Salivary Antigen-5/CAP Family Members Are Cu2+-dependent Antioxidant Enzymes That Scavenge O2⨪ and Inhibit Collagen-induced Platelet Aggregation and Neutrophil Oxidative Burst*

Background: The function of most salivary antigen-5/CAP members has remained elusive for decades. Results: Antigen-5 members bind Cu2+ and exhibit antioxidant activity by scavenging O2⨪. It inhibits platelet aggregation by collagen and neutrophil oxidative burst. Conclusion: Antigen-5 emerges as a novel family of antioxidant enzymes targeting O2⨪. Significance: Scavenging O2⨪ is conceivably associated with decreased inflammation in the microcirculation and may assist blood-sucking insects to successfully feed on blood. The function of the antigen-5/CAP family of proteins found in the salivary gland of bloodsucking animals has remained elusive for decades. Antigen-5 members from the hematophagous insects Dipetalogaster maxima (DMAV) and Triatoma infestans (TIAV) were expressed and discovered to attenuate platelet aggregation, ATP secretion, and thromboxane A2 generation by low doses of collagen (<1 μg/ml) but no other agonists. DMAV did not interact with collagen, glycoprotein VI, or integrin α2β1. This inhibitory profile resembles the effects of antioxidants Cu,Zn-superoxide dismutase (Cu,Zn-SOD) in platelet function. Accordingly, DMAV was found to inhibit cytochrome c reduction by O2⨪ generated by the xanthine/xanthine oxidase, implying that it exhibits antioxidant activity. Moreover, our results demonstrate that DMAV blunts the luminescence signal of O2⨪ generated by phorbol 12-myristate 13-acetate-stimulated neutrophils. Mechanistically, inductively coupled plasma mass spectrometry and fluorescence spectroscopy revealed that DMAV, like Cu,Zn-SOD, interacts with Cu2+, which provides redox potential for catalytic removal of O2⨪. Notably, surface plasmon resonance experiments (BIAcore) determined that DMAV binds sulfated glycosaminoglycans (e.g. heparin, KD ∼100 nmol/liter), as reported for extracellular SOD. Finally, fractions of the salivary gland of D. maxima with native DMAV contain Cu2+ and display metal-dependent antioxidant properties. Antigen-5/CAP emerges as novel family of Cu2+-dependent antioxidant enzymes that inhibit neutrophil oxidative burst and negatively modulate platelet aggregation by a unique salivary mechanism.

implying that it exhibits antioxidant activity. Moreover, our results demonstrate that DMAV blunts the luminescence signal of O 2 . generated by phorbol 12-myristate 13-acetate-stimulated neutrophils. Mechanistically, inductively coupled plasma mass spectrometry and fluorescence spectroscopy revealed that DMAV, like Cu,Zn-SOD, interacts with Cu 2؉ , which provides redox potential for catalytic removal of O 2 . . Notably, surface plasmon resonance experiments (BIAcore) determined that DMAV binds sulfated glycosaminoglycans (e.g. heparin, K D ϳ100 nmol/liter), as reported for extracellular SOD. Finally, fractions of the salivary gland of D. maxima with native DMAV contain Cu 2؉ and display metal-dependent antioxidant properties. Antigen-5/CAP emerges as novel family of Cu 2؉ -dependent antioxidant enzymes that inhibit neutrophil oxidative burst and negatively modulate platelet aggregation by a unique salivary mechanism.
The presence of anti-hemostatics in the salivary glands (SGs) 3 of hematophagous animals is an evolutionary strategy aimed to decrease the inflammatory tonus that emerges at sites of insect bite (1). These responses comprise a series of complex biochemical reactions that ultimately prevent blood loss. In this respect, several anticoagulants, platelet inhibitors, and vasodilators have been identified in most salivary secretions studied to date (1). Notably, a remarkable diversity of mechanisms accounts for the anti-hemostatic properties of saliva. For example, platelet aggregation is blocked by apyrases or ADP-, TXA 2 -, or serotonin-binding proteins; collagen inhibitors; integrin ␣ 2 ␤ 1 or ␣ IIb ␤ 3 antagonists; cAMP or cGMP intraplatelet-in-creasing molecules (PGI 2 , adenosine, and NO); anti-thrombins; and fibrin(ogen)olytic enzymes (2).
Although the anti-hemostatic functions of several molecules in SGs are mediated by well known family of proteins, such as serpins, lipocalins, apyrases, and kazal and kunitz domain-containing molecules, several other inhibitors of platelet aggregation or coagulation cascade have not been characterized molecularly. More recently, transcriptome and proteome analysis (sialomes) of numerous SGs from hematophagous sources revealed that this tissue is a source of an unprecedented number of uniquely expressed proteins (e.g. 30-kDa allergens, yellow proteins, and antigen-5) (1,3). Most often, the biological functions for members of these families have not been assigned. Notably, antigen-5 members, which belong to the CAP superfamily (cysteine-rich secretory proteins, antigen 5, and pathogenesis-related 1 proteins) of proteins, is inarguably one of the most, if not the most, ubiquitous family of salivary proteins, together with apyrases (3). The function of this family has remained elusive despite its presence in mosquitoes, sand flies, bugs, horse flies, Culicidae, Simulidae, ticks, and venomous animals (1). Only one salivary antigen-5 member has been unambiguously identified as an RGD-containing disintegrin by incorporation of a typical RGD triad flanked by cysteines not found in other antigen-5 members (4,5).
In our attempt to identify the function of this family of proteins, we discovered that antigen-5 members from the SG of the Triatomine vectors of Chagas disease, Dipetalogaster maxima (DMAV, D. maxima antigen-5) (6) and Triatoma infestans (TIAV, T. infestans antigen-5) (7) were inhibitors of low doses of collagen-induced platelet aggregation. DMAV did not interfere with aggregation triggered by other agonists or block the proximal interaction of collagen with platelets. This inhibitory profile resembles the inhibitory effects of antioxidants in platelet aggregation induced by collagen, which is the only agonistpromoting extracellular increase of O 2 . (8 -11), in addition to intracellular reactive oxygen species (ROS). For example, blockade of extracellular O 2 . by N-acetylcysteine is accompanied by inhibition of low doses of collagen-induced platelet aggregation (12). In addition, Zn,Cu-SOD prevents platelet recruitment induced by collagen in vitro (10). While the interplay of collagen and ROS has long been recognized (13)(14)(15)(16)(17)(18)(19)(20)(21)(22)(23), only more recently has it been explained molecularly (8,9). Collagen interacts with GPVI, which signals by tyrosyl phosphorylation and NADPH oxidase activation through TRAF4 (24). These steps result in the generation of O 2 . that is critical to prime the cyclo-oxygenase enzyme for kinase activation (13,24). Accordingly, intracellular O 2 . is important for collagen-induced platelet activation. NADPH oxidase inhibitor apocynin (12) or intracellular antioxidants, such as MnTMPyP, inhibit intracellular ROS formation by collagen and attenuate TXB 2 production and platelet aggregation (20). Furthermore, quercetin, which displays antioxidant properties, attenuates platelet aggregation by collagen (25). Mechanistically, it was found that DMAV binds Cu 2ϩ and catalytically removes O 2 . . Therefore, this report is the first to determine the function of the antigen-5 family of salivary proteins as antioxidant enzymes. This discovery is important, given the wide expression of antigen-5 family members in bloodsuck-ing animals, venomous creatures, and vertebrates (26). It also highlights the importance of scavenging O 2 . to attenuate the proinflammatory events associated with dysregulated production of ROS in the microcirculation.
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 DMAV (gi 344190589) at 280 nm was obtained using the ProtParam Web site, yielding a value of ⑀ 280 nm ϭ 58,705 mol liter Ϫ1 ⅐cm Ϫ1 A 0.1% 280 nm /cm (1 mg/ml) ϭ 1.217, predicted molar mass 25,341 Da (224 amino acids), and isoelectric point (pI) 8.57.
Clustal Alignment and Phylogenetic Tree Construction-A sequence similarity search to DMAV was performed using protein BLAST against Arthropoda organisms with a low complexity region filter. The sequences derived from the nonredundant protein database of the National Center for Biotechnology Information (NCBI) are represented by five letters followed by the NCBI gi accession number. The five letters derive from the first three letters of the genus and the first two letters from the species name. The protein sequences were aligned by the ClustalX software program. The phylogenetic tree was created with the Mega package after 10,000 bootstraps with the neighbor-joining algorithm, using the p-dis-tance model as described (6,29). The bar at the bottom of each tree represents 10% amino acid substitution.
