Journal of Cardiovascular Pharmacology and Therapeutics http://cpt.sagepub.com/

Characterization of the Adenosine Pharmacology of Ticagrelor Reveals Therapeutically Relevant Inhibition of Equilibrative Nucleoside Transporter 1 Duncan Armstrong, Claire Summers, Lorna Ewart, Sven Nylander, James E. Sidaway and J. J. J. van Giezen J CARDIOVASC PHARMACOL THER 2014 19: 209 originally published online 10 January 2014 DOI: 10.1177/1074248413511693 The online version of this article can be found at: http://cpt.sagepub.com/content/19/2/209

Published by: http://www.sagepublications.com

Additional services and information for Journal of Cardiovascular Pharmacology and Therapeutics can be found at: Open Access: Immediate free access via SAGE Choice Email Alerts: http://cpt.sagepub.com/cgi/alerts Subscriptions: http://cpt.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.com/journalsPermissions.nav

>> Version of Record - Feb 17, 2014 OnlineFirst Version of Record - Jan 20, 2014 OnlineFirst Version of Record - Jan 19, 2014 OnlineFirst Version of Record - Jan 10, 2014 What is This?

Downloaded from cpt.sagepub.com at University of Waikato Library on July 10, 2014

Experimental Study

Characterization of the Adenosine Pharmacology of Ticagrelor Reveals Therapeutically Relevant Inhibition of Equilibrative Nucleoside Transporter 1

Journal of Cardiovascular Pharmacology and Therapeutics 2014, Vol. 19(2) 209-219 ª The Author(s) 2014 Reprints and permission: sagepub.com/journalsPermissions.nav DOI: 10.1177/1074248413511693 cpt.sagepub.com

Duncan Armstrong, PhD1, Claire Summers, BSc2, Lorna Ewart, PhD4, Sven Nylander, PhD3, James E. Sidaway, PhD2, and J. J. J. van Giezen, PhD3

Abstract Introduction: Studies have shown that ticagrelor has a further adenosine-mediated mechanism of action in addition to its potent inhibition of the P2Y12 receptor, which may explain some of ticagrelor’s clinical characteristics. This study aimed to further characterize the adenosine pharmacology of ticagrelor, its major metabolites, and other P2Y12 receptor antagonists. Methods: Inhibition of nucleoside transporter-mediated [3H]adenosine uptake by ticagrelor, its major metabolites, and alternative P2Y12 antagonists was examined in recombinant Madin-Darby canine kidney (MDCK) cells. The pharmacology of ticagrelor and its major metabolites at adenosine A1, A2A, A2B, and A3 receptor subtypes was examined using in vitro radioligand binding and functional assays and ex vivo C-fiber experiments in rat and guinea pig vagus nerves. Results: Ticagrelor (and less effectively its metabolites) and the main cangrelor metabolite inhibited [3H]adenosine uptake in equilibrative nucleoside transporter (ENT) 1-expressing MDCK cells, whereas cangrelor and the active metabolites of prasugrel or clopidogrel had no effect. No significant inhibitory activity was observed in MDCK cells expressing ENT2 or concentrative nucleoside transporters 2/3. Ticagrelor demonstrated high affinity (inhibition constant [Ki] ¼ 41 nmol/L) for ENT1. In adenosine receptor-binding experiments, ticagrelor and its major circulating metabolite, AR-C124910XX, had low affinity (Ki > 6 mmol/L) for each of the adenosine A1, A2A, and A2B receptors, whereas ticagrelor had a submicromolar (Ki ¼ 190 nmol/L) affinity for the adenosine A3 receptor. However, in functional assays, at high concentrations (10 mmol/L) ticagrelor only partially inhibited 3 mmol/L adenosine-induced depolarizations in the guinea pig and rat vagus nerve preparations (by 35% and 49%, respectively). Conclusions: Ticagrelor inhibits cellular adenosine uptake selectively via ENT1 inhibition at concentrations of clinical relevance. However, the low-binding affinity and functional inhibition of adenosine receptors observed with ticagrelor or its metabolites indicate that they possess a negligible adenosine-like activity at clinically relevant concentrations. Keywords ticagrelor, antiplatelet, adenosine, transporter, receptor

Introduction Ticagrelor is a direct-acting, reversibly binding oral P2Y12 antagonist1,2 approved for treatment of patients with acute coronary syndromes (ACSs). The antiplatelet and clinical activity of ticagrelor is predominantly mediated through its potent and reversible inhibition of the platelet P2Y12 receptor.2,3 In the phase III PLATelet inhibition and patient Outcomes (PLATO) trial (n ¼ 18 624 patients with ACS), ticagrelor reduced the rate of the composite end point of myocardial infarction/stroke/death from vascular causes without an increase in overall major bleeding compared with clopidogrel. Moreover, all-cause mortality was also reduced in patients receiving ticagrelor compared with clopidogrel,4 an effect not observed with other contemporary antiplatelet

1

Safety Pharmacology, Global Safety Assessment, AstraZeneca, Alderley Park, Macclesfield, Cheshire, United Kingdom 2 Discovery Safety, Drug Safety and Metabolism, AstraZeneca, Alderley Park, Macclesfield, Cheshire, United Kingdom 3 CVMD, AstraZeneca R&D, Mo¨lndal, Sweden 4 Translational Safety, Drug Safety and Metabolism, AstraZeneca, Alderley Park, Macclesfield, Cheshire, United Kingdom Guest Editor: Boris Simkhovich Manuscript submitted: July 23, 2013; accepted: October 10, 2013. Corresponding Author: Sven Nylander, CVMD, AstraZeneca R&D Mo¨lndal, 431 83 Mo¨lndal, Sweden. Email: [email protected]

Downloaded from cpt.sagepub.com at University of Waikato Library on July 10, 2014

