Structure-Based Discovery of a Series of 5H-Pyrrolo[2,3-b]pyrazine FGFR Kinase Inhibitors

Fibroblast growth factor receptors (FGFRs), a subfamily of receptor tyrosine kinases, are aberrant in various cancer types, and considered to be promising targets for cancer therapy. We started with a weak-active compound that was identified from our internal hepatocyte growth factor receptor (also called c-Met) inhibitor project, and optimized it with the guidance of a co-crystal structure of compound 8 with FGFR1. Through rational design, synthesis, and the biological evaluation of a series of 5H-pyrrolo[2,3-b]pyrazine derivatives, we discovered several potent FGFR kinase inhibitors. Among them, compound 13 displayed high selectivity and favorable metabolic properties, demonstrating a promising lead for further development.


Introduction
Fibroblast growth factors receptors (FGFR1-4) are a subfamily of receptor tyrosine kinases (RTKs) that are involved in many cellular processes such as angiogenesis, embryogenesis, tissue homeostasis, wound repair, and cancer [1][2][3]. In the human genome, 22 fibroblast growth factors (FGFs) have been identified as ligands of FGFRs, and are usually classified into seven subfamilies, according to their biochemical function sequence similarity and evolutionary relationships [4,5]. Upon binding with fibroblast growth factors, FGFRs induce dimerization and autophosphorylation on tyrosine residues, resulting in the activation of kinases downstream signaling, including MEK-ERK, PI3K-Akt, and PLCγ [6,7]. Numerous evidences highlight that the activation of FGF/FGFR signaling plays a critical role in tumor progression and growth [8][9][10]. Moreover, aberrant signaling of FGF/FGFR has been frequently found in various cancers, making FGFR a hot therapeutic target in anticancer drug development [11]. Selective and potent FGFR inhibitors are needed, because, in general, compounds that are selective to one intended target kinase can potentially claim a more favorable safety profile than multi-target compounds [12][13][14]. Currently, several FGFR-targeted small molecules have been evaluated in clinical trials for cancer treatment, such as NVP-BGJ398 (1) [15], AZD4547 (2) [16,17], LY2874455 (3) [18], and CH5183284 (4) [19] (Figure 1). Those FGFR inhibitors exhibited distinct effects on different mutants, as they may provide unique therapeutic benefits for certain patients, so that developing FGFR inhibitors with a novel scaffold is in constant demand.
Previously, we described a series of 1-sulfonylpyrazolo [4,3-b]pyridines as potent and selective hepatocyte growth factor receptor (c-Met) inhibitors [20]. In the development of c-Met kinase inhibitors, we found that several 1-sulfonylpyrazolo [4,3-b]pyridines showed definite activity against FGFR1 at 10 µM, as shown in Table 1. Since the 1-sulfonylpyrazolo [4,3-b]pyridines as selective c-Met inhibitors have been reported [20], we are inquisitive about the binding mode of this chemotype as FGFR inhibitors. Therefore, we solved the X-ray crystal structure (PDB ID: 5Z0S) of compound 8 with FGFR1, as illustrated in Figure 2. Interestingly, the co-crystal structure of compound 8 bound to the FGFR1 kinase domain revealed novel binding interactions, which has not been reported. In this paper, we present our structure optimization of this series of compounds as FGFR inhibitors, in the hope that the investigation can stimulate new ideas for developing selective FGFR inhibitors as anticancer drugs. claim a more favorable safety profile than multi-target compounds [12][13][14]. Currently, several FGFR-targeted small molecules have been evaluated in clinical trials for cancer treatment, such as NVP-BGJ398 (1) [15], AZD4547 (2) [16,17], LY2874455 (3) [18], and CH5183284 (4) [19] (Figure 1). Those FGFR inhibitors exhibited distinct effects on different mutants, as they may provide unique therapeutic benefits for certain patients, so that developing FGFR inhibitors with a novel scaffold is in constant demand. Previously, we described a series of 1-sulfonylpyrazolo [4,3-b]pyridines as potent and selective hepatocyte growth factor receptor (c-Met) inhibitors [20]. In the development of c-Met kinase inhibitors, we found that several 1-sulfonylpyrazolo [4,3-b]pyridines showed definite activity against FGFR1 at 10 μM, as shown in Table 1. Since the 1-sulfonylpyrazolo [4,3-b]pyridines as selective c-Met inhibitors have been reported [20], we are inquisitive about the binding mode of this chemotype as FGFR inhibitors. Therefore, we solved the X-ray crystal structure (PDB ID: 5Z0S) of compound 8 with FGFR1, as illustrated in Figure 2. Interestingly, the co-crystal structure of compound 8 bound to the FGFR1 kinase domain revealed novel binding interactions, which has not been reported. In this paper, we present our structure optimization of this series of compounds as FGFR inhibitors, in the hope that the investigation can stimulate new ideas for developing selective FGFR inhibitors as anticancer drugs. Superimposed the crystal structure of a sulfoamide ligand bound to a hepatocyte growth factor receptor (c-Met) (PDB ID: 2WDI) to the structure of compound 8 bound to FGFR1. The protein was depicted in cartoon style, while the inhibitors and interacted residues were shown as sticks.

