"V体育ios版" Platelet aggregation Inhibitors from Hametophagous Animals

2.2.1. RPAI-1

Rhodnius prolixus aggregation inhibitor 1 (RPAI-1) is a 19-kDa lipocalin isolated from the salivary gland of R. prolixus (Francischetti, et al., 2000). RPAI-1 was the first sialogenin whose function was identified as ADP-binding protein (K_D ∼ 50 nM). RPAI-1 also displays high-affinity binding to ATP, AMP, adenosine (Ado), AP4A, and α,β Met ADP but does not bind to inosine, guanosine, uridine, or cytidine, implying that Ado structure is necessary for binding (Francischetti, et al., 2002). Despite its high affinity, it only inhibits platelet aggregation triggered by low doses of ADP (< 0.5 μM) (Figure 2A) or appropriate concentrations of collagen (Figure 2B), arachidonic acid, U46619, TRAP, PAF, and A23187 (Francischetti, et al., 2002, Francischetti, et al., 2000), which is ADP dependent (Kahner, et al., 2006, Watson, et al., 2005). In fact, high doses of ADP or other pro-aggregatory substances will consistently not (or only marginally) be affected by this family of inhibitors, as the amount of ADP released rapidly saturates the mopping up ability of the inhibitor that occurs at a 1:1 stoichiometry (Francischetti, et al., 2000). This is important to keep in mind to avoid a given inhibitor being erroneously classified as a specific inhibitor of collagen, an agonist that—at the usual in vitro concentration of ∼ 1 μg/ml—is particularly sensitive to ADP scavengers. RPAI-1 also inhibits platelet aggregation under high shear observed in the microcirculation as demonstrated by platelet function analyzer (PFA-100) or clot signature analyzer (Francischetti, et al., 2002).

Two questions are often raised about the true biological function of lipocalins such as RPAI-1 in vivo. The first is whether the concentration of the inhibitor found in saliva is compatible with the concentrations of pro-aggregatory components produced upon injury at the microcapillary level. The second is whether a high-affinity ADP-binding protein is needed in the saliva, because apyrases effectively inhibit platelet aggregation. Accordingly, computer simulation experiments were performed where the rate of scavenging or hydrolysis of ADP by RPAI-1 or R. prolixus (RP) apyrase (a 5/-nucleotidase) —which have distinct K_m and K_cat for ADP—were compared. The simulation results indicate that RPAI-1 (high affinity for ADP, no hydrolysis) is functionally more effective at low ADP concentrations (< 1 μM). In contrast, effective degradation of ADP by the RP apyrase (low affinity for ADP and enzymatic degradation) occurs when the ADP concentration is at least 1000- to 5000-fold higher (50 μM) than that readily removed by RPAI-1. Of note, ADP physiological plasma concentration is ∼ 0.1 μM, and it can reach 1–2 μM at the site of injury provoked by a Simplate device; 0.2–0.4 μM ADP is known to be biologically active (Francischetti, et al., 2002, Francischetti, et al., 2000). Accordingly, RPAI-1 and RP apyrase coexisting in the same secretion have precise and complementary biologic functions in preventing platelet aggregation in vivo.

2.3.1. Moubatin and TSGP3

Moubatin is a sialogenin cloned from the salivary gland of the soft tick Ornithodorus moubata salivary gland (Keller, et al., 1993, Waxman and Connolly, 1993). Recombinant moubatin inhibited collagen-stimulated aggregation of washed human platelets with an IC₅₀ of about 100 nM, while aggregation induced by other ligands was not affected. Moubatin did not inhibit platelet adhesion to collagen, supporting the view that it uniquely inhibited a step involved in platelet activation; however, at higher concentrations (2–6 μM), moubatin diminished the second phase of aggregation induced by ADP, inhibited aggregation in response to submaximal concentrations of the TXA2 mimetic U46619, and competed for the binding of a TXA2 receptor antagonist to platelet membranes. (Keller, et al., 1993). Therefore, moubatin behaved similarly to RPAI-1, because at high concentrations it affects ADP and U46619-induced platelet aggregation (Francischetti, et al., 2000). It has been suggested that moubatin interferes with a pathway associated with TXA₂ receptor (Keller, et al., 1993), but the molecular target of moubatin remained elusive.

