sPLA₂-V: a new player in thrombosis?

The protein C anticoagulant system plays a critical role in the regulation of haemostasis and inflammation. Binding of protein C to endothelial cell protein C receptor (EPCR) and the subsequent activation of EPCR-bound protein C by thrombin/thrombomodulin are the key steps in the protein C anticoagulant system. Activated protein C (APC) -mediated cell signalling, supported by its binding to EPCR or independent of EPCR, activates anti-inflammatory and cytoprotective pathways [1]. Therefore, pathophysiologic conditions that impair the interaction of protein C with EPCR have the potential to decrease APC generation. Such reduced binding to EPCR may not only contribute to thrombotic disorders, but also result in the loss of endogenous APC-mediated cytoprotection..

EPCR shares sequence and three-dimensional homology with the major histocompatibility class 1/CD1 family of proteins, particularly CD1d [2,3]. Similar to the CD1 family proteins, EPCR has a tightly bound phospholipid in the antigen presenting groove [3]. This similarly led to an initial belief that EPCR may play a role in presenting protein C/APC or lipid antigen in inflammatory cells [2]. However, at present, there is no evidence that, similar to CD1d, EPCR plays a role in presenting lipid antigens to inflammatory cells. However, the presence of the lipid in EPCR seems to be essential for EPCR binding to protein C, as extraction of the lipid from EPCR abolishes protein C binding [3]. Molecular dynamic simulations of phosphatidylethanolamine-bound and -unbound forms of EPCR reveal that the lipid likely maintains the conformation of EPCR that is necessary for the interaction with its ligands [4]. Recently, Hermida and colleagues showed that phosphatidylcholine (PC) is the major phospholipid bound to human EPCR, and that lipid exchange can occur in EPCR, just as it can in CD1d [5]. They also showed that the exchange of PC in EPCR for lyso PC or platelet activating factor (PAF) impaired the ability of EPCR to bind protein C and FVII, indicating that EPCR function could be modulated by a change in the identity of the phospholipid in the hydrophobic groove of EPCR. Secretory group V phospholipase A2 (sPLA₂-V), an enzyme that can be upregulated in a variety of inflammatory conditions and that metabolises PC into lyso PC, is capable of modulating both the binding of protein C to EPCR and the generation of APC on endothelial cells [5]. These data have raised the possibility that sPLA₂-V may exert prothrombotic and proinflammatory effects through the modification of the bound lipid in EPCR.

In a study published in this issue of Journal of Thrombosis and Haemostasis, Tamayo et al. provide evidence that sPLA₂-V, indeed, plays a thrombogenic role in vivo [6]. The data presented in the manuscript show that overexpression of sPLA₂-V in mice by hydrodynamic gene delivery impairs the ability of mice to activate protein C. More importantly, sPLA₂-V overexpression accelerates thrombus formation in a carotid artery laser thrombosis model. When EPCR was blocked with a blocking antibody, sPLA₂-V overexpression no longer had a significant influence upon APC generation or thrombus formation. Furthermore, administration of manoalide, an inhibitor of sPLA₂-V, significantly increased APC generation and moderately reduced thrombus formation in wild-type mice. From these data, the authors conclude that sPLA₂-V downregulates protein C activation in vivo by encrypting EPCR, and thus promotes thrombus formation. The authors raise the possibility of targeting sPLA₂-V activity as an antithrombotic strategy.

The present in vivo study builds on the authors’ earlier study using yeast expressed purified human soluble EPCR (sEPCR) and EPCR expressed on endothelial cells. To understand the true significance of the present study and its limitations, one should first know what the earlier study had shown, and more importantly, what it did not show. The earlier study showed the followings: (1) PC is the phospholipid located in the hydrophobic pocket of EPCR; (2) delipidated sEPCR had reduced ability to interact with its ligands; (3) lyso PC and PAF can locate into the hydrophobic pocket of sEPCR; (4) sEPCR containing lyso PC or PAF has impaired APC binding; (5) inhibition of sPLA₂-V on endothelial cells, either by treating cells with sPLA₂-V inhibitor or silencing sPLA₂-V gene, increased the ligand binding to EPCR and enhanced APC generation on endothelial cells. The authors strongly imply from this data that sPLA₂-V modulates the EPCR ability to generate APC by hydrolysing the PC in the EPCR to lyso PC. It is important to note here that there is no evidence presented in this study to show that the inhibition of the sPLA₂-V activity actually changed the lipid in EPCR on endothelial cells.

