Fragment-Based Lead Generation of 5-Phenyl-1H-pyrazole-3-carboxamide Derivatives as Leads for Potent Factor Xia Inhibitors

FXIa is suggested as a major target for anticoagulant drug discovery because of reduced risk of bleeding. In this paper, we defined 5-phenyl-1H-pyrazole-3-carboxylic acid derivatives as privileged fragments for FXIa inhibitors’ lead discovery. After replacing the (E)-3-(5-chloro-2-(1H-tetrazol-1-yl)phenyl)acrylamide moiety in compound 3 with 5-(3-chlorophenyl)-1H-pyrazole-3-carboxamide, we traveled from FXIa inhibitor 3 to a scaffold that fused the privileged fragments into a pharmacophore for FXIa inhibitors. Subsequently, we synthesized and assessed the FXIa inhibitory potency of a series of 5-phenyl-1H-pyrazole-3-carboxamide derivatives with different P1, P1′ and P2′moiety. Finally, the SAR of them was systematically investigated to afford the lead compound 7za (FXIa Ki = 90.37 nM, 1.5× aPTT in rabbit plasma = 43.33 μM) which exhibited good in vitro inhibitory potency against FXIa and excellent in vitro coagulation activities. Furthermore, the binding mode of 7za with FXIa was studied and the results suggest that the 2-methylcyclopropanecarboxamide group of 7za makes 2 direct hydrogen bonds with Tyr58B and Thr35 in the FXIa backbone, making 7za binds to FXIa in a highly efficient manner.


