supporting information - pnas · 2013-09-03 · light transmission aggregometry was performed as...

Supporting InformationKirkby et al. 10.1073/pnas.1218880110SI Materials and MethodsPreparation of Washed Platelets. Blood was collected by veni-puncture from healthy volunteers who had provided writtenconsent (St. Thomas’s Hospital Research Ethics Committee,reference 07/Q0702/24) into trisodium citrate (0.32% final).Platelet-rich plasma was obtained from whole blood by centri-fugation (230 × g, 15 min), and platelets were isolated by furthercentrifugation (1,100 × g, 10 min) in the presence of prosta-glandin I2 (PGI2; 1 μg/mL). The resulting pellets were washedtwice in Tyrode’s buffer, at pH 7.4, containing Hepes (20 mM;Sigma), resuspended to 3 × 108 platelets/mL, and allowed to restfor at least 30 min for the effects of PGI2 to dissipate.

Light Transmission Aggregometry.Washedplatelet suspensionswereincubated (30 min, room temperature) with the P2Y12 receptorblocker prasugrel-active metabolite (PAM; 3 μM; a kind gift ofAstraZeneca) or vehicle (0.5%DMSO). Aggregation to thrombin(1 U/mL; Sigma), collagen-related peptide [CRP-XL 1 μg/mL (akind gift of Richard Farndale, University of Cambridge, Cam-bridge, United Kingdom), in the presence of 1 mg/mL human fibrin-ogen (Sigma)], Horm collagen (3 μg/mL; Nycomed; in the pre-sence of 1 mg/mL human fibrinogen) or epinephrine (10 μM;Chronolog) was measured in a Bio/Data PAP-8E turbidimetricaggregometer (37 °C, 1,200 rpm stir speed, 4min) after preincubation(1 min, 37 °C) with the NO donor diethlyammonium (Z)-1-(N,N-diethylamino)diazen-1-ium-1,2-diolate (DEA/NONOate, Sigma) orvehicle. Although readily soluble in water, DEA/NONOate spon-taneously decomposes to NO at physiological pH and was there-fore dissolved in 10 mM NaOH. The addition of NaOH (1 mMfinal) did not alter the pHof theplatelet suspensionbuffer.Thehalf-life ofDEA/NONOate (2min) is such that∼75%of totalNO releasewould occur over the course of a 4-min aggregation experiment.As required, and in parallel with PAM, platelets were additionally

pretreated with milrinone (1 μM; Tocris Bioscience), erythro-9-2-hydroxy-3-nonyladenine (10 μM; Tocris Bioscience), sildenafil (10μM; a kind gift of Adrian Hobbs, The William Harvey ResearchInstitute, London), wortmannin (200 nM; Tocris Bioscience),aspirin (30 μM; Sigma), or N5- [imino(nitroamino)methyl]-L-ornithine, methyl ester, monohydrochloride (L-NAME) (10 μM;Sigma). In some experiments, L-arginine (100 μM; Sigma) wasadded 5 min before thrombin (1 U/mL) to stimulate platelet NOSactivity. In other experiments, 8-(4-chlorophenylthio)guanosine-3′,5′cyclic monophosphorothioate, Rp-isomer (Rp-8-pCPT-cGMPS)(200 μM) or KT5823 (5 μM) was added 10 min before thrombin(1 U/mL) to inhibit cGK activity.

96-Well Plate Aggregometry. In some experiments, platelet aggre-gation was assessed in 96-well plates, as we have previouslyreported (1–4). Briefly, washed platelets were treated as describedearlier and then plated out as 50-μL aliquots into individual wellsof a 96-well microtiter plate (Nunc). DEA/NONOate (0.1 nM–100μM) or vehicle (0.01 M NaOH) was then added to the wells, andthe plate was gently mixed for 1 min. The platelets were thenactivated by the addition of agonists, thrombin (0.5 or 1 U/mL) orCRP (0.5 or 1 μg/mL in the presence of 1 mg/mL human fibrin-ogen), and vigorous shaking (1,200 rpm at 37 °C for 5 min) ona microplate shaker (BioShake IQ, QUANTIFOIL Instruments).Absorbance was then measured at 595 nm (Sunrise microplatereader; Tecan), and percentage platelet aggregation was calcu-lated with reference to the absorbance of Tyrodes-Hepes buffertaken as a surrogate for 100% aggregation.

