Discovery of new quinoline and isatine derivatives as potential VEGFR-2 inhibitors: design, synthesis, antiproliferative, docking and MD simulation studies

Abstract A new set of quinoline and isatine derivatives were synthesized as antiangiogenic VEGFR-2 inhibitors. On a biological level, the in vitro ability of the obtained candidates to inhibit VEGFR-2 was found to be strong with IC50 values in the range of 76.64–175.50 nM. To investigate the cytotoxicity and safety, all compounds were tested against a panel of four cancer cell lines (A549, Caco2, HepG2 and MDA) as well as two normal cell lines (Vero and WI-38). Interestingly, compound 12 exhibited noticeable cytotoxicity against A549, Caco2 and MDA with IC50 values of 5.40, 0.58 and 0.94 µM, respectively. These results were better and comparable to that of doxorubicin (0.70, 0.82 and 0.90 µM, respectively) with more than three folds higher selectivity index against the Caco2 cell lines. Compound 9 prevented the healing of the cancer cells at a low concentration. Also, the compound’s potential to induce programmed cell death in Caco-2 was proved through the significant down regulating of the expression of Bcl2, Bcl-xl and Survivin in addition to the slight upregulation of the TGF-β gene. The cell cycle analysis indicated that compound 9 arrested the Caco-2 cells in the G2/M phase. Interestingly, the molecular docking studies against VEGFR-2 revealed the correct binding of the targeted compounds similar to sorafenib. Furthermore, MD experiments validated the binding of compound 12 with VEGFR-2 over 100 ns, as well as MM-PBSA analysis that confirmed the precise binding with optimum energy. Finally, ADMET analysis showed the general drug-likeness and confirmed the safety of the tested compounds. Communicated by Ramaswamy H. Sarma


Introduction
According to the WHO, colon and rectum cancer was the second most common global cause of cancer deaths in 2020 with an estimated number of 935000 among 10 million (WHO, Cancer, 2022).Cancer comprises beyond 200 various types, each type has a particular method to be diagnosed and treated (NHS, Cancer, 2022).Almost 90% of treatments with chemotherapy got failed because of cancer cells' drug resistance (Mansoori et al., 2017).Cancer cells can develop drug resistance through different mechanisms that may act on the drug by inactivation (Papageorgis et al., 2011), target alteration (Cree & Charlton, 2017) and efflux (Kage et al., 2002).Also, drug resistance may occur via the cancer cell's ability to repair DNA damage and inhibit apoptosis (Vasan et al., 2019).
Protein kinases played a crucial role to modulate drug resistance in cancer cells through different mechanisms (Belal et al., 2022).Among Protein kinases, Vascular Endothelial Growth Factor Receptor-2 (VEGFR-2) has been recognized as one of the most essential proteins.The VEGF/VEGFR-2 signaling pathway was described as the most vital factor in the promotion of the angiogenesis process (Taghour et al., 2022).
Angiogenesis is a critical process to the malignant cells that need to be heavily vascularized to obtain more oxygen and nutrients in order to survive and proliferate (Lugano et al., 2020).It was evidenced that VEGF/VEGFR-2 signaling pathway stimulates several endothelial responses with a noticeable ability to increase the permeability of vessels and enhance proliferation, differentiation and migration of the cancerous endothelial cells (Shi et al., 2016).In consequence, the circumvention of angiogenesis through the inhibition of VEGFR-2 could be a very saucerful strategy in the fight against cancer and cancer's drug resistance.
The computer-aided/in silico drug design methods had an imperative and relevant role in drug discovery through their ability to aid an effective rational design to discover novel active as well as safe compounds.Additionally, these methods have the benefit of saving effort, time and costs besides animal lives (Marrone et al., 1997;Li et al., 2020;Abdel-Aziz et al., 2015;Elkaeed et al., 2022).In accordance, the in silico approach was a cornerstone in various reports that reported drug design (Suleimen et al., 2021;Belal et al., 2022), discovery (Imieje et al., 2021), ADMET (Elkaeed et al., 2022) and toxicity (Elkaeed et al., 2022;Elkaeed et al., 2022) of new drugs.

