Synthesis and in vitro biological evaluation of thiosulfinate derivatives for the treatment of human multidrug-resistant breast cancer

Abstract

Organosulfur compounds derived from Allium vegetables have long been recognized for various therapeutic effects, including anticancer activity. Allicin, one of the main biologically active components of garlic, shows promise as an anticancer agent; however, instability makes it unsuitable for clinical application. The aim of this study was to investigate the effect of stabilized allicin derivatives on human breast cancer cells in vitro. In this study, a total of 22 stabilized thiosulfinate derivatives were synthesized and screened for their in vitro antiproliferative activities against drug-sensitive (MCF-7) and multidrug-resistant (MCF-7/Dx) human adenocarcinoma breast cancer cells. Assays for cell death, apoptosis, cell cycle progression and mitochondrial bioenergetic function were performed. Seven compounds (4b, 7b, 8b, 13b, 14b, 15b and 18b) showed greater antiproliferative activity against MCF-7/Dx cells than allicin. These compounds were also selective towards multidrug-resistant (MDR) cells, a consequence attributed to collateral sensitivity. Among them, 13b exhibited the greatest anticancer activity in both MCF-7/Dx and MCF-7 cells, with IC₅₀ values of 18.54±0.24 and 46.50±1.98 μmol/L, respectively. 13b altered cellular morphology and arrested the cell cycle at the G₂/M phase. Additionally, 13b dose-dependently induced apoptosis, and inhibited cellular mitochondrial respiration in cells at rest and under stress. MDR presents a significant obstacle to the successful treatment of cancer clinically. These results demonstrate that thiosulfinate derivatives have potential as novel anticancer agents and may offer new therapeutic strategies for the treatment of chemoresistant cancers.

Introduction

Cancer is one of the primary causes of mortality worldwide, accounting for approximately 8 million deaths per year¹. While significant advances in diagnostic screening, surgical resection, and targeted therapies have been made in recent years, many types of malignancy remain difficult to cure. Currently, the most common forms of treatment include radiotherapy, surgery and chemotherapy; however, resistance to chemotherapy, whether intrinsic or acquired, presents a significant obstacle to the successful treatment of cancer.

Multidrug resistance (MDR) occurs when cancer cells become cross-resistant to a wide variety of structurally and functionally unrelated drugs and is predominately associated with increased expression of multidrug efflux transporters in the cell membrane, such as P-glycoprotein (P-gp)^2,3. P-gp facilitates ATP-dependent efflux of chemotherapeutic drugs, allowing MDR cells to maintain sub-lethal intracellular concentrations of these compounds, resulting in poor therapeutic response and prognosis in many types of cancer^4,5,6. A broad range of synthetic and natural P-gp inhibitors have been investigated, but due to lack of efficacy and dose-limiting toxicity, these have so far proven to be unsuitable for clinical application⁷.

Phytochemicals have long been known to have numerous health benefits in addition to showing potential as anticancer agents and modulators of MDR^8,9,10. Organosulfur compounds derived from Allium vegetables, such as onion (Allium cepa) and garlic (Allium sativum), are well documented as having a wide range of protective effects in numerous disease states. The consumption of these vegetables is associated with many health benefits, including a reduced risk of cancer^11,12,13,14.

Allicin, an allyl thiosulfinate, is the main biologically active compound derived from garlic, and it has been shown to exhibit various biological effects including antioxidant¹⁵, antiparasitic^16,17, antimicrobial¹⁸, antibacterial¹⁹, anti-inflammatory^20,21, and anticancer activity^22,23,24. In vitro studies have shown that allicin inhibits cancer cell proliferation through cell cycle arrest in gastric, breast and cervical cancer^25,26,27. Studies have also shown that allicin induces apoptosis in leukemia-derived cells through the depletion of cellular glutathione (GSH) and modulation of the cellular redox state²⁸, as well as significantly increasing reactive oxygen species (ROS) production in liver cancer, leading to reduced mitochondrial membrane potential, apoptosis, and cell death²⁹. Additionally, allicin and saturated short-chain thiosulfinate analogs are known to act on intracellular calpain, inhibiting invasion and migration in cancer cells³⁰. This activity inhibits the release of microparticles (MPs), which have been shown to transfer P-gp from MDR cells to drug-sensitive cells^31,32. These findings make these compounds attractive leads in targeting proliferation, metastasis and MDR in cancer^30,32,33,34. Importantly, allicin and thiosulfinate derivatives have been shown to be cancer-specific while being non-toxic to normal cells and have been reported to increase survival times in tumor-bearing mice^35,36.

The anticancer activity of allicin is widely attributed to the allylthio group; however, this group is highly unstable, and as a consequence, allicin is prone to decomposition under relatively mild conditions^37,38,39. To circumvent this instability and to increase potency and compound half-life, substitution of the labile allylthio group with saturated alkanes and/or benzyl moieties has been shown to prevent thiosulfinate degradation in storage and under biological conditions^39,40.

In the present work, we describe the synthesis and anticancer activity of a series of aromatic and aliphatic thiosulfinates against both the human adenocarcinoma breast cancer cell line MCF-7 and the MDR sub-line MCF-7/Dx. We demonstrate that these thiosulfinate derivatives show specificity towards MDR breast cancer cells and inhibit cell proliferation through the disruption of mitochondrial respiration, leading to the induction of apoptosis and cell cycle arrest.

Methods

Materials

IR spectra were recorded on an Agilent Cary 630 FTIR spectrometer (Agilent Technologies, Santa Clara, CA, USA). ¹H and ¹³C NMR spectra were recorded on an Agilent 500 MHz NMR spectrometer in deuterated chloroform (CDCl₃) unless otherwise stated. Chemical shifts are quoted relative to residual chloroform (δ 7.26 for ¹H NMR and δ 77.36 for ¹³C NMR) as an internal standard, and all chemical shifts (δ) are reported in parts per million (ppm). The coupling constants (J) are quoted in Hertz (Hz). The type of signal detected is indicated by the following abbreviations: s=singlet, d=doublet, t=triplet, q=quartet, tt=triplet of triplets, and m=multiplet. High-resolution mass spectrometry (HRMS) data were obtained using an Agilent 6510 Q-TOF LC/MS. DCM (dichloromethane), used under N₂ in the synthesis of asymmetric disulfides and the oxidation of thiosulfinates, was dried over molecular sieves. All other reagents and solvents were obtained from commercial suppliers and were used without further purification. All thiosulfinates were synthesized via a disulfide intermediate, except for allicin (1b), which was synthesized directly from diallyl disulfide purchased from Sigma-Aldrich (NSW, Australia). Thin-layer chromatography (TLC) was performed on Merck pre-coated silica gel plates (60 F₂₅₄), and spots were visualized by exposure to iodine vapor or short-wave UV light (l254 nm). Whenever required, column chromatography was performed using Scharlau silica gel 60 (230–400 mesh), with n-hexane and ethyl acetate used as eluents.

Synthesis of reaction catalysts

CsF-Celite

Celite 521 (3.20 g) was added to cesium fluoride (5.36 g, 32.0 mmol) dissolved in 80 mL of H₂O. The mixture was stirred for 30 min at RT, after which the H₂O was removed under reduced pressure. The resultant solid was then washed twice in CH₃CN, filtered, and dried in a desiccator at room temperature.

1-Chlorobenzotriazole

An 8%–12% NaOCl solution (commercial bleach) was added dropwise to a stirred solution of 50% acetic acid (32 mL) and benzotriazole (8 g, 67.2 mmol). Once the addition was complete, the reaction was stirred for 3 h. The product was then filtered and washed until the pH of the filtrate was neutral. The product was then dried under reduced pressure to obtain 1-chlorobenzotriazole (m-CBTA) as a white powder. The spectral data agree with values reported in the literature⁴¹. Yield: 94%; ¹H NMR δ (500 MHz, CDCl₃): 8.09 (1H, d, J=8.5 Hz), 7.61–7.47 (2H, m), 7.46–7.44 (1H, m) ppm; and ¹³C NMR δ (500 MHz, CDCl₃): 144.46, 134.35, 129.55, 125.35, 120.99, 109.54 ppm.

General procedures for the synthesis of symmetrical disulfides

CsF-Celite (1.59 g) was stirred in 10 mL of acetonitrile, the appropriate thiol (5 mmol) was then added, and the solution was stirred for 5–48 h. The reaction progression was monitored by TLC. Once the reaction was complete, the mixture was filtered, and the filtrate evaporated under reduced pressure. The resultant disulfide was characterized by ¹H and ¹³C and used in thiosulfinate synthesis without further purification.

1-(Ethyldisulfanyl)ethane (2a)⁴²

Yield: 91%, as a yellow oil; ¹H NMR δ (500 MHz, CDCl₃): 2.70 (4H, q, J=7.5 Hz, SCH₂), 1.32 (6H, t, J=7.5 Hz, CH₃) ppm; ¹³C NMR δ (500 MHz, CDCl₃): 33.27, 14.81 ppm.

1-(propyldisulfanyl)propane (3a)⁴²

Yield: 97%, as a yellow oil; ¹H NMR δ (500 MHz, CDCl₃): 2.67 (4H, t, J=7.2 Hz, SCH₂), 1.71 (4H, sextet, J=7.2 Hz, SCH₂CH₂), 1.00 (6H, t, J=7.2 Hz, CH₃) ppm; ¹³C NMR δ (500 MHz, CDCl₃): 41.54, 22.86, 13.47 ppm.

