Furanone-functionalized benzothiazole derivatives: synthesis, in vitro cytotoxicity, ADME, and molecular docking studies

2.1 Synthesis

2.1.1 Synthesis of 4-(benzo[d]thiazol-2-yl)-4-oxobutanoic acid (3)

To a mixture of o-aminothiophenol, (1) (6.5 mmol) and α-ketoglutaric acid (2) (7 mmol), polyphosphoric acid (10 g) was added and the resulting mixture was then heated along with vigorous stirring at 220 °C for 4 h (Scheme 1). The reaction mixture was then cooled and poured in 10% Na₂CO₃ solution to yield a suspension which was stirred until the evolution of the gas stopped and then filtered. The solid precipitate was collected and washed with 350 mL water. Recrystallization was done from MeOH–H₂O to get the pure product (R _f 0.71; toluene:ethyl acetate:formic acid = 5:4:1); nature: dark brown amorphous powder; Yield: 80%; m.p.: 270 °C. Elemental analysis [C₁₁H₉NO₃S] (%); Calc.: C, 56.16; H, 3.86; N, 5.95; found: C, 56.01; H, 3.37; N, 5.68. IR (cm⁻¹): 3021 (OH-str); 2962 (C═C Ar, str); 2881 (C–H str); 1721 (carboxylic acid C═O str); 1683 (C═O str); 1498 (C═N str); 1390, 1322 (C–O str). ¹H NMR (DMSO-d ₆ ): δ 2.28–2.51 (t, 2H, CH₂), 2.84–2.90 (t, 2H, CH₂), 6.38–7.51 (m, 3H, benzothiazole), 7.91 (s, 1H, benzothiazole), 12.47 (br s, 1H, –COOH). ¹³C NMR (DMSO-d ₆ ): δ 188.2 (COOH); 168.2 (C═O); 160.4 (C-thiazole); 155.6 (1C, arylidine C); 137.6 (1C, arylidine C); 121–125 (4C, arylidine C); 30.21 (2C, CH₂); ESI-MS: m/z: 235 (M⁺).

Scheme 1:

Synthetic route to title compounds 4a–j. (Ring systems-benzothiazole and furanone).

3.1 Synthesis

Ten new benzothiazoles coupled with furanones were synthesized via multi-step reactions from 4-(benzo[d]thiazol-2-yl)-4-oxobutanoic acid (3) as the starting material (Scheme 1). Compound 3 was obtained by condensing o-aminothiophenol (1) with α-ketoglutaric acid (2). (E)-5-(benzo[d]thiazol-2-yl)-3-arylidene furan-2(3H)-one) (4a–j) were synthesized by treating compound 3 with different substituted aryl aldehydes in Ac₂O along with the catalytic amount of triethylamine. Physical and analytical data of the synthesized compounds 4a–j are given in Table 1.

3.2 In vitro cytotoxicity

The IC₅₀ value of seven of the synthesized compounds (4a, 4c, 4e, 4f, 4h, 4i, and 4j) was estimated against the three cancer cell lines. The results of in vitro cytotoxicity against three cancer cell lines indicated that the di-substituted (3-methoxy, 4-hydroxy) compound (4i) demonstrated excellent activity against A549, MCF7, and DUI45 with IC₅₀ values of 7.2 ± 0.5 µM, 6.6 ± 1.4 µM, and 7.3 ± 0.1 µM, respectively. Also, the 3-nitro derivative (4c) exhibited promising activity against A549, MCF7, and DUI45 having IC₅₀ values of 9.6 ± 0.1 µM, 7.3 ± 0.2 µM, and 8.7 ± 0.2 µM, respectively. Furthermore, 3-chloro (4e) and unsubstituted compound (4f), also revealed good activity against all three cell lines. Doxorubicin was used as a standard drug, and exhibited IC₅₀ value 0.9 ± 0.2 µM, 0.9 ± 0.1 µM, and 0.8 ± 0.3 µM against A549, MCF7, and DUI45, respectively (Table 2). Molecular docking analyses of the synthesized compounds were in corroboration with the obtained results.

3.3 Structure-activity relationship

The following structure-activity relationship (SAR) could be drawn after analyzing the results of in vitro cytotoxicity of the synthesized compounds. All the compounds displayed substantial in vitro cytotoxicity (IC₅₀ of 7.2–19.3, 6.6–18.7, 7.3–19.2 µM) against three cancer cell lines (A549, MCF7, and DUI45). Compounds 4e and 4f, exhibited good activity against all three cancer cells. Substitution of the chloro (Cl⁻) group at position-3 makes the compound more effective against the human breast (MCF7) and prostate cancer (DUI45). The presence of halogen in the arylidine ring showed improvement in activity as compared to other substitutions in the arylidine ring. Compound 4c, substituted with m-nitro shows better activity against all three cell lines. Compound 4i, substituted with a methoxy (–OCH₃) group at the m-position and with a hydroxyl (–OH) group at the p-position shows excellent activity. From the IC₅₀ values, it may be inferred that substitution at the m-position in the benzaldehyde ring is necessary for better activity. The presence of an electron-donating methoxy (–OCH₃) group at the m-position gives a better result than an electron-withdrawing group. Further substitution at the p-position in the same ring with an -OH group (electron-donating group) improves the activity (Figure 2).

