Introduction

The most serious health challenge of humanity is cancer1. It was predicted that about 30 million people will be diagnosed as new cancer patients by 20401. As a result, identifying novel drug targets and developing more selective chemotherapeutic agents are important goals of current drug research2. Although recently developed single target drugs are more selective and safer than awkward conventional anticancer therapeutics, they can probably develop drug resistance3,4. One way to solve this issue is a combination therapy but this can increase undesired side effects and toxicity5. In contrast, several studies proved that multitarget small molecules can overcome this resistance with acceptable safety profile3,4. Compared to combination therapy, multitarget drugs enhance patient compliance, void detrimental off-target effects, decrease drug-drug interactions, and save excessive manufacturing costs5.

There are over five hundred protein kinases in our bodies representing the most abundant family amidst all function proteins. Protein kinases harmonically control vital cellular processes like cell division, proliferation, metabolism, migration, and apoptosis6. Protein kinases deregulation is responsible for a number of fatal illnesses, including cancer2. Accordingly, the inhibition of these deregulated protein kinases is a viable strategy to combat cancer6.

Binding of vascular endothelial growth factor (VEGF) and hepatocyte growth factor (HGF) to their receptors VEGFR-2 and c-Met, respectively results in conformational changes and dimerization of the receptors followed by phosphorylation of multiple tyrosine residues in the intracellular domain2. Consequently, a cascade of intracellular signaling pathways were sparked off2.

Dysregulated VEGFR-2 and c-Met tyrosine kinases (TKs) work together to promote angiogenesis that result in development and progression of different human cancers7. Both tumor growth and metastasis are depend mainly on angiogenesis which is controlled by VEGFR-2 and c-Met signaling8. Consequently, multitarget molecules that can inhibit both VEGFR-2 and c-Met simultaneously may be more effective than single target molecules since they can shut down many signaling pathways implicated in tumor angiogenesis, proliferation, and metastasis7,9. Additionally, drug resistance is more common with single target drugs than multitarget ones9.

Rationale and molecular design

There are four pharmacophoric elements required for type II dual VEGFR-2/c-Met inhibitors namely nitrogenous heterocycle, linker, hydrogen bonding (HB) domain and hydrophobic moiety9. The first is a planar nitrogenous heterocyclic system (in blue) mainly six-membered monocyclic rings like pyridine10,11,12,13, pyrimidine7,10,14, and their fused benzo rings like quinoline10,12, quinazoline9,10,12,15, as well as fused hetero- rings like thienopyridine12,16, thienopyrimidine2, pyrrolopyridine17,18, pyrrolopyrimidine18, triazolopyrazine3, pyrrolotriazine12,19. Quinazoline ring is incorporated in several FDA approved 4-anilinoquinazoline EGFR inhibitors like lapatinib, erlotinib, gefitinib, dacomitinib, vandetanib, afatinib and icotinib as well as an VEGFR-2 inhibitor, cediranib (Fig. 1)12,20.

Figure 1
figure 1

Clinically used quinazoline containing TK inhibitors.

Full size image

The heterocyclic ring occupies the hinge region of both TKs and forms HB via its nitrogen with the highly conserved Cys919 and Met1160 residues in VEGFR-2 and c-Met proteins, respectively. It is worth mentioning that several dual VEGFR-2/c-Met inhibitors contain HB group at α-position of the nitrogenous heterocycle that augment binding of inhibitor to the adenine region of the protein through forming additional HBs (Fig. 2). This α-HB group may be amino7,21, oxo9,22, amido19, carbamoyl11, triazolyl10,11, ureido10,19, or even the pyrrole nitrogen of pyrrolopyridine system23. In addition to HB of pyrimidine nitrogen with Cys919 and Met1160 in the adenine pocket of VEGFR-2 and c-Met TK, respectively, α-ureido group of compound II formed extra two HB with Cys919 in the hinge region of VEGFR-2 TK. Moreover, compound II formed HB with Met1160 in the hinge region of c-Met TK10. Carboxamido group of altiratinib III formed extra HB with Met1160 in the hinge region of c-Met TK24. α-Amino group in compound V act as hydrogen bond donor (HBD) forming HB with Met1160 in the hinge region of c-Met TK. Contrary, the N-methylated analogue of compound V showed no activity against both VEGFR-2 and c-Met TK7. The authors purposed that there is a relationship between HB of the α-amino with the adenine region of TKs and the inhibitory activity of the compound V7. α-Amino group of VII BMS-777607 act as HBD forming HB with Met1160 in the hinge region of c-Met TK25. Amino group of compound IX forms extra HB with Cys919 and Met1160 in the hinge region of VEGFR-2 and c-Met TKs, respectively14. Triazole ring of compound X formed extra HB with Cys919 in the hinge region of VEGFR-2 TK. In addition, It also form new HB with Tyr1159 in adenine pocket of c-Met TK11. Tang et al., synthesized several series of pyrrolo[2,3-b]pyridine c-Met inhibitors with considerable VEGFR-2 inhibitory activity23,26. Nitrogen of pyrrole ring in XI act as HBD forming HB with Met1160 in the hinge region of c-Met TK26.

Figure 2
figure 2

Some reported dual VEGFR-2/c-Met inhibitors with α-HB group (in green).

