Allicin (98% purity; Shaanxi Ciyuan Biotech Co., Ltd., Xi'an, China) was dissolved in deionized water. The same batch of allicin was used for all of the experiments. Minimum essential medium (MEM)-α and fetal bovine serum (FBS) were purchased from Gibco (Thermo Fisher Scientific, Inc., Waltham, MA, USA). H₂O₂ was purchased from Sigma-Aldrich (St. Louis, MO, USA). Murine osteoblastic MC3T3-E1 cells were purchased from American Type Culture Collection (Manassas, VA, USA) and cultured at 37°C under 5% CO₂ in MEM-α supplemented with 10% FBS (or 2% FBS for cell maintenance) and 1% penicillin/streptomycin (CSPC Pharmaceutical Group Limited, Shijiazhuang, China). The cells were treated with 200 µM H₂O₂ for 6 h in order to induce OS, and allicin treatment was performed with a concentration of 10, 30 or 100 µg/ml for 24 h.

The viabilities of the MC3T3-E1 cells were evaluated by MTT assay, subsequent to the various treatments. Briefly, MC3T3-E1 cells were seeded in 96-well plates, cultured to ~85% confluence and were then treated with 200 µM H₂O₂ for 6 h with or without allicin treatment at 10, 30 or 100 µg/ml for 24 h at 37°C. Next, MTT solution was added to each well and the plates were incubated for an additional 2 h at 37°C. Subsequent to the removal of the solutions from the well, dimethyl sulfoxide (Thermo Fisher Scientific, Inc.) was added to dissolve formazan products. Finally, the absorbance of each well was recorded on a microplate spectrophotometer (PowerWave XS2; BioTek Instruments, Inc., Winooski, VT, USA) at 570 nm.

Apoptosis of MC3T3-E1 cells subsequent to treatment was examined with an Annexin V-FITC Apoptosis Detection kit (Sigma-Aldrich). In brief, 5×10⁵ cells were collected subsequent to treatments and were stained with annexin V-FITC and propidium iodide at 25°C for 5–15 min (placed in the dark). Apoptotic cells were then detected using a flow cytometer (BRYTE HS; Bio-Rad Laboratories, Inc., Hercules, CA, USA) and were quantified by calculating the percentage of apoptotic cells relative to the total number of cells using FlowJo 7.6 (Tree Star, Inc., Ashland, OR, USA).

Following treatment with 200 µM H₂O₂ for 6 h with or without 10, 30 or 100 µg/ml allicin for 24 h, MC3T3-E1 cells were collected and lyzed with an ice-cold cell lysis reagent (FNN0091; Thermo Fisher Scientific, Inc.) according to the manufacturer's instructions, and were then supplemented with protease inhibitors (04693116001; Roche Diagnostics, GmbH, Mannheim, Germany). The protein concentration of each sample was measured using the Pierce bicinchoninic acid assay reagent (Thermo Fisher Scientific, Inc., Rockford, IL, USA). Samples were separated using 8–12% SDS-PAGE gel, with a sample volume of 25 µg each, on a protein electrophoresis apparatus (Mini-PROTEAN Tetra Cell; Bio-Rad Laboratories, Inc.) at 80 V for 25 min and 120 V for 75 min. Next, the separated proteins were transferred to a polyvinylidene difluoride membranes (EMD Millipore, Bedford, MA, USA) and were blocked with 2% bovine serum albumin (BioVision, Mountain View, CA, USA) at 4°C overnight. Membranes were inoculated with rabbit polyclonal antibody against caspase 3 (cat. no. 9662; 1:400; Cell Signaling Technology, Inc., Danvers, MA, USA) and poly adenosine diphosphate-ribose polymerase (PARP; cat. no. 5625, 1:800; Cell Signaling Technology, Inc.), cytochrome c (ab90529, 1:600; Abcam, Cambridge, UK), Akt and caspase 9 (C7729, 1:300; Sigma-Aldrich), cAMP response element-binding protein (CREB) with (sc-101663; 1:600) or without S133 phosphorylation (sc-25785; 1:800), extracellular signal-regulated kinases (ERKs) with (sc-101761; 1:600) or without T204 phosphorylation (sc-292838; 1:800; Santa Cruz Biotechnology, Inc.) or to tubulin (13918-T16-100; 1:1,000; Sino Biological, Inc., Beijing, China) at 4°C overnight. A goat anti-rabbit IgG conjugated to horseradish peroxidase secondary antibody was used (A27036; 1:600; Thermo Fisher Scientific, Inc.) and a Novex ECL Chemiluminescent Substrate Reagent Kit (WP20005; Thermo Fisher Scientific, Inc.) was used to detect the specific binding of each target protein to its antibody. The results were expressed as relative protein levels to tubulin.