Results and Discussion
The biological activity of compounds (5, 6, 7, 8) as c-Met inhibitors was reported in our previous work [20], the compounds with the scaffold of 1-sulfonylpyrazolo [4,3-b]pyridines showed definite FGFR1 activity at 10 μM, all with a more than 50% inhibition ratio towards FGFR1 ( Table 1). The claim a more favorable safety profile than multi-target compounds [12][13][14]. Currently, several FGFR-targeted small molecules have been evaluated in clinical trials for cancer treatment, such as NVP-BGJ398 (1) [15], AZD4547 (2) [16,17], LY2874455 (3) [18], and CH5183284 (4) [19] (Figure 1). Those FGFR inhibitors exhibited distinct effects on different mutants, as they may provide unique therapeutic benefits for certain patients, so that developing FGFR inhibitors with a novel scaffold is in constant demand. Previously, we described a series of 1-sulfonylpyrazolo [4,3-b]pyridines as potent and selective hepatocyte growth factor receptor (c-Met) inhibitors [20]. In the development of c-Met kinase inhibitors, we found that several 1-sulfonylpyrazolo [4,3-b]pyridines showed definite activity against FGFR1 at 10 μM, as shown in Table 1. Since the 1-sulfonylpyrazolo [4,3-b]pyridines as selective c-Met inhibitors have been reported [20], we are inquisitive about the binding mode of this chemotype as FGFR inhibitors. Therefore, we solved the X-ray crystal structure (PDB ID: 5Z0S) of compound 8 with FGFR1, as illustrated in Figure 2. Interestingly, the co-crystal structure of compound 8 bound to the FGFR1 kinase domain revealed novel binding interactions, which has not been reported. In this paper, we present our structure optimization of this series of compounds as FGFR inhibitors, in the hope that the investigation can stimulate new ideas for developing selective FGFR inhibitors as anticancer drugs. Superimposed the crystal structure of a sulfoamide ligand bound to a hepatocyte growth factor receptor (c-Met) (PDB ID: 2WDI) to the structure of compound 8 bound to FGFR1. The protein was depicted in cartoon style, while the inhibitors and interacted residues were shown as sticks.

Results and Discussion
The biological activity of compounds (5, 6, 7, 8) as c-Met inhibitors was reported in our previous work [20], the compounds with the scaffold of 1-sulfonylpyrazolo [4,3-b]pyridines showed definite FGFR1 activity at 10 µM, all with a more than 50% inhibition ratio towards FGFR1 ( Table 1). The X-ray crystal structure of compound 8 with FGFR1 revealed that the pyrazolo[4,3-b]pyridine scaffold forms a hydrogen bond, with the backbone of residue Ala564 at the hinge part. The P-loop of FGFR1 lowered down to form a pi-pi stacking interaction between residue Phe489 and the imidazo[1,2-b]pyridazine ring. Although these interactions similarly existed in the co-crystal structure of c-Met bound with inhibitors (PDB ID: 2WDI), the most striking difference is that the overall binding conformation of inhibitor 8 is inversed, giving that the methylpyrazole pointed towards the back pocket of the ATP site of FGFR1, which can be immediately noticed from the superimposed structures, as shown in Figure 2B. Due to the novel binding interactions of compound 8, we thought it should be possible to selectively optimize the FGFR activity of this series of compounds. Based on our previous study, changing the scaffold of 1H-pyrazolo [4,3-b]pyridine to 5H-pyrrolo [2,3-b]pyrazine can increase the binding activity of FGFR1 [21]; therefore, we synthesized the compounds 9 and 10. As listed in Table 2, the activity of compounds 9 and 10 has increased slightly compared with compounds 5 and 6, respectively. In addition, the scaffold of 5H-pyrrolo [2,3-b]pyrazine also has been reported to have Bruton's tyrosine kinase, focal adhesion kinase, JAK3, ataxia telangiectasia and Rad3-related protein (ATR), and serine/threonine kinase activity [22][23][24][25][26][27]. Therefore, we selected 5H-pyrrolo[2,3-b]pyrazine as the starting scaffold for further modification.   It is apparently noticed from compounds 5, 7, 8, and 9 that they all contained an imidazole ring that formed the π-π interaction with the residue Phe489. Therefore, we attempted to simplify the bicycle to imidazole so that compound 11 was synthesized, which maintained the activity against FGFR1, as shown in Table 2. Furthermore, substituting the methyl of imidazole to 2-fluoroethyl (12) could slightly increase the inhibitory activity. Based on the advantage of synthesis, we selected compound 12 as the starting point for further modification.