Isobolographic Analysis. Using data from 96-well platelet aggrega-tion experiments, curves of inhibition of thrombin/CRP-inducedplatelet aggregation byDEA/NONOate in the presence of increasedconcentrations of PAM or ticagrelor were produced. These datawere fitted to a logistic equation using least-squares method(GraphPad Prism 5), and for each curve, the concentration re-quired to inhibit aggregation by 25%, 50%, and 75% was cal-culated. In some cases, to demonstrate 50% or 75% inhibition,this required interpolation of the curve beyond experimental data.From these data, isobolograms were plotted for each level ofinhibition. To do this, for each inhibitor curve, the concentrationof PAM/ticagrelor was plotted on the X-axis, and the concen-tration of DEA/NONOate on the Y-axis. Thus, a scatter plotis produced showing the combinations of DEA/NONOate andPAM/ticagrelor that produce a given level of inhibition. Forcomparison, a theoretical additive relationship is plotted thatrepresents a straight line connecting the inhibitory concentrationsof either PAM or ticagrelor and DEA/NONOate applied alone.The construction of these isobolograms is detailed in Fig. S5.

Whole-Blood Aggregometry.Whole blood, collected into lepirudin(25 μg/mL final; Celgene), was incubated with PAM (3 μM) orvehicle (0.5% DMSO) for 30 min at room temperature. Plateletaggregation to the thrombin receptor activating peptides SFLLRN-amide (10 μM; Bachem) and AYPGKF-amide (30 μM; Bach-em) was monitored by impedance aggregometry (Multiplate;Dynabyte) in blood diluted 1:1 with saline, and preincubated(1 min, 37 °C) with DEA/NONOate or vehicle.

ADP+ATP Secretion.Washed platelets, preincubated with PAM orits vehicle, as described earlier, were incubated with DEA/NON-Oate or vehicle and then activated with thrombin (1 U/mL) andshaking (1,200 rpm, 37 °C; BioShake IQ) in a white 96-well mi-croplate. After 2 min, phosphocreatine (1.5 mM; Sigma), creatinephosphokinase (3,000 U/mL; Sigma), and Chrono-Lume reagent(1:5 vol/vol; Chronolog) were added, and plates were incubatedfor a further 2 min before the luminescence of each well was readin a Berthold Mithras LB940 multimode plate reader.

P-Selectin Expression and Glycoprotein IIb/IIIa Activation. Washedplatelets, preincubated with PAM or its vehicle, as describedearlier, were incubated with DEA/NONOate or vehicle and thenactivated with thrombin (1 U/mL) under nonstirring conditions at37 °C. After 2 min, the reaction was stopped by dilution witha 10-fold excess of 25 μg/mL lepirudin to halt thrombin activity.Platelets were immediately stained with anti–CD61-allophyco-cyanin (eBioscience), PAC-1-FITC (BD Bioscience), and anti–P-selectin-PE (eBioscience) for 15 min at 4 °C and then fixed in 2%(vol/vol) formalin (Sigma). PAC-1-FITC and anti–P-selectin-PEimmunoreactivity was measured by flow cytometry using a FACS-Calibur instrument (Becton Dickenson) by determining the meanfluorescence intensity in the appropriate channel of 15,000 CD61-APC+ events. The specificity of PAC-1 binding was determinedusing RGDS blocking peptide (Sigma) and by blocking plateletactivation with prostacyclin (1 μM). The specificity of anti–P-se-lectin staining was confirmed by replacement of the antibody withan irrelevant isotype-matched control and by blocking plateletactivation with prostacyclin (1 μM). Example control data areshown in Fig. S12.

Akt Phosphorylation. Light transmission aggregometry was per-formed as described earlier, and the reaction was stopped at 2 minby the addition of an equal volume of 25 mM Tris buffer con-

Kirkby et al. www.pnas.org/cgi/content/short/1218880110 1 of 12

www.pnas.org/cgi/content/short/1218880110

taining 1% (vol/vol) Nonidet P-40, 5% (vol/vol) glycerol, 150 mMNaCl, 5 mM MgCl2, and 2× protease inhibitor mixture (RocheApplied Bioscience). Samples were sonicated (15 s) and denatured(95 °C, 5 min), and protein concentration was determined bythe bicinchoninic acid method. Protein samples (20 μg/lane)were separated by SDS/PAGE, blotted to nitrocellulose mem-branes, and probed with antiphospho(Ser473)-Akt (New EnglandBiolabs) or anti–pan-Akt primary antibodies (New EnglandBiolabs). Immunoreactivity was visualized using horseradishperoxidase-conjugated goat anti-rabbit secondary antibodies (Dako)and enhanced chemiluminescent detection. Equal protein loadingwas confirmed by reprobing with anti–β-actin antibodies (Dako).Blots were quantified by densitometry with ImageJ software(National Institutes of Health) and expressed as ratio of intensityof phospho(Ser473)-Akt to total Akt after normalization of eachto β-actin band intensity.

