Thrombosis Research 127 (2011) 551–559

Contents lists available at ScienceDirect

Thrombosis Research

j ourna l homepage: www.e lsev ie r.com/ locate / th romres

Regular Article

Investigation into the mechanism(s) of antithrombotic effects of carbon monoxidereleasing molecule-3 (CORM-3)

Hitesh Soni a,b, Mukul Jain a, Anita A. Mehta b,⁎a Zydus Research Centre (ZRC communication No.# 331), Sarkhej-Bavla N.H 8A Moraiya, Ahmedabad-382210, Indiab Department of Pharmacology, L.M.College of Pharmacy, Navarangpura, Ahmedabad-380009, India

⁎ Corresponding author at: Department of PharmacolOpp. Gujarat University, Navarangpura, Ahmedabad, Gu9428418611(Mobile); fax: +91 79 26304865.

E-mail address: dranitalmcp@gmail.com (A.A. Mehta

0049-3848/$ – see front matter © 2011 Elsevier Ltd. Aldoi:10.1016/j.thromres.2011.02.009

a b s t r a c t

a r t i c l e i n f o

Article history:Received 23 November 2010Received in revised form 18 January 2011Accepted 10 February 2011Available online 4 March 2011

Keywords:Carbon monoxideCORM-3PlateletRatSoluble guanylate cyclaseNitric oxideAntithrombotic

Carbon monoxide (CO) like nitric oxide (NO) has been recognized as activator of soluble guanylate cyclase(sGC) in many physiological functions. Studies, which demonstrate the mechanisms by which CO inhibitsplatelet aggregation in in vivo models, are few. Here we investigated the possible involvement of sGC, NO,plasminogen activator inhibitor (PAI-1) and p38 MAP Kinase in antithrombotic effects of CO released by anovel, water-soluble, CO releasing molecule-3 (CORM-3) using rat. The effects of CORM-3 on in vitro andex vivo platelet aggregation induced by thrombin as well as in in vivo thrombosis models were studied. Whenadded to rat washed platelets in in vitro study, CORM-3 (100 and 200 μM) inhibited thrombin-inducedplatelet aggregation. Similarly, antiplatelet effect was also observed when 3 mg/kg i.v. infusion of CORM-3administered for 10 minutes in ex vivo study using rat. Interestingly, in presence of inhibitor of sGC (ODQ,10 mg/kg, i.p.) and inhibitor of nitric oxide synthase (L-NAME, 30 mg/kg, i.p.), inhibition of thrombin-inducedaggregation by CORM-3 was significantly blocked. Notably, in presence of inhibitor of KATP channel(glibenclamide, 10 mg/kg, i.p.) and p38 MAP Kinase (SCIO-469, 1 mg/kg, i.p.), inhibition of aggregation byCORM-3 was not blocked. In in vivo studies using animal models of thrombosis, we found that CORM-3-mediated antithrombotic effect was dependent on activation of sGC, NO and suppression of PAI-1 in arterialthrombosis and Arterio-Venous (A-V) shunt models. Therefore, we concluded that antithrombotic activity ofCORM-3 may be mediated by activation of sGC, NO and inhibition of PAI-1.

ogy, L. M. College of Pharmacy,jarat, 380 009, India. Tel.: +91

).

l rights reserved.

© 2011 Elsevier Ltd. All rights reserved.

1. Introduction

The use of CO as a potential therapeutic agent has emerged slowlydue to its negative connotation as a toxic gas for mammalians sincedecades [1]. In the last few years' research on CO for the regulation ofmany physiological processes has been reported. Endogenously, CO isproduced as a by-product during breakdown of heme moiety and thisreaction is facilitated by the enzyme heme oxygenase (HO). HO existsmainly in inducible, HO-1 and constitutive, HO-2 forms. Hemebreakdown by HO-1 produces bilirubin, iron (Fe++) and CO, out ofwhich CO has been recognized as cytoprotective gas and is mimickingthe role of HO-1 in many pathophysiological conditions [2]. Althoughthemechanism(s) underlying the cytoprotective actions of CO has notbeen elucidated, evidence suggests that this gas exerts some of itseffects via activation of the guanylate cyclase/ cGMP pathway [3,4]. Awater-soluble form of CO-releaser, tricarbonyldichloro (glycinato)ruthenium (II) (CORM-3), has been developed and demonstrated as

cardioprotective agent [5,6]. Recently we have reported that cardio-protection by CORM-2 (a lipid soluble fast CO-releaser), is highlyconcentration-dependent, independent of coronary endothelium andcardioprotective effect might be attributed to the activation of KATP

channel present on vascular smooth muscle cell (VSMC) [7].Earlier study using CO as gas showed that exposure of minipigs to

low concentration (160 and 185 ppm) CO significantly increasedplatelet aggregation. Elevating CO concentration to 420 ppm showedadhesions of shape-changed platelets on the arterial endothelium thatwas revealed under scanning electron microscopy [8]. However, laterstudies showed inhibition of platelet aggregation by CO. It has beenreported that CO inhibited release of ADP and serotonin from platelet[9]. Both endogenously derived and exogenously applied CO inhibitsplatelet aggregation by stimulating the activation of sGC [10,11].Emerging studies indicated that CO may also exert importantprotection against thrombosis. Further, COmitigates platelet adhesionto endothelium in response to inflammation [12]. Furthermore, COinhibits platelet aggregation and thrombosis following organ trans-plantation, andmay contribute to the inhibition of platelet-dependentthrombosis following the induction of HO-1 in a rodent artery injurymodel [13,14]. Moreover, inhalation of CO rescues mice from lethalischemic injury by preventing microvascular thrombosis and theaccumulation of fibrin [15]. Recently, it has been observed that the

