Characterization of the Structural Requirements for a Carbohydrate Based Anticoagulant with a Reduced Risk of Inducing the Immunological Type of Heparin-associated Thrombocytopenia

Abstract

SummaryHAT is the most frequent drug induced immune-thrombocytopenia. We recently identified multimolecular PF4/heparin complexes as the major antigen. In order to evaluate the structural requirements for formation of the antigenic complex, we chemically synthesized 13 glucan sulfates and used 5 heparin fractions (2.4-4.8 kD) and a synthesized pentasaccharide, representing the antithrombin III binding sequence of heparin, to further characterize the HAT antigen. In the presence of glucan sulfates and heparin, HAT antibodies caused platelet activation typically at low but not at high concentrations, as measured by 14C-5HT release. The concentration range giving the activation pattern depended on the degree of sulfation (DS) and molecular weight (MW) of the glucan sulfates but not on the type of glycosidic linkage of a polysaccharide. With linear glucan sulfates with a chain length of 35 monosaccharides, the critical DS to form the HAT antigen ranged between 0.60 and 1.20. Glycosidic branched glucan sulfates were able to form the HAT antigen at a lower DS and a lower MW than linear glucan sulfates. Platelet activation by HAT-antibodies in the presence of linear curdlan sulfate fractions was dependent on their MW. At a low concentration (0.01 µM) medium-size fractions (60 kD) caused platelet activation but neither small (12 kD) nor large fractions (>150 kD) did. At higher concentrations (2 µM) the opposite reaction pattern was observed. In the case of heparin, the optimal chain length for forming the HAT antigen is a hexadecasaccharide (4.8 kD). Antigen generation decreased with larger and smaller fractions. For 50% platelet activation by HAT antibodies increasing concentrations of heparin were necessary using heparins with decreasing MW: 0.02 ± 0.015 µM (4.8 kD), 0.09 ± 0.016 pM (3.0 kD), 0.8 ± 0.21 µM (2.4 kD). In the presence of the pentasaccharide, HAT antibodies did not cause platelet activation atany concentration tested, nor bound to PF4-pentasaccha- ride complexes in a PF4 based ELISA system. We conclude that generation of the HAT antigen is dependent on the ratio of MW and DS of a glucan sulfate. We conclude from this study that a carbohydrate based anticoagulant with a reduced risk of forming the HAT antigen should be linear with a DS<0.6 or a MW <2.4 kD. These data might be important for the design of new drugs for parenteral anticoagulation.