Dipetalodipin, a Novel Multifunctional Salivary Lipocalin That Inhibits Platelet Aggregation, Vasoconstriction, and Angiogenesis through Unique Binding Specificity for TXA2, PGF2α, and 15(S)-HETE*

Dipetalodipin (DPTL) is an 18 kDa protein cloned from salivary glands of the triatomine Dipetalogaster maxima. DPTL belongs to the lipocalin superfamily and has strong sequence similarity to pallidipin, a salivary inhibitor of collagen-induced platelet aggregation. DPTL expressed in Escherichia coli was found to inhibit platelet aggregation by collagen, U-46619, or arachidonic acid without affecting aggregation induced by ADP, convulxin, PMA, and ristocetin. An assay based on incubation of DPTL with small molecules (e.g. prostanoids, leukotrienes, lipids, biogenic amines) followed by chromatography, mass spectrometry, and isothermal titration calorimetry showed that DPTL binds with high affinity to carbocyclic TXA2, TXA2 mimetic (U-46619), TXB2, PGH2 mimetic (U-51605), PGD2, PGJ2, and PGF2α. It also interacts with 15(S)-HETE, being the first lipocalin described to date to bind to a derivative of 15-lipoxygenase. Binding was not observed to other prostaglandins (e.g. PGE1, PGE2, 8-iso-PGF2α, prostacyclin), leukotrienes (e.g,. LTB4, LTC4, LTD4, LTE4), HETEs (e.g. 5(S)-HETE, 12(S)-HETE, 20-HETE), lipids (e.g. arachidonic acid, PAF), and biogenic amines (e.g. ADP, serotonin, epinephrine, norepinephrine, histamine). Consistent with its binding specificity, DPTL prevents contraction of rat uterus stimulated by PGF2α and induces relaxation of aorta previously contracted with U-46619. Moreover, it inhibits angiogenesis mediated by 15(S)-HETE and did not enhance inhibition of collagen-induced platelet aggregation by SQ29548 (TXA2 antagonist) and indomethacin. A 3-D model for DPTL and pallidipin is presented that indicates the presence of a conserved Arg39 and Gln135 in the binding pocket of both lipocalins. Results suggest that DPTL blocks platelet aggregation, vasoconstriction, and angiogenesis through binding to distinct eicosanoids involved in inflammation.

The hemostatic process, a host defense mechanism to preserve the integrity of the circulatory system, remains inactive until vascular injury occurs, leading to activation of hemo-stasis. The first step in this cascade of events is platelet interaction with the exposed extracellular matrix (ECM), which contains a large number of adhesive macromolecules such as collagen. Under conditions of high shear, initial tethering of platelets to the ECM is mediated by interaction between the platelet receptor glycoprotein (GP) 2 Ib and vWF bound to collagen (1). This interaction allows platelet receptor GPVI to bind to collagen, triggering release of the so-called secondary mediators TXA 2 and ADP that are necessary for integrins ␣2␤1 and ␣ IIb ␤ 3 activation and completion of platelet aggregation (2). Vasoconstriction is another critical step triggered by injury and mediated by biogenic amines produced by adrenergic fibers or vasoactive components such as TXA 2 and serotonin released by platelets in an attempt to decrease blood flow at sites of injury and therefore prevent blood loss (3).
Because of the interface encountered by vectors upon interaction with their host, salivary glands from bloodsucking arthropods have evolved different mechanisms that counteract hemostasis and inflammation (4). At least thirteen different mechanisms for inhibition of platelet function are reported to explain how these molecules affect platelet function, thus assisting hematophagous animals to acquire a blood meal (5). These inhibitors have been classified as enzymes, small ligand binders, enzymes or enzyme inhibitors, nitric oxide (NO)-releasing molecules, and integrin antagonists. Among members of the lipocalin family (6), inhibitors have been reported to bind to ADP (7), biogenic amines (8 -9), and leukotrienes (10). Other salivary components interfere with hemostasis by targeting vasoconstriction, such as tachikinin-like peptides from Aedes aegypti (11) or peptides such as sandfly maxadilan, which specifically activates PAC1, the type I receptor for pituitary adenylate cyclase-activating peptide (PACAP) (12). Vasodilation is also mediated through release of NO by NO-carrying nitrophorins from Rhodnius prolixus (13). In this report, we have cloned, expressed, and studied the mechanism of action of a novel lipocalin, herein named dipetalodipin (DPTL). DPTL binds to TXA 2 , PGF 2␣ , 15(S)-HETE, and other prostanoids, and * This work was supported, in whole or in part, by the Division of Intramural Research, NIAID, National Institutes of Health. □ S The on-line version of this article (available at http://www.jbc.org) contains supplemental data. 1  was found to block platelet aggregation, vasoconstriction, and angiogenesis. The antiinflammatory and antihemostatic properties of DPTL may assist triatomines to successfully feed on blood and counteract host pro-inflammatory mechanisms triggered upon injury.
