TR-05486; No of Pages 7 Thrombosis Research xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
Thrombosis Research journal homepage: www.elsevier.com/locate/thromres
Regular Article
Zucker Diabetic Fatty rats exhibit hypercoagulability and accelerated thrombus formation in the Arterio-Venous shunt model of thrombosis Jin Shang ⁎, Zhu Chen ⁎⁎, Min Wang, Qiu Li, Wen Feng, Yangsong Wu, Weizhen Wu, Michael P. Graziano, Madhu Chintala Department of Cardiometabolic Disease, Merck Research Laboratories, Kenilworth, New Jersey
a r t i c l e
i n f o
Article history: Received 10 January 2014 Received in revised form 6 April 2014 Accepted 8 April 2014 Available online xxxx Keywords: Rat Thrombin generation Coagulation Platelet reactivity Type 2 Diabetes Thrombus formation
a b s t r a c t Introduction: Diabetes is a significant risk factor for thrombosis. The present study aimed at assessing coagulability, platelet reactivity, and thrombogenicity of the diabetic female Zucker Diabetic Fatty (ZDF) rat model and its relevance in studying antithrombotic mechanisms. Materials and Methods: The basal coagulant state in ZDF rats was evaluated by clotting times, thromboelastography, and thrombin generation assay. A 14-day treatment with dapagliflozin in ZDF rats was pursued to investigate if glycemic control can improve coagulability. Thrombus formation in the Arterio-Venous (A-V) shunt model and the FeCl3-induced arterial thrombosis model was studied, with the antithrombotic effect of apixaban in the former model further investigated. Results: ZDF rats exhibited significantly shortened clotting times, enhanced thrombin generation, and decreased fibrinolysis at baseline. Effective glycemic control achieved with dapagliflozin did not improve any of these parameters. ZDF rats displayed accelerated thrombus formation and were amenable to apixaban treatment in the A-V shunt model albeit with less sensitivity than normal rats. ZDF rats exhibited less platelet aggregation in response to ADP, collagen and PAR-4, and attenuated thrombotic response in the FeCl3 model. Conclusions: ZDF rats are at a chronic hypercoagulable and hypofibrinolytic state yet with compromised platelet reactivity. They display accelerated and attenuated thrombosis in the A-V shunt and FeCl3 model of thrombosis, respectively. Results shed new light on the pathophysiology of the ZDF rat model and illustrate its potential value in translational research on anticoagulant agents in diabetics. Caution needs to be exerted in utilizing this model in assessing antiplatelet mechanisms in diabetes-associated atherothrombosis. © 2014 Elsevier Ltd. All rights reserved.
Introduction Thrombosis, both venous and arterial, is a major cause of morbidity and mortality worldwide [1]. Diabetes is one important risk factor for increased thrombotic events in patients with cardiovascular diseases [2]. Metabolic changes, including hyperinsulinemia, hypertriglyceridemia, altered adipokines and low-grade systemic inflammation, have been identified as contributors to platelet hyperreactivity in diabetic
Abbreviations: aPTT, activated partial thromboplastin time; A-V, Arterio-Venous; Dapa, dapagliflozin; ETP, endogenous thrombin potential; FXa, coagulation factor Xa; MI, myocardial infarction; Min, minute; PRP, platelet-rich plasma; PT, prothrombin time; s, second; SD, Sprague–Dawley; TEG, Thromboelastography; TGA, Thrombin generation assay; VTE, venous thromboembolism; ZDF, Zucker Diabetic Fatty; ZL, Zucker Lean. ⁎ Correspondence to: J. Shang, Cardiometabolic Disease, Merck Research Laboratories K15-3C-309, Galloping Hill Road, Kenilworth, NJ 07033, USA. Tel.: +1 908 740 0131. ⁎⁎ Correspondence to: Z. Chen, Cardiometabolic Disease, Merck Research Laboratories K15-3A-309, Galloping Hill Road, Kenilworth, NJ 07033, USA. Tel.: +1 908 740 0642. E-mail addresses: [email protected] (J. Shang), [email protected] (Z. Chen).
patients [3]. Moreover, diabetes is associated with a hypercoagulant and hypofibrinolytic state, in which altered coagulation factors and endothelial dysfunction have been found [3]. It is also important to recognize that despite the current glucose lowering or antithrombotic therapies, diabetic patients remain at high risk for cardiovascular events [4,5]. Development of novel therapies that target the residual prothrombotic risk in diabetics is thus an unmet medical need. In preclinical studies, the efficacy of novel antithrombotic agents has been typically evaluated in normal and healthy animals in provoked models for thrombosis. In contrast, disease models that possess metabolic characteristics similar to diabetic patients may allow translational research in studying and evaluating novel antithrombotic agents that could be targeted to high-risk, diabetic patients. The ob/ob and db/db models are widely used for diabetes research. Mice homozygous for the obese spontaneous mutation (Lepob, leptin deficiency) are commonly referred to as ob/ob [6], and mice carrying the diabetes spontaneous mutation (Leprdb, a point mutation in the leptin receptor (Lepr) gene) are referred to as db/db [7]. It has been reported that the ob/ob
http://dx.doi.org/10.1016/j.thromres.2014.04.008 0049-3848/© 2014 Elsevier Ltd. All rights reserved.