552 H. Soni et al. / Thrombosis Research 127 (2011) 551–559

absence of HO-1 in aortic allograft recipient mice resulted in 100%mortality within 4 days due to arterial thrombosis. In contrast,recipient mice normally expressing HO-1 showed 100% graft patencyand survival [16].

Transition metal carbonyls have been shown to be a safe andeffective means of transporting and releasing CO groups. Recent studyshowed that CORM-3, a water-soluble CO-releaser inhibited humanplatelets by a mechanism independent of sGC [17]. The water-solubleproperties of CORM-3 (a fast CO releaser) suggest that this compoundmay have clinical utility. Accordingly, the goal of the present study wasto determine whether the antiplatelet effects of CORM-3 demonstratedin vitro are also present exvivo and in vivo. To this end,weutilized awell-established rat model of arterial thrombosis (platelet rich thrombus),artrio-venous thrombosis (mixed thrombus) and venous thrombosis(RBC rich thrombus) in which fundamental physiological variables thatmodulate thrombosis were carefully monitored and controlled. Wetested the hypothesis that CORM-3 may prevent agonist-inducedplatelet aggregation in vitro. Furthermore we tested the role of sGC,NO, KATP channel and MAPK in CORM-3-mediated antiplatelet activity.To study antiplatelet potential of CORM-3 and its mechanism(s), weemployed in vitro and ex vivo experiments using washed rat platelets.Further, we extended our observations into in vivo animal models ofthrombosis and mechanism(s).

2. Materials and methods

2.1. Animals

Male wistar rats (250–300 g body weight) were used in the study.The animals were kept in individually ventilated cages in a roomwith controlled temperature (23±2 °C), lighting (12:12 h light–darkcycle) and relative humidity (55±10%). Animals had free access tostandard rat chow and water. The protocol for use of animals forconducting these experiments has been reviewed and approved bythe Institutional Animal Ethics Committee (IAEC).

2.2. Chemicals

CORM-2 was obtained from Sigma chemical. CORM-3 wassynthesized and iCORM-3 (inactive CORM-3) was prepared asdescribed previously [6]. Thrombin was also purchased from SigmaChemical, USA and prepared in phosphate buffer (pH 7.4). Urethanewas purchased from Sigma Chemicals, USA. FeCl3 was purchased fromHi-Media, India. Clopidogrel bisulfate was the generous gift fromZydus Cadila, India. In all experiments, CORM-3 was dissolved inwater for injection. ODQ, glibenclamide and SCIO-469 were formu-lated in DMSO (0.5%)+saline. L-NAME, Clopidogrel and CORM-3were prepared in water for injection. CORM-3 was inactivated(termed as iCORM-3) by dissolving it in PBS and leaving it at roomtemperature for 24 h [6].

2.3. Preparation of rat washed platelets for aggregation studies

Rat blood samples were colleted from retro-orbital route underlight ether anesthesia in to the tubes containing 3.8% trisodium citrate.All blood samples were centrifuged at 200Xg for 20 min and plateletrich plasma (PRP) was carefully collected. PRP was again centrifugedat 800Xg for 10 minutes and supernatant was removed. The plateletpellet was resuspended and washed three times in CGS (0.12 Msodium citrate, 0.1 M dextrose, 0.1 M Nacl, pH 6.5). The finalresuspension was in modified Tyrode's buffer (138 mM Nacl,2.9 mM Kcl, 12 mM NaHCO3, 0.4 mM MgCl2, 5.5 mM dextrose,0.36 mM NaH2PO4, 1.8 mM CaCl2, pH 7.4). The platelet count,conducted on a Cell Dyn 3700 (Abbott Diagnostics), was adjustedwith Tyrode's buffer to 3×108 / ml. Platelet activation inhibitors werenot used during platelet isolation due to their potential effect on

signaling pathways involved in shear-induced platelet activation.Washed Platelet (WP) suspension was allowed to rest for 30 min at37 °C before conducting experiments.

