Introduction

Reactive oxygen species (ROS), known as mediators of intracellular signaling cascades, are chemically reactive molecules containing oxygen (Kotaiah et al., 2012). Most living organisms can produce ROS and metabolize excessive amounts through normal physiological processes. Accumulation of excessive ROS is generally responsible for damaging lipids, proteins, and the DNA of cells, leading to oxidative stress (Nordberg and Arnér, 2001). As scientific research unravels the root cause of cancer, evidence of oxidative stress leading to chronic inflammation of cells is believed to perpetuate this genetic instability (Kundu and Surh, 2012). In particular, chronic excessive ROS is one of the underlying reasons for the development of cancer (Wu et al., 2014). Cancer is still a global health concern, with 12.7 million new cases reported in 2008, and is projected to reach 22.2 million cases by 2030 (Vineis and Wild, 2014). Genetic instability is recognized as a hallmark of cancer development, resulting in abnormal cell growth. In cases where a cancerous tumor has metastasized, the survival rate is drastically reduced (Ferrari et al., 2010). In an effort to reduce the damaging effects of ROS, antioxidants capable of scavenging excess ROS have received much attention in cancer therapy (Fuchs-Tarlovsky, 2013). Phenolic compounds represent an important class of antioxidants (Tung et al., 2007). Some examples, such as curcumin, resveratrol, and epigallocatechin, possessing both antioxidant and anticancer properties, have received great scientific interest. Additionally, these compounds suppress the proliferation of cancer cells and induce cell apoptosis. Apoptosis is a mechanism of programmed cell death that occurs naturally when cells are beyond repair or has abnormalities (Prasad et al., 2014; Tang et al., 2011). Recent research indicates that the inhibitory effects of phenolic antioxidant compounds stem mainly from their free radical scavenging properties, and a large amount of evidence supports their chemoprotective effects against oxidative stress-mediated disorders (Soobrattee et al., 2005). The inhibitory effects of phenols are further demonstrated by the observation that the presence of sulphur compounds classified as secondary antioxidants, e.g., sulphides, can decompose peroxides that are intermediates in the oxidation reactions and regenerate the primary antioxidants (Pospíšil, 1993).

The 1,3,4-oxadiazoles are an important class of heterocyclic bioactive compounds (Omar et al., 1996). Their widespread use as scaffolds in medicinal chemistry establishes this moiety as a member of the class of privileged structures. Differently substituted 1,3,4-oxadiazoles have been found to exhibit anti-inflammatory,analgesic, anticancer, antioxidant, and antifungal activities (Rapolu et al., 2013; Husain et al., 2009; Desai et al., 2013; Revanasiddappa and Subrahmanyam, 2009; Zhang et al., 2013). Additionally, Schiff base-containing compounds have been widely investigated and have received significant attention in chemistry and biology due to their anti-inflammatory, antioxidant, and anticancer activities (Hegazy and Ali, 2012; Alam et al., 2012; Ren et al., 2002; T’ang et al., 1985).

Employing the early evaluation of drug-likeness in the discovery and design process has assisted medicinal chemists in selecting leads with satisfactory biopharmaceutical properties (Bergström et al., 2013). Lipinski’s Rule of Five (ROF) is a simple and widespread strategy employed in selecting leads. Developed by Chris Lipinski and his colleagues at Pfizer, the ROF is based on four influential properties identified from Pfizer’s library of compounds, namely, molecular weight (MW), clog P, number of hydrogen bond donors (HBD) and number of hydrogen bond acceptors (HBA). The authors noted that compounds with poor absorption or permeability violate the boundaries of more than one of the four properties, which are a MW >500, Clog P >5, number of HBD exceeding 5 and HBA exceeding 10 (Lipinski et al., 1997, 2001). However, oral bioavailability does not solely rely on passive permeability as observed in the ROF. It involves a complex interplay of factors, such as active transporters and first-pass metabolizing enzymes. This complexity explains the observation that not all FDA-approved drugs meet the ROF; in fact, approximately 30 % of the drugs violate the ROF (Zhang and Wilkinson, 2007). Veber and colleagues from GlaxoSmithKline improved upon the ROF with the addition of molecular rigidity (measured by the number of rotatable bonds (NROTB)) and molecule accessibility (measured by the polar surface area (PSA)) based on oral bioavailability in rats. Compounds with 10 or fewer NRTOB and PSAs equal to or less than 140 Å2 are assumed to have acceptable oral bioavailability (Veber et al., 2002). At the dawn of twenty-first century, the availability of human intestinal absorption data coupled with advancements in computational processing power brought about the construction of predictive computer models to assist medicinal chemists. One of the earliest was developed by Egan and colleagues in 2002 and has been a useful tool in predicting human intestinal absorption. The model predicts whether a compound lies within the regions of 95 % or 99 % confidence ellipses of well-absorbed compounds (Egan and Lauri, 2002).

With the previously mentioned hypothesis in mind, we aimed to combine the beneficial effects of phenols, sulphides, 1,3,4-oxadiazoles and Schiff bases in a compact structure with expected antioxidant and cytotoxic activities by synthesizing new 1,3,4-oxadiazole derivatives with the core structure shown in (Fig.1). Furthermore, a computational study predicting the absorption properties of the synthesized compounds was performed using the Lipinski rule, the Veber rule, and Egan’s model.