Please cite this article as: Shang J, et al, Zucker Diabetic Fatty rats exhibit hypercoagulability and accelerated thrombus formation in the ArterioVenous shunt model of thro..., Thromb Res (2014), http://dx.doi.org/10.1016/j.thromres.2014.04.008
2
J. Shang et al. / Thrombosis Research xxx (2014) xxx–xxx
and db/db mouse models of diabetes did not develop hypercoagulability and they exhibited less platelet activation and aggregation due to the lack of functional leptin signaling in platelets [8]. Diet-induced obese C57BL/6J mice did not demonstrate hypercoagulability or platelet hyperreactivity either [9]. These frequently used mouse models for studying metabolic diseases are perhaps suboptimal for thrombosis research. The ZDF rat model is a commonly used type 2 diabetic rodent model that carries a naturally occurring dysfunctional mutation in the Lepr gene as well as a genetic defect in β-cell transcription [10,11]. Unlike male ZDF rats, the female ZDF rats only develop hyperglycemia when fed a diabetogenic diet [12,13], which resembles the pathogenesis of type 2 diabetes typically resulted from the interaction of both genetic and acquired risk factors [14,15]. While platelet aggregation has been studied for the ZDF rat model [16], comprehensive characterization with regard to its hemostatic profile, platelet reactivity, or potential prothrombotic status has not been pursued. In this report, several aspects of the high-fat fed (HFF) female diabetic ZDF rat model were examined with the ultimate goal of assessing its potential value in thrombosis research. Basal hemostatic profile and platelet reactivity were profiled to evaluate whether it can serve as a translational model for studying chronic effects of anti-diabetic mechanisms on systemic markers of blood coagulation and thrombosis. A 14day study with one anti-diabetic agent, dapagliflozin, was conducted to initially explore such a treatment paradigm. The acute, provoked thrombotic process in the diabetic ZDF rats was also assessed in comparison with normal rats in the routinely utilized A-V shunt thrombosis and the oxidative injury (FeCl3-induced) arterial thrombosis models. The anticoagulant effect of a marketed antithrombotic drug, apixaban [17], was further investigated in the acute A-V shunt model in the ZDF rats in contrast with the normal rats. Materials and Methods Animals Female ZDF rats, age- and gender-matched genetic control rats Zucker Lean (ZL), and male Sprague–Dawley (SD) rats were purchased at the age of 7–10 weeks from the Charles River Laboratories (Wilmington, MA). The ZDF rats were fed a high-fat diet (Research Diets D10111701, 48% kcal from fat and 34% kcal from carbohydrate, Research Diets, New Brunswick, NJ) for 3–4 weeks. The ZL and SD rats were fed a chow diet (Research Diets D7012, Research Diets, New Brunswick, NJ) for the same period of time before studies were conducted. All animal care and experimental procedures were approved by the IACUC of Merck Research Laboratories. The Guide for the Care and Use of Laboratory Animals was followed in the conduct of the animal studies and veterinary care was given to any animals requiring medical attention. Blood glucose, plasma insulin, triglycerides, cholesterol, and TAT levels Rats were fasted overnight for approximately 18 hours. Blood glucose levels were measured by hand-held glucometers (Onetouch Ultra, LifeSpan, CA). Plasma insulin levels were measured using the Rat Insulin ELISA kit (Mercodia, Sweden). Plasma triglycerides and total cholesterol levels were measured using the Triglyceride kit (Roche, IN) and the Cholesterol E kit (Wako Chemicals, VA). Plasma levels for TAT (thrombin-antithrombin complex) were determined by the Enzygnost TAT ELISA kit (Siemens Healthcare Diagnostics, Deerfield, IL). Blood sample collection for the evaluation of clotting times, platelet aggregation, and coagulation parameters was conducted as follows. Rats were anesthetized in the non-fasting state by intraperitoneal administration of Inactin at 100 mg/kg body weight (Sigma-Aldrich, St. Louis, MO). A 4–5 cm midline abdominal incision was made to expose the abdominal aorta. Blood was drawn into 4.5 ml sodium citrate vacutainer tubes by puncturing the abdominal aorta with 21 G Vacutainer multiple sample needles (Becton, Dickinson and Company, Baltimore, MD).
Thrombin generation assay (TGA) TGA was performed using an automated fluorogenic measurement apparatus (Thrombinoscope) in a 96-well plate format (Thermo Labsystems, Helsinki, Finland) as previously described [18]. Each plasma sample had its own calibrator and was activated by different triggers (Stago, NJ) as indicated. The Thrombinoscope BV software (Maastricht, the Netherlands) was used to calculate thrombin generation. To each well, 60 μl of plasma and 15 μl of trigger reagent were added and mixed well. The plate was incubated in the machine for at least 5 minutes (min) at 37 °C. 15 μl of 16.4 mM CaCl2 and fluorescent substrate was then dispensed into each well by the instrument to initiate reaction. Thrombin generation was recorded by detecting fluorescent signals and five parameters were generated: lag time (the time for initial thrombin generation), thrombin peak (the maximal amount of thrombin generated), time to peak (the time to reach the thrombin peak), endogenous thrombin potential (ETP, the total amount of thrombin generated), and slope (the rate of thrombin generation). Thromboelastography (TEG) An aliquot of 340 μl of freshly collected sodium-citrated whole blood was recalcified with 20 μl of 0.2 M CaCl2 in a cuvette in the thromboelastograph coagulation analyzer (model 5000, Heamoscope Corp, Niles, IL). From the TEG output, three key parameters, R (reaction time), maximal amplitude (MA), and LY60 (% lysis at 60 min after MA) were utilized in representing clotting time, clot strength, and fibrinolytic activity, respectively [18]. Platelet P-selectin expression by flow cytometry Platelet P-selectin (CD62p) expression was measured by using a FACScalibur flow cytometer (Beckton Dickinson, San Jose, CA). In brief, 0.5 μl of citrated whole blood was added to 104 μl staining buffer composed of 100 μl FACS buffer (BD Biosciences, San Jose, CA) and either 4 μl of R-PE-conjugated mouse anti-rat CD62p/P-Selectin monoclonal antibody (SM 2262RT from Acris, Herford, Germany) or its isotype control (AM03095AF-N). With gentle mixing, the tubes were incubated at room temperature in the dark for 30 min. Subsequently, 2 ml of FACS buffer was added to the samples, which were finally fixed in 1% paraformaldehyde (BD Cytofix Fixation Buffer) in PBS and kept at 4 °C. Platelet P-selectin expression was measured by flow cytometer and data were analyzed with FlowJo V. 7.6.2 (Tree Star, Inc, Ashland, OR). Ex Vivo clotting times and Platelet Aggregation Blood samples were collected into sodium citrate Vacutainers (Becton Dickinson, Frankin lakes, NJ) from abdominal aorta of rats that were anesthetized with Inactin (100 mg/kg/ml, Sigma Aldrich, administered intraperitoneally). The blood samples were centrifuged at 2400 x g for 15 min at 4 °C for plasma preparation. The prothrombin time (PT) and activated partial thromboplastin time (aPTT), which assess the extrinsic and intrinsic coagulation pathways, respectively, were measured in the Coag-A-Mate instrument with reagents from the vendor (Trinity Biotech, Ireland). The ACT test measures the clotting time of whole blood samples in response to a trigger for the intrinsic pathway. It was performed in an automated coagulation timer (ACT II; Medtronic, MN) using the RACT cartridges. Platelet aggregation studies were performed ex vivo using citrated platelet-rich plasma (PRP) in an optical ChronoLog Aggregometer (Model 700; ChronoLog, Havertown, PA) as previously described [19]. Briefly, PRP was generated from citrated whole blood after centrifuging at 120 x g for 15 min. 250 μL PRP was incubated in a cuvette containing a stir bar for 2–3 min. Agonists for platelet aggregation in the study
Please cite this article as: Shang J, et al, Zucker Diabetic Fatty rats exhibit hypercoagulability and accelerated thrombus formation in the ArterioVenous shunt model of thro..., Thromb Res (2014), http://dx.doi.org/10.1016/j.thromres.2014.04.008
J. Shang et al. / Thrombosis Research xxx (2014) xxx–xxx
included PAR-4 peptide (2.5 mM, H-4348, Bachem Bioscience, PA), ADP (2.5 μM, ChronoLog, Havertown, PA), and collagen (10 μg/mL, ChronoLog, Havertown, PA). Platelet aggregation was monitored for 5–7 min after the addition of the agonists. Platelet-poor plasma was used as 100% transmittance for aggregation. The peak aggregation response was recorded in percentage of transmittance.
