1,3,4-Oxadiazole-naphthalene hybrids as potential VEGFR-2 inhibitors: design, synthesis, antiproliferative activity, apoptotic effect, and in silico studies

Abstract In the current work, some 1,3,4-oxadiazole-naphthalene hybrids were designed and synthesised as VEGFR-2 inhibitors. The synthesised compounds were evaluated in vitro for their antiproliferative activity against two human cancer cell lines namely, HepG-2 and MCF-7. Compounds that exhibited promising cytotoxicity (5, 8, 15, 16, 17, and 18) were further evaluated for their VEGFR-2 inhibitory activities. Compound 5 showed good antiproliferative activity against both cell lines and inhibitory effect on VEGFR-2. Besides, it induced apoptosis by 22.86% compared to 0.51% in the control (HepG2) cells. This apoptotic effect was supported by a 5.61-fold increase in the level of caspase-3 compared to the control cells. Moreover, it arrested the HepG2 cell growth mostly at the Pre-G1 phase. Several in silico studies were performed including docking, ADMET, and toxicity studies to predict binding mode against VEGFR-2 and to anticipate pharmacokinetic, drug-likeness, and toxicity of the synthesised compounds.


Introduction
Although there are many advances in drug discovery for cancer control and treatment, still cancer is one of the most serious diseases responsible for a huge number of deaths 1 . According to global statistics, cancer is a leading cause of death worldwide, accounting for nearly 10 million deaths in 2020 2 . The cancer problem is localised not only in its widespread and metastasis but also in the lack of selectivity of anticancer drugs which leads to many side effects 3 . Moreover, the problem increases due to the continuous mutations in many targets of anticancer drugs which produce some sort of resistance 4 . Recently, medicinal chemists tried to develop new anticancer agents with high selectivity and can overcome the generated drug resistance [5][6][7] .
Vascular endothelial growth factor receptors (VEGFRs) are a group of tyrosine kinases that caught the attention of scientists for the discovery of new anticancer agents 8,9 . VEGFRs group is subdivided into three categories. (i) VEGFR-1 is responsible for the control of angiogenesis in embryos 10 . (ii) VEGFR-2 is the key element in tumour growth due to its crucial role in the formation of new vasculatures and angiogenesis 11 . (iii) VEGFR-3 controls the process of lymphangiogenesis 12 . VEGFR-2 took more interest due to its important role in tumour growth 13 .
In embryonic vasculogenesis, VEGFR1 plays a critical function. VEGFR2 is a protein that controls embryonic and tumour angiogenesis. VEGFR2 has mostly overexpressed in tumour vasculature endothelial cells, with reduced expression in normal endothelial cells 14 . Overexpressed VEGFR-2 is found in a variety of malignancies, including hepatocellular carcinoma and breast cancer [15][16][17] . Blocking VEGFR2 is a viable method for the identification of novel therapies for angiogenesis-dependent cancers 18 . VEGFR2 is now the most significant target of antiangiogenic therapy. During development, VEGFR3 is found in all endothelium, but it is only found in the lymphatic endothelium in adults 19 . It is up-regulated in the microvasculature of tumours and wounds 20 .
Till now, there are many drugs approved by the FDA targeting VEGFR-2 for the treatment of cancer 13 . Sorafenib 1 was approved for the treatment of thyroid cancer, metastatic renal cell cancer, and hepatocellular carcinoma. It is associated with many adverse effects as diarrhoea, renal dysfunction, and cardiovascular problems 21 . Sunitinib II is an oral VEGFR-2 inhibitor that exhibits potent antiangiogenic and antitumor activities. It exhibits a high bioavailability and potency against VEGFR in the nanomolar range. Small-cell lung cancer, GI stromal tumours, breast cancer, acute myelogenous leukaemia, multiple endocrine neoplasia types 2A and 2B, and familial medullary thyroid carcinoma were all treated with sunitinib 22 . Vorolanib III is a multi-target tyrosine kinase inhibitor that has successfully completed phase I studies. At a dose of 200 mg (once daily), it had an acceptable safety profile and a favourable therapeutic benefit for patients with advanced solid tumours 23 . Tivozanib IV is a powerful and highly selective orally accessible tyrosine kinase inhibitor with a long half-life (4 days) that targets VEGFR-1, VEGFR-2, and VEGF-3 at very low dosages 24 .
Four key pharmacophoric features of VEGFR-2 inhibitors have been identified [25][26][27][28][29][30] . Each feature has its own binding area in VEGFR-2's active site. The hinge region is occupied by a flat heteroaromatic moiety that forms one hydrogen bond with Cys917 26 . The second distinguishing feature is a linker moiety that sits between the hinge region and the DFG domain 31 . The pharmacophore moiety, which occupies the DFG domain and forms two crucial hydrogen bonding interactions with Glu883 and Asp1044, is the third feature. At least one H-bond acceptor (HBA) and one H-bond donor (HBD) group (e.g. amide or urea) must be present in the pharmacophore moiety 32 . The fourth feature is a terminal hydrophobic moiety that occupies the allosteric pocket, resulting in numerous hydrophobic interactions 33 ( Figure 1).
As an extension of our efforts to reach potent anti-VEGFR-2 agents 29,34-39 , a new series of 1,3,4-oxadiazole-naphthalene hybrids was synthesised as a modified form of the reported VEGFR-2 inhibitors. The synthesised hybrids were designed to have the main features of VEGFR-2 inhibitors and evaluated to confirm their VEGFR-2 inhibitory activities.