Figure 2:

Structure-activity relationship (SAR).

3.4 Acute toxicity

The results indicated no mortality in rats when the potent compounds 4c, 4e, 4f and 4i were administered in a dose of 500 mg/kg body weight. Consequently, the lethal dose (LD₅₀) of the tested compounds is >500 mg/kg body weight. Hence, the tested compounds are classified as non-toxic orally LD₅₀ (>500–<2000 mg/kg) based on the recommended guidelines by Organization for Economic Co-operation and Development (OECD).

3.5 Cardiomyopathy and hepatotoxicity

Cardiotoxicity in patients is a severe side effect of anticancer drugs [42]. Cardiotoxicity is believed to be instigated by the tyrosine kinase inhibitors viz., imatinib mesylate, dasatinib, nilotinib, sunitinib, sorafenib, and lapatinib (excluding gefitinib and erlotinib) [43]. While such drugs are linked with adverse cardiac events {viz., LV dysfunction (LVD), HF, cardiac ischemia, and myocardial infarction} are still approved owing to the availability of limited treatment opportunities [44, 45].

Hence, 4f and 4i being the most potent compounds were additionally tested for their toxicity in rat (heart and liver) in comparison to doxorubicin standard drug. Figure 3 shows photomicrographs of the normal architecture of cells of the liver (i) treated with control, (ii) effects of doxorubicin on liver cells, and (iii) effect of compound 4i on liver cells. Figure 4 illustrates the slides of the heart treated with compounds 4f and 4i revealed normal and well-maintained structures of myofibril in the subendothelial zone. No substantial difference in the structures of cardiac fiber was observed in comparison to the normal control group. However, the group treated with 4f and 4i substantially altered the structures of muscle fiber in comparison to the group treated with doxorubicin (Figure 4). Doxorubicin control group was found to induce cardiomyopathy, which was confirmed by a loss of cardiac fiber, vacuolation, and inflammation in left ventricle tissue.

Figure 3:

Photomicrographs of the normal architecture of cells of the liver (i) treated with control, (ii) effects of doxorubicin on liver cells, and (iii) effect of compound 4i on liver cells. Magnification 200×.

Figure 4:

A: Effects of compound 4i showing the normal architecture of cardiac fiber. B: Effects of compound 4f showing the normal architecture of cardiac fiber. C: Photomicrograph showing effects of normal control on cardiac fiber (well-maintained architecture of myofibrils). D: Effects of doxorubicin showing cardiomyopathy shown by loss of cardiac fiber, vacuolation, and inflammation in left ventricle tissue. Magnification 200×.

Tyrosine kinase inhibitors are generally associated with induction of liver injury and liver toxicity that varies in prevalence and severity by the use of different drugs. There is a noticeable increase in SGOT (serum glutamic oxaloacetic transaminase) and SGPT (serum glutamic pyruvic transaminase) levels caused by drug-induced liver toxicity [46]. The potent compounds 4f and 4i when studied for their liver toxicity exhibited approximately equal value (p < 0.01) of biochemical parameters viz., SGPT and SGOT in comparison to control while compound 4e revealed noticeable escalation in the serum hepatic enzyme levels (SGOT and SGPT) (Figure 5 and Table 3).

Figure 5:

Graphical representation of effect on (a) SGPT and (b) SGOT levels. (Dox, doxorubicin).

The most potent compound 4i studied for liver toxicity displayed analogous (p < 0.01) biochemical parameters; SGOT and SGPT as compared to control (Table 3). On the other hand, the animals treated with doxorubicin revealed substantial enhancement (p < 0.01) in SGOT and SGPT levels in comparison to control [47, 48].