Complex IV activity was measured with a Complex IV Rodent Enzyme Activity Microplate Assay kit (Abcam) according to the manufacturer's instructions. Briefly, MC3T3-E1 cell samples underwent detergent extraction with 1:10 volume detergent followed by 12,000 × g centrifugation for 20 min to extract the supernatants. Supernatant samples were quantified to an optimal concentration of 5 mg/ml, and were serially diluted using 1X phosphate-buffered saline. Samples were then loaded onto plates (100 µl per well), with a positive control sample and buffer control included as null reference. This step was performed to enable the binding of complex IV to the bound monoclonal antibody, included in the kit, which immobilized the enzyme in the wells. Following a 3-h incubation at 25°C, each well was rinsed twice with solution 1 (100 µl per well), then assay solution was added (100 µl per well). Finally, the reaction was measured for optical density at 550 nm absorbance at 1–5 min intervals for 2 h at 30°C using a spectrophotometer (PowerWave XS2). The results were expressed as a percentage of the complex IV activity of the control MC3T3-E1 cells.

A tetramethylrhodamine, ethyl ester (TMRE) Mitochondrial Membrane Potential Assay kit (Abcam) was used according to the manufacturer's instructions. TMRE is a cell-permeant, positively-charged, red-orange dye that readily accumulates in active mitochondria due to their relative negative charge. Depolarized or inactive mitochondria have a decreased membrane potential and fail to sequester TMRE. A total of 1×10⁵ MC3T3-E1 cells per treatment group were incubated with the mitochondrial membrane potential (MMP)-sensitive fluorescent TMRE for 20 min at 37°C, 5% CO₂ (1,000 nM FCCP was added to the positive control cells 10 min prior to TMRE addition). Following incubation, the cells were trypsinized (0.25%; Ameresco, Inc., Framingham, MA, USA), centrifuged and cell pellets were resuspended in 0.4 ml of Dulbecco's phosphate-buffered saline (Sigma-Aldrich) with 0.2% bovine serum albumin (BioVision) and analyzed for TMRE fluorescence by flow cytometry using FlowJo 7.6, with an excitation/emission fluorescence for TMRE of 549/575 nm.

The intracellular level of ROS was quantified by the oxidation of the ROS-sensitive fluorophore dichloro-dihydro-fluorescein diacetate (Thermo Fisher Scientific, Inc., Waltham, MA, USA) according to the manufacturer's instructions. In brief, confluent MC3T3-E1 cells in 6-well plates (5×10⁵ cells/well) were loaded with a 5 µM probe (5-chloromethyl-2,7-dichlorodihydrofluorescein diacetate) in Hanks' balanced salt solution (HBSS; Thermo Fisher Scientific, Inc.) and were incubated at 37°C and 5% CO₂ for 30 min. The cells were rinsed with HBSS subsequent to the removal of the probe and the 2,7-dichlorofluorescein fluorescence was measured using a luminescence spectrometer (2390–0000; PerkinElmer, Inc., Waltham, MA, USA) with an excitation source at 488 nm and an emission at 530 nm. The mitochondrial superoxide levels were detected using a Molecular Probes MitoSOX Red Mitochondrial Superoxide Indicator (Thermo Fisher Scientific, Inc.), which is a fluorogenic dye (excitation/emission = 510/580 nm) for highly selective detection of superoxide in the mitochondria of cells. Subsequent to the different treatments, cells were incubated with 5 µM MitoSOX Red at 37°C for 20 min and then the MitoSOX Red fluorescence was detected following removal and washing of the cells.