Expression of DMAV and TIAV in Escherichia coli-All percentage concentrations are expressed as the volume fraction of the chemical in water. Synthetic cDNA for DMAV (gi 344190589) was produced by BioBasic Inc. (Markham, Canada). The sequence displays an N-terminal NdeI and a C-terminal XhoI restriction site. 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 a stop codon. pET17b constructs were confirmed by sequencing before transformation by heat shock of E. coli strain BL21(DE3)pLysS cells. Cells were grown overnight in LB ampicillin/chloramphenicol agar plates, at 37°C. Colonies were picked and then used to generate glycerol stocks (25% volume fraction). For recombinant protein production, a protocol similar to that described was employed (30). Briefly, 30 ml of Luria-Bertani broth (with 34 g/ml chloramphenicol and 100 g/ml ampicillin) was inoculated and grown overnight for a maximum of 16 h at 37°C. Luria-Bertani broth (1 liter with 34 g/ml chloramphenicol and 100 g/ml ampicillin) was inoculated with 10 ml of the overnight culture and grown at 37°C with continuous shaking (26 rads/s) until an optical density of 0.6 -0.8 (A 600 nm ) was reached (ϳ3 h). Isopropyl-1-thio-␤-D-galactopypyranoside (1 mmol/liter) was added to induce expression, and the flask was shaken for 3 h under the same conditions; cells were harvested by centrifugation (6,000 ϫ g n ; 20 min) and washed once in 20 mmol/liter Tris-HCl, pH 8.0, before the cell pellet was frozen and stored until use. Pellets were sonicated several times, washed in 20 mmol/liter Tris, pH 8.3, centrifuged, and finally resuspended in 25 ml of 20 mmol/liter Tris, pH 8.3, containing 6 M guanidine and DTT (10 mmol/liter). After 1 h, the sample was centrifuged at 16,000 ϫ g for 30 min. Twenty-five ml was added slowly, drop-by-drop, into 4 liters of refolding buffer (20 mmol/liter Tris, 0.3 mol/liter arginine, 1 mmol/liter EDTA, 1 mmol/liter reduced glutathione (GSH), and 0.2 mmol/liter oxidized glutathione (GSSG), pH 9.3) under agitation. After 2 days under agitation at 4°C, the solution containing DMAV was concentrated to ϳ300 ml by tangential flow with a 10-kDa cut-off filter (Millipore, Inc., Billerica, MA). Sample was centrifuged (12,000 ϫ g n for 20 min) and concentrated to 30 ml in a Millipore ultrafiltration apparatus (membrane of 10 kDa). The sample was centrifuged (12,000 ϫ g n for 20 min), and the supernatant was used for protein purification (see below). Antigen-5 members from T. infestans (gi 149898844) were cloned in pET17b and expressed as above except that refolding was performed at pH 7.4. As a control, a lipocalin from D. maxima (DM146, gi 315270899) was cloned, expressed, and purified as above.
Protein Purification-DMAV (10 ml) was loaded into a HiPrep 16/60 Sephacryl S-100 HR (GE Healthcare) column connected to an AKTA purifier system (Amersham Biosciences) and equilibrated in 20 mmol/liter Tris-HCl, 0.15 mol/ liter NaCl, pH 9.3, with a flow of 1 ml/min. Protein was detected by peak absorbance at 280 and 220 nm, and fractions containing DMAV (estimated by SDS-PAGE) were combined (4 ml) and concentrated in a Centricon concentrator (10 kDa cut-off). Samples were loaded in a gel filtration column (Superdex 75 HR10/30, Amersham Biosciences) equilibrated in 20 mmol/liter Tris-HCl, 0.5 mol/liter NaCl, pH 9.3. Elution was carried out at a flow rate of 1 ml/min, and fractions containing DMAV were combined. Fractions were dialyzed against 20 mmol/liter Tris-HCl and loaded in a Mono Q anion exchange column (Amersham Biosciences). To remove bacterial DNA potentially contaminating DMAV, the fractions containing DMAV were combined and dialyzed against 10 mmol/liter HEPES, pH 6.0, centrifuged, and loaded into cation exchange chromatography using a HiPrep 16/10 SP FF column (Amersham Biosciences). Proteins were eluted with a linear gradient of NaCl (0 -1 mol/ liter) over 60 min at a flow rate of 0.5 ml/min. Fractions containing DMAV were combined and dialyzed against PBS, pH 7.4. DMAV was also purified in a reversed-phase C-18 chromatography column (Vydac, Carpenteria, CA), with a 60-min linear gradient of acetonitrile and 0.1% TFA at a flow rate of 1 ml/min. PAGE and Edman degradation were performed as described (30).
Mass Spectrometry of Purified DMAV-DMAV molecular weight was determined with a Sciex 4000 Q Trap ESI mass spectrometer.
Circular Dichroism-To examine secondary structure, samples were diluted in 1ϫ PBS. CD spectra were recorded over the wavelength range 190 -260 nm in a 1-mm path length quartz cuvette using a step size of 0.2 nm, a slit bandwidth of 1.0 nm, and a signal averaging time of 1.0 s using a Jasco J-815 circular dichroism spectrometer. Percentages of secondary structures were determined as described (31,32).
SEC-MALS-QELS-HPLC-To assess the identity, purity, and solution aggregation state of purified DMAV (17 mol/liter), analytical size exclusion chromatography (SEC) with in-line multiangle light scattering (MALS) and quasi-elastic light scatter detection (QELS) was performed. Chromatography was done on an Alliance HPLC system (Waters Corp., Milford MA) connected to a multiangle DAWN EOS MALS detector and a QELS detector (Wyatt Technology, Santa Barbara, CA).
Spectrum of DMAV-Spectrum of DMAV (500 l at 5 mol/ liter in TBS, pH 7.4) was analyzed by UV-visible spectroscopy wavelength scanning (Cary 100 BIO spectrophotometer; Varian Inc., Walnut Creek, CA) using the eluent buffer (TBS) as a blank. Aliquots of CuCl 2 (in TBS, pH 7.4) were added in 10-l increments, reaching a maximum concentration of 100 mol/ liter CuSO 4 .
Determination of Free Cysteine in DMAV-Reactions for thiol modification were done at room temperature in 100 l of 20 mmol/liter Tris-HCl, pH 8.0, 1 mmol/liter 4-fluoro-7-aminosulfonylbenzofurazan, 20% DMSO for 1 h. The reactions were read in a Tecan M200 plate reader at excitation and emission wavelengths of 386 and 516 nm, respectively. Reduction of DMAV was done with 10 mmol/liter TCEP at 40°C for 20 min. Non-reducing protein denaturation was done in 0.1% volume fraction SDS at 60°C for 30 min. In some experiments, DMAV was incubated with buffer or xanthine/xanthine oxidase (X/XO) followed by mass spectrometry (MS).
Anti-DMAV Polyclonal Antibody-Recombinant DMAV produced in E. coli was diluted to 1 mg/ml and injected in the presence of Freund's adjuvant in rabbits at Spring Valley Laboratories (Woodbine, MD). Additional injections were per-formed 3, 6, and 9 weeks later to boost the immune response. Before injection of DMAV, preimmune serum was obtained from all rabbits.
ELISA to Detect DMAV-Antibody titer was estimated by ELISA using bacterial expressed DMAV (0.5 g/well) immobilized overnight in PBS at 4°C. Wells were blocked with 200 l of TBS-BSA (BSA (2%, w/v) in TBS) for 2 h and then incubated for 1 h with a serial dilution of anti-DMAV antibody in TBS-BSA-Tween (BSA, 1% volume fraction; Tween, 0.05% volume fraction in TBS). After washing four times with TBS-BSA-Tween, the wells were incubated with a 1:2,000 dilution of anti-rabbit alkaline phosphatase-coupled antibody for 1 h and washed four times with TBS-BSA-Tween. Two hundred l of substrate (p-nitrophenyl phosphate liquid substrate system; Sigma) was added, and reactions were followed for 10 min and stopped with 100 l of NaOH (0.375 mol/liter, final concentration). Absorbance was detected at 405 nm.
