Life Sciences 93 (2013) 416–422

Contents lists available at ScienceDirect

Life Sciences

j ourna l homepage: www.e lsev ie r .com/ locate / l i fesc ie

Inhibition of human platelet aggregation by eosinophils

Aline Mendes Maziero a,⁎, Raquel Lorenzetti a, José Luiz Donato b, Sergio Lilla a, Gilberto De Nucci c

a Department of Pharmacology, Faculty of Medical Sciences, State University of Campinas (UNICAMP), Campinas, São Paulo, Brazilb ATCGen Biotechnology, Campinas, São Paulo, Brazilc Department of Pharmacology, Institute of Biomedical Sciences, University of São Paulo (USP), São Paulo SP, Brazil

⁎ Corresponding author at: Department of PharmacoloState University of Campinas (UNICAMP), 13084-971, C19 3521 9531; fax: +55 19 3289 2968.

E-mail address: maziero@fcm.unicamp.br (A.M. Mazie

0024-3205/$ – see front matter © 2013 Elsevier Inc. All rihttp://dx.doi.org/10.1016/j.lfs.2013.07.012

a b s t r a c t

a r t i c l e i n f o

Article history:

Received 15 May 2013Accepted 9 July 2013

Keywords:Eosinophil proteinsEosinophil cytosolic fractionPlatelet aggregation inhibitionEosinophil cationic protein

Aims: The relationship between the activity of eosinophils and platelets has been observed in recent decades bymany scientists. These observations include increased numbers of eosinophils associatedwith platelet disorders,including changes in the coagulation cascade and platelet aggregation. Based on these observations, the interac-tion between eosinophils and platelets in platelet aggregation was analyze.Main methods: Human platelets were incubated with eosinophil cytosolic fraction, promyelocytic human HL-60clone 15 cell lineage, and eosinophil cationic protein (ECP). Platelet rich plasma (PRP) aggregation was inducedby adenosine diphosphate, platelet activating factor, arachidonic acid, and collagen, and washed platelets (WP)were activated by thrombin.

Key findings: Aggregation induced by all agonists was dose dependently inhibited by eosinophil cytosolic fraction.This inhibition was only partially reversed by previous incubation of the eosinophils with L-Nitro-Arginine-Methyl-Ester (L-NAME). Previous incubationwith indomethacin did not prevent the cytosolic fraction induced inhi-bition. The separation of eosinophil cytosolic fraction by gel filtration on Sephadex G-75 showed that theinhibitory activity was concentrated in the lower molecular weight fraction. HL-60 clone 15 cells differentiatedinto eosinophils for 5 and 7 day were able to inhibit platelet aggregation. The ECP protein inhibited the plateletaggregation on PRP and WP. This inhibition was more evident in WP, and the citotoxicity MTT assay proved theviability of tested platelets, showing that the observed inhibition by the ECP protein does not occur simply by celldeath.Significance: Our results indicate that eosinophils play a fundamental role in platelet aggregation inhibition.

© 2013 Elsevier Inc. All rights reserved.

Introduction

Acquired platelet dysfunction with eosinophilia or nonthrombo-cytopenic purpura with eosinophilia is an acquired bleeding disorderof unknown etiology associated with platelet dysfunction and eosino-philia (Mitrakul, 1975; Suvatte et al., 1979).

Many authors have reported associations between the increasednumbers of eosinophils with platelet dysfunctions, such as increasedbleeding time, reduction in platelet aggregation induced by variousagonists, among others disorders (Chin and Koong, 1990; Hathiratet al., 1993; Laosombat et al., 2001; Lim et al., 1989; Lucas andSeneviratne, 1996; Muthiah et al., 1984; Poon et al., 1995; Ramanathanand Duraisamy, 1987; Suvatte et al., 1979).

In most cases the bleeding symptoms are mild and transient withspontaneous recovery, and eosinophilia remains only a few weeksafter the onset (Lee, 2012; Lim et al., 1989; Poon et al., 1995;

gy, Faculty of Medical Sciences,ampinas (SP), Brazil. Tel.: +55

ro).

ghts reserved.

Ramanathan and Duraisamy, 1987). The eosinophil count can varyfrom 3 to 6% of total white blood cells (Suvatte et al., 1979). Bleedingtime is prolonged in about 60% of patients (Suvatte et al., 1979;Hathirat et al. 1982), but the platelet count is normal. Platelet adhesive-ness is abnormally low in 60% of patients. Aggregation in response tostimulation by ADP, thrombin, and collagen is decreased, but the re-sponse to ristocetin is normal (Laosombat et al., 2001; Lim et al., 1989;Suvatte et al., 1979).

A distinctive group of proteins comprise the eosinophil granule(Gleich and adolphson, 1993; Peters et al., 1986). Among these, theeosinophil major basic protein (MBP) on a molar basis is the mostabundant eosinophil granule protein. The surrounding granule matrixcontains other proteins, including eosinophil-derived neurotoxin(EDN), eosinophil cationic protein (ECP), and eosinophil peroxidase(EPO) (Abu-Ghazaleh et al., 1992; Gleich et al., 1993).

The eosinophils are transported in the bloodstream in a resting stateand can be activated at sites of inflammation or when an activatedendothelium is found. Only a small amount of activated eosinophilsseems to be found in the bloodstream. However, in some circumstances,such as in idiopathic eosinophilic syndrome (HES), where there are anincrease of eosinophils, and consequently a higher level of activatedcells in the bloodstream (Egesten et al., 2001).

