Drugs and Reagents.

JNJ-9375, control human IgG4 (CNTO 3161), and buffer were obtained internally (Janssen Biotherapeutics, Malvern, PA). Apixaban was obtained from Selleckchem (Houston, TX). Human α-thrombin and prothrombin were obtained from either Haematologic Technologies (Essex Junction, VT) or Enzyme Research Laboratories (South Bend, IN). Mouse and rat thrombin were obtained from Enzyme Research Laboratories. Cynomolgus thrombin, meizothrombin variant R155Q/R271Q/R284Q (referred to as meizothrombin QQQ), and meizothrombin variant R155Q/R271Q/R284Q/S195A (referred to as meizothrombin QQQA) were recombinantly expressed and obtained from Cambridge Protein Works (Cambridge, UK). D-Phenylalanyl-prolyl-arginyl chloromethyl ketone–inactivated human α-thrombin was prepared internally. Coagulation reagents were from Diagnostica Stago (Parsippany, NJ): Thrombin 10 for TT, ecarin for ecarin clotting time (ECT), Neoplastine Cl Plus 5 for PT, and C.K. Prest 5 and CaCl₂ for aPTT. For the platelet aggregation analysis, ADP, collagen, and arachidonic acid were obtained from Chronolog (Havertown, PA); thrombin receptor agonist peptide-6 (TRAP-6) and H-Gly-Pro-Arg-Pro-OH-AcOH were obtained from Bachem (King of Prussia, PA).

Biacore Experiment for Substrate Affinity Determination.

Surface plasmon resonance analysis was performed using a BiacoreT200 instrument [GE healthcare, Uppsala, Sweden (previously Biacore AB)]. The experiments were performed in PBS buffer (pH 7.4, with 3 mM EDTA, 0.005% Tween 20, and 400 µg/ml bovine serum albumin). Goat anti-human Fcγ antibody (Jackson ImmunoResearch, West Grove, PA) was covalently immobilized onto a CM5 sensor chip per the manufacturer’s instructions. Approximately 13,000 response units of anti-human Fcγ antibody were covalently immobilized in a reaction and reference flow cell. The remaining reactive groups on the surface were deactivated by injection of ethanolamine-HCl. JNJ-9375 or CNTO 3161 were captured onto the anti-Fcγ–modified sensor chip (120–140 response units). The antibody capture was followed by injection of serially diluted possible binding substrates (human, cynomolgus monkey, mouse, rat thrombin, prothrombin, or meizothrombin) at 40 µl/min. To perform the kinetic experiments, the antigens were prepared in the PBS-based buffer described previously, following a 2-fold serial dilution to obtain a concentration range of 0.039–20 nM, and the interactions were monitored at 25°C. Regeneration of the anti-Fc surface after each cycle was achieved using 10 mM glycine (pH 2.0). The data collected were processed using the T200 evaluation software (version 2) and fitted to a Langmuir 1:1 binding model, with the results reported in the format of k_on (on rate), k_off (off rate), and K_D (equilibrium dissociation constant).

Coagulation Assays.

Assays were run on citrated plasma from the blood of healthy volunteers obtained via antecubital vein. All of the blood donors were required to be at least 2 weeks free of any medications having anticoagulant or antiplatelet effects, which included nonsteroidal anti-inflammatory drugs, warfarin, dabigatran, apixaban, and rivaroxaban. Rat blood was collected from the abdominal aorta and mouse blood was collected from the posterior vena cava. Male Sprague-Dawley rats (350–450 g) were purchased from Charles River Laboratories (Frederick, MD), and male C57 black mice (20–30 g) were purchased from Taconic Biosciences (Hudson, NY). Blood was mixed with 3.2% sodium to make a 1:10 dilution and centrifuged at 2500g for 15 minutes at 4°C. The resulting platelet poor plasma (PPP) was immediately stored at −80°C until use. Cynomolgus monkey and rabbit PPP were purchased from BioreclamationIVT (Westbury, NY).