Based on the X-ray structure of compound 8 bound to FGFR1, the 1-methyl-1H-pyrazole of compound 8 extended to the inside of the ATP site of FGFR1, where there are a large number of polar residues and a corresponding cavity for optimization. In order to investigate the structure-activity relationship (SAR) around the methylpyrazole part, we synthesized 13-26 by substituting the methyl group with different groups, and their FGFR1 enzymatic activities are summarized in Table 3. Interestingly, the compound 13, which bears an unsubstituted pyrazole ring, showed much higher activity than the others, and even at 10 nM concentration, it still has a more than 90% inhibition ratio (IC50 = 0.6 nM, as shown in Table 5). When isopropyl was induced to the structure, the compound (14) lost the FGFR1 activity. Compounds 15-17, containing a carbonyl group as hydrogen bond receptors, have moderate activity. Meanwhile, for the compounds with carbonyl groups that are deeper into the protein cavity (18)(19)(20)(21)(22), neither ester (19,20) nor acylamido (21,22) showed weaker activity than compound 15. Compared with the carbonyl group, the mesyl (23) caused a slight increase in inhibitory activity; however, replacing the mesyl with a larger cyclopropylsulfonyl group (24) decreased the inhibitory activity. Doubling the carbonyl group (25) showed stronger activity than monocarbonyl (15), which may be due to forming more hydrogen bonds. In summary, all of the attempts at optimizing the substitutes on pyrazole couldn't significantly improve the activity of FGFR1; therefore, we further carried out the optimization by focusing on compound 13. It is apparently noticed from compounds 5, 7, 8, and 9 that they all contained an imidazole ring that formed the π-π interaction with the residue Phe489. Therefore, we attempted to simplify the bicycle to imidazole so that compound 11 was synthesized, which maintained the activity against FGFR1, as shown in Table 2. Furthermore, substituting the methyl of imidazole to 2-fluoroethyl (12) could slightly increase the inhibitory activity. Based on the advantage of synthesis, we selected compound 12 as the starting point for further modification.
Based on the X-ray structure of compound 8 bound to FGFR1, the 1-methyl-1H-pyrazole of compound 8 extended to the inside of the ATP site of FGFR1, where there are a large number of polar residues and a corresponding cavity for optimization. In order to investigate the structure-activity relationship (SAR) around the methylpyrazole part, we synthesized 13-26 by substituting the methyl group with different groups, and their FGFR1 enzymatic activities are summarized in Table 3. Interestingly, the compound 13, which bears an unsubstituted pyrazole ring, showed much higher activity than the others, and even at 10 nM concentration, it still has a more than 90% inhibition ratio (IC50 = 0.6 nM, as shown in Table 5). When isopropyl was induced to the structure, the compound (14) lost the FGFR1 activity. Compounds 15-17, containing a carbonyl group as hydrogen bond receptors, have moderate activity. Meanwhile, for the compounds with carbonyl groups that are deeper into the protein cavity (18)(19)(20)(21)(22), neither ester (19,20) nor acylamido (21,22) showed weaker activity than compound 15. Compared with the carbonyl group, the mesyl (23) caused a slight increase in inhibitory activity; however, replacing the mesyl with a larger cyclopropylsulfonyl group (24) decreased the inhibitory activity. Doubling the carbonyl group (25) showed stronger activity than monocarbonyl (15), which may be due to forming more hydrogen bonds. In summary, all of the attempts at optimizing the substitutes on pyrazole couldn't significantly improve the activity of FGFR1; therefore, we further carried out the optimization by focusing on compound 13. It is apparently noticed from compounds 5, 7, 8, and 9 that they all contained an imidazole ring that formed the π-π interaction with the residue Phe489. Therefore, we attempted to simplify the bicycle to imidazole so that compound 11 was synthesized, which maintained the activity against FGFR1, as shown in Table 2. Furthermore, substituting the methyl of imidazole to 2-fluoroethyl (12) could slightly increase the inhibitory activity. Based on the advantage of synthesis, we selected compound 12 as the starting point for further modification.