Fig. 1
figure 1

Phenols, sulphides, 1,3,4-oxadiazoles and Schiff bases in a compact structure to incorporate active pharmacophore features for antioxidant and cytotoxic activities

Full size image

Materials and methods

Chemicals

All chemicals and solvents were of analytical grade and purchased from Sigma-Aldrich (St. Louis, MO, USA) and Merck (Frankfurter StraBe, Darmstadt, Germany). Thin layer chromatography was performed on pre-coated silica gel plates (Si, 60, F254) for monitoring reactions and determining purity. Melting points were determined using a MEL-TEMP II apparatus and are uncorrected. IR spectra were recorded from 4000–400 cm−1 using a Perkin Elmer 400 Fourier transform infrared spectrometer (Waltham, MA, USA). 1H NMR and 13C NMR spectra were recorded on a BRUKER-AVN III 400 MHz (Fallanden, Switzerland) instrument using DMSO-d6 and CDCl3 as solvents and TMS as an internal standard. Mass spectra were recorded on an Agilent 5975 (Santa Clara, CA, USA) for EI/MS and Finnegan TSQ 7000 for HREI/MS.

Synthesis

5-[2-(ethylsulfanyl)phenyl]-2,3-dihydro-1,3,4-oxadiazole-2-thione (1)

White powder (EtOH), this compound was prepared as follows: potassium hydroxide (0.21 g, 3.75 mmol) was added to a solution of 2-(ethylsulphanyl) benzohydrazide (1 g, 3.75 mmol) and an excess of carbon disulphide (0.8 g, 0.6 mL) in absolute ethanol. The mixture was heated at reflux for 18 h, and the solvent was evaporated. Distilled water (25 mL) was added to the resulting residue, which was then filtered. Hydrochloric acid (5 %, 2 mL) was added and the filtered white precipitate was washed with water and recrystallized from ethanol to give the pure compound. It was obtained as a white solid powder (Yield: 84 %); m.p. 178–180 °C; IR. (KBr) v max 3314, 3219, 3059, 2959, 1602, 1558, 1337 cm−1;1H NMR (DMSO-d6, 400 MHz): δ = 1.28 (3H, t, J = 7.02 Hz, S–CH2CH3), 3.06 (2H, q, J = 7.32 Hz, S–CH2), 7.35 (1H, ddd, J = 7.9 Hz, H6), 7.57 (2H, m, J = 8.5 Hz, H4, H5), 7.81 (1H, dd, J = 8.4 Hz, H7), 14.83 (1H, br s, NH);13 CNMR (DMSO-d6, 100 MHz): δ = 13.27(CH3, S–CH2CH3), 25.67 (CH2, S–CH2), 119.93 (C, C6), 124.87 (C, C4),126.70 (C, C8), 129.51 (C, C7), 132.06 (C, C5), 138.18 (C, C3), 159.46 (C, C9, C=N), 176.99 (C, C10, C=S); HREIMS m/z (pos): 237.0155 C10H9N2OS2 (calcd. 237.0162).

Ethyl 2-({5-[2-(ethylsulfanyl)phenyl]-1,3,4-oxadiazol-2-yl}sulfanyl)-acetate (2)

Yellowish solid (ethanol-ethyl acetate), this compound was synthesized when ethyl bromoacetate (1 g, 10 mmol) was added to a solution of compound 1 (2 g, 10 mmol) and anhydrous K2CO3 (1.3 g) in dry acetone. The mixture was stirred for 18 h at room temperature. The solvent was removed by evaporation and ethyl acetate (15 mL) was added. The mixture was left at room temperature to allow the product to precipitate. The solid product was filtered and recrystallized from ethanol-ethyl acetate (1:2). It was obtained as a pale yellow solid, (Yield: 78 %); m.p. 46 °C; IR. (KBr) v max2931, 2871, 1729, 1591 cm−1.1H NMR (CDCl3, 400 MHz): δ = 1.31 (3H, t, J = 7.2 Hz, S–CH2 CH 3 ), 1.39 (3H, t, J = 6.9 Hz, OCH2 CH 3 ), 3.02 (2H, q, J = 6.8 Hz, S–CH 2 ), 4.13 (2H, q, J = 7.0 Hz, OCH 2 ), 4.16 (2H, s, J = 6.9 Hz, SCH2), 7.26 (1H, m, J = 7.4 Hz, H6), 7.44 (2H, m, J = 7.6 Hz, H4, H5), 7.85 (1H, dd, J = 8.4 Hz, H7); 13C NMR (CDCl3, 100 MHz): δ = 13.45 (CH3, S–CH2CH3), 14.08 (CH3, C14, OCH2CH3), 26.91 (CH2, S–CH2CH3), 34.42 (CH2, C11, S–CH2), 62.39 (CH2, C13, OCH2), 121.64 (C, C6), 124.69 (C, C7), 126.93 (C, C4), 129.81 (C, C5), 131.38 (C, C8), 139.07 (C, C3), 162 (C, C9), 165.22 (C, C10), 167.45 (1C, C12); HREIMS m/z (pos): 325.0687 C14H17N2O3S2 (calcd. 325.0675).