D. maxima Salivary Gland Purification-For some experiments, the supernatant from the homogenate obtained from 50 pairs of D. maxima (reared at the Institute of Parasitology, Biology Centre, Ceske Budejovice, Czech Republic) was applied to a Sephadex G-75 column, equilibrated with PBS, pH 7.4, at a flow rate of 0.5 ml/min.
Mass Spectrometry and Protein Identification-Samples from selected fractions obtained above were solubilized in NU/PAGE sample buffer containing DTT and loaded into a 4 -12% NuPAGE gel with MES buffer (Invitrogen). The gel was cut in 20 slices of ϳ0.5 cm (from top to bottom of each lane), and identification was performed on reduced and alkylated, trypsin-digested samples prepared by standard MS protocols. The supernatant and two washes (5% formic acid in 50% acetonitrile) of the gel digests were pooled and concentrated by SpeedVac (Labconco, Kansas City, MO) to dryness directly in 200 l of polypropylene autosampler vials (Sun Sri, Rockwood, TN). Recovered peptides were resuspended in 5 l of solvent A (0.1% formic acid, 2% acetonitrile, and 97.9% water).
Prior to MS analysis, the resuspended peptides were chromatographed directly on the column without trap clean-up. Bound peptides were separated at 500 nl/min, generating 80 -120 bar pressure using an AQ C18 reversed-phase medium (3-m particle size and 200-m pore) packed in a pulled-tip, nanochromatography column (0.100-mm inner diameter ϫ 150-mm length; Precision Capillary Columns (San Clemente, CA)). Chromatography was performed in-line with an LTQ-Velos Orbitrap mass spectrometer (Thermo Fisher Scientific), and the mobile phase consisted of a linear gradient prepared from solvent A and solvent B (0.1% formic acid, 2% water, and 97.9% acetonitrile) at room temperature. Nano-LC-MS (LC-MS/MS) was performed with a ProXeon Easy-nLC II multidimensional liquid chromatograph and temperature-controlled Ion Max nanospray source (Thermo Fisher Scientific) in-line with the LTQ-Velos Orbitrap mass spectrometer.
Computer-controlled data-dependent automated switching to MS/MS by Xcalibur version 2.1 software was used for data acquisition and provided the peptide sequence information. Data processing and data bank searching were performed with Proteome Discoverer version 1.2 and SEQUEST (Thermo Scientific). The data were searched against protein sequences in the database generated with the salivary gland of D. maxima (6). Parsimony analysis was performed using ProteoIQ software (NuSep, Athens, GA), and protein probabilities were calculated using the ProteinProphet algorithm. All proteins were required to have at least two peptides and two spectra per peptide and a probability of 95% in at least one gel lane. MS/MS results were mapped into the Excel spreadsheet with indication of the slices containing the most abundant ions in sorted order, among other information.
Platelet Preparation, Aggregation, ATP Secretion, and TXA 2 Generation-Platelet-rich plasma was obtained under informed consent from medication-free platelet donors participating in a National Institutes of Health Institutional Review Board-approved protocol of the Department of Transfusion Medicine (DTM/NIH, Blood Bank). Platelet-rich plasma (100 l) was added to 200 l of Tyrode's buffer (137 mmol/liter NaCl, 27 mmol/liter KCl, 12 mmol/liter NaHCO 3 , 0.42 mmol/ liter NaH 2 PO 4 , 1 mmol/liter MgCl 2 , 5.55 mmol/liter Hepes, 0.25% bovine serum albumin, pH 7.4) giving a final concentration of 200,000 -400,000 platelets/l. Platelet aggregation was estimated using a Lumi-Aggregometer (Chrono-Log Corp.) under agitation as reported (30). In experiments using PGG 2 , platelets were washed and resuspended in Tyrode's buffer supplemented with gelatin (0.25% volume fraction). PAF was prepared as described (30). To estimate dense granule ATP release, peak luminescence was detected by adding luciferinluciferase (20 l) 4 min after the addition of collagen. For TXA 2 generation, platelets were activated by collagen, and after 4 min, the sample was transferred to an Eppendorf tube and centrifuged for 15 s at 14,000 ϫ g n . The supernatant was immediately frozen in dry ice, and then at Ϫ80°C. The next day, the sample was thawed and used at a 1:300 dilution for detection of TXB 2 by competitive ELISA (Cayman).
Platelet Adhesion-For platelet adhesion assays, platelet-rich plasma was incubated with calcein-AM (2 mol/liter, Calbiochem) for 30 min at room temperature and centrifuged in the presence of EDTA (5 mmol/liter) and apyrase (0.2 units/ml). Platelets were resuspended (2 ϫ 10 5 /l) in Tyrode's buffer (no additions). Inhibition of platelet adhesion to immobilized collagen was examined by fluorometry. Microfluor black microtiter 96-well plates (ThermoLabsystems, Franklin, MA) were coated with 50 l (1 g/well in PBS) of fibrillar (Horm) or soluble collagen I (BD Biosciences) overnight at 4°C in PBS, pH 7.4. Wells were washed with PBS and blocked with 2% (w/v) BSA (in PBS). Calcein-labeled platelets (2 ϫ 10 5 /l) were incubated with DMAV or EDTA for 20 min, and 50 l of platelets was added to each well. After 1 h, wells were washed three or four times with Tyrode-BSA, and adhesion was estimated by fluorescence (excitation 485 nm/emission 530 nm).
Thrombus Formation-Thrombus formation under flow was performed essentially as described (4).
Neutrophil Preparation-Polyprep medium (6.5 ml; MP Biomedicals, Solon, OH) was added to a 15-ml Falcon tube. Blood (7 ml, collected in 5 mmol/liter EDTA, final concentration) was carefully layered on the Polyprep medium. The tubes were centrifuged at 209 rads/s (863 ϫ g n ) for 45 min using acceleration and deceleration set at 5, at room temperature. Two rings were observed; the upper ring was discarded, and the lower ring was collected and transferred to a 50-ml Falcon tube. Then HBSS (Cellgro) (no calcium, no magnesium; 40 ml) was added, and cells were centrifuged again for 10 min at 104 rads/s. When contamination with red blood cells was visible, 1 ml of lysis buffer was added to the tubes for 40 -50 s, followed by the addition of HBSS (45 ml) and centrifugation as above. The pellet was resuspended in 3 ml and counted, and concentration was adjusted to 2 ϫ 10 6 /ml in HBSS (no Ca 2ϩ , no Mg 2ϩ ). Cells were kept on ice. . Diogenes reagent was diluted as indicated by the manufacturer, whereas lucigenin (40 mmol/liter) and luminol (500 mmol/liter) were diluted in DMSO followed by extensive vortexing.

Xanthine Oxidase Generation of O 2 . Estimated by Luminol or
Lucigenin-Fifty l of reaction mixture (RM) (100 mmol/liter phosphate buffer, pH 7.8, 500 mol/liter hypoxanthine, and 500 mol/liter luminol; final concentrations) was added to the wells, followed by the addition of DMAV (0 -10 l). Reactions started with xanthine oxidase (50 l, 1:2,500, or 1:5,000 final concentrations; ϳ5 milliunits/ml) using a multichannel pipette. Plates were immediately read using an LMAX II 384 Luminometer (Molecular Devices), and integrations of the signal were obtained with Softmax Pro software. Results are expressed as luminescence estimated under the curve. Uric acid formation was estimated by the increase in UV absorbance at 293 nm. Phosphate buffer was prepared by mixing 90 mmol/ liter Na 2 HPO 4 and 10 mmol/liter NaH 2 PO 4 . Reduction of Cytochrome c (cyt c)-In a 96-well flat microplate, 100 l of RM (0.1 mol/liter phosphate buffer, 0.5 mmol/ liter hypoxanthine, and 12.0 M (0.15 mg/ml) cyt c, final concentrations) was added to the wells, followed by the addition of DMAV (0 -3 l). Reactions started by the addition of 10 l of xanthine oxidase (1:2,500; ϳ5 milliunits/ml, final concentration). The reduction of cyt c was continuously detected at 550 nm using a Thermomax ELISA reader (Molecular Devices). For kinetic experiments, DMAV (10 mol/liter) was incubated with saturating concentrations (20 mol/liter) of Cu 2ϩ or Zn 2ϩ , followed by extensive dialysis against PBS. To calculate the K cat of DMAV for O 2 . , data were transformed as described (33). Measurement of H 2 O 2 -Measurement of H 2 O 2 was performed using Amplex Red reagent (Invitrogen) as described in the manufacturer's instructions. Briefly, an RM was prepared containing 2.35 ml of Krebs-Ringer phosphate (145 mmol/liter NaCl, 5.7 mmol/liter sodium phosphate, 4.86 mmol/liter KCl, 0.54 mmol/liter CaCl 2 , 1.22 mmol/liter MgSO 4 , 5.5 mmol/liter glucose, pH 7.35), 50 l of Amplex Red (50 mol/liter), and 100 l of horseradish peroxidase (HRP; 0.1 unit/ml). Then 50 l of RM was added to 50 l of neutrophils (0.5 ϫ 10 6 /ml, final concentrations), followed by the addition of PMA (20 nmol/liter to 40 nmol/liter). When a cell-free system was used, 5 l of hypoxanthine (500 mol/liter final concentrations) was added to 50 l of RM (which was diluted in HBSS), with and without DMAV. Reactions were started by the addition of 50 l of XO (6.4 milliunits/ml to 12.8 millunits/ml, final concentrations). As a control, Cu,Zn-SOD (0.4 milliunit/ml, final concentration) was tested. Reactions were followed with an excitation of 540 nm and emission of 590 nm using a SpectraMAX Gemini XPS Fluorimeter (Molecular Devices).