1-(Butyldisulfanyl)butane (4a)⁴²

Yield: 92%, as a yellow oil; ¹H NMR δ (500 MHz, CDCl₃): 2.69 (4H, t, J=7.5 Hz, SCH₂), 1.65 (4H, quintet, J=7.5 Hz, SCH₂CH₂), 1.41 (4H, sextet, J=7.5 Hz, CH₂CH₃), 0.92 (6H, t, J=7.5 Hz, CH₃) ppm; ¹³C NMR δ (500 MHz, CDCl₃): 39.25, 31.67, 22.01, 14.04 ppm.

1-(Hexyldisulfanyl)hexane (5a)⁴²

Yield: 62%, as a yellow oil; ¹H NMR δ (500 MHz, CDCl₃): 2.68 (4H, t, J=6.8 Hz, SCH₂), 1.67 (4H, quintet, J=6.8 Hz, SCH₂CH₂), 1.41–1.26 (12H, m, CH₂CH₂CH₂CH₃), 0.89 (6H, t, J=6.8 Hz, CH₃), ppm; ¹³C NMR δ (500 MHz, CDCl₃): 39.58, 31.79, 31.27, 29.54, 28.56, 14.37 ppm.

1-(4-Methylphenyldisulfanyl)-4-methylbenzene (6a)⁴³

Yield: 90%, as a white solid; ¹H NMR δ (500 MHz, CDCl₃): 7.38 (4H, d, J=8.0 Hz, ArH-2), 7.10 (4H, d, J=8.0 Hz, ArH-3), 2.32 (6H, s, CH₃) ppm; ¹³C NMR δ (500 MHz, CDCl₃): 137.80, 134.26, 130.14, 128.90, 21.41 ppm.

1-(4-Methoxyphenyldisulfanyl)-4-methoxylbenzene (13a)⁴³

Yield: 91%, as a yellow oil; ¹H NMR δ (500 MHz, CDCl₃): 7.40 (4H, d, J=9.0 Hz, ArH-2), 6.83 (4H, d, J=9.0 Hz, ArH-3), 3.80 (6H, s, OCH₃) ppm; ¹³C NMR δ (500 MHz, CDCl₃): 160.27, 133.02, 128.80, 114.96, 55.72 ppm.

1-(4-Tert-butylbenzyldisulfanyl)-4-tert-butylbenzene-1-methane (20a)⁴³

Yield: 82%, as an orange solid; ¹H NMR δ (500 MHz, CDCl₃): 7.34 (2H, d, J=8.0 Hz, ArH-2), 7.17 (2H, d, J=8.0 Hz, ArH-3), 3.60 (2H, s, CH₂), 1.31 (9H, s, C(CH₃)₃) ppm; ¹³C NMR δ (500 MHz, CDCl₃): 150.84, 134.57, 129.45, 125.76, 43.33, 34.90, 31.71 ppm.

General procedures for the synthesis of asymmetric disulfides

m-CBTA (0.61 g, 4.0 mmol) and benzotriazole (0.36 g, 3.0 mmol) were dissolved in dry DCM and stirred under a N₂ atmosphere at RT. The temperature was then reduced to -78 °C, and the appropriate thiol (R₁, 3.0 mmol, dissolved in dry DCM) was slowly added dropwise. The reaction mixture was then left to stir for 2 h. The second thiol (R₂, 3.2 mmol, dissolved in dry DCM) was then added slowly to the reaction. Once the addition was complete, the reaction mixture was allowed to warm to 0 °C and then stirred for a further 30 min. The reaction was then quenched with Na₂S₂O₃(0.5 g in 15 mL H₂O) and stirred for a further 10 min at 0 °C. The organic fraction was then extracted with aqueous NaHCO₃ and then twice more with H₂O. The product was dried over anhydrous MgSO₄, and the solvent was removed under reduced pressure. The resultant disulfide was characterized by ¹H and ¹³C and used in thiosulfinate synthesis without further purification.

1-(Ethyldisulfanyl)-4-methylbenzene (7a)⁴⁴

Yield: 76%, as a yellow oil; ¹H NMR δ (500 MHz, CDCl₃): 7.43 (2H, d, J=8.5 Hz, ArH-2), 7.13 (2H, d, J=8.5 Hz, ArH-3), 2.74 (2H, q, J=7.0 Hz, CH₂CH₃), 2.33 (3H, s, C₆H₄CH₃), 1.31 (3H, t, J=7.0 Hz, CH₂CH₃) ppm; ¹³C NMR δ (500 MHz, CDCl₃): 137.33, 134.57, 130.07, 128.69, 32.98, 21.37, 14.46 ppm.

1-(Propyldisulfanyl)-4-methylbenzene (8a)⁴⁰

Yield: 89%, as a yellow oil; ¹H NMR δ (500 MHz, CDCl₃): 7.43 (2H, d, J=8.5 Hz, ArH-2), 7.13 (2H, d, J=8.5 Hz, ArH-3), 2.71 (2H, t, J=7.3 Hz, SCH₂), 2.33 (3H, s, C₆H₄CH₃), 1.70 (2H, sextet, J=7.3 Hz, CH₂CH₃), 0.96 (3H, t, J=7.3 Hz, CH₂CH₃) ppm; ¹³C NMR δ (500 MHz, CDCl₃): 137.29, 134.55, 130.06, 128.64, 41.21, 22.48, 21.37, 13.44 ppm.

1-(Tert-butyldisulfanyl)-4-methylbenzene (9a)⁴⁵

Yield: 92%, as a yellow oil; ¹H NMR δ (500 MHz, CDCl₃): 7.44 (2H, d, J=8.5 Hz, ArH-2), 7.10 (2H, d, J=8.5 Hz, ArH-3), 2.32 (3H, s, C₆H₄CH₃), 1.30 (9H, s, C(CH₃)₃) ppm; ¹³C NMR δ (500 MHz, CDCl₃): 136.65, 135.71, 129.85, 127.76, 49.39, 30.22, 21.32 ppm.

1-(4-Tert-butylbenzyldisulfanyl)-4-methylbenzene (10a)

Yield: 54%, as a yellow oil; ¹H NMR δ (500 MHz, CDCl₃): 7.32 (2H, d, J=8.0 Hz, MeArH-2), 7.30 (2H, d, J=8.0 Hz, ArH-2), 7.21 (2H, d, J=8.0 Hz, ArH-3), 7.08 (2H, d, J=8.0 Hz, MeArH-3), 3.93 (2H, s, CH₂), 2.32 (3H, s, C₆H₄CH₃), 1.30 (9H, s, C(CH₃)₃) ppm; ¹³C NMR δ (500 MHz, CDCl₃): 150.83, 137.32, 134.57, 134.09, 129.99, 129.45, 128.88, 125.75, 43.33, 34.89, 31.70, 21.37 ppm.

1-(4-Methylphenyldisulfanyl)-4-methoxybenzene (11a)⁴⁶

Yield: 67%, as a yellow oil; ¹H NMR δ (500 MHz, CDCl₃): 7.41 (2H, d, J=8.5 Hz, MeOArH-2), 7.38 (2H, d, J=8.5 Hz, ArH-2), 7.11 (2H, d, J=8.5 Hz, ArH-3), 6.83 (2H, d, J=8.5 Hz, MeOArH-3), 3.79 (3H, s, OCH₃), 2.33 (3H, s, C₆H₄CH₃) ppm; ¹³C NMR δ (500 MHz, CDCl₃): 160.14, 137.96, 134.41, 132.30, 130.12, 128.88, 128.66, 114.00, 55.71, 21.43 ppm.

1-(Ethyldisulfanyl)-4-methoxybenzene (14a)

Yield: 76%, as a yellow oil; ¹H NMR δ (500 MHz, CDCl₃): 7.48 (2H, d, J=9.0 Hz, ArH-2), 6.86 (2H, d, J=9.0 Hz, ArH-3), 3.80 (3H, s, OCH₃), 2.74 (2H, q, J=7.5 Hz, SCH₂), 1.31 (3H, t, J=7.5 Hz, CH₂CH₃) ppm; ¹³C NMR δ (500 MHz, CDCl₃): 159.84, 131.99, 128.90, 114.97, 55.74, 32.91, 14.42 ppm.

1-(Butyldisulfanyl)-4-methoxybenzene (15a)⁴⁷

Yield: 69%, as a yellow oil; ¹H NMR δ (500 MHz, CDCl₃): 7.48 (2H, d, J=9.0 Hz, ArH-2), 6.86 (2H, d, J=9.0 Hz, ArH-3), 3.81 (3H, s, OCH₃), 2.73 (2H, t, J=7.5 Hz, SCH₂), 1.65 (2H, quintet, J=7.5 Hz, SCH₂CH₂), 1.38 (2H, sextet, J=7.5 Hz, CH₂CH₃), 0.89 (3H, t, J=7.5 Hz, CH₂CH₃) ppm; ¹³C NMR δ (500 MHz, CDCl₃): 159.83, 131.97, 128.90, 114.97, 55.75, 38.87, 31.13, 21.97, 13.98 ppm.