Rap1b Activation.Light transmission aggregometry was performedas described earlier, with a washed platelet concentration of 1 ×109/mL, and the reaction was stopped 15 s after the addition ofthrombin (1 U/mL) with a 5-fold excess of 25 mM Tris buffercontaining 1% Nonidet P-40, 5% glycerol, 150 mM NaCl, 5 mMMgCl2, and 1× protease inhibitor mixture. Samples were snapfrozen, and protein concentration was determined by the bicin-choninic acid method. Active Rap1-GTP was purified by RalGDSaffinity pull-down, using a commercially available kit (ThermoScientific). As a control, a sample of raw lysate was treated with anexcess of GTPγS before pull-down. Both the eluted Rap1-GTPand raw lysate were separated by SDS/PAGE, blotted to nitro-cellulose membranes, and probed with an anti-Rap1 antibody(Thermo Scientific). Immunoreactivity was visualized, usinghorseradish peroxidase-conjugated goat anti-rabbit secondaryantibodies (Dako) and enhanced chemiluminescent detection.Bands corresponding to Rap1-GTP were quantified by densi-tometry with ImageJ software (National Institutes of Health) andexpressed as a fraction of total Rap1 content of raw platelet lysate.

cAMP and cGMP Measurements. Light transmission aggregometrywas performed as described earlier, and at 5 min, platelets werelysed with Triton-X-100 (0.625% final) and treated with iso-butylmethylxanthine (500 μM final) and potassium fluoride (0.5

M final). cAMP and cGMP concentrations were determined byhomogenous time-resolved fluorescence-based competitive im-munoassays (Cisbio).

Vasodilator Stimulated Phosphoprotein Phosphorylation. PAM- orvehicle-treated washed platelets were stimulated with thrombin (1U/mL) in the presence of DEA/NONOate or vehicle. After 4 min,the reaction was stopped with methanol-free formaldehyde (2%final; Fisher Scientific). Platelets were permeabilized (0.2% TritonX-100; Sigma) and incubated with anti–vasodilator stimulatedphosphoprotein (VASP)-P(Ser239) primary antibody (Enzo Life-sciences), Alexa647-conjugated secondary antibody (Invitrogen),and FITC-conjugated anti-CD42b (eBiosciences), for 30 min each,in turn, before resuspension in 0.9% saline. VASP-P(Ser239) im-munoreactivity was measured by flow cytometry, using a FACS-Calibur instrument (Becton Dickenson) by determining the meanVASP-P-Alexa647 fluorescence of 15,000 CD42b-FITC+ events.The specificity of staining was confirmed using a specific blockingpeptide (Santa Cruz Biotechnology) and by omission of the pri-mary antibody. Example control data are shown in Fig. S12.

Calcium Mobilization. Platelets were loaded with Fura-2 by in-cubation with Fura-2 acetoxymethylester (Fura-2 AM) (5 μM,30 min, 37 °C; Sigma) in platelet rich plasma (PRP) beforewashing and treatment with PAM or vehicle, as described earlier.Thrombin (1 U/mL)-induced intracellular calcium mobilization,in the presence of DEA/NONOate or vehicle, was measured overthe course of 3 min as changes in the ratio of 510 nm emissionafter excitation at 340 nm and 380 nm in a Perkin-Elmer LS940Bfluorometer, as previously described (5). Total Ca2+ flux over thistime was measured as the area under the trace.

Statistics and Data Analysis. Unless otherwise stated, data wereanalyzed by two-way ANOVA, using GraphPad Prism 5 for MacOS X, and differences were considered significant if P < 0.05.Summary data (IC50, EC50) were obtained by fitting of data toa logistic equation and tested by Student t test (2 groups) or one-way ANOVA (>2 groups). The reader should note that in somecases where only partial responses were obtained at supra-physiological concentrations of DEA/NONOate, inhibitory po-tencies were interpolated.

1. Armstrong PC, et al. (2009) Utility of 96-well plate aggregometry and measurement ofthrombi adhesion to determine aspirin and clopidogrel effectiveness. Thromb Haemost102(4):772–778.

2. Armstrong PC, et al. (2008) Aspirin and the in vitro linear relationship between thromboxaneA2-mediated platelet aggregation and platelet production of thromboxane A2. J ThrombHaemost 6(11):1933–1943.

3. Chan MV, Armstrong PC, Papalia F, Kirkby NS, Warner TD (2011) Optical multichannel(optimul) platelet aggregometry in 96-well plates as an additional method of plateletreactivity testing. Platelets 22(7):485–494.

4. Chan MV, Warner TD (2012) Standardised optical multichannel (optimul) plateletaggregometry using high-speed shaking and fixed time point readings. Platelets 23(5):404–408.

5. Jones S, et al. (2010) The plasma membrane calcium ATPase modulates calciumhomeostasis, intracellular signaling events and function in platelets. J Thromb Haemost8(12):2766–2774.



