Alison Cameron-Vendrig,1,2,3,4,5 Adili Reheman,5 M. Ahsan Siraj,3

Xiaohong Ruby Xu,2,5,6 Yiming Wang,5,6 Xi Lei,5 Talat Afroze,2,3,4 Eric Shikatani,2,3,6

Omar El-Mounayri,2,3,4 Hossein Noyan,2,3,4 Ralph Weissleder,7,8 Heyu Ni,1,2,5,6,9,10

and Mansoor Husain1,2,3,4,6,9,11

Glucagon-Like Peptide 1 ReceptorActivation Attenuates PlateletAggregation and ThrombosisDiabetes 2016;65:1714–1723 | DOI: 10.2337/db15-1141

Short-term studies in subjects with diabetes receivingglucagon-like peptide 1 (GLP-1)–targeted therapies havesuggested a reduced number of cardiovascular events.The mechanisms underlying this unexpectedly rapid ef-fect are not known. We cloned full-length GLP-1 receptor(GLP-1R) mRNA from a human megakaryocyte cell line(MEG-01), and found expression levels of GLP-1Rs inMEG-01 cells to be higher than those in the human lungbut lower than in the human pancreas. Incubation withGLP-1 and the GLP-1R agonist exenatide elicited a cAMPresponse in MEG-01 cells, and exenatide significantlyinhibited thrombin-, ADP-, and collagen-induced plateletaggregation. Incubation with exenatide also inhibitedthrombus formation under flow conditions in ex vivoperfusion chambers using human and mouse wholeblood. In a mouse cremaster artery laser injury model,a single intravenous injection of exenatide inhibited throm-bus formation in normoglycemic and hyperglycemic micein vivo. Thrombus formation was greater in mice trans-planted with bone marrow lacking a functional GLP-1R(Glp1r2/2), compared with those receiving wild-type bonemarrow. Although antithrombotic effects of exenatidewere partly lost in mice transplanted with bone marrowfromGlp1r2/2 mice, they were undetectable in mice with agenetic deficiency of endothelial nitric oxide synthase.The inhibition of platelet function and the prevention of

thrombus formation by GLP-1R agonists represent poten-tial mechanisms for reduced atherothrombotic events.

Type 2 diabetes (T2D) is associated with a number of riskfactors that contribute to an increased risk of athero-thrombotic events, including hypertension, dyslipidemia,obesity, and chronic inflammation, as well as endothelialand platelet dysfunction (1). Platelets are small, versatile,anucleate cells in the circulation that play critical roles inboth early and late stages of atherothrombosis, contrib-uting also to cell-based thrombin generation and bloodcoagulation (2). Subjects with T2D exhibit a prothrom-botic state, including increased production of coagulationfactors; decreased production of fibrinolytic factors; and apropensity to platelet activation, aggregation, and adhe-sion (1,3,4). Compounding the latter, subjects with T2Dshow reduced sensitivity to antiplatelet drugs, such as aspi-rin and clopidogrel (5,6), and manifest a higher incidence ofcardiovascular events (1,6,7).

Although the currently available antidiabetic agentshave been effective at lowering blood glucose levels andpreventing microvascular disease, until the recent EMPA-REG study (8), it had been exceedingly difficult to demon-strate the beneficial effects of normalizing blood glucose

1Department of Physiology, University of Toronto, Toronto, Ontario, Canada2Heart and Stroke Richard Lewar Centre of Excellence, University of Toronto,Toronto, Ontario, Canada3Toronto General Research Institute, University Health Network, Toronto, Ontario,Canada4Peter Munk Cardiac Centre, University Health Network, Toronto, Ontario, Canada5Keenan Research Centre, Li Ka Shing Knowledge Institute, St. Michael’s Hos-pital, Toronto, Ontario, Canada6Laboratory Medicine and Pathobiology, University of Toronto, Toronto, Ontario,Canada7Center for Systems Biology, Massachusetts General Hospital, Harvard MedicalSchool, Boston, MA8Department of Systems Biology, Harvard Medical School, Boston, MA9Department of Medicine, University of Toronto, Toronto, Ontario, Canada

10Canadian Blood Services, Toronto, Ontario, Canada11Ted Rogers Centre for Heart Research, University Health Network, Toronto,Ontario, Canada

Corresponding author: Mansoor Husain, mansoor.husain@uhn.ca.

Received 14 August 2015 and accepted 22 February 2016.

This article contains Supplementary Data and a video online at http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db15-1141/-/DC1.

A.C.-V., A.R., H.Ni, and M.H. have contributed equally to this work.

© 2016 by the American Diabetes Association. Readers may use this article aslong as the work is properly cited, the use is educational and not for profit, andthe work is not altered.

See accompanying article, p. 1487.

1714 Diabetes Volume 65, June 2016

COMPLIC

ATIO

NS

levels in major adverse cardiovascular events (i.e., macro-vascular events). Indeed, in some studies, aggressive glu-cose lowering has been associated with an increase in theincidence of cardiovascular events, including death (9).In this context, significant interest has been generated byincretin-targeted therapies, which include glucagon-likepeptide 1 (GLP-1) receptor (GLP-1R) agonists and dipep-tidyl peptidase-4 inhibitors (DPP-4is), both of which areunder active investigation in long-term cardiovascular safetystudies. The first of those studies examining DPP-4is,namely the SAVOR-TIMI 53, EXAMINE, and TECOS stud-ies, have demonstrated no increase in cardiovascular events(10–12). By contrast, relatively short-term studies (13–16),in arguably lower-risk patients treated with GLP-1R ago-nists and DPP-4i, have suggested beneficial effects on car-diovascular event rates. Because the rather modest andbrief improvements in glycemic control observed in thesestudies are unlikely to explain the observed benefits, wewondered whether GLP-1–targeted therapies may reducemacrovascular event rates in these particular patient pop-ulations by directly inhibiting platelet aggregation andthrombus formation. Indeed, this premise could be extrap-olated from the one small uncontrolled study (17) thatdemonstrated the curious ability of DPP-4i sitagliptin toinhibit thrombin-induced platelet aggregation in vitro.