2.4. Measurement of platelet aggregation using thrombin as aggregatingagent

Platelet aggregation studies were performed on a SpectraMax 190microplate reader in 96-well, flat-bottomed, micro titer plates [18]using the SOFTmax Pro data acquisition software (Molecular DevicesCorp., California, USA). A 180-μl volume of WP was placed in eachwell, followed by addition of 20 μl of thrombin. For in vitro studies,WPs were incubated with various concentrations of CORM-3 (50, 100and 200 μM) for 2 minutes at 37 °C before addition of thrombin.Readings were taken every 1-minute over a 5-minute period at 405-nmwavelength. During the run time, the plate was incubated at 37 °Cand was shaken vigorously in a shaking mode at the maximal speedavailable. All platelet aggregation studies were performed in triplicate.Change in optical density (OD) was measured by taking OD of bufferas blank. Aggregations were performed using a modest concentrationof thrombin (0.5 IU/ml for rat platelets). % Aggregation was calculatedusing formula:

% Aggregation = Initial OD−Final ODð Þ= Initial OD½ �⁎100

2.5. Experimental protocol for ex vivo platelet aggregation study in rats(n=10)

Wistar rats were divided randomly on body weight basis in toseven groups as follows:

Group 1: Vehicle treated (0.5 ml/kg/min, i.v. infusion for 10 minutes,0.5% DMSO+Saline).Group 2: i CORM-3 (3 mg/kg/min, i.v.) for 10 minutes.Group 3: CORM-3 (3 mg/kg/min, i.v.) for 10 minutes.Group 4: ODQ (10 mg/kg, i.p.) before 30 min.+ CORM-3 (3 mg/kg/min, i.v.) for 10 minutes.Group 5: L-NAME (30 mg/kg, i.p.) before 30 min.+ CORM-3 (3 mg/kg/min, i.v.) for 10 minutes.Group 6: Glibenclamide (10 mg/kg, i.p.) before 30 min. + CORM-3(3 mg/kg/min, i.v.) for 10 minutes.Group 7: SCIO-469 (1 mg/kg, i.p.) before 30 min. + CORM-3 (3mg/kg/min, i.v.) for 10 minutes.

Blood samples were obtained through the retro-orbital routeunder anesthesia in tubes containing 3.8% trisodium citrate. Washedplatelets were prepared as described above and samples weresubjected to thrombin-induced platelet aggregation (0.5 IU/ml)assay. % Aggregation was calculated.

2.6. FeCl3-induced arterial thrombosis model in rats

Animals (n=10) were treated as per given protocol and thensubjected to FeCl3-induced arterial thrombosis. FeCl3 -inducedchemical injury was used as a model of arterial thrombosis aspreviously described [19]. Briefly, rats were anaesthetized withurethane (1.25 g/ kg, i.p.). A midline cervical incision was made onthe ventral side of the neck, and left carotid artery was isolated. A2×3 mmstrip ofWhatman filter paper No.#1 saturatedwith 35% (w/v)FeCl3 was kept on the carotid artery for 5 min. A temperature probe(Thermalert-TH8, Physitemp Instruments Inc., Clifton, N.J., USA) wasplaced distal to Whatman filter paper to monitor the temperature ofcarotid artery. A sudden fall in temperature (about 2 °C)was taken as anindication of cessation of blood flow as a consequence to thrombusformation. Time to occlusion (TTO) was defined as the time from FeCl3

Fig. 1. Effects of CORM-3 on thrombin-induced platelet aggregation using rat platelets.Platelet suspension was incubated for 2 min at 37 °C without CO-releaser (i CORM-3) orwith CO-releaser (CORM-3) before stimulation with thrombin (0.5 IU/ml). @ pb0.05,#pb0.01 Vs Vehicle control group.

553H. Soni et al. / Thrombosis Research 127 (2011) 551–559

application to time of thrombus formation. A cutoff time was fixed at60 min in case no thrombus formation was seen in drug-treatedanimals. Assessment of wet thrombus weight, plasma ProthrombinTime (PT) and plasma PAI-1 levels were also performed. Clopidogrelbisulfate (30 mg/kg p.o.) before 2 hour of blood collection (used aspositive control).

2.7. Assessment of Prothrombin Time (PT) in plasma

PT (Seconds) was measured from plasma samples obtained fromtreated animals by using commercially available kit Thromborel-S(Dade behring, Germany) using 4-channel coagulometer (Sysmex CA-50, Japan).

Fig. 2. Effects of CORM-3 on thrombin (0.5 IU/ml)-induced % change in ex vivo platelet aggr30 min prior to CORM-3 administration. * pb0.001 and @ pb0.05 Vs Vehicle control.

2.8. Assessment of active rat PAI-1 levels in plasma

Active rat PAI-1 levels were detected by ELISA kit as permanufacturer's instruction (Molecular Innnovations, USA) and levelsof PAI-1 (ng/ml) were calculated from standard curve.

2.9. Thrombus assessment in rat Arterio-Venous (AV) shunt model

Animals (n=10) were treated as per given protocol and thensubjected to A-V shunt model of thrombosis as previously de-scribed [20]. Briefly, wistar rats were anaesthetized with urethane,jugular vein and contralateral carotid artery were cannulated with12.5 cm long PE-10 polyethylene tube. Heparinised saline filledshunt was assembled by connecting the cannulae with 6 cm longPE-20 tubing containing 5 cm long 4–0 silk thread. The blood wasallowed to flow though the shunt for 5 min. The thread with as-sociated thrombus was removed and thrombus weight was cal-culated. The rat A-V shunt thrombosis model has been described asa ‘mixed’ thrombosis model. Thrombosis is initiated by plateletadherence to silk thread anchored in shunt. Clopidogrel was used ascomparator.