1-(Tert-butyldisulfanyl)-4-methoxybenzene (16a)⁴⁵

Yield: 84%, as a yellow oil; ¹H NMR δ (500 MHz, CDCl₃): 7.49 (2H, d, J=9.0 Hz, ArH-2), 6.84 (2H, d, J=9.0 Hz, ArH-3), 3.79 (3H, s, OCH₃), 1.29 (9H, s, C(CH₃)₃) ppm; ¹³C NMR δ (500 MHz, CDCl₃): 159.27, 130.63, 130.11, 114.77, 55.71, 49.28, 30.25 ppm.

1-(Hexyldisulfanyl)-4-methoxybenzene (17a)⁴⁶

Yield: 85%, as a yellow oil; ¹H NMR δ (500 MHz, CDCl₃): 7.48 (2H, d, J=9.0 Hz, ArH-2), 6.86 (2H, d, J=9.0 Hz, ArH-3), 3.81 (3H, s, OCH₃), 2.73 (2H, t, J=7.5 Hz, SCH₂), 1.66 (2H, quintet, J=7.5 Hz, SCH₂CH₂), 1.38-1.23 (6H, m, CH₂CH₂CH₂CH₃), 0.87 (3H, t, J=7.5 Hz, CH₃) ppm; ¹³C NMR δ (500 MHz, CDCl₃): 159.83, 132.00, 128.93, 114.96, 55.75, 39.25, 31.72, 29.02, 28.50, 22.85, 14.36 ppm.

1-(Cyclohexyldisulfanyl)-4-methoxybenzene (18a)⁴⁷

Yield: 89%, as a yellow oil; ¹H NMR δ (500 MHz, CDCl₃): 7.47 (2H, d, J=9.0 Hz, ArH-2), 6.86 (2H, d, J=9.0 Hz, ArH-3), 3.80 (3H, s, OCH₃), 2.80 (1H, tt, J=10.5, 3.7 Hz, H-1), 2.03-2.00 (2H, m, H-6), 1.78–1.75 (2H, m, H-2), 1.61–1.58 (1H, m, H-4_a), 1.39-1.20 (5H, m, H-3, H-5 and H-4_b) ppm; ¹³C NMR δ (500 MHz, CDCl₃): 159.46, 131.03, 129.69, 114.85, 55.70, 49.99, 32.90, 26.32, 25.97 ppm.

1-(4-Tert-butylbenzyldisulfanyl)-4-methoxybenzene (19a)

Yield: 64%, as a yellow oil; ¹H NMR δ (500 MHz, CDCl₃): 7.34 (2H, d, J=9.0 Hz, MeOArH-2), 7.30 (2H, d, J=9.0 Hz, ArH-2), 7.21 (2H, d, J=9.0 Hz, ArH-3), 6.80 (2H, d, J=9.0 Hz, MeOArH-3), 3.93 (2H, s, CH₂), 3.80 (3H, s, CH₃), 1.31 (9H, s, C(CH₃)₃) ppm; ¹³C NMR δ (500 MHz, CDCl₃): 159.81, 150.78, 134.06, 132.19, 129.44, 128.47, 125.79, 114.86, 55.70, 43.46, 34.87, 31.70 ppm.

4-(Butyldisulfanyl)phenol (21a)

Yield: 41%, as a yellow oil; ¹H NMR δ (500 MHz, CDCl₃): 7.43 (2H, d, J=8.5 Hz, ArH-2), 6.80 (2H, d, J=8.5 Hz, ArH-3), 2.73 (2H, t, J=7.5 Hz, SCH₂), 1.73 (2H, quintet, J=7.5 Hz, SCH₂CH₂), 1.41 (2H, sextet, J=7.5 Hz, CH₂CH₃), 0.88 (3H, t, J=7.5 Hz, CH₃) ppm; ¹³C NMR δ (500 MHz, CDCl₃): 155.94, 132.17, 117.49, 116.46, 39.23, 31.65, 21.00, 14.03 ppm.

4-(hexyldisulfanyl)phenol (22a)

Yield: 35%, as a yellow oil; ¹H NMR δ (500 MHz, CDCl₃): 7.43 (2H, d, J=8.5 Hz, ArH-2), 6.81 (2H, d, J=8.5 Hz, ArH-3), 2.73 (2H, t, J=7.5 Hz, SCH₂), 1.74 (2H, quintet, J=7.5 Hz, SCH₂CH₂), 1.44-1.23 (6H, m, CH₂CH₂CH₂CH₃), 0.87 (3H, t, J=7.5 Hz, CH₃) ppm; ¹³C NMR δ (500 MHz, CDCl₃): 155.90, 132.15, 117.46, 116.43, 39.24, 31.70, 29.13, 28.54, 22.85, 14.35 ppm.

General procedures for the synthesis of symmetric and asymmetric thiosulfinates

The selected disulfide (3.0 mmol) was dissolved in dry DCM (5 mL) under a N₂ atmosphere at -78 °C. m-CPBA (0.52 g, 3.0 mmol), dissolved in dry DCM (5 mL), was then slowly added dropwise. Once the addition was complete, the reaction was left to stir for 3 h, slowly warming to 0 °C. The reaction was quenched with saturated NaHCO₃, and the resulting aqueous solution was extracted 3 times with DCM. The combined organic fractions were then dried over anhydrous MgSO₄ and evaporated under reduced pressure. The crude product was then purified by silica gel flash column chromatography (n-hexane/ethyl acetate). After purification, all thiosulfinates were stored at -80 °C until required. All stock solutions were prepared in DMSO and stored at -20 °C.

Allicin, S-allyl prop-2-ene-1-sulfinothioate (1b)³⁹

Yield: 23%, as a yellow oil; IR v_max cm⁻¹: 1060 (S=O); ¹H NMR δ (500 MHz, D₂O): 6.05–5.94 (m, 2H, SCH₂CH), 5.53–4.79 (m, 4H, CH=CH₂), 4.05–3.82 (m, 4H, SCH₂) ppm; ¹³C NMR δ (500 MHz, D₂O): 136.01, 128.07, 127.87, 121.00, 61.07, 38.47 ppm; HRMS m/z: calcd. for C₆H₁₀OS₂ (M+H)⁺: 163.0246; found: 163.0245.

S-ethyl ethanesulfinothioate (2b)³⁹

Yield: 46%, as a yellow oil; IR v_max cm⁻¹: 1074 (S=O); ¹H NMR δ (500 MHz, CDCl₃): 3.18–3.11 (4H, m, SCH₂), 1.46 (3H, t, J=7.5 Hz, SCH₂CH₃), 1.40 (3H, t, J=7.5 Hz, S=OCH₂CH₃) ppm; ¹³C NMR δ (500 MHz, CDCl₃): 50.26, 27.38, 16.58, 8.12 ppm; HRMS m/z: calcd. for C₄H₁₀OS₂ (M+H)⁺: 139.0246; found: 139.0243.

S-propyl propane-1-sulfinothioate (3b)³⁹

Yield: 51%, as a yellow oil; IR v_max cm⁻¹: 1076 (S=O); ¹H NMR δ (500 MHz, D₂O): 3.26-3.19 (4H, m, SCH₂), 1.86-1.80 (4H, m, SCH₂CH₂), 1.07 (3H, t, J=7.5 Hz, (SCH₂CH₂CH₃), 1.02 (3H, t, J=7.5 Hz, S=OCH₂CH₂CH₃) ppm; ¹³C NMR δ (500 MHz, D₂O): 59.65, 38.20, 26.59, 19.70, 15.07, 15.00 ppm; HRMS m/z: calcd. for C₆H₁₄OS₂ (M+H)⁺: 167.0559; found: 167.0562.

S-butyl butane-1-sulfinothioate (4b)³⁹

Yield: 57%, as a yellow oil; IR v_max cm⁻¹: 1081 (S=O); ¹H NMR δ (500 MHz, CDCl₃): 3.15–3.10 (4H, m, SCH₂), 1.82–1.75 (4H, m, SCH₂CH₂), 1.50–1.43 (4H, m, CH₂CH₃), 0.98–0.93 (6H, m, CH₃) ppm; ¹³C NMR δ (500 MHz, CDCl₃): 56.33, 33.22, 32.95, 25.81, 22.21, 22.08, 14.01, 13.86 ppm; HRMS m/z: calcd. for C₈H₁₈OS₂ (M+H)⁺: 195.0872; found: 195.0873.

S-hexyl hexane-1-sulfinothioate (5b)¹⁷

Yield: 48%, as a yellow oil; IR v_max cm⁻¹: 1079 (S=O); ¹H NMR δ (500 MHz, CDCl₃): 3.16-3.08 (4H, m, SCH₂), 1.84–1.77 (4H, m, SCH₂CH₂), 1.46–1.41 (4H, m, S(CH₂)₂CH₂), 1.35–1.29 (8H, m, CH₂CH₂CH₃), 0.91–0.88 (6H, m, CH₃) ppm; ¹³C NMR δ (500 MHz, CDCl₃): 56.62, 33.26, 31.67, 31.57, 31.17, 28.63, 28.61, 23.78, 22.83, 22.73, 14.34, 14.30 ppm; HRMS m/z: calcd. for C₁₂H₂₆OS₂ (M+H)⁺: 251.1498; found: 251.1496.

