Glycosaminoglycans: Anticoagulant and Nonanticoagulant Actions

This issue of Seminars in Thrombosis and Hemostasis has the anticoagulant and nonanticoagulant actions of glycosaminoglycans (GAGs) as its topic. The most widely known GAGs are heparins (unfractionated [UFH] and low-molecular-weight [LMWH]) and they are best known for their anticoagulant properties. Although the clinical usefulness of heparins cannot be disputed, there are serious side effects associated with their use, more with UFH than with LMWH. These problems have spurred continuous efforts to develop modifications of UFH in order to make compounds safer without compromising the desired anticoagulant activities. Some of these developments are described in this issue.

Glycosaminoglycans interact with several plasma proteins to form complexes. One such protein is antithrombin (AT). AT in complex undergoes a profound configurational change in its structure that results in a 1000-fold greater activity to neutralize serine proteases, especially thrombin and factor Xa.

The first article by Hricovini and colleagues examines which changes GAGs have to undergo in order to permit binding to proteins; AT and fibroblast growth factors 1 and 2 are used as examples. Iduronic acid residues appear to play a major role in facilitating the docking of the heparin oligosaccharides to AT. The same was found for the binding of the two fibroblast growth factors. It is hoped that a better understanding of these conformational changes may lead to the design of better glycomimetics.

Casu and coworkers describe their experience with chemically modifying GAGs in order to obtain more active structures or structures with different biological activities. There are several ways in which sulfated GAGs can be modified. Desulfation of heparin, heparan sulfate, chondroitin sulfate, and dermatan sulfate results in different binding properties and different biological activities, not only anticoagulant but also antimetastatic properties. There is a good correlation between activities of these compounds and typical sulfation patterns along the backbone of the GAGs.

In the next contribution Harenberg et al describe their work on labeling certain GAGs with tyramine or fluorescein in order to better understand their interactions with coagulation enzymes and with leukocytes. Because direct fluorescein binding to GAGs alters their properties, the authors labeled them first with tyramine, to which they then attached fluorescein. This did not result in structural alterations. In vitro and in vivo, the endpoint-labeled compounds did not differ in their anticoagulant properties from their unlabeled precursors. The labeled products also bound to lymphocytes, monocytes, and granulocytes. Greatest fluorescent intensity was found on granulocytes. Neutralization experiments resulted in a greater amount of protamine needed to inactivate the labeled compounds. Displacement of the labeled compounds by other GAGs revealed a dependence on the degree of sulfation.

Sissi and coworkers studied the effect of differently sulfated heparins on the interference with their antifactor IIa and antifactor Xa activities by calcium ions at physiological concentrations. While the dissociation constants for heparin-antithrombin were not affected, the catalytic activity was strongly dependent on calcium ions. The catalysis increased with an intermediate degree of sulfation but was markedly decreased with supersulfation. Data suggest that the binding of antithrombin to heparin is the main determinant of the inhibition of factor Xa. The data obtained should be helpful in assessing the proper structure-function relationships of GAGs.

Next, Fenyvesi and his colleagues examined laboratory methods to monitor the anticoagulant effect of direct thrombin inhibitors, most notably lepirudin, argatroban, and melagatran. It is well recognized that the widely used activated partial thromboplastin time (aPPT) is not helpful in monitoring these compounds. Plasma samples of healthy donors were spiked in vitro with these direct thrombin inhibitors and assayed with the aPPT, prothrombin time (PT), ecarin clotting time (ECT), and a prothrombinase-induced clotting time (PiCT). The aPPT revealed nonlinearity at higher dose ranges of the three compounds and PT was poorly responsive to lepirudin and of limited value for the other direct thrombin inhibitors. The ECT was found to be useful, but it does not respond to heparins. The PiCT was found to be quite suitable for measuring the activity of all of the compounds. In addition it allows assessment of heparin activity. The authors recommend both the ECT and the PiCT for monitoring direct thrombin inhibitors.

Alban and associates report on a study conducted with a medium-molecular-weight heparin (MMWH). The major differences between UFH and LMWH reside, besides safety, in their ability to neutralize key enzymes of the clotting system. UFH inactivates both factor Xa and thrombin but has low bioavailability. LMWH inactivates predominantly factor Xa but not thrombin and has high bioavailability. The MMWH studied has both antifactor Xa and antithrombin properties but also high bioavailability. These features apparently allow a lower dosing regimen. This new compound could be of potential clinical usefulness once additional studies are completed.

Harenberg and coworkers next report on studies conducted in vitro in which a variety of anticoagulants were added to freshly drawn blood. After clotting they measured thrombin-antithrombin complexes (TAT) as an indicator of activation of the clotting system and platelet factor 4, a platelet release protein, as an indicator of platelet activation. The addition of recombinant hirudin, pegylated hirudin, and melagatran dose dependently reduced TAT levels and platelet factor 4, indicating that these direct thrombin inhibitors block not only clotting but also platelet activation. In comparison, similar results were obtained when unfractionated heparin and several LMW heparins were added.

In the next contribution Poggi and associates describe the effects of semisynthetic sulfaminoheparosan sulfates (SAHSs) derived from E. coli polysaccharides on the growth of B16-BL6 melanoma lung colonies. The data were compared with those obtained by using UFH or heparan sulfate. It is known that heparins and other GAGs inhibit cancer cell invasion by interfering with several steps of this process. The SAHSs were modified with respect to degree and distribution of sulfates. Several compounds were found to inhibit the melanoma lung colony formation. These actions appeared to be independent of the anticoagulant properties of these derivatives.

In the last article Petitou and coworkers describe the synthetic pentasaccharide fondaparinux, the first antithrombotic agent that specifically inhibits factor Xa. The new compound binds to antithrombin and then inactivates factor Xa. In animal models fondaparinux was as effective as heparin and LMW heparins but was associated with a lower incidence of bleeding. The main route of elimination is the kidneys. No interactions were noted with vitamin K antagonists or with aspirin. The compound does not cross the placenta. In clinical settings fondaparinux was more effective in preventing venous thromboembolism in patients undergoing orthopedic surgeries than LMW heparins, and promising results were found in its use for treating other thrombotic disorders. Fondaparinux is potentially improving the management of thromboembolism over present treatment modalities.

Great thanks and appreciation are expressed to all authors for their valuable contributions. The work by Professors Harenberg and Casu in assembling this valuable issue is gratefully acknowledged.