Dapagliflozin treatment Dapagliflozin (synthesized by Wuxi AppTec Inc) at 0.3 mg/kg or vehicle (0.4% methylcellulose) was administered via oral gavage to ZDF rats once daily for 14 days. Ambient blood glucose and body weight were measured every 2–3 days. Terminal blood was collected from anesthetized rats for various analyses as described in Results.
A-V shunt thrombosis model The model was developed based on previously described procedures [20]. Briefly, the right carotid artery and left jugular vein were cannulated with PE 100 tubing (Becton Dickinson, FRANKLIN LAKES, NJ). The right jugular vein was cannulated with PE-50 tubing for administration of compound or vehicle. After surgical procedures, an extracorporeal shunt was inserted into the ends of the right carotid artery and left jugular vein catheters. The shunt consisted of a 6-cm Tygon tubing (Thermo Fisher Scientific, Waltham, MA) that had been previously siliconized and dried. A 6-cm segment of silk thread (Size: 2–0, from Ethicon, Bridgewater, NJ) was secured inside shunt so that it remained longitudinally oriented during blood flow through the shunt. After initiating the circulation, the extracorporeal shunt remained in place for the indicated time. The shunt was then disconnected and the thread with its associated thrombus was removed from the shunt, blot dried, and weighed. Thrombus weight was calculated by subtracting the average weight of a 6-cm long silk thread. For measurement of the anti-thrombotic effect of apixaban in the A-V shunt model, apixaban (synthesized by Wuxi AppTec Inc) or vehicle (0.4% methylcellulose) was administered as a continuous i.v. infusion at 1 ml/kg/hour starting from 15 min before conducting a thrombosis procedure. At the end of the procedure with the shunt remaining in place for 15 min, blood was collected for measuring PT and aPTT.
FeCl3-induced arterial thrombosis model The model was established based on previously described procedures [21]. Briefly, in anesthetized rats, the left common carotid artery was surgically exposed and a Doppler flow probe (Model 1PRB; Transonic Systems, Ithaca, NY) was placed on the surface of the artery. Baseline blood flow was recorded using a Transonic Model T206 flow meter. A filter paper disk (3 mm in diameter) saturated with 10% FeCl3 was applied to the adventitial surface of the carotid artery, immediately proximal to the flow probe. Time to complete occlusion after initiation of artery injury was defined as the time required for blood flow to decline to 0.0 ml/min.
3
Results ZDF rats are in a hypercoagulant state The female ZDF rats on the high-fat diet were obese, hyperglycemic, hyperinsulinemic, and hypertriglyceridemic (Table 1). To determine their basal coagulability, we measured several coagulation parameters in ZDF rats and their genetic control, ZL rats. Compared to ZL rats, ZDF rats had significantly shortened PT (by 12%, p = 0.00002) and aPTT (by 28%, p = 0.004) (Table 1), suggesting both the extrinsic and intrinsic coagulation cascades are upregulated in the ZDF rats. In the thrombin generation assay (TGA), we observed significant differences in ZDF compared to ZL rats, including shortened lag time (by 27%), increased thrombin peak (by 169%), enhanced ETP (by 139%), and increased slope (by 264%) with 1 pM tissue factor as the trigger (Fig. 1A). Similar thrombin generation profiles were observed with higher tissue factor concentrations at 5 and 20 pM (Fig. 1B and C). The consistently increased thrombin generation at various tissue factor concentrations (Fig. 1D) suggests that the ZDF rats are in a hypercoagulant state. In thromboelastography, the ZDF rats' whole blood displayed increased clot strength (MA) and decreased fibrinolysis (LY60) compared to ZL rats’ whole blood (Table 1). The clotting times of whole blood samples measured by either thromboelastography or ACT did not show a significant difference between ZDF and ZL rats (Table 1). Plasma TAT levels had a numerical increase in ZDF rats without reaching statistical significance (Table 1). Taken together, the coagulation parameters using plasma samples indicate that ZDF rats are in a hypercoagulant state. Glycemic control does not normalize hypercoagulability in ZDF rats Hyperglycemia has been reported to contribute to accelerated arterial thrombus formation in animal models [22,23]. To address whether glycemic control can normalize hypercoagulability in ZDF rats, dapagliflozin (Dapa), a selective sodium-glucose co-transporter 2 (SGLT2) inhibitor that can rapidly and effectively lower blood glucose levels in the ZDF rat model [23,24] was utilized. Dapa improves glycemic control in diabetes by reducing renal glucose reabsorption [24]. When Dapa was orally administered at 0.3 mg/kg, blood glucose levels in ZDF rats were reduced from 300 to 150 mg/dl at 4 hours post the first dosing with the reduction sustained throughout the 14-day treatment period (Fig. 2A). There was no significant difference in body weight, plasma triglycerides or total cholesterol levels between the Dapatreated and vehicle-treated groups (Table 2). Coagulation parameters were examined at the end of the treatment, and no significant changes in the treatment group were observed in PT, aPTT, ACT, plasma TAT, or thromboelastography parameters (Table 2). The thrombin generation profiles of the Dapa-treated group were indistinguishable from the vehicle-treated group with or without various concentrations of tissue factor as the trigger (Fig. 2B). Platelet P-selectin levels did not exhibit appreciable change either (data not shown). Taken together, these results suggest that ZDF rats did not display reduced coagulability with glycemic normalization achieved with the SGLT2 mechanism for up to 2 weeks. ZDF rats exhibited accelerated thrombus formation in the A-V shunt model
Statistical analysis All data are presented as mean ± standard deviation (SD). Statistical analysis was conducted using either one-way ANOVA followed by Bonferroni post hoc test in GraphPad Prism (Version 4.0.3, Graphpad, La Jolla, CA) or Student’s t-test, as appropriate. For thrombus weight study comparing the ZDF, ZL and SD rats, the non-parametric KruskalWallis test (with Dunn’s post test) was applied. Statistical significance was defined as two-tailed p b 0.05.