S-4-tolyl 4-methylbenzenesulfinothioate (6b)⁴⁸

Yield: 24%, as a white solid; IR v_max cm⁻¹: 1075 (S=O); ¹H NMR δ (500 MHz, CDCl₃): 7.46 (2H, d, J=8.0 Hz, ArH-2), 7.24 (2H, d, J=8.0 Hz, ArH-2'), 7.21 (2H, d, J=8.0 Hz, ArH-3), 7.14 (2H, d, J=8.0 Hz, ArH-3'), 2.42 (3H, s, CH₃'), 2.38 (3H, s, CH₃-1) ppm; ¹³C NMR δ (500 MHz, CDCl₃): 144.91, 142.38, 140.84, 136.85, 130.54, 129.71, 127.96, 124.96, 22.01, 21.83 ppm; HRMS m/z: calcd. for C₁₄H₁₄OS₂ (M+H)⁺: 263.0559; found: 263.0557.

S-4-tolyl ethanesulfinothioate (7b)

Yield: 39%, as a yellow oil; IR v_max cm⁻¹: 1079 (S=O); ¹H NMR δ (500 MHz, CDCl₃): 7.49 (2H, d, J=7.5 Hz, ArH-2), 7.22 (2H, d, J=7.5 Hz, ArH-3), 3.10 (2H, q, J=7.5 Hz, S=OCH₂), 2.38 (3H, s, C₆H₄CH₃), 1.43 (3H, t, J=7.5 Hz, CH₂CH₃) ppm; ¹³C NMR δ (500 MHz, CDCl₃): 141.06, 135.73, 130.62, 125.58, 50.22, 21.70, 8.24 ppm; HRMS m/z: calcd. for C₉H₁₂OS₂ (M+H)⁺: 201.0402; found: 201.0404.

S-4-tolyl propane-1-sulfinothioate (8b)⁴⁰

Yield: 41%, as a yellow oil; IR v_max cm⁻¹: 1080 (S=O); ¹H NMR δ (500 MHz, CDCl₃): 7.50 (2H, d, J=8.0 Hz, ArH-2), 7.23 (2H, d, J=8.0 Hz, ArH-3), 3.08 (2H, m, S=OCH₂), 2.39 (3H, s, C₆H₄CH₃), 1.89 (2H, m, CH₂CH₃), 1.10 (3H, t, J=7.5 Hz, CH₂CH₃) ppm; ¹³C NMR δ (500 MHz, CDCl₃): 141.04, 135.66, 130.59, 125.83, 58.16, 21.68, 17.59, 13.58 ppm; HRMS m/z: calcd. for C₁₀H₁₄OS₂ (M+H)⁺: 215.0559; found: 215.0558.

S-4-tolyl 2-methylpropane-2-sulfinothioate (9b)⁴⁹

Yield: 17%, as a white solid; IR v_max cm⁻¹: 1078 (S=O); ¹H NMR δ (500 MHz, CDCl₃): 7.52 (2H, d, J=8.0 Hz, ArH-2), 7.20 (2H, d, J=8.0 Hz, ArH-3), 2.38 (3H, s, C₆H₄CH₃), 1.46 (9H, s, C(CH₃)₃) ppm; ¹³C NMR δ (500 MHz, CDCl₃): 140.52, 135.59, 130.60, 126.51, 60.47, 24.50, 21.63 ppm; HRMS m/z: calcd. for C₁₁H₁₆OS₂ (M+H)⁺: 229.0715; found: 229.0715.

S-4-tolyl (4-tert-butylphenyl)methanesulfinothioate (10b)

Yield: 33%, as a white solid; IR v_max cm⁻¹: 1078 (S=O); ¹H NMR δ (500 MHz, CDCl₃): 7.44 (2H, d, J=8.5 Hz, MeArH-2), 7.41 (2H, d, J=8.5 Hz, ArH-2), 7.30 (2H, d, J=8.5 Hz, ArH-3), 7.19 (2H, d, J=8.5 Hz, MeArH-3), 4.37 (1H, d, J=13.0 Hz, CH_a), 4.25 (1H, d, J=13.0 Hz, CH_b), 2.37 (3H, s, C₆H₄CH₃), 1.33 (9H, s, C(CH₃)₃) ppm; ¹³C NMR δ (500 MHz, CDCl₃): 152.09, 141.02, 135.80, 130.59, 130.44, 127.38, 126.21, 125.80, 62.50, 35.03, 31.62, 21.69 ppm; HRMS m/z: calcd. for C₁₈H₂₂OS₂ (M+H)⁺: 319.1185; found: 319.1187.

S-4-methoxyphenyl 4-methylbenzenesulfinothioate (11b)⁴⁸

Yield: 20%, as a white solid; IR v_max cm⁻¹: 1075 (S=O); ¹H NMR δ (500 MHz, CDCl₃): 7.53 (2H, d, J=8.0 Hz, MeOArH-2), 7.43 (2H, d, J=8.0 Hz, ArH-2), 7.29 (2H, d, J=8.0 Hz, ArH-3), 6.89 (2H, d, J=8.0 Hz, MeOArH-3), 3.84 (3H, s, OCH₃) 2.42 (3H, s, C₆H₄CH₃) ppm; ¹³C NMR δ (500 MHz, CDCl₃): 161.90, 142.48, 135.80, 129.94, 129.74, 124.67, 120.32, 115.18, 55.80, 21.87 ppm; HRMS m/z: calcd. for C₁₄H₁₄O₂S₂ (M+H)⁺: 279.0508; found: 279.0504.

S-4-tolyl 4-methoxybenzenesulfinothioate (12b)⁴⁸

Yield: 20%, as a white solid; IR v_max cm⁻¹: 1075 (S=O); ¹H NMR δ (500 MHz, CDCl₃): 7.60 (2H, d, J=8.5 Hz, MeOArH-2), 7.43 (2H, d, J=8.5 Hz, ArH-2), 7.19 (2H, d, J=8.5 Hz, ArH-3), 6.99 (2H, d, J=8.5 Hz, MeOArH-3), 3.87 (3H, s, OCH₃) 2.38 (3H, s, C₆H₄CH₃) ppm; ¹³C NMR δ (500 MHz, CDCl₃): 162.60, 141.08, 137.85, 130.42, 127.97, 126.53, 126.50, 114.76, 55.94, 21.73 ppm; HRMS m/z: calcd. for C₁₄H₁₄O₂S₂ (M+H)⁺: 279.0508; found: 279.0504.

S-4-methoxyphenyl 4-methoxybenzenesulfinothioate (13b)⁵⁰

Yield: 74%, as a yellow solid; IR v_max cm⁻¹: 1073 (S=O); ¹H NMR δ (500 MHz, CDCl₃): 7.56 (2H, d, J=8.5 Hz, ArH-2), 7.42 (2H, d, J=8.5 Hz, ArH-2'), 6.98 (2H, d, J=8.5 Hz, ArH-3'), 6.89 (2H, d, J=8.5 Hz, ArH-3), 3.86 (3H, s, OCH₃-2) 3.84 (3H, s, OCH₃-1) ppm; ¹³C NMR δ (500 MHz, CDCl₃): 162.56, 162.01, 137.88, 135.55, 126.51, 120.31, 115.15, 114.71, 55.94, 55.80 ppm; HRMS m/z: calcd. for C₁₄H₁₄O₃S₂ (M+H)⁺: 295.0457; found: 295.0457.

S-4-methoxyphenyl ethanesulfinothioate (14b)

Yield: 48%, as a yellow oil; IR v_max cm⁻¹: 1076 (S=O); ¹H NMR δ (500 MHz, CDCl₃): 7.53 (2H, d, J=9.0 Hz, ArH-2), 6.94 (2H, d, J=9.0 Hz, ArH-3), 3.84 (3H, s, OCH₃), 3.07 (2H, q, J=7.5 Hz, CH₂CH₃), 1.42 (3H, t, J=7.5 Hz, CH₂CH₃) ppm; ¹³C NMR δ (500 MHz, CDCl₃): 161.95, 137.76, 119.28, 115.42, 55.80, 49.98, 8.22 ppm; HRMS m/z: calcd. for C₉H₁₂O₂S₂ (M+H)⁺: 217.0351; found: 217.0350.

S-4-methoxyphenyl butane-1-sulfinothioate (15b)

Yield: 39%, as a yellow oil; IR v_max cm⁻¹: 1080 (S=O); ¹H NMR δ (500 MHz, CDCl₃): 7.53 (2H, d, J=9.0 Hz, ArH-2), 6.94 (2H, d, J=9.0 Hz, ArH-3), 3.84 (3H, s, OCH₃), 3.06 (2H, t, J=7.5 Hz, SCH₂), 1.82 (2H, m, SCH₂CH₂), 1.45 (2H, m, CH₂CH₃), 0.98 (3H, t, J=7.5 Hz, CH₂CH₃) ppm; ¹³C NMR δ (500 MHz, CDCl₃): 161.94, 137.72, 119.57, 115.41, 55.94, 55.80, 25.84, 22.29, 14.05 ppm; HRMS m/z: calcd. for C₁₁H₁₆O₂S₂ (M+H)⁺: 245.0664; found: 245.0665.