Full size image

The second pharmacophoric element is a hydrophobic linker (in black) which connects nitrogenous heterocycle and HB domain. This liker has little importance in protein binding, but it puts other pharmacophoric elements in their correct position and orientations27,28. The third is a HB domain (in red) which forms HB with both Asp1046 and Glu885 of the VEGFR-2 TK as well as both Asp1222 and Lys1110 of the c-Met TK. In the type II TKIs, the main difference between VEGFR-2 inhibitors and dual VEGFR-2/c-Met inhibitors is the relatively longer HB domain of dual inhibitors7,12. VEGFR-2 inhibitors contain 2–3 atom HB domain like carboxamide (e.g., sunitinib and nintedanib) or urea (e.g., sorafenib, regorafenib and lenvatinib)28. On the other hand, the dual VEGFR-2/c-Met inhibition is originated from the presence of an HB domain with 4 or more atoms7,12. HB domain of dual VEGFR-2/c-Met is usually a dicarboxamide structure which can be a widespread cyclopropane-1,1-dicarboxamide like clinically used cabozantinib and foretinib as well as hundreds of other investigated compounds2,4,12,14,15,19,24. The dicarboxamide structure can be also involved in heteroaromatic anilides like N-phenyl-2-oxoimidazolidine-1-carboxamide29, N-phenyl-3-oxo-2,3-dihydro-1H-pyrazole-4-carboxamide (e.g., compound IV)4,11, N-phenyl-2-oxo-1,2-dihydropyridine-3-carboxamide (e.g., VII BMS-777607 and merestinib)3,25, N-phenyl-4-oxo-1,4-dihydropyridine-3-carboxamide (e.g., VIII BMS-794833) (Fig. 3)18. Finally, N-acyl(thio)urea can surrogate a dicarboxamide structure and exhibited a dual VEGFR-2/c-Met inhibitory activity (Fig. 4)16,30,31,32,33,34. The forth is a hydrophobic moiety (in purple) that occupies allosteric pocket of the TKs3,4,6,35.

Figure 3
figure 3

HB domain requirements for dual VEGFR-2/c-Met TKs inhibition over VEGFR-2 single inhibition.

Full size image
Figure 4
figure 4

Some reported N-acyl(thio)urea containing c-Met and dual VEGFR-2/c-Met inhibitors.

Full size image

Consequently, cabozantinib was selected as a lead compound, building up the pharmacophore for dual VEGFR-2/c-Met inhibition through incorporation of N-acylthiourea moiety as HB domain that mimics dicarboxamide structure of cabozantinib Fig. 53. Quinazolin-2,4(1H,3H)-dione ring was chosen to occupy adenine pocket of the TKs. α-Oxo moiety was involved in several reported TKIs like dovitinib, nintedanib, orantinib, and sunitinib36. Therefore, α-oxo moiety of target compounds was designed to augment HB and increase the affinity for both the VEGFR-2 and c-Met TKs. Finally, we different phenyl and heteroaryl rings with variable substituents were designed to occupy allosteric pocket.

Figure 5
figure 5

Design of new quinazoline dual VEGFR-2/c-Met inhibitors.

Full size image

Result and discussion

Chemistry

The synthetic pathway of 3-phenylquinazolin-2,4(1H,3H)-dione derivatives 2a,b, 3a-h and 4a,b is depicted in Scheme 137,38. The first step involves the synthesis of intermediate benzoyl isothiocyanate by the reaction of 4-(2,4-dioxo-1,4-dihydro-2H-quinazolin-3-yl)-benzoyl chloride 1 and ammonium thiocyanate under reflux. In the second step, the latter intermediate got readily transformed to the final products by nucleophilic attack of the amino groups of various used reagents such as urea, thiourea, aromatic and hetero amines via a single pot three-component reaction.

Scheme 1
scheme 1

Synthetic routes of 3-phenylquinazolin-2,4(1H,3H)-dione derivatives 2a,b, 3a-h and 4a,b. Reagent and conditions: NH4SCN, (thio)urea or Ar-NH2, TEA, dioxane, reflux.

Full size image

The benzoyl isothiocyanate has been employed in synthesis of some new six membered heterocyclic skeletons like thioxo-triazin-2-one and triazine-2,4-dithione, by nucleophilic addition with urea and thiourea, respectively. Herein, an equimolar reaction of benzoyl isothiocyanate with urea and/or thiourea under reflux for 6–8 h, yields an intermediate which underwent cyclodehydration to furnish compounds 2a,b, respectively. The success of cyclization by water elimination was supported by spectral data of the resulting triazines 2a,b. Their IR spectra showed absorption bands at (3406, 3226), (1716, 1655) and 1237 cm−1 for NH’s, C=O’s, and C=S, respectively. The 1H-NMR spectrum of triazine 2a exhibited their presence of two tautomeric forms from the downfield signals at 7.98, and 8.10 ppm for the NH protons. In the same manner, 1H-NMR of triazine 2b represented 8.06 and 8.08 for NH protons.

