Discovery of new 3-methylquinoxalines as potential anti-cancer agents and apoptosis inducers targeting VEGFR-2: design, synthesis, and in silico studies

Abstract There is an urgent need to design new anticancer agents that can prevent cancer cell proliferation even with minimal side effects. Accordingly, two new series of 3-methylquinoxalin-2(1H)-one and 3-methylquinoxaline-2-thiol derivatives were designed to act as VEGFR-2 inhibitors. The designed derivatives were synthesised and evaluated in vitro as cytotoxic agents against two human cancer cell lines namely, HepG-2 and MCF-7. Also, the synthesised derivatives were assessed for their VEGFR-2inhibitory effect. The most promising member 11e were further investigated to reach a valuable insight about its apoptotic effect through cell cycle and apoptosis analyses. Moreover, deep investigations were carried out for compound 11e using western-plot analyses to detect its effect against some apoptotic and apoptotic parameters including caspase-9, caspase-3, BAX, and Bcl-2. Many in silico investigations including docking, ADMET, toxicity studies were performed to predict binding affinity, pharmacokinetic, drug likeness, and toxicity of the synthesised compounds. The results revealed that compounds 11e, 11g, 12e, 12g, and 12k exhibited promising cytotoxic activities (IC50 range is 2.1 − 9.8 µM), comparing to sorafenib (IC50 = 3.4 and 2.2 µM against MCF-7 and HepG2, respectively). Moreover, 11b, 11f, 11g, 12e, 12f, 12g, and 12k showed the highest VEGFR-2 inhibitory activities (IC50 range is 2.9 − 5.4 µM), comparing to sorafenib (IC50 = 3.07 nM). Additionally, compound 11e had good potential to arrest the HepG2 cell growth at G2/M phase and to induce apoptosis by 49.14% compared to the control cells (9.71%). As well, such compound showed a significant increase in the level of caspase-3 (2.34-fold), caspase-9 (2.34-fold), and BAX (3.14-fold), and a significant decrease in Bcl-2 level (3.13-fold). For in silico studies, the synthesised compounds showed binding mode similar to that of the reference compound (sorafenib).