S-4-methoxyphenyl 2-methylpropane-2-sulfinothioate (16b)

Yield: 27%, as a white solid; IR v_max cm⁻¹: 1072 (S=O); ¹H NMR δ (500 MHz, CDCl₃): 7.55 (2H, d, J=8.5 Hz, ArH-2), 6.92 (2H, d, J=8.5 Hz, ArH-3), 3.83 (3H, s, OCH₃), 1.45 (9H, s, C(CH₃)₃) ppm; ¹³C NMR δ (500 MHz, CDCl₃): 161.62, 137.63, 120.24, 115.44, 60.27, 55.78, 24.50 ppm; HRMS m/z: calcd. for C₁₁H₁₆O₂S₂ (M+H)⁺: 245.0664; found: 245.0660.

S-4-methoxyphenyl hexane-1-sulfinothioate (17b)

Yield: 51%, as a yellow oil; IR v_max cm⁻¹: 1078 (S=O); ¹H NMR δ (500 MHz, CDCl₃): 7.53 (2H, d, J=9.0 Hz, ArH-2), 6.94 (2H, d, J=9.0 Hz, ArH-3), 3.84 (3H, s, OCH₃), 3.05 (2H, t, J=7.5 Hz, SCH₂), 1.84-1.81 (2H, m, SCH₂CH₂), 1.46-1.31 (6H, m, CH₂CH₂CH₂CH₃), 0.90 (3H, t, J=7.5 Hz, CH₃) ppm; ¹³C NMR δ (500 MHz, CDCl₃): 161.94, 137.72, 119.59, 115.41, 56.21, 55.80, 31.69, 28.70, 23.80, 22.73, 14.31 ppm; HRMS m/z: calcd. for C₁₃H₂₀O₂S₂ (M+H)⁺: 273.0977; found: 273.0974.

S-4-methoxyphenyl cyclohexanesulfinothioate (18b)

Yield: 38%, as a white solid; IR v_max cm⁻¹: 1077 (S=O); ¹H NMR δ (500 MHz, CDCl₃): 7.52 (2H, d, J=8.5 Hz, ArH-2), 6.92 (2H, d, J=8.5 Hz, ArH-3), 3.82 (3H, s, OCH₃), 2.99 (1H, tt, J=11.5, 3.9 Hz, H-1), 2.17–2.14 (2H, m, H-6), 1.94–1.88 (2H, m, H-2), 1.71–1.50 (3H, m, H-3 and H-4_a), 1.41–1.25 (3H, m, H-5 and H-4_b) ppm; ¹³C NMR δ (500 MHz, CDCl₃): 161.71, 137.56, 119.86, 115.35, 63.74, 55.73, 27.50, 26.52, 25.86, 25.77, 25.69 ppm; HRMS m/z: calcd. for C₁₃H₁₈O₂S₂ (M+H)⁺: 271.0821; found: 271.0824.

S-4-methoxyphenyl (4-tert-butylphenyl)methanesulfinothioate (19b)

Yield: 29%, as a yellow oil; IR v_max cm⁻¹: 1074 (S=O); ¹H NMR δ (500 MHz, CDCl₃): 7.48 (2H, d, J=8.5 Hz, MeOArH-2), 7.41 (2H, d, J=8.5 Hz, ArH-2), 7.30 (2H, d, J=8.5 Hz, ArH-3), 6.91 (2H, d, J=8.5 Hz, MeOArH-3), 4.34, (1H, d, J=13.0 Hz, CH_a), 4.22 (1H, d, J=13.0 Hz, CH_b), 3.82 (3H, s, OCH₃), 1.33 (9H, s, C(CH₃)₃) ppm; ¹³C NMR δ (500 MHz, CDCl₃): 161.92, 152.06, 137.84, 130.42, 127.42, 126.21, 119.54, 115.39, 62.35, 55.79, 35.03, 31.62 ppm; HRMS m/z: calcd. for C₁₈H₂₂O₂S₂ (M+H)⁺: 335.1134; found: 335.1137.

S-4-tert-butylphenyl (4-tert-butylphenyl)methanesulfinothioate (20b)

Yield: 67%, as a white solid; IR v_max cm⁻¹: 1067 (S=O); ¹H NMR δ (500 MHz, CDCl₃): 7.38 (2H, d, J=8.5 Hz, ArH-2), 7.33 (2H, d, J=8.5 Hz, ArH-2'), 7.25 (2H, d, J=8.5 Hz, ArH-3'), 7.24 (2H, d, J=8.5 Hz, ArH-3), 4.29–4.25 (4H, m, CH₂), 1.31 (9H, s, C(CH₃)₃-1), 1.30 (9H, s, C(CH₃)₃-2) ppm; ¹³C NMR δ (500 MHz, CDCl₃): 152.04, 151.08, 133.82, 130.387, 129.17, 127.31, 126.11, 126.06, 62.15, 36.14, 34.97, 34.88, 31.61, 31.59 ppm; HRMS m/z: calcd. for C₂₂H₃₀OS₂ (M+H)⁺: 375.1811; found: 375.1811.

S-4-hydroxyphenyl butane-1-sulfinothioate (21b)

Yield: 17%, as a yellow oil; IR v_max cm⁻¹: 1036 (S=O); ¹H NMR δ (500 MHz, CDCl₃): 7.40 (2H, d, J=8.5 Hz, ArH-2), 6.75 (2H, d, J=8.5 Hz, ArH-3), 3.11 (2H, t, J=7.5 Hz, SCH₂), 1.84 (2H, dq, J=7.5, 3.5 Hz, SCH₂CH₂), 1.51 (2H, ds, J=7.5, 3.5 Hz, CH₂CH₃), 0.99 (3H, t, J=7.5 Hz, CH₃) ppm; ¹³C NMR δ (500 MHz, CDCl₃): 159.46, 138.09, 117.53, 117.39, 55.39, 25.84, 22.25, 14.03 ppm; HRMS m/ z: calcd. for C₁₀H₁₄O₂S₂ (M+H)⁺: 231.0508; found: 231.0507.

S-4-hydroxyphenyl hexane-1-sulfinothioate (22b)

Yield: 12%, as a yellow oil; IR v_max cm⁻¹: 1038 (S=O); ¹H NMR δ (500 MHz, CDCl₃): 7.40 (2H, d, J=8.5 Hz, ArH-2), 6.74 (2H, d, J=8.5 Hz, ArH-3), 3.10 (2H, t, J=7.5 Hz, SCH₂), 1.85 (2H, quintet, J=7.5 Hz, SCH₂CH₂), 1.49–1.46 (2H, m, SCH₂CH₂CH₂), 1.36–1.32 (4H, m, CH₂CH₂CH₃), 0.91 (3H, t, J=7.5 Hz, CH₃) ppm; ¹³C NMR δ (500 MHz, CDCl₃): 159.37, 138.09, 117.60, 117.37, 55.66, 31.65, 28.65, 23.80, 22.72, 14.30 ppm; HRMS m/z: calcd. for C₁₂H₁₈O₂S₂ (M+H)⁺: 259.0821; found: 259.0824.

Cell culture

The drug-sensitive human breast adenocarcinoma cell line MCF-7 and its drug-resistant sub-line, MCF-7/Dx (designated Dx for simplicity), were used in this study. Cells were grown in RPMI-1640 media (Sigma-Aldrich, NSW, Australia) supplemented with 10% heat-inactivated fetal bovine serum (Invitrogen, Life Technologies, Victoria, Australia) and maintained at 37 °C and 5% CO₂. Dx cells were initially developed from MCF-7 cells by incremental exposure to doxorubicin⁵¹. To maintain the MDR phenotype, Dx cells were cultured in the presence of 1 μg/mL doxorubicin; however, during experiments, cells were cultured in the absence of doxorubicin. Cells were routinely tested for mycoplasma contamination.

Cell antiproliferative activity assay

The anticancer activity of the synthesized compounds was determined using the MTS assay. Cells were seeded in 96-well plates (5×10³ cells per well) and, once adherent, were cultured for 72 h in the presence of selected thiosulfinates at the indicated concentrations. Due to the short half-life of allicin, cells were treated with fresh media supplemented with the appropriate concentration of allicin every 24 h. Control cells were incubated with DMSO at 0.1% (v/v). Cell viability was determined using the CellTiter 96^® AQueous One Solution Cell Proliferation Assay (Promega, USA) according to the manufacturer's recommendation. Briefly, following 72 h of treatment, 20 μL of the MTS reagent 3-(4,5-dimethyl-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium was added to each well and then incubated for 2 h in the dark at 37 °C. Absorbance values were then obtained at 490 nm (infinite M1000Pro, Tecan), and blank-subtracted absorbance values were normalized to the vehicle control, which was arbitrarily assigned as 100%. IC₅₀ values were defined as the drug concentration that prevented cell growth of more than 50% relative to the vehicle control and were determined using nonlinear regression analysis with Prism 7.0 (GraphPad Software, CA, USA).

Cell cycle analysis

Flow cytometric analysis was performed to calculate the distribution of the cell population through cell cycle phases. Cells were seeded in 6-well plates (5×10⁵ cells per well) and, once adherent, were cultured for 24 h in the presence of selected thiosulfinates at the indicated concentrations. Colchicine was used as a positive control at 1 μmol/L (Supplementary material), and control cells were incubated with DMSO at 0.1% (v/v). After incubation, the cells were harvested, washed twice with PBS, and fixed with 70% ice-cold ethanol for at least 24 h. The fixed cells were then washed twice with PBS, resuspended in fresh PBS containing 0.1 mg/mL RNase A, 20 μg/mL propidium iodide (PI) and 0.1% (v/v) Triton X-100, and then incubated in the dark at 37 °C for 30 min. After incubation, the cells were analyzed by flow cytometry (LSRFortessa X-20, BD Biosciences), with 1×10⁴ events per sample acquired. Cell cycle distribution was calculated using FlowJo version 10.1 (Tree Star, Inc, Australia).