2.10. Partial stasis combined with FeCl3-induced vessel injury in venousthrombosis model using rat

Male Wistar rats (n=10) were anesthetized with urethane(1.25 g/kg, i.p.). The abdomen was opened by making an incisionalong the linea alba towards the sternum, followed by exposition ofthe posterior vena cava. Partial stasis was induced in the posteriorvena cava by tying a cotton thread together with a blunt needle (21 G,BD) just caudally of the junction of the posterior vena cava and leftrenal vein. The needle was then removed. A round piece (5 mm) ofWattmann #1 filter paper saturated with 7 μl of 6% w/v ferric chloridesolution was then applied to the external surface of the posterior venacava caudally of the stenosis for 5 min and then removed. Warmsaline was sprayed over tissues, and muscle layer and skin wereprovisionally closed. One hour after removal of the filter paper,

egation after 10 min. of i.v. infusion in male wistar rats (n=10). Inhibitors were used

Fig. 3. Effects of CORM-3 on time to carotid artery occlusion (TTO) after 10 min. of i.v. infusion using FeCl3-induced arterial thrombosis in male wistar rats (n=10). Inhibitors wereused 30 min prior to CORM-3 administration. * pb0.001 and @ pb0.05 Vs Vehicle control.

554 H. Soni et al. / Thrombosis Research 127 (2011) 551–559

ligatures were applied near the bifurcation of the posterior vena cavaand around all side branches of the ligated posterior vena cavasegment. The ligated venous segment was excised, the thrombusremoved, blotted of excess blood and immediately weighed [21,22].

2.11. Tail-vein bleeding time (TVBT) in rat

CORM-3 was administered as 3-mg/kg/min. i.v. infusion for10 minutes to male Wistar rats (n=10) via femoral vein. The tailwas transected at 5 mm from tip using surgical blade under thio-pental (50 mg/kg, i.p.) anesthesia. Transected tail was immersedvertically in saline at 37 °C. The time until continuous blood flow

Fig. 4. Effects of CORM-3 on % reduction in thrombusweight after 10 min. of i.v. infusion inm* pb0.001 Vs i CORM-3 group.

ceased for N30 s was measured, with a maximum observation time of30 min (longer bleeding times were assigned a value of 30 min).Clopidogrel (30 mg/kg, p.o., 120 min pretreatment time) was used ascomparator [23].

2.12. Statistical analysis

Results were expressed as mean±SEM. Data were analyzed byOne-way ANOVA followed by Tukey's multiple comparison tests. Allanalysis was done using GraphPad Prism software version 4.0. Pb0.05was considered to be statistically significant.

alewistar rats (n=10). Inhibitors were used 30 min prior to CORM-3 administration.

Fig. 5. Effects of CORM-3 on Prothrombin Time (PT) after 10 min. of i.v. infusion using FeCl3-induced arterial thrombosis in male wistar rats (n=10). Inhibitors were used 30 minprior to CORM-3 administration.

555H. Soni et al. / Thrombosis Research 127 (2011) 551–559

3. Results

3.1. Effects of CORM-3 on thrombin-induced in vitro and ex vivo plateletaggregation using rat washed platelets and mechanisms

In in vitro study, CORM-3 produced significant reduction in %platelet aggregation using rat platelets at 100 and 200 μM concentra-tions when thrombin (0.5 IU/ml) was used as an aggregatory agent(Fig. 1). Further, we examined the mechanism of CORM-3-mediatedantiplatelet effect using thrombin as an aggregating agent and ex-plored the role of sGC, NO, KATP channel and p38 MAPK using

Fig. 6. Effects of CORM-3 on PAI-1 levels after 10 min. of i.v. infusion using arterial thrombadministration. # indicates pb0.01 Vs vehicle control.

pharmacological inhibitors in ex vivo platelet aggregation studies inrats. Administration of CORM-3 (3 mg/kg/min, i.v. infusion) for 10 minproduced significant reduction in % platelet aggregation (35.80±3.6,pb0.001) as compared to vehicle treated group (65.20±4.28, n=10).Pretreatment with ODQ (10 mg/kg, i.p.) attenuated the aggregatoryresponse (59.70±3.77). Similarly, L-NAME administration (30 mg/kg, i.p.) partially abolished the antiaggregatory response by CORM-3(49.10±3.68, pb0.05). On the other hand, administration of glib-enclamide (10 mg/kg, i.p.) and SCIO-469 (1 mg/kg, i.p.) did not affectthe inhibition of platelet aggregation by CORM-3. Clopidogrel (30 mg/kg, p.o.) was used as positive control in this experiment (Fig. 2).

osis model in male wistar rats (n=10). Inhibitors were used 30 min prior to CORM-3

Fig. 7. Effects of CORM-3 on % reduction in thrombus weight after 10 min. of i.v. infusion in A-V shunt model using male wistar rats (n=10). Inhibitors were used 30 min prior toCORM-3 administration. * pb0.001, # pb0.01, @pb0.001 Vs i CORM-3 group.

Fig. 8. Effects of CORM-3 on Venous Thrombus Wet Weight using Partial stasiscombined with FeCl3-induced venous thrombosis model in rat after 10 min. of i.v.infusion (n=10).