417A.M. Maziero et al. / Life Sciences 93 (2013) 416–422

The protein synthesis in eosinophils may occur in a constitutivemanner, where new proteins are constantly secreted or regulated(degranulation) in response to specific stimuli.

In recent years, non-classical forms of eosinophils degranulationhave been suggested. These include piecemeal degranulation, which isbased on morphological changes of granules or cytolytic degranulation(e.g., necrosis). This phenomenon is present in atopic dermatitis andallergic inflammation of the upper airways (Egesten et al., 2001).

Based on evidence about the involvement between eosinophils andplatelet disorders, this work describes the effect of eosinophil cytosolicfraction, HL-60 clone 15 cells lineage differentiated into eosinophil,and the ECP protein on the in vitro platelet aggregation induced bydifferent agonists.

Materials and methods

Materials

Adenosine diphosphate (ADP), platelet-activating factor (PAF),collagen, arachidonic acid, thrombin, Nω-nitro-L-arginine methyl ester(L-NAME), [3-(4,5-dimethythiazol-2-yl)-2,5 diphenyl tetrazolium bro-mide] (MTT), minimum essential medium (MEM), butyric acid andsephadex G-75 was purchased from Sigma Chemical Co. (St. Louis,MO, USA). The promyelocytic human HL-60 clone 15 cell lineage wasobtained from the American Type Culture Collection (Rockville, MD).Eosinophil cationic protein (ECP) was purchased from Lee biosolutionslaboratory (St. Louis, MO, USA). The other materials and chemicalswere obtained from commercial sources.

Platelet rich plasma (PRP) preparation

The platelet rich plasma was prepared as previously described withminor modifications (Donato et al., 1996).

Blood from healthy donors, who had not taken anymedication for atleast ten days, was anticoagulated with 3.8% sodium citrate (1:9 v/v).PRP was obtained by centrifuging whole blood at 200 g × 15 min atroom temperature. The supernatant (PRP) was collected and left atroom temperature until the aggregation assay was performed. Analiquot (1 ml) of PRP was centrifuged at 2000 ×g for 15 min at roomtemperature to obtain the platelet poor plasma (PPP) used to calibratethe aggregometer for the maximum aggregation.

Washed platelet (WP) preparation

Blood was collected in plastic tubes containing the ACD-C anticoag-ulant solution (citric acid/citrate/dextrose) at the proportion of 1:9(v/v). The mixture was centrifuged at 200 ×g for 15 min at room tem-perature, and the supernatant (PRP) was collected in a clean tube.Iloprost (0.8 μM) was added to the PRP/ACD-C, which was then centri-fuged at 800 ×g for 12 min. The supernatant was discarded and theplatelet pellet was carefully resuspended in calcium-free Krebs–Ringersolution (composition in mM: NaCl, 118; Mg SO4.7H2O, 1.7; KH2PO4,1.2; NaHCO3, 25.0; glucose, 5.6). Iloprost (0.8 μM) was added againand the final cell suspension was centrifuged at 800 ×g for 10 min.The precipitated platelets were finally resuspended in calcium-freeKrebs–Ringer solution to a final volume sufficient for a platelet suspen-sion containing 2 × 108 platelets/ml. The platelet suspension was keptat room temperature, and CaCl2 (1.0 mM) was added to the plateletsuspension immediately before the aggregation assay.

Platelet aggregation

Platelet samples (0.4 ml)were incubated at 37 °C for 1 min and con-stantly stirred at 900 rpm in a double-channel Lumi-Aggregometer(Chrono-log 560 CA; Chronolog Corp., Havertown, PA, USA), beforethe addition of 10 μl of different agonists. Platelet aggregation using

either PRP or WP was monitored for at least 5 min. The results wereexpressed as the percentage of maximum light transmission obtainedwhen the aggregometer was calibrated for 100% transmission withPPP (PRP platelet aggregation assays) or Krebs solution (WP aggrega-tion assays).

Preparation of the eosinophil cytosolic fraction

Eosinophils were obtained from the peritoneal cavity of 20–30 maleWistar rats (200–230 g) and purified on a discontinuous metrizamidegradient (Vadas et al., 1979). Briefly, the animals were killed with anoverdose of halothane anesthesia, and the peritoneal cavities werewashedwith 20 ml of Phosphate-buffered saline (PBS; pH 7.2) contain-ing heparin (20 units/ml). The peritoneal washings obtained from theanimals were collected in a clean plastic tube immersed in an ice bathand centrifuged at 1000 ×g for 10 min at 20 °C. The metrizamide dis-continuous gradient was prepared by carefully layering 2.5 ml of de-creasing concentrations of metrizamide dissolved in minimumessential media (MEM; pH 7.2) containing 0.1% gelatin (23.5, 20, and18%, wt/vol) into a conical propylene tube. The resulting leukocyte-rich pellet was gently resuspended in 2.5 ml of 18% metrizamide andlayered over the top of the metrizamide gradient. The gradient tubewas first centrifuged at 90 ×g (11 min at 4 °C) and then at 1000 ×g(14 min at 4 °C). The gradient zone containing the eosinophils(between the 23.5% and 20% gradients) was removed and washedtwice in MEM containing 1 mg/ml of hydrolyzed ovalbumin. TheMay–Grünwald dye exclusion test show that the final cell suspensioncontained 80–90% eosinophils and a cell viability above 90%. Beforetesting, the eosinophil suspension was diluted in Hanks balanced saltsolution (pH 7.2) to give a final concentration of 5 × 107 cells/ml.