Here we show that a human megakaryocyte cell lineexpresses a functional GLP-1R, and that treatment withthe GLP-1R agonist exenatide inhibits human and mouseplatelet aggregation in vitro, thrombus formation ex vivo,and mouse arterial thrombosis in vivo. Given the impor-tance of nitric oxide (NO) signaling, and more specificallytype III endothelial NO synthase (eNOS)–derived NO, toplatelet biology (18,19), and the known abilities of GLP-1and exenatide to activate eNOS in human endothelial cells(20,21), we here also define a role for eNOS in exenatide-inhibited mouse arterial thrombosis in vivo. Together,these findings provide potential mechanisms for observedimprovements in cardiovascular outcomes in patients withT2D who were treated with GLP-1R agonists, and a poten-tial rationale for their selective use in subjects with diabe-tes who are at particular risk of atherothrombosis.

RESEARCH DESIGN AND METHODS

Cloning and SequencingcDNA was synthesized from MEG-01 DNAse-treated totalRNA. Primers designed to span the 59 and 39 untranslatedregions (UTRs) of the human GLP-1R mRNA were used toamplify a 1.5-kb full-length GLP-1R cDNA using PhusionHigh-Fidelity PCR Master Mix (Thermo Scientific Finn-zymes Molecular Biology Solutions; ThermoFisher Scien-tific, Ottawa, Ontario, Canada). The amplicon was gel elutedusing a QIAquick Gel Extraction kit (catalog #28704;Qiagen, Toronto, Ontario, Canada) and cloned into the ex-pression vector pEF-IRES-puro 6. The resulting clone wassequenced to confirm the orientation of the insert withinthe expression vector as well as to validate that the clonedinsert contained full-length GLP-1R cDNA.

Fluorescent Exenatide ProbeMIN6 cells and MEG-01 cells were washed with Hanks’buffered salt solution and resuspended in binding buffer(as described previously) (22,23) with 100 nmol/L fluo-rescent probe (EP40-BF) or 100 nmol/L unlabeled exena-tide for 1 h in the dark. Cells were washed 23 with Hanks’buffered salt solution and resuspended in PBS with 2%BSA. Flow cytometry was performed using a MACSQuantAnalyzer (Miltenyi Biotec, San Diego, CA), and analysis wasperformed using FCS Express 4 software.

Quantitative RT-PCRTotal RNA was extracted from MEG-01 cells, human skinfibroblast (catalog #CC-2511; Lonza, Mississauga, Ontario,Canada), and human lung tissue using TRIzol (Invitrogen).Total RNA was then reverse transcribed into cDNA. Com-mercially available pancreatic cDNA was used as a positivecontrol for GLP-1R gene expression (catalog #HD-313;Zyagen, San Diego, CA). Previously published GLP-1Rprimers spanning the boundaries of exons 9–10 and 10–11were used to quantify GLP-1R expression via quantitativeRT-PCR (qRT-PCR) (24). The final primer concentrationsused were 0.4 mmol/L with 60°C annealing temperaturefor all primers. All samples were run in triplicate on a RocheLightCycler 480 System, and GLP-1R gene expression wasnormalized to the housekeeping gene TATA binding protein.

cAMP AssaysA cAMP Enzyme Immunoassay Kit (catalog #581001;Cayman Chemicals, Ann Arbor, MI) was used to measurechanges in intracellular cAMP induced by GLP-1. MEG-01cells were cultured to a concentration of 2 3 105/mL andtransferred to a 24-well plate (BD Falcon, San Jose, CA) with500 mL/well culture medium. 3-Isobutyl-1-methylxanthinewas added to each well at a concentration of 0.5 mmol/Land was allowed to incubate for 10 min at 37°C to inhibitphosphodiesterase activity. GLP-1(7–36), in concentrationsof 0.1, 1, 10, and 100 nmol/L, was added to the wells andincubated for 15 min at 37°C. As a positive control forcAMP generation in MEG-01 cells, prostacyclin (PGI2) wasadded at a concentration of 0.02 ng/mL to control wells andwas incubated for 10 min at 37°C. PBS was used as a neg-ative control. After incubation, cells were lysed by addingEDTA to a concentration of 10 mmol/L. Lysed cells weretransferred to 1.5 mL Eppendorf tubes and boiled for 5 minat 95°C. Samples were spun down at 10,000g for 15 min at4°C. Supernatant was removed and frozen at 280°C. ThecAMP Enzyme Immunoassay Kit assay was then performedaccording to manufacturer instructions.

Platelet AggregationAggregation of isolated mouse and human platelets inplatelet-rich plasma or gel-filtered platelet preparationswas assessed at 37°C using a computerized Chrono-Logaggregometer (Chrono-Log Corporation, Havertown, PA),as previously described (25,26). Two hundred fifty micro-liters of a 2–3 3 108/mL platelet solution was added to acylindrical glass cuvette (Chrono-Log Corporation) with asiliconized magnetic stir bar (Chrono-Log Corporation).

diabetes.diabetesjournals.org Cameron-Vendrig and Associates 1715

The aggregometer was set to a stir rate of 1,000 rpm.Both human and murine platelets were stimulated to ag-gregate with various agonists, such as thrombin (0.25–0.5 units), ADP (5 mmol/L), and collagen (5–10 mg/mL),and aggregation was assessed by aggregometer using Agro-Link software (Chrono-Log Corporation). Once stimulated,the assay was allowed to run until the percentage of lighttransmission reached a plateau, and the platelets had thusreached their maximum aggregation. Prior to stimulationwith agonists, platelets were incubated with exenatide(H-8730; Bachem) at increasing concentrations (0.1–10nmol/L) for 15 min at 37°C, or with PBS as a control.

Perfusion Chamber StudiesPerfusion assays were performed as previously described(25–27). Perfusion chamber slides (m-slide VI0.1; ibidi, Mar-tinsried, Germany) were coated with 100 mg/mL collagentype I (Kollagenreagens Horm; Nycomed) at 4°C for 18 h.Whole blood was labeled by adding 1 mmol/L DiOC6. Per-fusion was performed at a flow rate of 1,500 s21 for humanblood and 1,800 s21 for mouse blood using a syringe pump(Harvard Apparatus, Holliston, MA). Images of platelet ad-hesion aggregation and thrombus growth were recordedunder a Zeiss Axiovert 200M Inverted Fluorescent Micro-scope (603 W objective; Zeiss) using a CoolSNAP Camera(Photometrics). Platelet aggregation and thrombus growthwere quantitated by dynamics of platelet mean fluorescenceintensity over the period of 3 min of perfusion using theSlidebook program.