Based on the X-ray structure of compound 8 bound to FGFR1, the 1-methyl-1H-pyrazole of compound 8 extended to the inside of the ATP site of FGFR1, where there are a large number of polar residues and a corresponding cavity for optimization. In order to investigate the structure-activity relationship (SAR) around the methylpyrazole part, we synthesized 13-26 by substituting the methyl group with different groups, and their FGFR1 enzymatic activities are summarized in Table 3. Interestingly, the compound 13, which bears an unsubstituted pyrazole ring, showed much higher activity than the others, and even at 10 nM concentration, it still has a more than 90% inhibition ratio (IC 50 = 0.6 nM, as shown in Table 5). When isopropyl was induced to the structure, the compound (14) lost the FGFR1 activity. Compounds 15-17, containing a carbonyl group as hydrogen bond receptors, have moderate activity. Meanwhile, for the compounds with carbonyl groups that are deeper into the protein cavity (18)(19)(20)(21)(22), neither ester (19,20) nor acylamido (21,22) showed weaker activity than compound 15. Compared with the carbonyl group, the mesyl (23) caused a slight increase in inhibitory activity; however, replacing the mesyl with a larger cyclopropylsulfonyl group (24) decreased the inhibitory activity. Doubling the carbonyl group (25) showed stronger activity than monocarbonyl (15), which may be due to forming more hydrogen bonds. In summary, all of the attempts at optimizing the substitutes on pyrazole couldn't significantly improve the activity of FGFR1; therefore, we further carried out the optimization by focusing on compound 13. Table 3. The FGFR1 enzymatic activity of compounds 13-26. a (23) caused a slight increase in inhibitory activity; however, replacing the mesyl with a larger cyclopropylsulfonyl group (24) decreased the inhibitory activity. Doubling the carbonyl group (25) showed stronger activity than monocarbonyl (15), which may be due to forming more hydrogen bonds. In summary, all of the attempts at optimizing the substitutes on pyrazole couldn't significantly improve the activity of FGFR1; therefore, we further carried out the optimization by focusing on compound 13. (23) caused a slight increase in inhibitory activity; however, replacing the mesyl with a larger cyclopropylsulfonyl group (24) decreased the inhibitory activity. Doubling the carbonyl group (25) showed stronger activity than monocarbonyl (15), which may be due to forming more hydrogen bonds. In summary, all of the attempts at optimizing the substitutes on pyrazole couldn't significantly improve the activity of FGFR1; therefore, we further carried out the optimization by focusing on compound 13. (23) caused a slight increase in inhibitory activity; however, replacing the mesyl with a larger cyclopropylsulfonyl group (24) decreased the inhibitory activity. Doubling the carbonyl group (25) showed stronger activity than monocarbonyl (15), which may be due to forming more hydrogen bonds. In summary, all of the attempts at optimizing the substitutes on pyrazole couldn't significantly improve the activity of FGFR1; therefore, we further carried out the optimization by focusing on compound 13. To find the mechanism of this significant activity improvement of compound 13, we adopted the docking method to predict the binding interactions between compound 13 and FGFR1. As illustrated in Figure 3, compound 13 bound to FGFR1 with a very similar mode to how compound 8 bound, by using the nitrogen atom in the pyrazine ring to form an essential hydrogen bond with the FGFR1 hinge and using imidazole stacking with the residue Phe489. Besides, compound 13 also To find the mechanism of this significant activity improvement of compound 13, we adopted the docking method to predict the binding interactions between compound 13 and FGFR1. As illustrated in Figure 3, compound 13 bound to FGFR1 with a very similar mode to how compound 8 bound, by using the nitrogen atom in the pyrazine ring to form an essential hydrogen bond with the FGFR1 hinge and using imidazole stacking with the residue Phe489. Besides, compound 13 also 65. 6 14.9 12.5