Measurement of O 2 Consumption-O 2 consumption was monitored at 25°C using a Gilson 5/6 oxygraph fitted with a Clark electrode as described (33,34). The electrode was calibrated for absolute O 2 concentration. Reactants were injected using a long tip into the reaction chamber under continuous agitation with a stir bar placed in the base of the chamber. The RM, in a final volume of 1.5 ml, typically contained 100 mmol/ liter phosphate buffer, pH 7.8, 0.5 mmol/liter hypoxanthine, and 0.7 milliunit/ml xanthine oxidase. Consumption of O 2 was characterized by a deflection of the recorder. In some experiments, catalase (333 units/ml) or DMAV (0.5 M) was added to the chamber.
Inductively Coupled Plasma Mass Spectrometry (ICP-MS)-Aqueous samples from DMAV obtained directly from the last purification step (cation exchange column) were extensively dialyzed (cut-off of 10 kDa, Slide-A-Lyzer, Thermo Scientific) against 5 mM Hepes, pH 7.4, and transferred into a 15-ml autosampler tube. The mass of the sample was weighed on a four-place analytical balance by difference. An aliquot of 5 g of diluent consisted of 1.5% volume fraction. Optima grade nitric acid (HNO 3 ; Fisher) in water was weighed into the tube. The diluted sample was thoroughly mixed by gentle shaking. The contents of the samples were analyzed using the SemiQuant mode of Agilent (Santa Clara, CA) model 7500cs inductively coupled plasma mass spectrometer. The sample introduction system of the mass spectrometer consisted of a PFA microflow (0.1 ml/min) concentric nebulizer, a PFA Scott type double-pass spray chamber, and a sapphire injector. The SemiQuant mode is capable of quantifying elements of the entire periodic table by using the spectral information of the elements. The instrument was calibrated using a solution containing 20 g kg Ϫ1 each of 31 elements (lithium, beryllium, boron, sodium, magnesium, aluminum, calcium, scandium, vanadium, chromium, manganese, cobalt, nickel, copper, zinc, arsenic, selenium, strontium, molybdenum, silver, cadmium, antimony, barium, lanthanum, europium, holmium, ytterbium, thalium, lead, thorium, and uranium) prepared by diluting catalog no. ICP-MSCS ICP-MS calibration standard (High Purity Standards, Charleston, SC) with 1.5% HNO 3 . National Institute of Standards and Technology Standard Reference Material (SRM) 1643e, Trace Elements in Water, was measured with the samples. SRM 1643e served as the quality assurance for the measurement (35).
Reconstitution of DMAV with Divalent Ions and Detection by ICP-MS-Purified DMAV (8 ml, 10 mol/liter, in PBS) was initially incubated with EDTA (20 mmol/liter) for 30 min and dialyzed against 2 liters of PBS, pH 7.4 (eight changes over 4 days). Then 1 ml of DMAV (10 mol/liter) was incubated sep-arately with 30 mol/liter each of Cu 2ϩ , Mn 2ϩ , Ni 2ϩ , Co 2ϩ , and Zn 2ϩ for 40 min, followed by extensive dialysis against 2 liters of Hepes (2 mmol/liter), pH 7.4 (eight changes over 3 days). In the displacement assays, DMAV (10 M) was incubated first with a mixture containing 15 mol/liter each of Mn 2ϩ , Ni 2ϩ , Co 2ϩ , and Zn 2ϩ for 40 min, followed by the addition of 15 mol/liter Cu 2ϩ and dialysis as above. In the competition assays, DMAV (10 mol/liter) was incubated with a mixture of 15 mol/liter each of Mn 2ϩ , Ni 2ϩ , Co 2ϩ , Zn 2ϩ , and Cu 2ϩ for 40 min and dialyzed against 5 mM Hepes, pH 7.4. One ml of each ion-DMAV complex was used for ICP-MS.
Fluorescence Spectroscopy-Spectroscopy was carried out in a Jasco J-815 fluorospectrophotometer. Tryptophan fluorescence emissions were collected following excitation at 278 nm at a DMAV concentration of 20 mol/liter in 1 ml of TBS, pH 7.4, using quartz cuvettes (1-cm light path). Solutions of 2 mmol/liter of CuSO 4 , ZnSO 4 , or Li 2 SO 4 were prepared in TBS, pH 7.4. Each salt was added in 1-l increment aliquots, followed by emission scanning between 300 and 420 nm, with a maximum at 346 nm. Scanning was performed three times for each concentration, and the results were reported as the average for three runs. To calculate the apparent kDa, data points representing the fluorescence quenching values were fitted using non-linear regression using a single-site hyperbola model.
Isothermal Titration Calorimetry (ITC)-Prostanoids (in ethanol or methyl acetate) were placed in glass vials, and the vehicle was evaporated under a nitrogen atmosphere; the dried material was then resuspended in appropriate concentrations in 20 mmol/liter Tris-HCl, 0.15 mol/liter NaCl, pH 7.4, sonicated, and vortexed as described (30). Calorimetric assays for measuring DMAV binding to a number of ligands were performed using a VP-ITC microcalorimeter (Microcal, Northampton, MA) at 30°C. Titration experiments were performed by making successive injections of 10 l each of 40 mol/liter ligand into the 1.34-ml sample cell containing 4 mol/liter DMAV until near saturation was achieved. Prior to the run, the proteins were dialyzed against 20 mmol/liter Tris-HCl, 0.15 mol/liter NaCl, pH 7.4, for binding experiments. The calorimetric enthalpy (⌬H cal ) for each injection was calculated after correction for the heat of the DMAV dilution obtained in control experiments performed by titrating DMAV 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 (OriginLab, Northampton, MA).
Surface Plasmon Resonance (SPR)-All SPR experiments were carried out in a T100 instrument (Biacore Inc., Uppsala, Sweden) following the manufacturer's instructions. For immobilization using an amine-coupling kit (Biacore), CM5 chips were activated with 1-ethyl-3-(dimethylaminopropyl) carbodiimide and N-hydroxysuccinimide before injection of DMAV (10 g/ml) in acetate buffer, pH 5.0. Remaining activated groups were blocked with 1 mol/liter ethanolamine, pH 8.5, resulting in a final immobilization of 593.7 RU. Kinetic experiments were carried out by injecting different glycans, 500,000-Da dextran sulfate (DS500K), 50,000-Da dextran sulfate (DS50K), heparin (20 kDa), and heparin sulfate (14 kDa), for a contact time of 120 s at a flow rate of 30 l/min at 25°C. For all runs, HBS-P buffer was used (10 mmol/liter HEPES, 150 mmol/liter NaCl, 0.005% surfactant P20, pH 7.4). DMAV-sulfated glycan complex dissociation was monitored for 600 s, and the sensor surface was regenerated by a pulse of 30 s of 1 mol/ liter NaCl at 30 l/min. For all experiments, blank flow cells were used to subtract the buffer effect on sensorgrams. After subtraction of the contribution of bulk refractive index and nonspecific interactions with the CM5 chip surface, the individual association (k a ) and dissociation (k d ) rate constants were obtained by global fitting of data using interaction models (bivalent ligand or two-state reaction) using BIAevaluation TM (BIAcore AB, Uppsala, Sweden). Values were then used to calculate the equilibrium constant (K D ). Conditions were chosen so that the contribution of mass transport to the observed values of K D was negligible. In addition, models in the T100 evaluation software fit for mass transfer coefficient to mathematically extrapolate the true k a and k d . For some analytes (DS50 and DS500), the dissociation was too slow to calculate the dissociation rate constant, and the K D was calculated using affinity (steady-state) mode, which indicates concentration dependence binding at equilibrium levels. For some experiments, several sulfated glycans (diluted in HBS-P) were injected for 90 s with a dissociation time of 300 s. In some experiments, DMAV was used to interact with biotinylated heparin, immobilized as follows. Neutravidin (40 g/ml; Pierce) in acetate buffer, pH 5.0, was immobilized in a CM5 chip for 7 min (10 l/min), resulting in 2,000 RU, and blocked with ethanolamine. Then, biotinylated heparin (10 g/ml in HBS-P; Sigma-Aldrich) was applied to the chip at a flow rate of 20 l/min until it reached ϳ100 RU. DMAV was injected at a flow rate of 10 l/min, with an association time of 30 s and dissociation time of 300 s. The chip was regenerated by a 30-s pulse with 1 mol/liter NaCl.