The formation of the afforded N-benzoyl-N′-phenylthiourea derivatives 3a-h was carried out by the nucleophilic reaction of different aromatic amines with benzoyl isothiocyanate in dioxane and few drops of TEA for 6–10 h. Their chemical structures were supported by IR and 1H-NMR spectra. For example, the IR bands of compound 3a illustrated the presence of NH, C=O’s and C=S at 3243, 1718, 1663 and 1268, respectively. In addition, 1H-NMR spectrum of compound 3a exhibited three singlet signals for NH protons at 12.63, 11.72, and 11.66 ppm, respectively, along with aromatic protons at 7.26–8.11 ppm.

Similarly, the intermediate benzoyl isothiocyanate reacts with some heterocyclic amines namely 5-amino-1,3,4-thiadiazole-2-thiol and 2-aminobenzothiazole under reflux for 10–12 h to furnish thiadiazole and benzothiazole 4a,b, respectively.

Biology

In vitro antiproliferative activity against HCT-116

HCT-116 colorectal cancer cell line is characterized by overexpression of both VEGFR-2 and c-Met TKs. Consequently, we selected it to study the cytotoxic activity of the target derivatives9. We tested the effect of several concentrations of the designed derivatives on HCT-116 cells by using MTT assay. Table 1 represents the IC50 of all synthesized derivatives as well as that of the reference drug, cabozantinib.

Table 1 In vitro cytotoxicity of compounds 2a,b, 3a–h, 4a,b, and cabozantinib against HCT116 cell line.
Full size table

The designed compounds showed noticeable antiproliferative activities. 5 out of 12 derivatives namely, 2b, 3a, 3c, 3d, and 3e exhibited noteworthy cytotoxic activity (IC50 = 1.184–9.379 µM) that was more superior than that of the positive reference, cabozantinib. Additionally, compounds 4a and 4b (IC50 = 19.90 and 20.27 µM, respectively) displayed comparable cytotoxic activities to cabozantinib. The rest of target compounds, 2a, 3b, 3g, and 3h exhibited acceptable cytotoxic activity (IC50 = 27.030–53.39 µM). Compound 3f showed the least cytotoxic activity (IC50 = 113.500 µM). Compound 3c (with p-carboxyl moiety) represented higher cytotoxic activity than compound 3b (with m-carboxyl moiety) that suggests the regioisomerism effect on cytotoxic activity with p-substitution is superior to m-substitution. Apart from 3f, compounds with electron withdrawing group on phenyl ring represented higher cytotoxic activity as compared to unsubstituted phenyl.

In vitro toxicity against normal cells

The cytotoxic activity of selected two compounds namely, 3c and 3e against WI38 normal cell line was evaluated to study the safety of the designed compounds to the normal cell (Table 2). Although, both tested compounds exhibited higher cytotoxicity against normal cells than cabozantinib, they are better in terms of selectivity index. Compound 3c showed 20 times selective cytotoxicity against the colorectal cancer cells over the normal cells. Moreover, compound 3e showed more than 3 times selective cytotoxicity against the colorectal cancer cells over the normal cells. Compounds 3c and 3e as well as cabozantinib exhibited considerable safety as the values of SI of all of them are more than 239.

Table 2 In vitro cytotoxicity of compounds 3c and 3e against WI38 normal cell line.
Full size table

In vitro activity against c-Met and VEGFR-2 tyrosine kinases

Based on our rationale design of dual VEGFR-2/c-Met inhibition, we selected the most active cytotoxic derivatives to the HCT-116 cancer cells for scrutinization of their inhibition activity against both VEGFR-2 and c-Met TKs. Table 3 represents the inhibition activity of tested derivatives as well as cabozantinib. All tested derivatives showed acceptable inhibitory activity against the target TKs. Compound 3d exhibited comparable activity to cabozantinib against VEGFR-2 enzyme with (IC50 = 51 nM). In addition, compound 3e showed noticeable inhibitory activity against VEGFR-2 enzyme with (IC50 = 83 nM). Regarding the c-Met TK, both compounds 3c and 3e represented inhibition activity against c-Met TK with IC50 values 74 and 48 nM, respectively.

Table 3 Inhibitory activity of compounds 2b, and 3c-e against VEGFR-2 and c-Met TKs.
Full size table

Apoptosis assay

To go deeper in the mechanism of designed derivatives, the cell cycle arrest study of compounds 3c and 3e was conducted by using Annexin V-FITC/PI staining (Table 4 and Fig. 6). HCT-116 cells were treated by cabozantinib and the two tested compounds at their IC50 concentrations. Compared with the control group, HCT-116 cell cycle was blocked in G0/G1 phase after treatment with 3c (55.41%). On another hand, compound 3e (26.51%) had a higher ability to enhance the population of HCT-116 cells in G2/M process than cabozantinib (24.72%).

Table 4 Cell cycle analysis in HCT-116 colon cancer cell line treated with compounds 3c and 3e.
Full size table
Figure 6
figure 6

Flow cytometric analysis of cell cycle phases after treatment with 3c, 3e and cabozantinib.

Full size image

Moreover, we studied the apoptosis mechanism of HCT-116 that was induced by tested derivatives 3c and 3e as well as cabozantinib (Table 5 and Fig. 7). There was an increase in the number of both early and late apoptotic cells of both tested compounds 3c and 3e. The number of late apoptotic cells of 3e (35.39%) was higher than that of reference drug, cabozantinib (32.13%). Also, the number of late apoptotic cells of 3c (19.28%) was approximately equal to that of cabozantinib (19.61%). Totally, both designed compounds 3c and 3e were able to induce remarkable apoptosis. Moreover, compound 3e succeeded in apoptosis induction more efficiently than cabozantinib (Fig. 8).

