Pterostilbene exerts anticancer activity on non-small-cell lung cancer via activating endoplasmic reticulum stress

Pterostilbene (PT), the natural dimethylated analog of resveratrol (RSV), is a potent anticarcinogen for non-small-cell lung cancer (NSCLC), but its anti-NSCLC mechanisms remain unclear. In this study, we show that PT treatment time- and dose-dependently enhanced the endoplasmic reticulum stress (ERS) signaling (i.e., p-PERK, IRE1, ATF4, CHOP), thus decreasing the cell viability and inducing apoptosis in human PC9 and A549 NSCLC cell lines. Moreover, the decreased migratory and adhesive abilities, downregulation of intracellular glutathione (GSH) level, enhanced reactive oxygen species (ROS) generation, Caspase 3 activity and mitochondrial membrane depolarization were observed in NSCLC cells treated with PT. These effects were reversed by CHOP siRNA which inhibited the ERS signaling pathway, but were promoted by thapsigargin (a classical ERS inducer) in vitro. Besides, in vivo studies also verify that PT exerted anticancer activity by mobilizing ERS signaling and apoptosis-related proteins, and these effects were enhanced by thapsigargin. Therefore, ERS activation may represent a new mechanism of anti-NSCLC action by PT, and a novel therapeutic intervention for lung cancer.


Inhibition of cell viability and induction of apoptosis by PT treatment on PC9 and A549 cells.
To investigate whether PT has the anticancer effect on NSCLC, the CCK-8 assay was employed to evaluate its cytotoxic role on PC9 and A549 cells (Fig. 1). Treatment on cells for 24 h or 48 h with 20 μM, 40 μM, and 60 μM PT inhibited cell viability in a dose-and time-dependent manner, and the IC50 values of PT at 24 h and 48 h were approximately 50.09 μM and 27.35 μM in PC9 cells; and 52.01 μM and 24.12 μM in A549 cells, respectively. Microscopic images (PC9 in Fig. 1A and A549 in Fig. 1B) indicated that PT treatment resulted in significant cell shrinkage and decreased cellular attachment rate compared with their control groups. After treatment with 20, 40, and 60 μM PT for 24 h, the apoptotic index dose-dependently increased to 16.75 ± 3.98%, 35.96 ± 5.81%, and 53.18 ± 6.53% in PC9 cells; and 20.16 ± 4.05%, 39.84 ± 6.21%, and 50.07 ± 7.19% in A549 cells, respectively. This effect was both estimated by TUNEL assay and Annexin V-FITC/PI assay ( Fig. 2 and Supplementary Figure 2A). Moreover, to investigate whether PT has inhibitory role on HBE cells (normal human bronchial epithelial cells), we found that PT at low concentrations (20 and 40 μM) had weaker inhibitory effect on HBE cells than on  (20,40, and 60 μM) and assessed at time points of 24 and 48 h. The cell viability was analyzed by CCK-8 and expressed as OD values (% Control). The morphology of both cells were observed under an inverted phase-contrast microscope after the cells were treated for 24 h, and images were obtained. Significant cell shrinkage and a decreased cellular attachment rate were observed in the PT treatment groups with a dose dependent manner. All of the results were expressed as the mean ± SD; n = 6. a P < 0.05 vs. the control group, b P < 0.05 vs. the 20 μM PT-treated group, c P < 0.05 vs. the 40 μM PT-treated group.
NSCLC cells, but PT at higher concentration (60 μM) had similar inhibitory effect on both HBE and NSCLC cells (Supplementary Figure 1A).

Inhibition of cell migration and adhesion by PT treatment on PC9 and A549 cells.
The migratory and adhesive abilities of those two NSCLC cell lines via using a wound-healing assay, and an adhesion assay. Based on our preliminary experiments, the relatively low concentrations of PT (2, 4, and 6 μM) did not affect the cell viability in human NSCLC cells (Supplementary Figure 3A). So, after incubation with PT (2, 4, and 6 μM) for 24 h, the scratch wound distance is 135.85 ± 7.81%, 164.43 ± 8.52%, and 183.34 ± 10.63% in PC9 cells; and 128.33 ± 8.61%, 159.76 ± 9.30%, and 180.86 ± 11.09% in A549 cells, respectively (P < 0.05, compared Figure 2. Effect of PT treatment on the apoptosis of PC9 and A549 cells (24 h). After treatment on both cell lines (A for PC9, B for A549), the apoptotic index of PT-treated groups exerted a dose-dependent increase in induction of apoptosis. The upper panel showed the cell nucleus (blue) and the lower panel displayed the apoptotic cells (green). All of the results were expressed as the mean ± SD; n = 6. a P < 0.05 vs. the control group, b P < 0.05 vs. the 20 μM PT-treated group, c P < 0.05 vs. the 40 μM PT-treated group.