Interaction of DMAV with Heparin-Agarose-A HiTrapheparin-HP column (5 ml; GE Healthcare) was connected to an AKTA purifier system (Amersham Biosciences). Before use, the column was extensively washed with 2 mol/liter NaCl. The column was then equilibrated with 10 mmol/liter Tris, 0.15 mol/ liter NaCl (TBS) at a flow rate of 1 ml/min. DMAV (200 l at 12.5 mol/liter in TBS) was loaded into the column, followed by washing for 20 min with TBS. No increase in absorbance was observed in the flow-through. Then a 0 -2 mol/liter NaCl gradient (in TBS) at 1 ml/min was performed for 40 min.
Interaction of DMAV with Phospholipid Vesicles and Platelet Membranes-DMAV (3 mol/liter) was incubated with PC/PS (10 mol/liter) in 10 mmol/liter TBS (pH 7.4) supplemented with CaCl 2 (5 mM). After 30 min, the mixture was centrifuged to separate pellet and supernatant. NuPAGE sample buffer containing 5% ␤-mercaptoethanol was added to the samples and loaded in a 4 -12% NuPAGE gel. Gels were Coomassie Bluestained. DMAV that had not been incubated with PC/PS was used as a control. In other experiments, washed platelets (resuspended in Tyrode's buffer without BSA or gelatin) were incubated with PBS (control) or Horm collagen (10 g/ml) for 4 min under agitation in a Lumi-Aggregometer in the presence or absence of DMAV. Pellet and supernatant were obtained as above. DMAV that had not been incubated with platelets was used as control.
Statistical Analysis-Results are expressed as mean Ϯ S.E. (GraphPad Software, Inc., San Diego, CA). Statistical significance was set at p Ͻ 0.05 (t test or analysis of variance). For ICP-MS experiments, absolute expanded uncertainty for the results is 50% of the measured value, whereas the relative expanded uncertainty is 20% of the measured value for results of the same batch/measurement. Fig. 1A shows the alignment of antigen-5 family members from several bloodsucking arthropods. Most proteins are 20 -25 kDa with distinct pI values containing four cysteines and four histidines, which are highly conserved at positions His 52 , His 100 , His 160 , and His 215 (numbering system based on the sequence of DMAV). Other highly conserved clusters of amino acids are VGC and LVCNY. DMAV also exhibits a cluster of basic lysine residues in the C terminus, which may contribute to protein interaction with negatively charge molecules. Antigen-5 members belong to an expanded family of proteins from several blood-sucking arthropods whose phylogenetic associations are present in Fig. 1B.

Antigen-5 Family of Salivary Proteins and Phylogenetic, Biochemical, and Biophysical Properties-
To determine the function of the antigen-5 family of proteins, one member from D. maxima SGs was cloned, expressed, and refolded in a buffer containing 1 mmol/liter EDTA, 300 mmol/liter arginine, and oxidized and reduced glutathione at pH 9.3. The recombinant protein was purified through four chromatography steps, including two runs in gel filtration columns, one in an anion exchange column (Mono Q), and one in a cation exchange column (not shown). The purity of this preparation was assessed by one additional step in a reversed-phase column where only one peak was observed (not shown) and SDS-PAGE, which reveals one single band under denaturing or reducing conditions (Fig. 1C, inset). The N terminus of DMAV (MQXQNSNQFLGSLEITGKYR) was determined by Edman degradation and is identical to the sequence predicted by the cDNA for the mature protein. The mass of the recombinant protein (25,337 Da) was also confirmed by tandem MS (Fig. 1C) and is compatible with the molecular mass as predicted by the cDNA (25,341 Da), with an additional methionine used for expression purposes.
Next, DMAV was examined by analytical SEC-MALS-QELS-HPLC to determine solution state, molar mass, and hydrodynamic radius. Analysis of DMAV by SEC-HPLC indicated that 100% of the protein was monomeric in solution with a retention time of 19.018 min (Fig. 1D). Further analysis of DMAV by in-line MALS indicated that the weight average molar mass, 28.8 kDa, of the monomeric peaks is similar to the theoretical mass of 25,341 kDa (Fig. 1D), suggesting a globular shape of the protein. To further characterize the tertiary structure of DMAV, the hydrodynamic radius (R h ) of the monomeric recombinant protein was examined by QELS. The R h of DMAV was 2.8 nm, indicating that the monomeric form of the protein is globular-like. Next, CD spectra of DMAV were generated to examine its secondary structures, which were characterized by an increased intensity in negative ellipticity between 205 and 235 nm or decrease from 190 to 205 nm (Fig. 1E). Fig. 1E (inset) shows the percentage distribution of secondary structures in DMAV: ␣-helices (22.6 Ϯ 4.8%), ␤-sheet (30.0 Ϯ 4.4%), ␤-turn (16.1 Ϯ 4.2%), and random (30.7 Ϯ 4.8%).
DMAV was tested in a number of screening assays in an attempt to discover its function. DMAV (1 mol/liter) did not prolong the recalcification time or interfere with endothelial cell proliferation. It did not produce vasodilation or inhibit the complement cascade (data not shown).
DMAV Specifically Inhibits Collagen-induced Platelet Aggregation-When tested in platelets, DMAV was found to inhibit collagen-induced platelet aggregation. The specificity of DMAV as a platelet aggregation inhibitor was then verified. Fig.  2A shows that only low doses (Ͻ1 g/ml) of collagen-induced platelet aggregation are inhibited dose-dependently by DMAV. No effects on shape change were observed, suggesting that the inhibitor does not interfere with a proximal pathway of platelet activation (i.e. collagen-receptor interaction). It has also been noticed that platelets from some donors were more sensible to inhibition by DMAV than others. In addition, our results demonstrate that DMAV does not affect platelet aggregation by moderate to high doses of collagen or any concentration tested for a number of agonists, including convulxin, thrombin, arachidonic acid, TXA 2 /PGG 2 mimetic U46619, native PGG 2 , PGH 2 , ADP, and PAF.
Antigen-5 Member from T. infestans Salivary Gland Also Inhibits Platelet Aggregation-Besides DMAV, a second member of the antigen-5 family was expressed from T. infestans (TIAV) in order to verify whether the anti-platelet function was conserved. Fig. 3A compares the SDS-PAGE for both DMAV and TIAV, which exhibits the expected molar masses. Fig. 3B demonstrates that TIAV inhibits platelet aggregation by low doses of collagen (0.5 g/ml). Because collagen, but not ADP or thrombin, is the only platelet agonist that consistently promotes an increase of extracellular O 2 . (10, 11), it was hypothesized that antigen-5 members might affect the pro-aggregatory properties of ROS on platelets. Therefore, we compared the effects of TIAV with Cu,Zn-SOD. Fig. 3B also shows that Cu,Zn-SOD, like TIAV, attenuates platelet aggregation (but not the shape change) by collagen (0.5 g/ml). On the other hand, arachidonic acid (58 mol/liter)-induced aggregation is not affected by both inhibitors (Fig. 3C). We also tested the effects of TIAV in other parameters of platelet activation. Fig. 3, D and E, demonstrate that TIAV dose-dependently inhibits platelet ATP secretion and TXA 2 generation induced by collagen, respectively. Because it has been reported that NADPH oxidase inhibitors, such as diphenylene iodonium and apocynin, partially attenuate thrombus formation triggered by a collagen surface at high shear rates in vitro (38), we tested TIAV in this assay. Fig. 3F  3 kDa) and purified as described for DMAV were ineffective in the same assay (Fig. 4D).