More recently, moubatin and another lipocalin from O. savignyi (TSGP3) were expressed in Escherichia coli and shown to inhibit collagen-induced platelet aggregation (Mans and Ribeiro, 2008). As described before (Keller, et al., 1993), inhibition of platelet aggregation was partial, and shape change was still present as depicted in Figure 2C (Mans and Ribeiro, 2008). Of note, isothermal titration calorimetry (ITC) demonstrated that Moubatin and TSGP3 bind cTXA2 with high affinity (24 and 5 nM, respectively), but not ADP (Mans and Ribeiro, 2008). The implications of scavenging TXA₂ are many, taking into account that this prostanoid is a potent platelet aggregation agonist as well as a vasoconstrictor. In fact, moubatin (and TSGP3) relaxes rat aorta pre-constricted by U46619 and inhibits contraction induced by U46619 in a concentration-dependent manner. It is likely that the concentrations of moubatin and TSGP3 in saliva is enough to counteract TXA2 produced in vivo, inhibiting vasoconstriction and platelet aggregation at the site of feeding (Mans and Ribeiro, 2008). Accordingly, it was previously shown to exist as ∼5% of the total protein present in salivary gland extracts of O. savignyi. It is reasonable to suggest that other lipocalins including moubatin are expressed at a similar level. If it is assumed that only 50% of the salivary gland protein is secreted during feeding and that the feeding site is between 10–50 μl in volume, then moubatin would be present at ∼3–18 μM. This is well above the concentrations needed by TXA₂ to induce vasoconstriction or platelet aggregation and given the high affinities for TXA2 (Mans and Ribeiro, 2008).

2.3.2. Putative ADP or TXA2-binding proteins

2.4.1. Amine-binding protein (ABP)

A biogenic ABP that belongs to the nitrophorin group of lipocalins has been discovered in R. prolixus salivary gland. Using isothermal titration calorimetry and the Hummel-Dreyer method of equilibrium gel filtration, K_D values of 102, 24, and 345 nM were found for serotonin, norepinephrine, and epinephrine, respectively (Andersen, et al., 2003). While ABP does not block platelet aggregation by moderate concentrations of collagen or ADP, it inhibits platelet aggregation induced by a combination of low concentrations of ADP and either serotonin or epinephrine. Potentiation of aggregation induced by low concentrations of collagen along with serotonin (or epinephrine) is also inhibited as shown in Figure 2D (Andersen, et al., 2003). This occurs because serotonin activates the phospholipase C pathway via the 5-HT_2A receptor but not the G_i-dependent pathway and consequently induces only shape change when administered alone. Epinephrine binds to G_i-coupled adrenergic receptors, inhibiting adenylyl cyclase and eliciting no detectable aggregation response by itself (Andersen, et al., 2003).

The presence of ABP in the circulation would therefore cause a localized increase in the agonist threshold concentration for platelet aggregation. In R. prolixus salivary gland, this protein would act in concert with RPAI-1, NP7, and salivary apyrase, reducing the concentration of weak agonists in the vicinity of the feeding site and thereby attenuating the overall stimulus for platelet aggregation (Andersen, et al., 2003).

2.4.1. D7

The D7 protein family is one of the most abundantly expressed sialogenins of mosquitoes and one of the first to be cloned from the salivary glands of insects. Recombinant short D7 of An. gambiae were shown to bind serotonin and norepinephrine with high affinity, and it also binds histamine and epinephrine with lower affinity. D7 members from Aedes sp. have been identified as high-affinity serotonin-binding proteins with specificity similar to An. gambiae short D7 (Calvo, et al., 2006). Elucidation of the crystal structure and their binding pockets revealed that the amino domain of the long D7 protein of Aedes sp. could bind a hydrophobic compound, which was identified as cysteinyl leukotrienes (Calvo, et al., 2009). While the function of short D7 is reportedly to modulate tonus of vessels, it is plausible that scavenging serotonin may block potentiation of platelet activation by serotonin, as described above for ABP from R. prolixus.

2.4.2. Monotonin, TSGP-1 and SHBP

Monotonin has been isolated from the salivary gland of the soft tick Argas monolakensis and shown to bind to serotonin with high affinity (K_D < 2 nM) (Mans, et al., 2008); it may also block platelet aggregation as reported for ABP from R. prolixus (Andersen, et al., 2003). It has been calculated that the concentration of monotonin in the saliva is compatible with antihemostatic activity at the site of injection (Mans, et al., 2008). The crystal structure of monotonin has been determined and indicates that the protein has a single binding site for serotonin. From the conserved features of these proteins, a tick lipocalin biogenic amine-binding motif could be derived that was used to predict biogenic amine-binding (BAB motif) function in other tick lipocalins. Another serotonin-binding protein, TSGP1, has been characterized from O. savignyi, and its sequence also contains the BAB motif. It binds 5-HT (K_D ∼ 6 nM) with a stoichiometry of ∼1 (Mans, et al., 2008). Finally, SHBP from Dermacentor reticularis is another lipocalin that binds serotonin (and histamine) with high affinity and may share antiplatelet properties with monotonin and TSGP1 (Sangamnatdej, et al., 2002).