-7.5 -7.0 -6.5 -6.0 -5.5

0

20

40

60

80

100

[CRP] log g/ml

Agg

rega

tion

(%)

Veh

-1.5 -1.0 -0.5 0.0

0

20

40

60

80

100 Veh

[thrombin] log U/ml

Agg

rega

tion

(%)

-2.5 -2.0 -1.5 -1.0 -0.5 0.0

0

20

40

60

80 VehPAM

[thrombin] log U/mlA

ggre

gatio

n (%

)

-2.5 -2.0 -1.5 -1.0 -0.5 0.0

0.0

0.4

0.8

1.2 VehPAM

[thrombin] log U/ml

AD

P+A

DP

rele

ase

(nm

oles

/108 p

late

lets)

0

200

400

600

800 VehPAM

-7.0 -6.5 -6.0 -5.5 -5.0 -4.5

-7.0 -6.5 -6.0 -5.5 -5.0-7.5[TRAP-6] log M[PAR4-AP] log M

Agg

rega

tion

(AU

C)

(a) (b)

(c)

(d)

Light transmission aggregometry ADP+ATP secretion

Whole blood impedence aggregometry

96-well plate aggregometry

* *

**

-12 -10 -8 -6 -4

-20

0

20

40

60

80

100

[DEA/NONOate] log M

Inhi

bitio

n of

agg

rega

tion

to 1

μg/m

l CR

P (%

)

VehiclePAM *

(e) Light transmission aggregometry

Fig. S1. Effect of P2Y12 blockade alone on sensitivity to platelet agonists in different assays. Concentration-response curves to thrombin, protease-activatedreceptor (PAR)-activating peptides, and CRP in the presence and absence of PAM (3 μM) for washed platelet light transmission aggregometry (A), washedplatelet luminometric ADP+ATP secretion assays (B), whole-blood impedance aggregometry (C), or washed platelet 96-well plate aggregometry (D). Thesecurves demonstrate the requirement for high, but not supramaximal, concentrations of agonists in experiments with DEA/NONOate to provide a baselineaggregation that is only minimally altered by P2Y12 blockade. (E) CRP (1 μg/mL)-induced aggregation of washed platelets, determined by changes in lighttransmission, that was inhibited to a modest degree by the NO donor, DEA/NONOate. This inhibitory effect was greatly potentiated in platelets pretreatedwith the P2Y12 receptor blocker, PAM (3 μM). *P < 0.05 by two-way ANOVA for the effect of PAM (n = 3–4).



collagen 3μg/ml

1 min

Light transmission

Vehicle

collagen 3μg/ml

collagen 3μg/ml

collagen 3μg/ml

collagen 3μg/ml

collagen 3μg/ml

collagen 3μg/ml

collagen 3μg/ml

-1.5 -1.0 -0.5 0.0

0

20

40

60

80

100

[thrombin] log U/ml

Agg

rega

tion

(%)

-9 -8 -7 -6 -5 -4

0

20

40

60

80

100Vehicle

[collagen] log g/ml

Agg

rega

tion

(%) L-Arginine (100μM)

L-NAME (300μM)L-Arg + L-NAME

PAM (3μM)PAM + L-ArgPAM + L-NAMEPAM + L-Arg + L-NAME

Agg

rega

tion

(%)

to 3

μg/m

l col

lagen

0

20

40

60

80

PAM (3μM)L-arginine (100μM)

L-NAME (300μM)

--

-

-+

-

--

+

-+

+

+-

-

++

-

+-

+

++

+

Agg

rega

tion

(%)

to 1

U/m

l thr

ombi

n

0

20

40

60

80

PAM (3μM)L-arginine (100μM)

L-NAME (300μM)

--

-

-+

-

--

+

-+

+

+-

-

++

-

+-

+

++

+

(b) 96 well plate aggregometry

L-arginine (100μM) L-NAME (300μM) L-Arg + L-NAME

Vehi

cle

PAM

(a) Light transmission aggregometry

* * * *

Fig. S2. Endogenous NO does not contribute to platelet aggregation responses in this system in the presence or absence of P2Y12 blockade. Thrombin (1 U/mL)- orcollagen (3 μg/mL)-induced aggregation of washed platelets, measured by light transmission aggregometry (A), was not influenced by addition of L-arginine (100μM, 5 min incubation) or L-NAME (300 μM, 30min incubation) to boost or inhibit platelet-derived NO production. Collagen, but not thrombin-induced aggregation,was markedly inhibited by PAM (3 μM). P2Y12 receptor blockade did not influence the responses to L-arginine or L-NAME treatment. Parallel experiments using 96-well plate aggregometry (B) and a range of thrombin and collagen concentrations similarly found no effect of L-arginine or L-NAME on aggregation responses inthe presence or absence of P2Y12 receptor blockade. *P < 0.05 by one-way ANOVA vs. vehicle (n = 4).