Introduction
Cardiovascular (CV) disease continues to be the leading cause of death worldwide [1]. Thrombosis is the common underlying pathology of cardiovascular diseases and anticoagulants are the mainstay to prevent and/or treat thrombosis [2]. In clinical use, anticoagulants include antithrombin activators (heparins including unfractionated heparin, low molecular weight heparins and fondaparinux), vitamin K antagonists (coumarins such as warfarin), direct inhibitors of thrombin (hirudins, argatroban and dabigatran etexilate) and oral direct FXa inhibitors (rivaroxaban, apixaban, edoxaban and betrixaban) [3]. Although these agents possess high efficacy and relatively low cost to benefit ratio, they stillremain be associated with the life-threatening side effect of internal bleeding [4,5]. Therefore, despite the progresses made in past few years, there is also an urgent clinical need for developing new anticoagulants to prevent and/or treat thromboembolic diseases without the risk of bleeding or with low bleeding risk.
also an urgent clinical need for developing new anticoagulants to prevent and/or treat thromboembolic diseases without the risk of bleeding or with low bleeding risk.
Generally, it is known from coagulation cascade that proteins in the intrinsic pathway are more important for the amplification phase of coagulation, whereas those belonging to the extrinsic and common coagulation pathways are more involved in the initiation and propagation phases. Current anticoagulants used for treating thrombosis mainly target two key serine proteases, thrombin and factor Xa (FXa), and they both belong to the common pathway of the coagulation cascade. Meanwhile, it has been postulated that selective inhibition of intrinsic coagulation factors could provide antithrombotic benefits with low bleeding risk because this will keep the other pathways of coagulation intact for hemostasis [6][7][8]. It's shown by epidemiological and clinical studies that the inhibition of Factor XIa (FXIa) which belongs to the intrinsic pathway of the coagulation cascade has emerged as an excellent way to achieve anticoagulation without significant effects on hemostasis [9]. Human FXI deficiency (hemophilia C) was first described as a mild to moderate bleeding disorder [10]. However, these affected patients rarely suffer from spontaneous bleeding episodes. Furthermore, epidemiologic studies showed an increased risk of thrombosis in subjects with elevated FXI levels and some protection from thrombosis in subjects with reduced FXI levels [11,12].
Furthermore, the open-label, parallel-group Phase II study (NCT01713361) showed that reducing FXI levels specifically by an antisense oligonucleotide in patients undergoing elective knee arthroplasty was an effective method for postoperative venous thromboembolism prevention and appeared to be safe with respect to the risk of bleeding [13]. In addition, the Phase I study (NCT03197779) showed that the small molecule FXIa inhibitor BMS-962212 had good enough tolerability, pharmacokinetics and pharmacodynamics properties suitable for investigational use as an acute antithrombotic agent in Japanese or non-Japanese subjects [14,15]. In summary, FXIa is suggested as a major target for anticoagulant drug discovery because of reduced risk of bleeding [7].