Rationale of molecular design
The 1,3,4-oxadiazole moiety has various biological activities. The wide and potent activity of this moiety made it an important pharmacological scaffold for drug design especially in the field of cancer disease 5,40,41 . Additionally, naphthalene is a main building block in many anticancer agents 42,43 . The molecular hybridisation approach is one of the most interesting and efficient method for the design and discovery of new bioactive agents 44,45 . Depending on this approach and in continuation of our activities in the discovery of VEGFR-2 inhibitors, we synthesised a new series 1,3,4oxadiazole-naphthalene hybrids.
The 2-methoxynaphthalene moiety was used to occupy the hinge region of the VEGFR-2 binding site to validate the main pharmacophoric features of VEGFR-2 inhibitors. The naphthalene moiety's bicyclic structure is well-suited to the large size space of the hinge region 46 . Furthermore, the methoxy group serves as a hydrogen-bond acceptor, facilitating hydrogen bonding interactions in the hinge region. As a linker group, the 2-ethyl-1,3,4-oxadiazole moiety was used. In the linker region, the 1,3,4-oxadiazole moiety contains two nitrogen atoms that can form extra hydrogen bonds. As a pharmacophore moiety, we used an amide group to occupy the DFG region. Finally, different aromatic derivatives can occupy the allosteric hydrophobic region to investigate structureactivity relationships (Figure 1).

Chemistry
Different synthetic pathways were described in Scheme 1 for the preparation of final designed compounds. Firstly, the commercially available S(þ)naproxen 1 ((S)-2-(6-methoxynaphthalen-2-yl)propanoic acid) was esterified using absolute ethanol and conc. H 2 SO 4 to produce the ester form 2. Then, the ester derivative 2 was heated in ethanol and hydrazine hydrate to produce the corresponding hydrazide derivative 3.
The synthesis of the key starting material 1,3,4-oxadiazolyl scaffold 4 was achieved by the reaction of acid hydrazide 3 with carbon disulphide in an alcoholic potassium hydroxide solution. Different chloroacetinilides were obtained by the treatment of aromatic amines with chloroacetyl chloride 47 . The potassium salt of 4 was allowed to react with the substituted chloroacetinilides. Unfortunately, the required products couldn't be separated from the reaction mixture with the desirable purity. The presence of more than one product for each reaction was attributed to the high reactivity and/or basicity of the potassium salt. Alternatively, a less basic condition was adopted and the mercapto-containing structure 4 was directly allowed to react with chloroacetinilides in the presence of sodium acetate. This alternative route successfully afforded the final products in satisfying yields and reasonable purities.
The spectral and elemental analytical data of this group of novel compounds confirmed their structures (see experimental section). In all cases, the characteristic thiol stretching band at 2624 cm À1 vanished in all IR spectra and was replaced by a higher frequency one distinctive band of the primary amide NH moiety. Furthermore, in all IR spectra, typical carbonyl stretching bands between 1655 and 1675 cm À1 were observed. These findings support the idea that the acetanilide moiety is linked to the oxadiazole nucleus via S-linkage. The later NH group revealed a broad singlet, corresponding to one proton around 10.4 ppm. As an example, the 1 H NMR spectrum of compound 7 revealed a broad singlet signal, equivalent to one proton, at 10.36 ppm due to NH, and a multiplet signal, equivalent to ten protons, at 7.74-7.15 ppm due to aromatic protons. Finally, the ethyl linker between the naphthalene and oxadiazole rings produces a multiplet for one proton due (CH) and a doublet for three protons due to (CH 3 ) at 4.53 and 1.78 ppm, respectively. Aliphatic (S-CH2) protons were detected as a singlet signal at 4.01 ppm, a singlet signal equivalent to three protons at 3.81 ppm due to OCH3, and a singlet signal of three protons at 1.95 ppm due to the benzylic methyl moiety. The 13 C NMR spectrum of compound 7 revealed 24 signals.
The 1 H NMR spectra of compounds 12, showed, in addition to naphthalene aromatic protons, broad singlet due to NH group at 8.12 ppm, the ethyl group appeared as multiplet and doublet signals at 4.57 and 1.66 ppm, a singlet for two protons at 4.94 ppm due S-CH 2 group, another singlet for three protons at 3.83 ppm due to O-CH 3 . While the cyclohexyl side chain is represented by a multiplet signal between 1.66 and 1.04 ppm.
In general, mass spectra of compounds-containing halogen atoms showed the typical isotopic distribution patterns. The parent peak intensity of compounds 14 and 15 makes it easy to recognise the expected isotopic pattern of the two chlorine atoms; however, an anticipated isotopic ratio was observed with some fragments-containing chlorine; i.e. the anilinium fragment showed M þ , (Mþ 2) þ , and (M þ 4) þ isotopic peaks as detailed in the experimental section.