To avoid the interference of eosinophil nitric oxide (NO) release orPG1 synthesis on platelet aggregation, eosinophils were incubatedwith L-NAME (10 μM) and indomethacin (10 μM) for 60 and 20 minat 37 °C, respectively. Cells were prior disrupted by sonication (SonicDismembrator, Model 100; Fischer Scientific) in an ice bath using3 cycles of 10 s at full intensity and immediately stored at −80 °C.After complete freezing, cell homogenatewas removed from the freezerand kept at room temperature until the sampleswere thawed. After thisfreezing and thaw process, samples were centrifuged at 12,000 ×g for5 min at 4 °C and the supernatant was collected and used to assay theinhibitory effect over platelet aggregation. The two inhibitors wereused separately or in association before the eosinophil lysis.

To check the temperature stability of the active compound, theeosinophil cytosolic fraction was incubated at 100 °C for 20 min andsubsequently tested for activity over the platelet aggregation process.

Gel filtration

The active fraction was isolated using gel filtration chromatographyon Sephadex G-75. Lyophilized rat eosinophil cytosolic fraction wasdissolved in Tris–HCl (0.05 M), pH 7.2 and centrifuged at 10,000 ×gfor 5 min. The clear supernatant was applied to a Sephadex G-75 col-umn (1.5 × 24 cm) and eluted with the same buffer at a flow rate of10 ml/h. The eluted material was collected at 0.5 ml/fraction andassayed for the inhibitory activity on platelet aggregation. The activefractions were pooled and stored at −20 °C for further analysis.

Liquid chromatography analysis (RP-HPLC)

Prior to injecting the eosinophil sample into the analytical column, adesalting step (3 min at 30 μl/min with Water and 0.1% Trifluoroaceticacid; TFA) was performed using a pre-column cartridge packed withC-18 pepmap resin in line with an analytical column. Desalted samplewas loaded on a 75 μm C-18 pepmap column operating at 200 nl/min.The buffers used for separation was Water and Acetonitrile, both with

Fig. 1. Inhibition of WP-platelet aggregation. WP was incubated for 1 min with rat eosin-ophil cytosolic fraction before addition of thrombin 50–100 mU/ml. Results represent themean ± SD values of 4 independent experiments. * p b 0.001 compared with the control(thrombin) and ** p b 0.01 compared with platelet aggregation inhibition by eosinophilcytosolic fraction (6.3 × 106 cells/ml) without L-NAME.

418 A.M. Maziero et al. / Life Sciences 93 (2013) 416–422

0.1% of TFA, and for the elution, a linear gradient from 30% to 65% ofacetonitrile was used for 1 h.

Mass spectrometry

All mass spectra were acquired using a quadrupole time of flight(Q-TOF) hybrid mass spectrometer (Q-TOF Ultima, Micromass,Manchester, UK) equipped with a nano Z-spray source operating in thepositive ion mode. The ionization conditions used included a capillaryvoltage of 2.3 kV, a cone voltage and RF1 lens voltage of 30 and 100 V,

Fig. 2. Inhibition of PRP-platelet aggregation. PRP was incubated for 1 min with rat eosinophil(C), and collagen 40 μg/ml (D). Denatured cytosolic fraction was obtained by heating the at 10* p b 0.05, ** p b 0.01, *** p b 0.001 compared with the control and # p b 0.01 compared with

respectively, and collision energy of 10 eV. The source temperaturewas 70 °C and the cone gaswasN2 at aflow-rate of 80 L/h; no nebulizinggas was used to obtain sprays. Argonwas used for collisional cooling andfor fragmentation of ions in the collision cell. All spectra were acquiredwith the Q-TOF analyzer in “V-mode” (Q-TOF voltage D 9.1 kV) andthe MCP voltage set at 2150 V.

Aliquots of lyophilized RP-HPLC purified active fraction weredissolved in 10% acetonitrile with 0.1% TFA and introduced into themass spectrometer source with a syringe pump at a flow rate of500 nl/min. All masses are reported as average and were calculatedusing the “find manual” procedure or using the MassLynx-MaxEnt 1deconvolution algorithm.

Culture of HL-60 clone 15 cells

The promielocytic human HL-60 clone 15 cell lineage (HL-60 cells)was cultured as described by Zagai et al. (2004) at 0.5 to1 × 106 cells/ml in RPMI-1640 medium supplemented with 10% offetal bovine serum, 2 mM L-glutamine, 100 U/ml Penicillin, and100 μg/ml streptomycin in 5% CO2 at 37 °C. Differentiationwas inducedby adding 0.5 mM butyric acid to the medium during the maximumperiod of seven days (Fischkoff, 1988). Cell counting and viabilitywere determined using trypan blue exclusion technique.