In recent years, lots of small molecule FXIa inhibitors were reported, including compounds 1-6 [16][17][18][19][20][21] and BMS-962212 [15] (Figure 1). Nevertheless, small molecule FXIa inhibitors' research remained in clinical phase. Further development of novel small molecule FXIa inhibitors and investigation of their Structure-Activity Relationships (SAR) are required. It was reported compound 3 exhibited good FXIa affinity activities (FXIa Ki = 2nM) but short t1/2 lives and other drawbacks. As shown in Figure 2a, the structure 3 binds to FXIa in S1-S1′-S2′ mode. P1, P1 prime (P1′) and P2 prime (P2′) moieties of compound 3 occupy the S1, S1′and S2′ pocket of FXIa (Figure 2a). The carbonyl groups of the cinnamide linker in the P1 and P1′ moieties are well positioned within the oxyanion hole by forming key hydrogen bonds with Ser195 residue (Figure 2b); In addition, the amide NH functional group forms a water mediated hydrogen bond with Leu41 [18]. It's obvious that the mentioned two linkers are crucial to ligands' binding mode and it's necessary to retain the two linkers in the structure optimization. As shown in Figure 2a, the structure 3 binds to FXIa in S1-S1 -S2 mode. P1, P1 prime (P1 ) and P2 prime (P2 ) moieties of compound 3 occupy the S1, S1 and S2 pocket of FXIa (Figure 2a). The carbonyl groups of the cinnamide linker in the P1 and P1 moieties are well positioned within the oxyanion hole by forming key hydrogen bonds with Ser195 residue (Figure 2b); In addition, the amide NH functional group forms a water mediated hydrogen bond with Leu41 [18]. It's obvious that the mentioned two linkers are crucial to ligands' binding mode and it's necessary to retain the two linkers in the structure optimization. It's summarized that some known FXIa inhibitors such as compound 3, 4, 6 ( Figure 1) were consisted of 4 parts-P1, P1 prime (P1′), P2 prime (P2′) moieties and scaffold. It's the molecule which was developed as a FXIa inhibitor and each part of the molecule was necessary but might be not crucial for the overall activity generally. Therefore, it's reasonable to choose an useful building block as part of FXIa inhibitors. Meanwhile, it is shown clearly that compounds 1-2 and 8-10 ( Figure 3) were all based on the 5-phenyl-1H-pyrazole-3-carboxylic acid building block. Of them, compound 8 was a tissue-nonspecific alkaline phosphatase (TNAP) inhibitor [22], compound 9 was an enkephalinase inhibitor [23], compound 10 was a mGlu5 receptor negative allosteric modulator and compounds 1 and 2 both were FXIa inhibitors [16,17,24]. Thus, novel FXIa inhibitors might be developed based on the excellent 5-phenyl-1H-pyrazole-3-carboxylic acid derivatives by using the Fragment Based Lead Generation strategy. Furthermore, it was known that structure optimization using an optimal neutral P1 moiety could significantly facilitate identification of a potent, orally bioavailable FXIa inhibitor [25]. Taking these points into account, we replaced the P1 moiety in compound 3 with more neutral chloro-benzene leading to compound 7a and further modification of P1, P1′ and P2′ moieties in 7a furnished a series of 5-phenyl-1H-pyrazole-3-carboxamide FXIa inhibitors, of which the SAR was then systematically investigated ( Figure 4).  It's summarized that some known FXIa inhibitors such as compound 3, 4, 6 (Figure 1) were consisted of 4 parts-P1, P1 prime (P1 ), P2 prime (P2 ) moieties and scaffold. It's the molecule which was developed as a FXIa inhibitor and each part of the molecule was necessary but might be not crucial for the overall activity generally. Therefore, it's reasonable to choose an useful building block as part of FXIa inhibitors. Meanwhile, it is shown clearly that compounds 1-2 and 8-10 ( Figure 3) were all based on the 5-phenyl-1H-pyrazole-3-carboxylic acid building block. Of them, compound 8 was a tissue-nonspecific alkaline phosphatase (TNAP) inhibitor [22], compound 9 was an enkephalinase inhibitor [23], compound 10 was a mGlu5 receptor negative allosteric modulator and compounds 1 and 2 both were FXIa inhibitors [16,17,24]. Thus, novel FXIa inhibitors