Introduction
Cancer has been the most difficult and life-threatening illness to be treated 1 . After cardiovascular disease (CVD) cancer has been reported to be a significant cause of death worldwide 2 . At the end of 2018, cancer caused 9.6 million deaths 3 . The current anticancer therapy has many side effects arising from non-selectivity of the development of drug resistance 4 . Nonetheless, there is an urgent need to design new anticancer drugs that can prevent cancer cell proliferation even with minimal to no side effects on healthy cells.
At the level of molecular biology, protein tyrosine kinases have an important role in cell proliferation, migration, survival, and progression 5 . Tyrosine kinases phosphorylate the protein's tyrosine residues resulting in altered protein function. In some cases, tyrosine kinases become continuously active which ultimately leads to cancer 6 . Tyrosine kinase receptors (RTKs) are a panel of cell surface receptors that transfer the signal to polypeptides, hormones, and growth factors 7 . There have been numerous known RTKs such as Vascular endothelial growth factor receptors (VEGFRs) and endothelial growth factor receptors (EGFRs) 8 .
VEGFRs have been recognised as an outstanding medicinal target to discover new anticancer agents 9,10 . The class of VEGFRs comprises three subtypes; VEGFR-1, VEGFR-2, and VEGFR-3 11 . Among them, VEGFR-2 which has a crucial role in tumour angiogenesis. VEGFR-2 can be activated through binding with VEGF which starts the process of phosphorylation which boosts proliferation and migration of the endothelial cells 12 . VEGFR-2 is mainly overexpressed throughout endothelial cells of the tumour vasculature, with less expression in normal endothelial cells 13 . Hepatocellular carcinoma and breast cancer are well-known examples of tumours with overexpressed VEGFR-2 [14][15][16] .
VEGFR-2 inhibitors are small molecules that bind at the ATP binding site of VEGFR-2 to inhibit angiogenesis and lymphangiogenesis 17 . Beside many VEGFR-2 inhibitors approved by FDA or under clinical trials, there are a lot of effort to discover new ones for the management of cancer. Sorafenib I is a bi-aryl urea, has inhibitory effect against tyrosine kinases involved in tumour development, including VEGFR-2 18 . Sunitinib II is anti-tumour drug with dual activity against VEGFR-2 and PDGFR-b 19 . Telatinib III is an orally active anti-VEGFR-2 small-molecule 20 . Nintedanib IV is a potent small-molecule tyrosine kinase inhibitor with oral activity.
In addition, it has a triple angiokinase inhibitory effect as it inhibits the three major signalling pathways involved in angiogenesis 21 . Acrizanib V is a VEGFR-2 inhibitor with limited systemic exposure after topical ocular administration 22 . Vorolanib VI is a novel indolinone-based kinase inhibitor that targets the VEGFR-2 23 . It has fewer adverse effects and a wide therapeutic window 24 .
Essential Pharmacophoric features of VEGFR-2 inhibitors have been reported in many publications [25][26][27][28][29][30] . The reported pharmacophore includes: i) a flat hetero aromatic moiety which binds Cys917 via a hydrogen bonding interaction 26 , (ii) a spacer moiety which occupies the area between the ATP-binding domain and the DFG domain 31 , (iii) a pharmacophore moiety which consists of H-bond acceptor (HBA) and one H-bond donor (HBD) groups (e.g. amide or urea). Both HBA and HBD have a key binding role, as they form hydrogen bonding interactions with two crucial residues (Glu883 and Asp1044) 32 , and (iv) a terminal hydrophobic moiety can make many hydrophobic interactions in the allosteric hydrophobic pocket of VEGFR-2 33 (Figures 1 and 2).
In the current work, ligand-based drug design approach 34-37 was used to synthesise two series of 3-methylquinoxalin-2(1H)-one and 3-methylquinoxaline-2-thiol derivatives. This work is an extension of the earlier activities of our team to synthesise effective anticancer agents targeting VEGFR-2 9,38,39 . The synthesised derivatives were evaluated in vitro and in silico to assess their VEGFR-2 inhibitory activities.

Rationale of molecular design
VEGFR-2 inhibitors competitively block the ATP binding site which consists of four main regions. i) Hinge region which is occupied by the flat hetero aromatic ring of VEGFR-2 inhibitors. ii) Gatekeeper region which is occupied by the spacer moiety of VEGFR-2 inhibitors. iii) DFG-motif region which is occupied by the pharmacophore moiety of VEGFR-2 inhibitors. iv) Allosteric hydrophobic region which is occupied by the terminal hydrophobic moiety of VEGFR-2 inhibitors 33,40-42 ( Figure 2).
The main objective of our design was the synthesis of new compounds having the main pharmacophoric features of VEGFR-2 inhibitors. Such compounds comprise different bio-isosteres, each of them occupy a specific region at ATP binding site.
For the hinge region, two quinoxaline moieties were used; i) 3-methylquinoxalin-2(1H)-one (compounds 11a-j) and ii) 3-methylquinoxaline-2-thiol (compounds 12a-k). The bicyclic structure of quinoxaline moiety is suitable to the large size space of the ATP binding region 43 . In addition, the nitrogen atoms act as hydrogenbond acceptors to facilitate the hydrogen bonding interaction with the hinge region. The Gatekeeper region was targeted to be occupied by N-phenylacetamide moiety as spacer group. Regarding the DFG-motif region, an amide group (pharmacophore moiety) was selected to be buried in it. Finally, the allosteric hydrophobic region can be occupied by different aliphatic and aromatic derivatives to study the structure-activity relationships ( Figure 2).