Platelet aggregation assays using differentiated HL-60 clone 15 cells

Platelet aggregation assays were performed using cells after differ-ent periods of differentiation induced by butyric acid. It is well knownthat after 5 days of butyric acid treatment, these cells are fully differen-tiated in eosinophils (Fischkoff, 1988). Cell concentration was adjustedto 5 × 107 cells/ml and incubatedwith 10 μM indomethacin for 20 minat room temperature. This stepwas very important to inhibit the eosin-ophil cyclooxygenase activity and avoid inhibitory prostaglandin

cytosolic fraction before addition of ADP 5 μM (A), PAF 1 μM (B), arachidonic acid 1 mM0 °C for 20 min. Results represent the mean ± SD values of 4 independent experiments.of normal eosinophil cytosolic fraction (2.5 × 106 cells/ml).

Fig. 3.Gel filtration on SephadexG-75 of the rat eosinophil cytosolic fraction and inhibition of PAF-inducedplatelet aggregation. The columnwas equilibratedwith Tris–HCl 0.05 M, pH7.2.The material was eluted at 4 °C using flow of 10 ml/h and 0.5 ml samples were collected. Samples were pooled in two main fractions based on the absorbance at 280 nm, P-I and P-II(A). The fractions were tests with PAF agonist in the PRP-platelet aggregation (B).

419A.M. Maziero et al. / Life Sciences 93 (2013) 416–422

synthesis. Cells were subsequently centrifuged at 200 ×g for 5 min at20 °C and the pellet resuspended in Milli-Q water. To accomplish thecells lysis, including eosinophil granules, eosinophils suspension wassonicated in the sameway that the peritoneal eosinophil cytosolic frac-tion was performed. The final supernatant was also collected and usedto assay the inhibitory effect over the platelet aggregation.

Eosinophil cationic protein (ECP)

The pure protein from human eosinophils was purchased at concen-tration 2 mg/ml. Dilutions were performed using the buffer specific forthe protein (50 mM ammonium acetate with 0.2 NaCl buffer).

Citotoxicity MTT assay

The method was originally described by Mosmann (1983) andmodified by Xia et al. (2000).

MTT assay is a colorimetric assay for determining cellular viability onthe basis of the ability of metabolically active cells to reduce either ofthese salts in a colored formazan product.

Fifty μl of WP containing 1.2 × 108 platelets/ml was added in wellsof 96 plates. The assay was realized with activated and non-activatedplatelets with thrombin (50 mU/ml). The ECP protein was add intowells at 10 ng/ml final concentrations and incubated at 37 °C for 15and 60 min. After incubation period Krebs–Ringer solution and MTT(5 mg/ml) were added. Finally, after incubation for 3 h at 37 °C, thereaction was interrupted by adding Sodium dodecyl sulfate (SDS; 10%in HCl 0.01 M). The plate was incubated for 1 more hour at 37 °C. Theabsorbance at 570 nm of the 96 well plates was measured with aplate reader (SpectraMax 340,Molecular Devices, Sunnyvale, CA, USA.).

Statistical analysis

Data are expressed as means ± SD. The statistical significance be-tween groups was determined by using one-way ANOVA followed bythe Bonferroni test. A p value of less than 0.05 was considered statisti-cally significant.

Results

The thrombin-induced WP aggregation was dose-dependentlyinhibited by the eosinophil rat cytosolic fraction. Even with addition ofL-NAME in WP aggregation inhibition occurred, although not less thanL-NAME (Fig. 1).

Rat eosinophil cytosolic fraction inhibited the PRP platelet aggrega-tion induced by ADP (5 μM), PAF (1 μM), arachidonic acid (1 mM),and collagen (40 μg/ml) (Fig. 2). The most potent effect was observed

with PAF-stimulated platelets. The active factor present in the eosino-phil was inhibited by heating the cytosolic fraction to 100 °C for20 min (Fig. 2A). After this treatment, the inhibitory effect was drasti-cally reduced, from 76.5 ± 25.6% to 25.3 ± 6.3%.

After gel filtration of the rat eosinophil cytosolic fraction, two peakswere collected and classified with P-I and P-II (P-I shows proteins withmolecularweight larger than P-II, (Fig. 3)). The P-II fractionwas submit-ted to reverse phase chromatography, resulting chromatogram showedtwomain peaks labeled as P-II1 and PII-2 eluted at 30.2 and 25 min, re-spectively (Fig. 4). The mass spectrometry analysis of the two fractionsidentifies various polypeptide chains havingmolecular weight 15053.0,15070.0 and 13790.0 (Fig. 4). These masses are close to that of rat pro-teins: eosinophil cationic protein P70709 (ECP; 15334.43), eosinophil-derived protein P10153 (EDN; 15463.56) and Major basic proteinQ63189 (MBP; 13451.58) (Fig. 4). The intact mass measurementscannot be used alone to undoubtedly infer protein identity; furthermass spectrometric or immunological assays are required to fully char-acterize the active protein fraction.

To investigate if human eosinophils also present the platelet aggre-gation inhibition factor, we used model HL-60 cells and these cellsdifferentiated for 3 days induced 34.7 ± 12.8% inhibition. In addition,cells differentiated for 5 days inhibited 70.8 ± 11.8%, and those differ-entiated for 7 days inhibited 75.6 ± 1.8% (Fig. 5).

To verify a role of eosinophil major proteins, we test the human pureECP protein in the platelet aggregation using PRP and WP. In PRP, theECP (25 ng/ml) inhibited platelet aggregation after incubation for 15and 30 min. This inhibition was major significantly (54.3 ± 11.7%)with time incubation of 30 min (Fig. 6). In WP, the ECP (10 ng/ml)also inhibited platelet aggregation (74 ± 33.8%), and this inhibitionwas major that in PRP (Fig. 7). In both PRP and WP, the platelet aggre-gation inhibitionwith protein buffer did notwas considered statisticallysignificant (Figs. 6 and 7).