To confirm the antioxidant activity of DMAV using another detection method, we evaluated its effects in the cyt c reduction assay using X/XO to generate O 2 . in vitro. Fig. 4E demonstrates that DMAV produces partial inhibition of cyt c reduction, whereas three other recombinant proteins (ixolaris, dipetalodipin, and BSA) were without effect. We also determined that the enzymatic activity of X/XO, evaluated by production of uric acid, was unaffected by DMAV (Fig. 4F)

REACTIONS 2 and 3
To test whether DMAV also generates H 2 O 2 , peroxide production was estimated using Amplex Red reagent in reactions started with X/XO. In contrast to Cu,Zn-SOD, which increases H 2 O 2 formation, DMAV did not increase its production (Fig.  4G).
DMAV was also tested to determine whether it works as a catalase according to Reaction 4.

REACTION 4
An H 2 O 2 standard curve was constructed, followed by incubation for 15 min with PBS or DMAV. H 2 O 2 was detected with Amplex Red. Fig. 4H shows that no change in the concentration of H 2 O 2 was observed. We aimed to determine whether DMAV displays catalase activity using a direct method using a Clark electrode, which measures oxygen concentration in a solution (33). In this experiment, the solution containing xanthine exhibited an oxygen concentration of ϳ280 M. When XO was added to the chamber (under stirring conditions), the generation of O 2 . was accompanied by consumption of O 2 , detectable by a continuous downward inflection of the pen (Fig. 4I) (Fig. 4I). When the experiments were performed as above but DMAV was added instead of PBS, no change in oxygen consumption was noticed. Furthermore, after the addition of catalase, the same change in O 2 . formation was observed (relative to PBS), implying that DMAV is devoid of catalase activity. Proteins exhibit antioxidant activity through three distinct mechanisms (i.e. methionine oxidation (40), cysteine oxidation (41), and metal oxidation (39)). To verify whether free cysteines with the potential to scavenge O 2 . are present in DMAV, the protein was incubated with benzofurazone 4-fluoro-7-aminosulfonylbenzofurazan, which reacts with free cysteines. No increase over basal levels of the fluorescence of DMAV was observed. As a control, free cysteines or reduced DMAV produced a strong fluorescence signal (not shown). It is concluded that DMAV does not contain free cysteines that could mediate its antioxidant activity. To verify whether methionines in DMAV were oxidized, the protein was incubated with X/XO, and its mass was evaluated by MS. No change in mass compatible with methionine oxidation was observed (not shown). Next, DMAV was incubated with EDTA to chelate metals potentially bound to the protein. Fig. 4J shows that DMAV loses its antioxidant activity in the presence of EDTA (5 mmol/liter), which is consistent with a metal-dependent activity. In contrast, no inhibition of the antioxidant activity of Cu,Zn-SOD was detected, because it tightly interacts with Cu 2ϩ /Zn 2ϩ (42). The distinct sensitivity of DMAV and Cu,Zn-SOD for EDTA also implies that our preparation is devoid of contaminating E. coli Cu,Zn-SOD.
DMAV Binds Divalent Metal and ICP-MS Results-To estimate the metal content of recombinant DMAV, a highly purified sample was applied to ICP-MS, which is the most sensitive method available to evaluate metals and other ions present in a given sample (35). Fig. 5A presents the ratio (ion content in the sample, ng/g)/(ion content in the buffer, ng/g) in the preparation. The results indicate that DMAV binds Co 2ϩ , Zn 2ϩ , and Cu 2ϩ . In addition, Ni 2ϩ was found in our sample (not shown), which is often a contaminant detected by ICP-MS. No other ion of the periodic table was found to interact with DMAV. Next, we evaluated whether DMAV binds additional metals provided in purified form. For these experiments, DMAV was incubated with EDTA and extensively dialyzed against Hepes-saline. Each metal (30 mol/liter) was added to DMAV (10 mol/liter), and after 30 min, the mixture was again extensively dialyzed. Fig. 5B demonstrates that DMAV equally binds Cu 2ϩ , Mn 2ϩ , Ni 2ϩ , Co 2ϩ , and Zn 2ϩ . When no metal was added to DMAV, only trace amounts of Zn 2ϩ and Cu 2ϩ were identified in the sample, indicating DMAV bound to trace metals present in the buffer. Next, the relative specificity of DMAV for Cu 2ϩ and other metals was evaluated. In competition assays, DMAV was added to a mixture containing equal concentrations of Cu 2ϩ , Mn 2ϩ , Ni 2ϩ , Co 2ϩ , and Zn 2ϩ followed by dialysis. Only Cu 2ϩ was identified in the preparation (Fig. 5C). In displacement assays, DMAV was incubated with Mn 2ϩ , Ni 2ϩ , Co 2ϩ , and Zn 2ϩ , followed by the addition of Cu 2ϩ , and dialyzed. Fig. 5D demonstrates that only Cu 2ϩ was found associated with DMAV.
Fluorescence Spectroscopy for DMAV-Tryptophan fluorescence spectra of metal-free and Cu 2ϩ -complexed preparations of DMAV were employed to determine the relative affinity of the interaction identified by ICP-MS. The addition of CuSO 4 induced quenching of the fluorescence to 30% of the initial intensity, suggesting that tryptophan indoles moved to a more hydrophobic environment (Fig. 6A). In contrast, no change in intrinsic fluorescence was observed when Li 2 SO 4 was added to DMAV (Fig. 6A, dotted lines). Data points were fitted by nonlinear regression to a single-site hyperbolic model, which resulted in an apparent K D of 20.7 mol/liter (Fig. 6B). This value does not reflect the true affinity of DMAV for the ion because it varies with the concentration of DMAV used in the assay (20 mol/liter). As an additional control, no change in the intrinsic fluorescence spectroscopy was observed when CuSO 4 (60 mol/liter) was added to DM146 (gi315270899), another E. coli expressed salivary protein from D. maxima (Fig. 6C). The UV spectrum of DMAV was performed, and peak absorbance was detected at 280 nm. Additional peaks were not observed at longer wavelengths in the presence of Cu ϩ (data not shown), in contrast to Cu ϩ oxidases (so-called "blue copper proteins") (43,44).
Comparison of the Antioxidant Activities of Cu 2ϩ -DMAV and Cu,Zn-SOD-To verify the antioxidant potential of the protein, DMAV (10 mol/liter) previously incubated with (20 mol/liter) Cu 2ϩ was extensively dialyzed against PBS, pH 7.4.  . was generated by an X/XO system; luminescence was detected using luminol (500 mol/liter). D, several proteins (DM146, ixolaris, triplatin, 9.3 kDa, SVEP, and yellow protein; 2 mol/liter) were expressed and purified as DMAV and tested side by side. Detection of O 2 . was performed with Luminol (500 mol/liter). E, O 2 . was produced by the X (500 mol/liter)/XO (1.2 milliunits/ml) system and detected by the reduction of cytochrome c (12 mol/liter) at 550 nm. Other available salivary recombinant proteins (dipetalodipin (DPTL), ixolaris, and BSA; 2 mol/liter) were tested side by side with DMAV (5 mol/liter). F, DMAV (5 mol/liter) does not affect the enzymatic function of X/XO, determined by uric acid production at 293 nm. G, effects of DMAV on generation of H 2 O 2 . X (500 mol/liter)/XO (6.3 milliunits/ml) was used to generate H 2 O 2 , which was detected with the Amplex Red reagent. In control, spontaneous dismutation of O 2 . is observed, which is increased in the presence of Cu,Zn-SOD (4 milliunits/ml). DMAV (5 mol/liter) did not increase H 2 O 2 production. H, DMAV does not display catalase activity. Different concentrations of H 2 O 2 were incubated for 15 min with PBS or DMAV (2 mol/liter) and detected with Amplex Red reagent as described under "Experimental Procedures." I, catalase activity. XO (0.7 milliunits/ml) was added to a solution containing hypoxanthine (500 M), and O 2 consumption was detected with a Clark electrode. After 15 min, catalase (333 units/m) was added to the chamber, and a peak of O 2 is observed. No effects were observed after the addition of DMAV (0.5 M). J, DMAV displays metal-dependent antioxidant activity. DMAV (5 mol/liter) or Cu,Zn-SOD (0.42 unit/ml) was incubated with EDTA (5 mmol/liter) for 15 min and tested in the cytochrome c reduction assay. Antioxidant activity of DMAV was abolished by EDTA (n ϭ 3). Significance (*) was set at p Ͻ 0.05 (analysis of variance). A.U., arbitrary units. Error bars, S.E.