Chemistry
In order to synthesise the designed compounds, Schemes 1-4 were adopted. Initially, o-phenylenediamine 1 was refluxed with sodium pyruvate 2 to afford 3-methylquinoxalin-2(1H)-one 3 according to the reported procedure 44 . Subsequent heating of compound 3 with alcoholic potassium hydroxide gave the corresponding potassium salt 4 44 . To prepare 3-methylquinoxaline-2thiol 5, the previously prepared 3-methylquinoxalin-2(1H)-one 3 was refluxed with P 2 S 5 in pyridine as a solvent, then the reaction was acidified using HCl 45,46 . Heating of compound 5 with alcoholic potassium hydroxide gave the corresponding potassium salt 6 (Scheme 1).
The key intermediates were synthesised as described in Scheme 2. The commercially available p-amino benzoic acid 7 was treated with chloroacetyl chloride in dry DMF in the presence of NaHCO 3 to afford 4-(2-chloroacetamido)benzoic acid 8. Chlorination of 8 was achieved by its reflux with SOCl 2 in dichloroethane in the presence of catalytic amount of dry DMF to give the key compound 4-(2-chloroacetamido)benzoyl chloride 9. At the end, in acetonitrile and TEA mixture, compound 9 was stirred at room temperature with appropriate amines namely, methylamine, sec-butylamine, cyclopentylamine, 2-methoxyaniline, 3-methoxyaniline, 4-methoxyaniline, 4-aminoacetophenone, 4-fluoroaniline, 2-aminopyridine, m-toluidine, p-toluidine, and 2-aminothiazole to give the corresponding key intermediates 10a-l, respectively. The IR spectra of the key intermediates 10a-l exhibited the appearance of absorption bands at the ranges of 3254 À 3326 cm À1 and 1702 À 1625 cm À1 due to the NH and 2C¼O groups, respectively. In addition, 1 H NMR analyses exhibited the appearance of characteristic singlet signals for amidic NHs around d 10.00 ppm. Also, it showed singlet signals for CH 2 protons of acetamide moiety around d 4.30 ppm. Besides, such CH 2 group was detected around d 44.02 ppm in 13 C NMR spectra. The first series of the target compounds was prepared as described in Scheme 3. The potassium salt of 3-methylquinoxalin-2(1H)-one 4 was heated in dry DMF with the keys intermediates in the presence of catalytic amount of KI to give the titled compounds 11a-j.
The second series of the target compounds was synthesised depending on the synthetic pathway described in Scheme 4. The potassium salt of methylquinoxaline-2-thiol 6 was heated in dry DMF with the keys intermediates 10a-k in the presence of catalytic amount of KI to give the titled compounds 12a-k. 1 H NMR spectra exhibited the presence of singlet signals of CH 3 group of 3-methylquinoxalin-2(1H)-one moiety around d 2.49 ppm. In addition, 1 H NMR spectra showed characteristic signals for additional aromatic and aliphatic protons at the expected chemical shift. Taken compound 11c and 12c as representative examples, it showed many characteristic signals at aliphatic region corresponding to cyclopentyl moiety. Moreover, 13 C NMR spectra of such compounds confirmed the previous results as the aliphatic protons of cyclopentyl moiety appeared at the aliphatic region.

In vitro cytotoxic activities
Cytotoxic activities of the synthesised compounds were evaluated against MCF-7 (human breast cancer cell line) and HepG2 (human liver carcinoma cell line) using MTT assay 47 , using sorafenib as a reference standard (Table 1). Among the target compounds, 11e, 11g, 12e, 12g, and 12k exhibited promising cytotoxicity against the two cell lines with IC 50 values ranging from 2.1 to 9.8 mM, comparing to sorafenib (IC 50 ¼ 3.4 and 2.2 mM against MCF-7 and HepG2, respectively). Compound 11e exhibited a superior activity against MCF-7 and HepG2 with IC 50 values of 2.7 and 2.1, respectively. In addition, compounds 11f and 12f showed promising activity against HepG2 cells with IC 50 values of 9.6 and 7.5, respectively. Compounds 11a, 11b, 11c, 12c, and 12d showed moderate activity against HepG2 cells with IC 50 values of 16.5, 12.8, 18.7, 11.4, 18.7, and 17.6 mM, respectively. In addition, compounds 11f and 12f showed moderate activity against MCF-7 cells with IC 50 values of 12.4 and 10.8 mM, respectively. On the other hand, compounds 11d, 11h, 11i, 11j, 12a, and 12b showed weak cytotoxic activity against the two cell lines.