might be developed based on the excellent 5-phenyl-1H-pyrazole-3-carboxylic acid derivatives by using the Fragment Based Lead Generation strategy. It's summarized that some known FXIa inhibitors such as compound 3, 4, 6 ( Figure 1) were consisted of 4 parts-P1, P1 prime (P1′), P2 prime (P2′) moieties and scaffold. It's the molecule which was developed as a FXIa inhibitor and each part of the molecule was necessary but might be not crucial for the overall activity generally. Therefore, it's reasonable to choose an useful building block as part of FXIa inhibitors. Meanwhile, it is shown clearly that compounds 1-2 and 8-10 ( Figure 3) were all based on the 5-phenyl-1H-pyrazole-3-carboxylic acid building block. Of them, compound 8 was a tissue-nonspecific alkaline phosphatase (TNAP) inhibitor [22], compound 9 was an enkephalinase inhibitor [23], compound 10 was a mGlu5 receptor negative allosteric modulator and compounds 1 and 2 both were FXIa inhibitors [16,17,24]. Thus, novel FXIa inhibitors might be developed based on the excellent 5-phenyl-1H-pyrazole-3-carboxylic acid derivatives by using the Fragment Based Lead Generation strategy. Furthermore, it was known that structure optimization using an optimal neutral P1 moiety could significantly facilitate identification of a potent, orally bioavailable FXIa inhibitor [25]. Taking these points into account, we replaced the P1 moiety in compound 3 with more neutral chloro-benzene leading to compound 7a and further modification of P1, P1′ and P2′ moieties in 7a furnished a series of 5-phenyl-1H-pyrazole-3-carboxamide FXIa inhibitors, of which the SAR was then systematically investigated ( Figure 4).  Furthermore, it was known that structure optimization using an optimal neutral P1 moiety could significantly facilitate identification of a potent, orally bioavailable FXIa inhibitor [25]. Taking these points into account, we replaced the P1 moiety in compound 3 with more neutral chloro-benzene leading to compound 7a and further modification of P1, P1 and P2 moieties in 7a furnished a series of 5-phenyl-1H-pyrazole-3-carboxamide FXIa inhibitors, of which the SAR was then systematically investigated ( Figure 4). It's summarized that some known FXIa inhibitors such as compound 3, 4, 6 ( Figure 1) were consisted of 4 parts-P1, P1 prime (P1′), P2 prime (P2′) moieties and scaffold. It's the molecule which was developed as a FXIa inhibitor and each part of the molecule was necessary but might be not crucial for the overall activity generally. Therefore, it's reasonable to choose an useful building block as part of FXIa inhibitors. Meanwhile, it is shown clearly that compounds 1-2 and 8-10 ( Figure 3) were all based on the 5-phenyl-1H-pyrazole-3-carboxylic acid building block. Of them, compound 8 was a tissue-nonspecific alkaline phosphatase (TNAP) inhibitor [22], compound 9 was an enkephalinase inhibitor [23], compound 10 was a mGlu5 receptor negative allosteric modulator and compounds 1 and 2 both were FXIa inhibitors [16,17,24]. Thus, novel FXIa inhibitors might be developed based on the excellent 5-phenyl-1H-pyrazole-3-carboxylic acid derivatives by using the Fragment Based Lead Generation strategy. Furthermore, it was known that structure optimization using an optimal neutral P1 moiety could significantly facilitate identification of a potent, orally bioavailable FXIa inhibitor [25]. Taking these points into account, we replaced the P1 moiety in compound 3 with more neutral chloro-benzene leading to compound 7a and further modification of P1, P1′ and P2′ moieties in 7a furnished a series of 5-phenyl-1H-pyrazole-3-carboxamide FXIa inhibitors, of which the SAR was then systematically investigated ( Figure 4).