Rational
An appraisal of the literature revealed that the VEGFR-2 active pocket has four different regions that can be occupied by four significant pharmacophoric fragments.The key interactions inside the VEGFR-2 receptor were represented by sorafenib, a well-known VEGFR-2 inhibitor (Aziz et al., 2022) (Figure 1).
The present work focuses on the advancements of the reported VEGFR-2 inhibitors by taking nearly the same essential features that targeted the main active pockets of VEGFR-2.In niceties, as an aromatic moiety occupied the hinge region, the 2-oxoindoline and quinoline moieties were derived from sunitinib and cabozanitinib to prepare the indoline-containing compounds 12-14 and quinoline-containing compounds 8 and 9.The thiazolidine-2,4-dione moiety was designed to occupy the linker region.Regarding the HB region, an amide group was designed to bind with it.Finally, for the hydrophobic pocket, it was designed to be occupied with different aromatic moieties including methyl and dichloro derivatives Figure 2.
To complete the story, VEGFR-2 inhibition, cytotoxicity and safety evaluations of the synthesized compounds were presented.Additionally, flow cytometric analyses and apoptotic markers investigation were performed for the safest compound to understand its mechanism of action.Moreover, detailed in silico examinations including; molecular docking, ADMET, MD simulation and MM-PBSA were carried out to confirm the biological findings
IR, 1 H NMR, 13 C NMR, analyses can be used to identify the chemical structure of new molecules.The 1 H NMR characterizes the protons of the target compounds that appeared at certain regions.For example, singlet signals at 2.29, 2.24 and 2.28 ppm are due to the protons of methyl groups of compounds 8, 13 and 14, respectively.Also, 13 C NMR spectral data showed the expected signals of the aliphatic carbons of the synthesized compounds.For example, the signals of the methyl carbons of compounds 8, 13 and 14 appeared at d 21.30, 18.25 and 21.62 ppm, respectively.

In vitro VEGFR-2 inhibition assay
Initially, the in vitro evaluation for enzymatic inhibitory activity was done for all compounds (8, 9, 12, 13 and 14) using a VEGFR-2 kinase assay.The inhibition activity was recorded as IC 50 in nM and compared with sorafenib (standard inhibitor).The results displayed in Table 1 presented that the isatine derivatives 12 and 14 manifested good activities with IC 50 values of 76.64 and 80.15 nM, respectively.Controversy, compounds 8, 9 and 13 expressed moderate activities with IC 50 values of 122.90, 175.50 and 95.87 nM, respectively.

In vitro antitumor and safety evaluation (MTT assay)
The ability of the chemotherapeutic agent to attack the cancer cells specifically and not affect the normal cells is the key and essential feature in drug discovery.In this regard, the toxicities of all candidates in this study were tested against two normal cells, Vero and WI-38 cell lines using the MTT assay in the presence of doxorubicin as a reference molecule (Table 2).The obtained results indicated that compounds 8 and 9 were the safest members on Vero and WI-38 cell lines.
In particular, compounds 9 recorded the highest IC 50 values on Vero cell lines (IC 50 ¼ 28.50 mM) while compound 8 recorded the highest IC 50 values on WI-38 cell lines (IC 50 ¼ 11.75 mM).Also, the rest of the results revealed the general safety profile of the target compounds.
In order to examine the potentials of the synthesized members as anticancer agents, their inhibitory activities against four human cancer (adenocarcinoma A549, colorectal carcinoma, Caco2, hepatocellular carcinoma HepG2 and epithelial, human breast cancer, MDA) cell lines were examined.Doxorubicin, a universal anticancer agent, was used as a reference standard and displayed IC 50 values of 0.70, 0.82, 0.28 and 0.90 mM, against the cell lines A549, Caco-2, HepG2 and MDA, respectively.Cytotoxic results presented in Table 3 showed cell growth inhibition by all the synthesized compounds against the four cell lines It is worth noting that some of the investigated compounds appeared to have superior inhibitory effects.The hybrid compound 12 showed the most potent effects against A549, Caco-2 and MDA cell lines with IC 50 values of 5.40, 0.58 and 0.94 mM, respectively, while showing moderate effect against HepG2 cell line (IC 50 ¼ 14.45 mM).Interestingly, it displayed a promising activity against the Caco2 cell lines (IC 50 ¼ 0.58 mM), compared to doxorubicin against the same cell (IC 50 ¼ 0.82 mM).On the other hand, compound 13 was the most promising member against HepG2 cell lines with IC 50 values of 4.82 mM.Also, such compounds displayed moderate antiproliferative activities against Caco-2 and MDA cell lines with IC 50 values of 1.10 and 4.32 mM, respectively.Moreover, compounds 8, 9 and 14 showed moderate antiproliferative activities against Caco2, the most sensitive cell line, with IC 50 values of 9.87, 13.90 and 9.81 mM, respectively.The selectivity index (SI) was identified for each tested compound, as presented in Table 3, by scaling its IC 50 value against various cancer cell and IC 50 value against the normal cell line (Vero) (Koch et al., 2005).
As shown in Figure 3, compound 12 showed a selectivity index three-fold higher than doxorubicin.Excitingly, the same compound presented a comparable activity to that of doxorubicin against the MDA cell lines with also a higher selectivity index (more than three folds).Fascinatingly, compound 13 indicated strong activities to that of doxorubicin against the Caco2 and MDA cell lines displaying a much higher selectivity index (more than fifteen and three folds, respectively).
According to Table 2, the safest compound among the synthesized group is compound 9. Consequently, it was selected to conduct further biological examinations to understand the mechanism of action of such active compounds