Dipetalogaster maxima Salivary Gland cDNA Construction-This was done as described before (15) and in the supplemental data. Sequencing of cDNA indicate that DPTL is an abundant secreted lipocalin (data not shown).
Sequence Analysis-Sequence similarity searches were performed using BLAST. Cleavage site predictions of the mature proteins used the SignalP program. The molar extinction coefficient (⑀ 280 nm ) of mature DPTL at 280 nm was obtained at Expasy Protemics server, yielding for mature DPTL a value of ⑀ 280 nm ϭ 20315 M Ϫ1 ⅐cm Ϫ1 ; A 280 nm/cm 0.1% (1 mg/ml) ϭ 1.132, molecular weight 17,951.1 (165 aa), and pI 8. 49.
Expression of DPTL in Escherichia coli-Synthetic cDNA for DPTL was produced by Biobasics (Ontario, Canada). The sequence displays an N-terminal NdeI and a C-terminal XhoI restriction sites. The NdeI site adds a 5Ј-methionine codon to all sequences that acts as start codon in the bacterial expression system, whereas the XhoI site was incorporated after the stop codon. pET 17b constructs were confirmed before transformation of E. coli strain BL21(DE3)pLysS cells. Detailed description of expression of recombinant DPTL is available online in the supplemental data.
Protein Purification, PAGE, and Edman Degradation-These steps were performed as described in detail in the supplemental data available online.
High-throughput Ligand Binding Assay-To investigate putative ligands of DPTL, 50 l of 100 mM ammonium acetate, pH 7.4 (AA buffer) containing 1 M DPTL and 2 M each of arachidonic acid, 15(S)-HETE, PGE 2 , PGD 2 , PGF 2␣ , TXA 2 , U-46619, U-51605, leukotriene B 4 , and carbocyclic TXA 2 were injected into a 3.2 ϫ 250-mm Superdex peptide column (GE Healthcare) equilibrated with 100 mM AA buffer. A flow rate of 50 l/min was maintained with a P4000 SpectraSystem pump (Thermo Scientific, Rockford, IL). The absorbance at 280 nm was monitored using an ABI 785 detector (Applied Biosystems, Foster City, CA). Fractions were collected into a 96-well plate every minute using a Probot apparatus (Dionex, Sunnyvale, CA). Selected fractions (20 l) were mixed with 1 l of methanol containing 1 M HCl, centrifuged at 14,000 ϫ g for 10 min, and the supernatant injected into a 0.3 ϫ 150-mm C18 reverse phase column (Magic C18 200 Å; Michrom BioResources, Inc, Auburn CA) equilibrated with 10% methanol/water containing 0.1% acetic acid at a flow rate of 3 l/min maintained by an ABI 140D pump (Applied Biosystems). After 15 min, the methanol concentration was raised linearly to 90% in the course of 30 min. The column effluent was mixed with pure methanol at a rate of 4 l/min (to facilitate electrospray) using a syringe pump attached to a LCQ Deca XP Max mass spectrometer (Thermo Scientific). Mass spectrometry was performed in negative-ion mode to detect ligand masses. A similar protocol was used to detect positively charged agonists: PAF acether, leukotrienes C 4 , D 4 , and E 4 , histamine, serotonin, norepinephrine, epinephrine, and adenosine diphosphate, with the mass spectrometer running in positive-ion capture mode.