AnimalsAll protocols were approved by the Animal Care Commit-tees of the Toronto General Research Institute or KeenanResearch Centre and conformed to the guidelines of theCanadian Council on Animal Care. C57BL/6 wild-type micewere purchased from Charles River (Montreal, Quebec,Canada), eNOS knockout (KO) (Nos3tm1Unc) mice were pur-chased from The Jackson Laboratory (Bar Harbor, ME), andthe origin, characterization, and genotyping of Glp1r2/2

(KO) mice have been described previously (28). Experimentswere performed on 6- to 12-week-old mice housed for atleast 2 weeks.

Intravital Microscopy Laser Injury Thrombosis ModelMale adult mice were anesthetized, and a tracheal tubewas inserted to facilitate breathing. Antibodies, anestheticreagent (pentobarbital, 0.05 mg/kg body wt; AbbottLaboratories, Toronto, Ontario, Canada), and exenatide(60 nmol/kg body wt i.v.) (29,30) were administered by ajugular vein cannula. The cremaster muscle was preparedunder a dissecting microscope and superfused throughoutthe experiment with preheated bicarbonate buffer saline.Platelets were labeled by injecting a DyLight 649–conjugatedrat anti–mouse GP1bb antibody (0.1 mg/g; EMFRET Analyt-ics). Multiple independent upstream injuries were performedon a cremaster arteriole with the use of an Olympus BX51WIMicroscope with a pulsed nitrogen dye laser. The dynamicaccumulation of fluorescently labeled platelets within the

growing thrombus was captured and analyzed using Slide-book software (Intelligent Imaging Innovations). Blood glu-cose levels were monitored throughout the experiment andremained constant.

Bone Marrow TransplantationSix-week-old male C57BL/6 mice were irradiated with asingle 10-Gy dose of radiation (g Cell 40; Nordion Interna-tional) 3 h prior to transplantation. Bone marrow cells wereharvested from either Glp1r2/2 or wild-type mice femursand resuspended in PBS. A total of 2.0 3 106 bone marrowcells per mouse was injected intravenously with a 28-gaugeneedle. Mice were allowed to recover for 6 weeks prior toundergoing experiments.

Tail Bleeding TimeTo assess the effects of exenatide on a metric that dependson both platelet function and coagulation in vivo, murinetail bleeding assays were also performed (31,32). Mice weregiven a single injection of exenatide (60 nmol/L/kg i.v.),anesthetized, and placed on a heating pad for 10 min.The distal tail (5 mm from the tip) was amputated withsurgical scissors, and the remaining tail was immediatelyimmersed in 10 mL of 0.9% NaCl prewarmed to 37°C.The time required for spontaneous bleeding to cease wasrecorded.

DPP-4 Activity AssayThe CBA085 Innozyme DPP-4 Immunocapture ActivityAssay (Merck Millipore, Darmstadt, Germany) was usedto assess DPP-4 activity in human platelets and MEG-01cells. Protein from MEG-01 cells, human platelets, andlung tissue for use in the assay was extracted as describedabove. Blood was collected from a healthy donor into a BDVacutainer tube (3.2% buffered sodium citrate) and spundown at 190g for 10 min. The plasma fraction was re-moved and stored in aliquots at 280°C for use in theassay. As an experimental control, half of the plasmasamples were treated for 20 min with 20 mg/mL MK-626 (Merck Millipore) to block DPP-4 activity. The re-mainder of the assay was performed according to themanufacturer instructions.

RESULTS

Human Megakaryocytes and Platelets Express aFunctional GLP-1RGiven the variety of concerns with currently availableGLP-1R antibodies (33), we began our studies by explor-ing the expression of GLP-1R at the mRNA level. Usingprimers designed to span the entire GLP-1R coding se-quence, including both 59 and 39 UTRs, we were able toclone and sequence full-length human GLP-1R mRNAfrom a human megakaryocyte cell line (MEG-01). Analysisof the transcript size and sequence confirmed identitywith the known human GLP-1R (Fig. 1A) (GenBank Ac-cession #KR138540).

Next, qRT-PCR was used to measure the levels of GLP-1R mRNA expressed in MEG-01 cells relative to other humantissues. This analysis revealed GLP-1R mRNA expression

1716 Antithrombotic Effects of GLP-1R Agonists Diabetes Volume 65, June 2016

levels in MEG-01 cells to be,20% of the levels in the humanpancreas, but .10-fold more than in the human lung. NoGLP-1R signal was detected in negative control samples thatincluded human skin fibroblasts and non–reverse tran-scribed MEG-01 mRNA (Fig. 1B). Furthermore, the bind-ing of a fluorescently tagged exenatide peptide that wasdeveloped to probe GLP-1R expression (22) was evidentin MEG-01 cells (Fig. 1C).

The GLP-1R is a Gas-coupled receptor, and its activa-tion in the pancreatic b-cell leads to adenylate cyclase–dependent increases in cAMP (34). Accordingly, intracellularcAMP responses to GLP-1 and the GLP-1R agonist exena-tide were used to test the functional activity of the GLP-1Rexpressed in MEG-01 cells. These experiments showed dose-dependent increases in intracellular cAMP in response toGLP-1 (Fig. 2A) and exenatide (Fig. 2B), with maximalresponses observed at the 10 nmol/L concentration ofboth agonists. Although the maximal cAMP response pro-duced by these agents was considerably less than thatevoked by PGI2 (Fig. 2), the level of cAMP achieved withGLP-1 and exenatide is similar to that achieved in othercell types known to express the GLP-1R (35).