Apoptosis studies

The extent of apoptosis was quantitatively measured using the Annexin V binding assay according to the manufacturer's recommendation (BD Biosciences, Sydney, NSW, Australia). Cells were seeded in 6-well plates (5×10⁵ cells per well) and, once adherent, were cultured for 24 h in the presence of selected thiosulfinates at the indicated concentrations. Control cells were incubated with DMSO at 0.1% (v/v). After the incubation period, cells were harvested and washed twice with PBS and then resuspended in 1×binding buffer (1×10⁶ cells/mL). Next, 5 μL of Annexin V and 5 μL of PI (BD Pharmingen, North Ryde, NSW, Australia) were added to a 100-μL aliquot of the cell suspension, which was then vortexed and incubated at RT in the dark for 15 min. Subsequently, 200 μL of 1×binding buffer was added to the mixture, and the cells were analyzed immediately by flow cytometry (LSRFortessa X-20, BD Biosciences), with 1×10⁴ events per sample acquired. The percentages of viable (Annexin V⁻/PI⁻), early apoptotic (Annexin V⁺/PI⁻), late apoptotic (Annexin V⁺/PI⁺), and necrotic (Annexin V⁻/PI⁺) cells were calculated using FlowJo version 10.1 (Tree Star, Inc).

Measurement of mitochondrial function

Mitochondrial function was measured by determining the oxygen consumption rate (OCR) of cells with a Seahorse XF24 extracellular flux analyzer (Seahorse Bioscience, MA, USA) according to the manufacturer's protocol. Cells were seeded in 24-well XF cell culture microplates (2×10⁴ cells per well) and, once adherent, were cultured with selected thiosulfinates for 24 h at 37 °C and 5% CO₂. Prior to analysis, cells were washed, and the supernatant was replaced with buffered XF Base Medium supplemented with 2 mmol/L L-glutamine, 10 mmol/L glucose and 2 mmol/L sodium pyruvate at pH=7.4. The cells were incubated at 37 °C without CO₂ for an hour, and then, the OCR was measured utilizing an XF Cell Mito Stress Test Kit (Seahorse Bioscience, MA, USA). Oligomycin (final concentration 1 μmol/L), carbonylcyanide m-chlorophenylhydrazone (FCCP-final concentration 0.5 μmol/L), and rotenone/antimycin A (final concentration 0.5 μmol/L) were added to the sensor cartridge, and the OCR was measured using the standard cycling program.

Cell imaging

Cells were treated with thiosulfinates to examine effects on cellular morphology. Cells were seeded in 6-well plates (7×10⁵ cells per well) and, once adherent, were cultured for 24 h in the presence of selected thiosulfinates at the indicated concentrations. Control cells were incubated with DMSO at 0.1% (v/v). After the incubation period, images were captured on a digital camera (INFINITY 1, Lumenera, Ottawa).

Trypan blue exclusion assay

Cells were harvested and stained with Trypan blue, and viable cells were counted on a hemocytometer. Cell mortality was expressed as a percentage of viable cells relative to the vehicle control. The vehicle control DMSO concentration did not exceed 0.1% (v/v).

Calculation of drug-likeness properties and toxicity prediction

Toxicity risk parameters and drug-likeness properties were calculated using OSIRIS Property Explorer. OSIRIS analyzes molecules with respect to Lipinski's rule of five, generating both a drug-likeness score and an overall drug score in addition to identifying any toxicity risks^52,53. To comply with Lipinski's rule of five, compounds must have a molecular weight of less than 500 g/mol, LogP≤5, hydrogen bond donors ≤5, hydrogen bond acceptors ≤10, a solubility score between -5 and 1, and a polar surface area less than 120 Ų.

Statistical analysis

For in vitro assays, data were analyzed using one-way analysis of variance (ANOVA) followed by Tukey's post hoc analysis. P<0.05 were accepted as being statistically significant. All statistical analyses were performed using GraphPad Prism version 7.0 (GraphPad Software, CA, USA). The results are expressed as the mean±SEM unless otherwise noted.

Results

Chemistry

The symmetric disulfides 2a–6a, 13a and 20a were synthesized via oxidative coupling (Scheme 1, Eq 1) using CsF-Celite as a solid base⁵⁴. Synthesis of the short chain disulfides, 2a–4a, was carried out at room temperature, after which any remaining starting materials were removed under reduced pressure to generate the pure disulfides at good yields (>90%). For disulfides 5a–6a, 13a and 20a, it was found that at room temperature, only trace amounts of the product were formed after a 48-h period. These compounds were instead synthesized under reflux, which improved yields.

Figure 7

Synthetic methods for the preparation of symmetric thiosulfinates 1b–6b, 13b and 20b.

PowerPoint slide

The asymmetric disulfides 7a–11a, 14a–19a, and 21a–22a were synthesized by nucleophilic substitution (Scheme 2, Eq 3) via a sulfenyl chloride intermediate⁴⁶, and all symmetric and asymmetric disulfides were used in the next synthesis step without any further purification. Symmetric and asymmetric disulfides were converted to the corresponding thiosulfinate (Eq 2 and 4) by oxidation, using m-chloroperoxybenzoic acid, to provide our target thiosulfinates.

Figure 8

Synthetic methods for the preparation of asymmetric thiosulfinates 7b–12b, 14b–19b and 21b–22b.

PowerPoint slide

The formation of thiosulfonates (R₁-S-SO₂-R₂) was totally avoided when the reactions were performed at -78 °C⁵⁵. TLC and NMR analysis of the crude reaction mixtures confirmed the absence of thiosulfonate formation at this temperature (data not shown). Crude thiosulfinate mixtures were purified by flash column chromatography, with ethyl acetate and n-hexane used as an eluent. The structures of the final compounds 1b–22b (Figure 1) were confirmed by analysis of their spectral and analytical data (Supplementary material). Compounds 11b and 12b were produced from the oxidation of compound 11a, and due to their almost identical physical properties, separation was not possible during the purification step. 11b and 12b, formed at a ratio of 4:3, were treated as a mixture and will, for simplicity, henceforth be collectively referred to as 11b. To evaluate degradation associated with the storage conditions, compounds 7b and 8b were dissolved in chloroform and stored at room temperature for 3 weeks. Spectral analysis did not show decomposition, but allicin stored under similar conditions showed a significant increase in degradation by-products (data not shown). Compound 13b was also found to be significantly more stable than allicin. Spectral analysis of pure 13b did not show signs of decomposition after >12 months of storage at room temperature (data not shown).

Figure 1

Chemical structures of synthesised thiosulfinates.

PowerPoint slide

The half-life of allicin is drastically increased when stored at -80 °C, with no detectable loss of purity after 24 months of storage at this temperature⁵⁶. Consequently, all compounds were stored under these conditions until required. The stability of thiosulfinates is significantly improved when dissolved in polar solvents; therefore, all working stock solutions were dissolved in DMSO and stored at -20 °C⁵⁷. To ensure that these compounds were suitable for biological application, we analyzed each with OSIRIS Property Explorer to determine the toxicity risks and drug-likeness of each molecule (Supplementary material, Table S1). Each compound was found to be free of any toxicity risks, with all compounds free of mutagenic, tumorigenic, reproductive or irritant-like effects. Of the compounds tested, only 20b had Lipinski violations, due to a high predicted LogP and poor solubility, indicating that poor absorption or poor permeability is likely⁵⁸. This compound did show some anticancer activity against both MCF-7 and Dx cells (Supplementary material); however, in agreement with the OSIRIS prediction, it was poorly soluble in both DMSO and aqueous solutions and as such was excluded from further testing.

In vitro anticancer activity

The in vitro antiproliferative activities of the prepared compounds were evaluated against MCF-7 and MCF-7/Dx cells. All thiosulfinates were initially screened at a single high dose (50 μmol/L). Compounds that satisfied a predetermined level of cellular viability (<50%), relative to a vehicle control, were selected for further analysis.

It has previously been shown that after 24 h of incubation in water at 37 °C at a pH of 7.5, only 62% of pure allicin remains in a test sample⁵⁹. Therefore, for the 72-h viability assays, the cell culture media was replaced daily with fresh allicin medium. This considerably improved anticancer activity against both cell lines compared to experiments in which cells were treated in the same allicin stock for 72 h (Supplementary material).

In preliminary screenings, compounds allicin, 4b, 7b, 8b, 11b, and 13b–19b showed pronounced growth inhibition against MDR Dx cells at 50 μmol/L. However, at the same concentration, only allicin and 13b reduced the cellular viability of MCF-7 cells below 50%. The thiosulfinate derivatives showed pronounced selectivity towards the MDR Dx cells compared to the drug-sensitive MCF-7 cells (Figure 2A).

Figure 2

Antiproliferative activity of thiosulfinates. (A) Specificity of compounds 1b, 4b, 7b, 8b, 11b and 13b–19b at 50 μmol/L towards MDR MCF-7/Dx cells, compared to drug sensitive MCF-7 cells. (B) Comparison of molecule potency at 50 μmol/L towards MCF/Dx cells, where side chain R₂ is conserved and electron donating functional group (R₁) is altered. Data represents the mean±SEM of at least 3 independent experiments.