Chemistry
The synthetic route to target compounds 7a-7n is shown in Scheme 1. The coupling of compound 12 with a series of carboxylic acids 11a-11n afforded 13a-13n in the presence of 1-hydroxybenzotriazole, N-(3-dimethylaminopropyl)-N -ethylcarbodiimide hydrochloride and N,N-diisopropyl-ethylamine in DMF, which weretreated with LiOH·H 2 O in aqueous methanol at 40 • C and acidified with 1 M hydrochloric acid to pH 3-4 to afford target compounds 7a-7n.

Chemistry
The synthetic route to target compounds 7a-7n is shown in Scheme 1. The coupling of compound 12 with a series of carboxylic acids 11a-11n afforded 13a-13n in the presence of 1-hydroxybenzotriazole, N-(3-dimethylaminopropyl)-N′-ethylcarbodiimide hydrochloride and N,N-diisopropyl-ethylamine in DMF, which weretreated with LiOH·H2O in aqueous methanol at 40 °C and acidified with 1 M hydrochloric acid to pH 3-4 to afford target compounds 7a-7n. The synthetic route to target compounds 7o-7w is summarized in Scheme 2. The intermediates 15o-15w were prepared by the coupling of carboxylic acids 11f, 11k, 11m, 11r with amines 14a-14f in the presence of 1-hydroxybenzotriazole, N-(3-dimethylaminopropyl)-N′-ethylcarbodiimide hydrochloride and N,N-diisopropylethylamine in DMF, which were treated with LiOH·H2O in aqueous ethanol at 40 °C. The resultant mixture was acidified with 1 M hydrochloric acid to pH 3-4 to afford target compounds 7o-7w.  The synthetic route to target compounds 7x-7y is depicted in Scheme 3. Boc-Asp(OtBu)-OH (16) first reacted with ethyl 5-amino-1H-indole-2-carboxylate in the presence of POCl 3 and pyridine in CH 2 Cl 2 to give the adduct 17, which was deprotected with TFA/CH 2 Cl 2 (1/1 by v/v), and treated with (Boc) 2  with NaH in THF leading to 24zc. Target compounds 7za-7zc were synthesized from 24za-24zc according to the procedure described for the preparation of 7z.
The synthetic route to intermediates 11a-11n and 11r is shown in Scheme 5 according to relevant references [26]. 27a first reacted with diethyl oxalate in the presence ofLiHMDS in MTBE at 0 °C and the resultant mixture was acidified with hydrochloric acid (1 M) to yield 28a. The cyclization of 28a with hydrazine hydrate in EtOH at 80 °C yielded 29a. The hydrolysis of 29a with NaOH in EtOH and H2O and acidification with 1 M hydrochloric acid afforded 11a. The intermediates 11b-11n and 11r were synthesized from 27b-27n according to the procedure described for the preparation of 11a. The synthetic route to target compounds 7z, 7za-7zc is illustrated in Scheme 4. Coupling of carboxylic acid 11f with amine 21 in the presence of 1-hydroxybenzotriazole, N-(3-dimethyl-aminopropyl)-N -ethylcarbodiimide hydrochloride and N,N-diisopropylethylamine in DMF at room temperature, followed by the hydrolysiswithLiOH·H 2 O in aqueous methanol at room temperature, and subsequent acidification with 1 M hydrochloric acid afforded compound 22. The coupling of compound 22 with ethyl 5-amino-1H-indole-2-carboxylate in the presence of POCl 3 and pyridine in CH 2 Cl 2 , and next reduction of NO 2 with H 2 (1 atm) in MeOH and ethyl acetate in the presence of Pd/C yielded compound 23. Hydrolysis of ethyl ester with LiOH·H 2 O in aqueous ethanol at 40 • C, and acidification of the resultant mixture with 1 M hydrochloric acid afforded target compounds 7z.Compound 23 reacted with 2-methylcyclopropanecarboxylic acid in the presence of POCl 3 and pyridine in CH 2 Cl 2 leading to 24za. Reaction of 23 with 4-methylpiperazine-1-carbonyl chloride in the presence of pyridine and 4-dimethylaminopyridine in DMF afforded 24zb. Compound 23 was coupled with 2-(2-chloroethoxy)acetic acid in the presence of N-ethoxycarbonyl-2-ethoxy-1,2-dihydroquinoline in THF, and the resultant product was treated with NaH in THF leading to 24zc. Target compounds 7za-7zc were synthesized from 24za-24zc according to the procedure described for the preparation of 7z.
The synthetic route to intermediates 11a-11n and 11r is shown in Scheme 5 according to relevant references [26]. 27a first reacted with diethyl oxalate in the presence ofLiHMDS in MTBE at 0 • C and the resultant mixture was acidified with hydrochloric acid (1 M) to yield 28a. The cyclization of 28a with hydrazine hydrate in EtOH at 80 • C yielded 29a. The hydrolysis of 29a with NaOH in EtOH and H 2 O and acidification with 1 M hydrochloric acid afforded 11a. The intermediates 11b-11n and 11r were synthesized from 27b-27n according to the procedure described for the preparation of 11a. The synthetic route to intermediates 14a-14f and12 is depicted in Scheme 6 according to relevant references [18]. Compound 32 was prepared by the coupling of carboxylic acid 30 with amine 31 in the presence of POCl3 and pyridine in CH2Cl2 at −10 °C. Removal of N-Boc of 32 with hydrochloric acid in ethyl acetate yielded compound 12. The preparation of compound 14a-14f followed the procedure described for the preparation of compound 12. The synthetic route to intermediates 14a-14f and12 is depicted in Scheme 6 according to relevant references [18]. Compound 32 was prepared by the coupling of carboxylic acid 30 with amine 31 in the presence of POCl3 and pyridine in CH2Cl2 at −10 °C. Removal of N-Boc of 32 with hydrochloric acid in ethyl acetate yielded compound 12. The preparation of compound 14a-14f followed the procedure described for the preparation of compound 12. The synthetic route to intermediates 14a-14f and12 is depicted in Scheme 6 according to relevant references [18]. Compound 32 was prepared by the coupling of carboxylic acid 30 with amine 31 in the presence of POCl 3 and pyridine in CH 2 Cl 2 at −10 • C. Removal of N-Boc of 32 with hydrochloric acid in ethyl acetate yielded compound 12. The preparation of compound 14a-14f followed the procedure described for the preparation of compound 12.