Vegfr-2 inhibitory assay
VEGFR-2 inhibitory activity of the synthesised compounds was investigated using sorafenib as a reference drug. Table 1.
Summarised the IC 50 values of VEGFR-2 growth inhibitory concentration for all the synthesised members.

Statistical correlation between VEGFR-2 inhibition and cytotoxicity
To study the relation between cytotoxicity and VEGFR-2 inhibition, we plotted the values of VEGFR-2 inhibition against the corresponding cytotoxicity results using simple linear regression analysis. Co-efficient of determination (R 2 ) were calculated in this analysis. It was found that R 2 of VEGFR-2 inhibition and MCF-7 cytotoxicity is 0.887 with p values >0.0001. In addition, R 2 of VEGFR-2 inhibition and HepG2 cytotoxicity is 0.887 with p values >0.0001. The results indicated that there are high correlations between VEGFR-2 inhibition and cytotoxic activity on both cell lines, which reveals that the cytotoxicity may be a result of VEGFR-2 inhibition ( Figure 3).
Then, we investigated the effect of the terminal hydrophobic moiety. Comparing the activity of compounds 11a-c and 12a-c having aliphatic hydrophobic moieties with compounds 11e-g and 12e-g having aromatic hydrophobic moieties indicated that aromatic moieties are beneficial for activity. For aliphatic moieties, there is no great variation in the activity among small size (compound 11a and 12a), bulk (compound 11c and 12c), and branched (compound 11b and 12b) aliphatic moieties.
In addition, the effect of the substitution on the aromatic hydrophobic moieties has been investigated. Comparing the activity of compound 12k (incorporating 4-methylphenyl moiety) and compounds 11f and 12f (incorporating 4-methoxyphenyl moiety) with compounds 11h and 12h (incorporating 4-fluorophenyl) and compounds 11g and 12g (incorporating acetophenone moiety), revealed that grafting an electron donating group is more preferred biologically than electron withdrawing one. For electron withdrawing groups, it was found that acetyl incorporating members (11g and 12g) are more active than fluoro incorporating one (11h and 12h).
Next, the effect of the substitution on the aromatic hydrophobic moieties with electron donating group has been examined. For methylquinoxalin-2(1H)-one derivatives, the activities reduced in the order of 3-methoxy (11e) > 4-methoxy (11f) > 2-methoxy (11d). With regard to 3-methylquinoxaline-2-thiol derivatives, the activities decreased in the order of 4-methyl (12k) > 3-methoxy  Finally, the decreased IC 50 values of compound 11j (with thiazole moiety) against the tested cell lines and VEGFR-2 than the IC 50 values of compound 11i (with pyridine moiety), indicated that five-membered hetero aromatic hydrophobic moiety is more efficient than six-membered one.

In vitro cytotoxicity against normal cell
The cytotoxicities of the most active compounds (11e and 12e) against primary rat hepatocytes (normal hepatic cells) were evaluated in vitro ( Table 2). The results revealed that the tested compounds have low toxicity against the tested cells with IC 50 values of 15.0 and 16.7 lM, respectively. Sorafenib as a reference drug showed IC 50 value of 13.4 mM against the tested cells. These results revealed that the synthesised compounds have low toxicity against the normal cells as their toxicities are comparable to that of FDA approved drug (sorafenib).