To examine whether hypercoagulability results in increased thrombogenicity in ZDF rats, thrombus formation was measured in the A-V shunt model when blood flowed in the shunt for 5 min. The mean thrombus weight for the ZDF rats was 37 ± 4 mg, which was significantly greater than that of either ZL rats (12 ± 1 mg) or SD rats (15 ± 2 mg) (Fig. 3). There was no difference in this measurement between the ZL rats and SD rats. ZDF rats thus have accelerated thrombus formation compared to normal control rats in the A-V shunt model of thrombosis.
Please cite this article as: Shang J, et al, Zucker Diabetic Fatty rats exhibit hypercoagulability and accelerated thrombus formation in the ArterioVenous shunt model of thro..., Thromb Res (2014), http://dx.doi.org/10.1016/j.thromres.2014.04.008
4
J. Shang et al. / Thrombosis Research xxx (2014) xxx–xxx
Table 1 Basal State Characteristics of ZDF and ZL Rats.
Metabolic Characteristics
Coagulation Parameters
Platelet Parameters and ex vivo Platelet Responses
Strain
ZDF
Body Weight (g) Blood Glucose (mg/dl) Plasma Insulin (ng/ml) Plasma Triglycerides (mg/dl) Plasma Total Cholesterol (mg/dl) PT (s) aPTT (s) ACT (s) Plasma TAT (ng/ml) TEG R (min) TEG MA (mm) TEG LY60 (%) Platelet P-Selectin (MFI) Platelet Count (103/ml) Mean Platelet Volume (fL) ADP (2.5 µM) Collagen (10 µ/ml) PAR-4 TRAP (2.5 mM)
331 296 15.38 245 44 14.9 13.6 85 17.9 2.84 75.6 0.24 25.3 1219 6.6 34.8 48.0 71.2
ZL ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
19 *** 63 *** 10.35 * 12 *** 5 0.5 *** 2.5 ** 2 6.9 0.48 1.1 ** 0.23 * 4.5 ** 129 0.5 10.5 ** 7.0 * 4.5 **
172 91 0.33 29 45 16.9 19.0 84 11.4 3.26 72.4 1.04 17.7 1146 6.6 55.8 65.3 86.7
± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
3 5 0.07 6 6 0.5 3.2 2 3.3 0.37 1.3 0.54 1.5 42 0.4 6.5 5.6 7.2
Data are presented as mean ± SD (n = 5-6). Significantly different from ZL rats: * p b 0.05, ** p b 0.01 *** p b 0.001. ex vivo Responses: the peak response of platelet aggregation in terms of % transmittance.
agents [20] and they exhibited similar thrombus formation in the A-V shunt model as the ZL rats (Fig. 3), we compared the dose response to apixaban in ZDF rats and SD rats. Apixaban effectively and dosedependently reduced thrombus weight in the ZDF rats, albeit to less extent than in the SD rats (Fig. 4A). Apixaban dose-dependently prolonged PT in both SD and ZDF rats with the latter being at a smaller magnitude (Fig. 4B). Apixaban did not prolong aPTT for either group at the time
ZDF rats are amenable to apixaban treatment in the A-V shunt model with less sensitivity than normal rats Apixaban is among the class of new oral anticoagulants and is a highly selective and potent inhibitor for coagulation factor Xa (FXa) [20]. We evaluated the antithrombotic effect of apixaban in the A-V shunt model in ZDF rats. As SD rats are typically used to evaluate antithrombotic
A
B
TGA with 1 pM Tissue Factor 325
ZDF ZL
ZDF ZL
275
Thrombin (nM)
275
Thrombin (nM)
TGA with 5 pM Tissue Factor 325
225 175 125 75
225 175 125 75 25
25
-25
-25 0
2
4
6
8
10
12
14
16
18
20
0
2
4
Time (min)
C
8
10
12
14
16
18
20
Time (min)
D
TGA with 20 pM Tissue Factor
TGA Peak
***
250
Peak thrombin (nM)
ZDF ZL
200 150 100 50
***
***
300
300
Thrombin (nM)
6
ZDF ZL
250 200 150 100 50
0 0 0
2
4
6
8
10
12
Time (min)
14
16
18
20
ZDF HTF
ZL
ZDF MTF
ZL
ZDF
ZL
LTF
Fig. 1. Enhanced thrombin generation in ZDF rats. Thrombin generation assay was performed with plasma from ZDF and ZL rats in the presence of 1, 5 or 20 pM tissue factor as trigger (A, B, C). The peak thrombin values under the three conditions are graphed in D. *** p b 0.001 ZDF versus ZL. N = 5-6 per group.