556 H. Soni et al. / Thrombosis Research 127 (2011) 551–559

3.2. Effects of CORM-3 using in vivo models of thrombosis in rat andunderlying mechanisms

Subsequently, we also explored the above observations of ex vivostudy in in vivo models of thrombosis in rats. CORM-3 (3 mg/kg/min,i.v.) for 10 minutes showed increase in time to occlusion (TTO) inFeCl3-induced arterial thrombosis model. As shown in Fig. 3, thiseffect was inhibited by treatment with ODQ (10 mg/kg, i.p.) andpartially attenuated by L-NAME (30 mg/kg, i.p.). Similar observationswere found when reduction in arterial thrombus wet weight wasconsidered as an end point by CORM-3 (Fig. 4). We also found thatthere was no significant change in PT in all groups, which suggeststhat CORM-3 is not interfering with coagulation pathway (Fig. 5). Wealso measured plasma PAI-1 levels using ELISA assay and observedthat CORM-3 treatment causes significant reduction in plasma PAI-1level and this effect was attenuated when animals were treated withODQ or L-NAME (Fig. 6).

Another secondary model was rat A-V shunt model in which weobserved the significant (pb0.001) reduction in thrombus weightwhen animals were infused with CORM-3 (3 mg/kg/min, i.v.) for10 min. However, this effect was significantly reducedwhen rats werepretreated with ODQ or L-NAME (Fig. 7).

As indicated in Fig. 8, there was nonsignificant reduction of venousthrombus wet weight after treatment with CORM-3 (3 mg/kg/min, i.v.)for 10 min. Therefore, we did not consider for further mechanisticstudies.

Since bleeding profile plays a major role in safety assessment ofany antiplatelet or anticoagulant agents, we also investigated bleedingpotential of CORM-3 using TVBT model in rat. CORM-3 (3 mg/kg/min,i.v.) for 10 min, produced significant (pb0.05) bleeding in rats ascompared to vehicle control group but nonsignificant as compared to iCORM-3 group. Further, bleeding profile of CORM-3 was significantlylower (pb0.001) as compared to clopidogrel (comparator) (Fig. 9).

4. Discussion

Many clinical studies have demonstrated that increased plateletaggregation plays an important role in the pathogenesis of variouscardiovascular and thromboembolic diseases. [24]. NO, synthesizedfrom L-arginine by NOS, which then activates intracellular sGC and

subsequent cyclic GMP formation, plays an importantmodulatory rolein many physiological and pathological conditions [25]. CO, like NO,can activate sGC/cGMP pathway, which in turn elicits vasodilation andinhibits platelet activation and smooth muscle cell proliferation [10].Lindenblatt L et al. [26] have reported that vascular HO-1 with releaseof CO, particularly of bilirubin, attenuates thrombus formation, mostprobably via modulation of P-selectin expression on endothelial cells.Thus, local induction of HO-1 activity may be of preventive andtherapeutic value for clinical disorders with increased risk ofthrombotic events. The first line of evidence supporting the anti-thrombotic effect of HO-1 in vivo is the observation that COsuppressed vascular thrombosis occurring during the cardiac graftrejection likely through inhibiting platelet aggregation [13,27].Subsequently, there was a study showing that CO protected ischemiclung injury through down regulating the expression of PAI-1, theprincipal regulator of the fibrinolytic system, in macrophages and

Fig. 9. Effects of CORM-3 on bleeding time (Seconds) after 10 min. of i.v. infusion inmalewistar rats (n=10). Inhibitors were used 30 min prior to CORM-3 administration. @ pb0.05 ,# pb0.01, * pb0.001 Vs Vehicle control. a pb0.001 CORM-3 (3 mg/kg/min, i.v.) vs Clopidogrel (30 mg/kg, p.o.).

557H. Soni et al. / Thrombosis Research 127 (2011) 551–559

derepressing fibrinolysis [15]. In contrast to the numerous reports onthe vasoactive or anti-inflammatory effects mediated by CO, there areonly a limited number of studies on the mechanisms by which COinhibits platelet aggregation. Interestingly, all of them use gaseous COand reveal a similarity with NO in the ability of both gases to inhibit

Platelets fromVehicle control rats

Platelet activatiFibrin formation

Stimulus/Agonist addition

Platelets fromCORM-3 treated rats

Less fibrin forma

Stimulus/Agonist addition

CORM-3

Release of CO-mediated

Increase sGC and cGMP

Inhibition of PAI-1 (fibri

Fig. 10. Schematic diagram that indicates the possible mechanisms involved in CORM-3-mex vivo platelet aggregation studies in rat platelets. Disruption of endothelium by chemical stirat. CORM-3 showed antiplatelet activity by (i) CO-mediated NO release, (ii) activation ofguanylate cyclase, PAI-1=plasminogen activator inhibitor-1.

platelet function via a common target (i.e. sGC) [10,28]. It is importantto note that in platelets the biological antiaggregatory activity of COhas been considered relatively low compared to the effect elicited byother endogenous agents released from the endothelium, such as NOor PGI2. Indeed, only high concentrations of gaseous CO (100%)

on Platelet aggregation

tion Antiplatelet effect

NO

nolyticenzyme)

