Novel Inhibitor for Fibroblast Growth Factor Receptor Tyrosine Kinase

NP603, the 6-dimethoxy phenyl indolin-2-one, was designed as an FGF receptor 1 inhibitor through computational study. NP603 was synthesized and found to be more active against endothelial proliferation of HUVEC after the rhFGF-2 stimulation than SU6668, with a minimum effective dose of 0.4 μM but with similar potency as SU16g. NP603 inhibited the tyrosine phosphorylation in FGF receptor and the activation of extracellular signal-regulated kinase and c-Jun-N-terminal kinase after the rhFGF-2 stimulation. The increase in activity of NP603 supports the role of Lys514 movement in ligand–receptor binding in modeling study as the movement accommodates the hydrophobic interaction at the receptor pocket, leading to the enhancement of binding capacity.

Receptor tyrosine kinases (RTKs) have been shown to be important mediators of signal transduction in cells. These transmembrane molecules characteristically consist of an extracellular ligand-binding domain connected through a segment in the plasma membrane to an intracellular tyrosine kinase domain. Binding of the ligand to the receptor results in receptor dimerization and stimulation of the receptor-associated tyrosine kinase activity, which leads to phosphorylation of tyrosine residues on both the receptor and other intracellular molecules. These changes in tyrosine phosphorylation initiate a signaling cascade leading to a variety of cellular responses, one of which is angiogenesis. Growth factors, including the vascular endothelial growth factors (VEGF), the fibroblast growth factors (FGF), and the platelet-derived growth factors (PDGF) and their associated receptor tyrosine kinases, are major regulators of angiogenesis.

RTKs are important therapeutic targets for cancer drug discovery and development. Among the many inhibitors undergoing development, 3-substituted indolin-2-ones were found to exhibit high potency and selectivity against PDGF and VEGF (Flk-1) RTKs. SU5402 and SU5416 have been shown to be specific inhibitors of the kinase activity of the fibroblast growth factor receptor (FGFR) and vascular endothelial growth factor receptor (VEGFR), respectively, whereas SU6668 showed inhibitory activity against FGFR, VEGFR, and PDGFR. The pyrido[2,3-d]pyrimidine PD173074, a potential drug candidate, was similarly found to be highly selective for FGFR1. This molecule is a nanomolar inhibitor of FGFR1 and is also a submicromolar inhibitor of VEGFR2.

Structure–activity relationship (SAR) data for compounds in the pyrido[2,3-d]pyrimidine series indicate that binding affinity and selectivity for FGFR1 is attained through modification of the phenyl group attached to the 6-position of the pyrido[2,3-d]pyrimidine scaffold. In general, derivatization at the 3- and 5-positions of the phenyl ring increases selectivity for FGFR1, especially when the attached group is larger than a methyl. The SAR data are in good agreement with the surface observed in the structure for PD173074 bound in the FGFR1 kinase domain.

Through superposition of the SU5402-FGFR1 structure (1FGI) with 2FGI it was determined that the conserved Lys514 undergoes significant movement upon PD173074 binding. The side chain of Lys514 adopts an alternative conformation which allows the dimethoxy phenyl substituent of PD173074 access to the back of the ATP-binding cleft.

On comparison of indolinones such as SU4984, SU5402, and SU6668 to PD173074, certain structural similarities may be noticed, in particular, a NH (hydrogen donor) and an aromatic group. SAR studies on the indolin-2-one series suggested that modifications at the C-6 position would lead to compounds with different kinase inhibition profiles for VEGFR2 (Flk-1) and FGFR1. However, structural information from the protein binding study of PD173074 indicated that the contribution of the dimethoxy phenyl group to tight binding in FGFR1 is important. Therefore, structural modification of the indolin-2-ones to include this dimethoxy phenyl motif at position 6 should, in principle, lead to a potent FGFR1 inhibitor.

This reasoning suggested that the novel C-6 substituted oxindole analog (NP603) may be an interesting probe in modeling studies to ascertain the role of Lys514 movement in ligand–receptor binding. In this study, the proposed compound NP603 was flexibly docked to the model of the constructed FGFR1 receptor template. The goal was to study the effects of the added substituent at the 6-position on the structure and the orientation of the ligand–receptor complex. In addition to docking studies, NP603 was synthesized and evaluated for its activity.