Please cite this article as: Shang J, et al, Zucker Diabetic Fatty rats exhibit hypercoagulability and accelerated thrombus formation in the ArterioVenous shunt model of thro..., Thromb Res (2014), http://dx.doi.org/10.1016/j.thromres.2014.04.008
Thrombus weight (mg)/ 5 min
J. Shang et al. / Thrombosis Research xxx (2014) xxx–xxx
A 400
Glucose (mg/dL)
350 300 250
ZDF+Veh ZDF+Dapa
200 150
**
100
**
*
*
*
*
50 0 0
2
6
4
8
10
12
p
Contents lists available at ScienceDirect
Thrombosis Research journal homepage: www.elsevier.com/locate/thromres
Regular Article
Zucker Diabetic Fatty rats exhibit hypercoagulability and accelerated thrombus formation in the Arterio-Venous shunt model of thrombosis Jin Shang ⁎, Zhu Chen ⁎⁎, Min Wang, Qiu Li, Wen Feng, Yangsong Wu, Weizhen Wu, Michael P. Graziano, Madhu Chintala Department of Cardiometabolic Disease, Merck Research Laboratories, Kenilworth, New Jersey
a r t i c l e
i n f o
Article history: Received 10 January 2014 Received in revised form 6 April 2014 Accepted 8 April 2014 Available online xxxx Keywords: Rat Thrombin generation Coagulation Platelet reactivity Type 2 Diabetes Thrombus formation
a b s t r a c t Introduction: Diabetes is a significant risk factor for thrombosis. The present study aimed at assessing coagulability, platelet reactivity, and thrombogenicity of the diabetic female Zucker Diabetic Fatty (ZDF) rat model and its relevance in studying antithrombotic mechanisms. Materials and Methods: The basal coagulant state in ZDF rats was evaluated by clotting times, thromboelastography, and thrombin generation assay. A 14-day treatment with dapagliflozin in ZDF rats was pursued to investigate if glycemic control can improve coagulability. Thrombus formation in the Arterio-Venous (A-V) shunt model and the FeCl3-induced arterial thrombosis model was studied, with the antithrombotic effect of apixaban in the former model further investigated. Results: ZDF rats exhibited significantly shortened clotting times, enhanced thrombin generation, and decreased fibrinolysis at baseline. Effective glycemic control achieved with dapagliflozin did not improve any of these parameters. ZDF rats displayed accelerated thrombus formation and were amenable to apixaban treatment in the A-V shunt model albeit with less sensitivity than normal rats. ZDF rats exhibited less platelet aggregation in response to ADP, collagen and PAR-4, and attenuated thrombotic response in the FeCl3 model. Conclusions: ZDF rats are at a chronic hypercoagulable and hypofibrinolytic state yet with compromised platelet reactivity. They display accelerated and attenuated thrombosis in the A-V shunt and FeCl3 model of thrombosis, respectively. Results shed new light on the pathophysiology of the ZDF rat model and illustrate its potential value in translational research on anticoagulant agents in diabetics. Caution needs to be exerted in utilizing this model in assessing antiplatelet mechanisms in diabetes-associated atherothrombosis. © 2014 Elsevier Ltd. All rights reserved.
Introduction Thrombosis, both venous and arterial, is a major cause of morbidity and mortality worldwide [1]. Diabetes is one important risk factor for increased thrombotic events in patients with cardiovascular diseases [2]. Metabolic changes, including hyperinsulinemia, hypertriglyceridemia, altered adipokines and low-grade systemic inflammation, have been identified as contributors to platelet hyperreactivity in diabetic
Abbreviations: aPTT, activated partial thromboplastin time; A-V, Arterio-Venous; Dapa, dapagliflozin; ETP, endogenous thrombin potential; FXa, coagulation factor Xa; MI, myocardial infarction; Min, minute; PRP, platelet-rich plasma; PT, prothrombin time; s, second; SD, Sprague–Dawley; TEG, Thromboelastography; TGA, Thrombin generation assay; VTE, venous thromboembolism; ZDF, Zucker Diabetic Fatty; ZL, Zucker Lean. ⁎ Correspondence to: J. Shang, Cardiometabolic Disease, Merck Research Laboratories K15-3C-309, Galloping Hill Road, Kenilworth, NJ 07033, USA. Tel.: +1 908 740 0131. ⁎⁎ Correspondence to: Z. Chen, Cardiometabolic Disease, Merck Research Laboratories K15-3A-309, Galloping Hill Road, Kenilworth, NJ 07033, USA. Tel.: +1 908 740 0642. E-mail addresses: [email protected] (J. Shang), [email protected] (Z. Chen).
patients [3]. Moreover, diabetes is associated with a hypercoagulant and hypofibrinolytic state, in which altered coagulation factors and endothelial dysfunction have been found [3]. It is also important to recognize that despite the current glucose lowering or antithrombotic therapies, diabetic patients remain at high risk for cardiovascular events [4,5]. Development of novel therapies that target the residual prothrombotic risk in diabetics is thus an unmet medical need. In preclinical studies, the efficacy of novel antithrombotic agents has been typically evaluated in normal and healthy animals in provoked models for thrombosis. In contrast, disease models that possess metabolic characteristics similar to diabetic patients may allow translational research in studying and evaluating novel antithrombotic agents that could be targeted to high-risk, diabetic patients. The ob/ob and db/db models are widely used for diabetes research. Mice homozygous for the obese spontaneous mutation (Lepob, leptin deficiency) are commonly referred to as ob/ob [6], and mice carrying the diabetes spontaneous mutation (Leprdb, a point mutation in the leptin receptor (Lepr) gene) are referred to as db/db [7]. It has been reported that the ob/ob
http://dx.doi.org/10.1016/j.thromres.2014.04.008 0049-3848/© 2014 Elsevier Ltd. All rights reserved.