Exenatide Inhibits Platelet Aggregation In Vitro andThrombus Formation Under Flow Conditions Ex VivoTo assess the consequences of GLP-1R activation on plateletaggregation, both human and mouse platelets were in-cubated with increasing concentrations of exenatide (0.1–10nmol/L) and were stimulated to aggregate with thrombin.On aggregometry traces, the first wave of thrombin-inducedaggregation is known to correspond to platelet activation,

with the second steeper wave corresponding to plateleta-granule and dense-granule release (36). At 6 min afterstimulation with thrombin, during the second wave ofaggregation, exenatide-treated human (Fig. 3A and B)and mouse (Fig. 3C and D) platelets manifest decreasedaggregation in gel-filtered platelets compared with PBS-treated control cells, suggesting an inhibitory effect ofGLP-1R activation on platelet granule release. Similar in-hibitory effects of exenatide were also evident in humanplatelet aggregation stimulated with ADP and collagen inplatelet-rich plasma and gel-filtered platelets, respectively(Supplementary Fig. 1).

Using collagen-coated glass perfusion chambers, fluo-rescently labeled platelet aggregation and thrombus forma-tion in whole human and mouse blood were monitoredunder a high shear rate (human 1,500 s21, mouse 1,800 s21)in exenatide-treated (10 nmol/L) versus PBS-treated(equal volume) blood. Even greater than the effects ofexenatide on platelet aggregation in vitro, thrombus for-mation in whole blood from both human (Fig. 4A and B)and mouse (Fig. 4C and D) was markedly inhibited by theGLP-1R agonist ex vivo.

Exenatide Inhibits Thrombus Growth In VivoUsing our well-established quantitative laser injury cre-master arterial thrombosis model (26,31,32), we next de-termined that exenatide inhibited thrombus growth inadult male mice compared with PBS-injected controls.Moreover, the thrombi that did form in exenatide-treatedanimals were smaller, detached more easily, and embol-ized more rapidly compared with PBS-treated controls(Fig. 5A and B and Supplementary Video).

Similar experiments were performed in mice withstreptozotocin-induced diabetes, because this model has

Figure 1—The human megakaryocyte cell line MEG-01 expressesGLP-1R. A: RT-PCR results of MEG-01 RNA using primers span-ning the full-length sequence (59-39 UTR) of human GLP-1R. The leftpanel shows a 1-kb ladder, and the right panel shows the productof the PCR; an ;1.5-kb band corresponding to the size of knownGLP-1R mRNA was amplified. This band was gel eluted, cloned,and sequenced to reveal full-length human GLP-1R (GenBank Ac-cession #KR138540). B: qRT-PCR was used to confirm the expres-sion of GLP-1R mRNA in MEG-01 cells, revealing expression levelsrelative to other tissues known to express GLP-1R (n = 3). C: Bindingof a GLP-1R agonist to MIN6 (insulinoma cell line; positive control)and MEG-01 cells was assessed using flow cytometry with EP40-BF,a fluorescently labeled exenatide peptide. Error bars represent SEM.Two-way ANOVA was performed with Bonferroni post hoc tests;N = 3/group. *P < 0.05, #P < 0.01. MFI, mean fluorescence in-tensity. RT, reverse transcribed.

Figure 2—The human megakaryocyte cell line MEG-01 manifests acAMP response to GLP-1R agonists. Incubation of MEG-01 cellswith both native GLP-1 (A) and GLP-1R agonist exenatide (B) re-sults in an increase in intracellular cAMP. Error bars represent SEM.One-way ANOVAs were performed with Bonferroni post hoc tests;N = 3/group. *P < 0.05, #P < 0.01 vs. PBS (untreated control). PGI2represents a positive control.

diabetes.diabetesjournals.org Cameron-Vendrig and Associates 1717

previously been shown to have enhanced thrombosis (37).As expected, greater thrombus formation was observedin hyperglycemic mice compared with normoglycemiccontrols, with exenatide inhibiting thrombus growth toa similar degree in both models (Fig. 5C and D).

To dissect whether the antithrombotic effects ofexenatide observed in vivo were the result of GLP-1Ractivation on circulating cells, such as platelets, or werethe potential consequence of putative GLP-1R–mediatedeffects on endothelial or other cell types, lethally irradi-ated mice were transplanted with bone marrow fromGlp1r2/2 versus wild-type mice. After successful bonemarrow reconstitution (6 weeks), recipient mice weresubjected to the cremaster arterial laser injury model.Importantly, mice reconstituted with bone marrow thatlacked a functional GLP-1R had greater thrombus for-mation than those that received wild-type bone marrow(Fig. 6, green vs. gray). Furthermore, the antithromboticeffect of exenatide was markedly attenuated in micelacking a functional GLP-1R in bone marrow–derivedcells (Fig. 6, pink vs. red). Together, these results suggest

a physiological role for GLP-1R in bone marrow–derivedcells in thrombosis.

However, several studies (38,39) have also shown thatexenatide is a vasodilator, and that exenatide treatmentin human umbilical vein cells causes the release of NO,which, in addition to being a vasodilator, is a potent in-hibitor of platelet function (20). As such, we also investi-gated whether exenatide would function in mice with agenetic absence of the endothelial NOS gene (eNOS2/2).Compared with PBS-injected controls, the extent of in vivothrombus formation after laser injury was not reduced byexenatide in eNOS2/2 mice (Fig. 7).

Finally, to exclude the possibility of an effect of exenatideon coagulation and hemostasis, mouse tail vein bleedingtimes were measured. No significant differences in bleedingtimes were observed between exenatide- and PBS-treatedanimals with (Supplementary Fig. 2A, wild-type) or without(Supplementary Fig. 2B, Glp1r2/2) a functional GLP-1R.

Neither Platelets Nor MEG-01 Cells Harbor DPP-4ActivityGiven that DPP-4is are widely used members of the incretinclass of antidiabetic drugs, we also sought to explore therole of DPP-4 in platelets and the megakaryocyte cell lineused in this study. DPP-4 is an aminopeptidase that isexpressed in many tissue cell types as a membrane-spanningprotein, and is also found in the plasma in a soluble form. Animmunocapture activity assay (Merck Chemicals) revealedthat neither MEG-01 cells nor platelets harbor any endog-enous DPP-4 activity (Supplementary Fig. 3).