The data presented in the current report [6] clearly demonstrates that overexpression of sPLA₂-V inhibits APC generation, whereas inhibition of endogenous sPLA₂-V increases APC generation in vivo. sPLA₂-V overexpression did not alter the expression levels of EPCR or TM, and appears to have no significant effect on overall haemostatic balance as measured in whole blood thromboelastometry. However, this study does not provide evidence that sPLA₂-V actually modifies the lipid bound to EPCR, or that such lipid modification is responsible for the observed decrease in the APC generation in mice overexpressing sPLA₂-V. With the current existing methodologies, it may not be feasible to show in vivo that sPLA₂-V alters the lipid in the EPCR on the endothelium. However, lack of direct evidence for such a mechanism even in the in vitro cell system makes it difficult to conclude definitively that sPLA₂-V-mediated lipid modification in EPCR is responsible for the reduced APC generation thus results in accelerated thrombus formation in mice overexpressing sPLA₂-V. There are a number of other caveats and limitations in the present study. Some of them are probably unavoidable, but others are avoidable. Lack of a suitable method to effectively measure the plasma levels of sPLA₂-V makes it difficult to determine any correlation exists between sPLA₂-V and APC levels in the above group of mice. There is no information in the present study whether sPLA₂-V overexpression actually impairs protein C binding to EPCR on the endothelium. This could have been easily examined by measuring protein C levels in circulation in wild-type and sPLA₂-V overexpressing mice. This study measures APC levels in mice following PC/phosphatidylserine/factor Xa challenge and not in control unchallenged mice. Since the reduction of APC levels in sPLA₂-V overexpressing mice is proposed to be responsible for the accelerated thrombus formation following the laser injury in mice that are not challenged with PC/phosphatidylserine/factor Xa prior to the laser injury, it is more appropriate to measure APC levels in unchallenged mice. To substantiate that EPCR is needed for sPLA₂-V to exert its prothrombotic effect, sPLA₂-V overexpressing mice were administered with anti-EPCR mAb that blocks protein C/APC binding to EPCR. This experiment showed that sPLA₂-V overexpression was unable to accelerate the thrombus formation when EPCR was blocked with the antibody. But the EPCR blocking antibody treatment exhibits a stronger prothrombotic effect than sPLA₂-V overexpression (i.e., greater reduction in APC generation and increased thrombus formation). Thus, blocking protein C binding to EPCR is not an appropriate experiment to assess the role of EPCR in sPLA₂-V-mediated prothrombotic effect.

Despite the above caveats and limitations, the present study provides valuable and novel information, i.e. sPLA2-V plays a thrombogenic role in vivo. sPLA₂-V is constitutively expressed by endothelial cells, and its expression is increased by inflammatory stimuli, such as tumour necrosis factor-α and vascular endothelial growth factor [7,8]. sPLA₂-V is expressed in all stages of atherosclerosis development and thought to play an atherogenic role through degradation of phospholipids [9,10]. Therefore, it is entirely possible that upregulation of sPLA₂-V in inflammation and other diseases may contribute, at least partly, to thrombotic disorders associated with these diseases. If so, inhibition of sPLA₂-V activity in such patients may represent an effective antithrombotic therapeutic strategy. To further investigate this possibility, it is important to determine first whether sPLA₂-V levels correlate with thrombotic risk in specific patient groups. This requires development of a suitable assay to measure sPLA₂-V levels in plasma.

APC plays an influential role in many disease processes, including thrombosis, sepsis, inflammation, diabetes and cancer [1,11–13]. Protein C binding to EPCR is critical for the generation of APC. APC-mediated anti-inflammatory and cytoprotective effects most frequently require APC binding to EPCR. In addition to protein C and APC, many other ligands bind EPCR. They include factor VIIa, proteinase-3, β2 integrin Mac-1, Plasmodium falciparum erythrocyte membrane protein 1, and γδ T cell antigen receptor [1]. Recent studies elucidating the importance of interactions between these novel ligands and EPCR indicate new roles for EPCR in haemostasis, the pathogenesis of severe malaria, innate immunity and cancer [1]. Interestingly, most of the ligands that bind EPCR do so at sites that overlap with the binding site for protein C. It will therefore be very interesting to ascertain whether EPCR lipid modification also influences the binding of other ligands to EPCR. If the proposed mechanism of sPLA₂-V-mediated impairment of protein C activation is proven to be correct, i.e. sPLA₂-V-mediated lipid modification in EPCR results in the loss of the ligand binding to EPCR, it opens an exciting new area of research investigation related to regulation of EPCR function through lipid modifications within EPCR. If the present findings were confirmed and extended, use of sPLA₂-V inhibitors to treat various thrombotic, inflammatory and vascular diseases may become a viable option in the future. For this, we have to build a stronger basic foundation by convincingly demonstrating that sPLA₂-V modifies the lipid in EPCR in vivo, and this lipid modification is responsible for EPCR encryption.

Fig. 1.

A hypothetical model of sPLA₂-V impairment of protein C activation. Functional EPCR contains a phosphatidylcholine (PC) in its hydrophobic groove. In this conformation, EPCR can bind protein C. Thrombin (T):thrombomodulin (TM) complex activates the protein C bound to EPCR to generate activated protein C (APC). Secreted phospholipase A2, group V (sPLA₂-V) hydrolyses the PC in the EPCR to lyso PC. The lipid modification in the hydrophobic groove of EPCR leads to a structural rearrangement in the helical region of EPCR, which probably results in narrowing of the pocket of the ligand-binding groove. The structurally rearranged EPCR is unable to bind protein C. The loss of protein C binding to EPCR translates into impaired protein C activation by thrombin:thrombomodulin complex.