Activated Partial Prothrombin Time (aPTT) In Vitro Coagulation Assays
In order to assess the in vitro coagulation activity of compound 7za, activated partial prothrombin time (aPTT) of 7za and compound 4 was compared.As indicated in Table 2, 7za showed good in vitro coagulation activity with 1.5× aPTT value of 43.33 μM in rabbit plasma.

Activated Partial Prothrombin Time (aPTT) In Vitro Coagulation Assays
In order to assess the in vitro coagulation activity of compound 7za, activated partial prothrombin time (aPTT) of 7za and compound 4 was compared.As indicated in Table 2, 7za showed good in vitro coagulation activity with 1.5× aPTT value of 43.33 μM in rabbit plasma.  phenyl

Activated Partial Prothrombin Time (aPTT) In Vitro Coagulation Assays
In order to assess the in vitro coagulation activity of compound 7za, activated partial prothrombin time (aPTT) of 7za and compound 4 was compared.As indicated in Table 2, 7za showed good in vitro coagulation activity with 1.5× aPTT value of 43.33 μM in rabbit plasma.  phenyl

Activated Partial Prothrombin Time (aPTT) In Vitro Coagulation Assays
In order to assess the in vitro coagulation activity of compound 7za, activated partial prothrombin time (aPTT) of 7za and compound 4 was compared.As indicated in Table 2, 7za showed good in vitro coagulation activity with 1.5× aPTT value of 43.33 μM in rabbit plasma.  phenyl

Activated Partial Prothrombin Time (aPTT) In Vitro Coagulation Assays
In order to assess the in vitro coagulation activity of compound 7za, activated partial prothrombin time (aPTT) of 7za and compound 4 was compared.As indicated in Table 2, 7za showed good in vitro coagulation activity with 1.5× aPTT value of 43.33 μM in rabbit plasma.  phenyl

Activated Partial Prothrombin Time (aPTT) In Vitro Coagulation Assays
In order to assess the in vitro coagulation activity of compound 7za, activated partial prothrombin time (aPTT) of 7za and compound 4 was compared.As indicated in Table 2, 7za showed good in vitro coagulation activity with 1.5× aPTT value of 43.33 μM in rabbit plasma.

Activated Partial Prothrombin Time (aPTT) In Vitro Coagulation Assays
In order to assess the in vitro coagulation activity of compound 7za, activated partial prothrombin time (aPTT) of 7za and compound 4 was compared. As indicated in Table 2, 7za showed good in vitro coagulation activity with 1.5× aPTT value of 43.33 µM in rabbit plasma.

Molecular Model Studies on the Interaction of Compound 7za with FXIa
Molecular docking method was used to study the binding mode of compound 7za in the active site of FXIa. Two different binding modes of compound 7za with FXIa were obtained (Figure 5a,b). Superposition of these two binding modes with the known compound 3 (Figure 5c,d) showed that 7za in the second binding mode and known compound 3 overlay well (Figure 5d), indicating that this kind of binding model of 7za with FXIa is more reasonable (Figure 5b). As designated, the 5-(3-chloro-2-fluorophenyl)-1H-pyrazole moiety is located deeply inS1 pocket of FXIa with the Cl atom forming an interaction with Tyr228, and the carbonyl group of 5-(3-chloro-2-fluorophenyl)-1H-pyrazole-3-carboxamide is well positioned within the oxyanion hole by forming key hydrogen bonds with Gly193 residues which is similar to the binding model of compound 3 with FXIa ( Figure 5d). Furthermore, the 2-methylcyclopropanecarboxamide group of 7za makes 2 direct hydrogen bonds with Tyr58B and Thr35 in the FXIa backbone, making 7za binds to FXIa in a highly efficient manner (Figure 5b).

Molecular Model Studies on the Interaction of Compound 7za with FXIa
Molecular docking method was used to study the binding mode of compound 7za in the active site of FXIa. Two different binding modes of compound 7za with FXIa were obtained (Figure 5a,b). Superposition of these two binding modes with the known compound 3 (Figure 5c,d) showed that 7za in the second binding mode and known compound 3 overlay well (Figure 5d), indicating that this kind of binding model of 7za with FXIa is more reasonable (Figure 5b). As designated, the 5-(3-chloro-2-fluorophenyl)-1H-pyrazole moiety is located deeply inS1 pocket of FXIa with the Cl atom forming an interaction with Tyr228, and the carbonyl group of 5-(3-chloro-2-fluorophenyl)-1H-pyrazole-3-carboxamide is well positioned within the oxyanion hole by forming key hydrogen bonds with Gly193 residues which is similar to the binding model of compound 3with FXIa (Figure 5d). Furthermore, the 2-methylcyclopropanecarboxamide group of 7za makes 2 direct hydrogen bonds with Tyr58B and Thr35 in the FXIa backbone, making 7za binds to FXIa in a highly efficient manner (Figure 5b).