To study the kinetics of the reaction of Cu 2ϩ -DMAV with O 2 .
generated by X/XO, the rate of formation of cyt c reduction was spectrophotometrically monitored at pH 7.8 in the absence (V) and presence (v) of various Cu 2ϩ -DMAV concentrations. Results obtained by this method compare well with direct measurement of O 2 . estimated by pulse radiolysis (45). Fig. 7, A and B, show the progress curves of cyt c reduction with increasing concentrations of Cu 2ϩ -DMAV or Cu,Zn-SOD respectively. Fig. 7, C and D, shows the dose-response curves for the results presented in 7, A and B, respectively, with a calculated IC 50 of ϳ50 nmol/liter for Cu 2ϩ -DMAV and 1 nM for Cu,Zn-  SOD. Data were analyzed by plotting V/v as a function of [Cu 2ϩ -DMAV] according to Krishna et al. (33) (Fig. 7E). Fig. 7F shows the same transformation for Cu,Zn-SOD. From the slope of the line and knowing Kcyt-c III ϩ superoxide (33,46) at pH 7, the k app for the Cu 2ϩ -DMAV interaction with O 2 . was calculated to be 0.31 ϫ 10 8 mol liter Ϫ1 s Ϫ1 . This is ϳ30 -50 times less than the K cat (2-3 ϫ 10 9 mol liter Ϫ1 s Ϫ1 ) observed for [Cu,Zn-SOD] (33,39,(45)(46)(47). The antioxidant activity of Cu 2ϩ -DMAV was not affected by the presence of catalase in the assay (not shown). McCord and Fridovich (48) defined the specific activity of SOD in units/mg, where 1 unit is the amount in mg of protein that inhibits 50% of the reduction of 10 mol/liter cytochrome c by O 2 . (generated by the X/XO system) in a 3.0-ml reaction volume containing 50 mmol/liter phosphate buffer at pH 7.8. This is ϳ430 units/mg Cu 2ϩ -DMAV. DMAV was devoid of antioxidant activity when preincubated with Zn 2ϩ (data not shown). DMAV Binds to Heparin and Sulfated Glycosaminoglycans-It has been reported that extracellular SOD interacts with heparin through its C terminus (49). To verify whether DMAV interacts with sulfated glycosaminoglycans, it was immobilized in a CM5 sensor chip, followed by injection of sulfated glycosaminoglycans. Fig. 8A shows that DS50K and DS500K avidly interact with DMAV. Binding was also detectable with heparin and heparan sulfate. In contrast, no signal was observed when non-sulfated dextran 12K, non-sulfated dextran 670K, chondroitin sulfate, or enoxaparin were used as analytes (not shown). Next, the kinetics of DMAV interaction with DS500K and DS50K were studied. Sensorgrams are shown in Fig. 8, B and C, and data were fitted with a bivalent analyte model. In both cases, the association phase was followed by a very slow dissociation rate, which does not allow precise calculation of the k off . In these cases, the K d can be determined by steady-state kinetics. Fig. 8D shows that DS500K and DS50K interact with immobilized DMAV in a saturable manner with a K d of ϳ0.41 nmol/liter Ϯ 0.05 nmol/ liter and 1.96 nmol/liter Ϯ 0.2 nmol/liter, respectively. Fig. 8, E and F, also demonstrates that DMAV interacts with heparan sulfate or heparin, respectively. To estimate the K d of heparin-DMAV interactions with accuracy, biotinylated heparin was coupled to neutravidin immobilized in a CM5 chip, and DMAV was used as an analyte. Sensorgrams are depicted in Fig. 8G. The inset presents the fit of the data using steady-state kinetics, which resulted in a K d of 97.0 Ϯ 4.5 nmol/liter. The interaction of DMAV with heparin was also tested using a heparin-agarose column. Fig. 8H demonstrates that elution of DMAV was observed at ϳ1 mol/liter NaCl. This result provides additional evidence to conclude that DMAV is a heparin-binding protein. In order to evaluate whether DMAV binds to membranes, it was incubated with PC/PS or activated platelets as described under "Experimental Procedures." In both cases, no interaction was observed (not shown), excluding DMAV as a phospholipidbinding protein.
Antioxidant Activity Is Present in the SG of D. maxima-To examine whether the SG of D. maxima exhibits antioxidant activity, salivary gland homogenate (SGH) was prepared, and the supernatant was obtained after centrifugation. Fig. 9A shows that SGH inhibits the signal of O 2 . generated by X/XO with an IC 50 of ϳ0.05 pairs/assay. In addition, SGH dose-dependently attenuated collagen (0.53 g/ml)-induced platelet aggregation (Fig. 9B). Because SGH is a complex mixture of proteins, and different molecules may exhibit antioxidant activity or inhibit platelet aggregation, our goal was to verify whether specific fractioncontaining DMAV also exhibited the activities. SGH was prepared from 100 pairs of glands and supernatant loaded into a G-75 gel filtration column. Fig. 9C shows the elution profile of the homogenate, which was characterized by a large peak where most lipocalins (the most abundant protein family in triatomine saliva) are found. Recombinant DMAV loaded onto the same column under identical experimental conditions elutes around 11.5-12.5 min, which is compatible with its molar mass of 25 kDa (Fig. 9D). To determine the elution profile of native DMAV, an aliquot of each fraction was used to immobilize the protein, and an ELISA was performed using polyclonal anti-DMAV antibody as a probe. Fig. 9E shows that the peak of DMAV recognized by the antibodies coincides with the elution pattern of recombinant protein. To confirm the presence of DMAV in the SGH, fractions 21 and 23, which flank the maximum antibody titles indicative of DMAV by ELISA, were chosen. These fractions were loaded into a one-dimen- sional SDS-PAGE followed by gel-tryptic digestion and MS/MS of the bands. Fig. 9, F (for fraction 21) and G (for fraction 23), shows that DMAV was identified without ambi-guity as one of the most abundant components in addition to lipocalins. Fractions 21 and 23 were also evaluated for the Cu 2ϩ content by ICP-MS. Fig. 9H demonstrates that Cu 2ϩ is detectable in fraction 21, and higher levels are observed in fraction 23. Finally, fractions 21 and 23 exhibit antioxidant activity using the cyt c reduction assay, which was metal-dependent because it was blocked by EDTA (Fig. 9I).

DISCUSSION
Studying the function of the antigen-5 family of proteins revealed that members of this family exhibit antioxidant activity. This was concluded based on several enzymatic or cell- Control experiments excluded the possibility that free thiols (41) or methionine oxidation (40) mediates the antioxidant function of DMAV. In contrast, EDTA completely abolished its antioxidant function, indicating a metal-dependent mechanism. ICP-MS was used to determine the metal content in DMAV. Results revealed that DMAV binds divalent ions Mn 2ϩ , Co 2ϩ , Ni 2ϩ , Cu 2ϩ , and Zn 2ϩ . Through competition and displacement studies, it was concluded that DMAV has higher specificity for Cu 2ϩ over other metals. Moreover, the addition of Cu 2ϩ to apo-DMAV decreased the intrinsic fluorescence of the protein, indicating a conformational change upon binding. This experiment estimated an apparent K D in the micromolar range, which is an overestimation because it is of the same order of magnitude as the protein concentration used in the assay. Under identical experimental conditions, Li 2 SO 4 did not induce protein intrinsic fluorescence change. Furthermore, CuSO 4 does not induce intrinsic fluorescence change in DM146, an unrelated recombinant salivary protein.