Wound healing assay of compound 9 on caco-2 cells
The antiangiogenic properties of compound 9 (its inhibitory effect on cell migration and healing) were then assessed using a wound-healing assay.The idea of this assay is to create a wound abrasion in the cancer cells with a defined gap (0.36 mm) then the migratory cells can invade the wound field and close it.
After 24 h, the wound was examined; the diameter of wound Scheme 1. Synthesis of compounds 8 and 9.
closure was calculated for the treated cell lines against the untreated ones.The results showed that the scratch of untreated Caco-2 cells (control) was significantly decreased from 0.36 to 0.06 mm and was fully closed within 24 h as shown in (Figure 4B).While the wound diameter of CaCo-2 cells that were treated with compound 9 (sub IC 50 value) slightly decreased from 0.36 to 0.304 mm (Figure 4C).These results reflect the ability of compound 9 to cause an extreme reduction of the Caco-2 migration potential and consequently, confirm its antiangiogenic characters.

The genetic mechanism of compound 9 against caco-2 cells using RT-qPCR
The possible mechanisms of compound 9 to induce programmed cell death in the treated Caco-2 cells have been studied through the investigation of the apoptotic regulator genes.The expression levels of (Bcl2, Bcl-xl, Survivin and TGF-b) were investigated by the RT-qPCR technique.
The overall results indicated that compound 9 showed significant variations in the gene expression levels of the four cancerrelated genes.As displayed in Figure 5, cellular treatment with compound 9 caused downregulating of the expression of Bcl2, Bcl-xl and Survivin and slightly upregulation of TGF-b genes.

Cell cycle analysis through flow cytometry
Next, the treated Caco-2 cell lines were subjected to a flow cytometric analysis of their cell cycle pattern comparing the untreated cells (Figure 6).Compared with the untreated cells (Figure 6A), the treated Caco-2 cell lines (Figure 6B) showed cellular arrest in the G2/M phase (increased cell population to 28.1% compared with 9.0% in the untreated cells).Also, decreased number of cells was reported in G0/G1 and S phases (26.4 and 12.5%, respectively).The same values were (51.7 and 24.7%, respectively) in the untreated cells.

LDH assay
The cytotoxicity assay was confirmed using LDH assay by treating WI-38, Vero and CaCo-2 cell lines with the IC 50 value of the tested compounds.After incubation, the exhausted media were collected by centrifugation to quantify the concentration of LDH after and before treatment using the LDH cytotoxicity detection kit (Promega) following the manufacturer's protocol and instructions.The cytotoxicity assay of the IC 50 values of the compound 9 on both cancerous and noncancerous cell lines are indicated in Figure 7.The results indicated a significant increase in the concentration of LDH after treatment from 2.7 U/L in the untreated CaCo-2 cell to 57.4 U/L in the treated cells and 20.25 after H 2 O 2 induction using the same concentration.Also, Wi-39 and Vero cell treated cells with the IC 50 concentration of the tested compound showed an increase in LDH concentration (from the untreated cells) from 3.1 to 23.45 U/L and 2.2 to 29.25 U/L, respectively, and comparing with the positive control (H 2 O 2 at the same concentration) that recorded 18.25 and 10.79 U/L, respectively.