Isothermal Titration Calorimetry (ITC)-Prostanoids (in ethanol or methyl acetate) were placed in glass vials and the vehicle evaporated under nitrogen atmosphere; the dried material was then resuspended in appropriate concentrations in 20 mM Tris-HCl, 0.15 M NaCl, pH 7.4, sonicated, and vortexed. Calorimetric assays for measuring DPTL binding to a number of ligands were performed using a VP-ITC microcalorimeter (Microcal, Northampton, MA) at 35°C. Titration experiments were performed by making successive injections of 10 l each of 40 M ligand into the 1.34-ml sample cell containing 4 M DPTL until near-saturation was achieved. Prior to the run, the proteins were dialyzed against 20 mM Tris-HCl, 0.15 M NaCl, pH 7.4, for binding experiments. The calorimetric enthalpy (⌬H cal ) for each injection was calculated after correction for the heat of DPTL dilution obtained in control experiments performed by titrating DPTL into buffer.
The binding isotherms were fitted according to a model for a single set of identical binding sites by nonlinear squares analysis using Microcal Origin software. Enthalpy change (⌬H), and stoichiometry (n) were determined according to Equation 1, where Q is total heat content of the solution contained in the cell volume (V o ), at fractional saturation , ⌬H is the molar heat of ligand binding, n is the number of sites, and M t is the bulk concentration of macromolecules in V o . The binding constant, K a , is described as Equation 2, where [X] is the free concentration of ligand.
Free-energy (⌬G) and entropy term (ϪT⌬S) of association were calculated according to Equations 3 and 4. Platelet Aggregation and ATP Release Assays-Platelet-rich plasma was obtained by plateletpheresis from medication-free platelet donors at the DTM/NIH blood bank. Aggregation and ATP release were performed as described (16) and in the supplemental data.
Platelet Adhesion Assay under Static Conditions-Inhibition of platelet adhesion to immobilized collagen was examined by fluorometry. Microfluor black microtiter 96-well plates (ThermoLabsystems, Franklin, MA) were coated with 2 g of fibrillar (Horm) or soluble collagen overnight at 4°C in PBS, pH 7.2, essentially as described (16) and in the supplemental data.
Contraction of Rat Aorta-Contraction of rat aortic ring preparations by U-46619 was measured isometrically and recorded with transducers from Harvard Apparatus Inc. (Holliston, MA). A modified Tyrode solution (with 10 mM HEPES buffer) that was oxygenated by continuous bubbling of air was used in the assays (17). In the first assay, aortic rings were suspended in a 0.5-ml bath kept at 30°C and were pre-constricted by 100 nM U-46619 before addition of proteins to give final concentrations of 1 M. In the second assay, aortic ring preparations were preincubated with 100 nM of DPTL, and increments of 100 nM U-46619 were added until maximum contraction was reached. Additions to the bath were never greater than 5% of the volume of the bath.
Contraction of Rat Uterus-Wistar female rats were injected intraperitonially with 0.1 mg of estradiol in 1 ml of phosphatebuffered saline. 24 h later, they were killed, and the uterus removed into a modified De Jalon solution (NaCl 154 mM, KCl 5.6 mM, D-glucose 2.8 mM, NaHCO 3 6 mM, CaCl 2 0.4 mM, Hepes 5 mM, 0.1 M dexamethasone, final pH 7.4). About 1.5-cm pieces of the uterus were attached to a 1-ml bath kept at 35°C, and their contractions recorded isotonically (Harvard Apparatus Inc.) under a 2 g load. Rhythmic contractions were induced by addition of PGF 2␣ at the indicated concentrations.
Human Dermal Microvascular Endothelial Cell (HMVEC) Culture-HMVEC (CC-2643) were purchased from Clonetics (San Diego, CA) and grown at 37°C, 5% CO 2 in T-25 flasks in the presence of EBM-2 Plus as described (17) and in the supplemental data.
Tube Formation Assay-Tube formation assay was done as described with modifications (18). Costar culture plates (96well; Corning, NY) were coated with 30 l of growth factorreduced Matrigel (BD Biosciences) and allowed to solidify at room temperature. 100 l of MVEC suspension (5 ϫ 10 5 /ml) were added to each well in the presence of vehicle, or 15(S)-HETE, or DPTL, or DPTL plus 15(S)-HETE at the concentrations indicated in the figure legends. Plates were incubated at 37°C, 5% CO 2 for 5-6 h, and formation was observed under an inverted microscope coupled to a digital camera (Axiovert 200; Carl Zeiss, Inc., Thornwood, NJ). Images were captured with AxioCamHR color camera (model 412-312) attached to the microscope. Tube length was measured by outlining the tubes (and converted to pixels) using AxioVision 4.6.3 software.