DISCUSSION

The current study shows that human megakaryocytesexpress a functional GLP-1R, and that GLP-1R activationwith the GLP-1R agonist exenatide plays a role in inhibitingmouse and human platelet aggregation in vitro, mouse andhuman thrombus formation under flow conditions ex vivo,and mouse thrombus formation in vivo. This may haveimportant implications for subjects with T2D, who haveplatelet dysfunction and are at an increased risk of adversemacrovascular events.

In a meta-analysis, Monami et al. (40) suggested thatGLP-1–targeted therapies may reduce the risk of acute myo-cardial infarction and cardiovascular mortality in T2D pa-tients. Moreover, they suggested that these beneficial effectsare the result of mechanisms other than reductions in con-ventional risk factors such as obesity, blood pressure, andlipids. In this context, our report on the antiplatelet andantithrombotic effects of the GLP-1 analog exenatide repre-sents a potential mechanism underlying these short-termclinical findings, in what were arguably smaller groups inlower-risk populations than those being studied in severallarge cardiovascular safety studies currently underway.

The investigators of ELIXA, a cardiovascular safetystudy of the GLP-1 analog lixisenatide in a high-risk post–acute coronary syndrome population of subjects with

Figure 3—Treatment with a GLP-1R agonist attenuates thrombin-induced platelet aggregation in vitro. Representative aggregometrytraces of human (A) and mouse (C) platelets incubated for 15 minwith exenatide (0.1–10 nmol/L) and stimulated to aggregate with0.25 units (A) or 0.5 units (C) of thrombin. Incubation for 15 minwith exenatide (Ex) inhibits the second wave of aggregation, mea-sured 6 min after thrombin stimulation in human (B) and mouse(D) platelets. B and D: Error bars represent the SEM. One-wayANOVA was performed with Dunnett multiple-comparisons test;N = 3–7/group. *P < 0.05; #P < 0.01 vs. PBS.

1718 Antithrombotic Effects of GLP-1R Agonists Diabetes Volume 65, June 2016

diabetes, have recently published results (41) stating thatlixisenatide was noninferior to placebo for cardiovascularsafety. Based on this early finding, we can only speculateas to whether the biology we report here will translateinto reductions in thrombosis-dependent major adversecardiovascular events in other studies.

To our knowledge, there exist no published data onthe effects of GLP-1 or its analogs on platelet functionand thrombosis. Regrettably, neither the small mecha-nistic clinical studies of GLP-1 analogs published to datenor the larger cardiovascular safety studies currentlyunderway have included substudies aimed at plateletfunction. Having said that, our search of clinicaltrials.govrevealed one study (NCT01408862, yet to be published)of 20 healthy participants, whose updated informationsuggests that GLP-1R is expressed on human platelets.However, this online update also suggests that neitherGLP-1 nor its metabolite GLP-1(9–36) had an effect onthe aggregation of platelets isolated from these healthyvolunteers. Given our results, we believe that humanplatelet function should be tested in clinical studies ofGLP-1 analogs, particularly those including exenatide orits longer-acting formulations.

Interestingly, Gupta et al. (17) proposed that sitaglip-tin, a DPP-4i, inhibits platelet aggregation by interfering

with tyrosine phosphorylation of the platelet plasmamembrane Ca2+ ATPase channel, thereby limiting theaccumulation of intracellular Ca2+. Given that thereshould have been no GLP-1 present in their plateletpreparations, and that we have been unable to documentany DPP-4 activity in platelets (or MEG-01 cells) (Sup-plementary Fig. 3), the results of Gupta et al. (17) aresomewhat surprising. We can only speculate as towhether their results in vitro suggest a “direct” actionof sitagliptin on platelet biology (i.e., independent ofGLP-1 or DPP-4). With regard to their demonstrationof an in vivo effect of sitagliptin on platelet aggregation,their use of a small group of subjects with diabetes (N =50) treated with sitagliptin for 1 and 3 months wasneither blinded nor controlled with placebo or an activecomparator.

Our study also has several limitations. First, we havebeen unable to directly demonstrate the existence ofGLP-1R protein on megakaryocytes or platelets. Recently,commercially available antibodies for GLP-1R have beenshown to be nonspecific (42). Although we did obtain anew antibody believed to recognize human GLP-1R (33),this reagent did not show sufficient specificity for GLP-1Rin our hands (data not shown). Having said that, ourevidence implicating GLP-1R–dependent platelet biology

Figure 4—Treatment with a GLP-1R agonist inhibits collagen-induced thrombus growth under flow conditions ex vivo. Incubation for15 min with exenatide (10 nmol/L) significantly inhibited thrombus growth in a perfusion chamber. Whole human (A) and mouse (C) bloodwas perfused through collagen-coated chambers at shear rates of 1,500 s21 (A) and 1,800 s21 (C ) for 3 min. Fluorescently labeled plateletaccumulation was recorded in real time, and the platelet mean fluorescence intensity was quantified. The panels show representativeimages captured after 1 and 3 min of perfusion. After incubation with exenatide (10 nmol/L), human (B) and mouse (D) thrombus growth inthe perfusion chamber was reduced compared with PBS-treated controls. N = 3 separate experiments. P < 0.05 by Student t testcomparing area under the curve.

diabetes.diabetesjournals.org Cameron-Vendrig and Associates 1719

is substantive, including the cloning and sequencing offull-length GLP-1R mRNA from MEG-01 cells, the bindingof a fluorescent exenatide to MEG-01 cells, and the func-tional studies described above. Second, the results of ourbone marrow transplant experiments suggest that a celltype other than bone marrow–derived cells may also con-tribute to the antithrombotic effects of GLP-1R activa-tion. Although in vivo thrombus formation still occurredin mice receiving Glp1r2/2 bone marrow, the extent ofthrombosis was considerably less than that observed inmice receiving wild-type bone marrow (Fig. 6). As such,we have not excluded the possibility that exenatide me-diates antithrombotic effects, at least in part, via effectson the endothelium. Although our results in eNOS2/2

mice suggest that this putative downstream target ofGLP-1R activation is central to the in vivo antithromboticeffects of exenatide (Fig. 7), it remains controversial as towhether platelets express eNOS (18,43). Because of this,the contribution of endothelial cells to the platelet-inhibiting effects of exenatide will require future studieswith an endothelial cell–specific KO of GLP-1R.