Docking study
Molecular docking is an interesting and vital tool for scientists who work in drug discovery as it allows them to predict the binding of certain molecules against a specific molecular target (El-Helby et al., 2019;Belal et al., 2022;Abdallah et al., 2021).In this investigation, docking studies of the synthesized members against their target enzyme, VEGFR-2, were achieved.First, the native ligand was docked into the active site of VEGFR-2 and the results indicated the validity of the docking protocol (Figure 8).In addition, sorafenib's binding mode was in agreement with the reported before (Figure 9).
As shown in Figure 10, compound 8 as a representative compound occupied the linker region via its thiazolidine-2,4dione moiety to form many hydrophobic interactions (HIs) with Cys1043, Phe1045, Val914 and Val846.Additionally, it was further stabilized within the active site of the enzyme via the key HB interactions between the quinoline amide moiety and Cys917.Also, the oxoquinoline moiety formed five hydrophobic and one electrostatic interaction between the hinge amino acid residues Leu1033, Leu838, Ala864 and Cys917.Finally, the 4-methylphenyl moiety was buried into the hydrophobic pocket forming many HIs with Leu887, Ile886 and Val897.The potential of the obtained molecules to bind and interact with the essential amino acids within the active site of VEGFR-2 could explain their activity.The binding modes of the other compounds were illustrated in the Supplementary material.

MD Simulations
The molecular docking studies are in silico experiments that can clearly reveal the binding of a ligand within a protein.In docking studies, the interaction among the binding system described the target protein as a rigid unit.Hence, docking experiments do not consider what happens to the protein's structure after binding to the active compound (Sousa et al., 2006).An MD simulation experiment, on the other hand, will be able to unambiguously examine and explore at an atomic level the behavior and changes of the protein's structure (Liu et al., 2018).The MD simulations experiment, therefore, describes accurately proteins' conformational changes in response to ligand binding.
The binding of a compound into an enzyme's active site affects the conformation (structure) of that enzyme (Kuzmanic & Zagrovic, 2010).As a result, conformational changes of the emerged compound 12-VEGFR-2 complex as well as its dynamics were studied as RMSD to better understand its stability after binding.Excitingly, the emerged complex fluctuated slightly till 30 ns � and got stabilized later till the end of the study (Figure 11A).Based on the RMSF, the flexibility of the VEGFR-2 in the emerged complex was investigated.Advantageously, compound 12's binding makes the protein flexible to some extent in 1025-1075 residue areas (Figure 11B).The radius of gyration (R g ) is a vital element relating the consistency of the investigated enzymes to their volume change.The R g is the RMSD of a weighted mass unit of atoms of the considered enzyme from the mass center.The R g indicates the 3 D changes in a protein, as well as its   compactness, and the degree of fluctuation during the simulation time inversely correlates with the stability and compactness of the system (Liu et al., 2017;Kumar et al., 2014).Captivatingly, the Rg of the emerged complex was found to be less than the start time confirming stability and compactness (Figure 11C).The interaction of the compound 12-VEGFR-2 complex with the enclosing solvents was analyzed by SASA throughout the simulations.The computed SASA value indicates the conformational changes that occurred throughout the study.A lower SASA value indicated a lowering of surface area and therefore, a stable complex.Rivetingly, as shown in Figure 11D, the compound 12-VEGFR-2 complex displayed lower SASA values that were also stable all the study.Finally, the MD simulation experiments were employed to compute the created hydrogen bonds between compound 12 and VEGFR-2 enzyme.The MD simulations confirmed that the highest H-bonding conformations numbers through the compound 12-VEGFR-2 were up to four hydrogen bonds (Figure 11E).
Additionally, we analyzed the conformational (structure) changes that occurred as a result of the binding compound 12 and VEGFR-2 enzyme during the 1 and 100 ns of the MD simulations production run.As illustrated in Figure 12, the results indicated those changes, as well as authenticated the stability of compound 12 inside the VEGFR-2 enzyme active pocket during the whole experiment.

MM-PBSA
The accurate binding free energy of the compound 12-VEGFR-2 complex was calculated during the last stable 20 ns of the MD simulations' production run by means of  MM/PBSA.Engrossingly, compound 12 displayed an extremely low binding free energy of À 120 KJ/mol against the VEGFR-2 enzyme; the binding free energy was stable (Figure 13A) through the duration of the examination, indicating an accurate bonding of the complex.
In the following, the binding energy of the emerged complex was examined to reveal the various components that contribute to the energy and to determine the essential amino acid residues that positively contribute to the interaction process.
Four residues (VAL-846, LEU-887, VAL-914 and CYS-1043) residues of the VEGFR-2 (Figure 13 B) contributed higher than À 5 KJ/mol, and these four residues were considered key or essential residues that positively contributed to binding with compound 12.