The MTT reduction assay showed that neither the short (15 min)nor the prolonged (60 min) exposure of human platelets to ECP protein(10 ng/ml) caused any toxic effect (Fig. 8).

Discussion

For the effects of eosinophils on haemostasis, a relationship has beenobserved between the presence and activity of these cells and the onsetof the coagulation cascade and platelet aggregation (Takai et al., 1992).In both cases of eosinophilia, there was a decrease of platelet aggrega-tion induced by ADP, collagen, or epinephrine.

A study carried out on a group of 168 children presenting high num-ber of eosinophils, show that the platelet count was normal, but in 8% ofpatients, the symptoms of bleeding were considered severe, while in53%, bleeding time showedmarkedly increased. The platelet aggregation

Fig. 4. Panel (A): Reverse phase liquid chromatography profile of the active fraction P-II.Deconvoluted mass spectrum obtained for fractions minutes P-II1 and P-II2 are showedin panels (B) and (C) respectively.

Fig. 5.Differentiation of HL-60 clone 15 and inhibition of PAF-induced platelet aggregationon PRP. The PRP-platelets were incubated with cell cytosolic fraction and incubated for5 min at 37 °C before induction of platelet aggregation using PAF. Results represent themean ± SD values of 3 independent experiments. * p b 0.01 compared with the respec-tive platelet aggregation inhibition in the presence of undifferentiated HL-60 clone 15cells.

Fig. 6. ECP inhibition of ADP-induced platelet aggregation on PRP. The PRP-platelets wereincubatedwith ECP and incubated for 15 and30 min at 37 °C using ADP. Results representthe mean ± SD values of 3 independent experiments. * p b 0.01 compared with the ADPcontrol.

420 A.M. Maziero et al. / Life Sciences 93 (2013) 416–422

was lower in 33% of subjects studied. Platelet aggregation induced bycollagen, ADP, and epinephrine proved to be reduced or not observed,and in the second wave of aggregation (irreversible phase), typicalaggregations were induced by ADP and epinephrine (Laosombat et al.,2001).

In another study conducted in Venezuela, platelet dysfunction asso-ciated with eosinophilia was evaluated in a group of children, who hadmoderate bleeding episodes or mucosal epithelial intense eosinophilicreaction (N650 cells/μl) and intestinal infection (Ruiz-Sáez et al.,2005). In this study, all childrenhad a platelet dysfunction characterizedby the constant absence of platelet aggregation induced by collagen

(100%). In four patients, decrease of platelet aggregation was inducedby epinephrine (66%) and ADP (66%), and the absence of the secondwave of aggregation was observed. Three children in the group showedprolonged bleeding time (N360 s).

Previous reports have mostly concentrated on the interaction ofpolymorfonuclear cells (PMNs) with platelets. This interaction maycause the increase of inflammation and platelet response, for examplearachidonic acid metabolites, serotonin (5-HT), oxygen radicals, andadenosine triphosphate (ATP) release (Maugeri et al., 1992; Schattneret al., 1990; Zatta et al., 1990).

Our results demonstrated that eosinophil cytosolic fraction showedinhibitory activity over platelet aggregation on both PRP andWP. How-ever, inhibition in PRP (PAF agonist)wasmore pronounced, 99%, than inWP (thrombin agonist) with 70% inhibition. An assay with non-stimulated PMNs (95–98% neutrophils) showed inhibition aggregationin 100% fromWP, whereaswith PRP, the inhibition effectwas less effec-tive (74% inhibition), bothwith collagen agonist (Schattner et al., 1990).PAF was able to induce inhibition aggregation platelet in leukocyte-depleted whole blood, which suggests that unstimulated PMNsmay re-lease factor(s) that inhibit platelet aggregation and β-thromboglobulinrelease (Zatta et al., 1990).

Fig. 7. ECP inhibition of Thrombin-induced platelet aggregation onWP. The WP-plateletswere incubated with ECP for 15 and 30 min at 37 °C using Thrombin. Results representthe mean ± SD values of 5 independent experiments. * p b 0.01 compared with theThrombin control.

421A.M. Maziero et al. / Life Sciences 93 (2013) 416–422

Other cell inflammation such non-stimulated mononuclear leuko-cytes (ML) also inhibit aggregation at WP induced by thrombin andcollagen (59 and 31% inhibition, respectively) (Schattner et al., 1994).

Aggregation induced by different agonists was inhibited after plate-let incubation with increasing amounts of eosinophil cytosolic fraction.Maximum inhibition effect was observed at concentration of 1.25 ×106 cells/mlwith PAF agonist, on account PAF is aweak agonist of plate-let aggregation (Pelagalli et al., 2002). Higher doses of eosinophilscytosolic fraction (5 × 106 cells/ml) were necessary to inhibit 100%aggregation when used agonist collagen. Studies with neutrophils,only 2.5 × 106 cells/ml was effective in 100% inhibition; however, forML, 16 × 106 cells/ml was necessary for inhibition when the agonistwas collagen (Schattner et al., 1990, 1994). One possible explanationis that proteins such as MBP may be present in other PMNs (Malikand Batra, 2012).