3.6 Molecular docking

Molecular docking can help to predict the most energetically favorable binding pose of a ligand to its receptor in terms of the binding energy [41, 49], [50], [51], [52], [53]. Molecular docking studies were implemented to ascertain the molecular binding pattern of the synthesized benzothiazole clubbed furanone derivatives within the active pocket of the crystal structure of VEGFR (PDB ID: 2QU5) using Schrodinger program Maestro 10.5 [54]. Various interactions including hydrogen bonding, hydrophobic and π-interactions, etc., with the active pocket of the enzyme were taken into account while calculating the docking scores. Since, VEGFR-2 plays a significant role in angiogenesis, lymphangiogenesis, tumor growth, and metastasis, it is considered a promising target for anticancer therapy [55]. Table 4 displays the docking score and the binding energies of compounds 4a–j. Compounds 4a and 4i revealed the highest docking score −10.64 and −10.33, respectively in the active pocket of the VEGFR. Compound 4a forms five hydrogen bonds-two with amino acid LYS 920 (–H···O═C and O–H···O═C), two with CYS 919 (H···O═C and C═O–···H) and one with GLU 885 (H···O═C). While, compound 4i forms four hydrogen bonds, one with amino acid ASP 1046 (O···HN), two with CYS 919 (H···O═C and O–···HN) and one with GLU 885 (O–H···O═C). It was observed that the furanone ring of compounds 4a and 4i play an essential role in the hydrogen bonding with the –NH of the principal chain of CYS 919, a key amino acid residue of VEGFR found in its flexible hinge region. The aromatic benzene ring of both the compounds 4a and 4i interacts with the hydrophobic pocket along with hydrogen bond formation with GLU 885. The surface view of the compounds 4a and 4i revealed that both fit entirely in the receptor pocket of the VEGFR enzyme (Figure 6). The docking simulation methodology was validated by redocking of the co-crystal ligand of PDB ID: 2QU5 with compound 4i in the same binding domain of VEGFR and investigating the overlying pattern of 4i with doxorubicin as well as a co-crystal ligand as shown in Figure 7.

Figure 6:

Docked posed conformer of compound (a) 4a and (b) 4i.

Figure 7:

Superimpose of compound (a) 4i (reddish–brown) and doxorubicin (green) displayed one common H-bond with CYS 919 (yellow) (b) 4i (green) and co-crystal ligand (reddish–brown) displayed two common H-bonds with CYS 919 and ASP 1046 (yellow).

The surface view of the compounds 4a and 4i revealed that both 4a and 4i fit entirely in the receptor binding pocket of VEGFR and is shown in Figure 8. XP visualization of the hydrophobic enclosure pose of compounds 4a and 4i is displayed in Figure 9. The hydrophobic part of compounds 4a and 4i is enclosed by LEU 1035, LEU 840, VAL 848, PHE 918, LYS 868, ALA 866, VAL 914, LEU 882, and LEU 889. These hydrophobic interactions delivered optimum lipophilicity to compounds 4a and 4i enabling them to cross blood-brain barrier like membrane leading to improved interaction with the receptor pocket of VEGFR-2 kinase and therefore better VEGFR-2 kinase inhibition [56]. Thus, from these results, it is concluded that compounds 4a and 4i have the potential to act as VEGFR inhibitors.

Figure 8:

Molecular surface view of compound 4a (upper panel) and compound 4i (lower panel).

Figure 9:

XP visualization of hydrophobic enclosures of compounds (a) 4a and (b) 4i. A ball and stick model (green) was used to represent the hydrophobic atoms of compounds 4a and 4i and hydrophobic residues (amino acids) shown in the turquoise CPK presentation.

3.7 MM-GBSA binding free energy

The binding energy results from several drug interactions with the pockets of the receptor including polar, electrostatic, van der Waals and hydrophobic interaction, π–π stacking, etc. [57]. The compounds, 4a and 4i exhibited higher docking scores and also higher binding energies of −45.43 and −51.56 kcal/mol, respectively. Amongst all, compound 4i exhibited the highest DG binding (−51.56 kcal/mol), even higher than standard drug doxorubicin −36.75 kcal/mol. The binding energies of all other compounds fall in the range of −44.71 to −51.56 kcal/mol.

3.8 ADME (adsorption, distribution, metabolism, excretion)

ADME properties of potent best-docked compounds such as QPlogKP for the permeability of the skin, QPlogBB for overall CNS activity, QPlogKhsa represent the binding of human serum albumin (HSA), etc., are represented in Table 5. All the compounds of benzothiazole linked furanone are predicted to possess acceptable as well as excellent molecular descriptors that include molecular weight, hydrogen bond donor, hydrogen bond acceptor, QPlogKP, Percent human Oral Absorption, QPlogBB, QPlogKhsa, and QPlogPo/w values thereby satisfying the Lipinski’s Rule of Five (Table 5). However, both compounds 4a and 4i displayed the highest value of QPlogPw, H-bond acceptor, QPlogPoct, and % of human oral absorption, compound 4i possessing additional hydroxyl and methoxy group demonstrated the best ADME parameters. Therefore, along with the results obtained from the docking scores, MMGB assay, and ADME prediction, compound 4i was found to be the best compound displaying better pharmacokinetic and pharmacodynamic properties.