Cell cycle analysis
The cell cycle is a well-maintained process by which cells of eukaryotes replicate themselves. The homeostatic balance between cell loss and cell gain must be achieved to produce and conserve the complex architecture of tissues. One way in which this connection may be achieved is through the coupling of the cell cycle and apoptosis 48 .
Since compound 11e effectively inhibited the growth of HepG2 cells, it was expected that this inhibitory effect was due to its capability to hinder the cell cycle progression. Therefore, cell cycle process was analysed after exposure of HepG2 cells to 11e with a concentration of 2.1 mM (IC 50 value of compound 11e) after 24 h. HepG2 cells were used as a control without treatment by compound 11e. Flow cytometry data revealed that the percentage of cells arrested at the G2/M phase increased from 18.24% (in control cells) to 46.62 (in 11e treated cells). In addition, the percentage of HepG2 cells mild increased at the S phase from 25.48 to 31.80%. Oppositely, the percentage of HepG2 cells decreased at G1 phase from 55.03% to 20.34%. Such findings revealed that compound 11e arrested the HepG2 cell growth at G2/M phase (Figure 4 and Supplementary Data).

Apoptosis analysis
To quantify the apoptosis triggered by 11e, Annexin-V/propidium iodide (PI) staining assay was conducted. In such procedure, compound 11e at a concentration of 2.1 mM was applied on HepG2 cells. Then, the media were incubated for 24 h. As shown in Table  3, the apoptotic effect of 11e in HepG2 cells was five times more than observed in control cells. In details, compound 11e induced apoptosis by 49.14% (early apoptosis ¼ 48.87% & late apoptosis ¼ 0.27%), compared to 9.71% in the control cells (early apoptosis ¼ 9.58% & late apoptosis ¼ 0.13%).

Western blot analysis
Apoptosis is a programmed cell death characterised by some biological processes including condensation of nuclear chromatin, loss of plasma membrane phospholipid asymmetry, DNA cleavage into small fragments, and formation of membrane-bound apoptotic bodies 49 .
During intrinsic apoptosis, caspase-9 is activated to produce a subsequent activation of other effector caspases. It was reported that caspase-9 is a highly specific protease that only cleaves a few proteins, whereas caspase-3 contributes to the majority of cleavage that takes place during apoptosis 50,51 . Additionally, the mitochondrial apoptosis is largely mediated through Bcl-2 family proteins, which include. i) Pro-apoptotic members such as BAX that promote mitochondrial permeability and cell death. ii) Antiapoptotic members such as Bcl-2 that inhibit the mitochondrial release of cyt c and suppress cell death 52 . According to these reports, a cell with a high BAX/Bcl-2 ratio will be more sensitive to some given apoptotic stimuli when compared to a similar cell type with a low BAX/Bcl-2ratio 53 .
2.2.8.1. Caspase-3 and caspase-9 determination. To investigate the effect of the synthesised compounds on caspase-3 and caspase-9 levels, the most promising member 11e was applied on the most sensitive cells (HepG2) at a concentration of 2.1 mM for 24 h. Western blot analyses revealed that compound 11e produced a significant increase in the level of caspase-3 (2.34-fold) compared to the control cells. Moreover, compound 11e showed a significant increase in the level of caspase-9 (2.34-fold) compared to the control cells ( Figure 5 and Supplementary Data).
2.2.8.2. BAX and Bcl-2 determination. Compound 11e as a promising member was investigated to evaluate its effect on BAX and  Bcl-2 after 24 h of its application on HepG2 cells using Western blot technique. The results showed that compound 11e produced an up-regulation of BAX and down-regulation of Bcl-2. In details, BAX level increased by 3.14-fold, while Bcl-2 level decreased by 3.13-fold. In addition, BAX/Bcl-2 ratio was 9.17, which indicated that compound 11e had a significant effect on apoptosis pathway ( Figure 5 and Supplementary Data).

