates C3 and C5 into C3a and C5a [1,4]. Thus, one explanation of the OSCS impact on C3 fixation and complement lysis would be the consumption of available C3 by OSCS. To test if this assumption is true, FXII-deficient plasma was incubated with OSCS. As expected, there was no C3a generation and therefore no C3 consumption (data not shown). However, the inhibition of C3 fixation by OSCS was still present as shown in Figure 3. This result indicates that FXII activation by OSCS and a resultant consumption of C3 is not the explanation for complement inhibition by OSCS.
Results Oversulfated Chondroitin Sulfate Inhibits the Complement Classical Pathway
Natural polyreactive antibody clone 2E4 can recognize E. Coli BL21 bacteria and induce bactericidal lysis through the complement classical pathway [16]. To test if OSCS impacts the complement classical pathway, OSCS or CSA were added to this model of the complement classical pathway activation system. Complement fixation on the bacteria, as determined by C3 deposition, was dramatically inhibited by OSCS but not by CSA (Figure 1A). Quantitative bacterial killing was determined by Live/Dead staining, and as shown in Figure 1B, 25.5% of bacteria were dead after addition of antibody and complement and 26.5% of bacteria were dead if the complement was treated with CSA. However, the percentage of dead bacteria was reduced to a baseline level of ,8% if the complement was treated with OSCS. Bacterial lysis was also determined by 3H-TdR release in addition to Live/Dead staining. OSCS inhibited the bactericidal activity mediated by natural antibody clone 2E4 through the complement classical pathway (Figure 1C). Flow cytometry indicated that bacterial morphology changed after incubation with antibody 2E4 and complement. FSC (proportional to cell size) decreased while SSC (proportional to cell granularity) increased, indicating cell death. These morphological changes occurred in the CSA-treated sample as well as in the complement-treated positive control but were not observed in the OSCS-treated sample (Figure 1D).
OSCS Inhibition of Complement Fixation on Bacteria is C1 Inhibitor-dependent
As described above, FXII activation and C3 consumption could not explain the OSCS inhibition of complement. In Figure 4, OSCS inhibits C3 fixation on bacteria with both normal (C1inh+/ FXII+) and FXII-deficient (C1inh+/FXII2) complement plasma. However, when C1inh was depleted from either plasma, the OSCS inhibition disappeared (Figure 4). The control glycosaminoglycan CSA does not inhibit C3 fixation in any plasma tested. This result indicates that the inhibition of complement fixation by OSCS is dependent on C1inh.OSCS Enhances the Binding of C1 Inhibitor with C1s
We have demonstrated C1inh is important in the OSCS mediated inhibition of complement. As the C1inh is the only inhibitor of C1s, we evaluated the interactions of both C1inh and C1s with OSCS and two other glycosaminoglycans, heparin and CSA. In order to evaluate these interactions using surface plasmon resonance, the GAGs were biotinylated and immobilized on NeutrAvidin-immobilized CM5 sensor chips to serve as surface ligands. As shown in Figure 5A, the direct interactions of C1inh with GAGs differ, with more C1inh binding OSCS than CSA or heparin (Fig. 5A). Figure 5B shows GAG interaction with C1s, with the highest binding (RU) for OSCS, followed by heparin and then CSA. The protein binding to GAGs may be due to a charge difference and may not impact the association of C1inh with C1s. To directly evaluate the impact of GAGs in the binding of C1inh with C1s, an SA sensor chip was immobilized with biotin-labeled anti-human C1inh antibody and then C1inh was captured generating a surface with C1inh as the ligand. Then different concentrations of C1s were dissolved in a low concentration (200 nM, ,0.6 mg/ml) of CSA, heparin, or OSCS, and injected over the C1inh surface. As shown in Figure 5C, the binding of C1inh to C1s was dramatically increased in the presence of 200 nM of OSCS as compared to heparin or CSA.OSCS does not Block Antigen-antibody Binding in the Complement Classical Pathway
Antibody/antigen binding is the first step that initiates the complement classical pathway. To rule out the possibility that OSCS interferes with antibody binding to bacteria rather than blocking the activation of complement components, OSCS or CSA were added to bacterial samples together with the natural antibody 2E4. Before the addition of complement, the bacteria were washed with PBS to remove any unbound reagents. As shown in Figure 2A, the presence of OSCS during antibody/ antigen interaction step did not block complement fixation while the presence of OSCS after the addition of the complement plasma dramatically inhibited C3 fixation (Figure 2A).