2-[2-(Ethylsulfanyl)phenyl]-5-hydrazinyl-1,3,4-oxadiazole (3)

White powder: this compound was synthesized from the addition of a solution of hydrazine hydrate (80 %, 2 mL) to compound 2 (2 g, 5 mmol) in dioxane (25 mL) and the mixture was allowed to stir for 48 h at room temperature). The solvent was evaporated under reduced pressure resulting in the formation of a pure white solid (Yield: 69 %); m.p. 86 °C; IR. (KBr) v max 3490, 3328, 3020, 3081, 2975, 1662, 1642 cm−1.1H NMR (DMSO-d6, 400 MHz): δ = 1.26 (3H, t, J = 7.3 Hz, S–CH2CH3), 3.01 (2H, q, J = 7.4 Hz, S–CH2), 4.51 (2H, br s, J = 5.0 Hz, NH2), 7.3 (1H, m, J = 7.9 Hz, H6), 7.49 (2H, m, J = 8.0 Hz, H4, H5), 7.68 (1H, d, J = 8.2 Hz, H7), 8.62 (1H, s, NH); 13C NMR (DMSO-d6, 100 MHz): δ = 13.21 (CH3, S–CH2CH3), 25.59 (CH2, S–CH2), 121.53 (C, C6), 124.78 (C, C7), 126.41 (C, C4), 128.93 (C, C5), 130.99 (C, C8), 137.25 (C, C3), 157.17 (C, C9) 165.72 (C, C10); HREIMS: m/z (pos): 236.0733 C10H12N4O1S1 (calcd. 236.0732).

General procedure for the synthesis of compounds (4a–e)

The appropriately substituted hydroxybenzaldehyde (1 mmol) was added in small portions with stirring to a solution of Compound 3 (0.25 g, 1 mmol) in absolute ethanol (15 mL), and the mixture was heated at reflux for 6 h. The mixture was cooled and the precipitate that formed was washed and recrystallized from ethanol.

4-[(1E)-(2-{5-[2-(ethylsulfanyl)phenyl]-1,3,4-oxadiazol-2-yl}hydrazin-1-ylidene)methyl]phenol (4a)

Using 4-hydroxybenzaldehyde, colorless solid (Yield: 90 %); m.p. 198–200 °C; IR. (KBr) v max 3676 (OH phenol), 3322, 2981, 1633 cm−1;1H NMR (DMSO-d6, 400 MHz): δ = 1.29 (3H, t, J = 7.3 Hz, S–CH2CH3), 3.04 (2H, q, J = 7.4 Hz, SCH2), 6.84 (2H, m, J = 7.5 Hz, H14), 7.34 (1H, m, J = 7.4 Hz, H6), 7.53 (4H, m, J = 7.7 Hz, H4, H5, H13), 7.79 (1H, d, J = 8.5 Hz, H7), 8.06 (1H, s, CH=N), 9.9 (1H, s, OH), 11.88 (1H, s, NH); 13C NMR (DMSO-d6, 100 MHz): δ = 13.84 (CH3, S–CH2CH3), 26.19 (CH2, S–CH2), 116.15 (2C, C14), 122.37 (C, C6), 125.30 (C, C12), 125.62 (C, C7), 127.19 (C, C4), 128.82 (2C, C13) 129.50 (C, C5), 131.36 (C, C8), 138.04 (C, C3), 145.55 (C, C11), 157.89 (C, C15), 159.52 (C, C9), 161.94 (C, C10); HREIMS: m/z (pos): 340.0992 C17H16N4O2S1 (calcd. 340.0994).

4-[(1E)-(2-{5-[2-(ethylsulfanyl)phenyl]-1,3,4-oxadiazol-2-yl}hydrazine-1-ylidene)methyl]-2-methoxyphenol (4b)

Using 4-hydroxy-3-methoxy benzaldehyde, colorless solid (Yield: 85 %); m.p. 100–102 °C; IR. (KBr) v max 2951, 1631 cm−1.1H NMR (DMSO-d6, 400 MHz): δ = 1.28 (3H, t, J = 7.2 Hz, S–CH2CH3), 3.04 (2H, q, J = 7.3 Hz, S–CH2), 3.82 (3H, s, OCH2CH3), 6.83 (1H, d, J = 7.2 Hz, H14), 7.1 (1H, dd, J = 7.1 Hz, H6), 7.32 (2H, m, J = 7.4 Hz, H5, H13), 7.52 (2H, m, J = 7.3 Hz, H4, H17), 7.79 (1H, d, J = 7.7 Hz, H7) 8.05 (1H, s, CH=N), 9.51 (1H, s, OH), 11.89 (1H, s, NH); 13C NMR (DMSO-d6, 100 MHz): δ = 18.58 (CH3, S–CH2CH3), 30.98 (CH2, S–CH2), 60.76 (CH3, OCH3), 114.36 (C, C17), 120.68 (C, C14), 126.44 (C, C13), 127.07 (C, C6), 130.14 (C, C7), 130.76 (C, C4) 132.01 (C, C12), 134.35 (C, C5), 136.27 (C, C8), 142.70 (C, C3), 150.79 (C, C11), 153.17 (C, C16), 153.72 (C, C15), 162.73 (C, C9), 166.65 (C, C10); HREIMS m/z (pos): 370.1086 C18H18N4O3S1 (calcd. 370.1100).

2-Ethoxy-4-[(1E)-(2-{5-[2-(ethylsulfanyl)phenyl]-1,3,4-oxadiazol-2-yl}hydrazine-1-ylidene)methyl]phenol (4c)