(a) (b)GPIIb/IIIa activation (PAC-1 binding) P-selectin expression

-12 -10 -8 -6 -4

-20

0

20

40

60

80

100

VehiclePAM (3uM)

[DEA/NONOate] log M

P-se

lect

in ex

pres

sion

(% in

hibi

tion)

Freq

uenc

y

0

20

40

60

80

100

P-selectin Fluorescence

100 101 102 104103

Freq

uenc

y

0

20

40

60

80

100

Freq

uenc

y

0

20

40

60

80

100

PAC-1 Fluorescence

100 101 102 104103

Freq

uenc

y

0

20

40

60

80

100 Vehicle Vehicle

+PAM +PAM

Thrombin (1U/ml)

Thrombin + DEA/NONOate (1µM)

Unstimulated

Thrombin (1U/ml)

Thrombin + DEA/NONOate (1µM)

Unstimulated

-12 -10 -8 -6 -4

-20

0

20

40

60

80

100

VehiclePAM (3uM)

[DEA/NONOate] log M

PAC

-1 b

indi

ng(%

inhi

bitio

n)

* *

Fig. S3. Blockade of the P2Y12 receptor sensitizes platelets to the inhibitory effect of NO against thrombin-induced GPIIb/IIIa activation and alpha granulesecretion. Using flow cytometry thrombin (1 U/mL), stimulated GPIIb/IIIa activation was measured by PAC-1 binding (A), and alpha granule secretion wasmeasured by P-selectin expression (B). DEA/NONOate produced concentration-dependent inhibitions of both of these responses, and these inhibitions weregreatly increased by pretreatment of platelets with the P2Y12 receptor blocker PAM (3 μM). *P < 0.05 by two-way ANOVA for the effect of PAM (n = 4).



-12 -10 -8 -6 -4 -2

0

20

40

60

80

100

Veh

1.2∝M4∝M

0.4∝M

PAM

PAM

PAM

[DEA/NONOate] log M

Inhi

bitio

n of

1U

/ml t

hrom

bin-

indu

ced

aggr

egat

ion

in w

ashe

d pl

atel

ets (

%)

Concentration of DEA/NONOate required for 25% inhibition in the presence of a given concentration of PAM

Concentration of PAM required for 25% inhibition in the absence of DEA/NONOate

Concentration of DEA/NONOate required for

25% inhibition in the absence of PAM

Theoretical curve predicted for additive effect of DEA/NONOate and PAM

Theoreticpredictedeffect of ffNONOat

0 2 4 6 8

0

2000

4000

6000

[PAM] uM

[DE

A/N

ON

Oat

e ]

μM

Data curving towards the axes indicates a synergistic inhibition of platelet aggregation by DEA/NONOate and PAM

Concentration Response CurvesCurves show inhibitory effect of DEA/NONOate in the absence, and presence of increasing concentrations of PAM

Isobologram (25%)Combined concentrations of DEA/NONOate and PAM required to produce 25% inhibition of platelet aggregation

Fig. S4. Construction of isobolograms from matched concentration-response curves. To perform isobolographic analysis, multiple matched 96-well plateaggregation experiments were performed, and inhibition curves were generated showing the concentrations of DEA/NONOate required to inhibit aggre-gation responses in the presence of specific concentrations of P2Y12 receptor blocker. From curves fitted to these data, the concentration of DEA/NONOaterequired to produce 25%, 50%, and 75% inhibition in the presence of each concentration of P2Y12 receptor blocker was calculated. The values were used toconstruct isobolograms, in which the concentration of P2Y12 receptor blocker was plotted on the X-axis and the concentration of DEA/NONOate required toproduce a given effect (e.g., 25% inhibition) plotted on the Y-axis. This gives a curve showing the combination concentrations of DEA/NONOate and P2Y12

receptor blocker required to produce a given response (e.g., a 25% isobole). This curve is compared with the theoretical linear relationship connecting theconcentrations of DEA/NONOate and P2Y12 receptor blocker required individually to produce that same effect; that is, the theoretical line corresponding to anadditive relationship. If the experimental data are distorted toward the axes away from the theoretical additive line, a synergistic (supraadditive) relationship isdemonstrated.



0 2 4 6

0

50000

100000

150000

200000

Experimental data

[Ticagrelor] μM

[DE

A/N

ON

Oat

e ] μ

M

Theoretical Additive Line

0 5 10 15

0

50000

100000

150000

Experimental data


[PAM] μ M[D

EA

/NO

NO

ate ]

μM

0 1 2 3 4 5

0

20

40

60

80

100

[PAM] μM

[DE

A/N

ON

Oat

e ] μ

M

Experimental data


0 1 2 3

0.000

0.005

0.010

0.015

[PAM] μM

[DE

A/N

ON

Oat

e ]

μM Experimental data


0 5 10 15

0.0

0.2

0.4

0.6

0.8

Experimental data

[PAM] μM

[DE

A/N

ON

Oat

e ]

μM


0.0 0.1 0.2 0.3 0.4 0.5

0

50

100

150

Experimental data

[Ticagrelor] μM

[DE

A/N

ON

Oat

e ] μ

M


0.00 0.05 0.10

0.000

0.005

0.010

0.015

0.020

Experimental data

[Ticagrelor] μM

[DE

A/N

ON

Oat

e ]