210

Journal of Cardiovascular Pharmacology and Therapeutics 19(2)

agents. Furthermore, during the clinical development program of ticagrelor, dyspnoea and ventricular pauses were observed in some patients.3-5 These observations are suggestive of an additional mechanism of action for ticagrelor. Several other findings support such an additional mode of action via adenosine. For example, ticagrelor augmented both endogenous and exogenous adenosine-induced coronary blood flow in a canine model6 and augmented exogenous adenosineinduced coronary blood flow increases and dyspnoea in healthy volunteers.7 In vitro, ticagrelor has been shown to inhibit adenosine uptake into human erythrocytes as well as other human, rat, and canine cell lines6; ticagrelor also inhibited ex vivo human platelet aggregation in whole blood via increases in adenosine levels.8 Several clinical studies have examined the cardiovascular and respiratory effects of adenosine.9-12 Adenosine acting via A1, A2A, A2B, and A3 receptors may have beneficial effects in patients with ACS, including cardioprotection, vasodilatation, inflammatory regulation, and platelet function inhibition.11 Adenosine may also mediate bradycardia through direct effects on the sinoatrial and atrioventricular nodes10 and tachycardia via increased cardiac sympathetic drive and decreased vagal tone.9,10 In addition, adenosine has been shown to cause dyspnoea in healthy volunteers, potentially through direct stimulation of pulmonary C-fiber endings,12 a mechanism supported by in vivo data in rats13 and guinea pigs.14 Although ticagrelor is neither an adenosine analog, nor is it metabolized into adenosine,15 it is important to characterize the pharmacology of ticagrelor and its metabolites toward adenosine transporters and receptors and compare it with that of other known P2Y12 antagonists. Adenosine uptake is mediated by members of the sodium-dependent concentrative nucleoside transporter (CNT) 2/3 family16 and sodium-independent equilibrative nucleoside transporter (ENT) 1/2 family.17 Identification of the transporter that ticagrelor inhibits to reduce adenosine uptake may also explain ticagrelor’s adenosinemediated mechanism of action. Hence, the objective of the present in vitro and ex vivo studies was to further characterize ticagrelor’s pharmacology with respect to adenosine. The effect of ticagrelor on specific adenosine transporters was examined using recombinant cell systems. Second, receptor–ligand binding and functional assays were performed, as a direct effect of ticagrelor and its metabolites at adenosine receptors has not been ruled out. Finally, as dyspnoea could result from activation of pulmonary C-fibers, the effects of adenosine and ticagrelor were assessed in ex vivo guinea pig and rat C-fiber preparations.

Materials and Methods Materials Cell culture media were sourced from Invitrogen (Paisley, United Kingdom). Transfection reagents were purchased from Life Technologies. [3H]S-(4-Nitrobenzyl)-6-thioinosine ([3H]NBTI; specific activity 8.4 Ci/mmol, 11.9 mmol/L) and [2,8 3H]adenosine (specific activity 40 Ci/mmol, 25 mmol/L)

were purchased from ARC Inc (Saint Louis, Missouri). All other reagents and chemicals were purchased from Sigma (Dorset, UK) unless specified. Ticagrelor, AR-C124910XX (main circulating metabolite of ticagrelor and active vs P2Y12), ARC133913XX (main urinary metabolite of ticagrelor and inactive vs P2Y12),15 and elinogrel, clopidogrel active metabolite, prasugrel active metabolite and cangrelor main metabolite were synthesized in house (AstraZeneca, Mo¨lndal, Sweden).

Cell Lines for Adenosine Uptake Experiments Plasmids encoding human ENTs (hENTs) and CNTs and green fluorescent protein (protein IDs ENT1 [M1-V456], ENT2 [M1L456], CNT1 [M1-Q649], CNT2 [M1-A658], CNT3 [M1F691]) were synthesized and inserted into pcDNA3.1 by GeneArt gene synthesis service. Transient transfection of Madin-Darby canine kidney (MDCK) cells with the constructs was performed using Lipofectamine 2000. A control cell line was created by transient transfection of MDCK cells with an empty pcDNA3.1 expression construct (MOCK) using MaxCyte electroporation equipment. Transiently transfected cells were maintained in growth medium containing G418 (400 mg/mL). Cell membrane preparations for competitive binding experiments with NBTI required the dissociation of the transfected MDCK cells with accutase, resuspension in Dulbecco phosphate-buffered saline (DPBS) containing Ca2þ and Mg2þ and protease inhibitor cocktail, and disruption via sonication (Soniprep 150, MSE (UK) Ltd, London, UK). Subsequent centrifugation at 3000g for 3 minutes at 4 C and further centrifugation of the homogenate at 30 000g for 30 minutes at 4 C preceded resuspension of the crude membranes in DPBS (pH7.4) without protease inhibitor.

Examination of [3H]Adenosine Uptake by ENTs and CNTs The effect of the following test compounds (Figure 1) on [3H]adenosine uptake in MDCK cell lines expressing specific ENTs and CNTs was examined: ticagrelor, AR-C124910XX, AR-C133913XX, cangrelor, cangrelor main metabolite, elinogrel, clopidogrel active metabolite, prasugrel active metabolite, dipyridamole (adenosine transport inhibitor), and NBTI (adenosine transport inhibitor); unless stated all compounds were examined at 0.1, 1, 10, and 100 nmol/L, and 1, 10, 30, and 100 mmol/L. Transfected cells were seeded into 96-well scintiplates (PerkinElmer ScintiPlate-96 TC, PerkinElmer, Cambridge, UK; ENT1/2transfected cells: 2  104 cells/well; CNT2/3-transfected cells 3  104 cells/well). Equilibrative nucleoside transporter 1/2 cells after 24 hours or CNT2/3 cells after 72 hours were exposed to either test compound or control in the presence of 90 nmol/L [3H]adenosine substrate in sodium-free (ENT1/2)6 or sodiumcontaining buffer (CNT2/3). The reaction was stopped after 30 minutes with sodium-free buffer containing 100 mmol/L dipyridamole (ENT1/2) or 5 mmol/L cold adenosine (CNT2/3). [3H]Adenosine uptake was measured using proximity scintillography and normalized for total protein concentration. All assays

Downloaded from cpt.sagepub.com at University of Waikato Library on July 10, 2014

Armstrong et al

211

A.

F

B. F

HN N

N

N HO

N

O

N

F

N

NH2 N N

O

H

OH

CI

N

S

CI

O P OHO

H

N

O HO

O P

O

O

OH H

H

H O

F

F F

S

O H

H

H

OH

H N

O F

N

F

F. N

N

H.

O O

N H

O

O

I.

OH

OH

SH HO

N

N CI

CI S

S N

N H

H

H N

O

G.

O

F

S

S

N N

F N

O

OH

E.

