Construction of a synthetic methodology-based library and its application in identifying a GIT/PIX protein-protein interaction inhibitor

Abstract

In recent years, the flourishing of synthetic methodology studies has provided concise access to numerous molecules with new chemical space. These compounds form a large library with unique scaffolds, but their application in hit discovery is not systematically evaluated. In this work, we establish a synthetic methodology-based compound library (SMBL), integrated with compounds obtained from our synthetic researches, as well as their virtual derivatives in significantly larger scale. We screen the library and identify small-molecule inhibitors to interrupt the protein-protein interaction (PPI) of GIT1/β-Pix complex, an unrevealed target involved in gastric cancer metastasis. The inhibitor 14-5-18 with a spiro[bicyclo[2.2.1]heptane-2,3'-indolin]-2'-one scaffold, considerably retards gastric cancer metastasis in vitro and in vivo. Since the PPI targets are considered undruggable as they are hard to target, the successful application illustrates the structural specificity of SMBL, demonstrating its potential to be utilized as compound source for more challenging targets.

Fig. 1. Construction and analysis of SMBL.

a Representative scaffold of the entity library. The spheres show different substituents. b Derivation of the existing compounds to construct the virtual library. Two compounds from one paper (Chin. J. Chem. 2012, 30, 2669) were taken as examples to show the processes. The spheres show the derivable sites. c, d Similarity range between the SMBL and commercial libraries.

Fig. 2. Screening methods to identify 14-5-18 and 15-4-26 as inhibitory molecules for GIT1/β-Pix interaction.

a Co-immunoprecipitation (co-IP) experiment to detect the interaction between β-Pix and GIT1 in both the wild-type or mutant forms. HA-GIT1 (wide-type and mutant forms) and Flag-β-Pix were transfected into HEK-293T and the cell lysates were collected to perform the co-IP assay. The demonstrated blots were representatives from three independent experiments with similar results. b Transwell invasion assay to detect the invasion abilities of MKN45 cells with GIT1 wide-type or mutants (n = 3 biologically independent samples). Scale bar = 100 µm. c Cartoon model showing the interaction of GIT1 (magenta) and the key peptide β-Pix (cyan). Key residues Leu271 and Leu279 were indicted in sticks (upper), and the pocket was generated for binding (down). d Schematic diagram of fluorescence polarization (FP) assay performed between GIT1 protein and FITC-β-Pix (5-FITC-Acp-ALEEDAQILKVIEAYCTSAKT). e Workflow of screening in both entity and virtual libraries to identify 14-5-18 and 15-4-26 as potential inhibitors. The Ki values were calculated according to IC₅₀ values of balance in the FP assay. Data are presented as mean values ± SD and error bars indicate SD. Statistical analysis: One-way ANOVA, Tukey’s multiple comparisons tests, each group was compared with the other group. Source data are provided as a Source Data file.

Fig. 3. Compound 14-5-18 interrupted the GIT1/β-Pix interaction.

a Kinetic measurement of the binding affinity of 14-5-18 to GIT1 by BLI assay. Concentrations ranging from 25 to 400 μM were shown with the real-time response for each step of the kinetic assay. b Co-IP assay to indicate the interaction of GIT1 and β-Pix and the inhibitory effects of 14-5-18. HA-GIT1 and Flag-β-Pix were transfected into HEK-293T cells, and the cell lysates were collected to perform the co-IP assay at concentrations of 0, 5, 20, and 50 μM. The demonstrated blots were representatives from three independent experiments with similar results. c Fluorescence colocalization experiment to visualize the cell distribution of GIT1 and β-Pix with or without treatment of 14-5-18. GFP-GIT1 (green) and Myc-β-Pix were expressed followed by treatment with a fluorescent secondary antibody (red) in the MGC803 cell line to construct the fluorescence system. The cells were observed under confocal microscopy after 14-5-18 was added at concentrations of 0 and 20 μM. Plot fluorescence intensities of the area marked by the white lines in the left panels are shown in the right graphs. Green and red curves represent the GIT1 and β-Pix, respectively. Scale bar = 10 μm. The demonstrated figures were representatives from three independent experiments with similar results. d The percentage of GIT1/β-Pix colocalized cells with or without treatment of 14-5-18. The analysis was performed in n = 30 cells from three independent experiments. Statistical analysis: Two-sided Fisher’s exact test (d). Source data are provided as a Source Data file.

Fig. 4. Compound 14-5-18 inhibited invasion of gastric cancer cells in vitro and in vivo.

a Transwell assays were employed to detect the anti-invasion abilities of 14-5-18 (10 and 50 μM) in the MGC803 (left) and MKN45 (right) cell lines for 24 h. The right columns show quantification, n = 3 biologically independent samples. Scale bar = 100 µm. b Expression levels of total Rac1/Cdc42 and the activated forms GTP-Rac1/Cdc42 were detected after 14-5-18 was added to MKN45 and MGC803 at concentrations of 5, 20 and 50 μM for 24 h, n = 3 biologically independent samples. c In vivo imaging of nude mice after intravenous injection of luciferase-expressing MGC803 cells. Compound 14-5-18 were administrated at the indicated dosages by gavage once a day for 18 days (n = 6 biologically independent animals), and solvent DMSO (2%) + Tween 20 (5%) in water was used as control. Otherwise noted, DMSO was used as solvent control. Data are presented as mean values ± SD, and error bars indicate SD. Statistical analysis: One-way ANOVA, Dunnett’s multiple-comparisons test (a–c, each group was compared with the control group). Source data are provided as a Source Data file.

Fig. 5. Compound 14-5-18 targets the GIT1/β-Pix interaction at GIT1 residues Leu271 and Leu279.

a Cartoon model showing the binding pose of GIT1 (magenta) and 14-5-18 (cyan). Magnification of the binding pocket (middle) indicates the key residues (in lines) adjacent to 14-5-18 (in sticks). The graph (right) shows the energy decomposition of each adjacent residue. b Transwell invasion assay to detect the invasion abilities of MKN45 cells with wide-type or mutant GIT1, with or without 14-5-18 (50 µM) for 24 h (n = 3 biologically independent samples). Scale bar = 100 µm. c The cellular activity of Rac1 in MKN45 cells with wide-type or mutant GIT1 with or without treatment of 14-5-18 (50 µM) for 24 h (n = 3 biological independent samples). d The graphic abstract. Otherwise noted, DMSO was used as solvent control. Data are presented as mean values ± SD, and error bars indicate SD. Statistical analysis: unpaired two-tailed t-tests. Source data are provided as a Source Data file.