Please cite this article as: Shang J, et al, Zucker Diabetic Fatty rats exhibit hypercoagulability and accelerated thrombus formation in the ArterioVenous shunt model of thro..., Thromb Res (2014), http://dx.doi.org/10.1016/j.thromres.2014.04.008
2
J. Shang et al. / Thrombosis Research xxx (2014) xxx–xxx
and db/db mouse models of diabetes did not develop hypercoagulability and they exhibited less platelet activation and aggregation due to the lack of functional leptin signaling in platelets [8]. Diet-induced obese C57BL/6J mice did not demonstrate hypercoagulability or platelet hyperreactivity either [9]. These frequently used mouse models for studying metabolic diseases are perhaps suboptimal for thrombosis research. The ZDF rat model is a commonly used type 2 diabetic rodent model that carries a naturally occurring dysfunctional mutation in the Lepr gene as well as a genetic defect in β-cell transcription [10,11]. Unlike male ZDF rats, the female ZDF rats only develop hyperglycemia when fed a diabetogenic diet [12,13], which resembles the pathogenesis of type 2 diabetes typically resulted from the interaction of both genetic and acquired risk factors [14,15]. While platelet aggregation has been studied for the ZDF rat model [16], comprehensive characterization with regard to its hemostatic profile, platelet reactivity, or potential prothrombotic status has not been pursued. In this report, several aspects of the high-fat fed (HFF) female diabetic ZDF rat model were examined with the ultimate goal of assessing its potential value in thrombosis research. Basal hemostatic profile and platelet reactivity were profiled to evaluate whether it can serve as a translational model for studying chronic effects of anti-diabetic mechanisms on systemic markers of blood coagulation and thrombosis. A 14day study with one anti-diabetic agent, dapagliflozin, was conducted to initially explore such a treatment paradigm. The acute, provoked thrombotic process in the diabetic ZDF rats was also assessed in comparison with normal rats in the routinely utilized A-V shunt thrombosis and the oxidative injury (FeCl3-induced) arterial thrombosis models. The anticoagulant effect of a marketed antithrombotic drug, apixaban [17], was further investigated in the acute A-V shunt model in the ZDF rats in contrast with the normal rats. Materials and Methods Animals Female ZDF rats, age- and gender-matched genetic control rats Zucker Lean (ZL), and male Sprague–Dawley (SD) rats were purchased at the age of 7–10 weeks from the Charles River Laboratories (Wilmington, MA). The ZDF rats were fed a high-fat diet (Research Diets D10111701, 48% kcal from fat and 34% kcal from carbohydrate, Research Diets, New Brunswick, NJ) for 3–4 weeks. The ZL and SD rats were fed a chow diet (Research Diets D7012, Research Diets, New Brunswick, NJ) for the same period of time before studies were conducted. All animal care and experimental procedures were approved by the IACUC of Merck Research Laboratories. The Guide for the Care and Use of Laboratory Animals was followed in the conduct of the animal studies and veterinary care was given to any animals requiring medical attention. Blood glucose, plasma insulin, triglycerides, cholesterol, and TAT levels Rats were fasted overnight for approximately 18 hours. Blood glucose levels were measured by hand-held glucometers (Onetouch Ultra, LifeSpan, CA). Plasma insulin levels were measured using the Rat Insulin ELISA kit (Mercodia, Sweden). Plasma triglycerides and total cholesterol levels were measured using the Triglyceride kit (Roche, IN) and the Cholesterol E kit (Wako Chemicals, VA). Plasma levels for TAT (thrombin-antithrombin complex) were determined by the Enzygnost TAT ELISA kit (Siemens Healthcare Diagnostics, Deerfield, IL). Blood sample collection for the evaluation of clotting times, platelet aggregation, and coagulation parameters was conducted as follows. Rats were anesthetized in the non-fasting state by intraperitoneal administration of Inactin at 100 mg/kg body weight (Sigma-Aldrich, St. Louis, MO). A 4–5 cm midline abdominal incision was made to expose the abdominal aorta. Blood was drawn into 4.5 ml sodium citrate vacutainer tubes by puncturing the abdominal aorta with 21 G Vacutainer multiple sample needles (Becton, Dickinson and Company, Baltimore, MD).
Thrombin generation assay (TGA) TGA was performed using an automated fluorogenic measurement apparatus (Thrombinoscope) in a 96-well plate format (Thermo Labsystems, Helsinki, Finland) as previously described [18]. Each plasma sample had its own calibrator and was activated by different triggers (Stago, NJ) as indicated. The Thrombinoscope BV software (Maastricht, the Netherlands) was used to calculate thrombin generation. To each well, 60 μl of plasma and 15 μl of trigger reagent were added and mixed well. The plate was incubated in the machine for at least 5 minutes (min) at 37 °C. 15 μl of 16.4 mM CaCl2 and fluorescent substrate was then dispensed into each well by the instrument to initiate reaction. Thrombin generation was recorded by detecting fluorescent signals and five parameters were generated: lag time (the time for initial thrombin generation), thrombin peak (the maximal amount of thrombin generated), time to peak (the time to reach the thrombin peak), endogenous thrombin potential (ETP, the total amount of thrombin generated), and slope (the rate of thrombin generation). Thromboelastography (TEG) An aliquot of 340 μl of freshly collected sodium-citrated whole blood was recalcified with 20 μl of 0.2 M CaCl2 in a cuvette in the thromboelastograph coagulation analyzer (model 5000, Heamoscope Corp, Niles, IL). From the TEG output, three key parameters, R (reaction time), maximal amplitude (MA), and LY60 (% lysis at 60 min after MA) were utilized in representing clotting time, clot strength, and fibrinolytic activity, respectively [18]. Platelet P-selectin expression by flow cytometry Platelet P-selectin (CD62p) expression was measured by using a FACScalibur flow cytometer (Beckton Dickinson, San Jose, CA). In brief, 0.5 μl of citrated whole blood was added to 104 μl staining buffer composed of 100 μl FACS buffer (BD Biosciences, San Jose, CA) and either 4 μl of R-PE-conjugated mouse anti-rat CD62p/P-Selectin monoclonal antibody (SM 2262RT from Acris, Herford, Germany) or its isotype control (AM03095AF-N). With gentle mixing, the tubes were incubated at room temperature in the dark for 30 min. Subsequently, 2 ml of FACS buffer was added to the samples, which were finally fixed in 1% paraformaldehyde (BD Cytofix Fixation Buffer) in PBS and kept at 4 °C. Platelet P-selectin expression was measured by flow cytometer and data were analyzed with FlowJo V. 7.6.2 (Tree Star, Inc, Ashland, OR). Ex Vivo clotting times and Platelet Aggregation Blood samples were collected into sodium citrate Vacutainers (Becton Dickinson, Frankin lakes, NJ) from abdominal aorta of rats that were anesthetized with Inactin (100 mg/kg/ml, Sigma Aldrich, administered intraperitoneally). The blood samples were centrifuged at 2400 x g for 15 min at 4 °C for plasma preparation. The prothrombin time (PT) and activated partial thromboplastin time (aPTT), which assess the extrinsic and intrinsic coagulation pathways, respectively, were measured in the Coag-A-Mate instrument with reagents from the vendor (Trinity Biotech, Ireland). The ACT test measures the clotting time of whole blood samples in response to a trigger for the intrinsic pathway. It was performed in an automated coagulation timer (ACT II; Medtronic, MN) using the RACT cartridges. Platelet aggregation studies were performed ex vivo using citrated platelet-rich plasma (PRP) in an optical ChronoLog Aggregometer (Model 700; ChronoLog, Havertown, PA) as previously described [19]. Briefly, PRP was generated from citrated whole blood after centrifuging at 120 x g for 15 min. 250 μL PRP was incubated in a cuvette containing a stir bar for 2–3 min. Agonists for platelet aggregation in the study
Please cite this article as: Shang J, et al, Zucker Diabetic Fatty rats exhibit hypercoagulability and accelerated thrombus formation in the ArterioVenous shunt model of thro..., Thromb Res (2014), http://dx.doi.org/10.1016/j.thromres.2014.04.008
J. Shang et al. / Thrombosis Research xxx (2014) xxx–xxx
included PAR-4 peptide (2.5 mM, H-4348, Bachem Bioscience, PA), ADP (2.5 μM, ChronoLog, Havertown, PA), and collagen (10 μg/mL, ChronoLog, Havertown, PA). Platelet aggregation was monitored for 5–7 min after the addition of the agonists. Platelet-poor plasma was used as 100% transmittance for aggregation. The peak aggregation response was recorded in percentage of transmittance.