S N

N

OH

OH

H

HO

H

HN

HO H

H

S

D.

P H

H

N

N

N HO

N

OH

S

H

C.

N

N

OH

HO

F

HN

N HO

O

OH

SH

N

N O

N

N N

N

N

F O

O OH

Figure 1. Structures of compounds used in the adenosine uptake studies: (A) ticagrelor, (B) AR-C124910XX, (C) AR-C133913XX, (D) cangrelor, (E) cangrelor main metabolite, (F) elinogrel, (G) clopidogrel active metabolite, (H) dipyridamole, and (I) prasugrel active metabolite.

were performed on 3 occasions except the control MOCK experiments for the CNT2/3 studies where N ¼ 1.

Examination of the Dissociation of [3H]NBTI by Ticagrelor and Its Metabolites and Other P2Y12 Receptor Antagonists Competitive binding studies were performed to examine whether ticagrelor was interacting directly with ENT1 to inhibit

[3H]adenosine uptake. Compounds that demonstrated inhibition of adenosine uptake by ENT1 were assessed for their ability to displace [3H]NBTI in MDCK cells expressing hENT1 and MDCK-MOCK. Membrane homogenates (final concentration 50 mg/mL) were incubated in a total 1.8 mL of DPBS (with Ca2þ and Mg2þ) with each compound and 2 nmol/L [3H]NBTI for 60 minutes at room temperature. Incubation was terminated by rapid filtration through Whatman 350 GF/B filter plates, and the remaining radioactivity was measured by liquid scintillation

Downloaded from cpt.sagepub.com at University of Waikato Library on July 10, 2014

212

Journal of Cardiovascular Pharmacology and Therapeutics 19(2) Calliper Life Sciences (Hanover, Maryland). To further investigate the functional consequence of ticagrelor interaction with the A2A receptor, a single concentration of either the reference agonist, adenosine-50 N-ethylcarboxamide (NECA; 10 nmol/L), or ticagrelor (1 mmol/L) was used to stimulate the A2A receptor and a range of 10 concentrations of the A2A selective reference antagonist, 8-(3-chlorostyryl) caffeine (CSC), were used to inhibit these responses in human embryonic kidney cells. All experiments were performed on 3 occasions.

a)

4 3 2

The Role of Ticagrelor at Adenosine Receptor Subtypes in an Ex-Vivo Vagus Preparation

0

As adenosine may mediate dyspnoea by its reported action on pulmonary C-fibers, the effects of adenosine and ticagrelor were assessed in ex-vivo guinea pig and rat C-fiber preparations. Male Dunkin Harley guinea pigs (Harlan, United Kingdom) and male Wistar rats (Harlan) were acclimatized for 7 days prior to sacrifice via cervical dislocation or pentobarbitone overdose, respectively. Vagus nerves were removed as described previously13 and placed in Krebs solution and bubbled with 95% O2/5% CO2. Nerves were cleared of connective tissue and desheathed; nerve trunks remained in oxygenated Krebs solution throughout. The methodology for the depolarization recording from the vagus was as reported previously.18,19 Drugs were applied into the perfusion solution of the first channel only and the response recorded via a chart recorder. Responses to 100 nmol/L to 30 mmol/L adenosine were determined in the guinea pig vagus (n  6), and responses to 1 to 10 mmol/L adenosine were determined in the rat vagus (n ¼ 2). Control responses to capsaicin (1 mmol/L, perfused for 2 minutes) were obtained at the beginning and end of each experiment. Selective antagonists for the adenosine A1 (DPCPX), A2A (SCH58261), A2B (MRS1706), and A3 (MRS1220) receptors were used to further characterize the guinea pig (n ¼ 6) and rat nerves (n ¼ 4) using the following sequence: (1) response to 1 mmol/L capsaicin perfusion for 2 minutes, (2) 2 reproducible responses to 3 mmol/L adenosine perfusion for 2 minutes, (3) 10-minute incubation with 3 mmol/L selective antagonist, (4) 2-minute stimulation with 3 mmol/L adenosine in presence of antagonist, (5) 2-minute stimulation with 3 mmol/L adenosine, (6) final response to 1 mmol/L capsaicin perfusion for 2 minutes, and (7) washout until responses returned to baseline. The above-mentioned protocol was used to test the effects of ticagrelor on 3 mmol/L adenosine-induced depolarization of guinea pig (ticagrelor doses 0.3, 1, 3, and 10 mmol/L; n ¼ 4) and rat (ticagrelor dose 10 mmol/L; n ¼ 4) vagus nerves; direct activation of vagus C-fibers by ticagrelor was carefully monitored during the 10-minute incubation period (n ¼ 4, both models). The effect of an adenosine deaminase inhibitor erythro-9-(2-hydroxy-3-nonyl) adenine hydrochloride (EHNA; n ¼ 2) and an adenosine transport inhibitor (dipyridamole; n ¼ 4) on the adenosine dose–response curve was also investigated in the guinea pig nerve preparations. Adenosine solutions, vehicle controls (water for EHNA; 0.1% dimethyl sulfoxide for dipyridamole), 10 mmol/L EHNA, or 1 or 10 mmol/L dipyridamole were added via the Krebs solution.

EN T1

C O M

EN T2

1

K

Uptake (pmol/mg/min)

5

b)

Uptake (pmol/mg/min)

5

Sodium-free buffer Sodium-containing buffer

4 3 2 1

T3 C

N

T2 N C

M

O

C

K

0

Figure 2. Rate of uptake of [3H]adenosine (90 nmol/L) in MDCK cells transiently transfected with (A) ENT1 or ENT2 (sodium-free buffer only) or (B) CNT2 or CNT3; in each series of experiments nontransporting (MOCK) constructs were assessed for control purposes. Data shown are mean of 3 experiments + SEM (except MOCK in (B) where n ¼ 1). CNT indicates concentrative nucleoside transporter; ENT, equilibrative nucleoside transporter; MDCK, Madin-Darby canine kidney; SEM, standard error of the mean.

(Ultima GOLD F, Perkin Elmer) counting. Nonspecific binding was defined in the presence of 2 mmol/L unlabeled NBTI. Specific binding was calculated by subtracting the nonspecific binding from the total binding. All assays were performed on 3 occasions.