Table 5 Apoptosis induction analysis for compounds 3c, 3e and cabozantinib.
Full size table
Figure 7
figure 7

Effect of 3c, 3e and cabozantinib on HCT-116 cells in annexin V-FITC staining test.

Full size image
Figure 8
figure 8

Representative cytograms of apoptotic HCT-116 cells induced by 3c and 3e compared to cabozantinib for 24 h.

Full size image

In silico studies

Molecular modeling studies

Molecular docking of the target compounds was carried out in the active site of c- VEGFR-2 (PDB: 4ASD) and c-Met (PDB: 3LQ8) TKs. Docking styles of compounds 3c and 3e in the active site of VEGFR-2 TK are shown in Figs. 9 and 10, respectively. The thiourea group of compound 3c showed HB with a highly conserved residue Asp1046 in the HB region and α-oxo group of quinazoline formed HB with backbone amide of Phe918 in the hinge region. Moreover, compound 3c exhibited hydrophobic interactions with both Leu840 and HB with Lys868. 4-Oxo moiety of compound 3e declared HB with the conserved residue Cys919 in the hinge region. Compound 3e showed another two HB with Asp1046 and Lys868 in addition to hydrophobic interaction with Leu84027,28.

Figure 9
figure 9

Docking style of compound 3c with VEGFR-2 TK (PDB: 4ASD).

Full size image
Figure 10
figure 10

Docking style of compound 3e with VEGFR-2 TK (PDB: 4ASD).

Full size image

On the other hand, Figs. 11 and 12 represent docking representations of compounds 3c and 3e it the active site of c-Met TK. α-Oxo moiety of quinazoline ring of compound 3c showed HB with highly conserved residue Met1160 in the hinge region as well as thiourea group formed HB with the another highly conserved residue Lys1110 in the HB region. Further, compound 3c declared hydrophobic interaction with Ile1084. N-acylthiourea group of compound 3e exhibited dual HB with both Lys 1110 and Asp1222 in addition to hydrophobic interactions with both Ile1084 and Met1211 in the hinge region.

Figure 11
figure 11

Docking style of compound 3c with c-Met TK (PDB: 3LQ8).

Full size image
Figure 12
figure 12

Docking style of compound 3e with c-Met TK (PDB: 3LQ8).

Full size image

In silico prediction of physicochemical and pharmacokinetic properties

To reach advanced clinical phases, drug candidates have to possess an acceptable pharmacokinetic profile. Therefore, the physicochemical and pharmacokinetic properties of target derivatives were predicted using SwissADME (Tables 6, 7 and Fig. 13). All target compounds showed good druglikeness with no Lipinski violations except compound 3h. Regarding Abbott bioavailability score, most target derivatives have accepted oral absorption. All target compounds as well as cabozantinib were predicted not to cross BBB. Target compounds were predicted to have little effect on CYP450 enzymes like CYP1A2, CYP2C19, CYP2C9, CYP2D6 and CYP3A4 which minimizes the suspected drug-drug interactions.

Table 6 Physicochemical parameters and drug-likeness of target compounds and cabozantinib.
Full size table
Table 7 Solubility and pharmacokinetics of target compounds and cabozantinib.
Full size table
Figure 13
figure 13

Rader for 3c, 3e and cabozantinib.

Full size image

Conclusion

In brief, based on the structure of a dual VEGFR-2/c-Met inhibitor, cabozantinib, we designed and synthesized a new series of novel 3-phenylquinazolin-2,4(1H,3H)-diones with (thio)urea scaffold. The cytotoxic activity of the target compounds was conducted against HCT-116 colorectal cancer cell line. Compounds contain electron-withdrawing groups on phenyl ring at position-4, declared the highest cytotoxic activity. The inhibitory activity was performed against both VEGFR-2 and c-Met TKs. α-Oxo moiety in quinazoline ring form HB with Met1160 residue in the adenine region of c-Met TK. Compound 3e has the highest inhibitory activity against both VEGFR-2/c-Met (IC50 = 83 and 48 nM, respectively).

Experimental

Chemistry

All reactions were observed on silica gel GF254 plate with thin layer chromatography (TLC). Melting points (uncorrected) were recorded on an electrothermal melting apparatus. FT-IR spectra were recorded on a Shimadzu 8101 PC spectrometer. 1H- and 13C-NMR spectra were determined on a Varian Mercury 400 MHz spectrophotometer in DMSO-d6. Chemical shifts are reported in parts per million with tetramethylsilane as an internal standard and are given in δ units. Electron impact mass spectra were obtained at 70 eV using a GCMS-QP 1000 EX spectrometer. Elemental analyses were carried out at the microanalytical center at Cairo University.