Modeling of DPTL and Pallidipin-Structures of DPTL and pallidipin (gi388359) were modeled using the alignment mode in SWISS-MODEL (19). The template structure was the ammonium complex of nitrophorin 2 (PDB accession number 1EUO).

RESULTS
Analysis of a Dipetalogaster maxima salivary gland cDNA library indicates that members of the lipocalin family of proteins is highly abundant, representing more than 90% of predicted secreted molecules (not shown). One of these sequences, DPTL, displays high sequence similarity to pallidipin, a lipocalin from Triatoma pallidipennis, which has been claimed as a specific inhibitor of collagen-induced platelet aggregation (20); however, the molecular target for pallidipin and its exact mechanism of action have remain elusive thus far. Fig. 1A shows a CLUSTAL alignment of DPTL and pallidipin in addition to other salivary antihemostatic proteins including RPAI-1 (7), triplatin (21), moubatin (22), and TSGP3 (23). DPTL was found to be more closely related to pallidipin, as depicted by clade I (Fig. 1B). The presence of mature DPTL in the salivary gland was supported by a one-dimensional gel of the gland homogenate, which displays intense staining at ϳ19 kDa mol wt, corresponding to salivary lipocalins. To identify the N terminus of DPTL, gels were electroblotted to PVDF membranes and bands submitted to Edman degradation. Fig. 1C shows the N terminus of the most abundant proteins. One of them was identified as an 18 kDa protein with N-terminal sequence KEcTLMAAaSNFNSDKYfDV (lowercase indicates ambiguous identification), which is in agreement with the corresponding cDNA coding for DPTL. The other sequence, GSISEcKTPKPMDDFSGTKF, was identified as procalinlike (24). These two sequences together represent at least 70% of the proteins loaded in the gel. Therefore, DPTL is a particularly abundant protein that is found in mature form in the salivary gland of D. maxima (Fig. 1C).
To study the effects of DPTL on platelet aggregation, its corresponding cDNA was cloned in pET17b vector, followed by transformation and expression of recombinant protein in BL21(DE3)pLysS cells. DPTL was purified through a series of chromatographic procedures and eluted as a single peak in the last step, performed in a gel-filtration column (Fig. 1D). PAGE of purified DPTL demonstrated that it migrates as a ϳ18 kDa protein under nonreducing conditions, and slightly higher mol wt in the presence of DTT, consistent with reduction of disulfide bridges. The N terminus was submitted to Edman degradation, and resulted in the sequence MxExTLMAAASNFNSD-KYFD (Fig. 1E). This sequence is identical to the N-terminal sequence predicted by the cDNA coding for DPTL with the exception of methionine in position 1, which has been added for expression purposes. In addition, MS results identified a mass of 17,948 Da for recombinant DPTL, which is in agreement with the theoretical mass of the molecule with an extra Met (17,951.1 Da) (Fig. 1F).