PowerPoint slide

Amongst the aromatic thiosulfinates, there was a clear correlation between anticancer activity and the substituent attached to the sulfenyl sulfur (R₁), with the more electron-rich 4-methoxybenzyl group showing higher antiproliferative effects compared to the 4-methylbenzyl group (Figure 2B). Additionally, for the asymmetric thiosulfinates, the longer and/or bulkier the substituent on the sulfinyl sulfur (R₂), the less potent the compound's anticancer activity. This trend was also apparent at lower concentrations (Supplementary material). Accordingly, 4b, 7b, 8b, 11b, and 13a–19b were selected for further analysis. The IC₅₀ values of allicin, 4b, 7b, 8b, 11b, and 13a–19b were determined and are summarized in Table 1. As with the initial screening, the compounds with the most electron-rich group on the sulfenyl sulfur exhibited stronger antiproliferative activity, with 13b (both R groups occupied by the 4-methoxybenzyl moiety) exhibiting the greatest anticancer activity in both MCF-7 and Dx cells. Due to its promising cytotoxic activity against both MCF-7 and Dx cells, 13b was selected for further studies.

Table 1 In-vitro anticancer activity (IC₅₀)^a of selected active molecules.

Cellular morphology

To observe the effects of thiosulfinate treatment on cellular morphology, MCF-7/Dx cells were treated with various concentrations of 13b for 24 h. The morphology of Dx cells treated with 25 μmol/L 13b was not significantly affected after 24 h. However, at 37.5 and 50 μmol/L, cells displayed distinct membrane blebbing, cell shrinkage, detachment and an apparent disaggregation (Figure 3). This reduction in viable cells was confirmed using the Trypan blue exclusion assay, which after treatment with 37.5 and 50 μmol/L 13b resulted in a cell viability reduction to 43.13%±2.07% and 27.29%±5.79%, respectively.

Figure 3

Effects of thiosulfinate 13b on the cellular morphology and viability of MCF-7/Dx cells. Images show morphological changes in MCF-7/Dx cells treated for 24 h with A) vehicle control, B) 25 μmol/L, C) 37.5 μmol/L and D) 50 μmol/L. There was no decrease in cell viability at 12.5 μmol/L (98.33%±1.86%), a slight decrease at 25 μmol/L (91.75%±4.31%), and significant decreases at 37.5 μmol/L and 50 μmol/L (43.13%±2.07% and 27.29%±5.79%, respectively). Data and images represent the mean±SEM of 3 independent experiments.

PowerPoint slide

Cell cycle analysis

Many cytotoxic compounds inhibit cell growth by arresting cells at particular stages in the cell cycle. To investigate whether the synthesized thiosulfinates had this effect, MCF-7/Dx cells were incubated with various concentrations of 13b for 24 h and then stained with PI as described above in the Materials and Methods section. Consistent with its effect on cell growth inhibition, 13b caused cell cycle arrest in Dx cells in a dose-dependent manner (Figure 4). Cells treated with the vehicle control showed a typical distribution of 44.30% of cells in the G₀/G₁ phase, 38.13% in the S phase, and 16.70% of cells in the G₂/M phase (Table 2). Treatment of cells with 12.5 μmol/L 13b increased the proportion of cells in the G₀/G₁ phase to 50.54% and caused a small decrease in cells in the S phase. At 37.5 μmol/L and 50 μmol/L, there were reductions in the proportion of cells in the G₀/G₁ and S phases, while the number of cells in G₂/M increased significantly. After being treated with 50 μmol/L, 29.37% of cells were found to be in the G₀/G₁ phase, 26.17% in the S phase, and 44.83% arrested in the G₂/M phase. During the screening, many thiosulfinates were also found to significantly arrest Dx cells in the G₂/M phase (Supplementary material). These data demonstrate that synthesized thiosulfinates primarily arrest Dx cells in the G₂/M phase, significantly reducing the proportion of cells in the G₀/G₁ and S phases.

Figure 4

Effect of 13b on the cell cycle progression of MCF-7/Dx cells. Cells were treated for 24 h with (A) DMSO (vehicle control), (B) 25 μmol/L, (C) 37.5 μmol/L and (D) 50 μmol/L. Images are representative of 4 independent experiments.

PowerPoint slide

Table 2 Effect of compound 13b on the cell cycle progression of MCF-7/Dx cells^a. ^*P<0.05, ^**P<0.01 compared with the indicated compound to control cells in the corresponding cell cycle phase.

Induction of apoptosis

The translocation of phospholipid phosphatidylserine (PS) to the outer leaflet of the cell plasma membrane is one of the hallmarks of apoptosis⁶⁰. This is followed by a loss of membrane integrity and eventual cell death. To establish whether thiosulfinates induce cell death via this mechanism, MCF-7/Dx cells were treated with 12.5–50 μmol/L of 13b for 24 h and then stained with Annexin V-450 (which binds to exposed PS) and PI to determine the proportion of cells undergoing early apoptosis (Annexin-positive, PI-negative), late apoptosis (Annexin-positive, PI-positive), and total apoptosis (early+late apoptosis) (Figure 5).

Figure 5

Effect of compound 13b on the induction of apoptosis in MCF-7/Dx cells. MCF-7/Dx cells were treated with various concentrations of 13b for 24 h, stained with Annexin V-450 and PI, and the apoptotic effect was assessed by flow cytometry. Representative results are shown, with quadrants indicating the proportion of cells that are necrotic: Q1, late apoptotic: Q2, early apoptotic: Q3, and viable: Q4. Images are representative of 3 independent experiments.

PowerPoint slide

At 12 μmol/L, there was no significant increase in the total number of apoptotic cells relative to the solvent control. A small increase was observed in cells treated with 25 μmol/L, and significant increases were observed in Dx cells treated with 37.5 and 50 μmol/L 13b. The proportion of cells undergoing apoptosis at 25 μmol/L, 37.5 μmol/L and 50 μmol/L was 7.06%±0.82%, 31.32%±8.06% and 57.52%±2.92%, respectively, indicating that 13b induces apoptosis in a dose-dependent manner (Table 3).

Table 3 Quantitative apoptosis assay of MCF-7/Dx cells treated with 13b^a. ^**P<0.01 vs the % of apoptotic cells of the control.

Measurement of mitochondrial function

Mitochondria play a central role in many biological processes, especially in the initiation of apoptosis. To further investigate the cytotoxic effects of thiosulfinates on MCF-7/Dx cells, we sought to determine whether 13b disrupted mitochondrial respiratory function, which was investigated using the XF24 extracellular flux analyzer. As allicin has previously been shown to interfere with mitochondrial functions in various cell lines, including MCF-7, we included it for comparison^25,29.

Dx cells were treated with 10–50 μmol/L of allicin and 13b for 24 h, with both OCR and extracellular acidification rate (ECAR) measured using the “mitochondrial stress test” (Agilent Technologies, CA, USA). Sequential addition of the mitochondrial complex inhibitors oligomycin, FCCP, antimycin A and rotenone allowed for the determination of several mitochondrial function parameters. Treatment with allicin and 13b inhibited the basal OCR relative to the vehicle control in a dose-dependent manner (Figure 6). The addition of oligomycin, an ATP synthase inhibitor, allows for the determination of ATP-linked respiration, with both allicin and 13b inducing a decrease. The addition of FCCP, an electron transport chain uncoupler, allows for the measurement of maximum cellular respiration rate in response to increased energy demand. There was a significant dose-dependent decrease observed for both compounds. Likewise, spare respiratory capacity was significantly impaired after treatment with both thiosulfinates.

Figure 6

Effect of thiosulfinates on mitochondrial function. Cells were seeded in XF24 well plates and treated for 24 h with (A) 13b and (B) allicin at the indicated concentrations. Medium was then replaced with unbuffered XF assay medium, supplemented with sodium pyruvate, glucose and L-glutamine, and allowed to equilibrate for 1 h. Baseline measurements were recorded then oligomycin, FCCP, antimycin A and rotenone were injected sequentially at the indicated time points. Data represents the mean±SEM of at least 4 independent experiments. ^*P<0.05, ^**P<0.01, compared with the indicated compound to the vehicle control.

PowerPoint slide

When cells lose the ability to maintain oxidative phosphorylation/respiration, they often compensate by increasing glycolysis to maintain ATP production⁶¹. Due to the decrease observed in the OCR, we investigated the effects of allicin and 13b on the ECAR, a surrogate marker for glycolysis⁶². The glycolytic potential was also measured by the addition of oligomycin, which forces the cells to compensate by increasing glycolysis (Supplementary material). Despite decreases in the OCR at all tested concentrations, there was no significant effect on basal ECAR or glycolytic potential.

