Effective natural inhibitors targeting IGF-1R by computational study

Virtual filtering of ZINC15 database for IGF-1R

ZINC is a free public resource for ligand discovery. The database contains over twenty million commercially available molecules in biologically relevant representations that may be downloaded in popular ready-to-dock formats and subsets. The Web site also enables searches by availability, bioactive and drugs, biogenic, reactivity, and others. ZINC connects purchasable compounds to high-value ones such as metabolites, drugs, natural products, and annotated compounds from the literature [12, 13]. Natural products have evolved over millions of years and acquired a unique chemical diversity, resulting in the diversity of their biological activities and drug-like properties. Therefore, even before the rise of modern chemical pharmacology, natural products have been used for centuries as components of traditional medicines, particularly as active components of herbal remedies. In modern pharmacology too, natural products have become one of the most critical resources for developing new lead compounds and scaffolds [14]. Therefore, we chose to select ligands from the subset of natural products in the biogenic classification of the ZINC database, and these ligands must have been named and for sale, which can facilitate our experimental studies. In the end, we selected 17931 ligands for follow-up research through the screening of the above conditions.

The structure of IGF-1R will be used to construct the receptor, and then we will perform a LibDock docking operation between the receptor and 17931 ligands. By molecularly docking 17931 ligands with IGF-1R one by one, 5169 ligands that could be successfully docked with it were selected. ZINC38433168 is a new class of IGF-1R inhibitors related to hydantoins that was discovered through high-throughput screening [15]. It can combine with IGF-1R, making it difficult for IGF-1R to activate various signal pathways so as to achieve the purpose of treating Ewing sarcoma [15]. Libdock is a fast molecular docking tool for fast and accurate virtual screening of large-scale databases. The higher the score, the better the effect of molecular docking. In accordance with LibDock score, the compounds with the top 20 scores proceeded to the next step (Table 1).

ADME properties of ligands

The ADME module predicted the pharmacological properties of 17931 ligands and the reference drug ZINC38433168. Pharmacological properties listed in Table 2 include water-soluble level, plasma protein binding (PPB) properties, liver toxicity, blood-brain barrier (BBB) level, binding to CYP2D6, and intestinal absorption level. By definition of water-soluble, it is found that most compounds can be soluble in water. For the blood-brain barrier, ZINC000040865806 and ZINC000013451932 have low permeability, ZINC000027646038 and ZINC000027646060 have high permeability, and the levels of all other compounds are uncertain. 17 compounds and the reference drug ZINC38433168 are predicted to act as non-CYP2D6 inhibitors that significantly affect drug metabolism. 8 compounds were also found to have no hepatotoxicity, similar to the reference drug ZINC38433168. 6 compounds were found to have good intestinal absorption characteristics. 9 compounds were strongly absorbent based on the plasma protein binding properties.

At the same time, we should also attach great importance to the safety of natural compounds. As far as the safety of the compounds is concerned, we used the TOPKAT module to study the toxicity of these 20 compounds and the reference drug ZINC38433168, such as rodent carcinogenicity, Ames mutagenicity, and potential characteristics of developmental toxicity (Table 3). Among the results, it was found that 7 compounds were nonmutagenic, and 5 compounds had no developmental toxicity. The results in Table 3 predict that the reference drug ZINC38433168 has high rodent carcinogenicity in mice but no rodent carcinogenicity in rats. Considering all of the above results, ZINC000014946303 and ZINC000006003042 have no hepatotoxicity.

Moreover, their rodent carcinogenicity, Ames mutagenicity, and developmental toxicity are at a low level compared to the remaining compounds. Overall, we can identify ZINC000014946303 and ZINC000006003042 as safe drugs. ZINC000014946303 and ZINC000006003042 and the reference drug ZINC38433168 have similarities in their chemical structures and were selected for the following study (Figures 1, 2).