Biological testing
2.2.1. In vitro cytotoxic activities MTT assay was used to assess the cytotoxic activities of the synthesised compounds against MCF-7 (human breast cancer cell line) and HepG2 (human hepatocellular carcinoma cell line). And Sorafenib was used as a control drug (Table 1). When compared to sorafenib (IC 50

VEGFR-2 inhibitory assay
The inhibitory effect of the most cytotoxic compounds 5, 8, 15, 16, 17, and 18 on VEGFR-2 was studied using sorafenib as a control drug. VEGFR-2 concentrations after inhibition by the aforementioned compounds against HepG2 cells were summarised in Table 2     (378.7 pg/ml). Compounds 15 and 16 also had moderate effects (864.5 and 834.9 pg/ml, respectively). Compounds 17 and 18, on the other hand, had low effects (1067 and 1004 pg/ml, respectively).

In vitro cytotoxicity against normal cell
The cytotoxic effects of the most active compounds 5, 8, 15, 16, 17, and 18 against normal adult liver epithelial cells (Transformed Human Liver Epithelial-2, THLE-2 cells) were assessed in vitro ( Table 3). The findings revealed that these compounds have low cytotoxicity against the normal THLE-2 cells with IC 50 values of 33.7, 16.7, 34.9, 29.7, 22.8, and 28.6 mM, respectively. While IC 50 value of sorafenib as a reference drug was 27.8 mM. The results showed that these synthesised compounds have low cytotoxicity against the normal cells in comparison to their cytotoxicity against cancer cell lines.

Structure-activity relationship (SAR)
The results of different biological analyses (cytotoxic activity and VEGFR-2 inhibitory assay) gave a valuable SAR. Comparing the cytotoxic activity of compounds 5 (incorporating benzyl moiety as a hydrophobic tail) and 6 (incorporating phenyl moiety as a hydrophobic tail), indicated that benzyl moiety enhances the cytotoxic activity. Modification of compound 6 by insertion of the methyl group at ortho-position (compound 7) gave a mild change in activity. While insertion of the methyl group at meta-position (compound 8) afforded a significant increase in the cytotoxic activity. On the other hand, the methyl substitution of phenyl moiety at para-position (compound 9) produced a dramatic decrease in the cytotoxic activity. Insertion of the methyl group at the two ortho-positions of phenyl ring (compound 10) did not produce a significant change in the cytotoxic activity. Comparing the cytotoxicity of compound 8 (with 3-methylphenyl moiety) and compound 11 (with 3-methoxyphenyl moiety), indicated that methyl moiety is more advantageous than the methoxy group. Insertion of the electron-withdrawing group at para-position (compound 13) gave a mild increase in the cytotoxic activity. Insertion of chloro atom at both 2 and 4 positions of phenyl ring (compound 14) did not produce a significant change in the cytotoxic activity. While insertion of the chloro atom at both 2 and 6 positions of the phenyl ring (compound 15) produced a significant change in the cytotoxic activity. Insertion of the nitro group at ortho-position of phenyl ring (compound 16) or para-position (compound 17) gave a dramatic increase in the biological activity. Modification of compound 17 by insertion of the hydroxyl group at position À2 of the phenyl ring to give compound 18 afforded a mild decrease in activity. Changing the phenyl ring of compound 6 by cyclohexyl moiety (compound 12), produced a dramatic decrease in the cytotoxic activity. This indicated that aromatic hydrophobic moiety is more efficient than not aromatic one.