In silico ADMET analysis
Five parameters related to the ADMET profile were computed for compounds 8, 9, 12, 13 and 14 using Discovery studio 4.0and displayed in Table 4. Sorafenib was used as a reference molecule.
Agreeably, all the considered compounds presented very low (except compounds 13 and 14 were low) potential to pass the BBB.Although all the considered compounds manifested low aqueous solubility (Aq-S), all of them indicated good intestinal absorption (In-A) levels (except compound 9 presented a moderate level).Fascinatingly, all of them demonstrated noninhibitory potential against the cytochrome P450, CYP2D6, (I-CYP2D6) and were expected to bind to the plasma protein (B-PB) with a ratio of more than 90% (Figure 14).

In silico toxicity studies
Seven computational models of toxicity were performed with Discovery Studio software (Xia et al., 2004;BIOVIA, QSAR, ADMET and Predictive Toxicology, 2020).The obtained results were displayed in Table 5. Sorafenib was used as a reference molecule.
All compounds exhibited noncarcinogenic patterns for mouse females based on the FDA mouse carcinogenicity model, (FDA-C).Besides, all compounds were nonmutagen against the Ames mutagenicity model (AM-MUT).For the rat  oral LD 50 model, (R-O-LD 50 ), all compounds except compounds 8 and 9 displayed rat oral LD 50 values ranging from 1.029 to 2.880 mg/kg/day.Such values are more than that of sorafenib (0.890 mg/kg/day).Also, all compounds except compound 9 showed rat chronic LOAEL, (RC-LOAEL), values ranging from 0.004 to 0.043 g/kg which were equal to or more than that of sorafenib (0.004 g/kg).Additionally, all the compounds expressed a predicted mild and nonirritation in ocular and skin irritancy models, respectively.

Conclusion
Five quinoline and isatine derivatives were synthesized as antiangiogenic VEGFR-2 inhibitors.The considered candidates inhibited VEGFR-2 with IC 50 values in the range of 76.64-175.50nM and administrated strong activities against cell lines (A549, Caco2, HepG2 and MDA) conjugated with a high degree of safety against two normal cell lines (Vero and WI-38).In particular, compound 12 inhibited the growth of A549, Caco2 and MDA with IC 50 values of 5.40, 0.58 and 0.94 mM, respectively.These results were better and comparable to doxorubicin (0.70, 0.82 and 0.90 mM, respectively) showing a higher selectivity index.Also, compound 13 was stronger than doxorubicin against the Caco2 and MDA cell lines with much higher selectivity indices.Compound 9, (the safest candidate), prevented cancer cells' healing, induced programmed cell death and arrested the Caco-2 cells in the G2/M phase.On the in silico level, the molecular docking  studies revealed the correct binding of the targeted compounds against VEGFR-2 (PDB ID: 2OH4).Also, the MD and MM-PBSA experiments validated the accurate binding of compound 12, the most VEGFR-2 inhibitor in this study, against VEGFR-2 over 100 ns.Finally, ADMET analysis indicated the general drug-likeness as well as the safety of the target compounds.

Chemistry
All the reagents, chemicals and apparatus were held in the Supplementary material.The targeted compounds were furnished following the reported methods.

In vitro VEGFR-2 kinase assay
VEGFR-2 ELISA kit was applied.Detailed discussions were held in the Supplementary material (Abou-Seri et al., 2016).0 is very high, � is high, �� is medium, ��� is low, ���� is very low.b � is very low, �� is low, 3 is good, 4 is optimal.c ���� is good, ��� is moderate, �� is poor, � is very poor.d TR is inhibitor, FS is non inhibitor.e FS is less than 90% and TR is more than 90%.

Figure 1 .
Figure 1.Diagram of sorafenib showing its key interactions with VEGFR-2 TK active site.

Figure 2 .
Figure 2. Schematic summary of the rational.
b SI ¼ Cytotoxic effect on Vero/Cytotoxic effect on A549 cell line.c SI ¼ Cytotoxic effect on Vero/Cytotoxic effect on Caco-2 cell line.d SI ¼ Cytotoxic effect on Vero/Cytotoxic effect on HepG2 cell line.e SI ¼ Cytotoxic effect on Vero/Cytotoxic effect on MDA-MB-231 cell line.

Figure 3 .
Figure 3. Activity and selectivity indices of the synthesized compounds against the four cancer cell lines.

Figure 4 .
Figure 4. Wound healing assay.(A) Microscopic image illustrates the initial wound insert.(B and C) The inhibitory effect of negative control and compound 9 on the migration of Caco-2 cells.(D) The graphical illustration of the diameter of the wound against control and treated cells.

Figure 6 .
Figure 6.Cell cycle analysis of Caco-2 cell line after treatment with compound 9.

Figure 8 .
Figure 8. Validation of the docking protocol.

Table 2 .
In vitro anti-proliferative activities of 8