Results presented in Fig. 1 indicated that recombinant DPTL was pure, soluble, and suitable for further experimentation. In a first step toward studying its mechanism of action, DPTL was shown to dose-dependently inhibit low doses of collagen-induced platelet aggregation, with an IC 50 ϳ30 nM. Inhibition was abolished when high doses of collagen were employed. DPTL also blocked in a dose-dependent manner ATP release triggered by collagen (Fig. 2). Notably, no effect on shape change was observed, suggesting that DPTL did not target collagen itself, nor collagen receptors integrin ␣2␤1 or GPVI. This was confirmed through platelet adhesion assays carried out with calcein-labeled platelets incubated with immobilized soluble (integrin ␣2␤1-mediated) or fibrillar (GPVI and integrin ␣2␤1dependent) collagen (25)(26). Results reported in Table 1 show that adhesion of platelets to fibrillar or soluble collagen was not inhibited by DPTL (1-10 M). As a positive control, EDTA prevented platelet adhesion to fibrillar collagen ϳ60% and abolished adhesion to soluble collagen (26). These results, in addition to the lack of inhibition of platelet shape change, demonstrated that DPTL is not a specific collagen inhibitor and suggested that secondary mediators might be the target of the molecule. TXA 2 and ADP are two important mediators of platelet aggregation that are, respectively, generated and released by platelets upon stimulation by collagen (2). In an attempt to verify the inhibitory profile of DPTL toward other agonists that activate platelets independently of secondary mediators, it was tested as an inhibitor for U-46619 (TXA 2 mimetic) and AA-induced platelet aggregation. Fig. 2 shows that DPTL dose-dependently inhibits U-46619 and AA-induced platelet aggregation in a dose-dependent manner, corroborating the notion that DPTL targets TXA 2 (or ADP) mediated platelet responses. Because collagen, TXA 2 , and AAinduced aggregation is particularly sensitive to 5Ј-nucleotidases (27), ADP receptor antagonists (28), or ADP-binding proteins (7), it was of interest to exclude ADP as a potential target for DPTL. Fig. 2 shows that DPTL was ineffective as an inhibitor when ADP was employed at low or moderate concentrations, excluding this agonist as a target for the inhibitor. In addition, DPTL did not affect platelet aggregation triggered  by strong agonists such as PMA (PKC activator), convulxin (GPVI agonist), and ristocetin (vWF-dependent platelet agglutinator), which characteristically induce platelet aggregation/ agglutination independently of ADP or TXA 2 .
Whereas results suggested that DPTL targets secondary mediators of platelet aggregation, they did not formally identify TXA 2 as the (sole) ligand or establish whether other ligands involved in pro-hemostatic events unrelated to platelet function were targets for DPTL. Therefore, an experiment was optimized to broaden our search, where the inhibitor was incubated with small compounds (e.g. biogenic amines, prostaglandins and endoperoxides, leukotrienes, HETEs, epoxides and lipids) that may affect platelet function, vessel tonus, angiogenesis, or neutrophil function. The mixture was loaded into a gel-filtration column that excludes protein with mol wt higher than 20 kDa but retains small ligands. If DPTL binds to a given ligand(s), complex formation will occur and elute in the void (Ͼ20 kDa), while free, unbound ligands remain in the column. Accordingly, Fig. 3A shows a peak eluted at 20 min that represents DPTL and potentially bound ligands. This fraction was acidified to precipitate DPTL. The sample was centrifuged, and the supernatant containing ligands was applied to a RP-HPLC column followed by elution using a gradient of methanol. Each fraction was collected and submitted to reverse phase HPLC/ mass spectrometry that scanned for different masses compatible with the test mixture. A negative mass detection of 351 was found when fraction 20 was sprayed (Fig. 3C), this mass was compatible with PGD 2 or PGE 2 , which have masses of 352.4. As a control, fractions 17 (Fig. 3B) and 24 (Fig. 3D), which eluted before and after a peak corresponding to DPTL, were devoid of ligands.
Calorimetry results therefore indicated that DPTL binds TXA 2 , PGF 2␣ , or 15(S)-HETE among other ligands. This was also tested through additional pharmacologic assays. Fig. 5A shows that inhibition of collagen (5.2 g/ml)-induced platelet aggregation by DPTL was identical to inhibition by SQ 29548, a TXA 2 antagonist or by indomethacin which blocks TXA 2 production. Further, when DPTL was added to platelets incubated with SQ 29548 and indomethacin, no additional inhibition was observed. These results indicated a common target, i.e. TXA 2 pathway. Additionally, Fig. 5B shows that DPTL suppressed rhythmic contractions of the rat uterus induced by 0.2 M PGF 2␣ , being the inhibitory effect surmounted by high concentrations of the prostaglandin. Moreover, DPTL induces relaxation of aorta previously contracted with U-46619 (Fig. 5C).