Docking studies
In this work, the synthesised compounds were docked against VEGFR-2 using sorafenib as a reference drug. This investigational work was performed to get further insight into the binding modes of the synthesised compounds against VEGFR-2 binding site (PDB ID: 2OH4). The binding free energies (DG) for all the synthesised compounds against the target receptor were calculated and reported in Table 4. The reported key binding site of VEGFR-2 consists of Glu883 and Asp1044 33,54 . Validation of the docking procedure and the binding mode of the reference drug (sorafenib) 33,54 were showed in Supplementary data.
The synthesised compounds exhibited binding mode inside the binding sites of VEGFR-2 similar to that of sorafenib. Compound 11e was completely buried inside VEGFR-2 active site with similar binding mode to sorafenib. The docking score of such compound was À28.81 kcal/mol. The pharmacophore moiety (amide group) was incorporated in hydrogen bonding interaction forming a hydrogen bond with Glu883 and another one with Asp1044. The phenyl group (spacer) formed three hydrophobic interactions with Val914 and Cys1043. The 3-methylquinoxalin-2(1H)-one nucleus was inserted in hinge region of the binding pocket forming five hydrophobic interactions with Leu1033, Phe916, Ala864, and Leu838. In addition, the terminal methoxyphenyl group (hydrophobic tail) formed one hydrophobic bond with Leu887. Additionally, it formed two electrostatic interactions with Asp1044 ( Figure 6).
Regarding compound 11a (incorporating methyl group as hydrophobic tail) showed binding energy of À22.37 kcal/mol. It showed binding mode similar to that of sorafenib with some deviations. Firstly, due to lack of bulk aromatic moiety (as appeared in compound 11e), this led to disappearance of hydrophobic interactions at the allosteric binding pocket of VEGFR-2. In addition, the orientation of 3-methylquinoxalin-2(1H)-one nucleus at the hinge region prevent the hydrogen bonding interaction with Cys917 ( Figure 7).
With respect to compound 12a, it exhibited a binding mode like that of sorafenib with binding energy of À27.98 kcal/mol. The  different features of compounds 12a occupied the same regions which occupied by sorafenib. However, elongation of the linker moiety exerted mild change in the orientation of 3-methylquinoxaline-2-thiol nucleus at the hinge region preventing the hydrogen bonding interaction with Cys917 ( Figure 8).

In silico ADMET studies
The in silico ADMET parameters were assessed via Discovery studio 4.0 using Sorafenib as a reference molecule.
The results revealed that the tested compounds have low or very low BBB penetration levels except for compounds 12h, 12j, and 12k which were predicted to have medium levels. Accordingly, most compounds were anticipated to be safe against CNS. Furthermore, compounds 11a-g, 11i, and 11j were predicted to have good levels of aqueous solubility, while compounds 12a-k were predicted to have low levels. Moreover, intestinal absorption levels of all the tested compounds were predicted to be good. The cytochrome P4502D6 inhibition was predicted using CYP2D6 model 55 . All the tested compounds were predicted as non-inhibitors of CYP2D6. So that, these compounds are expected to be safe for the liver. The plasma protein binding (PPB) model predicts the degree of molecule binding to PP. If it is > ¼ 90%, it means that a molecule can bind the PP at high concentration 56 . Compounds 11c-e, 11g-i, 12h, 12i, and 12k were expected to bind plasma protein over 90%, while compounds 11a, 11b, 11f, 11j, 12ag, and 12i were predicted to bind plasma protein less than 90% (Table 5, Figure 9).