Ligand-binding analysis

We will use IGF-1R to study the ligand binding mechanism of candidate compounds. Firstly, we docked ZINC000014946303, ZINC000006003042, and ZINC38433168 into the molecular structure of IGF-1R through the CDOCKER module, calculated and displayed the CDOCKER potential energy. The reference drug ZINC38433168 has higher CDOCKER potential energy than ZINC000014946303 and ZINC000006003042 as calculated by the CDOCKER module (Table 4). In contrast to the reference drug ZINC38433168, IGF-1R has a higher binding affinity for ZINC000014946303 and ZINC000006003042. The hydrogen bond and π-related interactions were studied by structural calculation (Figures 2, 3). Characteristic pharmacodynamic masses refer to the active conformation that produces geometric and energy matching when the drug molecule interacts with the receptor target. As shown in Figures 2, 3, there were 21 Characteristic pharmacodynamic masses in ZINC000014946303 and 32 Characteristic pharmacodynamic masses in ZINC000006003042. In the complex, they form several pairs of π-related interactions and hydrogen bonds with IGF-1R. In Tables 5, 6, 3 hydrogen bond pairs were found in the complex formed by the combination of ZINC000006003042 and IGF-1R, by O11 of ZINC000006003042 with LYS1033:HZ1 of IGF-1R, by O23 of ZINC000006003042 with LYS1033:HZ3 of IGF-1R, by O23 of ZINC000006003042 with LYS1033:HE2 of IGF-1R (Table 5). We also show the residue and the hydrogen bond in the structural figures (Figure 3D–3F). ZINC000006003042 and IGF-1R also formed a 4-pair π-related interaction, by 2 pairs of MET1079 of IGF-1R with ZINC000006003042, by 1 pair of MET1054 of IGF-1R with compound, by 1 pair of VAL1063 of IGF-1R with compound (Table 6). ZINC000014946303 and IGF-1R formed 9 pairs of π-related interaction and 1 pair of hydrogen bonds. When the reference drug ZINC38433168 is combined with IGF-1R, 6 pairs of π-related interactions and 3 pairs of hydrogen bonds will be formed. The interaction between ZINC000014946303 and ZINC000006003042 and IGF-1R is shown in Figure 4.

Molecular dynamics simulations

We use molecular dynamics simulation to evaluate the stability of the ligand-IGF-1R in the natural environment and use the CDOCKER module to carry out molecular docking experiments and then obtain the original conformation. The potential energy and RMSD images for each complex are listed in Figure 5. After 20 ps, the tracks of ZINC000014946303 and ZINC000006003042 come into equilibrium. In time, the potential energy and RMSD of these 2 complexes tend to be stable. In brief, both ZINC000014946303 and ZINC000006003042 can bind to IGF-1R to form complexes that can exist stable in the natural environment inhibiting IGF-1R.

ZINC000014946303 and ZINC000006003042 reduces the proliferation of Ewing sarcoma cells

We examined the effect of ZINC000014946303 and ZINC000006003042 on Ewing sarcoma cells in vitro. A 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay showed that the viability of RD-ES cells decreases as the concentration of the drug increases (P < 0.05) (Figure 6A). In a colony formation assay, RD-ES cells treated with ZINC000014946303 and ZINC000006003042 exhibited lower clonogenicity in both number and size than the control group cells (Figure 6B, 6C).

ZINC000014946303 and ZINC000006003042 reduces the migration of Ewing sarcoma cells

The effect of ZINC000014946303 and ZINC000006003042 on Ewing sarcoma cells invasion and migration is determined by a scratch assay. After 24 h, the scratch area decreased significantly in the control group, but only slightly in the ZINC000014946303 group and the ZINC000006003042 group (Figure 6D, 6E). Although the scratch area in the control group, the ZINC000014946303 group, and the ZINC000006003042 group all decreased over time, the reduction was greater in the control group. These results suggest that ZINC000014946303 and ZINC000006003042 reduce the migration of Ewing sarcoma cells.