Discussion

Allicin has numerous health benefits and has been shown to be effective against many diseases. Use as a therapeutic agent is limited, however, due to a propensity for allicin to degrade under relatively mild conditions. In considering this, we developed a library of allicin derivatives and assessed their anticancer activity against the breast adenocarcinoma cell line MCF-7 and the MDR sub-line MCF-7/Dx. In an effort to stabilize the functionality of the thiosulfinate group, we substituted the labile allylthio moiety with various substituents. Spectral analysis of aryl thiosulfinates stored in chloroform at room temperature for several weeks showed minimal decomposition, while allicin showed distinct increases in impurities (data not shown). Additionally, compound 13b was found to be stable after storage at room temperature for over a year, whereas allicin is known to be incredibly unstable when stored in its pure state under these conditions⁶³. These results reflect previous findings that electron-rich aryl-substituted thiosulfinates are much more stable than allicin under similar conditions^40,57. These results were also reflected in our 72-h antiproliferative assay (Figure 2A). The majority of thiosulfinates tested showed improved activity compared to cells incubated with allicin, which we propose is in part due to the increased stability of these compounds. To maintain activity for the duration of the assay, allicin was replenished with fresh stock and media every 24 h. Conversely, when incubated with the same allicin stock for 72 h, cell viability remained largely unaffected in both cell lines (Supplementary material). This suggests that the antiproliferative activity observed in this assay is due to allicin and not its degradation by-products.

Increasing the chain length of the symmetric alkyl thiosulfinates showed a corresponding reduction in the proportion of viable cells in both cell lines. However, when one alkyl chain was replaced with a benzyl functional group, the compounds with shorter and less bulky alkyl groups showed more favorable results, with compounds containing the 4-methoxybenzyl moiety showing the greatest activity (Figure 2). We propose that this improved activity of the benzyl group over the straight chain substituents is due in part to the benzyl group acting as a bioisostere for the allyl group.

Several thiosulfinate derivatives exhibited cytotoxic activity towards Dx cells (Table 1). However, of the compounds tested, only 13b displayed an IC₅₀ <50 μmol/L in MCF-7 cells. These results highlight the disparity in the anticancer activity (IC₅₀) of the thiosulfinates against the drug-sensitive MCF-7 cells vs MDR Dx cells. Since Dx cells express P-gp, it was expected that they would be less sensitive to treatment, as Dx cells exhibit up to a 200-fold resistance to conventional chemotherapeutics in vitro⁶⁴.

A possible explanation for this selectivity could be linked to caspase-3 expression. MCF-7 cells do not express caspase-3, a key player in the apoptosis cascade⁶⁵. Conversely, Dx cells do express caspase-3 and, in direct comparison with MCF-7 cells, have been shown to be significantly more susceptible to certain apoptotic stimuli^64,66. Whether this accounts for the disparity in the anticancer activity shown in Table 1 is currently unknown. As allicin has been shown to initiate apoptosis through both caspase-dependent and caspase-independent mechanisms in various cell lines^27,67, further work is required to determine whether Dx cell death is executed via this pathway.

Allicin has previously been shown to be cancer specific, inducing significantly lower levels of cytotoxicity against normal cells^18,25,35. Additionally, short chain thiosulfinates, containing both allyl and saturated chains, show good activity in vivo, significantly increasing the survival times of sarcoma-bearing mice³⁶. Our most potent molecule, 13b, did not show specificity toward cancer cells, with effects observed in non-malignant human brain epithelial cells (HBEC) (Supplementary material). Interestingly, when treated with allicin, HBEC cells showed levels of cytotoxicity similar to those observed in Dx cells (data not shown), despite the numerous reports of allicin's cancer-specific activity. In light of this, we are currently undertaking further studies to improve the specificity of these molecules through modification of the pharmacophore. Additionally, it would be interesting to investigate the anticancer effect of these molecules on hematological cell lines, as it has previously been shown that allicin is more toxic to cells in suspension compared to cells in monolayers⁶⁸.

Allicin has previously been shown to rapidly induce morphological changes in colon carcinoma cells, gastric cancer cells and MCF-7 breast cancer cells, as well as reduce cell proliferation and induce growth arrest^26,69,70. Similarly, the thiosulfinates described in this study altered Dx cellular morphology. Cell shrinkage, detachment, and fragmentation were observed 24 h after treatment (Figure 3). This change in morphology was evident as early as 6 h after initial treatment (Supplementary material). To further investigate the mechanisms behind this reduced viability and altered cellular morphology, we tested for apoptosis, as allicin and saturated short-chain thiosulfinates have been shown to induce apoptosis in a number of cell lines^26,27,28. In the present study, thiosulfinates were found to induce apoptosis in Dx cells to varying degrees (Supplementary material), with 13b showing the most activity, inducing apoptosis in a dose-dependent manner (Table 3).

The effect of allicin on viability and apoptosis has been linked to a reduction in intracellular GSH, which indicates an alteration in the cellular redox state^25,28,71. GSH reduction occurs in response to sustained production of ROS upon exposure to allicin, which then leads to a decrease in mitochondrial potential, a reduction in ATP generation and subsequent cell death^16,29,72. ROS are generated in cells under normal conditions; however, production is elevated when there is a disruption either in the electron transport chain or ATP generation. Tumor cells inherently show elevated metabolic activity and oxygen consumption and produce high levels of ROS under normal conditions, making them more susceptible to damage from excess ROS production^73,74. As mitochondria are major mediators of apoptosis and cellular bioenergetics, modulation of mitochondrial respiratory function in an effort to kill cancer cells is attracting interest as a therapeutic strategy^74,75,76. To investigate the possible mitochondrial effects of the synthesized thiosulfinates on Dx cells, we used the XF24 extracellular flux analyzer, allowing for the measurement of multiple parameters associated with mitochondrial function⁷⁷.

In this study, we found that allicin and compound 13b cause similar responses in Dx cells, with significant reductions in oxygen consumption both at basal levels and under stress, as well as a decrease in ATP turnover (Figure 6). This suggests that these compounds reduce the cell's ability to generate a membrane potential and its capacity to synthesize ATP⁷⁷. Spare respiratory capacity was also significantly impaired after treatment with both thiosulfinates, which indicates a loss of mitochondrial membrane potential as well as a significantly reduced capacity to respond to stressors, such as ROS^78,79. We hypothesize that these compounds damage the mitochondrial electron transport chain through an elevation in ROS, which initiates the apoptosis cascade leading to cell death. Our investigations are ongoing to determine the precise molecular mechanisms underlying this process.

These disruptions to mitochondrial respiration offer another possible explanation for the observed sensitivity of MDR Dx cells to thiosulfinate treatment relative to the drug-sensitive MCF-7 cells. Dx cells may be more susceptible to thiosulfinate treatment due to the effects of a phenomenon known as collateral sensitivity. Collateral sensitivity is the ability of a compound to preferentially kill MDR cells over drug-sensitive cells and is generally associated with the over-expression of P-gp⁸⁰. Various mechanisms have been proposed for the effects of CS, such as the down-regulation of P-gp expression, disruption of oxygen consumption, elevated ROS, and depletion of cellular ATP^81,82,83. CS is represented as a ratio, which is determined by dividing the IC₅₀ of the MDR sub-line by the IC₅₀ of the drug-sensitive parental line to produce a resistance ratio ≤0.5⁸⁴. For 13b, the IC₅₀ was 46.50 μmol/L for MCF-7 cells and 18.54 mmol/L for Dx cells, providing a resistance ratio of 0.40. In conjunction with the ability of 13b to reduce ATP-linked respiration and to disrupt OCR under stress and at basal levels, this result suggests that 13b is a potential CS agent.

Garlic-derived compounds, including allicin, are known to inhibit cancer cell growth by disrupting the cell cycle, causing cells to accumulate predominately in the G₂/M phase in direct response to the overproduction of ROS and disruption to mitochondrial respiration^85,86. Treatment with compound 13b resulted in a dose-dependent accumulation of cells in G₂/M (Table 2). Additionally, during initial screenings at 50 μmol/L, the cell cycle was found to be arrested at the G₂/M phase for most of the thiosulfinates tested (Supplementary material). For allicin, we did not find any change in cell cycle distribution at this concentration; however, higher concentrations increased accumulation in G₀/G₁ and G₂/M, with these results reflecting a previous study on MCF-7 cells²⁵.

In addition to antiproliferative effects, thiosulfinates have also shown promise as antimetastatic agents. It has recently been reported that blocking mitochondrial ATP production inhibits breast cancer cell migration and invasion. Likewise, allicin has been shown to induce cell cycle arrest through microtubule disruption and to suppress invasion and metastasis of breast cancer in vitro^30,70,87. As we have shown that allicin and 13b inhibit ATP-linked mitochondrial respiration, we hypothesize that these compounds are likely to influence cellular migration and therefore warrant further investigation.

In conclusion, allicin has previously been shown to exhibit many beneficial biological effects, including strong anticancer activity. In light of the instability of allicin in vitro and in vivo, we sought to improve its activity by developing a library of thiosulfinate derivatives with increased half-life and improved efficacy. We found that these thiosulfinate derivatives have improved antiproliferative activity in drug-sensitive and MDR breast cancer cell lines. Several compounds showed specificity for the MDR cells, and Dx cells showed collateral sensitivity to treatment with 13b. We report that these derivatives alter cellular morphology, induce apoptosis and cell cycle arrest, and inhibit mitochondrial bioenergetics, presenting a promising lead for future studies in the treatment of MDR cancer.

Author contribution

Mary BEBAWY and Alison UNG designed the research; Ariane ROSEBLADE performed the research; Ariane ROSEBLADE, Mary BEBAWY and Alison UNG analyzed the data and wrote the paper.