The paradigm of endothelium-derived signaling mole-cules playing important roles in platelet function andthrombosis is well established. For example, endothelium-derived PGI2 inhibits platelet aggregation by activating aG-protein–coupled receptor in platelets, increasing theirintracellular cAMP levels through adenylate cyclase, withsubsequent cAMP/cAMP-dependent protein kinase sig-naling inhibiting virtually all platelet-activating mecha-nisms (44,45). In this context, intracellular cAMP levelsin platelets play a key role in maintaining hemostasis.Given our demonstration of both GLP-1 and exenatidecausing increased cAMP levels in megakaryocytes, webelieve that similar mechanisms mediate the ability ofexenatide to directly inhibit platelet aggregation in vitro.In addition to this, the more potent antithrombotic ef-fects of exenatide observed in vivo appear to depend onits activation of eNOS.

The expression of a functional GLP-1R on megakar-yocytes and the inhibitory effects of GLP-1 and exena-tide on platelet aggregation and thrombus formationhave several implications. First, this could help to

Figure 5—Treatment with a GLP-1R agonist inhibits laser injury–induced thrombus growth in vivo. A laser injury was induced in thecremaster artery of anesthetized normoglycemic and hyperglycemic mice, and thrombus growth at the site of injury was monitored andquantified using fluorescence-tagged platelets. After a single injection of exenatide (60 nmol/kg i.v.), fluorescently labeled platelet accu-mulation at the site of the laser injury was reduced in both wild-type normoglycemic mice (A) and wild-type mice with streptozotocin-induced hyperglycemia (C). Both wild-type normoglycemic (B) and hyperglycemic (D) mice had smaller and less durable thrombus growthafter exenatide (60 nmol/kg) treatment compared with PBS-treated controls. N = 3 separate experiments. P < 0.01 for all comparisonsby Student t test.

1720 Antithrombotic Effects of GLP-1R Agonists Diabetes Volume 65, June 2016

explain early clinical data suggesting improved cardio-vascular outcomes in GLP-1 analog–treated patients(16). Second, T2D is associated with increased plateletreactivity and risk of thrombus formation, higher inci-dences of myocardial infarction and stroke (46), and di-minished sensitivity to widely used antiplatelet drugs,such as aspirin and clopidogrel (5). In this context, oursuggestion that exenatide might reverse the “prothrom-botic” phenotype of T2D is of clinical significance. In-deed, if our findings translate to subjects with T2D who

have high cardiovascular risk, treatments that include GLP-1R activation may help to reduce the burden of major ad-verse cardiovascular events in this population. Third, becauseplatelets also play a role in atherosclerotic plaque formationand leukocyte activation (47–49), inhibiting platelet activa-tion with a GLP-1 analog may also contribute to antiathero-genic and anti-inflammatory effects (30,50). Fourth,because intravenous exenatide did not prolong mousetail-bleeding time, its antithrombotic actions are notaccompanied by effects on coagulation or hemostasis.

Figure 6—Enhanced thrombosis and only partial loss of the antithrombotic effects of a GLP-1R agonist in mice with bone marrow lacking afunctional GLP-1R. Lethally irradiated wild-type (WT) mice reconstituted with bone marrow transplant (BMT) from WT (A, upper panel) orGlp1r2/2 (KO, A, lower panel) mice were subjected to the laser injury–induced cremaster arteriole thrombosis model under intravitalmicroscopy. A: Upper panel: Representative pictures of thrombus formation in BMT-WT mice treated with control buffer (left) or exenatide(right). Lower panel: BMT-KO mice treated with control buffer (left) or exenatide (right). B: Platelet mean fluorescence intensity of thrombi.Mice that received GLP-1R KO bone marrow had increased thrombus growth (green) compared with mice receiving WT bone marrow(gray). Treatment with exenatide inhibited thrombus growth to a greater extent in mice with WT bone marrow (red) than in mice with GLP-1RKO bone marrow (pink), although exenatide still reduced thrombus growth in mice transplanted with GLP-1R KO bone marrow (pink)compared with PBS-treated control mice with GLP-1R KO bone marrow (green). The kinetic curves represent platelet mean fluorescenceintensity, and the shaded regions are representative of SEM. N = 3 separate experiments. P < 0.01 for all comparisons by Student t test.

Figure 7—eNOS regulates the antiplatelet activity of exenatide. A laser injury was induced in the cremaster artery of anesthetized eNOS2/2

mice, and thrombus growth at the site of injury was monitored and quantified using fluorescence-tagged platelets. After a single in-travenous injection of exenatide (Ex; 60 nmol/kg), fluorescently labeled platelet accumulation at the site of the laser injury was not reducedin eNOS2/2 mice compared with saline-treated controls. A: Representative images captured 1 min after laser injury. B: Average of N = 5thrombi. P = NS.

diabetes.diabetesjournals.org Cameron-Vendrig and Associates 1721

This important finding argues against the possibilityof exenatide causing an excess of bleeding complications.Taken together, these implications may confer significanttherapeutic advantage to GLP-1R agonists.

