Honokiol was purchased from the National Institute for the Control of Pharmaceutical and Biological Products (Beijing, China) with >99% purity. Honokiol power was dissolved in DMSO at 20 mg/ml (75 mM). z-VAD-fmk was from Beyotime Institute of Biotechnology (Shanghai, China). PrimeScript^® RT reagent Kit (Perfect Real-Time) and SYBR^® Premix Ex Taq™ II (Tli RNase H Plus) were purchased from Takara Biotechnology Co. Ltd (Dalian, China). DNA Ladder Detection Kit was from KeyGEN Biotech (Nanjing, China). M-PER^® Mammalian Protein Extraction Reagent was from Thermo Scientific (Pierce Biotechnology, Rockford, IL, USA). RIP1 antibody and Annexin V-FITC/PI kit were from R&D Systems (Minneapolis, MN, USA). RIP3 and Bcl-2 antibodies were purchased from Epitomics (Burlingame, CA, USA), PTEN, Bcl-xl and GAPDH antibodies were from Cell Signaling Technology (Danvers, MA, USA).

Human breast carcinoma cell lines were obtained from Cancer Institution of Zhejiang University (Zhejiang, China). MCF-7 was maintained in RPMI-1640 medium. MDA-MB-231 was cultured in L-15. Bcap-37, T47D and SKBR-3 were maintained in DMEM. All media were supplemented with 10% FBS. The multidrug resistant human breast carcinoma cell line MCF-7/ADR and Bcap-37/ADR was developed as reported previously (12) and maintained in RPMI-1640 (supplemented with 10% FBS) with 1 and 0.5 μg/ml doxorubicin.

Breast cancer cells of cultured cell lines were seeded in 96-well culture plates at density of 6,000–8,000 cells/well and cultured for 24 h pretreatment. Then, cells were incubated with drug-free medium, honokiol of different concentrations for different time durations. Cell toxicity assays were preformed using MTT method as described previously (12).

MCF-7 cells were cultured in 6-well culture plates at 8–10×10⁵/well in 2 ml medium for 24 h. Cells were treated with 0.1% DMSO (vehicle control) and honokiol at different concentrations (liquid volume: 1.5 ml/well) for different time durations. After treatment, cells were stained with Annexin V-FITC/PI and assayed by flow cytometry as previously described (12).

Cells were seeded and treated as in flow cytometric analysis for cell apoptosis and necrosis. Cells were fixed in 4% phosphate-buffered paraformaldehyde for 10 min and were washed twice with PBS subsequently. Then cells were incubated with 10 μg/ml Hoechst 33342 for 10 min in dark at room temperature. Nuclei stains of Hoechst 33342 were examined under a confocal microscope (Leica TCS SP5). Hoechst reagent was taken up by the cell nuclei, and apoptotic cells exhibited a bright blue fluorescence.

Cells were seeded and treated as in flow cytometric analysis for cell apoptosis and necrosis. After treatment, cells were collected and washed twice with chilled PBS. DNA gel electrophoresis was done after DNA extraction according to the manufacturer’s instructions of DNA Ladder Detection Kit. HL60 cells treated with 20 μg/ml VP-16 for 6 h were used as positive control.

Cells were seeded and treated as in flow cytometric analysis for cell apoptosis and necrosis. Total RNA was extracted using TRIzol reagent (Invitrogen, Carlsbad, CA, USA). Concentration and quality of RNA were measured by Nanodrop 1000 spectrophotometer (Thermo Scientific). Total RNA were reverse transcribed using PrimeScript^® RT reagent Kit (Takara Biotechnology Co. Ltd). Subsequently, real-time PCR reactions were performed using SYBR^® Premix Ex Taq™ II (Tli RNaseH Plus) Kit (Takara Biotechnology Co. Ltd) on Stepone Plus System (Applied Biosystems, Foster City, CA, USA). Detailed primer sequences are summarized in Table I. PCR reaction conditions were: 95°C for 30 sec, 40 cycles with 95°C for 5 sec and 60°C for 30 sec. β-actin was applied as internal reference for normalization. All the above operation procedures were performed according to the manufacturer’s protocols. ΔΔCt method was introduced to determine the relative mRNA expression level.

Cells were treated and harvested as in flow cytometric analysis for cell apoptosis and necrosis, protein was extracted using M-PER^® Mammalian Protein Extraction Reagent. Protein concentrations were determined with BCA method. A total of 20 μg of each sample was loaded on the gel and procedure of electrophoresis and western blot analysis were performed as previously described (11). After PVDF membranes were ready for antibody staining, they were incubated with appropriate primary antibodies at 4°C overnight and HRP-conjugated second antibodies at room temperature for 1 h in sequence. ECL detection kit was introduced for signal development. Images were acquired using chemiluminesence. Jurkat cells were used as positive control in analysis RIP3 protein expression.

