Drug Screening of Flavonoids as Potential VEGF Inhibitors Through Computational Docking and Cell Models

Abstract

Vascular endothelial growth factor (VEGF), also known as VEGF-A, has been linked to various diseases, such as wet age-related macular degeneration (wAMD) and cancer. Even though there are VEGF inhibitors that are currently commercially available in clinical applications, severe adverse effects have been associated with these treatments. There is still a need to develop novel VEGF-based therapeutics against these VEGF-related diseases. Here, we established a series of VEGF-based computational docking analyses and cell models, such as a wound healing assay in HaCaT cells and an evaluation of NF-κB performance in macrophages, to screen a large library of flavonoid-type phytochemicals. Three flavonoids, namely, farrerol, ononin and (-)-epicatechin, were shown to express binding affinities to VEGF protein and inhibit VEGF-mediated biological activities. The investigation evidently suggested that the three flavonoids above could be considered potential anti-VEGF agents for the following drug development against VEGF-mediated diseases.

Keywords: VEGF inhibitor; angiogenesis; computational docking; drug screening; flavonoids.

Figure 1

Computational docking analysis of farrerol, ononin and (−)-epicatechin against VEGF protein. Protein structure was downloaded from the Protein Data Bank (PDB code: 1FLT), and chemical structures were obtained from Chemdraw software v. 20.0. Axitinib was used as a positive control, and the binding energy was −17.4 kJ/mol.

Figure 2

(A) Farrerol, ononin and (−)-epicatechin showed cell absorption in Caco-2 cells at concentrations of 1 µM and 10 µM. Caco-2 cells were cultured until fully differentiated after being cultured for 21 days, and the integrity of cell monolayer was determined by the transepithelial electrical resistance (TEER). * p < 0.05. Binding degrees of farrerol (B), ononin (C) and (−)-epicatechin (D) were calculated through the following equation: binding degree = (C_pre − C_post)/C_pre, where C_pre was the initial concentration of the tested ligand and C_post was the unbound chemical in the filtrate. Red curve: initial concentration of chemicals; black curve: concentration of chemicals binding to the protein (C_pre − C_post). n = 4.

Figure 3

Wound healing evaluation of farrerol, ononin and (−)-epicatechin. Photos of each cell scrape at 0 (At₀) and 20 h (At₂₀) were captured with microscope and imaging software (Zen, https://www.zeiss.com/microscopy/en/home.html, accessed on 1 September 2024) at a 10× magnification. The data indicate the mean ± SD (n = 4) fold change compared with the blank group, and the asterisks represent statistically significant differences such that * p < 0.05 and ** p < 0.01 compared with the blank group. A: positive control, Avastin; F: farrerol; O: ononin; E: (−)-epicatechin.

Figure 4

Farrerol, ononin and (−)-epicatechin attenuated VEGF-induced NF-κB activities. The pNF-κB constructs were transported by jetPRIME (Polyplus, Illkirch-Graffenstaden, France) reagent on the fourth day, and the chemicals were tested at concentrations of 1 µM, 10 µM and 25 µM. A: avastin (10 ng/mL). n = 4. * p < 0.05, compared with the blank group. A: positive control, Avastin; F: Farrerol; O: ononin; E: (−)-epicatechin.