Acknowledgments. The authors thank Dr. Tianru Jin, Dr. MichaelWheeler, and Dr. Daniel Drucker of the University of Toronto for helpful dis-cussion and access to specific reagents.Funding. This work was funded in part by an Operating Grant to H.Ni from theCanadian Institutes of Health Research (MOP 119540), an Investigator-InitiatedResearch grant to H.Ni and M.H. from Merck Canada (IISP #40363), and a Grant-in-Aid to M.H. from the Heart and Stroke Foundation of Canada (T6757). M.H.was supported by a Career Investigator Award from the Heart and Stroke Foun-dation, Ontario Provincial Office.Duality of Interest. The discoveries described in this report form part ofU.S. provisional patent application US61/721,819. M.H. has received other in-vestigator-initiated research support from Amylin/BMS/Astra-Zeneca and NovoNordisk. M.H. has received consulting fees and/or speaking honoraria from Astra-Zeneca, Boehringer-Ingelheim, Eli Lilly, Janssen, Merck, Novo Nordisk, and Roche.No other potential conflicts of interest relevant to this article were reported.Author Contributions. A.C.-V., A.R., and M.A.S. conceived, designed, andconducted the experiments; analyzed the data; and wrote and revised themanuscript. X.R.X., X.L., O.E.-M., and H.No. conducted the experiments andanalyzed the data. Y.W. and T.A. conceived, designed, and conducted theexperiments and analyzed the data. E.S. conducted the experiments, analyzed thedata, and wrote and revised the manuscript. H.Ni and M.H. conceived anddesigned the experiments and wrote and revised the manuscript. R.W. contributedkey reagents and expertise for the project. H.Ni and M.H. are the guarantors ofthis work and, as such, had full access to all the data in the study and takeresponsibility for the integrity of the data and the accuracy of the data analysis.

References1. Vinik AI, Erbas T, Park TS, Nolan R, Pittenger GL. Platelet dysfunction intype 2 diabetes. Diabetes Care 2001;24:1476–14852. Hou Y, Carrim N, Wang Y, Gallant RC, Marshall A, Ni H. Platelets in he-mostasis and thrombosis: novel mechanisms of fibrinogen-independent plateletaggregation and fibronectin-mediated protein wave of hemostasis. J Biomed Res.30 October 2015 [Epub ahead of print]. DOI: 10.7555/JBR.29.201501213. Festa A, D’Agostino R Jr, Mykkänen L, et al.; The Insulin Resistance Ath-erosclerosis Study (IRAS). Relative contribution of insulin and its precursors tofibrinogen and PAI-1 in a large population with different states of glucose tol-erance. Arterioscler Thromb Vasc Biol 1999;19:562–5684. Ni H. The platelet “sugar high” in diabetes. Blood 2012;119:5949–59515. Angiolillo DJ, Fernandez-Ortiz A, Bernardo E, et al. Platelet function profilesin patients with type 2 diabetes and coronary artery disease on combined aspirinand clopidogrel treatment. Diabetes 2005;54:2430–24356. Mylotte D, Kavanagh GF, Peace AJ, et al. Platelet reactivity in type 2 di-abetes mellitus: a comparative analysis with survivors of myocardial infarctionand the role of glycaemic control. Platelets 2012;23:439–4467. Angiolillo DJ, Bernardo E, Sabaté M, et al. Impact of platelet reactivity oncardiovascular outcomes in patients with type 2 diabetes mellitus and coronaryartery disease. J Am Coll Cardiol 2007;50:1541–15478. Zinman B, Wanner C, Lachin JM, et al.; EMPA-REG OUTCOME Investigators.Empagliflozin, cardiovascular outcomes, and mortality in type 2 diabetes. N EnglJ Med 2015;373:2117–21289. Gerstein HC, Miller ME, Genuth S, et al.; ACCORD Study Group. Long-termeffects of intensive glucose lowering on cardiovascular outcomes. N Engl J Med2011;364:818–82810. Scirica BM, Bhatt DL, Braunwald E, et al.; SAVOR-TIMI 53 Steering Com-mittee and Investigators. Saxagliptin and cardiovascular outcomes in patientswith type 2 diabetes mellitus. N Engl J Med 2013;369:1317–1326

11. White WB, Cannon CP, Heller SR, et al.; EXAMINE Investigators. Alogliptinafter acute coronary syndrome in patients with type 2 diabetes. N Engl J Med2013;369:1327–133512. Green JB, Bethel MA, Armstrong PW, et al.; TECOS Study Group. Effect ofsitagliptin on cardiovascular outcomes in type 2 diabetes. N Engl J Med 2015;373:232–24213. Monami M, Dicembrini I, Nardini C, Fiordelli I, Mannucci E. Effects of glu-cagon-like peptide-1 receptor agonists on cardiovascular risk: a meta-analysis ofrandomized clinical trials. Diabetes Obes Metab 2014;16:38–4714. Frederich R, Alexander JH, Fiedorek FT, et al. A systematic assessment ofcardiovascular outcomes in the saxagliptin drug development program for type 2diabetes. Postgrad Med 2010;122:16–2715. Engel SS, Golm GT, Shapiro D, Davies MJ, Kaufman KD, Goldstein BJ.Cardiovascular safety of sitagliptin in patients with type 2 diabetes mellitus: apooled analysis. Cardiovasc Diabetol 2013;12:316. Monami M, Dicembrini I, Nardini C, Fiordelli I, Mannucci E. Effects of glu-cagon-like peptide-1 receptor agonists on cardiovascular risk: a metanalysis ofrandomised clinical trials. Diabetes Obes Metab 2014;16:38–4717. Gupta AK, Verma AK, Kailashiya J, Singh SK, Kumar N. Sitagliptin: anti-platelet effect in diabetes and healthy volunteers. Platelets 2012;23:565–57018. Randriamboavonjy V, Fleming I. Endothelial nitric oxide synthase (eNOS)in platelets: how is it regulated and what is it doing there? Pharmacol Rep 2005;57(Suppl.):59–6519. Naseem KM, Riba R. Unresolved roles of platelet nitric oxide synthase.J Thromb Haemost 2008;6:10–1920. Ding L, Zhang J. Glucagon-like peptide-1 activates endothelial nitric oxidesynthase in human umbilical vein endothelial cells. Acta Pharmacol Sin 2012;33:75–8121. Erdogdu O, Nathanson D, Sjöholm A, Nyström T, Zhang Q. Exendin-4stimulates proliferation of human coronary artery endothelial cells througheNOS-, PKA- and PI3K/Akt-dependent pathways and requires GLP-1 receptor.Mol Cell Endocrinol 2010;325:26–3522. Clardy SM, Keliher EJ, Mohan JF, et al. Fluorescent exendin-4 derivativesfor pancreatic b-cell analysis. Bioconjug Chem 2014;25:171–17723. Reiner T, Thurber G, Gaglia J, et al. Accurate measurement of pancreaticislet beta-cell mass using a second-generation fluorescent exendin-4 analog.Proc Natl Acad Sci USA 2011;108:12815–1282024. Gupta NA, Mells J, Dunham RM, et al. Glucagon-like peptide-1 receptor ispresent on human hepatocytes and has a direct role in decreasing hepaticsteatosis in vitro by modulating elements of the insulin signaling pathway.Hepatology 2010;51:1584–159225. Yang H, Reheman A, Chen P, et al. Fibrinogen and von Willebrand factor-independent platelet aggregation in vitro and in vivo. J Thromb Haemost 2006;4:2230–223726. Reheman A, Yang H, Zhu G, et al. Plasma fibronectin depletion enhancesplatelet aggregation and thrombus formation in mice lacking fibrinogen and vonWillebrand factor. Blood 2009;113:1809–181727. Ni H, Denis CV, Subbarao S, et al. Persistence of platelet thrombus for-mation in arterioles of mice lacking both von Willebrand factor and fibrinogen.J Clin Invest 2000;106:385–39228. Scrocchi LA, Brown TJ, MaClusky N, et al. Glucose intolerance but normalsatiety in mice with a null mutation in the glucagon-like peptide 1 receptor gene.Nat Med 1996;2:1254–125829. Goto H, Nomiyama T, Mita T, et al. Exendin-4, a glucagon-like peptide-1receptor agonist, reduces intimal thickening after vascular injury. Biochem Bio-phys Res Commun 2011;405:79–8430. Arakawa M, Mita T, Azuma K, et al. Inhibition of monocyte adhesion toendothelial cells and attenuation of atherosclerotic lesion by a glucagon-likepeptide-1 receptor agonist, exendin-4. Diabetes 2010;59:1030–103731. Lei X, Reheman A, Hou Y, et al. Anfibatide, a novel GPIb complex antagonist,inhibits platelet adhesion and thrombus formation in vitro and in vivo in murinemodels of thrombosis. Thromb Haemost 2014;111:279–289