2.5.1. Aegyptin and Anopheline antiplatelet protein (AAPP)

Aegyptin is a 30-kDa salivary sialogenin from Ae. aegypti that displays a unique sequence characterized by glycine, glutamic acid, and aspartic acid repeats. It specifically blocks collagen-induced human platelet aggregation and granule secretion (IC₅₀ ∼ 50 nM) without affecting aggregation induced by other agonists. Figure 2E shows tracings of collagen-induced platelet aggregation where, in the presence of 60 nM aegyptin, the onset time for shape change is increased, while at higher doses (120 nM), shape change was abolished. Surface plasmon resonance (SPR) experiments demonstrate that aegyptin binds to collagen (K_D ∼ 1 nM) but does not interact with vitronectin, fibronectin, laminin, fibrinogen, or vWF (Calvo, et al., 2007). Aegyptin inhibits collagen-induced platelet aggregation (IC₅₀ 100 nM) and blocks vWf interaction with collagen type III under static conditions and platelet adhesion to collagen under flow conditions at high shear rates with similar IC₅₀ (∼ 300 nM) (Calvo, et al., 2007). More recently, the sequence RGQOGVMGFO, which mediates collagen interaction with vWF, has been identified as a high-affinity binding site for aegyptin. Because it also interacts with the linear peptide RGQPGVMGFP and heat-denatured collagen, triple-helix and hydroxyproline residues are not a prerequisite for binding. Aegyptin also recognizes (GPO)₁₀ and GFOGER peptides with low-affinity, which represent GPVI and integrin α2β1 binding sites in collagen, respectively (Calvo et al., unpublished observations). Consistent with these results, aegyptin interferes with platelet interaction with soluble collagen and prevents collagen binding to recombinant GPVI. In vivo experiments show that aegyptin prevents laser-induced carotid thrombus formation in the presence of Rose Bengal, a model where collagen reportedly plays an important role. Inhibition of thrombus formation is observable at 50 μg/kg, while occlusion of the carotid took more than 80 min at doses of 100 μg/kg. Rats treated at antithrombotic concentrations did not bleed significantly, suggesting that aegyptin is a suitable molecule to inhibit platelet-collagen interactions in vivo (Calvo et al., unpublished observations).

Anopheline antiplatelet protein (AAPP) from Anopheles stephensi saliva is a 30-kDa protein that displays strong sequence similarity to aegyptin. Recombinant AAPP directly binds to immobilized collagen and specifically inhibits collagen-induced platelet aggregation (IC₅₀ ∼ 25 nM) and intracellular Ca²⁺ increase. It blocks both platelet integrin α2β1-dependent and GPVI-expressing Jurkat cell-mediated adhesion to collagen. Intravenous injection of AAPP in rats inhibited collagen-induced platelet aggregation ex vivo (Yoshida, et al., 2008).

2.5.2. Leech antiplatelet protein (LAPP)

LAPP was identified in the salivary glands of the leech Haementeria officinalis. Recombinant LAPP inhibits collagen-mediated platelet aggregation under test-tube stirring conditions (IC₅₀ ∼ 60–100 nM) without affecting other agonists. It also blocks platelet adhesion to soluble collagen under static conditions (IC₅₀ ∼ 80 nM), a step mediated by integrin α2β1 (Connolly, et al., 1992, Keller, et al., 1992). Consistent with these results, rLAPP prevents integrin α-I domain binding to collagen with IC₅₀ ∼ 2 μg/ml (125 nM)(Depraetere, et al., 1999). rLAPP also inhibits platelet adhesion to collagen type I (IC₅₀ 70 nM) at high shear rate (1600 s⁻¹) and prevents binding of vWF to collagen type III (van Zanten, et al., 1995). While rLAPP inhibits platelet deposition to cross sections of human atherosclerotic coronary arteries (van Zanten, et al., 1995) it did not block collagen graft thrombosis in baboons, suggesting that inhibition of collagen alone is not enough to prevent thrombosis (Schaffer, et al., 1993), possibly because tissue factor exposure plays an important role in graft models. The structure of LAPP has been determinedand consists of a compactly folded C-terminal domain and an N-terminal region that is disordered in the crystal. The C-terminal domain folds similarly to the N-domain of hepatocyte growth factor and has been classified as the so-called PAN domain (Huizinga, et al., 2001). A pattern of four conserved cysteines forming two disulfide bonds and of five residues with a conserved hydrophobic character may be characteristic of this type of fold.