Synthesis of 4-(2,4-dioxo-1,4-dihydro-2H-quinazolin-3-yl)-benzoyl chloride 1

4-(2,4-dioxo-1,4-dihydro-2H-quinazolin-3-yl)-benzoic acid (0.013 mol, 1 gm) was dissolved in thionyl chloride, then the reaction mixture was refluxed for 5 h. After completion of the reaction, the excess of thionyl chloride was evaporated and the residue was allowed to stand at room temperature, then recrystallized from benzene to yield the product as yellow crystal. Yield: 85%. M.P: 270 °C. FT-IR (KBr, υ, cm−1) = 3195 (NH), 1763, 1725 (C = O’s), 751 (C–Cl). 1H-NMR (DMSO-d6, 400 MHz): δ (ppm) = 11.65 (s, 1H, NH), 7.22–7.96 (m, 4H, Ar–H (quinazoline)), 7.49–8.07 (dd, 4H, Ar–H). 13C-NMR (DMSO): δ 114.74, 115.77, 123.09, 127.52, 128.04, 129.99, 130.30, 131.95, 135.81, 140.30, 150.43, 162.57, 167.35. MS (EI) (RT: 3.56–3.58 min): m/z = 300.22 (obs.), 300.03 (expected) [M+], 302.22 (obs.), 302.27 (expected) [M+ + 2]. Anal. Calcd for C15H9ClN2O3: C, 59.91; H, 3.02; Cl, 11.79; N, 9.32%. Found C, 60.16; H, 3.26; Cl, 11.84; N, 9.18%.

General procedures for the synthesis of 3-[4-(6-mercapto-4-oxo/thioxo-1,4-dihydro-[1,3,5]-triazin-2-yl)-phenyl]-1H-quinazolin-2,4-dione 2a,b

4-(2,4-dioxo-1,4-dihydro-2H-quinazolin-3-yl)-benzoyl chloride 1 (0.003 mol, 0.32 g) was allowed to react with ammonium thiocyanate (0.003 mol, 0.32 g) in dioxane (20 ml) under reflux for 2 h. After completion of the reaction, the formed precipitate of ammonium chloride was removed, then urea and/or thiourea was added to the filtrate in presence of few drops of TEA under reflux for 6 h. The solid formed was filtered off, dried, and recrystallized from ethanol to afford compounds 2a, b, respectively.

3-(4-(4-oxo-6-thioxo-1,4,5,6-tetrahydro-1,3,5-triazin-2-yl)phenyl)quinazolin-2,4(1H,3H)-dione 2a

Brown crystals. Yield 60%. MP > 300 °C. FT-IR (KBr, υ, cm−1) = 3406, 3226, 3060 (NH’s), 1717, 1654 (C = O’s), 1610 (C=N), 1280 (C=S). 1H-NMR (DMSO-d6, 400 MHz): δ (ppm) = 11.64 (s, 1H, NH), 8.10 (s, 1H, NH), 7.98 (s, 1H, NH), 7.25–7.75 (m, 4H, Ar–H (quinazoline), 7.42–7.96 (dd, 4H, Ar–H). 13C-NMR (DMSO): δ 114.76, 115.75, 123.06, 128.05, 128.47, 129.52, 129.71, 129.78, 134.63, 135.78, 138.78, 140.31, 150.50, 162.60, 167.94. MS (EI) (RT: 2.90–3.00): m/z = 365.00 (obs.), 365.37 (expected) [M+], 367.00 (obs.), 367.37 (expected) [M+ + 2]. Anal. Calcd for C17H11N5O3S: C, 55.88; H, 3.03; N, 19.17; S, 8.78%. Found C, 55.99; H, 3.14; N, 19.10; S, 8.88%.

3-(4-(4,6-dithioxo-1,4,5,6-tetrahydro-1,3,5-triazin-2-yl)phenyl)quinazolin-2,4(1H,3H)-dione 2b

Pale brown crystals. Yield 60%. MP > 300 °C. FT-IR (KBr, υ, cm−1) = 3406, 3235, 3063 (NH's), 1716, 1655 (C=O’s), 1607 (C=N), 1275, 1237 (C=S’s). 1H-NMR (DMSO-d6, 400 MHz): δ (ppm) = 11.64 (s, 1H, NH), 8.08 (s, 1H, NH), 8.06 (s, 1H, NH), 7.23–8.04 (m, 4H, Ar–H (quinazoline)), 7.42–7.98 (dd, 4H, Ar–H). 13C-NMR (DMSO): δ 114.76, 115.75, 123.06, 128.05, 128.48, 129.72, 132.45, 134.64, 135.68, 138.48, 140.30, 150.22, 162.17, 167.96, 182.55. MS (EI) (RT: 1.60–1.63): m/z = 381.00 (obs.), 381.44 (expected) [M+], 383.00 (obs.), 383.44 (expected) [M+ + 2]. Anal. Calcd for C17H11N5O2S2: C, 53.53; H, 2.91; N, 18.36; S, 16.81%. Found C, 53.66; H, 3.09; N, 18.25; S, 16.97%.

General procedures for the synthesis of arylquinazolin-2,4-diones 3a-h

Multi-component reaction of 4-(2,4-dioxo-1,4-dihydro-2H-quinazolin-3-yl)-benzoyl chloride 1 with ammonium thiocyanate and aromatic amines namely aniline, m-aminobenzoic acid, p-aminobenzoic acid, ethyl 4-aminobenzoate, p-chloroaniline, p-nitroaniline, p-aminophenol, and 4-amino-N-pyrimidin-2-yl-benzenesulfonamide in dioxane (20 ml) and few drops of TEA was heated under reflux for 10–12 h (monitored by TLC) to afforded compounds 3a–h, respectively.