μM


0.000 0.005 0.010 0.015

0.0000

0.0002

0.0004

0.0006

0.0008

0.0010

Experimental data

[Ticagrelor] μM

[DE

A/N

ON

Oat

e ] μ

M


(a) (b)

(c) (d)

PAM; Thrombin 1U/ml; IC50 isobologram

PAM; CRP 1µg/ml; IC50 isobologram

(e) (f)

(g) (h)

Ticagrelor; Thrombin 1U/ml; IC50 isobologram

Ticagrelor; CRP 1µg/ml; IC50 isobologram

PAM; Thrombin 0.5U/ml; IC50 isobologram

PAM; CRP 0.5µg/ml; IC50 isobologram

Ticagrelor; Thrombin 0.5U/ml; IC50 isobologram

Ticagrelor; CRP 0.5µg/ml; IC50 isobologram

Fig. S5. IC50 isobolograms demonstrate that the interaction between P2Y12 receptor blockade and NO is synergistic for inhibition of thrombin- and CRP-induced aggregation. Isobolographic analyses of the combinations of the P2Y12 receptor blockers PAM (A–D) or ticagrelor (E–H) and DEA/NONOate required toproduce a 50% inhibition of platelet aggregation stimulated by (A and E) thrombin 1 U/mL, (B and F) thrombin 0.5 U/mL, (C and G) CRP 1 μg/mL, or (D and H)CRP 0.5 μg/mL demonstrate, in all cases, a highly synergistic relationship. This high level of synergism is demonstrated by the markedly greater effect of thedrug combinations (isoboles curved toward the axis) than the linear relationship predicted by the arithmetic sum of the effect of either drug alone (n = 5–6 foreach point).



0 2 4 6 8

0

2000

4000

6000 Theoretical additive lineExperiemental data

[PAM] uM[D

EA

/NO

NO

ate

] µM

(a) (b)

(c) (d)



(e) (f)

(g) (h)







0 5 10 15 20

0

500000

1000000

1500000

Experimental data

Theoretical additive line

[PAM] uM

[DE

A/N

ON

Oat

e ] µ

M

0.0 0.5 1.0 1.5

0.0

0.1

0.2

0.3

0.4

0.5

Experimental data


[PAM] uM

[DE

A/N

ON

Oat

e ] µ

M

0 2 4 6 8 10

0

5000

10000

15000

Experimental data


[PAM] uM

[DE

A/N

ON

Oat

e ] µ

M

0.0 0.5 1.0 1.5 2.0

0.000

0.005

0.010

0.015

0.020 Theoretical additive line

Experimental data

[PAM] uM

[DE

A/N

ON

Oat

e ] µ

M

0 20 40 60

0

50

100

150

200 Theoretical additive line

Experimental data

[PAM] uM

[DE

A/N

ON

Oat

e ] µ

M

0.0 0.2 0.4 0.6 0.8

0.000

0.001

0.002

0.003


Experimental data

[PAM] uM

[DE

A/N

ON

Oat

e ] µ

M

0 2 4 6 8

0.00

0.02

0.04


Experimental data

[PAM] uM

[DE

A/N

ON

Oat

e ] µ

M

Fig. S6. IC25 and IC75 isobolograms demonstrate that the interaction between P2Y12 receptor blockade and NO is synergistic for inhibition of thrombin- andCRP-induced aggregation. Isobolographic analyses of the combinations of PAM and DEA/NONOate required to produce 25% (A, C, E, and G) and 75% (B, D, F,and H) inhibition of platelet aggregation stimulated by thrombin 1 U/mL (A and B), thrombin 0.5 U/mL (C and D), CRP 1 μg/mL (E and F), or CRP 0.5 μg/mL (Gand H), display a highly synergistic relationship. This high level of synergism is demonstrated by the markedly greater effect of the drug combinations (isobolescurved toward the axis) than the linear relationship that is predicted by the arithmetic sum of the effect of either drug alone (n = 5–6 for each point).



-12 -10 -8 -6 -4

-20

0

20

40

60

80

100

[DEA/NONOate] log M

Inhi

bitio

n of

thro

mbi

n-in

duce

dag

greg

atio

n in

was

hed

plat

elet

s (%

)

VehPAM (3µM)

Epinephrine (10µM)Epinephrine + PAM

-12 -10 -8 -6 -4

-20

0

20

40

60

80

100

[DEA/NONOate] log M

Inhi

bitio

n of

thro

mbi

n-in

duce

dag

greg

atio

n in

was

hed

plat

elet

s (%

)

VehPAM (3µM)Aspirin (100µM)Aspirin + PAM

(a)

(b)

*

*

*

#

-12 -10 -8 -6 -4

-20

0

20

40

60

80

[DEA/NONOate] log (M)

Inhi

bitio

n of

agg

rega

tion

(%)

to 1

U/m

l thr

ombi

n (%

)

Vehicle

PAM (3uM)

KT5823 (5µM)