Dipetalodipin
Collagen Finally, DPTL was found to inhibit by Ͼ85% tube formation evoked by 15(S)-HETE (18), suggesting that it could negatively modulate angiogenesis (Fig. 5D). Because DPTL is a lipocalin with sequence homology to nitrophorin 2 (NP2), a NO-binding protein from another triatomine species whose structure has been determined (13), we constructed a molecular model for the inhibitor using the NP2 structure as a template. Fig. 6A shows that DPTL displays structural features typical of the lipocalin family of proteins, whose structure consists of eight-stranded antiparallel ␤-barrel forming a central hydrophobic cavity; ligands are normally bound at a site located in the center of the ␤-barrel. Fig. 6B shows a comparison of this putative binding pocket in the models of DPTL and pallidipin. Many of the

TABLE 2 Affinity, stoichiometry, and thermodynamic parameters of DPTL interaction with different ligands
Interaction was detected by ITC as depicted in Fig. 4. residues predicted to lie in this pocket are shared by the two proteins and may play a role in stabilizing the bound ligand. Among these are a number of hydrophobic and aromatic residues that could be important in interactions with the hydrocarbon chain of eicosanoid ligands. Hydrogen bonding interactions with polar functional groups on the ligand are also normally essential, and their presence is suggested by the highly favorable enthalpies measured in ITC experi-ments. A number of residues that could potentially form hydrogen bonding or ionic interactions with bound ligands are also present in the binding pocket. Most notable among these are the conserved residues Arg 39 and Gln 135 in both (mature) DPTL and pallidipin. Remarkably high sequence similarity for the full-length sequences of both lipocalins is also depicted by the CLUSTAL alignment presented in Fig. 6C.

DISCUSSION
Salivary secretions are rich sources of bioactive molecules that counteract host defenses in distinct ways. Many of these molecules have turned out to display unique and specific pharmacologic properties that in several instances contribute to our understanding of vertebrate biology (4). In this report, we have identified DPTL as a novel salivary lipocalin that binds to distinct prostanoids. Accordingly, ITC experiments demonstrated that DPTL binds to cTXA 2 , U-46619 (stable TXA 2 mimetic), and TXB 2 which is the metabolic end product of TXA 2 . Binding occurred with a K D ϳ100 -200 nM and was compatible with 1:1 stoichiometry. Binding to TXA 2 was consistent with the inhibitory profile for platelet aggregation observed in the presence of the inhibitor and suggests that native TXA 2 generated by platelets is a target for the inhibitor. Accordingly, DPTL affected only platelet responses induced by low concentrations of collagen or by AA and U-46619, which are TXA 2 dependent. A similar inhibitory profile is observed for aggregation of platelets from a patient with a mutation of the TXA 2 receptor (29), or for mice with a gene deletion of Gq (30). Targeting of the TXA 2 pathway by DPTL is also corroborated by a lack of additional effects of the inhibitor on collagen-induced platelet aggregation in the presence of SQ29548 and indomethacin (Fig. 5A). In contrast, DPTL did not interfere with aggregation triggered by high doses of collagen, which occurs via tyrosine kinase-dependent PLC␥2 activation (2,31), or by strong agonists such as convulxin and PMA or vWF-dependent agglutinating agent ristocetin.
Targeting TXA 2 has several implications, as it is the major contributor of platelet aggregation by collagen, which is the most atherogenic protein of the vessel wall (32). Upon platelet adhesion to collagen, TXA 2 is generated and activates platelets through the TP receptors that are coupled to G q and G 12/13 (33). This promotes shape change, and activation of an intracellular pathway that leads to granule secretion, and ADP release. ADP is critical for completion of platelet aggregation by TXA 2 through binding to P2Y12-coupled G i activation and decrease of intraplatelet cAMP. This explains the sensitivity of TXA 2induced platelet aggregation to apyrases (2), ADP receptor antagonists (28), and ADP-binding proteins such as RPAI-1 (7). TXA 2 (and ADP) are also critical for collagen-induced platelet aggregation, because they induce integrin ␣2␤1 activation, a step necessary for firm platelet adhesion to collagen (2,34). Finally, TXA 2 (and ADP) promote integrin ␣ IIb ␤ 3 activation, leading to fibrinogen binding to activated platelets, and are necessary for recruitment of platelets to the site of injury (35). Therefore, targeting TXA 2 increases the threshold for platelet aggregation by a number of agonists, and appears to be an effective strategy developed by a hematophagous arthropod to attenuate platelet function and associated inflammation (36 -37).
