Based on this structure, surface charge pattern, heparin affinity measurements, and docking of a heparin disaccharide, a heparin binding site is proposed in the contact area of the serpinproteinase encounter complex [25]. Beinrohr et al proposed that by binding to C1inh and neutralizing its positively charged surface patches the polyanions facilitate the C1inh-C1s interaction (a “sandwich” mechanism) [25]. This can explain how the inhibitory activity of C1 inhibitor toward proteases, such as C1s or activated factor XI, can be greatly enhanced by heparin and other glycosaminoglycans [18]. Our data support this model for the enhancement of C1inh C1s interaction by GAGs that we observed in the current study, as well as in our earlier work [26]. Heparin lots contaminated with OSCS can inhibit complement activity in vitro. However the sustained levels of 10?0 microgram/ mL are unlikely with intravenous dosing of heparin although subcutaneous administration [27] of contaminated heparin may have allowed for higher local levels of OSCS. A veterinary drug, polysulfated glycosaminoglycan (PSGAG) that is very similar in structure to OSCS is still used in animals and administered locally (e.g. intramuscularly). PSGAG is also a polysulfated chondroitin sulfate with 3 to 4 sulfate groups per disaccharide unit and is considered to be a disease-modifying veterinary drug for osteoarthritis. PSGAG is anti-inflammatory and many mechanisms have been postulated from preservation of joint glycosaminoglycans to inhibition of PGE2 synthesis, toxic oxygen radical generation, and complement activation. Studies have shown an impact of PSGAG at relatively higher doses on complementmediated lysis of red blood cells [28,29] without a clear mechanism of action. Our in vitro experiments using bacteria as model indicate PSGAG is a very strong inhibitor of complement fixation of bacteria. The potentiation of C1inh interaction with C1s by OSCS can also explain the effect of PSGAG on complement lysis and provide a mechanism for studies suggesting an increased likelihood of infections with intra-articular injection of PSGAG and low levels of bacteria [30,31]. Although there was an increase in the absolute numbers of infections reported during the 2007?008 timeframe of the OSCS contamination, the relative numbers decreased [24] . It may be of value to further assess GAG related products including PSGAG, for infection related adverse events, although adverse event reporting has many limitations. Based on the concentrations heeded to inhibit complement, there may not be an in vivo effect unless high doses are administered locally (e.g. subcutaneous administration) rather than systemically. There have been suggestions that glycosaminoglycans can be used to inhibit the complement activity in situations such as
autoimmune diseases [32]. PSGAG treatment of animal arthritis is an example of such a use for a GAG. As PSGAG is not administered intravenously, a kallikrein mediated adverse effect, such as seen with OSCS contaminated heparin, may be less likely. A human version of such a product was marketed in Europe and withdrawn. Heparin has recently been shown to prevent fetal loss in a model of anti-phospholipid syndrome by inhibiting complement activation [33]. A more potent inhibition of complement, such as seen with OSCS, may be useful. Although OSCS complement inhibition was demonstrated with the classical complement pathway, we also observed OSCS inhibition of Factor B after treatment with complement serum (data not shown). This indicates OSCS may also modulate the alternative pathway. The potential interactions between OSCS and alternative pathway factors (e.g., Factor B, Factor H and properdin) need further investigation. Since OSCS activates the contact system in humans as well as inhibiting complement, it is unlikely to be used in the future for the purpose of complement inhibition. However, it is unclear whether the same structural attributes are responsible for both effects. Development of a GAG which separates the anti-complement activity from the pro-kallikrein activity of OSCS could be of value in treatment of inflammatory disease. In conclusion, OSCS can inhibit the complement classical pathway by potentiating the binding of C1inh with C1s. This potentiation is much stronger with OSCS than heparin. A veterinary drug, PSGAG, has similar effects to OSCS on bacterial lysis by complement. C1inh potentiation may explain the antiinflammatory properties of PSGAG as well as experimental studies showing an increased likelihood of infections with intra-articular injection of PSGAG and low levels of bacteria.
Methods Materials
OSCS-contaminated and un-contaminated heparin lots were obtained by the FDA from Baxter Healthcare (1000 U/ml or 5000 U/ml in 10 ml and 30 ml vials) [34]. Synthetic OSCS and commercial veterinarian drug polysulfated glycosaminoglycan (PSGAG) were obtained from the Division of Pharmaceutical Analysis, FDA at St. Louis. Chondroitin sulfate A (CSA) was purchased from Sigma (St. Louis, MO). Purified human complement C1 esterase inhibitor (C1inh) was purchased from Sigma and EMD Chemicals USA (Gibbstown, NJ). Normal pooled human plasma was purchased from Innovative Research (Novi, Michigan). Purified human Factor XIIa and Factor XII deficient plasma supplied by Hyphen BioMed (France) was purchased through Innovative Research. Pooled complement-preserved plasma from horse, donkey, pig and beagle were purchased from Bioreclamation Inc (Westbury NY). Activated C1s was obtained from EMD chemicals USA (Gibbstown, NJ). Anti-complement component C1 inhibitor mouse monoclonal antibody was obtained from the AntibodyShop (Demark) and purified peroxidase-conjugated goat anti-human C1inh IgG was purchased from Cedarlane laboratories (Canada). Protein A/G-Sepharose was purchased from BioVision (Mountain View, CA). PE anti-human complement C3 monoclonal antibody was purchased from Cedarlane laboratories. Aamine-PEO3-biotin and NeutraAvidin were obtained from Pierce (Rockford, IL); Biotinylated GAGs were prepared as described by Li et al [4] .CM5 sensor chips, Streptavidin (SA) sensor chips, 1-ethyl-3-(3-dimethylaminopropyl)-carbodiimide hydrochloride (EDC), N-hydroxysuccinimide (NHS), ethanolamine?HCl, HBS�P buffer (0.01 M HEPES, pH 7.4, 0.15 M NaCl, 0.005% surfactant P20), acetate buffer (pH 5.0), Glycine(pH 2.5), NaOH (50 mM), NaCl (4 M) Mg2Cl2 (5 M) and deionized water were from GE Healthcare (Piscataway, NJ).