PAM + KT5823

-12 -10 -8 -6 -4

-20

0

20

40

60

80


Inhi

bitio

n of

agg

rega

tion

(%)

to 1

U/m

l thr

ombi

n (%

)

Vehicle

Rp-8-pCPT-cGMPS (200µM)

PAM (3uM)

PAM + Rp-8-pCPT-cGMPS

(c)

(d)

**

**

Fig. S7. The synergistic inhibition produced by NO and P2Y12 receptor blockade is independent of thromboxane formation but dependent on Gαi/z signalingand roles of cGMP-dependent protein kinases. Thrombin (1 U/mL)-induced aggregation of washed platelets was studied by light transmission aggregometry.As previously observed, PAM (3 μM) potentiated the inhibitory effect of DEA/NONOate in this system. Blockade of thromboxane A2 formation by the cyclo-oxygenase inhibitor aspirin (100 μM) had no effect on the sensitivity of platelets to DEA/NONOate either in the absence or presence of PAM (A). In contrast,addition of epinephrine (10 μM), which restores Gαi/z signaling in the presence of P2Y12 receptor blockade, completely reversed the ability of PAM to po-tentiate the antiplatelet effects of DEA/NONOate in this assay (B). Two putative cGMP-dependent protein kinase (cGK) inhibitors, Rp-8-pCPT-cGMPS (200 μM)(C) and KT5823 (5 μM) (D) failed to reverse the inhibition of platelet aggregation produced by the combination of DEA/NONOate and PAM. This may reflecteither the involvement of cGK-independent platelet inhibition pathway or the limitations of these experimental tools. *P < 0.05 by two-way ANOVA vs.control; #P < 0.05 by two-way ANOVA vs. the same treatment in the absence of epinephrine (n = 3–4).



-12 -10 -8 -6 -4 -2

-20

0

20

40

60

80

100


Inhi

bitio

n of

thro

mbi

n-in

duce

dag

greg

atio

n in

was

hed

plat

elet

s (%

)

VehPAMMilMil+PAM

-12 -10 -8 -6 -4 -2

-20

0

20

40

60

80

100


Inhi

bitio

n of

thro

mbi

n-in

duce

dag

greg

atio

n in

was

hed

plat

elet

s (%

)

VehPAMSilSil+PAM

-12 -10 -8 -6 -4 -2

0

2

4

6

8

10 VehPAMEHNAEHNA+PAM


cAM

P (n

M)

-12 -10 -8 -6 -4 -2

0

10

20

30 VehiclePAMEHNAEHNA+PAM


cGM

P (n

M)

-12 -10 -8 -6 -4 -2

-20

0

20

40

60

80

100

VehPAMEHNAEHNA+PAM

[DEA/NONOate] log (M)In

hibi

tion

of th

rom

bin-

indu

ced

aggr

egat

ion

in w

ashe

d pl

atel

ets (

%)

-12 -10 -8 -6 -4 -2

0

2

4

6

8

10 VehPAMMilMil+PAM


cAM

P (n

M)

-12 -10 -8 -6 -4 -2

0

2

4

6

8

10 VehiclePAMSilSil+PAM

[DEA/NONOate] log (M)cA

MP

(nM

)

-12 -10 -8 -6 -4 -2

0

50

100

150

200 VehiclePAM

Sil

Sil+PAM


cGM

P (n

M)

-12 -10 -8 -6 -4 -2

0

10

20

30

40 VehPAM

MilMil+PAM

[DEA/NONOate] log (M)cG

MP

(nM

)

PDE

2; E

HN

APD

E3;

Milr

inon

ePD

E5;

Sild

enaf

il

*

*

*

*

##

*

*#

#

Aggregation cGMP cAMP

Fig. S8. Inhibition of PDE2, PDE3, or PDE5 does not affect the ability of P2Y12 receptor blockade to potentiate the inhibitory effect of NO. Aggregation wasmeasured in thrombin (1 U/mL)-stimulated washed platelets as changes in light transmission, after which, platelets were lysed and cAMP and cGMP contentmeasured by immunoassay. Pretreatment of platelets with the PDE2 inhibitor erythro-9-2-hydroxy-3-nonyladenine (10 μM) or the PDE3 inhibitor, milrinone(Mil; 10 μM), had no effect on the ability of DEA/NONOate to inhibit platelet aggregation or stimulate platelet cGMP and cAMP formation. P2Y12 receptorblockade by PAM (3 μM) potentiated the antiplatelet effect of DEA/NONOate regardless of concurrent PDE2/3 inhibition. In contrast to erythro-9-2-hydroxy-3-nonyladenine and milrinone, the PDE5 inhibitor sildenafil (Sil; 10 μM) enhanced the ability of DEA/NONOate to inhibit aggregation; this was accompanied byenhanced cGMP, but not cAMP, accumulation. Addition of PAM further increased the sensitivity of platelets to DEA/NONOate, but the magnitude of thisincrease in potency was similar to that seen in the absence of sildenafil, suggesting an additive interaction. *P < 0.05 by two-way ANOVA vs. the sametreatment in absence of PAM. #P < 0.05 by two-way ANOVA vs. the same treatment in absence of PDE inhibitor (n = 4–8).



