Heparinplatelet Interactions In The Pathogenesis Of

Nonimmune heparin-platelet interactions are central to the pathogenesis of HIT because of the key role of platelet factor 4 (PF4). This cationic chemokine is secreted from activated platelets and binds to GAGs on the surface of platelets and endothelium (Dawes et al., 1982; Rao et al., 1983; O'Brien et al., 1985; Capitanio et al., 1985; Cines et al., 1987; Visentin et al., 1994). PF4 also binds to soluble GAGs, especially highly anionic heparin, leading to a competition between cell-bound and soluble GAGs for PF4 (Horne, 1993; Newman et al., 1998).

When complexed with GAG, PF4 exposes one or more neoantigens that stimulate formation of HIT antibodies (Amiral et al., 1992, 1995; Kelton et al., 1994; Newman and Chong, 1999). Recent evidence suggests that the neoantigen is formed by close approximation of two PF4 tetramers, which can happen when the positive charge of the PF4s is neutralized by GAGs (Greinacher et al., 2006). However, to be immunogenic, the PF4-GAG complexes presumably must be soluble and thereby accessible to the immune system. Perhaps this explains why PF4-GAG that is constitutively present on the endothelial surface is not immuno-genic, but soluble PF4-heparin complexes are.

Once stimulated by exposure to PF4-heparin, HIT antibodies can bind to PF4 complexed with other GAGs (e.g., heparan sulfate and chondroitin sulfate) on cell membranes. By this mechanism, they could stimulate platelets (Rauova et al., 2006) and also (directly or indirectly) endothelial expression of tissue factor (Cines et al., 1987; Herbert et al., 1998). Such heparin-independent binding of HIT antibodies to platelets and endothelium may explain the appearance or persistence of thrombocytopenia in HIT after heparin exposure has ceased (see Chapter 2). Activation of platelets in the absence of heparin, however, appears to require extensive saturation of the platelet surface with PF4, since antibody binding in vitro is observed only with PF4 concentrations >300 nM, whereas the Kd for the binding of PF4 to platelets is reported to be about 30 nM (Loscalzo et al., 1985; Rauova et al., 2006). Such concentrations would be rarely, if ever, achieved in vivo. On the other hand, PF4 binds to endothelium even at normal plasma concentrations less than 1 nM and is readily displaced by heparin. Therefore, stimulation of endothelium by HIT antibodies seems a more probable mechanism for appearance or persistence of HIT after heparin has been discontinued.

When heparin is present, it forms soluble complexes with PF4 that it displaced from endothelium or that was secreted by activated platelets. These complexes also bind HIT antibodies, and they have the potential for docking at the platelet surface, attaching via their heparin at cationic sites rather than at the GAG (chondroitin sulfate) naturally found in the platelet cell membranes (Greinacher et al., 1993; Horne and Hutchison, 1998).

The ability of HIT-immune complexes to stimulate platelets appears to depend on the size of the PF4-heparin component (Rauova et al., 2005). The largest PF4-heparin complexes have the greatest chance of binding to platelets, and they can also carry several HIT-IgG molecules (Rauova et al., 2005). Therefore, when such a complex attaches to a platelet, it brings an especially rich trove of HIT-IgG to activate the cell through its Fcy receptors (Horne and Alkins, 1996).

The size of PF4-heparin complexes depends upon the length of the heparin chains and upon the molar ratio of heparin to PF4. A heparin molecule of Mr ~11,000 can bind about four PF4 molecules, only partially saturating each one (Loscalzo et al., 1985). Heparin molecules about half this size (Mr 5000-7000) can crosslink two or three PF4s (Bock et al., 1980). Therefore, LMWH (Mr 300010,000 Da) does not form ultralarge complexes as readily as unfractionated heparin (UFH, Mr 4000-30,000 Da), and the pentasaccharide fondaparinux (Mr 1728) does not form them at all (Rauova et al., 2005).

When UFH and PF4 are present in roughly equal molar amounts, large lattices (>670,000 Da) of heparin and PF4 can form (Bock et al., 1980; Rauova et al., 2005) (Fig. 2, middle panel). As the molar concentration of PF4 exceeds that of heparin, the size of the complexes becomes smaller and smaller until each contains only a single heparin molecule saturated with PF4 (Fig. 2, upper panel). A consequence of this is that these complexes cannot attach to platelets because the PF4 sites that might bind to platelet chondroitin sulfate are blocked by heparin, while there is no heparin free to bind to platelet cationic sites (Horne and Hutchison, 1998). If the molar ratio shifts the other way, so that heparin is in excess, the complexes will become limited to one heparin and one PF4 molecule each (Fig. 2, bottom panel). In this situation, the complexes are unlikely to bind to platelets because of competition with free heparin molecules and because the negative charge of heparin, which affects its affinity for the platelet, is partially neutralized by binding to PF4.

The critical importance of the molar ratio of heparin to PF4 probably explains why most patients who develop HIT antibodies never develop the clinical syndrome: molar ratios of PF4 and heparin favorable for platelet binding are rare or transient in most clinical situations (Amiral et al., 1996; Visentin et al., 1996; Kappers-Klunne et al., 1997; Warkentin et al., 2000; Rauova et al., 2006). While the plasma concentration of heparin is dose-dependent, the plasma concentration of PF4 depends on the level of platelet activation, which is affected by the degree of stimulation by HIT antibodies and on displacement of PF4 from the endothelial and platelet surfaces by heparin (O'Brien et al., 1985; Horne, 1993). Furthermore,

Binds to Platelets?

PF4 > Heparin

0 0

Post a comment