Whereas platelet aggregation assays suggested that DPTL targets TXA 2 , it was possible that the inhibitor interacted with other compounds not necessarily involved with platelet function. Therefore, we broadened our search for ligands involved in pro-hemostatic events other than platelet activation. This was carried out through experiments that combined chroma-tography of DPTL added to a mixture of eicosanoids, biogenic amines, endoperoxides, and lipids, followed by mass spectrometry of potentially bound molecules. Results indicated that a molecule compatible with the mass of PGD 2 bound to DPTL; formal identification of the compound was subsequently attained by ITC. Results demonstrated that DPTL interacts not only with TXA 2 mimetic U-46619, but also displays high affinity to other prostanoids such as PGF 2␣, which is a potent endothelium-dependent vasoconstrictor (38). Consistent with this specificity, DPTL induced relaxation of PGF 2␣ -dependent uterus contraction. Because DPTL also interacts with PGH 2 , an endoperoxide that behaves as a vasoconstrictor (39), and also converts spontaneously to PGF 2␣ , it is evident that DPTL behaves as a potential vasodilator increasing blood flow at sites of feeding. Our results also demonstrate high-affinity interaction of DPTL to PGD 2 , a major mediator of mast cells and thought to be involved in allergic reactions and immune responses (40 -41). Furthermore, DPTL binds to 15(S)-HETE and is the first lipocalin reported to date that targets a bioactive component generated by the 15-LOX pathway. This interaction may be relevant, taking into account the role of 15(S)-HETE in monocyte-endothelial cell interaction (42), ICAM-1 expression by endothelium cells (43), and other pro-inflammatory events (44). 15(S)-HETE also modulates angiogenesis (18), and this activity was inhibited by DPTL, evidenced by reduced endothelial tube formation over growth factor-reduced Matrigel. Interference with angiogenesis may potentially attenuate inflammation and granulation tissue formation at site of bite, as was reported previously for tick saliva (17).
These results suggest that DPTL potentially displays multifunctional antihemostatic properties, through attenuation of platelet aggregation mediated by TXA 2 , negative modulation of vessel tonus by PGF 2␣, and blockade of angiogenesis by 15(S)-HETE (Fig. 5). It is important to recognize that DPTL is a very abundant lipocalin in the salivary gland, accounting for at least 30% of total salivary lipocalins estimated by SDS/PAGE (Fig.  1C). Assuming a molecular mass of ϳ20 kDa for DPTL and release of 50% of the salivary contents (ϳ1 g/salivary gland pair) upon feeding being ϳ30% DPTL, a concentration of at least 1 M of the inhibitor could exist in the feeding environment (ϳ15 l); this concentration is clearly above the K D for TXA 2 and other prostanoids. In addition, D. maxima expresses an apyrase and several other uncharacterized salivary proteins that may interfere with hemostasis in a distinct yet redundant manner.
Results obtained by pharmacologic assays were compatible with DPTL being a lipocalin that evolved with a binding pocket adapted to accommodate small eicosanoids. To get further insights into the mechanism of binding of DPTL to TXA 2 , a molecular model based on NP2 was constructed (13). The putative binding pocket of the model exhibits a generally hydrophobic structure that would be consistent with the binding of eicosanoids. Also present are residues that could potentially form hydrogen bonds with polar groups on the ligand. Many of these residues are conserved in the pallidipin, which has been claimed as a specific inhibitor of collagen-mediated platelet aggregation, suggesting that the two proteins may have similar functions; however, it is clear that DPTL (and pallidipin) are not specific collagenbinding proteins or receptor antagonists because they do not affect platelet shape change or platelet adhesion. Accordingly, it is plausible that pallidipin exerts its antiplatelet activity through a similar mechanism characterized here for DPTL as binding to TXA 2 ; interaction with other eicosanoids is also likely. In conclusion, DPTL displays unique ligand specificities that may assist the triatomine D. maxima to successfully feed on blood. DPTL is a potentially useful tool to understand the contributions of TXA 2 , PGF 2␣ , and 15(S)-HETE in different pathways leading to platelet aggregation, vasoconstriction, angiogenesis, and oxidative stress in health or disease (49 -50).