Standard clotting assays including TT, ECT, PT, and aPTT were performed in plasma samples spiked with JNJ-9735 as described previously (Douxfils et al., 2012), or from ex vivo plasma samples obtained from rats that were subjected to the AV shunt model of thrombosis. Coagulation assays were performed on a four-channel coagulometer STA rt4 (Diagnotica Stago). Briefly, for the TT assay, 100 µl PPP containing different concentrations of JNJ-9375 was incubated for 1 minute at 37°C in a cuvette and coagulation was initiated by adding 100 µl thrombin [1.5 U/ml (National Institutes of Health units)]. The onset of clotting was recorded as the coagulation time. For the ECT assay, 100 µl PPP containing different concentrations of JNJ-9375 was incubated for 4 minutes at 37°C in a cuvette and coagulation was initiated by adding 50 µl ecarin (5 U/ml, reconstituted in H₂O). For the PT assay, 50 µl PPP containing different concentrations of JNJ-9375 was incubated for 1 minute at 37°C in a cuvette and coagulation was initiated by adding 100 µl Neoplastine Cl reagent. For the aPTT assay, 50 µl C.K. Prest reagent was incubated with 50 µl PPP containing different concentrations of JNJ-9375 for 3 minutes at 37°C in a cuvette and coagulation was initiated by adding 50 µl 25 mM CaCl₂. The ratio change in clotting time induced by JNJ-9375 was calculated as the fold increase over the baseline value from the vehicle control group.

Platelet Aggregation.

Citrated blood, as in the clotting test, was processed to yield platelet rich plasma (PRP) by centrifuging the blood samples at 160g for 10–12 minutes at room temperature. The resultant PRP was transferred to 15 ml falcon tubes. The remaining samples were then further centrifuged at 2500g for 15 minutes at room temperature to collect PPP. The PRP was adjusted to a platelet count of 300 × 10³/µl with addition of the PPP from the donor. The PRP was kept at room temperature until use and used within 3 hours of preparation. Rat PRP and PPP were prepared using the same protocol except that the blood was drawn from the abdominal aorta and the final PRP platelet count was adjusted to 600 × 10³/µl (Derian et al., 1995).

Platelet aggregation was measured using the 700 Whole Blood/Optical Lumi-Aggregometer from the Chrono-log Corporation. Platelet aggregation was determined by measuring the optical density change of stirred PRP after adding the aggregating agonist (Zhou and Schmaier, 2005; Hawes et al., 2015). Briefly, 250 µl of PRP was added to a cuvette with a stir bar and prewarmed at 37°C on the aggregometer sample storage area for at least 5 minutes. JNJ-9375 was added to the cuvette for pretreatment and incubated for 5 minutes before adding agonist. When thrombin was used as the agonist, PRP was pretreated with H-Gly-Pro-Arg-Pro-OH-AcOH (3 mM final concentration) for 2–4 minutes to prevent fibrin polymerization. The cuvette with PPP was then placed in the background chamber to represent 100% light transmittance. The PRP test cuvettes were transferred to the test chamber and the test was started by adding the corresponding agonists. The reaction was allowed to run for 5 minutes. Platelet aggregation led to an increase in the light transmittance, which was recorded and the percentage of aggregation was determined as the highest point of deflection from the baseline. The inhibition of platelet aggregation was determined by comparing the aggregation responses in the control, vehicle-treated, and antagonist-treated samples.

Thrombin Generation Assay.

Thrombin generation was measured in PPP according to the Thrombinoscope guidelines using a calibrated automated thrombogram (CAT) system (Diagnostica Stago). Briefly, 80 µl of citrated PPP was mixed with 20 µl PPP reagent in round-bottom 96-well plates and thrombin generation was initiated after 10-minute preincubation at 37°C by adding 20 µl of FluCa reagent (Diagnostica Stago). Fluorescence (excitation: 360 nm; emission: 460 nm) was measured every 20 seconds for 60 minutes on a Fluoroskan Ascent plate reader (Thermo) and thrombin concentration was determined by the software using a thrombin calibrator reagent. Parameters including lag time, peak thrombin, and endogenous thrombin potential (ETP) were derived and plotted as a function of thrombin inhibitor concentration.

Thromboelastograph.

Thromboelastograph (TEG) analysis was performed with TEG-5000 analyzers (Haemonetics, Braintree, MA) following modification of the citrated RapidTEG protocol. Briefly, freshly drawn citrated blood from healthy volunteers was mixed with test compound or buffer control and incubated at 25°C for 5 minutes. The blood (340 µl) was transferred into a TEG sample cup preloaded with 20 µl of 0.2 M calcium chloride and 10 µl of RapidTEG reagent diluted 1:10 in HEPES-buffered saline [20 mM HEPES and 150 mM NaCl (pH 7.4)], and the TEG recording was initiated immediately. The 1:10 dilution of RapidTEG used was determined for optimal reproducibility of TEG parameters between multiple rounds of analysis. Two TEG-5000 analyzers (four channels) were in operation simultaneously for each titration series, with one channel used as a buffer control during each measurement. The interday and interchannel variability was minimal and met the standard set by the manufacturer (data not shown). From the TEG output, three key parameters, the reaction time, maximal amplitude, and percentage of lysis at 30 minutes after maximal amplitude, were extracted to assess clotting time, clot strength, and fibrinolytic activity, respectively.