Using 4-hydroxy-3-ethoxybenzaldehyde, colorless solid (Yield: 89 %); m.p. 204 °C; IR. (KBr) v max 3545, 2931, 1634 cm−1. 1H NMR (DMSO-d6, 400 MHz): δ = 1.28 (3H, t, J = 7.3 Hz, S–CH2CH3), 1.36 (3H, t, J = 7.0 Hz, OCH2CH3), 3.04 (2H, q, J = 7.3 Hz, S–CH2), 4.07 (2H, q, J = 6.9 Hz, OCH2), 6.84 (1H, d, J = 7.3 Hz, H14), 7.09 (1H, dd, J = 8.2 Hz, H6), 7.25 (1H, d, J = 1.8 Hz, H5), 7.34 (1H, m, H13) 7.54 (2H, m, J = 8.2 Hz, H4, H17), 7.78 (1H, d, J = 8.0 Hz, H7) 8.04 (1H, s, CH=N), 9.42 (1H, s, OH), 11.88 (1H, s, NH). 13C NMR (DMSO-d6, 100 MHz): δ = 13.29 (CH3, S–CH2CH3), 14.47 (CH3, OCH2CH3) 25.77 (CH2, S–CH2), 63.92 (CH2, OCH2), 110.27 (C, C17), 115.45 (C, C14), 121.33 (C, C13), 121.65 (C, C6), 124.97 (C, C7), 125.49 (C, C4) 126.80 (C, C12), 129.18 (C, C5), 131.21 (C, C8), 137.37 (C, C3), 145.93 (C, C11), 147.01 (C, C16), 148.57 (C, C15), 157.57 (C, C9), 161.40 (C, C10); HREIMS m/z (pos): 384.1257 [M+] (calc. for C19H20N4O3S1 384.1256).

2,6-di-tert-butyl-4-[(1E)-(2-{5-[2-(ethylsulfanyl)phenyl)-1,3,4-oxadiazol-2-yl)-hydrazin-1-ylidene)methyl]phenol (4d)

Using 3,5-di-tert-butyl-4-hydroxybenzaldehyde, colorless solid, (Yield: 90 %); m.p 216 °C; IR. (KBr) v max 3630, 3313, 3214, 2956, 1643 cm−1.1H NMR (DMSO-d6, 400 MHz): δ = 1.28 (3H, t, J = 7.3 Hz, S–CH2CH3), 1.41 (18H, s, 2× t-Bu), 3.04 (2H, q, J = 7.4 Hz, CH2), 7.35 (2H, m, phenolic OH and H4), 7.46 (2H, s, H13), 7.53 (2H, d, J = 3.8 Hz, H5, H6), 7.77 (1H, d, J = 7.7 Hz, H7) 8.08 (1H, s, CH=N), 11.86 (1H, s, NH).13C NMR (DMSO-d6, 100 MHz): δ = 13.42 (CH3, S–CH2CH3), 25.77 (CH2, S–CH2), 30.11 (6C, 2× –C(CH3)3), 34.4 (2C, 2× –C(CH3)3, 122.04 (2C, C13), 123.38 (C, C12), 124.92 (C, C6), 125.47 (C, C7), 126.91 (C, C4), 129.05 (C, C5), 130.97 (C, C8), 137.34 (C, C14), 139.24 (C, C3), 146.04 (C, C=N, C11), 155.37 (C, C15), 157.53 (C, C=O, C9), 161.49 (C, C=N, C10); HREIMS m/z (pos): 452.2239 C25H32N4O2S1 (calcd. 452.2246).

2-Bromo-4-chloro-6-[(1E)-(2-{5-[2-(ethylsulfanyl)phenyl]-1,3,4-oxadiazol-2-yl}-hydrazin-1-ylidene)methyl]phenol (4e)

Using 3-bromo-5-chloro-2-hydroxybenzaldehyde, colorless solid (Yield: 88 %); m.p. 202–204 °C; IR. (KBr) v max 3545, 2949, 1632 cm−1.1H NMR (DMSO-d6, 400 MHz): δ = 1.3 (3H, t, J = 7.3 Hz, CH3), 3.06 (2H, q, J = 7.4 Hz, CH2), 7.36 (1H, ddd, J = 7.9, H15), 7.55 (2H, m, H5, H6), 7.69 (2H, dd, J = 9.2 Hz, H4, H13), 7.8 (1H, m, H7), 8.34 (1H, br.s, CH=N), 11.46 (1H, br.s, OH), 12.69 (1H, s, NH); 13C NMR (DMSO-d6, 100 MHz): δ = 13.32 (CH3, S–CH2CH3), 25.75 (CH2, SCH2), 111.02 (C, C16), 121.19 (C, C12), 121.52 (C, C6), 123.62 (C, C7), 124.87 (C, C4), 126.80 (C, C14) 128.22 (C, C13), 129.10 (C, C5), 131.22 (C, C8), 132.52 (C, C15), 137.71 (C, C3), 144.09 (C, C11), 152.36 (C, C17), 158.01 (C, C9), 160.56 (C, C10); HREIMS m/z (pos): 451.9707 C17H14N4BrClO2S1 (calcd. 451.9709).

In silico drug-likeness evaluation

Discovery Studio 4.0 software (Accelrys Incorporation, San Diego, CA, USA) was used to calculate the MW, calculated log P (clog P), number of HBA and HBD for the Lipinski rule; NROTB and PSA for the Veber rule; and, predicting the human intestinal absorption based on Egan’s model, which were available in the ADMET module of the software.

Antioxidant activities

DPPH radical scavenging assay

DPPH assay is one of the tests used to prove the ability of the compounds to act as donors of hydrogen atoms. As the DPPH free radical molecule has a single electron, it gives rise to a deep violet color, characterized by an absorption band in methanol solution at about 515 nm. When a solution of DPPH is mixed with an antioxidant compound that can donate a hydrogen atom, this gives rise to the reduced form of DPPH (non radical DPPH) with the loss of its deep violet color (Prior et al., 2005). The scavenging activity of stable 2,2 Diphenyl-1-picryl dyhydrazyl free radical was determined according to the method of (Gerhäuser et al., 2003), with slight modifications. (1 mg/1 mL) of the synthesized compound and the reference standards were prepared as stock solutions then a series of dilution solutions with five varying concentrations were tested. 5 μL of samples/standards were loaded, and followed by 195 μL of DPPH reagent, the mixtures were then mixed vigorously and incubated in the dark at room temperature for 2 h, and the absorbance was measured spectrophotometrically at 515 nm at intervals of 20 min for 3 h. The percentage of DPPH free radical scavenging activity was calculated as: DPPH (%) = [(Abs of blank−Abs of sample)/absorbance of sample] × 100. The results were expressed as (half-maximum inhibitory concentration (IC50) value) the compounds concentration that required to reduce 50 % of the hydroxyl radical produced.

