Nnan 18 i 60H

FIGURE 1 Biologic role of PF4 in stabilizing thrombus formation after endothelial injury. (A) Animals were subjected to FeCl3-induced carotid injury and formation of a stable thrombus was determined. The percentage of animals developing a stable occlusive thrombus (■), transient occlusive thrombus (■), and no occlusive thrombus is shown. The animals are arranged from no PF4 expression on the left to excess PF4 expression on the right. The number tested of each phenotype is indicated at the bottom (P< 0.005 mPF4+/~ vs. mPF4+/+; P< 0.0001 for mPF4_/~ vs. mPF4+/+ and hPF4+ vs. mPF4+/+). (B) Mean thrombosis score of wild-type and littermate mPF4_/~ mice treated with increasing doses of recombinant hPF4 infusion (n = 5 animals at each point). A thrombosis score of 2 = stable occlusive thrombus, 1 = transient occlusive thrombus, and 0 = no occlusive thrombus developed. The zero concentration data points are extrapolated from panel (A). Numbers indicate levels of heparin infusion in which occlusive thrombi occurred at a significantly different rate from non-heparin-treated mice of that phenotype (P < 0.01, Student t test; 1 = mPF4+/+, and 2 = mPF4_/~. Abbreviations: PF4, platelet factor 4; hPF4, human PF4; mPF4, murine PF4. Source: Reproduced from Eslin et al., 2004. (Blood, 2004 vol. 104, p. 3176 by copyright permission of the author and American Society of Hematology.)

(B) hPF4 (mg/kg)

to the level of PF4 expression and does not require exogenous heparin. Interference with cell-surface PF4/GAG assembly with high dose heparin or protamine sulfate, ameliorates the severity of the thrombocytopenia (Rauova et al., 2006). These findings suggest that the level of PF4 on the vasculature and other cell surfaces may affect antigen assembly and thereby contribute to the clinical variability seen in patients with anti-PF4-heparin antibodies. Thrombosis may be fostered in patients who release relatively large amounts of PF4 from their platelets, either due to constitutive overexpression or due to platelet-activating effects of atherosclerosis and vascular injury. Heparin may stabilize or propagate thrombosis by neutralizing the charge effects of "excess" cell-surface PF4. Assembly of cell-surface antigenic complexes capable of binding HIT antibodies in the absence of drug may contribute to the development of HIT when only low sensitizing doses of heparin are employed and to the persistent hypercoagulable state after heparin has been withdrawn.

Anticoagulant properties of PF4 have also been described. PF4 binds to TM, a 60.3 kDa protein constitutively expressed on the surface of ECs. Binding of thrombin to TM alters its substrate specificity, such that proteolytic cleavage of protein C is accelerated 20,000-fold (Esmon, 1989). TM is posttranslationally modified by association with a chondroitin sulfate A-like GAG, which invests it with the capacity to bind cationic peptides at physiological pH. The binding of eosinophilic cationic protein, major basic protein, and histidine-rich glycoprotein to these GAG residues inhibits the function of TM, whereas the binding of PF4

(but not b-thromboglobulin or thrombospondin) increases protein C-cofactor activity 25-fold in a cell-free system (Slungaard and Key, 1994; Dudek et al., 1997). Addition of PF4 to cultured ECs accelerates APC generation approximately 5- to 10-fold depending on vascular origin (Slungaard et al., 2003). Injection of PF4 into primates infused with thrombin increases APC generation two- to three-fold and prolongs the baseline aPTT (Slungaard et al., 2003). Additional studies should clarify whether HIT antibodies interfere with the anticoagulant function of PF4 and thereby may predispose to warfarin-associated venous limb gangrene.

Not all of PF4's biologic effects are mediated by electrostatic interactions. Two putative PF4 cellular receptors have been identified. Like other chemokines, PF4 binds to the Duffy antigen/receptor for chemokines (DARC) (Tournamille et al., 1997), which has been identified on ECs in postcapillary venules and in the splanchnic bed, even in individuals who do not express the antigen on their erythrocytes (Peiper et al., 1995). The distribution of Duffy on ECs in other vascular beds is less well studied. The binding site of PF4 on DARC likely coincides with the binding sites of other chemokines located on the NH2-terminal domain (Tournamille et al., 1997). The role of DARC on red cell surfaces and on various vascular beds in the natural clearance of PF4 remains to be defined.

PF4 also binds to an alternatively spliced isoform of the chemokine receptor CXCR3, CXCR3B (Lasagni et al., 2003). CXCR3, now termed CXCR3A, is a seven transmembrane chemokine receptor that binds Mig (CXCL9), IP-10 (CXCL10), and I-TAC (CXCL11). CXCL -9, -10 and -11 are chemokines that regulate lymphocyte chemotaxis and inhibit angiogenesis. CXCR3B, derived from an alternative splicing site within the intron of the CXCR3 gene, has high affinity for PF4 (Kd = 4 nmol/L) and mediates the anti-angiogenic effects of PF4. CXCR3B is expressed on microvascular ECs of the heart, kidney, liver, and skeletal muscle (Lasagni et al., 2003). Whereas overexpression of CXCR3A into human microvascular ECs increases survival, overexpression of CXCR3B is associated with low proliferation and increased apoptotic cell death. It is presumed that binding of PF4 to the CXCR3B receptor mediates the described anti-angiogenic properties of PF4 (Yamaguchi et al., 2005; Strieter et al., 2006) though the need for micromolar concentrations of PF4 to see an antigenic effect with PF4 in most settings is more consistent with its binding to surface HSPG.

A pathogenic role of PF4 in atherogenesis has been recently elucidated. Human atherosclerotic lesions are invested with PF4 (Pitsilos et al., 2003) (Fig. 2; see color insert). PF4 is found not only along the overlying endothelium, but also in foam cells and acellular portions of the plaque. In vitro, PF4 binds to the low density lipoprotein (LDL) receptor and to proteoglycans, forming ternary complexes that show limited migration into clathrin-coated pits, thereby retarding endocytosis and catabolism of LDL (Sachais et al., 2002). PF4 binds directly to oxidized LDL, promoting foam cell formation (Nassar et al., 2003) and upregulat-ing expression of E-selectin, an adhesion molecule implicated in atherogenesis (Yu et al., 2005). In mice, activated platelets deposit PF4 on endothelium and mono-cytes, potentiating effects of P-selectin on platelet-leukocyte aggregate formation and atherosclerotic development (Huo et al., 2003). In a murine HIT model, hypercholesterolemic diet is associated with increased platelet reactivity and EC activation, as indicated by elevated levels of soluble vascular cell adhesion molecule (VCAM) (Reilly et al., 2006). Antibodies to PF4-heparin complexes have recently been identified as an independent predictor of myocardial infarction at 30 days in patients presenting with acute coronary ischemic syndromes (Williams

0 0

Post a comment