Synthesis and evaluation of anthranilamide-based derivatives as FXa inhibitors

Factor Xa (FXa) plays a significant role in the blood coagulation cascade and is a promising target for anticoagulation drugs. Three oral FXa inhibitors have been approved by FDA for treating thrombotic diseases. In this study, 43 novel compounds were synthesized anthranilamide-based FXa inhibitors aiming to ameliorate the toxicity of traditional FXa inhibitors in clinic. The data indicated that the compounds 6a, 6a-b, 6a-e, 6k, 6k-a and 6k-b showed remarkable FXa inhibitory activity and excellent selectivity over thrombin in vitro. Selected compounds also exhibited anticoagulant activities in vitro consequently and were potent novel anti-coagulators in further.


INTRODUCTION
The coagulation and anticoagulation process was considered as a dynamic equilibrium process, the breakage of this balance was capable to induce thrombosis [1]. Warfarin was employed as oral anticoagulant for decades [2] however the clinical utility of warfarin was limited by its narrow therapeutic index, dietary restrictions, slow onset of action and the need for regular monitoring [3,4]. In order to overcome this clinical obstacle of warfarin, researchers focused on developing novel synthetic molecules to inhibit specific enzymes such as protease factor Xa (FXa) [5]. FXa is a junction of the intrinsic and extrinsic pathways, and it was convinced to be a key component in the coagulation factor activation and thrombosis formation [6]. Up to date, three oral direct FXa inhibitors such as Rivaroxaban have been approved to treat the venous thrombosis in clinics and several other candidates were in the variant stages of clinical studies or biological testing (the structures of three approved drugs were in Figure 1) [7]. However, these three medicines still possess specific obstacles in clinical utility, such as risk of bleeding (including intracranial bleeding, gastrointestinal bleeding, epistaxis and other fatal bleeds), no antidote and higher incidence of thromboembolic events after ceasing treatment [8]. Therefore, the development of novel antithrombotic drugs is still attracted many attention for unmet clinical demand.
According to structural analysis, the carboxamide group at both rivaroxaban and betrixaban was considered to play a crucial role for connecting the scaffold, P1 and P4. Betrixaban interacts with residues Gly218 and Gly216 in FXa by two hydrogen bonds [7]. All these structural information supported a hypothesis that the carboxamide group might be crucial component for the docking of FXa inhibitors with human FXa protein. In this study, we synthesized 43 novel compounds in which carboxamide group was involved.
Additionally, the clinical pharmacokinetics investigations of rivaroxaban revealed that the metabolic pathway of rivaroxaban is to hydrolyze the amide bond and then eliminate via renal and biliary/fecal routes rapidly [9]. The metabolic characterizations of rivaroxaban suggested less toxicity by reversing the order of chemical group in betrixaban which was presumed to result in totally different metabolic pathways in humans even.

Research Paper
In brief, we designed and evaluated novel synthesized compounds as inhibitors of FXa and their structural information upon computational simulation approaches.

Synthesis
The compounds listed in Tables 1 and 2 are prepared as shown in Figure 2. The compound 3 was synthesized from the anilines 2 and o-nitrobenzoyl chloride 1, Then the anilines 4 was prepared by reduction reaction. Compound 6 was accessed by treatment of compound 4 with compound 5. The compound's Activity for inhibit FXa was reported as IC50 and we also reported the Ki of several more potent compounds.

In Vitro inhibition activity studies on FXa
We extensively investigated different P1 group. Important SAR findings are summarized in Table 1. Compounds 6a-6o that contained same scaffold and P4 group were used for choosing potent P1 group. Compounds 6a, 6h, 6k and 6l exhibited a promising inhibitory activity against FXa with IC 50 values of 28. 7, 277.5, 23.8 and 181.5 nM. The P1 group of 6a, 6h, 6k and 6l were chosen for structural modification of next stage and provided an opportunity to improve potency.
Then, the SAR of substituent on benzene ring and different P4 group were examined (Table 2). Unfortunately, the 2-chloropyridine analogs (6h-a -6h-g) and 4-methoxybenzene analogs (6l-a -6l-g) still did not exhibited good activity. However, 5-bromothiophene analogs and 2, 4-dichlorobenzene analogs both showed excellent activity. In particular compounds 6a-b and 6k-b showed some validity with the IC 50 value of 18.8 and 35.5 nM. From these results we can find that 3-methylsubstituted scaffold (6a-c, 6a-g, 6k-c and 6k-g) was not suitable, the related compounds displayed poor IC 50 values at a micromole level no matter what kind of R 2 and R 3 . Compounds with 5-electron donating group-substituted scaffold exhibited almost 10 fold better than 5-electron withdrawing group substituted scaffold. The IC 50 value of compounds with non-substituted scaffold was almost at the same level with electron donating group scaffold. Table 3 shows the thrombin selectivity comparison and the extension of the prothrombin time (PT) of 6a, 6a-b, 6a-e, 6k, 6k-a and 6k-b. The IC50 of rivaroxaban against thrombin is 6.9 μM [10], our compounds were also highly selective, IC 50 were all far higher than 10 μM. The assay of PT also showed that our compound possessed improvement on anticoagulant selectivity.

Docking simulation studies
In order to analyze the detailed interactions between compounds 6a, 6a-b, 6a-e, 6k, 6k-a and 6k-b to FXa, the computational docking simulation studies were performed. As showed in Figure 3, There are residue Gly216 and Gly218 in S1 pocket of FXa, and some hydrogen bonds were taked shape betweed conpound and these amino-acid residue. the unit of pyridone or morpholino are access  the S4 pocket of FXa, which was composed of Tyr99 , Phe174 and Trp215. The p-p conjugation was formed between the chloro-or bromo-substituent and S1 pocket through amino-acid residue Tyr228. The mode of action was similar with rivaroxaban [7].