22
Molecules 2018, 23 To find the mechanism of this significant activity improvement of compound 13, we adopted the docking method to predict the binding interactions between compound 13 and FGFR1. As illustrated in Figure 3, compound 13 bound to FGFR1 with a very similar mode to how compound 8 bound, by using the nitrogen atom in the pyrazine ring to form an essential hydrogen bond with the FGFR1 hinge and using imidazole stacking with the residue Phe489. Besides, compound 13 also 15 To find the mechanism of this significant activity improvement of compound 13, we adopted the docking method to predict the binding interactions between compound 13 and FGFR1. As illustrated in Figure 3, compound 13 bound to FGFR1 with a very similar mode to how compound 8 bound, by using the nitrogen atom in the pyrazine ring to form an essential hydrogen bond with the FGFR1 hinge and using imidazole stacking with the residue Phe489. Besides, compound 13 also formed a salt bridge interaction between the pyrazole and Asp641. We assumed this interaction To find the mechanism of this significant activity improvement of compound 13, we adopted the docking method to predict the binding interactions between compound 13 and FGFR1. As illustrated in Figure 3, compound 13 bound to FGFR1 with a very similar mode to how compound 8 bound, by using the nitrogen atom in the pyrazine ring to form an essential hydrogen bond with the FGFR1 hinge and using imidazole stacking with the residue Phe489. Besides, compound 13 also formed a salt bridge interaction between the pyrazole and Asp641. We assumed this interaction 29. 5 13.5 16 To find the mechanism of this significant activity improvement of compound 13, we adopted the docking method to predict the binding interactions between compound 13 and FGFR1. As illustrated in Figure 3, compound 13 bound to FGFR1 with a very similar mode to how compound 8 bound, by using the nitrogen atom in the pyrazine ring to form an essential hydrogen bond with the FGFR1 hinge and using imidazole stacking with the residue Phe489. Besides, compound 13 also formed a salt bridge interaction between the pyrazole and Asp641. We assumed this interaction  To find the mechanism of this significant activity improvement of compound 13, we adopted the docking method to predict the binding interactions between compound 13 and FGFR1. As illustrated in Figure 3, compound 13 bound to FGFR1 with a very similar mode to how compound 8 bound, by using the nitrogen atom in the pyrazine ring to form an essential hydrogen bond with the FGFR1 hinge and using imidazole stacking with the residue Phe489. Besides, compound 13 also formed a salt bridge interaction between the pyrazole and Asp641. We assumed this interaction dramatically increased the binding affinity of compound 13. To find the mechanism of this significant activity improvement of compound 13, we adopted the docking method to predict the binding interactions between compound 13 and FGFR1. As illustrated in Figure 3, compound 13 bound to FGFR1 with a very similar mode to how compound 8 bound, by using the nitrogen atom in the pyrazine ring to form an essential hydrogen bond with the FGFR1 hinge and using imidazole stacking with the residue Phe489. Besides, compound 13 also formed a salt bridge interaction between the pyrazole and Asp641. We assumed this interaction dramatically increased the binding affinity of compound 13.