-12 -10 -8 -6 -4

-20

0

20

40

60

80

100


Inhi

bitio

n of

thro

mbi

n-in

duce

d ag

greg

atio

n in

was

hed

plat

elet

s (%

)

VehiclePAM

-10 -9 -8 -7 -6

-20

0

20

40

60

80

100

[prostacyclin] log (M)

Inhi

bitio

n of

thro

mbi

n-in

duce

d ag

greg

atio

n in

was

hed

plat

elet

s (%

) VehiclePAM

-12 -10 -8 -6 -4

-20

0

20

40

60

80

100

[DEA/NONOate] log M

Inhi

bitio

n of

thro

mbi

n-in

duce

d A

TP+

AD

P re

leas

e (%

)

VehiclePAM

-11 -10 -9 -8 -7 -6 -5

-20

0

20

40

60

80

100

[prostacyclin] log (M)In

hibi

tion

of th

rom

bin-

indu

ced

AT

P+A

DP

rele

ase

(%)

VehiclePAM

DEA/NONOate Prostacyclin

Was

hed

Plat

elet

Agg

rega

tion

AD

P+AT

P Se

cret

ion

DEA/NONOate IC50; Veh: 3mM (extrapolated)DEA/NONOate IC50; PAM: 10nMPotency shift: 300,000-fold

DEA/NONOate IC50; Veh: 40nMDEA/NONOate IC50; PAM: 2nMPotency shift: 20-fold

DEA/NONOate IC50; Veh: 0.3mM (extrapolated)DEA/NONOate IC50; PAM: 20nMPotency shift: 15,000-fold

DEA/NONOate IC50; Veh: 200nMDEA/NONOate IC50; PAM: 4nMPotency shift: 50-fold

*

*

*

*

Fig. S9. The interaction between P2Y12 receptor blockade and NO is much greater than that between P2Y12 receptor blockade and prostacyclin. Thrombin(1 U/mL)-induced aggregation of washed platelets was studied by light transmission aggregometry, and ADP+ATP secretion by luminometry. As previouslyobserved, DEA/NONOate produced only a weak inhibition of these responses, whereas prostacyclin was a more potent and effective inhibitor. Blockade ofP2Y12 receptors by PAM (3 μM) increased the sensitivity of platelets to both DEA/NONOate and prostacyclin in both assays. The magnitude of this potentiationwas, however, far greater for DEA/NONOate (15,000- to 30,000-fold shift) than that observed with prostacyclin (20- to 50-fold shift). *P < 0.05 by two-wayANOVA for the effect of PAM (n = 4–8).



Freq

uenc

y

0

20

40

60

80

100

VASP-P Fluorescence

100 10 102 103

Vehicle

DEA/NONOate (100µM)

DEA/NONOate (100µM), primary antibody omitted

DEA/NONOate (100µM), blocking peptide

Freq

uenc

y

0

20

40

60

80

100

Freq

uenc

y

0

20

40

60

80

100

P-selectin Fluorescence

100 101 102 104103

PAC-1 Fluorescence

100 101 102 104103

Vehicle

Thrombin (1U/ml)

Thrombin (1U/ml), isotype control

Thrombin (1U/ml), prostacyclin (1µM)

Vehicle

Thrombin (1U/ml)

Thrombin (1U/ml), blocking peptide (RGDS)

Thrombin (1U/ml), prostacyclin (1µM)

(a)

(b)

(c)

1

Fig. S10. Representative control data for flow cytometry experiments. VASP phosphorylation (Ser239) was measured in washed platelets by flow cytometry(A). An increase in VASP-P(Ser239)-associated fluorescence was observed after treatment of platelets with DEA/NONOate (100 μM), but this increase influorescence intensity could be abolished by either omission of the primary antibody or preincubation of the primary antibody with a specific blocking peptide.Alpha granule secretion was measured in washed platelet by flow cytometry for P-selectin expression (B). An increase in P-selectin-associated fluorescence wasobserved after treatment of platelets with thrombin (1 U/mL), but this could be prevented by treatment of platelets with prostacyclin (1 μM) or restored toalmost baseline levels by substituting the primary antibody with an irrelevant isotype-matched control antibody. GPIIb/IIIa activation was measured in washedplatelets by flow cytometry for PAC-1 binding (C). An increase in PAC-1-associated fluorescence was observed after treatment of platelets with thrombin (1 U/mL),but this could be reversed by treatment of platelets with prostacyclin (1 μM) or by preincubating the PAC-1 antibody with RGDS blocking peptide. Histograms arerepresentative of n = 3.



supporting information - pnas · 2013-09-03 · light transmission aggregometry was performed as...

Documents