2.3.3.
In silico toxicity studies Discovery studio 4.0 was used to determine the expected toxicity potential of the synthesised compounds 57,58 .
As shown in Table 6, most compounds showed in silico low toxicity profile against the tested models. Compounds 11a-j and 12a,b were predicted to have carcinogenic potency TD 50 values ranging from 19.427 to 81.588 mg/kg body weight/day, which were higher than that of sorafenib (carcinogenic potency TD 50 ¼ 19.236 mg/kg body weight/day). While compounds 12c-k were estimated to have low carcinogenic potency TD 50 values (from 7.026 to 17.638 mg/kg body weight/day). In addition, the rat maximum tolerated doses of compounds 11h, 12b, and 12h-k were estimated to be between 0.133 and 0.096 g/kg body weight, which were higher than that of sorafenib (rat maximum tolerated dose ¼ 0.089 g/kg body weight). The other derivatives were predicted to have fewer rat maximum tolerated doses. Moreover, compounds 11a-c, 11g, 11i, 11j, 12a, 12c, 12g, 12i, and 12k were predicted to be non-toxic against developmental toxicity potential model. For rat oral LD 50 model, the tested compounds showed oral LD 50 values ranging from 2.102 to 18.807 g/kg body weight/day. Such values are far more than that of sorafenib (oral LD 50 ¼ 0.823 g/kg body weight/day). Moreover, all the tested compounds were predicted to be mild irritant against ocular irritancy model, and non-irritant against skin irritancy model.
Physical, elemental, and spectral data of the intermediate compounds 10a-l were depicted in Supplementary data. General procedure for synthesis of compounds 11a-j To a solution of the potassium salt of 3-methylquinoxalin-2(1H)one 4 (320 mg, 0.002 mol) in DMF (20 ml) the appropriate 4-(2-chloroacetamido)-N-substituted-benzamides 10a-i and 10l (0.002 mol) was added. The mixture was heated on a water bath for 10 h. After cooling to room temperature, the reaction mixture was poured onto crushed ice. The precipitated solids were filtered, dried and crystalised from ethanol to give the target compounds 11a-j.

In vitro cytotoxic activity
In vitro cytotoxicity was carried out using MTT assay protocol 47,65-67 as described in Supplementary data.

In vitro VEGFR-2 kinase assay
In vitro VEGFR-2 inhibitory activity was assessed against. Human VEGFR-2 ELISA kit as described in Supplementary data 68,69 .

In vitro cytotoxicity against normal cell
The toxicity of compounds 11e and 12e was assessed against normal cell lines (primary rat hepatocytes) according to method of two-steps in situ collagenase perfusion as described by Seglen 70 (Supplementary data).

Cell cycle analysis
The effect of compound 11e on cell cycle distribution was performed using propidium iodide (PI) staining technique as described in Supplementary data [71][72][73] .

Apoptosis analysis
The effect of compound 11e on cell apoptosis was investigated as described in Supplementary data [74][75][76] .

Western blot analysis
Western blot technique was applied to assess the potential effect of compound 11e on the expression of caspase-9, caspase-3, BAX, and Bcl-2 as reported in Supplementary data 77-79 .

4.3.
In silico studies 4.3.1. Docking studies Crystal structure of VEGFR-2 [PDB ID: PDB ID: 2OH4, resolution: 2.05 Å] was obtained from Protein Data Bank. The docking investigation was accomplished using MOE2014 software. At first, the crystal structure of VEGFR-2 was prepared by removing water molecules. Only one chain was retained beside the co-crystallized ligand (sorafenib). Then, the selected chain was protonated and subjected to minimisation of energy process. Next, the active site of the target protein was defined.
Structures of the synthesised compounds and sorafenib were drawn using ChemBioDraw Ultra 14.0 and saved as MDL-SD format. Such file was opened using MOE to display the 3 D structures which were protonated and subjected to energy minimisation. Formerly, validation of the docking process was performed by docking the co-crystallized ligand against the isolated pocket of active site. The produced RMSD value indicated the validity of process. Finally, docking of the tested compounds was done through the dock option inserted in compute window. For each docked molecule, 30 docked poses were produced using ASE for scoring function and force field for refinement. The results of the docking process were then visualised using Discovery Studio 4.0 software 34,80-83 .

ADMET studies
ADMET descriptors were determined using Discovery studio 4.0 as according the reported method 34,80,84 (Supplementary data).

Toxicity studies
Discovery studio 4.0 software was used to predict the toxicity potential of the synthesised compounds as reported in Supplementary data 85 .
Pharmacy, Al-Azhar University, Cairo, Egypt for his support in biological testing.