We further optimized the right side of compound 13. The designed and synthesized compounds 27-35 with modifications on the imidazole moiety of compound 13 were assessed for their FGFR1 inhibitory activities (Table 4). Firstly, substituents from small to large groups, namely methyl (28), ethyl (29), isopropyl (30), isobutyl (31), and cyclopentylmethyl (32), were induced to imidazole. The results indicated that the substitution of the imidazole ring with ethyl (29, FGFR1 IC 50 = 3.0 nM) and isopropyl (30, FGFR1 IC 50 = 3.0 nM) showed better inhibitory activity among the compounds (27)(28)(29)(30)(31)(32). Incorporation of the methylacetamide (33) group reduced the enzymatic potency. The activity of compounds 34 and 35 showed that the bicycles were weaker than the imidazole. potency. The activity of compounds 34 and 35 showed that the bicycles were weaker than the imidazole.  Based on the structure diversity and FGFR1 enzymatic inhibition ratio, we selected four compounds to obtain the accurate IC50 values and assess their antiproliferation activity In KG1 cells. As listed in Table 5, compounds 13 and 29 showed excellent biological activities. Since this series of compounds stemmed from our c-Met inhibitors, compound 13 was selected as the representative to potency. The activity of compounds 34 and 35 showed that the bicycles were weaker than the imidazole.  Based on the structure diversity and FGFR1 enzymatic inhibition ratio, we selected four compounds to obtain the accurate IC50 values and assess their antiproliferation activity In KG1 cells. As listed in Table 5, compounds 13 and 29 showed excellent biological activities. Since this series of compounds stemmed from our c-Met inhibitors, compound 13 was selected as the representative to potency. The activity of compounds 34 and 35 showed that the bicycles were weaker than the imidazole.  Based on the structure diversity and FGFR1 enzymatic inhibition ratio, we selected four compounds to obtain the accurate IC50 values and assess their antiproliferation activity In KG1 cells. As listed in Table 5, compounds 13 and 29 showed excellent biological activities. Since this series of compounds stemmed from our c-Met inhibitors, compound 13 was selected as the representative to potency. The activity of compounds 34 and 35 showed that the bicycles were weaker than the imidazole.  Based on the structure diversity and FGFR1 enzymatic inhibition ratio, we selected four compounds to obtain the accurate IC50 values and assess their antiproliferation activity In KG1 cells. As listed in Table 5, compounds 13 and 29 showed excellent biological activities. Since this series of compounds stemmed from our c-Met inhibitors, compound 13 was selected as the representative to potency. The activity of compounds 34 and 35 showed that the bicycles were weaker than the imidazole.  Based on the structure diversity and FGFR1 enzymatic inhibition ratio, we selected four compounds to obtain the accurate IC50 values and assess their antiproliferation activity In KG1 cells. As listed in Table 5, compounds 13 and 29 showed excellent biological activities. Since this series of compounds stemmed from our c-Met inhibitors, compound 13 was selected as the representative to potency. The activity of compounds 34 and 35 showed that the bicycles were weaker than the imidazole.  Based on the structure diversity and FGFR1 enzymatic inhibition ratio, we selected four compounds to obtain the accurate IC50 values and assess their antiproliferation activity In KG1 cells. As listed in Table 5, compounds 13 and 29 showed excellent biological activities. Since this series of compounds stemmed from our c-Met inhibitors, compound 13 was selected as the representative to potency. The activity of compounds 34 and 35 showed that the bicycles were weaker than the imidazole.  Based on the structure diversity and FGFR1 enzymatic inhibition ratio, we selected four compounds to obtain the accurate IC50 values and assess their antiproliferation activity In KG1 cells. As listed in Table 5, compounds 13 and 29 showed excellent biological activities. Since this series of compounds stemmed from our c-Met inhibitors, compound 13 was selected as the representative to potency. The activity of compounds 34 and 35 showed that the bicycles were weaker than the imidazole.  Based on the structure diversity and FGFR1 enzymatic inhibition ratio, we selected four compounds to obtain the accurate IC50 values and assess their antiproliferation activity In KG1 cells. As listed in Table 5, compounds 13 and 29 showed excellent biological activities. Since this series of compounds stemmed from our c-Met inhibitors, compound 13 was selected as the representative to Based on the structure diversity and FGFR1 enzymatic inhibition ratio, we selected four compounds to obtain the accurate IC 50 values and assess their antiproliferation activity In KG1 cells. As listed in Table 5, compounds 13 and 29 showed excellent biological activities. Since this series of compounds stemmed from our c-Met inhibitors, compound 13 was selected as the representative to test 17 kinases' (include c-Met) enzymatic activity. As shown in Table 6, compound 13 demonstrated nearly no inhibition towards c-Met, even at a concentration of 1 µM, indicating that compound 13 may be a selective FGFR inhibitor.   To further assess the druggability of compound 13, it was further subjected to in vitro metabolic stability assays and an in vivo pharmacokinetic test. Compound 13 showed excellent metabolic stability, as it has a low clearance ratio of 1 µL/min/mg in the human liver microsome. In addition to the liver microsome assay, five human cytochrome P450 (CYP) enzymes commonly metabolizing exogenous chemicals were used to test the direct inhibition of compound 13. As shown in Table 7, the compound showed favorable metabolic properties, as the inhibition ratio for five CYPs were less than 50%, even at the compound concentration of 10 µM.   Compounds 27-33 were prepared according to the procedure in Scheme 3. Compounds 44-50 were also generated by a substituent reaction from compound 42; then, 1H-pyrazole-4-yl was introduced at the 3-position of the corresponding compounds (44-50) via a Suzuki coupling reaction to provide compounds 27-33. Scheme 2. Reagents and conditions: (a) NaH, Dimethylformamide, r.t., 4 h, 84% yield; (b) NaH, Dimethylformamide, 1-fluoro-2-iodoethane, 80 °C, 4 h, 90% yield; (c) 4-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)-1H-pyrazole, K2CO3, Pd(PPh3)Cl2, Dioxane: H2O (v:v = 3:1), 80 °C, 4 h, 87% yield; (d) K2CO3, Dimethylformamide, r.t., 8 h, 70-89% yield.