The Role of Ticagrelor and Its Metabolites at Adenosine Receptor Subtypes in Recombinant Cells The activity of ticagrelor, AR-C124910XX, and ARC133913XX at adenosine receptor subtypes was investigated using standard radioligand binding and functional assays in Chinese hamster ovary cells by Ricerca BioSciences LLC (Taipei, Taiwan), Cerep (Celle L’evescault, France), and

Downloaded from cpt.sagepub.com at University of Waikato Library on July 10, 2014

Armstrong et al

213

Table 1. Effect of P2Y12 Receptor Antagonists and Dipyridamole on Adenosine Transport as Measured by Inhibition of [3H]Adenosine Uptake in MDCK Cell Lines Expressing Specific ENTs and CNTs.a [3H]Adenosine Uptake, pIC50 Compound Ticagrelor AR-C124910XX AR-C133913XX Cangrelor Cangrelor main metabolite Elinogrel Clopidogrel active metabolite Prasugrel active metabolite Dipyridamole

ENT1 6.59 + 0.09 (0.26) 5.67 + 0.18 (2.2) 5.45 + 0.41 (3.5) 100) 5.77 + 0.12 (1.7) 100) 100) 100) 8.13 + 0.17 (0.0074)

ENT2 4.76 4.70 4.10 100) + 0.11 (0.73)

CNT2 30) tested + 0.55 (2.4)

CNT3 30) 30) 30) 30) 30) 30) 30) not tested 4.56 + 0.38 (27)

Abbreviations: CNTs; concentrative nucleoside transporters; ENTs, equilibrative nucleoside transporters; MDCK, Madin-Darby canine kidney; SEM, standard error of the mean; pIC50, negative logarithm of the IC50; IC50, the concentration causing 50% inhibition of signal. a Data are mean of 3 experiments (pIC50 + SEM); mean IC50 (mmol/L) is shown in parentheses.

Table 2. Effect of P2Y12 Receptor Antagonists and Dipyridamole on the Inhibition of [3H]NBTI Binding to ENT1 in MDCK Cells Expressing ENT1.a Compound Dipyridamole (control) NBTI (control) Ticagrelor AR-C124910XX AR-C133913XX Cangrelor metabolite

pKi 8.59 + 0.6 (0.0026) 9.57 + 0.74 (0.0003) 7.38 + 0.2 (0.041) 6.48 + 0.38 (0.33) 5.65 + 0.32 (2.3) 100)

Abbreviations: ENT, equilibrative nucleoside transporter; MDCK, Madin-Darby canine kidney; [3H]NBTI, [3H]S-(4-Nitrobenzyl)-6-thioinosine; SEM, standard error of the mean; pKi, negative logarithm of the equilibrium dissociation constant; the concentration causing 50% inhibition of signal. a Data are the mean of 3 experiments (pKi + SEM); mean Ki (mmol/L) is shown in parentheses.

Data Analysis and Statistics Prism 5.0 software (Graphpad Software Inc, San Diego, California) was used to calculate the micromolar concentration of test compound producing 50% inhibition (IC50) + standard error of the mean (SEM) using nonlinear regression (log(inhibitor) vs response). Inhibition constant (Ki) values were derived from IC50 values based on the equation of Cheng and Prusoff.20 The effect of each antagonist, or ticagrelor, on adenosine-induced depolarization of C-fiber preparations was analyzed using a paired student t test comparing adenosine responses before and after incubation with antagonist or ticagrelor. Differences were deemed statistically significant when P < 0.05.

Results Effect of Ticagrelor and Its Metabolites and Other P2Y12 Receptor Antagonists on [3H]adenosine Uptake in Transfected MDCK Cells [3H]Adenosine uptake was similar (1-4 pmol/mg/min) in cells expressing ENT1, ENT2, CNT2, or CNT3 (Figure 2). In cells

expressing CNT-2/3, [3H]adenosine uptake was shown to be sodium dependent (Figure 2). [3H]Adenosine transport was negligible in cells expressing CNT1 (data not shown), and subsequent screening against this transporter was not performed. Minimal [3H]adenosine uptake was observed in control MOCK-transfected cells (Figure 2). Dipyridamole, a known inhibitor of the nucleoside transporters, inhibited [3H]adenosine uptake in each cell line (data not shown). In cells expressing ENT1, ticagrelor inhibited [3H]adenosine uptake with an IC50 of 260 nmol/L. AR-C124910XX, AR-C133913XX, and the metabolite of cangrelor also inhibited adenosine uptake (Table 1). No other P2Y12 antagonists inhibited [3H]adenosine uptake at an IC50 cangrelor main metabolite > AR-C124910XX > AR-C133913XX. However, ticagrelor was about 35-fold less potent than that of dipyridamole (IC50 7.4 nmol/L). In cells expressing ENT2, CNT2, or CNT3, [3H]adenosine uptake was not inhibited by ticagrelor or any P2Y12 antagonist, with all IC50 values less than 10 mmol/L (Table 1). To confirm that the observed inhibition of [3H]adenosine uptake was mediated through inhibition of ENT1, we tested whether each compound could cause dissociation of the ENT1 ligand [3H]NBTI from ENT1 (Table 2). Ticagrelor demonstrated a high affinity for ENT1 (Ki ¼ 41 nmol/L), which was approximately 15-fold less than that of dipyridamole, 2.6 nmol/L (Table 2). AR-C124910XX and AR-C133913XX also displaced [3H]NBTI from ENT1 (Table 2 and Figure 3) but with lower affinities (0.3 mmol/L and 2.3 mmol/L, respectively). The dissociation constant (Kd) of [3H]NBTI was 0.29 nmol/L.