2.5.3. Saratin

Saratin, isolated from the saliva of the leech Hirudo medicinalis, binds to collagen with high affinity (K_D ∼ 50 nM; SPR experiments). It is a potent inhibitor of vWF binding to collagen (IC₅₀ ∼ 1 μg/ml; ∼ 100 nM) according to adhesion experiments performed at high shear rates (Barnes, et al., 2001). Platelet adhesion to collagen at 300 s⁻¹ is not inhibited at 10 μg/ml (1 μM) saratin, and only very high concentrations of saratin (200 μg/ml; ∼ 20 μM) inhibit platelet aggregation by collagen. Therefore, it has been suggested that saratin, in contrast to aegyptin, LAPP, and calin, is specific for the vWF-collagen interaction. More recently, however, saratin was shown to inhibit binding of the α2 integrin subunit I domain to collagen and prevents platelet adhesion in the presence of ADP and TXA2 inhibitors, suggesting that it interferes with integrin α2β1 binding to collagen in addition to inhibiting vWF-collagen binding (White, et al., 2007).

Structural determination of saratin-collagen complex in solution has been revealed by NMR spectroscopy. Saratin has high structural homology to LAPP, which is a distant member of the PAN domain superfamily. In fact, when comparing the secondary structure elements of the solution structure of saratin with the structure of LAPP, both proteins show a very similar arrangement. The structural homology in core regions of the two proteins suggests that both bind to collagen in a similar way (Gronwald, et al., 2008).

Because of its therapeutic potential, saratin has been tested in vivo and found to significantly decrease platelet adhesion, intimal hyperplasia, luminal stenosis, and thrombosis after carotid endarterectomy in rats without increasing bleeding time. It also decreases venous anastomotic intimal hyperplasia in a canine dialysis model. Further, under stenotic shear conditions of 800 s⁻¹, saratin reduced (40% to 60%) platelet deposition triggered by human denuded vessel walls, fatty streaks, severely damaged vessels, and atherosclerotic plaque (Cruz, et al., 2001, Vilahur, et al., 2004). These results support the view that collagen-binding proteins may be beneficial as antithrombotic agents under certain pathologic conditions.

2.5.4. Calin

Calin is a semipurified substance (∼ 65 kDa.) from H. medicinalis that has not yet been molecularly characterized. It inhibits platelet aggregation by collagen (IC₅₀ ∼ 6.5 to 13 μg/ml; 100–200 nM) and adhesion to immobilized collagen (IC₅₀ ∼ 22 μg/ml; 400 nM) (Deckmyn, et al., 1995). Notably, it prevents binding of vWF to coated collagen under static conditions (IC₅₀ 10 μg/ml) and high shear rates (1300s⁻¹; ∼ 80 μg/ml) (Depraetere, et al., 1999, Harsfalvi, et al., 1995). Therefore, calin inhibits direct platelet-collagen interactions and vWF-binding to collagen. This dual effect may contribute to prevention of thrombus formation observed under flow (Harsfalvi, et al., 1995). In vivo activity of calin has been tested in a thrombosis model in hamsters based on vessel clamp damage to the femoral vein. Results show that intravenous calin inhibited thrombus formation in this model with an ED₅₀ of 0.07 mg/kg and complete inhibition with 0.2 mg/kg. No effects were seen on coagulation tests or bleeding times, whereas ex vivo aggregation induced by collagen was inhibited dose dependently (Deckmyn, et al., 1995). Because the molecular identity of calin is not known, the precise mechanism of action remains unclear.

2.7.1. Nitrophorins

In R. prolixus, NO is stored in the lumen of the salivary glands as a stable nitrophorin(NP)-NO complex. Binding is tight at the pH of the salivary gland (∼ 5), but affinity is lower at the pH of the host (∼ 7.5), a property that facilitates release of NO upon saliva injection (Montfort, et al., 2000). Of interest, the ability to transport NO has evolved independently in the bedbug C. lectularius, and is again heme-based but apparently does not involve the lipocalin fold (Valenzuela, et al., 1995). The structure of the NP4-NO complex has been determined. Binding of NO induces a conformational change in the spacious distal pocket, and the NO moiety becomes completely buried and surrounded by hydrophobic groups. By releasing NO at the site of an arthropod bite, NP conceivably inhibits platelet aggregation. This assumption has been confirmed in vitro in experiments where one NP family member (NP7) loaded with NO reportedly inhibits ADP-induced platelet aggregation and promotes disaggregation of platelets (Figure 2F) in addition to being an anticoagulant through binding to phosphatidylserine (Andersen, et al., 2004).