General Information
Reagents and solvents were purchased from commercial suppliers and used without further purification. Reactions were monitored by thin layer chromatography. 1 H-NMR spectra (400 MHz) and 13

Chemistry
Ethyl 2,4-dioxo-4-phenylbutanoate (28b): To a stirred solution of acetophenone (27b) (2.0 g, 16.5 mmol) in MTBE (30 mL) was added lithium hexamethyldisilazide (1.3 M, 12.7 mL, 16.5 mmol) dropwise at 0 • C; After addition, the reaction mixture was stirred at 0 • C for 0.5 h and diethyl oxalate (3.0 g, 20.8 mmol) was added dropwise. Then, the mixture was stirred at room temperature overnight. TLC analysis showed reaction was complete and the reaction mixture was extracted with H 2 O (20 mL). The aqueous layer was separated, acidified by hydrochloric acid (1 M) to pH 6 and extracted by ethyl acetate (10 mL × 2). The combined organic layer was concentrated in vacuum to give 28b as yellow oil, which was used for next step without further purification (3.4 g, 92.7% yield).
Ethyl3-phenyl-1H-pyrazole-5-carboxylate (29b): To a solution of 28b (3.4 g, 15.4 mmol) in EtOH (15 mL) was added hydrazine hydrate (1.2 g, 24.0 mmol) and the mixture was stirred at 50 • C for 2 h when TLC analysis indicated completion of reaction. Then the reaction mixture was evaporated to get crude 29b as brown oil, which was used for next step without further purification (2.6 g, 77.9% yield). Compounds 11a, 11r and 11c-11n were synthesized according to the procedure described for the preparation of 11b. (S)-Ethyl 5-(2-((tert-butoxycarbonyl)amino)-3-phenylpropanamido)-1H-indole-2-carboxylate (35a): To a mixture of Boc-Phe-OH (33a) (5 g, 18.9 mmol), ethyl 5-amino-1H-indole-2-carboxylate (34) (3.9 g, 18.9 mmol) and pyridine (5 mL) in CH 2 Cl 2 (50 mL) at −10 • C was added POCl 3 (2.9 g, 18.9 mmol) dropwise. After addition, the reaction mixture was stirred at −10 • C for 2 h when TLC analysis indicated completion of reaction, then H 2 O (10 mL) was added and the organic layer was separated, washed by hydrochloric acid (1 M, 10 mL), dried by Na 2 SO 4 , filtered. The filtrate was evaporated in vacuum to get crude 35a as a brown solid, which was used for next step without further purification (6.6 g, 77.6% yield). Compounds 14b~14f were synthesized according to the procedure described for the preparation of 14a. and the reaction mixture was stirred at room temperature overnight.Then TLC analysis indicated reaction was complete, and H 2 O (40 mL) was added.The mixture was stirred for 10 min and filtered to get crude product 13a as a yellow solid, which was used for next step without further purification.  Compounds 7b~7n were synthesized according to the procedure described for the preparation of 7a.          Compounds 7p~7w were synthesized according to the procedure described for the preparation of 7o.      A mixture of compound 22 (3.00 g, 6.93 mmol), ethyl 5-amino-1H-indole-2-carboxylate (1.42 g), N-(3-dimethylaminopropyl)-N -ethylcarbodiimide hydrochloride (2.66 g, 13.9 mmol), 1-hydroxybenzotriazole (1.87 g, 13.9 mmol) and N,N-diisopropylethylamine (1.79 g, 13.9 mmol) in DMF(20 mL) was stirred at room temperature overnight when TLC analysis indicated completion of reaction, then H 2 O (200 mL) was added. The suspension was stirred for 10 min and filtered. The filter cake was dried at room temperature overnight and resolve in MeOH (200 mL) and ethyl acetate (100 mL). To the solution was added Pd/C (10%, 0.30 g) and the reaction mixture was stirred at the atmosphere of H 2 overnight when TLC analysis indicated completion of reaction.The suspension was filtered and the filtrate was concentrated in vacuum to afford intermediate 23 as a grey solid (3.1 g, 75.6% yield), m.p.: 120-122 • C,