The Role of Ticagrelor at Adenosine Receptor Subtypes in Recombinant Cells In radioligand-binding experiments, ticagrelor had low affinity (Ki > 6 mmol/L) for the A1, A2A, and A2B receptors (Table 3) and a submicromolar affinity for the adenosine A3 receptor

Downloaded from cpt.sagepub.com at University of Waikato Library on July 10, 2014

214

Journal of Cardiovascular Pharmacology and Therapeutics 19(2)

b)

150

% vehicle control

% vehicle control

a)

100

50

0

150

100

50

0 -10

-8

-6

-4

-10

log [ticagrelor] (M)

d)

150

% vehicle control

% vehicle control

c)

-8

-6

-4

log [AR-C124910XX] (M)

100

50

0

150

100

50

0 -10

-8 -6 log [AR-C133913XX] (M)

-4

-10

-8 -6 log [cangrelor main metabolite] (M)

-4

Figure 3. Inhibition of [3H]NBTI binding in MDCK-ENT1 cells by (A) ticagrelor, (B) AR C124910XX, (C) AR-C133913XX, and (D) cangrelor main metabolite. Data shown are mean of 4 experiments + SEM. [3H]NBTI indicates [3H]S-(4-nitrobenzyl)-6-thioinosine; MDCK, Madin-Darby canine kidney; ENT, equilibrative nucleoside transporter; SEM, standard error of the mean. Table 3. Effect of Ticagrelor and its Metabolites at Adenosine Receptor Subtypes in CHO Cell Lines.a Radioligand binding pKi Ticagrelor Adenosine A1 Adenosine A2A Adenosine A2B Adenosine A3 AR-C124910XX Adenosine A1 Adenosine A2A Adenosine A2B Adenosine A3 AR-C133913XX Adenosine A1 Adenosine A2A Adenosine A2B Adenosine A3

Antagonist pIC50

Agonist pEC50

5.20 4.86 100) + 0.16 (0.19)

4.66 + 0.13 (22) 100) 4.64 + 0.5 (23) 5.19 + 0.16 (6.4)

100)

5.19 5.19 100) + 0.07 (0.25)

4.87 + 0.4 (14) 100) 4.89 + 0.41 (13) 5.25 + 0.15 (5.6)

100)

5.72 3.86 100) + 0.13 (0.5)

4.78 + 0.33 (17) 100) 3.72 + 0.11 (>100) 4.33 + 0.11 (46)

100)

Abbreviations: CHO, Chinese hamster ovary; SEM, standard error of the mean; pKi, logarithm of inhibition constant; the concentration causing 50% inhibition of signal; EC50, the concentration causing 50% maximum effect; negative logarithm of the IC50; pEC50, negative logarithm of the EC50. a Data are the mean of 3 experiments (mean pKi, pIC50, and pEC50, each + SEM); mean Ki, IC50, and EC50 (mmol/L) are shown in parentheses.

(Ki ¼ 0.19 mmol/L). In functional assays, ticagrelor inhibited the action of the reference agonists at the A1, A2B, and A3 receptors with IC50 values of 22 mmol/L, 23 mmol/L, and

6 mmol/L, respectively; AR C124910XX inhibited these receptor subtypes with a similar potency to ticagrelor, whereas the potency of AR-C133913XX was generally lower (Table 3).

Downloaded from cpt.sagepub.com at University of Waikato Library on July 10, 2014

Armstrong et al

215 10 mmol/L EHNA (an inhibitor of adenosine deaminase) did not appear to change the magnitude of depolarization induced by 1, 3, or 10 mmol/L of adenosine in the guinea pig, relative to vehicle (n ¼ 2; data not shown). Similarly, addition of 1 mmol/L or 10 mmol/L dipyridamole (an adenosine transport inhibitor) did not appear to change the magnitude of adenosine-induced depolarizations (0.3-3 mmol/L) in the guinea pig vagus relative to vehicle (n ¼ 4; data not shown). A1, A2B, and A3 adenosine receptor antagonists (3 mmol/L DPCPX, 3 mmol/L MRS1706, and 3 mmol/L MRS1220, respectively) inhibited 3 mmol/L adenosine-induced depolarizations in guinea pig (n ¼ 6; Figure 5C) and rat (n ¼ 4; Figure 5D) vagus nerve preparations. Addition of ticagrelor (10 mmol/L) partially inhibited 3 mmol/L adenosine-induced depolarizations in both the guinea pig (n ¼ 4; Figure 6A) and the rat (n ¼ 4; Figure 6B) vagus nerve preparations by 35% and 49%, respectively. Small depolarizations were also observed in response to 10 mmol/L ticagrelor in 3 of 6 guinea pig vagus nerve preparations (0.02 + 0.01 mV) and 3 of 4 rat vagus nerve preparations (0.03 + 0.01 mV).

Discussion

Figure 4. Inhibition of (A) 1 mmol/L ticagrelor-induced and (B) 10 nmol/L NECA-induced stimulation of the adenosine A2A receptor by the selective antagonist CSC in HEK cells. Data shown are mean of 3 experiments + SEM. CSC indicates 8-(3-chlorostyryl) caffeine; NECA, 50 N-ethylcarboxamide; HEK, human embryonic kidney; SEM, standard error of the mean.

The concentration of ticagrelor causing 50% maximal effect (EC50) at the A2A subtype was 680 nmol/L (Table 3). To confirm that the agonist effect seen in the functional assay was due to interaction with the A2A receptor, a selective antagonist, CSC, was used to demonstrate that the signal caused by stimulation with NECA (reference agonist) or ticagrelor could be inhibited. Addition of CSC inhibited 10 nmol/L NECA- and 1 mmol/L ticagrelor-induced stimulation of the adenosine A2A receptor (Figure 4A and B).

The Role of Ticagrelor at Adenosine Receptor Subtypes in an Ex Vivo Vagus Preparation Perfusion of adenosine onto guinea pig (n  6; Figure 5A) and rat (n ¼ 2; Figure 5B) vagus nerve preparations resulted in reproducible depolarizations in a dose-dependent manner, confirming the activity of adenosine in this model; a maximum response was observed with a 10 mmol/L dose. Addition of