- L-NAME

- ODQ

ediated antiplatelet effect. Agonists (such as thrombin) lead to platelet aggregation inmulus (such as FeCl3) leads to platelet aggregation in in vivomodels of thrombosis usingsGC and (iii) inhibition of fibrinolytic enzyme, PAI-1. NO=nitric oxide, sGC=soluble

558 H. Soni et al. / Thrombosis Research 127 (2011) 551–559

appear to inhibit aggregation of human platelets via the activation ofguanylate cyclase [28]. Furthermore, YC-1 [3-(5 V-hydroxymethyl-2Vfuryl)-1-benzyl indazole)], a compound that sensitizes guanylatecyclase to the action of NO, was also shown to amplify the action ofgaseous CO on guanylate cyclase activity in platelets [29]. Recently, anovel class of compounds, termed carbon monoxide-releasingmolecules (CORMs), has been discovered and their chemical andbiochemical features characterized [30]. These compounds have beendemonstrated to liberate CO in biological systems providing a usefulresearch tool for exploring the mechanism by which CO exerts itspharmacological activities. CORM-3 ((tricarbonylchloro(glycinato)ruthenium(II)) has a unique feature of being fully water-soluble andhas been shown to simulate the bioactivities of gaseous CO includingvessel relaxation [31], protection against I/R injury [6] , prevention oforgan rejection following transplantation [6] and inhibition of theinflammatory response [32]. Brune B et al. [10] tested the functionaleffect of CO on platelets as well as on cyclic nucleotides. In comparisonto gassing with nitrogen, CO resulted in an increase in activity of sGCto 402%. These observations are consistent with the view thatinhibition of platelet aggregation induced by CO is mediated via anelevation of intracellular cGMP content [15]. Chlopicki S et al. [17]reported that CORM-3 releasing CO in a time dependent andconcentration dependent manner displays anti-aggregatory effect inhuman platelets in vitro and this effect was independent of sGCactivation. A recent report showed that aortic transplantation in HO-1deficient mice results in 100% mortality within 4 days owing tosevere arterial thrombosis. Notably, treatment of HO-1-deficient micewith CORM-2 improved survival (62% survival at N56 days) [16].These reports are emphasizing the pleiotropic properties of CO andCO-releasing compounds in the resolution of vascular disorders.

Although there are few reports which suggests the antithromboticeffects of CO, the direct role of CORM-3 in in vivo thrombosis modelhas not been studied. Therefore, we studied the antiplatelet potentialof CORM-3 and also studied its antithrombotic activity in rat model ofthrombosis and explored its mechanism(s). We found that CORM-3produced antiplatelet effect in in vitro rat platelet aggregation studies.We studied the role of sGC (by using ODQ as sGC inhibitor), NO (byusing L-NAME as NOS inhibitor), KATP channel (by using glibencla-mide as KATP channel inhibitor) and p38MAPK (by using SCIO-469 asp38MAPK inhibitor) in ex vivo and in vivo models of thrombosis. Wefound that CORM-3 (3 mg/kg i.v.) produced significant antiplateletactivity in ex vivo assay and also observed that there was significantreduction in antiplatelet activity when animals were pretreated witheither ODQ or L-NAME, which suggests that antiplatelet activity byCORM-3 may be mediated partly via sGC and partly via NO-mediatedpathway. In response to low-concentration CO stimulation, vascularendothelial cells and blood platelets release NO [33]. Intracellularredistribution of NO may be one hypothesis for CO-mediated NOrelease. It is well known that arterial thrombosis is a critical event inthe pathogenesis of lesion development in many cardiovasculardiseases. Therefore, we further confirmed these observations inarterial thrombosis, venous thrombosis and arterio-venous (A-V)shunt model in rats. We observed that CORM-3 administrationshowed significant increase in time to occlusion (TTO) in arterialthrombosis model. Interestingly, CORM-3 administration also sup-presses the plasma PAI-1 levels, which suggests that along with sGCand NO, there might be the role of PAI-1 in CORM-3 mediatedantithrombotic activity. Similarly, there was significant reduction inthrombus weight in A-V shunt model and this activity abolished uponpretreatment with ODQ and L-NAME. Virchow described three maincomponent of thrombus (i) vessel injury, (ii) blood constituents and(iii) dynamics of flow. Historically, arterial thrombus (platelet richwhite clot) and venous thrombus (RBC rich red clot) are considered asseparate pathophysiological entities. This difference is furtherevidenced by the different use of antiplatelets (such as aspirin,clopidogrel, prasugrel etc.) and anticoagulants (such as heparin, low

molecular weight heparins, fondaparinux, factor Xa and IIa inhibitorsetc.) in the management of arterial and venous thrombosis respec-tively. We found that CORM-3 has antiplatelet activity and having noeffect on coagulation pathway as evident by our results onprothrombin time and it was expected that CORM-3 might have littleor no effect on venous thrombosis. As expected, we found that CORM-3 has non-significant effect on thrombusweight in venous thrombosismodel. However, there was a trend for reduction in venous thrombusweight but it was not statistically significant. These results indicatedthat platelets are also involved in venous thrombus but not aspredominant as RBCs. Since bleeding profile plays a major role insafety assessment of any antiplatelet or anticoagulant drug, we alsoinvestigated bleeding profile of CORM-3 using tail vein bleeding time(TVBT) model in rat. We found that CORM-3 produced significantbleeding at 3 mg/kg, i.v. administration but interestingly bleedingtendency was less as compared to marketed antiplatelet agent,clopidogrel.