MATERIALS AND METHODS
Reagents and solvents were obtained from commercial suppliers and used as received without further purification. All reactions were monitored by thin layer chromatography. 1

Synthesis of 3 (3a-3h)
To a stirred solution of compound 2 (4.8 mmol) (2a-2b), K 2 CO 3 (0.80 g, 5.8 mmol) and DMAP (0.05 g, 0.4 mmol) in THF (20 mL), solution of compound 1 > 1000 (6.24 mmol) (1a-1d) in THF (5 mL) was added at room temperature and the mixture was refluxed for 2 h. The reaction mixture was cooled down to room temperature and concentrated under reduced pressure. Then water (100 mL) was added to the mixture and stirred for 10 min at room temperature. The resulting precipitate was collected by filtration. The reaction was monitored by TLC with EA.

Synthesis of 4 (4a-4h)
To a 250 mL round bottom flask, compound 3 (3.9 mmol) (3a-3h), zinc powder (2.05 g, 31.2 mmol), NH 4 Cl (2.11 g, 39 mmol), methanol (30 mL), THF (30 mL) and water (15 mL) were added. The mixture stirred at 40°C for 2 h. The reaction mixture was filtered, washed with DMF and the filtrate was concentrated under reduced pressure. Then water (200 mL) was added to the mixture and stirred for 0.5 h. The residue was filtered and washed with water to yield the title compound 4 (4a-4h). The reaction was monitored by TLC with EA.

Inhibition activity measurement against FXa
The inhibition of FXa was measured using human FXa (Hyphen BioMed, city, Paris, FRA) and chromogenic substrate CS-11(22) (Hyphen BioMed, Paris, FRA) in 384well microtiter plates at room temperature. The synthesized compounds 6 and rivaroxaban were dissolved in DMSO at a concentration of 10 mM and then serially diluted to spanning a range of 30 nM to 100 μM, respectively. 2 μL of FXa (56.8 nM), 16 μL of Tris buffer (adjust to pH 7.4 with HCl containing 0.3 M NaCl and 50 mM Tris) and 3 μL of test compound were added to the well, respectively. The negative control was composed of the same mixed solutions except replacing test compound with DMSO. The positive control was composed of the same mixed solutions except replacing test compound with rivaroxaban. After incubated at 37°C for 5 min, 8 μL of FXa substrate solution (3.5 mM) was added and then incubated at 37°C for 25 min. The FXa activity was measured at 405 nm using a SpectraMax M5 (Molecular Devices, Sunnyvale, CA, USA). The IC 50 was calculated by the software named SPSS (IBM, North Castle, NY, USA) and the Probit function in it.
Thrombin inhibition activity of 6a, 6a-b, 6a-e, 6k, 6k-a and 6k-b The inhibition of thrombin was measured using human FIIa (Hyphen BioMed, Paris, FRA) and chromogenic substrate CS-01(38) (Hyphen BioMed, Paris, FRA) in 384-well microtiter plates at room temperature. The compounds 6a, 6a-b, 6a-e, 6k, 6k-a, 6k-b and rivaroxaban were dissolved in DMSO to a concentration of 10 mM and then serially diluted to spanning a range of 10 μM to 100 μM, respectively. 2 μL of FIIa (3 NIH/mL), 20 μL of Tris buffer (adjust to pH 7.4 with HCl) containing 0.3 M NaCl and 50 mM Tris and 2 μL of test compound were added to the well, respectively. The negative control was composed of the same mixed solutions except replacing test compound with DMSO. The positive control was composed of the same mixed solutions except replacing test compound with rivaroxaban. After incubated at 37°C for 5 min, 3 μL of FIIa substrate solution (4 mM) was added and then incubated at 37°C for 25 min. The FIIa activity was measured at 405 nm using a SpectraMax M5 (Molecular Devices).

Prothrombin time (PT) assay
A commercially available automatic coagulometer (Steellex Science Instrument Co., Ltd., Beijing, China) was employed to measure PT. The clotting times were also measured using the instrument itself, in accordance with the manufacturer's instructions. Increasing concentrations of inhibitor or solvent were added to human (39 Years old, male, Chinese) plasma and incubated for 3 min at 37°C. Prothrombin time (PT) was determined by automatic coagulometer.

Docking simulation
FXa structure was selected from the protein data bank (PDB code: 2xbv) and prepared using Protein Preparation Wizard in Schrödinger package, including assigning bond orders, adding hydrogen atoms, deleting water molecules, creating disulfide bonds and capping terminals. The original ligand of the protein structure-XBV was used as the docking center to generate the receptor grid parameters. The box size was set as 12 Å. Compounds 6a, 6a-b, 6a-e, 6k, 6k-a and 6k-b were prepared using the LigPrep module in Schrödinger. Epik method was used to determine possible ionization state of ligands at pH 7.0 ± 2.0 and low-energy conformers were produced using OPLS-2005 force field. Molecular docking calculations were performed by using Glide module with default parameters at standard precision in Schrödinger.

CONCLUSIONS
In conclusion, the synthetic anthranilamide compounds were evaluated as novel inhibitors of FXa. Among 43 compounds, 6a, 6a-b, 6a-e, 6k, 6k-a and 6k-b showed excellent activity and selectivity over thrombin. In further, the compound 6a-e showed best anticoagulant activity (3.8 μM) in the six compounds significantly. The computational docking simulation study clarified the interactions mode of compounds. The results in this study indicated that compound 6a-e exhibited remarkable thrombin inhibitory effect in intro and it might be a potent novel anti-coagulator for further in vivo studies.

ACKNOWLEDGMENTS AND FUNDING
This project was supported by the Scientific and Technological Plan Projects of Tianjin (No. 12ZCZDSY01100).