1722 Antithrombotic Effects of GLP-1R Agonists Diabetes Volume 65, June 2016

32. Wang Y, Reheman A, Spring CM, et al. Plasma fibronectin supports he-mostasis and regulates thrombosis. J Clin Invest 2014;124:4281–429333. Pyke C, Knudsen LB. The glucagon-like peptide-1 receptor—or not? En-docrinology 2013;154:4–834. Peyot ML, Gray JP, Lamontagne J, et al. Glucagon-like peptide-1 inducedsignaling and insulin secretion do not drive fuel and energy metabolism in pri-mary rodent pancreatic beta-cells. PLoS One 2009;4:e622135. Ramos LS, Zippin JH, Kamenetsky M, Buck J, Levin LR. Glucose and GLP-1stimulate cAMP production via distinct adenylyl cyclases in INS-1E insulinomacells. J Gen Physiol 2008;132:329–33836. Practical-Haemostasis.com. Platelet function testing: light transmissionaggregometry [LTA] [article online], 2012. Available from http://www.practical-haemostasis.com/Platelets/platelet_function_testing_lta.html. 17 May 201537. Stolla MC, Li D, Lu L, Woulfe DS. Enhanced platelet activity and thrombosisin a murine model of type I diabetes are partially insulin-like growth factor1-dependent and phosphoinositide 3-kinase-dependent. J Thromb Haemost2013;11:919–92938. Sélley E, Kun S, Szijártó IA, et al. Exenatide induces aortic vasodilationincreasing hydrogen sulphide, carbon monoxide and nitric oxide production.Cardiovasc Diabetol 2014;13:6939. Dandona P, Chaudhuri A, Dhindsa S. A novel antihypertensive effect ofexenatide, a GLP-1 agonist. Am J Hypertens 2010;23:22840. Monami M, Ahrén B, Dicembrini I, Mannucci E. Dipeptidyl peptidase-4 in-hibitors and cardiovascular risk: a meta-analysis of randomized clinical trials.Diabetes Obes Metab 2013;15:112–12041. Pfeffer MA, Claggett B, Diaz R, et al.; ELIXA Investigators. Lixisenatide inpatients with type 2 diabetes and acute coronary syndrome. N Engl J Med 2015;373:2247–2257

42. Panjwani N, Mulvihill EE, Longuet C, et al. GLP-1 receptor activation in-directly reduces hepatic lipid accumulation but does not attenuate developmentof atherosclerosis in diabetic male ApoE(-/-) mice. Endocrinology 2013;154:127–13943. Gambaryan S, Kobsar A, Hartmann S, et al. NO-synthase-/NO-independentregulation of human and murine platelet soluble guanylyl cyclase activity. JThromb Haemost 2008;6:1376–138444. Raslan Z, Aburima A, Naseem KM. The spatiotemporal regulation of cAMPsignaling in blood platelets-old friends and new players. Front Pharmacol 2015;6:26645. Beck F, Geiger J, Gambaryan S, et al. Time-resolved characterization ofcAMP/PKA-dependent signaling reveals that platelet inhibition is a concertedprocess involving multiple signaling pathways. Blood 2014;123:e1–e1046. Alexopoulos D, Xanthopoulou I, Tsigkas G, et al. Intrinsic platelet reactivityand thrombus burden in patients with ST-elevation myocardial infarction. ThrombRes 2013;131:333–33747. Lievens D, von Hundelshausen P. Platelets in atherosclerosis. ThrombHaemost 2011;106:827–83848. Lievens D, Zernecke A, Seijkens T, et al. Platelet CD40L mediatesthrombotic and inflammatory processes in atherosclerosis. Blood 2010;116:4317–432749. Murphy AJ, Bijl N, Yvan-Charvet L, et al. Cholesterol efflux in megakaryocyteprogenitors suppresses platelet production and thrombocytosis. Nat Med 2013;19:586–59450. Kodera R, Shikata K, Kataoka HU, et al. Glucagon-like peptide-1 receptoragonist ameliorates renal injury through its anti-inflammatory action withoutlowering blood glucose level in a rat model of type 1 diabetes. Diabetologia 2011;54:965–978

diabetes.diabetesjournals.org Cameron-Vendrig and Associates 1723