Because NO can inhibit platelet aggregation (Wallis, 2005), wetested the possibility that the inhibitory effect was due to NO secretionby eosinophil (Persson et al., 2001; Rimele et al., 1988). PMN-derivedNO is an inhibitor of platelet aggregation (McCall et al., 1989;Salvemini et al., 1989). This hypothesis was discarded first by the factthat NO is a very unstable molecule in aqueous solution. In addition,we pre-incubated the eosinophils with the L-NAME, an inhibitor of NOsynthesis. In WP, the inhibitory effect was not significantly inhibitedby previous incubation of the eosinophil with L-NAME. In our experi-ment, L-NAME at concentrations of 0.1 and 1 mMwas also added to in-hibit NO production by platelets, suggesting that the effect of inhibiting

Fig. 8. MTT assay to evaluate the cytotoxicity of ECP protein on platelets. Platelets(1.2 × 108 platelets/ml) were incubated with the protein ECP (10 ng/ml) and MTT(5 mg/ml) for 15 and 60 min in the presence or absence of thrombin (50 mU/ml). Resultsrepresent the mean ± SD the values of optical density (570 nm) of 3 independent exper-iments in duplicate.

the cytosolic fraction does not occur by nitric oxide pathway; however,further studies are necessary to elucidate the mechanism of inhibition.

Heating of the substances contained in the eosinophil cytosolicfraction at 100 °C did not cause inhibition of platelet aggregation,demonstrating the involvement of thermosensitive proteins in thisprocess.

Themass spectrometry analysis suggests the existence of known eo-sinophil cationic protein in the active cytosolic fraction (P-II). Is knownfrom literature that these proteins shows various forms of glycosylation,(Kita, 2011; Peterson et al., 1988; Rosenberg and Tiffany, 1994) and thisdifference can be critical in regulating their biological functions (Boixet al., 2001; Venge et al., 1999).

HL-60 cells lineage differentiate primarily to eosinophils instead ofneutrophils when cultured with butyric acid, if they have previouslybeen cultured under alkaline conditions (pH 7.6) (Fischkoff, 1988). Itis well known that HL-60 cells cultured in the presence of 0.5 mM bu-tyric acid for 5 days leads to the differentiation process and dramaticallyincreases the expression of the eosinophil peroxidase (EPO) and themajor basic protein MBP genes, which is similar to that obtained withthe eosinophilic inducer IL-5 or interferon-gamma (IFN-γ; 100 U/ml)and tumor necrosis factor-alpha (TNF-α; 1000 U/ml) (Lopez et al.,2003).

RT-PCR, using specific primer for EPO and MBP, demonstrated thatduring differentiation with 0.5 mM butyric acid for 7 days, the mRNAexpressions of these proteins dramatically increased (Lopez et al.,2003).

Our results demonstrated that HL-60 cells were effectively differen-tiated in eosinophil, and the expression of cationic proteins was signifi-cantly increased.

The kinetics of HL-60 cells differentiation to eosinophils was parallelto the increase in platelet inhibition effect, strongly indicating that theappearance of specific eosinophil proteins is responsible for this activity.The activity of EPOwas discarded because itsmolecularweight does notfit with that of the proteins contained in the active fraction. Whetherthis inhibitory effect of these proteins may explain the low plateletaggregation observed in the nonthrombocytopenic purpurawith eosin-ophilia remains to be further evaluated (Mitrakul, 1975; Suvatte et al.,1979).

To confirm that one of the eosinophil cationic proteins may beresponsible for the inhibition effect of platelet aggregation, we usedthe ECP protein in the aggregation tests. Our results demonstratedthat the protein ECP causes a platelet aggregation inhibition on PRP,and this inhibition is time dependent. This percent inhibition, 54.33 ±11.68%, is less than that observed with the eosinophil cytosolic fraction,most probably because the ADP is considered a strong platelet agonist,while PAF is considered a weak platelet agonist.

In WP, the platelet inhibition was higher than in PRP, 74 ± 33.8%,and this inhibition also occurred in a time dependent manner.

We decided to check if the effect of platelet aggregation inhibitioncould be occurring because of the cytotoxic protein ECP, for this reasonwe assessed cell viability using theMTT assay.We found that the inhibi-tion effect is not related to the death of platelets.

Other authors demonstrated that a positive association existsbetween platelet activation and eosinophil activation in human airwaysfrompatientswith asthma (Benton et al., 2010). In this study, they dem-onstrated a significant positive association between p-selectin and ECP,both markers of platelets and eosinophils, respectively. The concentra-tion of ECP protein found in nasal wash samples was 11.6 ng/ml. Thisconcentration is very close to that used to test inhibition of plateletaggregation in WP (Benton et al., 2010).

The MBP and EPO proteins have been described as strong plateletagonists inducing granules secretion (Rohrbach et al., 1990). These pro-teins were classified as platelet agonists with a distinct mechanism ofactivation. Both MBP and EPO evoked a dose-dependent nonlyticsecretion of platelet 5-hydroxytryptamine (5-HT) in unstirred plateletsuspensions even in the presence of 10 μM indomethacin. EDN and

422 A.M. Maziero et al. / Life Sciences 93 (2013) 416–422

ECP released low amounts of 5-HT, and this observation suggests thatthe release of 5-HT is not due to the highly cationic property of theseproteins, because the EDN and ECP have pI essentially identical(Rohrbach et al., 1990).