Animals.

Male Sprague-Dawley rats (350–450 g) were used for the in vivo studies. All animal care and experimental procedures were approved by the Institutional Animal Care and Use Committee at Janssen Research & Development, LLC, and all animal studies were conducted in accordance with the Guide for the Care and Use of Laboratory Animals.

Rat AV Shunt Model of Thrombosis.

JNJ-9375 and apixaban were evaluated on the antithrombotic efficacy in the rat AV shunt model. Rats were anesthetized by intramuscular administration of 100 mg/kg of ketamine plus 10 mg/kg of xylazine (Hanna’s Pharmaceutical Supply CO., Wilmington, DE). Rats were placed on a heating pad to maintain body temperature and allowed to breathe normally on their own. Surgery was performed to isolate and catheterize the right carotid artery and both jugular veins to allow for administration of drugs, blood collection, and induction of thrombosis as previously described (Shang et al., 2014). After collecting a baseline blood sample (1 ml), drugs or vehicle were administered to the animals as an intravenous bolus 15 minutes before induction of thrombosis. An external tube was connected from the right carotid artery to the left jugular vein to initiate thrombi formation. Briefly, the shunt was prepared from a 6 cm section of Tygon tubing (Cole Parmer, Vernon Hills, IL) that connects on one end to the right carotid artery cannula and on the other end to the left vein cannula with 8 mm of special Masterflex Tygon tubing. A 5 cm length of surgical silk (size 2-0) (Ethicon, Bridgewater, NJ) was secured inside the shunt. All of the pieces of the shunt (except the 2-0 silk tread) were siliconized at least 24 hours prior to the assembly. Blood was allowed to flow through the tubing for 15 minutes and a blood clot was allowed to form on the thread. At the end of 15 minutes, the external tubing with the thread and blood clot was removed and the blood clot was weighed. A second blood sample (5 ml) was collected from the abdominal aorta and the animal was euthanized while still under anesthesia. Blood samples were used for drug level measurement and clotting assays.

Rat Tail Transection Bleeding Model.

JNJ-9375 (0.3, 1, 3, and 10 mg/kg) and apixaban (0.3, 1, and 3 mg/kg) were tested in a rat tail transection model to evaluate bleeding risks. Five minutes after receiving the intravenous injection of compounds, the rat tails were transected at a diameter of 3 mm and the cut ends of the tails were immediately immersed into 50-ml tubes filled with 40 ml of prewarmed saline. Bleeding was continuously monitored, and the tube was changed with prewarmed saline every 5 minutes for a total of 30 minutes. The readouts included the time to stop bleeding, total bleeding time, and total blood loss determined by measuring the hemoglobin mass in the saline using the QuantiChrom hemoglobin assay kit in which hemoglobin is converted into a colored end product and the intensity of the color is proportional to the hemoglobin concentration.

JNJ-9375 Concentration Measurement.

Measurement of the JNJ-9375 pharmacokinetic concentration in rat plasma was determined using an immunoassay method employing a homogenous sandwich assay format with electrochemiluminescent detection. The capture and detection reagents in the assay were both anti-human Fc monoclonal antibodies. Rat plasma samples were analyzed in the assay at 1/500 dilution. Briefly, diluted standards, quality controls, and samples containing the analyte were combined with a master mix containing the final working concentrations of the capture and detection reagent and incubated for 45 minutes at room temperature. The Meso Scale Discovery plate was washed and read on the Meso Scale Discovery sector imager (Meso Scale Diagnostics, Rockville, MD). The Meso Scale Discovery output files with the raw electrochemiluminescent counts were swept into Cyberlab and imported into Watson LIMS regression software for analysis. The Watson study regression used a five-parameter logistic (autoestimate) fit with a 1/Y² weighting factor.

Data Analysis and Statistics.

All of the data are expressed as mean ± S.E.M. Statistical analyses were performed using GraphPad Prism software (version 6.0; La Jolla, CA). Significance was defined as P < 0.05 as determined by one-way ANOVA followed by Tukey’s or Dunnett’s post hoc test for multiple comparisons. R package nlme was used to calculate the EC_2X values for in vitro clotting assays using the polynomial mixed-effects model.