Ferric ion reducing antioxidant power (FRAP) assay

The reducing capacities of the synthesized compounds were measured by the method of Benzie and Strain (Benzie and Strain, 1996), with modifications. A total of 10 mL of 300 mM acetate buffer was adjusted to pH 3.6 by mixing with 3.1 g CH3COONa.3H2O and 16 mL of glacial acetic acid. The TPTZ (2,4,6-tripyridyl-s-triazine) solution was prepared by dissolving 10 mM TPTZ in 40 mM HCl. Then, 1 mL of the TPTZ solution was mixed with the ferric ion reducing antioxidant power (FRAP) solution and 1 mL of 20 mM ferric chloride hexahydrate in distilled water. The FRAP solution was warmed to 37 °C, and the synthesized compounds were added and allowed to react in the dark. Absorbance was monitored spectrophotometrically at 593 nm. After that 10 μL of compound or standard and 90 μL of distilled water were added to 300 μL of the working FRAP reagent. Absorbance was measured (593 nm) at 0 min immediately upon addition of the working FRAP reagent after vortexing. Thereafter, absorbance reading was taken after 4 min. All results were expressed as µM ferrous per g dry mass and were compared with that of reference compounds. The results are expressed as mean ± standard deviation of triplicates.

Cytotoxic screening

Cell culture

Human cancer and normal cell lines were obtained from American Type Culture Collection (ATCC) (Manassas, VA, USA). Dulbecco’s modified Eagle’s medium (DMEM), Hanks’ Balanced Salt Solution (HBSS), 100 mM non-essential amino acids, phosphate buffer solution (pH 7.2), gentamycin and amphotericin B were purchased from Invitrogen Corporation (Carlsbad, CA, USA). L-glutamine, fetal bovine serum (FBS), 0.25 % trypsin-EDTA, dimethyl sulphoxide (DMSO), curcumin, vinblastinesulfate and 4,5-dimethylthiazol-2-yl-2,5-diphenyltetrazoliumbromide (MTT) were purchased from Sigma-Aldrich (St. Louis, MO, USA).

Cytotoxicity assay

The cytotoxicity of the compounds was evaluated against three types of cancer cell lines, namely, pancreatic (BxPC-3), non-metastatic (MCF-7) and metastatic breast (MDA-MB-231), as well as a normal pancreatic cell line (hTERT-HPNE). Cell lines were cultured in DMEM media supplemented with 2 mM L-glutamine, 10 % FBS, 50 µg/mL gentamycin and 2.5 µg/mL amphotericin B, and maintained in a 37 °C humidified atmosphere of 5 % CO2 in a cell incubator. Compounds 1–4e and the positive control (curcumin and vinblastine sulfate) were dissolved in DMSOand further diluted with DMEM media to yield a final DMSO concentration of <0.5 % (v/v). Cells were plated into 96-well microplates at a density of 104 cells per well and maintained in a cell incubator for 24 h. Then 100 µL of samples were introduced in triplicate to give a final concentration range of 100–1.56 µM. For the negative control,cells were treated with vehicle (0.5 % v/v DMSO) only. Cells were further incubated for 72 h, and then, cell viability was determined. Culture media were carefully refreshed with 100 µL of HBSS, followed by 50 µL per well of MTT reagent (2 mg/mL). Microplates were returned to the incubator for 2 h, after which the supernatant was removed and the formazan crystals were dissolved in 100 µL of DMSO. The absorbance of the formazan product was read on a microplate reader at 570 nm, with 620 nm as the background wavelength (Infinite 200, Tecan, Männedorf, Switzerland). The IC50’s of the samples and the positive control were determined from dose-response curves using Prism 5.02 software (GraphPad Software Inc., La Jolla, CA, USA) (Looi et al., 2013). The results are expressed as mean ± standard deviation of triplicates.

Apoptosis and cell cycle analysis

Apoptosis and cell cycle analysis were performed separately using commercial Annexin V: FITC Apoptosis Detection Kit I and CycleTest™ Plus DNA Reagent kit, respectively (BD Bioscience, NJ, USA). Cells were seeded at a density of 5 × 105 cells per well in six-well plates. After overnight attachment, cells were treated with compound at 13, 26 and 52 µM for 72 h in apoptosis analysis and 13 µM for 24, 48 and 72 h in cell cycle analysis. After treatment, cells were harvested, centrifuged at 1500 × g for 5 min and cell staining was performed according to the manufacturer’s instructions. DNA QC Particles kit was used to calibrate the FACSCanto II flowcytometer (BD Biosciences, NJ, USA). Cell populations were subjected to cytometry analysis and the quadrants were set according to the population of viable cells in untreated samples. Three independent experiments were performed and in each experiment, a total of 10,000 events were collected for both apoptosis and cell cycle analysis. FacsDiva 5.0.3 software was used to calculate the percent of cells in the respective quadrants for apoptosis analysis and Mod Fit LT software (Verity Software House Inc., Topsham, ME, USA) for cell cycle analysis (Looi et al., 2013). The results are expressed as mean ± standard deviation and statistical significant were calculated using Prism 5.02 software’s paired Student’s t-test against controls (GraphPad Software Inc., La Jolla, CA, USA).

