Abstract
Glycosaminoglycans (GAGs) are negatively charged carbohydrate of linear polymer. They are composed of irregular repeat disaccharide units with various sulfation patterns. They are widely distributed at different tissues and cells and are attached on cell membranes through various protein cores in the form of proteoglycans. Heparin or heparan sulfate are one of glycosaminoglycans with unique sulfation at 2-hydroxy position of glucosamine. Their heterogeneity in chemical structures and locations make them extremely important as multifunctional regulator of many GAG-binding proteins. They are responsive many biological processes, such as cell differentiation / proliferation, tumor metastasis and viral infection, through structural specific mechanism. Cardiotoxins (CTXs) are major components of cobra snake venoms, which account for approximate 50% in weight. They are highly basic □-sheet polypeptide resembling three finger loops. They exhibit strong tendency to cause cytotoxicity, necrosis at local tissue and systolic heart arrest. Pathological observation is the bitten victim of cobra suffered from severe inflammation in local tissue, a phenomena known to be mediated by the release of histamine, increasing capillary permeability, infiltration of immune cells and induction of many GAG-binding chemokines. Strong retention of CTXs, but not other venom proteins, at local tissues or rapid clearance from plasma implies unknown cell surface components might be responsive for CTXs retention. In this work, the interactions between heparin and CTXs are studies by many biophysical, biochemical and cell methods. Of all the binding assays, surface plasmon resonance (SPR) gives sensitive and accurate results in both kinetic and thermodynamic point of views. Accompanied with various CTX homologues and various heparin derivatives in different chemical structure, it is shown that the interactions between heparins and CTXs are structural specific. The retention capability of CTX is demonstrated to be heparan sulfate-dependent by retention test on immobilized CHO cells. Such retention capability is consistent with the results at immobilized heparin surface by SPR. Upon comparison of all studied CTX homologues, it is suggested that the structural important loop2 region determines the heparin binding affinity, binding mode and binding specificity. The N-sulfate on heparin is exclusively important and is correlated with the presence of Lys31 as compared by various CTX homologues. The possible role of loop2 region in heparin binding mode might associate with the connectivity of hydrophobic patch by three tips of finger loops. The CTXA3-heparin complex is further stabilized and gain retention capability by the presence of citrate ion, which is also major constituent (~50mM) in Taiwan cobra venoms. Both SPR studies and cell retention test support that the citrate-mediated retention of CTXA3 is heparin-dependent. The co-crystal structure of CTXA3 with heparin hexasaccharide shows dimeric packing of CTXs with one citrate ion at putative charged pocket composed of Lys23 and Lys31. This crystal structure explains the retention of CTXs on heparin through CTX dimerization, and the role of charged/non-charged residue at tip of loop2 in different heparin binding modes. The structural determinants of CTXA3-heparin interaction are further studied to confirm their binding specificities. Heparin length in hexasaccharide is the least size for CTXA3 binding, while the length up to dp12-14 is functional important as investigated by its inhibition ability on CTX cytotoxicity. The thermodynamic investigation of CTXA3 binding to high molecular weight heparin shows extremely entropy-driven interaction, which implies the strong tendency to protein aggregation. Although SPR studies shows no significant retention on heparin surface, the chemical cross-linking and fluorescent self-quenching experiments show potential oligomerization of CTXA3 upon heparin binding. Due to the heterogeneity of heparin, the kinetic of CTXA3-heparin interaction also reveals heterogeneous binding. This heterogeneous binding is reflected by biphasic / sequential dissociation rates with CTX-occupancy dependent manner. The biphasic dissociation of CTXA3 from heparin surface is observed over 3~4% CTX-occupancy, which implies heterogeneous domains with different binding modes on heparin are responsive for this behavior. Enzymatic degradations under affinity-protection with different CTX/heparin ratios generate various CTX-protected domains. Heparin disaccharide compositions of CTXA3-protected domains at different occupancies show different specificities toward N-acetylate at low occupancy but N-sulfate at high occupancy. They indicated diverse heparin binding sequences for CTXA3 binding. The affinity-protected domain at low CTX-occupancy shows the size in dp12-14 and low charge/mass characteristics. The removal of N-acetylate eliminate biphasic dissociation with reduced binding affinity / cooperativity and reduced ANS-binding fluorescence, which implies the possible role of hydrophobic interaction in N-acetylate groups to promote CTXA3 dimerization locally. In summary, the structural factors of CTX-heparin interaction are determined both on CTXs and heparin. The heterogeneity of heparin and regulated heparin binding mode might determine different tissue specificity for different CTX homologues and their biological significance.