4-(2,4-dioxo-1,4-dihydroquinazolin-3(2H)-yl)-N-(phenylcarbamothioyl)benzamide 3a

White crystals. Yield 75%. MP 210 °C. FT-IR (KBr, υ, cm−1) = 3243 (NH), 1716, 1663 (C=O’s), 1268 (C=S). 1H-NMR (DMSO-d6, 400 MHz): δ (ppm) = 12.63 (s, 1H, NH), 11.72 (s, 1H, NH), 11.66 (s, 1H, NH), 7.26–8.05 (m, 9H, Ar–H), 7.72–8.11 (dd, 4H, Ar–H). 13C-NMR (DMSO): δ 114.78, 115.79, 123.10,124.89, 126.85, 128.06, 128.81, 129.18 129.71, 129.80, 132.45, 135.84, 140.33, 140.48, 150.43, 162.59, 168.27, 179.58. MS (EI) (RT: 4.47–4.50): m/z = 416.34 (obs.), 416.34.46 (expected) [M+]. Anal. Calcd for C22H16N4O3S: C, 63.45; H, 3.87; N, 13.45; S, 7.70%. Found C, 63.56; H, 4.01; N, 13.35; S, 7.83%.

3-(3-(4-(2,4-dioxo-1,4-dihydroquinazolin-3(2H)-yl)benzoyl)thioureido)benzoic acid 3b

Pale brown crystals. Yield 75%. MP: > 300 °C. FT-IR (KBr, υ, cm−1) = 3405 (COOH), 3222, 3061 (NH's), 1716, 1656 (C=O’s), 1279 (C=S). 1H-NMR (DMSO-d6, 400 MHz): δ (ppm) = 12.70 (s, 1H, COOH), 11.65 (s, 1H, NH), 11.61 (s, 1H, NH), 11.40 (s, 1H, NH), 7.23–8.12 (m, 8H, Ar–H), 7.42–7.98 (dd, 4H, Ar–H). 13C-NMR (DMSO): δ 114.75, 115.75, 123.06, 123.15, 128.05, 128.48, 129.52, 129.72, 129.78, 129.86, 134.63, 135.77, 138.79, 140.30, 150.51, 162.61, 167.97, 179.94, 182.54. MS (EI) (RT: 2.73–2.76): m/z = 460.48 (obs.), 460.47 (expected) [M+]. Anal. Calcd for C23H16N4O5S: C, 59.99; H, 3.50; N, 12.17; S, 6.96%. Found C, 60.12; H, 3.60; N, 12.09; S, 7.11%.

4-(3-(4-(2,4-dioxo-1,4-dihydroquinazolin-3(2H)-yl)benzoyl)thioureido)benzoic acid 3c

Orange crystal crystals. Yield 70%. MP 258 °C. FT-IR (KBr, υ, cm−1) = 3355 (COOH), 3044 (NH), 1721, 1681 (C=O’s), 1264 (C=S). 1H-NMR (DMSO-d6, 400 MHz): δ (ppm) = 12.80 (s, 1H, COOH), 12.68 (s, 1H, NH), 11.82 (s, 1H, NH), 11.65 (s, 1H, NH), 7.25–8.16 (m, 8H, Ar–H), 7.52–8.02 (dd, 4H, Ar–H). 13C-NMR (DMSO): δ 113.01, 114.77, 115.79, 117.30, 123.10, 124.04, 128.05, 128.57, 129.85, 130.42, 131.69, 132.39, 140.54, 142.43, 150.43, 162.59, 167.20, 168.18, 179.74. MS (EI) (RT: 4.09–4.40): m/z = 460.48 (obs.), 460.47(expected) [M+]. Anal. Calcd for C23H16N4O5S: C, 59.99; H, 3.50; N, 12.17; S, 6.96%. Found C, 60.12; H, 3.60; N, 12.09; S, 7.11%.

Ethyl4-(3-(4-(2,4-dioxo-1,4-dihydroquinazolin-3(2H)-yl)benzoyl)thioureido)benzoate 3d

White crystals. Yield 80%. MP > 300 °C. FT-IR (KBr, υ, cm−1) = 3359, 3065, 2932 (NH's), 1715, 1666 (C=O’s), 1270 (C=S). 1H-NMR (DMSO-d6, 400 MHz): δ (ppm) = 11.65 (s, 1H, NH), 11.45 (s, 1H, NH) 10.72 (s, 1H, NH), 7.24–8.00 (m, 8H, Ar–H), 7.54–8.08 (dd, 4H, Ar–H), 4.35 (q, 2H, CH2), 1.36 (t, 3H, CH3). 13C-NMR (DMSO): δ 14.72, 60.98, 114.78, 115.78, 120.02, 123.08, 125.11, 128.06 128.87, 129.80, 130.59, 134.96, 135.81, 139.36, 140.34, 144.07, 150.49, 162.62, 165.25, 165.84, 166.16. MS (EI) (RT: 1.91–1.95): m/z = 488.05 (obs.), 488.53 (expected) [M+]. Anal. Calcd for C25H20N4O5S: C, 61.47; H, 4.13; N, 11.47; S, 6.56%. Found C, 61.56; H, 4.29; N, 11.35; S, 6.66%.