Results and discussion

Chemistry

The 2-[2-substituted-hydrazinyl]-5-[2-(ethylsulphanyl)phenyl]-1,3,4-oxadiazole derivatives 4a–e were synthesized according to the reaction sequence outlined in Scheme 1. 5-[2-(Ethylsulphanyl)phenyl]-1,3,4-oxadiazole-2(3H)-thione (1) was prepared by reaction of 2-(ethylsulphanyl)benzohydrazide with CS2 in KOH. Treatment of (1) with ethyl bromoacetate gave the ethyl acetate analog (2) in good yield. Reaction of the ethyl acetate (2) with 80 % N2H4 .H2O yielded the proposed acid hydrazide (3), which was further reacted with differently substituted hydroxybenzaldehydes in the presence of ethanol to afford compounds 4a–e. The structures of the synthesized compounds were confirmed by IR, 1H NMR, and 13C NMR spectroscopy and mass spectrometry.

Scheme 1
scheme 1

Reaction scheme for the formation of compounds 4a–e. Reagents and conditions: a (i) CS2/KOH, EtOH, reflux 18 h; (ii) HCl; b ethyl bromoacetate, K2CO3, acetone, r.t., 18 h; c NH2NH2.H2O, dioxane, r.t., 18 h; d ArCHO, EtOH, reflux 6 h

Full size image

In silico drug-likeness evaluation

Performing early evaluations of drug-likeness would be advantageous in filtering compounds with predicted poor biopharmaceutical properties. Drug-likeness encompasses the evaluation of structural or physical properties of a compound that is consistent with established drugs (Yehye et al., 2012). The drug-likeness of parent 1 and intermediates 2 and 3, and derivatives 4a–4e were evaluated based on the Lipinski rule. None of the compounds violated any of the rules, with the exception of compounds 4d and 4e, which violated only the clog P rule, where their values were >5 (Table 1). According to the Lipinski rule, violation of one parameter is still acceptable (Lipinski et al., 2001). However, it should be mentioned that approximately 30 % of FDA-approved drugs violate the Lipinski rule, and this is attributed to the inherent rigidity of the strict boundaries set by the rule (Zhang and Wilkinson, 2007; Petit et al., 2012).

Table 1 Calculated molecular properties based on the Lipinski and Veber rules
Full size table

On the other hand, based on the Veber rule, all of the compounds are within the rule’s boundaries (Table 1). The differences in the Veber rule are the addition of molecular rigidity (NROTB) and molecule accessibility (PSA), developed after data mining the results of oral bio availabilities, in rats, of 1100 drug candidates (Veber et al., 2002). However, both the Lipinski and Veber rules share a common weakness in considering drug absorption as a univariate relationship. Therefore, it does not take into account interactions between the molecular properties or the multivariate relationships for the prediction of drug absorption (Egan and Lauri, 2002). Egan and colleagues built a computational model looking at multivariate factors that influence drug absorption. When they plotted the descriptors atom-based log P (AlogP) vs. PSA, they were able to cluster compounds within 95 % or 99 % confidence ellipses. They further explained that both lipophilicity (measured by Alog P) and hydrogen bonding ability (measured by PSA) are nonlinearly related to permeability. This enables the construction of a clustering model (ellipses) more useful than a quantitative model (Egan et al., 2000). When we assessed our compounds using Egan’s model, compounds 1–4a–c were within the 95 % confidence ellipse and compound 4e was within the 99 % confidence ellipse for good absorption, whereas compound 4d was predicted to have poor absorption (Fig. 2). It was noted that compound 4d has a high log P value (7.38) (Table 1). A published report explained that a highly lipophilic compound will be highly partitioned in the lipid portion of a membrane and fail to pass through the aqueous portion of the membrane, which leads to its poor permeation across the intestinal barrier (Wils et al., 1994).

Fig. 2
figure 2

Plot of AlogP versus PSA for compounds 1–4e and curcumin based on Egan’s model. AlogP atom-based log P,PSA polar surface area, CE confidence ellipse

Full size image

Antioxidant activities

All of the newly synthesized compounds were evaluated for their free radical scavenging ability and compared with five positive controls—quercetin, BHT, trolox, rutin, and ascorbic acid, and the results are shown in Table 2. Compounds 1 and 3 showed the highest radical scavenging activities. Compounds 4a, 4b, 4d, and 4e demonstrated greater radical scavenging activities than BHT, trolox, rutin, and ascorbic acid. Compound 4c exhibited lower IC50 than that of BHT but higher than those of the other tested compounds. Compound 2 was the only compound that did not show any radical scavenging activity in this assay.

Table 2 DPPH scavenging activity (EC50 µg/mL), and FRAP values (μM/g) for the synthesized compounds (1, 2, 3, 4a, 4b, 4c, 4d, and 4e)
Full size table

The assay for reduction of iron (III) to iron (II) at low pH was performed to determine the FRAP values (μM/100 g) of the synthesized compounds in comparison with the positive controls. The results are shown in (Table 2). Compounds 3, 4b, 4c, and 4d displayed excellent FRAP values compared to the reference compounds. Compounds 2 and 4a possessed greater FRAP reducing power than BHT, while compounds 1 and 4e showed very poor FRAP values when compared to the reference compounds.