Dapagliflozin treatment Dapagliflozin (synthesized by Wuxi AppTec Inc) at 0.3 mg/kg or vehicle (0.4% methylcellulose) was administered via oral gavage to ZDF rats once daily for 14 days. Ambient blood glucose and body weight were measured every 2–3 days. Terminal blood was collected from anesthetized rats for various analyses as described in Results.
A-V shunt thrombosis model The model was developed based on previously described procedures [20]. Briefly, the right carotid artery and left jugular vein were cannulated with PE 100 tubing (Becton Dickinson, FRANKLIN LAKES, NJ). The right jugular vein was cannulated with PE-50 tubing for administration of compound or vehicle. After surgical procedures, an extracorporeal shunt was inserted into the ends of the right carotid artery and left jugular vein catheters. The shunt consisted of a 6-cm Tygon tubing (Thermo Fisher Scientific, Waltham, MA) that had been previously siliconized and dried. A 6-cm segment of silk thread (Size: 2–0, from Ethicon, Bridgewater, NJ) was secured inside shunt so that it remained longitudinally oriented during blood flow through the shunt. After initiating the circulation, the extracorporeal shunt remained in place for the indicated time. The shunt was then disconnected and the thread with its associated thrombus was removed from the shunt, blot dried, and weighed. Thrombus weight was calculated by subtracting the average weight of a 6-cm long silk thread. For measurement of the anti-thrombotic effect of apixaban in the A-V shunt model, apixaban (synthesized by Wuxi AppTec Inc) or vehicle (0.4% methylcellulose) was administered as a continuous i.v. infusion at 1 ml/kg/hour starting from 15 min before conducting a thrombosis procedure. At the end of the procedure with the shunt remaining in place for 15 min, blood was collected for measuring PT and aPTT.
FeCl3-induced arterial thrombosis model The model was established based on previously described procedures [21]. Briefly, in anesthetized rats, the left common carotid artery was surgically exposed and a Doppler flow probe (Model 1PRB; Transonic Systems, Ithaca, NY) was placed on the surface of the artery. Baseline blood flow was recorded using a Transonic Model T206 flow meter. A filter paper disk (3 mm in diameter) saturated with 10% FeCl3 was applied to the adventitial surface of the carotid artery, immediately proximal to the flow probe. Time to complete occlusion after initiation of artery injury was defined as the time required for blood flow to decline to 0.0 ml/min.
3
Results ZDF rats are in a hypercoagulant state The female ZDF rats on the high-fat diet were obese, hyperglycemic, hyperinsulinemic, and hypertriglyceridemic (Table 1). To determine their basal coagulability, we measured several coagulation parameters in ZDF rats and their genetic control, ZL rats. Compared to ZL rats, ZDF rats had significantly shortened PT (by 12%, p = 0.00002) and aPTT (by 28%, p = 0.004) (Table 1), suggesting both the extrinsic and intrinsic coagulation cascades are upregulated in the ZDF rats. In the thrombin generation assay (TGA), we observed significant differences in ZDF compared to ZL rats, including shortened lag time (by 27%), increased thrombin peak (by 169%), enhanced ETP (by 139%), and increased slope (by 264%) with 1 pM tissue factor as the trigger (Fig. 1A). Similar thrombin generation profiles were observed with higher tissue factor concentrations at 5 and 20 pM (Fig. 1B and C). The consistently increased thrombin generation at various tissue factor concentrations (Fig. 1D) suggests that the ZDF rats are in a hypercoagulant state. In thromboelastography, the ZDF rats' whole blood displayed increased clot strength (MA) and decreased fibrinolysis (LY60) compared to ZL rats’ whole blood (Table 1). The clotting times of whole blood samples measured by either thromboelastography or ACT did not show a significant difference between ZDF and ZL rats (Table 1). Plasma TAT levels had a numerical increase in ZDF rats without reaching statistical significance (Table 1). Taken together, the coagulation parameters using plasma samples indicate that ZDF rats are in a hypercoagulant state. Glycemic control does not normalize hypercoagulability in ZDF rats Hyperglycemia has been reported to contribute to accelerated arterial thrombus formation in animal models [22,23]. To address whether glycemic control can normalize hypercoagulability in ZDF rats, dapagliflozin (Dapa), a selective sodium-glucose co-transporter 2 (SGLT2) inhibitor that can rapidly and effectively lower blood glucose levels in the ZDF rat model [23,24] was utilized. Dapa improves glycemic control in diabetes by reducing renal glucose reabsorption [24]. When Dapa was orally administered at 0.3 mg/kg, blood glucose levels in ZDF rats were reduced from 300 to 150 mg/dl at 4 hours post the first dosing with the reduction sustained throughout the 14-day treatment period (Fig. 2A). There was no significant difference in body weight, plasma triglycerides or total cholesterol levels between the Dapatreated and vehicle-treated groups (Table 2). Coagulation parameters were examined at the end of the treatment, and no significant changes in the treatment group were observed in PT, aPTT, ACT, plasma TAT, or thromboelastography parameters (Table 2). The thrombin generation profiles of the Dapa-treated group were indistinguishable from the vehicle-treated group with or without various concentrations of tissue factor as the trigger (Fig. 2B). Platelet P-selectin levels did not exhibit appreciable change either (data not shown). Taken together, these results suggest that ZDF rats did not display reduced coagulability with glycemic normalization achieved with the SGLT2 mechanism for up to 2 weeks. ZDF rats exhibited accelerated thrombus formation in the A-V shunt model
Statistical analysis All data are presented as mean ± standard deviation (SD). Statistical analysis was conducted using either one-way ANOVA followed by Bonferroni post hoc test in GraphPad Prism (Version 4.0.3, Graphpad, La Jolla, CA) or Student’s t-test, as appropriate. For thrombus weight study comparing the ZDF, ZL and SD rats, the non-parametric KruskalWallis test (with Dunn’s post test) was applied. Statistical significance was defined as two-tailed p b 0.05.