PI3K activity in lysates of MC3T3-E1 cells following the various treatments was evaluated using a PI3-Kinase Activity ELISA kit (Echelon Biosciences, Inc., Salt Lake City, UT, USA). The assay was a competitive ELISA in which the signal is inversely proportional to the amount of phosphatidylinositol (3,4,5) trisphosphate (PIP₃) produced. Subsequent to the completion of the PI3K reactions, the reaction products were first mixed and incubated with a PIP₃ antibody and then added to a PIP₃-coated microplate for competitive binding. A peroxidase-linked secondary antibody and colorimetric detection were used to detect the PIP₃ detector binding to the plate. The colorimetric signal is inversely proportional to the amount of PIP₃ produced by PI3K.

All the experiments were conducted in triplicate and the results are expressed as the mean ± standard error of at least three independent experiments. The statistical significance was determined by analysis of variance and subsequently applying Student's t-test. P<0.05 was considered to indicate a statistically significant difference.

In order to investigate the anti-OS effects of allicin in MC3T3-E1 cells, they were exposed to H₂O₂ to mimic ROS elevation, and the effects of the exposure on cell viability and apoptosis were evaluated. The MC3T3-E1 cells were treated with 200 µM H₂O₂ for 6 h and the cell viability was examined by MTT assay. The results indicated that the cell viability was significantly reduced by H₂O₂ treatment for 6 h (P=0.00082; Fig. 1A). Next, the apoptosis levels were examined by flow cytometry of the MC3T3-E1 cells following 200 µM H₂O₂ treatment for 6 h, and consistent with the cell viability results, Fig. 1B indicated that there was a significantly higher percentage of apoptotic cells in the 200 µM H₂O₂-treated group (P=0.00034). The MC3T3-E1 cells were treated with 200 µM H₂O₂ and with 0, 10, 30 or 100 µg/ml allicin and then the cell viability and apoptosis levels were measured. Compared with the H₂O₂ treatment alone, the combined treatment of H₂O₂ with 10 (P=0.033), 30 or 100 µg/ml (P=0.00090) allicin significantly ameliorated the cell viability. This effect was increased in the 30 µg/ml group compared with the 10 µg/ml group (P=0.042; Fig. 1A). Furthermore, a combined treatment of H₂O₂ with allicin demonstrated that H₂O₂-induced apoptosis was significantly inhibited by the treatment of 10 (P=0.040), 30 P=0.0076) or 100 µg/ml (P=0.0065) allicin and the inhibition was significantly increased in the 30 µg/ml group compared with the 10 µg/ml group (P=0.022) (Fig. 1B).

In order to confirm the effect of allicin on H₂O₂-induced apoptosis in MC3T3-E1 cells, cytochrome c release and caspase 3 activation (catalyzed from pro-caspase 3 into 17- and 12-kDa subunits) were measured in the H₂O₂-treated cells by western blot assay. It was demonstrated that the 200 µM H₂O₂ treatment for 6 h promoted increased levels of cytochrome c release P=0.00041), activated caspase-3 (P=0.00037) and lyzed PARP (P=0.00083) by activated caspase-3 (Fig. 1C–F). However, allicin treatment with 10, 30 and 100 µg/ml partially reversed the effect of H₂O₂ treatment and led to reduced cytochrome c release (P=0.041, P=0.0085 and P=0.00047, respectively; Fig. 1D), caspase 3 activation (P=0.040, P=0.028 and P=0.0066, respectively; Fig. 1E) and PARP lysis (P>0.05, P=0.0080 and P=0.0043, respectively; Fig. 1F). These effects were also observed to increase in a dose-dependent manner.