In the search for more potent leads through the incorporation of different pharmacophores into one structure, it is important to determine the structure–activity relationship of the resulting compounds. In this study, two different in vitro antioxidant assays were carried out to evaluate the biological activities of the newly synthesized compounds – the DPPH radical scavenging assay and the FRAP assay. The total antioxidant capacity value should include methods applicable to both lipophilic and hydrophilic antioxidants with regards to the similarities and differences of both hydrogen atom transfer (HAT) and electron transfer (ET) mechanisms (Karadag et al., 2009). The DPPH assay involves both HAT and ET mechanisms, whereas the FRAP assay involves only the ET mechanism (Prior et al., 2005). HAT-based methods measure the classical ability of an antioxidant to scavenge free radicals by donating hydrogen to form stable compounds. ET-based methods detect the ability of a potential antioxidant to transfer one electron and reduce any compound, including metals, carbonyls, and radicals (Karadag et al., 2009).

Based on the results observed, we believe that the HAT mechanism is involved in the antioxidant activity. Compounds 1, 3, and 4a–e contain an NH group that could easily transfer a H atom onto the DPPH radical thus demonstrated good radical scavenging activity. Compound 2 does not possess a NH or OH group and therefore, it could not transfer a H atom onto the DPPH radical, leading to its poor activity as an antioxidant.

In the case of compound 1, the mechanisms of antioxidants are shown in Scheme 2. Assuming that the bond-dissociation energy of N–H is low, the scavenging of DPPH• by 1 is proposed in Scheme 2, in which hydrogen radicals are extracted from the N–H of the oxadiazole ring. This mechanism showed that compound 1 is able to suppress two DPPH• radicals.

Scheme 2
scheme 2

Proposed HAT mechanism between compound 1 and the DPPH radical

Full size image

The result shown in Table 2 indicates that the substituents on the benzene ring of the Schiff bases have little influence on the radical scavenging ability (except for compound 4c). The IC50 value of compounds 4a, 4b, 4d and 4e does not differ very much with each other. Compound 4c demonstrated the lowest radical scavenging activity. We could not offer any explanation at the moment as to why 4c, which is very similar in structure with 4b, but show poor radical scavenging activity.

For the monophenolic compounds 4a–e, we proposed a different mechanism. Compound 4b demonstrated the highest radical scavenging activity. The proposed mechanism involves the dimerization between the resulting radicals as shown in Scheme 3. Brand-Williams has proposed a similar mechanism for phenoxy compounds with free ortho- and para- positions (Brand-Williams et al., 1995). After dimerization the two hydroxyl groups would be regenerated intramolecularly and could again interact with the DPPH radical.

Scheme 3
scheme 3

Proposed HAT mechanism between compound 4b and the DPPH radical

Full size image

From the FRAP assay data, comparison between compounds 4a–e show that the presence of an electron donating group on the phenolic ring increases the reducing power of the compounds. On the other hand, the presence of electron withdrawing groups reduces the reducing power of the compounds. Compounds 4b and 4c contain strong electron donating ability through resonance by OMe and OEt group, respectively and therefore have strong reducing power compared to 4a, which does not possess a substituent, as well as 4d, which possesses a weak electron donating group. In contrast, compound 4e, containing two strong electron withdrawing groups (Br and Cl), shows very poor reducing power.

Cytotoxic activity

The cytotoxic effects of the derivatives were investigated in vitro against the BxPC-3, MCF-7, and MDA-MB-231 cancer cell lines and the hTERT-HPNE normal cell line using the MTT assay. This colorimetric assay measures the percent of viable cells after exposure to a test compound (Malich et al., 1997). A dose-response curve is observed from cytotoxic compounds where the percentage of viable cells decreases with increasing concentrations of the compounds. The 50 % inhibitory concentration (IC50) is obtained from the dose-response curve at the concentration where 50 % of the cells remain viable (Table 3). Positive controls, curcumin and vinblastine sulfate were used for comparison. Curcumin was selected because it contains two benzene rings that are substituted with –OH and –OMe groups, which has structural resemblance to our compounds (4a–e). Furthermore, curcumin has been reported to show anticancer and antioxidant potencies, which are related to the activities we are investigating in this study (Prasad et al., 2014). On the other hand, vinblastine sulfate is an anticancer drug clinically being used for the treatment of cancers. The derivatives revealed a moderate inhibitory effect against the cancer cell lines, comparable to curcumin and vinblastine sulfate. The monophenolic derivatives (4a, 4b, 4c, and 4e) gave an inhibitory effect against the cancer cell lines, with the most potent being compound 4e against MDA-MB-231 (IC50 = 21.40 ± 1.22 µM) and 4c against BxPC-3 (IC50 = 26.17 ± 1.10 µM). However, the halogenated compound 4e showed higher cytotoxicity against normal pancreatic cells (hTERT-HPNE IC50 = 1.88 ± 0.29 µM) than pancreatic cancer cells (BxPC-3 IC50 = 31.07 ± 2.76 µM). Compounds 1, 2, and 4d exhibit low cytotoxic activity against the cell lines tested, as their IC50 values were greater than 100 µM.