ZINC000014946303 and ZINC000006003042 reduce IGF-1R expression in Ewing sarcoma cells

To verify that the effects of the ZINC000014946303 and ZINC000006003042 were due to their inhibition of IGF-1R in Ewing sarcoma cells, we assessed levels of IGF-1R using Western blotting. Western blotting demonstrated that IGF-1R expression decreases with increasing drug concentrations (Figure 6F). These findings suggest that ZINC000014946303 and ZINC000006003042 induce apoptosis in Ewing sarcoma cells by inhibiting IGF-1R.

ZINC database and discovery studio software

Discovery Studio software is characterized by simulating small and macromolecular systems. Its primary function is to proceed with the screening, design, and modification of potential drugs by utilizing calculations of structural biology and chemistry, right after identifying and refining lead compounds and candidate drug pathways [16].

LibDock, CDOCKER, ADME, TOPKAT belong to the Discovery Studio4.5 system. The functions of each module are as follows, LibDock is a fast molecular docking tool for fast and accurate virtual screening of large-scale databases. CDOCKER is a CHARMm-based flexible docking program that uses soft-core potentials and optimization grid representation to interface ligand molecules with active receptor sites. ADME provides six computational models for predicting absorption, distribution, metabolism, excretion, and toxicological properties from chemical structure. TOPKAT is a tool for predicting the toxicological properties of compounds. The Ames test (Salmonella typhimurium reverse mutation assay) is a bacterial short-term test for identification of carcinogens using mutagenicity in bacteria as an end point. Previous researchers have conducted drug selection studies through this method to verify the reliability of these modules [17–19]. At UCSF (University of California, San Francisco), Irwin and shochet laboratories provide the ZINC15 database available for unrestricted commercial use.

LibDock for virtual filtering

The high-throughput screening (HTS) initiation occurs in the search for IGF-1R inhibitors. HTS focuses on compounds that are in affinity to IGF-1R. Since the reference drug ZINC38433168 is produced in a complete drug library, synthetic aperture radar is established for hundreds of analogs [15]. LibDock is able to act upon the virtual filtering step [20]. It can utilize probes and grid sites to calculate hot spots in protein. Then protein hot spots are further used for ligand arrangement and intelligent minimization algorithm, and CHARMm field of force is applied to minimization of ligand so that favorable interactions can be formed [21]. All ligand positions are sorted according to the ligand fraction when the ligand is minimized. The protein database identifier of IGF-1R is 3O23, and its chemical structure is listed in Figure 2. The protein database identifier of the reference drug is ZINC38433168. We got 3O23's crystal structure from the Protein Database (PDB), and the ligand-binding region of 3O23 was chosen. It was prepared by removing the crystal water and other heteroatoms from the surrounding area, followed by protonation and hydrogen addition. Following that, the produced protein was utilized to determine the binding site, which was then picked as the best docking site. The next step is to bind the ready ligand pair to the active site and perform virtual screening. Docking sites are sorted and grouped according to the LibDock score by the name of the corresponding compound.

Toxicity indicators and ADME characteristics of ligands

With respect to the detections of distribution and metabolic profiles, excretion and absorption levels for candidate compounds, and potential compounds' toxicity and other properties, our researchers take advantage of the ADME module and TOPKAT module in DS4.5. Apropos of BBB and PPB levels, water-solubility, CYP2D6, hepatotoxicity, Ames mutagenicity, rodent carcinogenicity, and DTP, it is essential to thoroughly investigate the aforesaid pharmacological characteristics before we can begin to select IGF-1R inhibitors.