The influence of allicin on the H₂O₂-induced osteoblast mitochondrial dysfunction was also evaluated. ROS generation in MC3T3-E1 cells with or without the treatment of 200 µM H₂O₂ was assessed. A significantly elevated ROS release from H₂O₂-treated cells was indicated (P=0.00093; Fig. 2A). However, this increase was blocked by 10 (P=0.038), 30 (P=0.0086) or 100 µg/ml (P=0.0075) allicin, and the 30 µg/ml group was significantly reduced compared with the 10 µg/ml group (P=0.048). Secondly, mitochondrial superoxide was measured using MitoSOX Red, a live-cell-permeable and mitochondrial-localizing superoxide indicator that further assessed the effect of allicin on mitochondrial superoxide production. It was demonstrated that H₂O₂ treatment enhanced mitochondrial MitoSOX Red fluorescence (P=0.00071; Fig. 2B) whereas treatment with allicin with a concentration of 10 (P=0.0094), 30 (P=0.00087) and 100 µg/ml (P=0.00092) substantially inhibited the stimulatory effect of H₂O₂ to the mitochondrial superoxide production (Fig. 2B).

In addition, the activity of the mitochondrial respiratory chain complex IV and mitochondrial membrane potential was examined. The complex IV activity decreased dramatically in the H₂O₂-treated cells (P=0.0032; Fig. 2C). However, allicin significantly ameliorated this effect (10 µg/ml, P=0.044; 30 µg/ml, P=0.0063; or 100 µg/ml, P=0.0053; Fig. 2C) and the activity was significantly higher in the 30 µg/ml group compared with the 10 µg/ml group (P=0.028; Fig. 2C). MMP, as a marker of mitochondrial oxidative phosphorylation activity, is a well-established indicator of mitochondrial function. As TMRE is a cell-permeable, positively-charged, red-orange dye that readily accumulates in active mitochondria due to their relative negative charge, the mitochondrial function was assessed by examining the TMRE accumulation in mitochondria. TMRE accumulation was significantly reduced by the H₂O₂ treatment (P=0.0037), and it was reversed by 10, 30 or 100 µg/ml allicin (Fig. 2D). In conclusion, the results of the present study confirmed that allicin may protect from mitochondrial damage in MC3T3-E1 cells following OS.

The PI3K/AKT pathway is important in cell survival, particularly against OS (29,30). To further deduce the mechanism of the cytoprotective effect of allicin, the present study investigated whether PI3K/AKT activation was implicated in the cytoprotective effect of allicin in osteoblastic MC3T3-E1 cells. It was demonstrated that PI3K activity was significantly decreased in the MC3T3-E1 cells subsequent to H₂O₂ treatment (P=0.0036; Fig. 3A). The treatment with 30 (P=0.024) or 100 µg/ml (P=0.0035) allicin markedly ameliorated this effect (Fig. 3A). Furthermore, western blotting indicated that phosphorylated (p)-AKT was also significantly downregulated (P=0.00037), whereas the cleaved caspase 9 (C Casp 9) was markedly promoted (P=0.00040) by H₂O₂ treatment (Fig. 3B–D). Notably, 30 and 100 µg/ml allicin significantly increased the level of p-AKT (P=0.0057 and P=0.00048, respectively; Fig. 3C) and reduced levels of C casp 9 (P=0.0063 and 0.00056, respectively; Fig. 3C) compared with the H₂O₂-only group. In summary, allicin was suggested to ameliorate the H₂O₂-induced PI3K/AKT pro-survival signal inhibition in osteoblasts by H₂O₂.

CREB/ERK signaling was also demonstrated to be involved in OS (31). In the present study, the activation of CREB/ERK signaling was also demonstrated, and p-CREB and p-ERK levels were altered in the MC3T3-E1 cells following the treatment with H₂O₂ and allicin. The H₂O₂ treatment significantly reduced the level of p-CREB (P=0.00071) and p-ERK (P=0.00033) in the MC3T3-E1 cells (Fig. 4). Allicin treatment with 10, 30 or 100 µg/ml markedly increased p-CREB (P=0.028, P=0.015 and P=0.0022, respectively; Fig. 4B) and p-ERK (P=0.036, P=0.0035 and P=0.00045, respectively; Fig. 4C) levels in MC3T3-E1 cells compared with the H₂O₂-only group. In summary, it was suggested that allicin ameliorates the H₂O₂-induced CREB/ERK pro-survival signaling inhibition in osteoblasts via H₂O₂.