In summary, CO has been shown to inhibit platelet aggregationand to relax vascular smooth muscle via cGMP-dependent pathway[10,34]. Our results also suggest that antiplatelet activity of CORM-3may be due to enhancement of sGC activity in ex vivo and in vivomodel of thrombosis. Brune B et al. [10] have reported that CO hasbeen considered to be less potent as compared to endogenoussubstances such as NO or PGI2 for inhibition of platelet aggregation.They have also proposed that higher concentration of CO has shownantiaggregatory effect and this effect was attributed via activation ofsGC [28]. We also believe that CO released from CORM-3 may be lesspotent than NO and therefore we considered role of NO and also PAI-1for in vivo antithrombotic effect of CORM-3. We found that activationof NO and inhibition of PAI-1 may also contribute to the in vivoantithrombotic effect of CORM-3 (Fig. 10). Numerous reportsdemonstrated CO-dependent protection in various animal models ofdisease and suggested that CO may be applicable as a molecularmedicine in corresponding human diseases. Progress in this area hasbeen slow due to regulatory and safety concerns associated withhuman trials with inhalation gases. An ongoing phase II trial hasaddressed the safety of inhaled CO during renal transplantation(Clinicaltrials.gov #NCT00531856). Pharmacological application of COwith the transition-metal carbonyl CO-releasing molecules mayprovide an additional therapeutic avenue. Whether direct applicationof CO by either pharmacologic administration or inhalation willprovide a safe and effective modality for the treatment of humandisease requires further research directed at understanding thepharmacokinetics and toxicology of CO application in humans. Weenvision that this class of molecules may be used as pharmaceuticalagents for the treatment of various cardiovascular and other diseases.

Conflicts of interest

The authors have no conflicts of interest.

Acknowledgements

Authors thank Dr. Ajay Sharma (Masson Eye Institute, Missouri)for his valuable guidance and support. Authors also thank manage-ment of Zydus Research Centre for all support.

References

[1] Chance B, Erecinska M, Wagner M. Mitochondrial responses to carbon monoxidetoxicity. Ann NY Acad Sci 1970;174:193–204.

[2] Otterbein LE, Soares MP, Yamashita K, Bach FH. Heme oxygenase-1: unleashingthe protective properties of heme. Trends Immunol 2003;24:449–55.

[3] Morita T, Perrella MA, LeeME, Kourembanas S. Smoothmuscle cell-derived carbonmonoxide is a regulator of vascular cGMP. Proc Natl Acad Sci USA 1995;92:1475–9.

[4] Wang R, Wang Z, Wu L. Carbon monoxide-induced vasorelaxation and theunderlying mechanisms. Br J Pharmacol 1997;121:927–34.

559H. Soni et al. / Thrombosis Research 127 (2011) 551–559

[5] Motterlini R, Clark JE, Foresti R, Sarathchandra P, Mann BE, Green CJ. Carbonmonoxide-releasing molecules: characterization of biochemical and vascularactivities. Circ Res 2002;90:E17–24.

[6] Clark JE, Naughton P, Shurey S, Green CJ, Johnson TR, Mann BE, et al.Cardioprotective actions by a water-soluble carbon monoxide-releasing molecule.Circ Res 2003;93:E2–8.

[7] Soni H, Patel P, Rath AC, Jain M, Mehta AA. Cardioprotective effect with carbonmonoxide releasing molecule-2 (CORM-2) in isolated perfused rat heart: Role ofcoronary endothelium and underlying mechanism. Vascul Pharmacol 2010;53:68–76.

[8] Marshall M, Hess H. Acute effects of low carbonmonoxide concentrations on bloodrheology, platelet fuction and the arterial wall in the minipig. Res Exp Med (Berl)1981;178:201–10.

[9] Mansouri A, Perry CA. Inhibition of platelet ADP and serotonin release by carbonmonoxide and in cigarette smokers. Experientia 1984;40:515–7.

[10] Brune B, Ullrich V. Inhibition of platelet aggregation by carbon monoxide ismediated by activation of guanylate cyclase. Mol Pharmacol 1987;32:497–504.

[11] Wagner CT, Durante W, Christodoulides N, Hellums JD, Schafer AI. Hemodynamicforces induce the expression of heme oxygenase in cultured vascular smoothmuscle cells. J Clin Invest 1997;100:589–96.

[12] Morisaki H, Katayama T, Kotake Y, Ito M, Tamatani T, Sakamoto S, et al. Roles ofcarbon monoxide in leukocyte and platelet dynamics in rat mesentery duringsevoflurane anesthesia. Anesthesiology 2001;95:192–9.