All the experiments were done in triplicate, and similar results were obtained in three different experimental setups. The data represent means ± standard deviation. One-way analysis of variance (ANOVA) was used for evaluations of time- and dose-response curves. Student’s t-test was used in single group comparisons. P<0.05 was considered statistically significant.

Our data revealed that HNK repressed cell viability of all tested breast cancer cell lines (MCF-7, MDA-MB-231, Bcap-37, T47D, SKBR-3, MCF-7/ADR and Bcap-37/ADR) after 24-h treatment with IC₅₀ ranged from 18.389 to 21.940 μg/ml (Table II). HNK showed parallel cytotoxical effect on multidrug resistant cells and their corresponding sensitive cells (Table II). Further, we demonstrated that HNK had time- and dose-dependent cytotoxic effects on MCF-7 cells (Fig. 1).

A range of chemicals trigger different cell death modes in a dose-dependent manner. Frequently, lower-dose exposure induces autophagy or apoptosis, whereas higher-dose induces necrosis (13). Our group previously reported that HNK at a concentration of 40 μg/ml triggered programmed necrosis in MCF-7 cells (11). Accordingly, we hypothesized HNK may trigger apoptosis to programmed necrosis transition in a time- and dose-dependent manner.

Cell death modes triggered by HNK at a fixed concentration (40 μg/ml) of different exposure durations were characterized using Hoechst 33342 staining, DNA laddering and flow cytometric analysis of Annexin V-FITC/PI staining. MCF-7 cells were incubated with 40 μg/ml HNK for 30 min, 1 and 6 h, respectively, and then labeled with Annexin V-FITC (AV) and PI (Fig. 2). Even for the shortest incubating time (30 min), cell viability was significantly reduced by 40 μg/ml HNK (P<0.001) (Fig. 2A and C). Cell death was accelerated with the extension of incubating time (Fig. 2A and C). Necrotic cell death (PI⁺) triggered by 40 μg/ml HNK increased (P<0.001) with the extension of incubating time, whereas early apoptotic cell death (AV staining positive and PI negative) manifested contrary tendency (Fig. 2C and D). Internucleosomal DNA fragmentation is one of the hallmarks of apoptosis at later stage (or degradation phase), which are simply detectable as a ladder pattern in electrophoresis of isolated DNA. Nuclear chromatin condensation and apoptotic bodies are both morphological characteristic manifestations of apoptosis and are detectable using nuclei staining by Hoechst 33342. No obvious ladder pattern was observed in any of the treatment groups (40 μg/ml HNK incubation for 30 min, 1, 2, 4 and 6 h, respectively) (Fig. 3A). Confocal imaging of Hoechst 33342-stained nuclei revealed no remarkable chromatin condensation or apoptotic bodies in any of the treatment groups, but anomalous diffuse staining were detected in groups with longer treatment durations (4 and 6 h) (Fig. 3B). Both DNA ladder and Hoechst 33342 staining assays demonstrated later-stage apoptosis were not the predominant cell death mode of 40 μg/ml HNK triggered cell death. Taken together, 40 μg/ml HNK triggered a cell death mode transition from early-stage apoptosis to programmed necrosis in a time-dependent manner.

Death modes triggered by different concentrations (20, 25, 30 and 40 μg/ml) of HNK for 6 h were further analyzed. Cell fraction of early-stage apoptosis decreased, fraction of necrotic cell death conversely increased (P<0.001) with the extension of doses (Fig. 2B–D), which demonstrated similar tendency to that of time-dependent manner. However, no obvious DNA ladders or apoptotic bodies were detected using DNA laddering or Hoechst 33342-staining in any of the treatment groups (Fig. 3C and D). Such results also demonstrated HNK triggered cell death mode transition from early-stage apoptosis to programmed necrosis in a dose-dependent manner.

Collectively, these experimental results support the original hypothesis that HNK triggered cell death mode transition from apoptosis to programmed necrosis in a time- and dose-dependent manner.

Cyclophilin D (CypD) is a vital protein involved in mitochondrial permeability transition (mPT) and critical regulator in HNK-induced programmed necrotic cell death (11). Recent reports demonstrated cyclophilin D-dependent mPT regulates some necrotic but not apoptotic cell death (14), and comports as apoptotic repressor (15,16). In consequence, we explored whether HNK-induced cell death mode transition was specifically regulated by CypD. As we previously demonstrated, cellular CypD expression levels of mRNA and protein coincided with each other perfectly (11). We then measured the expression levels of CypD mRNA after HNK exposure in MCF-7 cells using RT-PCR. CypD mRNA levels in MCF-7 cells after HNK exposure was evaluated in all groups with different incubation durations (Fig. 4A and B). CypD mRNA levels increased with the extension of incubating time in the early death phase and declined in the late-phase (6 h). A possible explanation for the declined CypD mRNA levels in late-phase is that cells are under cellular collapse. Fig. 4B provides details of CypD mRNA levels between groups at different doses of incubation. CypD mRNA levels manifested a gradually increasing tendency with the extension of doses (Fig. 4B) (P<0.001). CypD mRNA level was suppressed in apoptosis-predominant group (HNK 20 μg/ml), was parallel in apoptosis and necrosis-balanced group (HNK 25 μg/ml), whereas elevated in necrosis-predominant group (HNK 30 and 40 μg/ml) (Fig. 4B). These results indicated CypD was a potential modulator of HNK-triggered cell death mode transition time- and dose-dependently.