N-((4-chlorophenyl)carbamothioyl)-4-(2,4-dioxo-1,4-dihydroquinazolin-3(2H)-yl) benzamide 3e

Pale violet crystals. Yield 81%. MP 250 °C. FT-IR (KBr, υ, cm−1) = 3323, 3248 (NH's), 1724, 1671 (C=O’s), 1245 (C=S). 1H-NMR (DMSO-d6, 400 MHz): δ (ppm) = 11.67 (s, 1H, NH), 10.60 (s, 1H, NH), 10.36 (s, 1H, NH), 7.29–8.11 (m, 8H, Ar–H), 7.43–7.97 (dd, 4H, Ar–H). 13C-NMR (DMSO): δ 114.74, 115.79, 123.05, 128.02, 128.48, 129.02, 129.41, 129.50, 129.74, 132.17, 134.61, 135.76, 138.78, 140.33, 150.47, 162.60, 167.47, 167.93. MS (EI) (RT: 1.74–1.76): m/z = 450.33 (obs.), 450.91 (expected) [M+], 452.33 (obs.), 452.91 (expected) [M+ + 2]. Anal. Calcd for C22H15ClN4O3S: C, 58.60; H, 3.35; Cl, 7.86; N, 12.43; S, 7.11%. Found C, 58.74; H, 3.46; Cl, 7.91; N, 12.38; S, 7.22%.

4-(2,4-dioxo-1,4-dihydroquinazolin-3(2H)-yl)-N-((4-nitrophenyl)carbamothioyl) benzamide 3f.

Brown crystals. Yield 65%. MP 255 °C. FT-IR (KBr, υ, cm−1) = 3393, 3247, 3071 (NH's), 1718, 1667 (C = O's), 1237 (C = S). 1H-NMR (DMSO-d6, 400 MHz): δ (ppm) = 11.90 (s, 1H, NH), 11.64 (s, 1H, NH), 11.40 (s, 1H, NH), 7.24–7.97 (m, 8H, Ar–H), 7.50–8.06 (dd, 4H, Ar–H). 13C-NMR (DMSO): δ 114.77, 115.79, 120.61, 122.10, 123.08, 125.06, 128.05, 129.69, 129.78, 132.55, 135.80, 135.82, 139.75, 140.34, 150.41, 162.57, 167.75, 182.55. MS (EI) (RT: 5.60–5.85): m/z = 459.00 (obs.), 459.46 (expected) [M+-2], 461.00 (obs.), 461.46 (expected) [M+]. Anal. Calcd for C22H15N5O5S: C, 57.26; H, 3.28; N, 15.18; S, 6.95%. Found C, 57.40; H, 3.40; N, 15.09; S, 7.08%.

4-(2,4-dioxo-1,4-dihydroquinazolin-3(2H)-yl)-N-((4-hydroxyphenyl)carbamothioyl) benzamide 3g

Dark brown crystals. Yield 75%. MP 260 °C. FT-IR (KBr, υ, cm-1) = 3386 (OH), 3246, 3067, 2919 (NH's), 1716, 1665 (C=O's), 1268 (C=S). 1H-NMR (DMSO-d6, 400 MHz): δ (ppm) = 12.44 (s, 1H, OH), 11.64 (s, 1H, NH), 11.62 (s, 1H, NH), 11.40 (s, 1H, NH), 6.81–8.10 (m, 8H, Ar–H), 7.50–8.06 (dd, 4H, Ar–H). 13C-NMR (DMSO): δ 114.78, 115.57, 115.78, 120.04, 123.09, 126.47, 128.05, 128.48, 129.70, 129.78, 132.55, 135.82, 140.33, 150.42, 162.57, 167.76, 168.10, 182.54. MS (EI) (RT: 3.40–3.75): m/z = 432.00 (obs.), 432.45 (expected) [M+]. Anal. Calcd for C22H16N4O4S: C, 61.10; H, 3.73; N, 12.96; S, 7.41%. Found C, 61.22; H, 3.85; N, 12.85; S, 7.55%.

4-(2,4-dioxo-1,4-dihydroquinazolin-3(2H)-yl)-N-((4-(N-(pyrimidin-2-yl)sulfamoyl) phenyl)carbamothioyl)benzamide 3h

Pale brown crystals. Yield 80%. MP 300 ºC. FT-IR (KBr, υ, cm−1) = 3397, 3094, 2920 (NH's), 1718, 1666 (C=O’s), 15,835 (C=N), 1337 (O=S=O), 1263 (C=S). 1H-NMR (DMSO-d6, 400 MHz): δ (ppm) = 12.76 (S, 1H, NH), 11.83 (s, 1H, NH), 11.63 (s, 1H, NH), 11.40 (s, 1H, NH), 7.08–8.55 (m, 11H, Ar–H + pyrimidine protons), 7.50–8.05 (dd, 4H, Ar–H). 13C-NMR (CDCl3): δ 114.77, 115.78, 122.10, 123.09, 123.85, 124.35, 127.57, 128.05, 128.75, 129.71, 129.78, 129.86, 132.55, 135.82, 140.32, 140.35, 142.75, 150.42, 162.57, 167.76, 182.54. MS (EI) (RT: 2.50–2.75): m/z = 573.00 (obs.), 573.54 (expected) [M+], 575.00 (obs.), 575.54 (expected) [M+ + 2]. Anal. Calcd for C26H19N7O5S2: C, 54.44; H, 3.34; N, 17.09; S, 11.18%. Found C, 54.66; H, 3.64; N, 17.00; S, 11.33%.