The present in vitro and ex vivo studies were performed to further characterize the pharmacology of ticagrelor with respect to its adenosine-mediated effects. Our findings confirm that ticagrelor inhibits adenosine uptake via the ENT1 transporter and demonstrate that ticagrelor is unlikely to act directly at adenosine receptors to a clinically relevant extent. These data are in line with, and provide a mechanism for, the adenosinemediated activity of ticagrelor as shown in previous studies.6-8 Ticagrelor inhibited adenosine uptake in MDCK cells expressing ENT1, with a potency (IC50) approximately 35fold less than that of dipyridamole (a known inhibitor of ENT1).21 These findings are generally consistent with those from a previous study in human breast carcinoma cells, where the potency of ticagrelor with respect to inhibition of adenosine uptake was approximately 10-fold less than that of dipyridamole.6 Further, we have previously demonstrated that ticagrelor can augment adenosine-induced platelet inhibition by prolonging the half-life of adenosine added to whole blood. Via this mechanism, ticagrelor in effect increases the concentration of adenosine, which can subsequently interact via the platelet adenosine A2A receptor; the potency (EC50) of ticagrelor for this effect was approximately 5 mmol/L.8 Here, we have also demonstrated that ticagrelor can displace [3H]NBTI from ENT1 in MDCK cells confirming that ticagrelor binds to ENT1 directly to inhibit adenosine uptake. The affinity of ticagrelor for ENT1 was 41 nmol/L, and the dissociation constant (Kd) of [3H]NBTI in this system was similar (0.29 nmol/L) to that reported previously.17 Ticagrelor demonstrated low affinity (ie, Ki > 6 mmol/L) for the adenosine A1, A2A, and A2B receptor subtypes and a submicromolar affinity for the A3 subtype. Generally, we also showed low affinity interactions of the major circulating and urine metabolites of ticagrelor (AR-C124910XX and AR-C133913XX, respectively) with the A1, A2A, and A2B receptor subtypes.

Downloaded from cpt.sagepub.com at University of Waikato Library on July 10, 2014

216

Journal of Cardiovascular Pharmacology and Therapeutics 19(2)

b)

0.10 0.08

Depolarisation (mV)

Depolarisation (mV)

a)

0.06 0.04 0.02 0.00 -8 -6 log [adenosine] (M)

0.06 0.04 0.02

-4

1.0

d)

100 75 50 *

*

25

3.0 10.0 [Adenosine] (mM)

100 75

* % inhibition

% inhibition

0.08

0.00 -10

c)

0.10

*

50

*

*

25 0

0

-25

-25 Vehicle

A1 A2a A2b Adenosine receptor

A3

Vehicle

A1 A2a A2b Adenosine receptor

A3

Figure 5. Effect of adenosine on (A) guinea pig and (B) rat vagus nerve preparations. Data shown are mean + SEM; n  6 and n ¼ 2 at each adenosine concentration in guinea pig and rat, respectively. The effect of the selective adenosine A1, A2A, A2B, and A3 receptor antagonists on adenosine-mediated depolarizations is shown in (C) guinea pig and (D) rat preparations. Data shown are mean + SEM; n ¼ 4 for each experiment; *P < .05 versus vehicle. SEM indicates standard error of the mean.

These findings were confirmed for ticagrelor in a different model; ex vivo guinea pig and rat C-fiber preparations. As expected, and in line with findings from earlier studies,13,14 adenosine depolarized vagal C-fibers (guinea pig and rat) via the A1, A2B, and A3 receptors. Ticagrelor inhibited adenosine-mediated depolarization of vagal C-fibers at high concentrations only. Small depolarizations were also observed in response to 10 mmol/L ticagrelor alone. It is unlikely that A1 and A3 receptor subtypes were responsible for this effect, as the functional studies suggested that ticagrelor acts as a weak antagonist at these receptors. The results indicate that the A2A receptor was not involved in vagus nerve depolarization, and ticagrelor may possibly be acting as a partial agonist at the A2B receptor. To examine this theory, antagonism of the response to ticagrelor in the guinea pig vagus was tested with the A2B inhibitor, MRS1706. However, due to the lack of a reproducible response to ticagrelor, these experiments were not successful. A potential explanation for the variable effects of ticagrelor observed in the receptor experiments reported herein is that any effect for compounds with a low affinity for the receptor may be overshadowed by the inherent variability in signal in such experiments. Therefore, we conclude that it is unlikely that ticagrelor or its metabolite will

induce a clinical response via specific adenosine receptor interactions. A limitation in the interpretation of these findings is that low species homology between rat and human receptors22 and prior knowledge that the xanthine class of adenosine receptor ligands displays different affinities between species (see Jacobson for review)23 suggest that extrapolation between rat or guineapig models and humans must be made with caution. Indeed, it is not possible to directly extrapolate findings from the in vitro preparation studies to human pharmacology. For example, the extent of protein binding to ticagrelor is not compensated for in these in vitro experiments. In patients with ACS who received ticagrelor (90 mg twice daily) for 4 weeks, the steady state mean maximum plasma concentration was 1.5 mmol/L.24 However, ticagrelor and its major circulating metabolite are extensively plasma protein bound in vivo (>99.7%),25 meaning that the unbound concentration is in the low nanomolar range. In comparable in vitro experiments where plasma protein binding is also negligible, the Ki for ticagrelor versus 2MeS-ADP on the P2Y12 receptor is approximately 4 nmol/L.2 Based on the affinity of ticagrelor for the ENT1 transporter (Ki ¼ 41 nmol/L) and the strong plasma protein binding, we

Downloaded from cpt.sagepub.com at University of Waikato Library on July 10, 2014

Armstrong et al

217

a) 100

% inhibition

75 50

*

25 0 -25 Vehicle

0.3

1.0

3.0

10.0

Ticagrelor (µM)

b) 100 75 % inhibition

* 50 25 0 -25 Vehicle

Ticagrelor (10µM)

Figure 6. Effect of ticagrelor on adenosine-induced (3 mmol/L) depolarizations in (A) guinea pig and (B) rat vagus preparations. Data shown are mean + SEM; n ¼ 4 for each experiment; *P < .05 versus vehicle. SEM indicates standard error of the mean.