This work together with results reported by other authors highlightthe contribution of eosinophil cationic proteins in the regulation ofplatelet aggregation. Further studies are required to characterize theaction mechanisms. This can help in the treatment of patients witheosinophilia, who presents a clear evidence of changes in plateletclotting.

Conclusion

In conclusion,we described a new eosinophil inhibitory activity overplatelet aggregation. Our results show new evidence of the relationshipbetween eosinophils and platelets. They also demonstrated that a pro-tein of eosinophils, ECP, can be potentially responsible for the inhibitionof platelet aggregation. Further investigations are still required toelucidate the mechanisms involved in platelet aggregation inhibition.

Conflict of interest statement

The authors declare no conflict of interest.

Acknowledgments

The authors are grateful to Fundação de Amparo à Pesquisa doEstado de São Paulo (FAPESP) for providing financial support.

We thank Prof. Dr. Antonio Condino Neto and the biologist JussaraRehder (from CIPED, State University of Campinas, SP, Brazil) forproviding the HL-60 clone 15 cells lineage.

References

Abu-Ghazaleh RI, Kita H, Gleich GJ. Eosinophil activation and function in health anddisease. Immunol Ser 1992;57:137–67.

Benton AS, Kumar N, Lerner J, Wiles AA, Foerster M, Teach SJ, et al. Airway plateletactivation is associated with airway eosinophilic inflammation in asthma. J InvestigMed 2010;58(8):987–90. [Dec].

Boix E, Carreras E, Nikolovski Z, Cuchillo CM, Nogués MV. Identification and characteriza-tion of human eosinophil cationic protein by an epitope-specific antibody. J LeukocBiol 2001;69(6):1027–35. [Jun].

Chin NS, Koong PL. A study of 31 patients with easy bruising from University Hospital,Kuala Lumpur. Med J Malaysia 1990;45(4):325–8. [Dec].

Donato JL, Marcondes S, Antunes E, Nogueira MD, Nader HB, Dietrich CP, et al. Role ofchondroitin 4-sulphate as a receptor for polycation induced human platelet aggrega-tion. Br J Pharmacol 1996;119(7):1447–53. [Dec].

Egesten A, Calafat J, Janssen H, Knol EF, Malm J, Persson T. Granules of human eosinophilicleucocytes and their mobilization. Clin Exp Allergy 2001;31(8):1173–88. [Aug].

Fischkoff SA. Graded increase in probability of eosinophilic differentiation of HL-60promyelocytic leukemia cells induced by culture under alkaline conditions. LeukRes 1988;12(8):679–86.

Gleich GJ, Adolphson C. Bronchial hyperreactivity and eosinophil granule proteins. AgentsActions Suppl 1993;43:223–30.

Gleich GJ, Adolphson CR, Leiferman KM. The biology of the eosinophilic leukocyte. AnnuRev Med 1993;44:85–101.

Hathirat P, Bintadish P, Isarangkura P, Sasanakul W, Chiewsilp P, Ratanasirivanich P.Platelet transfusion in acquired platelet dysfunction with eosinophilia. SoutheastAsian J Trop Med Public Health 1982;13(2):220–4. [Jun].

Hathirat P, Isarangkura P, Pintadit P, Chuansumrit A, Mahaphan W. Immunoglobulins inacquired platelet dysfunction with eosinophilia. Southeast Asian J Trop Med PublicHealth 1993;24(Suppl. 1):206–9.

Kita H. Eosinophils: multifaceted biological properties and roles in health and disease.Immunol Rev 2011;242(1):161–77. [Jul].

Laosombat V, Wongchanchailert M, Sattayasevana B, Kietthubthew S, Wiriyasateinkul A.Acquired platelet dysfunction with eosinophilia in children in the south of Thailand.Platelets 2001;12(1):5–14. [Feb].

Lee AC. Unusual hematologic disease affecting Caucasian children traveling to SoutheastAsia: acquired platelet dysfunctionwith eosinophilia. Hematol Rep 2012;4(1). [Jan 2].

Lim SH, Tan CE, Agasthian T, Chew LS. Acquired platelet dysfunction with eosinophilia:review of seven adult cases. J Clin Pathol 1989;42(9):950–2. [Sep].

Lopez JA, Newburger PE, Condino-Neto A. The effect of IFN-gamma and TNF-alpha on theeosinophilic differentiation and NADPH oxidase activation of human HL-60 clone 15cells. J Interferon Cytokine Res 2003;23(12):737–44. [Dec].

Lucas GN, Seneviratne TR. A case of acquired platelet dysfunction with eosinophilia.Ceylon Med J 1996;41(3):122. [Sep].

Malik A, Batra JK. Antimicrobial activity of human eosinophil granule proteins: involve-ment in host defence against pathogens. Crit Rev Microbiol 2012;38(2):168–81.[May].

Maugeri N, Evangelista V, Piccardoni P, Dell'Elba G, Celardo A, de Gaetano G, et al.Transcellular metabolism of arachidonic acid: increased platelet thromboxane gener-ation in the presence of activated polymorphonuclear leukocytes. Blood 1992;80(2):447–51. [Jul 15].

McCall TB, Boughton-Smith NK, Palmer RM, Whittle BJ, Moncada S. Synthesis of nitricoxide from L-arginine by neutrophils. Release and interaction with superoxideanion. Biochem J 1989;261(1):293–6. [Jul 1].