Molecule docking and pharmacophore prediction

When conducting molecular docking research, we apply the CDOCKER module in DS4.5. CDOCKER is characterized by the ability to analyze the posture of the molecule under test and the high precision docking with the help of the force field of CHARMm. The interaction energy generated by the force field of CHARMm acts on the receptor and ligand and is reflected in their binding affinity. When the ligand is docked with the receptor, the water molecule changes from a crystalline state to an immobilized state, continuously changing the formation state of the complex [22, 23]. It is essential to remove the water molecules and add hydrogen atoms to the protein. When the binding of ligand and receptor reaches the optimal docking state, the complex will dock with IGF-1R and then initiate the CDOCKER process [24, 25].

Dynamic simulation experiment

For a start, we chose the optimal binding conformation of each compound to conduct molecular dynamics simulation experiments. The first step is to establish an orthogonal box that can accommodate the ligand-receptor complex and solvate it using a water model. The system was then injected with solid chloride to simulate the physiological environment, and the force field of CHARMm and energy was adjusted to a minimum. The temperature and pressure were adjusted throughout the operation of the production module, with a duration and step size of 25 ps and 1 fs. The linear constraint solving algorithm and the particle grid Ewald algorithm respectively correspond to fixation of the hydrogen bond and calculation of the long-distance electrostatic field. Ultimately, we define the initial complex settings as the reference. The trajectory's structure, energy, and RMSD can be analyzed using the DS4.5 trajectory protocol.

Cell lines

The American Type Culture Collection provided Ewing sarcoma cells (RD-ES). The cells were cultured in Dulbecco's modified Eagle medium (DMEM, Hyclone, USA) supplemented with 10% fetal bovine serum (Gibco, MD, USA). At 37° C, the culture dishes were kept in 5% CO2 and 95% air.

MTT assay

Ewing sarcoma cells (RD-ES) were plated at a density of 500 cells per well in 96-well culture plates and treated with different doses of ZINC000014946303 and ZINC000006003042. MTT (Sigma, St. Louis, MO, USA) was dissolved in phosphate-buffered saline (5 mg/mL) to test cell viability. The media was replaced with fresh DMEM supplemented with 10% fetal bovine serum and diluted MTT (1:10, 10% MTT) on the day of measurement, and the cells were incubated for 3.5 hours at 37° C. The formazan crystals were then dissolved in 200 L of dimethyl sulfoxide solution after removing the incubation medium. The decrease of MTT was measured using an ELx800 absorbance microplate reader (BioTek Instruments, VT, USA) based on light absorbance at 570 nm.

Colony formation assay

At a density of 50 cells/cm², Ewing sarcoma cells (RD-ES) were seeded in Petri dishes. After 24h of culture, treated the cells with different doses of ZINC000014946303 and ZINC000006003042, respectively. Colonies were counted and characterized after 10 days of in vitro culture. Then rinsed the colonies with phosphate-buffered saline, fixed the 4% paraformaldehyde for 15 min, added 5% crystal violet staining for 30 min after discarding the fixative solution, and slowly washed off the staining solution with running water.

In vitro scratch assay

RD-ES cells were first cultured on 24-well Permanox™ plates until the entire plate was covered. Used a 10-μL pipette tip to evenly stroke consistent cell-free areas in each well along a ruler and slowly rinse loose cells using DMEM. Cells were then treated with different doses of ZINC000014946303 and ZINC000006003042. An image of the scraped area was captured using a phase-contrast microscope at three-time points of 0, 12, and 24 hours. The remaining wounded areas and scratch width were measured at six different points in each image.

Western blotting

Ewing sarcoma cells (RD-ES) in the log growth phase were seeded at a density of 2 × 10⁵ cells/well in a six-well plate, followed by treatment with different doses of ZINC000014946303 and ZINC000006003042. After 48 h of culture, the cells were lysed to extract the total protein and electrophoretically separated. Transferred the protein onto the membranes, cut the protein band of interest, treated it with primary antibodies against IGF-1R and β-actin, then washed the membranes and incubated with secondary antibodies. The membranes were visualized by an enhanced chemiluminescence detection system (Pierce; Thermo Fisher Scientific, Inc.).