The role of platelets in decrypting monocyte tissue factor

Localization of TF in cells 

The TF activity associated with lipopolysaccharide (LPS)-stimulated cells is known to be strongly encrypted. Although about 80% of the TF extracellular domain antigen in activated monocytes is present at the surface, only 10% to 15% of the total TF activity available upon lysis is actually available on intact cells.25 Recently, three pools of TF antigen constituting the total TF were quantified by lysing smooth muscle cells with β-octyl glucoside.35 It was estimated that about 20% of the total cellular TF is available on the surface, about 30% is intracellular, and approximately 50% is latent. The intracellular material was found to be sufficiently associated with the cell membrane to be biologically active if released.20 Thus, when TF-bearing cells are subjected to a succession of freeze-thaw cycles, calcium ionophore, or phorbol 12-myristate 13-acetate (PMA), TF activity increases markedly without significant changes in TF mRNA or protein.4, 8

In order to explain the encryption phenomenon, it has been suggested that TF molecules are clustered in the membrane and only reach their full potential when dispersed.20 Calcium ionophore has been shown to cause a 100-fold rise in the TF activity of intact cells,2 an affect that can be blocked by pretreating the cells with calmodulin inhibitors. It was hypothesized that the rise in TF activity upon decryption may stem from the exposure of an essential macromolecular substrate-binding site in the TF/FVIIa complex, as a direct consequence of conversion of TF dimers to monomers. A recent study concluded that the effect of calcium ionophore in decrypting TF activity is not solely the result of increased availability of exposed PS, and that the FVIIa/TF substrate specificity is not altered by the decryption process.37

Interactions of monocytes with other blood cells may therefore not only affect their TF activity at the level of synthesis or degradation, but also at the level of activity encryption or decryption. TF activity expression in blood may be regulated by the profusion of microparticles derived from platelets and monocytes, which may be one of the most efficient and important mechanisms involved in decrypting monocyte-derived TF activity in vivo.

Role of platelets in the expression of TF in monocytes of whole blood 

Even before it was established that monocytes produce TF, it was shown that platelets enhance the expression of TF in leukocytes.21 This was later confirmed in monocyte cell cultures, where a mechanism for the upregulation of TF by 12-hydroxyeicosatetraenoic acid (HETE) derived from activated platelets was suggested.16, 30 Correspondingly, a platelet effect in our whole blood ex vivo system was demonstrated.26 In this model, comprising blood anticoagulated with heparin or hirudin and stimulated with LPS (5 ng/mL), TF activity varied about 50-fold between individuals.24 Part of this variation was due to individual differences in platelet function as when platelets from a high responder were added to a reconstituted cell/plasma system (platelet-depleted blood cell fraction combined with platelet poor plasma [PPP]) from a low responder, we observed an upregulation of TF expression.26

Using a more refined system, which consisted of separated and then recombined fractions of blood cells and plasma (ie, mononuclear cells, granulocytes, heparinized platelet-rich plasma [PRP], and PPP), we were able to demonstrate that the platelet effect depended on the presence of granulocytes.10 Furthermore, this platelet/granulocyte effect was mediated through the interaction of P-selectin on activated platelets with the CD15 ligand on monocytes and granulocytes.

We then examined whether this platelet effect reflected upregulation of the procoagulant activity of TF or was due to increased TF synthesis. Platelet-depleted blood (platelets 10% of normal) was recombined with either PPP or PRP and then stimulated with LPS (5 ng/mL) alone or in combination with tumor necrosis factor (TNF) or the phorbol ester PMA. The presence of platelets did not affect TF antigen synthesis under any of these conditions, but did increase mean TF activity by 190% to 240%.27 In contrast, platelet lysate not only enhanced LPS-induced TF activity by about threefold, but also substantially increased LPS-induced TF antigen synthesis.

Models for the decryption of monocyte TF by platelets 

There are at least two mechanisms that could explain the platelet effect observed in our ex vivo blood model. First, a platelet-monocyte fusion event with subsequent exchange of available PS. By acquiring PS from the platelet membrane, the specific activity of TF antigen present in the monocyte membrane would be upregulated. Second, the generation of monocyte-derived microparticles with decrypted TF activity. This explanation is supported by the observation that monocyte-derived TF-containing microparticles may be associated with platelet microparticles, a phenomenon that might be conducive to platelet upregulation of TF activity, indirectly via its microparticles.5 It is noteworthy that PS is also exposed on the surface of monocyte-derived microparticles.34 Hence, this surface may facilitate the activation of FVII, leading to optimal generation of TF/FVIIa complexes, which in turn bind and activate FIX and FX. Regardless of which mechanism predominates, there is mounting evidence to support a role for platelets in decrypting TF activity in monocyte membranes.

Monocyte shedding of microparticles with membrane associated procoagulant activities has been obtained by LPS stimulation of monocytes.34 On these microparticles exposed TF, PS (the active template in the coagulation enzyme complex), and the adhesion molecules CD14, CD11a, and CD18 have all been detected.

An atherogenic role for microparticles, probably of mainly monocyte origin, can be indirectly inferred from a study where high levels of shed membrane apoptotic microparticles expressing both TF and PS were detected in extracts from atherosclerotic plaques.17 It is known that microparticles derived from platelets are relatively thrombogenic.11, 15 They are released from activated platelets and express functional adhesion receptors, including P-selectin, on their surface. Platelet microparticles provide a catalytic surface comprising PS that accelerates coagulation,29, 33, 36 and they can bind to neutrophils and monocytes.13 Platelet microparticles have been shown to be present in various disease states.1, 9, 12, 15, 17, 22, 23, 38

A recent study found that thrombogenic TF present on leukocyte-derived microparticles was incorporated into spontaneously formed human thrombi.31 It was suggested that monocytes and possibly polymorphonuclear (PMN) leukocytes are the sources of the circulating plasma TF, which will be transferred to activated platelets, in effect producing TF-positive platelets capable of triggering and propagating thrombosis. This transfer process was mediated by the interaction of CD15 with platelets and also by TF, which seems to act as an adhesion molecule.32 This observation concurs with our own findings that anti-CD15 antibodies abolish about 80% of LPS-induced TF activity in monocytes of cell suspensions recombined with PRP.10

In another study using monocytes recombined with various amounts of PRP and incubated for 24 hours in the presence of LPS plus PMA, only a small amount of monocyte TF activity was generated in the absence of platelets, increasing almost 20-fold when the monocytes were recombined with platelets.5 Furthermore, isolated platelets possessed significant amounts of TF after stimulation of blood with LPS alone for 2 or 6 hours, whereas no such activity was discernible after 24-hour incubation. Platelet TF activity at 2 and 6 hours was further enhanced by the addition of PMA and was still detectable at 24 hours (≈50% of the 6-hour value). Microparticles isolated from blood stimulated with LPS alone had no detectable TF activity as measured by a sensitive clotting assay, whereas microparticles isolated from blood stimulated with LPS and PMA expressed significant amounts of TF activity at 6 hours and particularly at 24 hours.

Conclusion 

These results indicate that platelets play a central role in the decryption of TF and the transfer of TF from monocytes to microparticles. The TF reported to be associated with PMN leukocytes may also be accounted for by a mechanism involving the generation and transferral of TF-bearing microparticles.