To examine whether hypercoagulability results in increased thrombogenicity in ZDF rats, thrombus formation was measured in the A-V shunt model when blood flowed in the shunt for 5 min. The mean thrombus weight for the ZDF rats was 37 ± 4 mg, which was significantly greater than that of either ZL rats (12 ± 1 mg) or SD rats (15 ± 2 mg) (Fig. 3). There was no difference in this measurement between the ZL rats and SD rats. ZDF rats thus have accelerated thrombus formation compared to normal control rats in the A-V shunt model of thrombosis.
Please cite this article as: Shang J, et al, Zucker Diabetic Fatty rats exhibit hypercoagulability and accelerated thrombus formation in the ArterioVenous shunt model of thro..., Thromb Res (2014), http://dx.doi.org/10.1016/j.thromres.2014.04.008
4
J. Shang et al. / Thrombosis Research xxx (2014) xxx–xxx
Table 1 Basal State Characteristics of ZDF and ZL Rats.
Metabolic Characteristics
Coagulation Parameters
Platelet Parameters and ex vivo Platelet Responses
Strain
ZDF
Body Weight (g) Blood Glucose (mg/dl) Plasma Insulin (ng/ml) Plasma Triglycerides (mg/dl) Plasma Total Cholesterol (mg/dl) PT (s) aPTT (s) ACT (s) Plasma TAT (ng/ml) TEG R (min) TEG MA (mm) TEG LY60 (%) Platelet P-Selectin (MFI) Platelet Count (103/ml) Mean Platelet Volume (fL) ADP (2.5 µM) Collagen (10 µ/ml) PAR-4 TRAP (2.5 mM)
331 296 15.38 245 44 14.9 13.6 85 17.9 2.84 75.6 0.24 25.3 1219 6.6 34.8 48.0 71.2
ZL ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
19 *** 63 *** 10.35 * 12 *** 5 0.5 *** 2.5 ** 2 6.9 0.48 1.1 ** 0.23 * 4.5 ** 129 0.5 10.5 ** 7.0 * 4.5 **
172 91 0.33 29 45 16.9 19.0 84 11.4 3.26 72.4 1.04 17.7 1146 6.6 55.8 65.3 86.7
± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
3 5 0.07 6 6 0.5 3.2 2 3.3 0.37 1.3 0.54 1.5 42 0.4 6.5 5.6 7.2
Data are presented as mean ± SD (n = 5-6). Significantly different from ZL rats: * p b 0.05, ** p b 0.01 *** p b 0.001. ex vivo Responses: the peak response of platelet aggregation in terms of % transmittance.
agents [20] and they exhibited similar thrombus formation in the A-V shunt model as the ZL rats (Fig. 3), we compared the dose response to apixaban in ZDF rats and SD rats. Apixaban effectively and dosedependently reduced thrombus weight in the ZDF rats, albeit to less extent than in the SD rats (Fig. 4A). Apixaban dose-dependently prolonged PT in both SD and ZDF rats with the latter being at a smaller magnitude (Fig. 4B). Apixaban did not prolong aPTT for either group at the time
ZDF rats are amenable to apixaban treatment in the A-V shunt model with less sensitivity than normal rats Apixaban is among the class of new oral anticoagulants and is a highly selective and potent inhibitor for coagulation factor Xa (FXa) [20]. We evaluated the antithrombotic effect of apixaban in the A-V shunt model in ZDF rats. As SD rats are typically used to evaluate antithrombotic
A
B
TGA with 1 pM Tissue Factor 325
ZDF ZL
ZDF ZL
275
Thrombin (nM)
275
Thrombin (nM)
TGA with 5 pM Tissue Factor 325
225 175 125 75
225 175 125 75 25
25
-25
-25 0
2
4
6
8
10
12
14
16
18
20
0
2
4
Time (min)
C
8
10
12
14
16
18
20
Time (min)
D
TGA with 20 pM Tissue Factor
TGA Peak
***
250
Peak thrombin (nM)
ZDF ZL
200 150 100 50
***
***
300
300
Thrombin (nM)
6
ZDF ZL
250 200 150 100 50
0 0 0
2
4
6
8
10
12
Time (min)
14
16
18
20
ZDF HTF
ZL
ZDF MTF
ZL
ZDF
ZL
LTF
Fig. 1. Enhanced thrombin generation in ZDF rats. Thrombin generation assay was performed with plasma from ZDF and ZL rats in the presence of 1, 5 or 20 pM tissue factor as trigger (A, B, C). The peak thrombin values under the three conditions are graphed in D. *** p b 0.001 ZDF versus ZL. N = 5-6 per group.
Please cite this article as: Shang J, et al, Zucker Diabetic Fatty rats exhibit hypercoagulability and accelerated thrombus formation in the ArterioVenous shunt model of thro..., Thromb Res (2014), http://dx.doi.org/10.1016/j.thromres.2014.04.008
Thrombus weight (mg)/ 5 min
J. Shang et al. / Thrombosis Research xxx (2014) xxx–xxx
A 400
Glucose (mg/dL)
350 300 250
ZDF+Veh ZDF+Dapa
200 150
**
100
**
*
*
*
*
50 0 0
2
6
4
8
10
12
p