Mitrakul C. Transient, spontaneous bruising with long bleeding time and normal plateletcount. Am J Clin Pathol 1975;63(1):81–6. [Jan].

Mosmann T. Rapid colorimetric assay for cellular growth and survival: application toproliferation and cytotoxicity assays. J Immunol Methods 1983;65(1-2):55–63.[Dec 16].

Muthiah MM, Mitchell TR, Thorn CT. Acquired platelet dysfunction with eosinophilia. BrMed J (Clin Res Ed) 1984;289(6451):1044. [Oct 20].

Pelagalli A, Lombardi P, d'AngeloD, DellaMorte R, Avallone L, StaianoN. Species variabilityin platelet aggregation response to different agonists. J Comp Pathol 2002;127(2-3):126–32. [Aug-Oct].

Persson T, Andersson P, Bodelsson M, Laurell M, Malm J, Egesten A. Bactericidal activity ofhuman eosinophilic granulocytes against Escherichia coli. Infect Immun 2001;69(6):3591–6. [Jun].

Peters MS, Rodriguez M, Gleich GJ. Localization of human eosinophil granule majorbasic protein, eosinophil cationic protein, andeosinophil-derived neurotoxin byimmunoelectron microscopy. Lab Invest 1986;54(6):656–62. [Jun].

Peterson CG, Jörnvall H, Venge P. Purification and characterization of eosinophilcationic protein from normal human eosinophils. Eur J Haematol 1988;40(5):415–23. [May].

Poon MC, Ng SC, Coppes MJ. Acquired platelet dysfunction with eosinophilia in whitechildren. J Pediatr 1995;126(6):959–61. [Jun].

RamanathanM, Duraisamy G. Acquired platelet dysfunction with eosinophilia (APDE): anunderrecognised condition. Med J Malaysia 1987;42(1):53–5. [Mar].

Rimele TJ, Sturm RJ, Adams LM, Henry DE, Heaslip RJ, Weichman BM, et al. Interaction ofneutrophils with vascular smooth muscle: identification of a neutrophil-derivedrelaxing factor. J Pharmacol Exp Ther 1988;245(1):102–11. [Apr].

Rohrbach MS, Wheatley CL, Slifman NR, Gleich GJ. Activation of platelets by eosinophilgranule proteins. J Exp Med 1990;172(4):1271–4. [Oct 1].

Rosenberg HF, Tiffany HL. Characterization of the eosinophil granule proteins recognizedby the activation-specific antibody EG2. J Leukoc Biol 1994;56(4):502–6. [Oct].

Ruiz-Sáez A, Sifontes LN, Feijoo R, Certad G, Arenas-Pinto A, Pocaterra L, et al. Plateletdysfunction-eosinophilia syndrome in parasitized Venezuelan children. Am J TropMed Hyg 2005;73(2):381–5. [Aug].

Salvemini D, de Nucci G, Gryglewski RJ, Vane JR. Human neutrophils and mononuclearcells inhibit platelet aggregation by releasing a nitric oxide-like factor. Proc NatlAcad Sci U S A 1989;86(16):6328–32. [Aug].

Schattner MA, Geffner JR, Isturiz MA, Lazzari MA. Inhibition of human platelet activationby polymorphonuclear leukocytes. Br J Pharmacol 1990;101(2):253–6. [Oct].

Schattner MA, Finiasz MR, Notrica JA, Lazzari MA. Platelet aggregation inhibition bymononuclear leukocytes. Thromb Res 1994;73(3–4):205–14. [Feb 15].

Suvatte V, Mahasandana C, Tanphaichitr V, Tuchinda S. Acquired platelet dysfunctionwith eosinophilia: study of platelet function in 62 cases. Southeast Asian J TropMed Public Health 1979;10(3):358–67. [Sep].

Takai Y, Kawada K, Kohriyama K. Abnormalities of coagulation, fibrinolysis andkallikrein-kinin system in 2 cases with remarkable eosinophilia. Rinsho Ketsueki1992;33(3):288–96. [Mar].

Vadas MA, David JR, Butterworth A, Pisani NT, Siongok TA. A new method for the purifi-cation of human eosinophils and neutrophils, and a comparison of the ability of thesecells to damage schistosomula of Schistosoma mansoni. J Immunol 1979;122(4):1228–36. [Apr].

Venge P, Byström J, Carlson M, Hâkansson L, Karawacjzyk M, Peterson C, et al. Eosinophilcationic protein (ECP): molecular and biological properties and the use of ECP as amarker of eosinophil activation in disease. Clin Exp Allergy 1999;29(9):1172–86.[Sep].

Wallis JP. Nitric oxide and blood: a review. Transfus Med 2005;15(1):1–11. [Feb].Xia Y, Li J, Bertino A, Kuter DJ. Thrombopoietin and the TPO receptorduring platelet

storage. Transfusion 2000;40(8):976–87. [Aug].Zagai U, Sköld CM, Trulson A, Venge P, Lundahl J. The effect of eosinophils on collagen gel

contraction and implications for tissue remodelling. Clin Exp Immunol 2004;135(3):427–33. [Mar].

Zatta A, Prosdocimi M, Bertelé V, Bazzoni G, Del Maschio A. Inhibition of platelet functionby polymorphonuclear leukocytes. J Lab Clin Med 1990;116(5):651–60. [Nov].