General procedures for the synthesis of heterylquinazolin-2,4-diones 4a-b

A mixture of 4-(2,4-dioxo-1,4-dihydro-2H-quinazolin-3-yl)-benzoyl chloride 1 (0.005 mol, 1.5 gm) and ammonium thiocyanate (0.005 mol, 0.39 gm) in dioxane (20 ml) and few drops of TEA was refluxed for 2 h. Afterward, the formed ammonium chloride precipitate was removed by filtration, then hetero amines namely 5-amino-[1,3,4]-thiadiazole-2-thiol, and 2-aminobenzothiazole (0.005 mol) were added to the filtrate and the reaction mixture was refluxed for 8–10 h. The formed solids were filtered off, purified by recrystallization using benzene/ethanol to furnish compounds 4a-b, respectively.

4-(2,4-dioxo-1,4-dihydroquinazolin-3(2H)-yl)-N-((5-mercapto-1,3,4-thiadiazol-2-yl)carbamothioyl)benzamide 4a

Yellow crystals. Yield 82%. MP > 300 °C. FT-IR (KBr, υ, cm−1) = 3408, 3212, 3064 (NH's), 2723 (SH), 1717, 1655 (C=O’s), 1608 (C=N), 1275 (C=S). 1H-NMR (DMSO-d6, 400 MHz): δ (ppm) = 13.16 (s, 1H, SH), 11.68 (s, 1H, NH), 11.63 (s, 1H, NH), 11.61 (s, 1H, NH), 7.24–8.08 (m, 4H, Ar–H(quinazoline)), 7.42–7.98 (dd, 4H, Ar–H). 13C-NMR (DMSO): δ 114.76, 115.75, 123.06, 128.05, 128.47, 129.52, 130.28, 131.04, 134.64, 135.77, 138.78, 140.30, 150.43, 162.60, 167.35, 167.94. MS (EI) (RT: 3.30–3.50): m/z = 455.00 (obs.), 455.53 (expected) [M+-1], 456.00 (obs.), 456.53 (expected) [M+]. Anal. Calcd for C18H12N6O3S3: C, 47.36; H, 2.65; N, 18.41; S, 21.07%. Found C, 47.50; H, 2.78; N, 18.32; S, 21.21%.

N-(benzo[d]thiazol-2-ylcarbamothioyl)-4-(2,4-dioxo-1,4-dihydroquinazolin-3(2H)-yl)benzamide 4b

Brown crystals. Yield 75%. MP > 300 °C. FT-IR (KBr, υ, cm−1) = 3406, 3234, 3063 (NH's), 1716, 1655 (C=O’s), 1608 (C=N), 1280 (C=S). 1H-NMR (DMSO-d6, 400 MHz): δ (ppm) = 11.57 (s, 2H, 2NH), 11.40 (s, 1H, NH), 7.23–8.08 (m, 8H, Ar–H), 7.42–7.98 (dd, 4H, Ar–H). 13C-NMR (CDCl3): δ 114.76, 115.75, 123.06, 124.05, 128.05, 128.46, 129.52, 129.71, 129.78, 132.51, 134.63, 135.08, 135.78, 138.78, 140.30, 146.19, 150.42, 150.50, 162.60, 167.76, 167.93. MS (EI) (RT: 1.90–2.10): m/z = 473.00 (obs.), 473.54 (expected) [M+], 474.00 (obs.), 474.54 (expected) [M+ + 1]. Anal. Calcd for C23H15N5O3S2: C, 58.34; H, 3.19; N, 14.79; S, 13.54%. Found C, 58.50; H, 3.29; N, 14.62; S, 13.70%.

Biological evaluation

Supporting information includes all experimental details for the MTT assay, and in vitro enzyme inhibition of VEGFR-2 and c-Met TKs.

In silico studies

Molecular modeling study

Molecular docking and visualizations of 2D and 3D styles were conducted by using MOE 2014.0901 and Discovery Studio visualizer softwares. The structures of co-crystallized TKs were retrieved from protein data bank (VEGFR-2: 4ASD & c-Met: 3LQ8). The structures of compound 3c and 3e were standardized and their energies were minimized. After that, the protein structures were prepared using MOE standard protocol27. Validation of docking approach was performed by redocking process of the original ligands (VEGFR-2: sorafenib & c-Met: foretinib) into the binding sites of both proteins9. Docking of compounds 3c and 3e were performed using default MOE docking setting. The 2D and 3D representations of the most stable poses were selected for further studies.

In silico prediction of physicochemical and pharmacokinetic properties

The freely available SwissADME website was utilized to predict the physicochemical and pharmacokinetics of the target compounds. Supporting information includes how different filters and parameters were calculated9.