Cyclosporin A (CsA) targets and binds to the mitochondria CypD, which subsequently inhibites modulated-function of CyPD in cell death (17). We found CsA dramatically inhibited HNK induced cell death using MTT assay (Fig. 4C) at doses of 15 and 25 μM. Further, we used AV/PI staining methods to determine whether HNK induced time-dependent necrotic cell death was inhibited by CsA. We increased total dose (through increasing volume of drug, 2 ml/well) in each well with a similar concentration of HNK (40 μg/ml) in order to amplify its necrotic death induction effects. Pretreatment with 15 μM CsA for 2 h significantly inhibited HNK-induced necrotic cell death (PI staining-positive cells) with HNK incubation of 2, 4 and 6 h, respectively (P<0.001) (Fig. 4D). The evidence supports CypD was a potential modulator of HNK-triggered cell death mode transition.

Programmed necrosis is defined as necrosis highly regulated by RIP1 and RIP3 (2). RIP3 is essential and crucial component in the initiation of programmed necrosis (18,19), but has no role in apoptosis (20,21). Enhanced expression of RIP3 is strongly associated with increased programmed necrosis (2) and enable the switch of TNF-induced cell death from apoptosis to necrosis (20). We further assessed RIP3 expression during HNK-triggered cell death mode transition using western blot analysis. Untreated MCF-7 did not express any detectable RIP3, which accords with a previous report (21). Expressions of RIP3 increased significantly after HNK treatment in a dose- and time-dependent manner (P<0.0001) (Fig. 5A, B, D and F). Furthermore, in programmed necrosis predominant groups, RIP3 expressions were approximate or more than 2 times compared with early-apoptosis predominant groups (HNK 25 μg/ml for 6 h or HNK 40 μg/ml for 1 h) (Fig. 5A, B, D and F).

RIP1 is a critical switch in regulation cell fate, including NF-κB activation associated cell survival and proliferation, apoptosis and programmed necrosis (22–24). Under caspase-inhibited condition, RIP1 and RIP3 form a so-called ‘necrosome’ complex to initiate programmed necrosis signaling pathway (20). We further assessed RIP1 expression during HNK treatment using western blot analysis. Results indicated MCF-7 cells expressed reduced-RIP1 in all treatment groups compared with baseline level of normal MCF-7 cells (Figs. 5A, 6B). The RIP1 expression inhibition after HNK treatments was dose- and time-dependent (Fig. 5E and G).

We also detected, despite pretreatment with 15 μM CsA for 2 h significantly inhibited HNK-induced necrotic cell death, it did not affect enhanced RIP3 expression triggered by HNK (Fig. 5C). Overall, these results specifically revealed RIP3 was a potential regulator in HNK-induced programmed necrosis and cell mode transition from apoptosis to programmed necrosis.

Bcl-2 and Bcl-xl are pro-survival proteins of the Bcl-2 family. They excert anti-apoptotic effect through sequestering and inhibiting pro-apoptotic Bcl-2 proteins, which may trigger mitochondrial outer membrane permeabilization (25–27). Overexpression of these anti-apoptotic Bcl-2 proteins protects cells from apoptosis induction from a broad range of apoptotic stimuli (26). We aimed to detect the relevance between HNK-triggered apoptosis to programmed necrotic cell death transition and expression of Bcl-2 and Bcl-xl. No obvious changes in expressions of Bcl-2 and Bcl-xl proteins were detected paralleling with increased doses or treatment durations using western blot analysis (Fig. 6A and B).

Loss of expression or inactivation of PTEN amplifies the PI3K-Akt-mTOR pathway, which promotes cell survival and proliferation (28–30). PTEN protein expression during cell death mode transition triggered by increased doses or duration of incubation was also determined. Western blot analysis results revealed that PTEN protein expressions increased paralleling with HNK incubated-duration (Fig. 6A) and apparently enhanced in necrosis-predominant group (HNK 40 μg/ml for 2, 4 and 6 h). Similarly, intensive expression of PTEN was also detected with increased doses of HNK, especially in necrosis-predominant group (HNK 30 and 40 μg/ml for 6 h) (Fig. 6B).