Table 3 IC50 values of compounds 1–4a–e and curcumin obtained from their respective dose-response curves after 72 h treatment
Full size table

Apoptosis and cell cycle analysis

The poor survival rate of cancer patients in the early stages of chemotherapy could be attributed to the loss of potency when the cancer progresses into the metastatic stage, as well as non-specificity, which can cause serious side effects on normal healthy cells (Hoelder et al., 2012). Bearing this in mind, in addition to non-metastatic cancer cell lines (BxPC-3 and MCF-7), we employed normal cell lines (hTERT-HPNE) and metastatic cancer cell lines (MDA-MB-231) in our approach to evaluate the specificity and potency of the compounds. Although compound 4e is highly cytotoxic towards metastatic cancer cells, non-selectivity deter us from proceeding with this compound for further investigation of anticancer mechanism. Instead, we have selected the compound 4c to demonstrate that it induces apoptosis through flow cytometry. This was accomplished by Annexin V-FITC/PI dual-staining assay of BxPC-3 cells after treatment with 4c (13, 26, and 52 µM) for 72 h. Compound concentrations were selected based on the IC50 obtained previously. As illustrated in Fig. 3, a significant (p < 0.01) dose-independent increase from 15.00 to 47.20 % and finally to 46.33 % of apoptotic cells is observed (right lower and upper sections of the fluorocytogram), in contrast to the control which had only 1.7 %. This result is in a decreasing percent of viable cells from 96.97 % in control to 45.28 % in treated cells (p < 0.01) at the highest concentration tested. In a separate experiment, cell cycle arrest was demonstrated through flow cytometry by single-staining (PI) assay of BxPC-3 treated cells at time intervals of 24, 48, and 72 h. The cell cycle profiles and percent of each phase are shown in Fig. 4. Analysis of the DNA profiles revealed a significant (p < 0.05) time-dependent increase in cell populations at G0/G1 phase in comparison with control, displaying cell cycle arrest and suppression of cell proliferation. Taken together, both flow cytometry experiments show 4c induce apoptosis with cell cycle arrest at G0/G1-phase; which is in agreement with the anticancer mechanisms reported for other derivatives of 1,3,4-oxadiazoles (Suan et al., 2013).

Fig. 3
figure 3

Effects of compound 4c on cell apoptosis of BxPC-3cells. Apoptosis was detected by AnnexinV-FITC/PI dual-staining after treatment with compound 4c (13, 26, and 52 µM) for 72 h and 10,000 events were analyzed by flow cytometer. Results are expressed as mean ± standard deviation of three independent experiments, significant difference in comparison to controls is indicated as **p < 0.01

Full size image
Fig. 4
figure 4

Effects of compound 4c on cell cycle of BxPC-3cells. Cells were with treated with compound 4c (13 µM) for 24, 48, and 72 h, and the DNA content of 10,000 events was analyzed by flow cytometer. Results are expressed as mean ± standard deviation of three independent experiments, significant difference in comparison to controls are indicated as *p < 0.05 and **p < 0.01

Full size image

In relating the antioxidant and cytotoxic activities, parent compound 1, containing only a 1,3,4-oxadiazole moiety, exhibited high DPPH radical scavenging activity, with an IC50 = 2.22 ± 0.01 µg/mL, and did not show any cytotoxic potency in the tested cell lines. The same was observed regarding the cytotoxic potency of the intermediate compound 2, which contains an ethyl ester moiety, and showed a FRAP value higher than the BHT positive control. Replacing the ester moiety in 2 with a hydrazide group in compound 3 enhanced the antioxidant activity, but with no notable change in the cytotoxic potency. This observation may be due to the presence of the free –NHNH2 group, the free electron pair of the NH2 does not participate in the resonance with the aromatic ring, thus it could act as an electron donor (Gozzo et al., 1999). Compounds 4a–e are derivatives of compound 3, formed after the introduction of the phenolic OH moiety with varying substitutions on the aromatic ring. Interestingly, the introduction of the phenolic OH confers cytotoxic potency against the tested cell lines, with the exception of compound 4d (Table 3). It was reported that the phenolic OH moiety found in natural polyphenolic compounds, such as trolox and quercetin, is important towards suppressing ROS (Derochette et al., 2013; Priyadarsini et al., 2003). Because chronic inflammation from excessive ROS is one of the reasons for cancer development, suppressing it is a potential strategy for cancer therapy. Based on our antioxidant results, higher activity is observed for the phenolic compounds in DPPH than in the FRAP assay. Therefore, we infer that the scavenging activity of the reactive species by the compounds 4a–e is caused by the phenolic OH moiety, which has hydrogen that can be donated. With regards to compound 4d, the di-tert-butyl moiety flanking the phenolic OH may have caused steric hindrance to the hydrogen donating ability of this compound. Our structure–activity relationship data are summarized in Fig 5. These results indicate that 1,3,4-oxadiazole derivatives with phenolic OH and NH group have better scavenging activity, while in order to have cytotoxic activity, the 1,3,4-oxadiazole derivatives should contain a Schiff base group connected to a substituted benzene ring.

Fig. 5
figure 5

The effects of various substituents toward antioxidant and cytotoxicity activity

Full size image

Conclusion

Eight new 1,3,4-oxadiazoles were synthesized and determined to have bioactive pharmacophores meeting the essential requirements for antioxidant and cytotoxic activities. The results revealed that compounds 1 and 3 showed strong antioxidant activities but display low cytotoxic potency. Oxadiazole derivatives 4a–e showed variability in their antioxidant activities, and in addition to some cytotoxicity that could be due to the phenolic OH. The introduction of steric (di-tert-butyl) or electron withdrawing (Cl and Br) groups to the aromatic ring caused reduced potency and selectivity, respectively, in the cancer cell lines. Preliminary investigation of the anticancer mechanism for these compounds revealed compound 4c induces apoptosis and causes cell cycle arrest at G0/G1 phase. Drug-likeness is observed in all of the compounds 1–4e, based on the Lipinski and Veber rules, but poor human intestinal absorption was predicted with compound 4d in Egan’s model.