Cell cycle analysis
Because compound 5 effectively inhibited the growth of HepG2 cells, it was assumed that this inhibitory effect was due to its ability to obstruct cell cycle progression. As a result, the cell cycle process was investigated after HepG2 cells were exposed to compound 5 at a concentration of 8.8 mM (IC 50 value of compound 5). As a control, HepG2 cells were not treated with compound 5. According to flow cytometry data, the percentage of cells arrested in the S phase increased from 27.59% (in control cells) to 42.05% (in compound 5 treated cells). Furthermore, the percentage of HepG2 cells increased from 1.76 to 34.12% during the Pre-G1 phase. In contrast, the percentage of HepG2 cells decreased slightly during the G0/G1 phase, from 51.31 to 43.91%. These findings revealed that compound 5 primarily inhibited HepG2 cell growth during the Pre-G1 phase (Figure 3).

Apoptosis analysis
The apoptosis induced by compound 5 was quantified using an Annexin-V/propidium iodide (PI) staining assay. Compound 5 was applied to HepG2 cells at a concentration of 8.8 mM. As shown in Table 4 and Figure 4, compound 5 had a significantly higher apoptotic effect in HepG2 cells (22.86%) than in control cells (0.51%).
Percentages equal mean ± SEM of three experiments. Ã p < 0.05 indicates a statistically significant difference from the corresponding control (HepG2) group using unpaired t-tests.

Caspase-3 determination
To test the effect of the synthesised compounds on caspase-3 levels, the most promising member 5, at a concentration of 8.8 mM, was applied to the most sensitive cells (HepG2) for 24 h. The results showed that compound 5 significantly increased the level of caspase-3 (5.61-fold) when compared to control cells ( Figure 5).

Docking studies
Docking studies were performed in this study to gain a better understanding of the binding modes of the synthesised compounds into the VEGFR-2 binding site (PDB ID: 2OH4). Sorafenib was used as a control drug. Table 5 displays the binding free energies (DG). Asp1044 and Glu883 have been identified as key amino acids involved in the binding of VEGFR-2 inhibitors 33,48 .
Sorafenib had a binding affinity of À22.46 kcal/mol. Three hydrogen bonding interactions occurred between the urea moiety and Glu883 and Asp1044. Four hydrophobic interactions were formed by the central phenyl group with Val914, Val846, Cys1043, and Phe1045. The hinge region was occupied by the N-methylpicolinamide moiety, which formed one hydrogen bond with Cys917 and five hydrophobic interactions with Val846, Leu838, Leu1033, Phe1045, and Ala864. The allosteric binding pocket was occupied by the terminal 1-chloro-2-(trifluoromethyl)benzene moiety, which formed five hydrophobic interactions with His1024, Ile890, Ile886, and Leu887. It also had one electrostatic interaction with Asp1044 ( Figure 6). The findings revealed that the synthesised compounds have a binding mode similar to that of sorafenib. Table 5 summarises the generated DG (binding energies) against VEGFR-2.
Compounds 17 and 18 exhibited binding energies of À23.78 and À23.40 kcal/mol, respectively. The pharmacophore groups in each compound formed two hydrogen bonds with Glu883 and Asp1044. The 2-ethyl-1,3,4-oxadiazole moiety in each compound formed many hydrophobic interactions in the linker region with extra hydrogen bonding with Cys1043. The 2-methoxynaphthalene moiety of compounds 17 and 18 formed four and seven hydrophobic bonds, respectively. The p-nitrophenyl of compound 17 and 2-hydroxy-4-nitrophenyl of compound 18 were oriented into the allosteric binding pocket-forming two electrostatic interactions with Asp1044 and one hydrophobic interaction with Leu887 (Figures 9 and 10).

In silico ADMET studies
Discovery studio 4.0 was used to investigate the pharmacokinetic properties (ADMET studies) of the synthesised compounds. As a control molecule, sorafenib was used. Table 6 summarises the ADMET parameter values. Compounds 14-18 exhibited a very low level of BBB penetration, while compounds 5-13 had medium to high levels. The aqueous solubility of the synthesised compounds ranged from very low to very low. For absorption parameters, compounds 5-13 showed good levels, while compounds 14-17 were expected to have moderate absorption levels. Compound 18 showed poor absorption. In addition, all the tested compounds were anticipated to have a non-inhibitory effect against CYP2D6, and plasma protein binding ability of more than 90% (Figure 11).