would therefore anticipate that at clinical doses, ticagrelor is likely to only partially inhibit ENT1. However, such mild inhibition is expected to be sufficient to increase local concentrations of adenosine in the circulation; we have previously demonstrated that the addition of 1 mmol/L ticagrelor to whole blood in vitro is sufficient to significantly conserve extracellular concentrations of adenosine, which were 3.4-fold higher in the presence of ticagrelor relative to vehicle only.8 Given that adenosine is a directacting agonist at its receptors, any local elevation would be expected to increase receptor signaling and the resulting response. Indeed, ticagrelor (180 mg) augmented adenosine-induced coronary blood flow increases in healthy volunteers, an effect that correlated with plasma exposure and was more evident at plasma concentrations greater than 1 mmol/L.7 Based on the results of the receptor binding studies, we conclude that a direct interaction of ticagrelor with adenosine

receptors is unlikely to be of clinical relevance. The affinity of ticagrelor for the adenosine A3 receptor, the subtype with the highest affinity interaction observed with ticagrelor (Ki ¼ 190 nmol/L), is about 5-fold lower than the affinity of ticagrelor for the ENT1 transporter (Ki ¼ 41 nmol/L). A lack of clinically relevant effect for ticagrelor at the adenosine receptors is supported by data from Wittfeldt et al7 who showed that the effect of ticagrelor on coronary blood flow velocity correlated with the plasma concentration of ticagrelor only in the presence of adenosine. Indeed, if the effect of ticagrelor was mediated by a direct effect at the adenosine receptors, an effect on coronary blood flow with ticagrelor alone would have been expected and this was not observed. The adenosine-mediated mechanism of action of ticagrelor via inhibition of ENT1 could potentially contribute to the overall clinical benefit of ticagrelor, including the reduction in all cause and cardiovascular mortality observed in ticagrelortreated patients.4 For example, adenosine has been shown to have several effects that may be beneficial for patients with ACS, including cardioprotection, vasodilatation, inflammatory regulation, and platelet function inhibition.8,11 Indeed, local adenosine increases via inhibition of ENT1 could also contribute to the dyspnoea or bradyarrhythmias that have been observed in some patients receiving ticagrelor.3,4 Of the other P2Y12 antagonists examined, only the cangrelor main metabolite inhibited adenosine uptake at ENT1, with an IC50 of 1.7 mmol/L. No data are available on the plasma concentration of the cangrelor main metabolite following infusion of cangrelor; however it is interesting to note that dyspnoea is also a side effect associated with the use of this agent.26-28 No effect was observed for the active metabolite of clopidogrel, consistent with the observation that ‘‘adenosine-mediated’’ effects were not observed in the phase III trials of clopidogrel,29 although dyspnoea, judged by the investigator to be likely or possibly due to the study drug, occurred in 3.7% of patients receiving clopidogrel and 0% of patients receiving placebo in the ONSET/OFFSET trial.5 Despite reports of dyspnoea in patients receiving elinogrel at an incidence greater than those receiving clopidogrel,30 elinogrel demonstrated no significant inhibition of adenosine uptake in transporter-expressing cells in the present study. We were unable to test the demethylated metabolite of elinogrel that has been reported to circulate at a concentration 10% of parent.31 However, a direct effect for elinogrel on adenosine receptors has not been investigated and can therefore not be excluded. Alternative mechanisms by which all P2Y12 antagonists (including ticagrelor) could induce dyspnoea include an increased release of adenosine triphosphate from red blood cells32 or nonplatelet P2Y12 receptor-mediated effects (eg, smooth muscle or neuronal P2Y12).33,34

Conclusion Ticagrelor inhibits adenosine uptake via inhibition of ENT1. This additional mechanism of action for ticagrelor could potentially be of clinical relevance and help explain the clinical

Downloaded from cpt.sagepub.com at University of Waikato Library on July 10, 2014

218

Journal of Cardiovascular Pharmacology and Therapeutics 19(2)

profile of ticagrelor in patients with ACS. Generally, ticagrelor showed low affinity for the adenosine receptor subtypes suggesting that a direct effect of ticagrelor on adenosine receptors at clinically relevant levels is unlikely. Acknowledgments The authors would like to thank employees of Ricerca BioSciences LLC, Calliper Life Sciences, and Cerep for performing the receptor binding and functional assays. We thank Drs Maria Belvisi and Mark Birrell of Imperial College, London, for performing isolated vagus experiments. We also thank James Goodman for the generation of the CNT2/3 data and Dr Marie South for statistical advice. Medical writing support was provided by David Evans (Gardiner-Caldwell Communications, Macclesfield, United Kingdom) and this was funded by AstraZeneca. The present studies were funded by AstraZeneca, of which all authors are employees. The data presented in this manuscript have not been published or presented previously.

Declaration of Conflicting Interests The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: All authors are employees of AstraZeneca.

Funding The author(s) received the following financial support for the research, authorship, and/or publication of this article: The studies reported in this paper were sponsored by AstraZeneca.

References 1. Husted S, Emanuelsson H, Heptinstall S, Sandset PM, Wickens M, Peters G. Pharmacodynamics, pharmacokinetics, and safety of the oral reversible P2Y12 antagonist AZD6140 with aspirin in patients with atherosclerosis: a double-blind comparison to clopidogrel with aspirin. Eur Heart J. 2006;27(9):1038-1047. 2. van Giezen JJ, Nilsson L, Berntsson P, et al. Ticagrelor binds to human P2Y(12) independently from ADP but antagonizes ADPinduced receptor signaling and PA. J Thromb Haemost. 2009; 7(9):1556-1565. 3. Cannon CP, Husted S, Harrington RA, et al. Safety, tolerability, and initial efficacy of AZD6140, the first reversible oral adenosine diphosphate receptor antagonist, compared with clopidogrel, in patients with non-ST-segment elevation acute coronary syndrome: primary results of the DISPERSE-2 Trial. J Am Coll Cardiol. 2007;50(19):1844-1851. 4. Wallentin L, Becker RC, Budaj A, et al. Ticagrelor versus clopidogrel in patients with acute coronary syndromes. N Engl J Med. 2009;361(11):1108-1111. 5. Storey RF, Bliden KP, Patil SB, et al. Incidence of dyspnea and assessment of cardiac and pulmonary function in patients with stable coronary artery disease receiving ticagrelor, clopidogrel, or placebo in the ONSET/OFFSET study. J Am Coll Cardiol. 2010;56(3):185-193. 6. van Giezen JJ, Sidaway J, Glaves P, Kirk I, Bjo¨rkman JA. Ticagrelor inhibits adenosine uptake in vitro and enhances adenosinemediated hyperemia responses in a canine model. J Cardiovasc Pharmacol Ther. 2012;17(2):164-172.