In Vitro Metabolic Stability Study
Microsomes (Human microsome: Xenotech, Lot No.H0610; Rat microsome: Xenotech, Lot No. R1000) (0.5 mg/mL) were preincubated with 1 µM of test compound for 5 min at 37 • C in 0.1 M phosphate buffer (pH 7.4) with 1 mmol Ethylenediaminetetraacetic acid (EDTA), and 5 mmol MgCl 2 . The reactions were initiated by adding prewarmed cofactors (1 mmol NADPH). After 0-min, 5-min, 10-min, and 30-min incubations at 37 • C, the reactions were stopped by adding an equal volume of cold acetonitrile. The samples were vortexed for 10 min, and then centrifuged at 10,000× g for 10 min. Supernatants were analyzed by LC/MS/MS for the amount of the parent compound remaining, and the corresponding loss of the parent compound was also determined by LC/MS/MS.
The Cytochromes P450 (CYP) enzymatic activities were characterized based on their probe reactions: CYP3A4 (midazolam), CYP2D6 (dextromethorphan), CYP2C9 (Diclofenac), CYP1A2 (phenacetin) and CYP2C19 (mephenytoin). Incubation mixtures were prepared in a total volume of 100 µL as follows: 0.2 mg/mL of microsome (Human microsome: Xenotech, Lot No.H0610), 1 mmol of NADPH, 100 mmol of phosphate buffer (pH 7.4), probe substrates cocktail (10 µM of midazolam, 100 µM of testosterone, 10 µM of dextromethophan, 20 µM of diclofenac, 100 µM of phenacetin, 100 µM of mephenytoin), and 10 µM of the tested compound or positive control cocktail (10 µM of ketoconazole, 10 µM of quinidine, 100 µM of sulfaphenazole, 10 µM of naphthoflavone, and 1000 µM of tranylcypromine) or negative control (PBS). The final concentration of the organic reagent in the incubation mixtures was less than 1% v/v. There was a 5-min preincubation period at 37 • C before the reaction was initiated by adding a nicotinamide adenine dinucleotide phosphate (NADPH)-generating system. Reactions were conducted for 20 min for CYPs. For each probe drug, the percentage of the metabolite conversion was less than 20% of the substrate added. The inhibition rate was calculated as: (the formation of the metabolite of probe substrates with 10 µM of the tested compound)/(the formation of the metabolite of the probe substrates with PBS) × 100%.

Conclusions
Starting with the scaffold of 1-sulfonylpyrazolo [4,3-b]pyridine discovered in c-Met inhibitor development, we developed a series of new pyrrolo[2,3-b]pyrazine analogues. Based off of the X-ray crystal structure of compound 8 bound with FGFR1, we found that it showed novel binding conformation, which significantly differed to c-Met ligand binding. Extensive structure-activity relationships were conducted, and compound 13 was identified as possessing high enzymatic and cellular potency against FGFR1. The activity difference between FGFR and c-Met indicated that the series of inhibitors might be selective FGFR inhibitors. Moreover, through in vitro metabolic stability study, compound 13 shown favorable metabolic properties. All of these data indicated that compound 13 would be a promising lead compound for further drug development.