The crystal structure of FGFR1 bound to the inhibitor PD173074 (Protein Data Bank code: 2FGI) was selected for the construction of the docking template. In order to prepare the target protein as a template, the ligand and crystallographic water were removed. Hydrogens and Kollman unified charges were added using AutoDockTools. Affinity grid maps for each atom type in the ligand set (plus an electrostatics map) centered on the cavity were computed using AutoGrid. The constructed FGFR1 template was validated by redocking with the native ligand, PD173074. AutoDock was employed to perform the docking calculation. All ligands were created and initially optimized, charges were assigned, and ligand energies minimized. During final preparations, charges were added, hydrogens merged, and rotatable bonds defined. PD173074 docked in the binding pocket in the conformation found in the crystal structure, with RMSD of 1.6 Å.

To study the binding mode, SU6668, NP603, and related analogs were docked with the validated FGFR1 template. The binding cleft for the ATP domain of FGFR1 was confirmed as the target area for the inhibitors. The binding mode of NP603, designed by substitution of dimethoxy phenyl group at the 6-position of indolin-2-one to enhance binding via hydrophobic interaction, was verified. The docking energy of NP603 was −13.3 kcal/mol, the lowest among the inhibitors studied.

NP603’s orientation showed that the substituted moiety at position 6 located near to the active binding site of the substrate. This molecule makes three hydrogen bonds to the protein backbone of FGFR1: between N-1 of the oxindole and the carbonyl oxygen of Glu562, between the carbonyl oxygen of the oxindole and the amide nitrogen of Ala564, and between N-1 of the pyrrole and the carbonyl oxygen of Ala564. Hydrophobic residues Val559 and Val561 are involved in interaction with the 3,5-dimethoxy group of NP603. The oxygen atoms of the methoxy groups in NP603 contribute two more hydrogen bonds. van der Waals contacts are made between the dimethoxy phenyl group and the side chains of multiple residues, including Lys514.

According to the docking model of SU6668 (which lacks the C-6 substitution), two hydrogen bonds were formed between the N and O of the oxindole and Ala564. However, the docked conformation was flipped compared with the crystal structure of SU5402. The FGFR1 template used was developed from PD173074 structure with alternative conformation involving Lys514 movement to accommodate hydrophobic interaction.

As NP603 demonstrated good binding with FGFR1 in silico, NP603 and related compounds SU6668, SU16f, and SU16g were synthesized and evaluated for their activity in vitro. NP603 was synthesized through multistep synthesis involving Suzuki coupling, displacement reactions, hydrolysis, and condensation. The structure of the synthesized compound was confirmed by analytical techniques.

The in vitro kinase assay of NP603 showed an IC50 against FGFR1 of 0.4 μM, three times less than SU16g. All tested compounds inhibited rhFGF-2-induced proliferation of HUVEC. NP603 was the most effective, with SU16g also showing strong activity. The inhibitory potency correlated with docking energy. NP603’s cytotoxic concentration in the absence of rhFGF-2 was found to be 50 μM, indicating selective activity.

Western blot analysis demonstrated that NP603 inhibited FGF-2-induced activation of ERK and JNK. It showed potent inhibitory activity on the FGF-2-induced tyrosine phosphorylation of FGF receptor, leading to inhibition of downstream signaling and cellular proliferation.

In angiogenesis assays, NP603 suppressed HUVEC tube formation in a dose-dependent manner. Treated cells formed incomplete, narrow structures compared to elongated robust tubes in FGF-2-treated controls. NP603 inhibited tube formation more effectively than SU6668.

The activity observed supports the role of Lys514 movement in ligand–receptor binding, allowing the 3,5-dimethoxyphenyl group at C-6 position in NP603 to locate in the receptor pocket, enhancing binding. The dimethoxyphenyl moiety increased inhibitory potency as predicted. Analog compounds mimicking C-6 substitution (L1–L8) showed similar or improved docking energies. L4 and L8, in particular, showed better interaction due to resonance and inductive effects, with L8 showing the lowest docking energy.

Thus, NP603 is a potent FGFR1 inhibitor whose design validates computational modeling predictions regarding Lys514 and C-6 substitution. Further evaluation of Orantinib its analogs is warranted.