7. Wittfeldt A, Emanuelsson H, Brandup-Wognsen G, et al. Ticagrelor enhances adenosine-induced coronary vasodilatory responses in humans. J Am Coll Cardiol. 2013;61(7):723-727. 8. Nylander S, Femia E, Scavone M, et al. Ticagrelor inhibits human platelet aggregation via adenosine in addition to P2Y12 antagonism. J Thromb Haemost. 2013;11(10):1867-1876. 9. Biaggioni I, Olafsson B, Robertson RM, Hollister AS, Robertson D. Cardiovascular and respiratory effects of adenosine in conscious man. evidence for chemoreceptor activation. Circ Res. 1987; 61(6):779-786. 10. Rongen GA, Brook SC, Pollard MJ, et al. Effect of adenosine on heart rate variability in humans. Clin Sci (Lond). 1999;96(6):597-604. 11. Fredholm BB. Adenosine, an endogenous distress signal, modulates tissue damage and repair. Cell Death Differ. 2007;14(7):1315-1323. 12. Burki NK, Dale WJ, Lee LY. Intravenous adenosine and dyspnea in humans. J Appl Physiol. 2005;98(1):180-185. 13. Hong JL, Ho CY, Kwong K, Lee LY. Activation of pulmonary C fibers by adenosine in anaesthetized rats: role of adenosine A1 receptors. J Physiol. 1998;508(pt 1):109-118. 14. Chuaychoo B, Lee MG, Kollarik M, Pullmann R Jr, Undem BJ. Evidence for both adenosine A1 and A2A receptors activating single vagal sensory C-fibres in guinea pig lungs. J Physiol. 2006; 575(pt 2):481-490. 15. Teng R, Oliver S, Hayes MA, Butler K. Absorption, distribution, metabolism, and excretion of ticagrelor in healthy subjects. Drug Metab Dispos. 2010;38(9):1514-1521. 16. Plagemann PG. Naþ-dependent, concentrative nucleoside transport in rat macrophages. Specificity for natural nucleosides and nucleoside analogs, including dideoxynucleosides, and comparison of nucleoside transport in rat, mouse and human macrophages. Biochem Pharmacol. 1991;42(2):247-252. 17. Ward JL, Sherali A, Mo ZP, Tse CM. Kinetic and pharmacological properties of cloned human equilibrative nucleoside transporters, ENT1 and ENT2, stably expressed in nucleoside transporter-deficient PK15 cells. Ent2 exhibits a low affinity for guanosine and cytidine but a high affinity for inosine. J Biol Chem. 2000;275(12):8375-8381. 18. Rang HP, Ritchie JM. Depolarization of nonmyelinated fibers of the rat vagus nerve produced by activation of protein kinase C. J Neurosci. 1988;8(7):2606-2617. 19. Birrell MA, Crispino N, Hele DJ, et al. Effect of dopamine receptor agonists on sensory nerve activity: possible therapeutic targets for the treatment of asthma and COPD. Br J Pharmacol. 2002; 136(4):620-628. 20. Cheng Y, Prusoff WH. Relationship between the inhibition constant (K1) and the concentration of inhibitor which causes 50 per cent inhibition (I50) of an enzymatic reaction. Biochem Pharmacol. 1973;22(23):3099-3108. 21. Molina-Arcas M, Casado FJ, Pastor-Anglada M. Nucleoside transporter proteins. Curr Vasc Pharmacol. 2009;7(4):426-434. 22. Salvatore CA, Jacobson MA, Taylor HE, Linden J, Johnson RG. Molecular cloning and characterization of the human A3 adenosine receptor. Proc Natl Acad Sci U S A. 1993;90(21): 10365-10369. 23. Jacobson KA. Adenosine A3 receptors: novel ligands and paradoxical effects. Trends Pharmacol Sci. 1998;19(5):184-191.

Downloaded from cpt.sagepub.com at University of Waikato Library on July 10, 2014

Armstrong et al

219

24. Storey RF, Husted S, Harrington RA, et al. Inhibition of platelet aggregation by AZD6140, a reversible oral P2Y12 receptor antagonist, compared with clopidogrel in patients with acute coronary syndromes. J Am Coll Cardiol. 2007;50(19):1852-1856. 25. European Medicines Agency. Assessment report for Brilique; 2011. www.ema.europa.eu/docs/en_GB/document_library/EPAR_-_ Public_assessment_report/human/001241/WC500100492.pdf. Accessed 11th November 2013. 26. Bhatt DL, Lincoff AM, Gibson CM, et al. Intravenous platelet blockade with cangrelor during PCI. N Engl J Med. 2009; 361(24):2330-2341. 27. Harrington RA, Stone GW, McNulty S, et al. Platelet inhibition with cangrelor in patients undergoing PCI. N Engl J Med. 2009; 361(24):2318-2329. 28. Bhatt DL, Stone GW, Mahaffey KW, et al. Effect of platelet inhibition with cangrelor during PCI on ischemic events. N Engl J Med. 2013;368(14):1303-1313. 29. Yusuf S, Zhao F, Mehta SR, et al. Effects of clopidogrel in addition to aspirin in patients with acute coronary syndromes without ST-segment elevation. N Engl J Med. 2001;345(23):494-502.

30. Welsh RC, Rao SV, Zeymer U, et al. A randomized, double-blind, active-controlled phase 2 trial to evaluate a novel selective and reversible intravenous and oral P2Y12 inhibitor elinogrel versus clopidogrel in patients undergoing nonurgent percutaneous coronary intervention: the INNOVATE-PCI trial. Circ Cardiovasc Interv. 2012;5(3):336-346. 31. Siller-Matula JM, Krumphuber J, Jilma B. Pharmacokinetic, pharmacodynamic and clinical profile of novel antiplatelet drugs targeting vascular diseases. Br J Pharmacol. 2010; 159(3):502-517. 32. Ohman J, Kudira R, Albinsson S, Olde B, Erlinge D. Ticagrelor induces adenosine triphosphate release from human red blood cells. Biochem Biophys Res Commun. 2012;418(4): 754-758. 33. Wihlborg AK, Wang L, Braun OO, et al. ADP receptor P2Y12 is expressed in vascular smooth muscle cells and stimulates contraction in human blood vessels. Arterioscler Thromb Vasc Biol. 2004;24(10):1810-1815. 34. Cattaneo M, Faioni EM. Why does ticagrelor induce dyspnea? Thromb Haemost. 2012;108(6):1031-1036.

Downloaded from cpt.sagepub.com at University of Waikato Library on July 10, 2014