Kinetic studies using a cytochrome c reduction assay suggest that Cu 2ϩ -DMAV operates as a catalyst and not as a simple scavenger of O 2 . (33,45). Our interpretation is that antigen-5 members belong to a novel family of metal-dependent antioxidant enzymes. Of note, identification of Cu 2ϩ as a ligand for DMAV is congruent with the identification of metal (i.e. Zn 2ϩ ) coordination in the structure of antigen-5 members from hookworm and snake venoms (50,51). However, it remains to be determined how DMAV interacts with Cu 2ϩ and the mechanism of metalation. The finding that Cu 2ϩ preferentially binds to DMAV is also particularly relevant because Zn 2ϩ does not exhibit redox potential and therefore cannot confer antioxidant properties to a protein. Structural information will be critical to gain further insight on the mechanism of action of DMAV. With respect to metal requirements for antioxidant activity, Cu 2ϩ -DMAV resembles Cu,Zn-SOD, which binds Cu 2ϩ and also exhibits one of the highest k cat values (k cat for O 2 . , ϳ2-3 ϫ 10 9 mol liter Ϫ1 s Ϫ1 ) (39,45,47). It has also been shown that several Cu 2ϩ salts or low molar mass Cu 2ϩ chelates exhibit a high catalytic redox potential and scavenge O 2 . with efficiency close to that of Cu,Zn-SOD (52)(53)(54)(55). Cu 2ϩ -binding proteins, such as ceruloplasmin (56) Cu,Zn-SOD and Cu 2ϩ -DMAV differ in several respects. On a molar basis, Cu 2ϩ -DMAV is about 30 -50 times less active than Cu,Zn-SOD. EDTA abolishes the antioxidant activity of Cu 2ϩ -DMAV and other complexes (52, 53), but not Cu,Zn-SOD. This indicates that the interaction of DMAV with Cu 2ϩ is not tight as observed with Cu,Zn-SOD (42,45), as confirmed experimentally (Fig. 6B). Moreover, we did not detect accumulation of H 2 O 2 production in the presence of DMAV. It is concluded that the products of spontaneous dismutation are identical to the products of the DMAV-catalyzed reaction, as initially reported for Cu,Zn-SOD (48). Furthermore, the Cu 2ϩbinding proteins such as tyrosinase, laccase, ascorbate oxidase (66) . was generated by the X (500 mol/liter)/XO (1.2 milliunits/ml) in the presence of SGH of D. maxima. Superoxide was detected by chemiluminescence using luminol. B, effects on platelet aggregation. Platelets were incubated with the SGH of D. maxima followed by the addition of collagen (0.53 g/ml). Aggregation was detected by turbidimetry in a Lumi-Aggregometer. C, fractionation of the SG of D. maxima. SGH corresponding to 100 pairs of D. maxima was applied to a Sephadex-G75 column as eluted with PBS, pH 7.4, at a flow rate of 0.5 ml/min. D, recombinant DMAV (100 g) was loaded into a Sephadex G-75 column equilibrated in PBS and eluted at 0.5 ml/min. E, ELISA for DMAV. Fifty l of each fraction collected in C was added to Immulon 2HB plates, followed by incubation of anti-DMAV polyclonal antibodies and alkaline anti-rabbit phosphatase-coupled secondary antibodies as described under "Experimental Procedures." Fraction 21 (F) and fraction 23 (G) were separated by one-dimensional NuPAGE, and the bands were used for in-gel tryptic digestion and MS. Each peptide mass was searched against a database for the SG proteins from D. maxima. Results are indicated by arrows. 3-PGK, 3-phosphoglycerol kinase. H, Cu 2ϩ content of fractions 21 and 23 was estimated by ICP-MS. Expanded uncertainty of the bars are 50% of the value. I, antioxidant activity. Fractions 21 and 23 were tested for inhibition of cytochrome c reduction using X (500 mol/liter)/XO (1.2 milliunits/ml) with and without EDTA (5 mmol/liter). Significance (*) was set at p Ͻ 0.05 (Student's t test). Error bars, S.E. suppresses production of TXA 2 (12). In support of our findings that DMAV inhibits extracellular O 2 . is the result presented in Fig. 3B showing that Cu,Zn-SOD inhibits platelet aggregation by low doses of collagen only, but, as reported before, it does not interfere with aggregation induced by high concentrations, such as 6 g/ml (10). Furthermore, the increase in extracellular O 2 . levels occurs after 2 min of collagen addition and coincides with the appearance of large aggregates (8). Our results demonstrated that the anti-aggregatory effect of DMAV is only observable at later time points (no effects in shape change), when the production of O 2 . has already started. Also, thrombus formation under flow in a collagen surface is only partially attenuated by negative modulators of intracellular O 2 . production, such as NADPH oxidase inhibitors (e.g. diphenylene iodonium, apocyanin, and 2-acetylphenothiazine) or Cu,Zn-SOD mimetic (e.g. MnTMPyP) (38,68). In agreement with these results, TIAV, which acts extracellularly, marginally affected thrombus formation in vitro (Fig. 3F).
It is important to emphasize that although convulxin (Cvx) generates intracellular ROS (24,69), it does not promote ROS increase extracellularly; nor does Cu,Zn-SOD inhibit platelet activation by Cvx (69). These results are therefore in agreement with the lack of effects of DMAV in aggregation caused by low doses of Cvx ( Fig. 2A). Actually, this result is not surprising, because Cvx is a strong agonist that promotes direct PLC␥2 activation and that relies only marginally on secondary mediators (e.g. TXA 2 and ADP) to induce full platelet aggregation (70 -72). Although thrombin or thrombin-receptor activating peptide (TRAP) induce ROS production intracellularly (38,69) and aggregation is negatively modulated by intracellular antioxidants MnTMPyP (20) or Tempol (73), extracellular O 2 . in the supernatant of thrombin-activated platelets has not been detected by Krotz et al. (10) or Caccese et al. (11). Plausibly, this explains why DMAV does not affect thrombin-induced platelet aggregation ( Fig. 2A). Finally, our experiments using ITC and platelet aggregation assays excluded the possibility that pro-aggregatory molecules involved in platelet aggregation by collagen, such as ADP or TXA 2 , PGG 2 /PGH 2 , and 12-HETE (37), are targets for DMAV. It is concluded that DMAV interrupts platelet aggregation by a mechanism never reported before for a salivary protein that is explained . also plays an important role in inflammation, as evidenced by the inhibitory role of extracellular SOD (EC-SOD or SOD3) in neutrophilic inflammation in the lung and in the skin (78 -83). Conceivably, DMAV interferes with the proinflammatory effects of extracellular ROS produced by neutrophils, platelets, and other cell types (84,85), resulting in attenuation of expression of procoagulant tissue factor, NET formation, cytokine release, lipid peroxidation, and conversion of nitric oxide to peroxynitrite in the vasculature (86 -90). In other words, scavenging of extracellular O 2 . down-modulates a myriad of proinflammatory reactions. SPR results and affinity chromatography experiments demonstrated that DMAV interacts with immobilized heparin (K D ϳ100 nmol/liter) and other sulfated compounds (e.g. heparan sulfate and dextran sulfate). This finding reveals a remarkable similarity of DMAV with extracellular Cu,Zn-SOD, which binds heparin sulfate proteoglycans on cell surfaces via its heparin-binding domain (86). It is conceivable that binding of DMAV to endothelial cells increases the effective concentration of the enzyme and the antioxidant tonus at bite sites. Further, the presence of DMAV in the salivary gland was supported by the identification of metal-dependent antioxidant activity in chromatographic fractions containing native DMAV, confirmed by mass spectrometry and ELISA. Nevertheless, purification of native DMAV from the gland will be necessary in order to fully characterize it as an enzyme in nature, through the determination of the metal content and the kinetics and products of the reaction with O 2 . . Finally, it is notable that the salivary glands of blood-sucking arthropods reportedly contain at least one antiplatelet, one anticoagulant, and one vasodilator (1). Based on the results presented herein, and because antigen-5 members are ubiquitously found and relatively abundant, it is fair to say that at least one antioxidant might also be present in these glands. The discovery of novel protein families in insect salivary protein or other exogenous secretions has proven particularly important in our understanding of vertebrate vascular biology and in providing tools for biochemistry and cell biology or to predict the function of endogenous proteins (1,2). By analogy with DMAV, we speculate that other endogenous CAP/antigen-5 members (e.g. GLIPR1, GAPRI, PI15, and CRISP) (26) might also behave as metal-binding proteins with pro-or antioxidant functions. In conclusion, expression of antigen-5 members in the SGs of insects and ticks counteracts the diverse proinflammatory function of O 2 . in the microcirculation, contributing to the successful feeding of blood-sucking arthropods.