In silico toxicity studies
The synthesised compounds were examined in silico to reach the expected toxicity profile using Discovery Studio software 49,50 . As a   (Table 7).
Except for compound 7, all compounds were predicted to be non-carcinogenic. TD 50 values for compounds 6,7,8,16, and 18 were 20.304, 23.695, 18.535, 22.826, and 21.184 mg/kg body weight/day, respectively, when compared to sorafenib (TD 50 ¼ 17.535 mg/kg body weight/day). Except for compound 18 (rat maximum tolerated dose ¼ 0.132 g/kg body weight), all compounds had lower rat maximum tolerated doses (ranging from 0.042 to 0.053 g/kg body weight) than sorafenib. The rat chronic LOAEL ranged from 0.012 to 0.041 g/kg body weight, which was higher than sorafenib (0.004 g/kg body weight). Furthermore, all compounds were predicted to be non-irritants in skin irritancy models and mild irritants in ocular irritancy models.  Compounds 5,8,15,16,17, and 18 showed VEGFR-2 inhibitory effects with concentrations of 610.4, 583.7, 864.5, 834.9, 1067, and 1004 pg/ml, respectively. SAR study revealed that substitution at the hydrophobic tail with electron-withdrawing is more beneficial than substitution with the electron-donating group for cytotoxicity. Compound 5, the most active counterpart, induced a significant increase in apoptosis (22.86% compared to 0.51% in the control) and arrested the HepG2 cell growth mostly at the Pre-G1 phase. Additionally, compound 5 exerted a significant increase in the level of caspase-3 (5.61-fold). Docking studies revealed that the

In vitro cytotoxic activity
The MCF-7, HepG2, and THLE-2 cells were tested for cytotoxicity using the 3-(4, 5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay [61][62][63][64][65] . The ability of living cells to reduce the yellow product MTT to a blue product, formazan, via a reduction reaction that occurs in the mitochondria is used to assess cell population growth. In this assay, 5000 cells/well were plated in a 96-well plate and allowed to grow for 24 h before being treated with over-mentioned suitable media containing increased concentrations of tested compounds (0, 0.1, 1, 10, 100, and 1000 mM).
Each experiment was conducted in triplicate. After removing the Figure 11. The expected ADMET studies. media, 100 mL of MTT was added to each well and incubated for 4 h. Following that, 100 mL of dimethyl sulfoxide (DMSO) solution was added to solubilise the resulting formazan product, and absorbance at 570 nm was measured using an ELISA microplate reader (Epoc-2 C micro-plate reader, Bio Tek, VT, USA). The IC 50 values [the concentration required to inhibit cell viability by 50%] were calculated.

Cell cycle analysis using flow cytometry
The effect of compound 5 on the cell cycle phases was investigated using propidium iodide staining and flow cytometric analysis according to the cell cycle kit (PN C03551). Briefly, HepG2 cells were allowed to grow in 25 cm 3 flask, then treated with compound 5 for 48 h. After that, the cells were harvested and cell fixation was performed, cells were centrifuged at 2000 rpm for 5 min then, the supernatant was aspirated. The pellet of fixed cells was resuspended in a 0.5 ml cell cycle kit, vortexed, and incubated at 25 C for 15 min. Finally, DNA was stained with 50 mg/ml propidium iodide for 30 min. Flow cytometric analysis of cell cycle performed on a COULTER V R EPICS V R XL TM Flow Cytometer (USA) 67-70 .

Flow cytometric analysis for detection of apoptosis
To assess the effect of compound 5 on cell apoptosis, Annexin V-FITC Kit was used according to the kit protocol (PN IM3546), followed by flow cytometric analysis. In brief, HepG2 cells were allowed to grow in 25 cm 3 flask, after that treated for 48 h. then, washed in phosphate-buffered saline and suspended to 5 Â 10 5 -5 Â 10 6 cells/mL in 1X binding buffer. Then we added to 100 mL of the cell suspensions, 5 mL of dissolved PI, and 1 mL of annexin VFITC solution and incubated in the dark for 15 min. Next to that, we added 400 mL of ice-cold 1X binding buffer and mixed gently. Apoptotic cells were determined by flow cytometric analysis on a COULTER V R EPICS V R XL TM Flow Cytometer (USA) 67,71,72 .

Caspase-3 determination
The effect of compound 5 on the Caspase-3 level was assessed using an ELISA kit (Catalog # KHO1091) according to manufacturer instructions.

ADMET studies
The ADMET descriptors were calculated using Discovery Studio 4.0 in accordance with the reported method 73,74,77 (Supplementary Data).

Toxicity studies
The toxicity potential of the synthesised compounds was predicted using the Discovery studio 4.0 software, as reported in Supplementary Data 78 .