[13] Sato K, Balla J, Otterbein LE, Smith RN, Brouard S, Lin Y, et al. Carbon monoxidegenerated by heme oxygenase-1 suppresses the rejection of mouse-to-rat cardiactransplants. J Immunol 2001;166:4185–94.

[14] Peng L, Mundada L, Stomel JM, Liu JJ, Sun J, Yet SF, et al. Induction of hemeoxygenase-1 expression inhibits platelet-dependent thrombosis. Antioxid RedoxSignal 2004;6:687–90.

[15] Fujita T, Toda K, Karimova A, Yan S-F, Naka Y, Yet SF, et al. Paradoxical rescue fromischemic lung injury by inhaled carbon monoxide driven by derepression offibrinolysis. Nat Med 2001;7:598–604.

[16] Chen B, Guo L, Fan C, Bolisetty S, Joseph R, Wright MM, et al. Carbon monoxiderescues heme oxygenase-1-deficient mice from arterial thrombosis in allogeneicaortic transplantation. Am J Pathol 2009;175:422–9.

[17] Chlopicki S, Olszanecki R, Marcinkiewicz E, Lomnicka M, Motterlini R. Carbonmonoxide released by CORM-3 inhibits human platelets by a mechanismindependent of soluble guanylate cyclase. Cardiovasc Res 2006;71:393–401.

[18] Fratantoni JC, Poindexter BJ. Measuring platelet aggregation with microplatereader. A new technical approach to platelet aggregation studies. Am J Clin Pathol1990;94:613–7.

[19] Soni H, Sharma A, Bhatt S, Jain MR, Patel PR. Antithrombotic effects due topharmacological modulation of thrombin-activatable fibrinolysis inhibitor in rats.Pharmacology 2008;82:304–9.

[20] Wong PC, Crain Jr EJ, Nguan O, Watson CA, Racanelli A. Antithrombotic actions ofselective inhibitors of blood coagulation factor Xa in rat models of thrombosis.Thromb Res 1996;83:117–26.

[21] Gustafsson D, Nyström J, Carlsson S, Bredberg U, Eriksson U, Gyzander E, et al. Thedirect thrombin inhibitor melagatran and its oral prodrug H 376/95: intestinalabsorption properties, biochemical and pharmacodynamic effects. Thromb Res2001;101:171–81.

[22] Peternel L, Drevensek G, Cerne M, Stalc A, Stegnar M, Budihna MV. Evaluation oftwo experimental venous thrombosis models in the rat. Thromb Res 2005;115:527–34.

[23] Dejana E, Callioni A, Quintana A, deGaetanoG. Bleeding time in laboratory animals.II - A comparison of different assay conditions in rats. Thromb Res 1979;15:191–7.

[24] Andrews RK, Berndt MC. Platelet physiology and thrombosis. Thromb Res2004;114:447–53.

[25] Moncada S, Palmer RMJ, Higgs EA. Nitric oxide: physiology, pathophysiology, andpharmacology. Pharmacol Rev 1991;43:109–42.

[26] Lindenblatt N, Bordel R, Schareck W, Menger MD, Vollmar B. Vascular hemeoxygenase-1 induction suppresses microvascular thrombus formation in vivo.Arterioscler Thromb Vasc Biol 2004;24:601–6.

[27] Otterbein LE, Zuckerbraun BS, Haga M, Liu F, Song R, Usheva A, et al. Carbonmonoxide suppresses arteriosclerotic lesions associated with chronic graftrejection and with balloon injury. Nat Med 2003;9:183–90.

[28] Brune B, Schmidt KU, Ullrich V. Activation of soluble guanylate cyclase by carbonmonoxide and inhibition by superoxide anion. Eur J Biochem 1990;192:683–8.

[29] Friebe A, Mullershausen F, Smolenski A, Walter U, Schultz G, Koesling D. YC-1potentiates nitric oxide- and carbon monoxideinduced cyclic GMP effects inhuman platelets. Mol Pharmacol 1998;54:962–7.

[30] Motterlini R, Mann BE, Johnson TR, Clark JE, Foresti R, Green CJ. Bioactivity andpharmacological actions of carbon monoxide-releasing molecules. Curr PharmDes 2003;9:2525–39.

[31] Foresti R, Hoque M, Bains S, Green CJ, Motterlini R. Haem and nitric oxide:synergism in the modulation of the endothelial haem oxygenase-1 pathway.Biochem J 2003;372:381–90.

[32] Sawle P, Foresti R, Mann BE, Johnson TR, Green CJ, Motterlini R. Carbonmonoxide-releasing molecules (CO-RMs) attenuate the inflammatory response elicited bylipopolysaccharide in RAW264.7 murine macrophages. Br J Pharmacol 2005;145:800–10.

[33] Thom SR, Xu YA, Ischiropoulos H. Vascular endothelial cells generate peroxynitritein response to carbon monoxide exposure. Chem Res Toxicol 1997;10:1023–31.

[34] Cardell LO, Ueki IF, Stjärne P, Agusti C, Takeyama K, Lindén A, et al. Bronchodilatationin vivo by carbon monoxide, a cyclic GMP related messenger. Br J Pharmacol1998;124:1065–8.