Activation of ERS signaling, autophagy and apoptosis-related proteins induced by PT on PC9 and A549 cells.
To investigate the anticancer activity of PT on ERS signaling, ERS-related molecules were detected via western blot in PC9, A549, and HBE cells. Western blot revealed that relative low concentrations of PT treatment (2, 4, and 6 μM) had no effect on the expression of ERS-related proteins in NSCLC cells (P > 0.05, compared with the control group, Supplementary Figure 3A). PT treatment (20,40, and 60 μM) significantly enhanced the expression of p-PERK, IRE1, ATF4, and CHOP in both NSCLC cell lines in a dose-dependent manner (P < 0.05, compared with the control group, Fig. 5), but the pro-ERS effect on normal bronchial epithelial cells was weaker than on NSCLS cells. PT 20 μM had no effect on ERS signaling upregulation on HBE cells (P > 0.05, compared with the control group, Supplementary Figure 1B), and PT 40 μM slightly promoted ERS signaling expression (P < 0.05). Only PT at higher concentration (60 μM) obviously enhanced ERS signaling expression on HBE cells (Supplementary Figure 1B). Moreover, to investigate whether PT could induce autophagy on NSCLC cells, both western bot and immunocytochemistry analyses revealed PT induced a concentration-dependent accumulation of autophagy-symbolic LC3BII protein in NSCLC cells (Supplementary Figure 4). Furthermore, we also determined whether PT affects the apoptotic proteins in NSCLC cells via measuring the protein expression of Bcl2, Bax, Caspase 3, and p53. PT treatment decreased the expression of Bcl2 and increased the expression of Bax, Caspase 3 and p53 (P < 0.05, compared with the control group, Fig. 5).
Effect of PT treatment combined with CHOP siRNA on cell viability, Caspase 3 activity, ROS generation, CHOP and apoptosis-related protein levels on PC9 and A549 cells. A specific CHOP siRNA was used to explore the effect of ERS signaling downregulation on the anti-NSCLC activity of PT treatment in vitro. PC9 and A549 cells were first transfected with CHOP siRNA and then treated with PT (40 μM) for an additional 24 h. Transfection with CHOP siRNA significantly decreased the expression of the levels of CHOP in PC9 and A549 cells (P < 0.05, compared to transfection with the control siRNA, Fig. 6D). Moreover, the combination of CHOP siRNA and PT increased cell viability (Fig. 6A), decreased Caspase 3 activity (Fig. 6B) and generation of ROS (Fig. 6C) (P < 0.05, compared with the combination of control siRNA and PT); however, CHOP siRNA alone did not affect cell viability, Caspase 3 activity, and ROS generation compared with the control siRNA (P > 0.05). Moreover, Bcl2 was further upregulated by co-treatment with PT and CHOP siRNA, whereas Bax and Caspase 3 levels were significantly downregulated by co-treatment with CHOP siRNA and PT (P < 0.05, compared with the control siRNA and PT co-treatment, Fig. 6D). Though CHOP siRNA co-treatment decreased PT-induced promotion of Caspase 3 activity and ROS generation, Caspase 3 activity and ROS generation still remained at a high level, indicating CHOP is not the only mediator for apoptosis induction. Interestingly, co-treatment of CHOP siRNA had no effect on PT-induced upregulation of p53 expression (P > 0.05), suggesting p53 upregulation was independent of CHOP activation by PT treatment.

Effect of PT treatment on Ca 2+ dyshomeostasis in PC9 cells or combined with CHOP siRNA.
The average cytosolic Ca 2+ concentration in a single cell was determined using the fluorescent probe Fluo-3AM by laser confocal scanning microscopy. PC9 cells were scanned for 3 min to obtain a basal fluorescence intensity level of intracellular Ca 2+ (F0). Then treated with PT 40 μM and another 12 min under the treatments to obtain the real-time fluorescenceintensity (F). PT 40 μM treatment significantly increased the cytosolic Ca 2+ level in PC9 cells compared with the control group (P < 0.05, Supplementary Figure 2C), and nearly at 10 min up to the Ca 2+ peak. Moreover, CHOP siRNA pre-treatment obviously reversed the PT 40 μM induced increase of cytosolic Ca 2+ level (P < 0.05, compared with PT 40 μM group, Supplementary Figure 2C), with only minor increase of Ca 2+ level compared with control group (P < 0.05). These results revealed that PT treatment could induce Ca 2+ dyshomeostasis thus promoting ERS.
Effect of PT treatment combined with THA on cell viability, Caspase 3 activity, ROS generation, CHOP and apoptosis-related protein levels on PC9 and A549 cells. THA, one of the classical ERS inducers, causes calcium leakage from ER into the cytosol, thereby exerting anticancer activity by decreasing cancer cell viability through apoptosis. The anti-NSCLC role of THA has also been proven in our previous studies 7 . We combined PT (40 μM) with THA (0.5 μM) and tested their cytotoxicity to PC9 and A549 cells for 24 h. Based on our preliminary experiments, a relatively low concentration and duration of THA treatment (0.5 μM for 24 h) were chosen so that the potential ERS activating effect could emerge more prominently.

Figure 4. Effect of PT treatment on the abilities of migration and adhesion in PC9 and A549 cells (24 h). (A)
Representative wound healing images of each cells were shown, and the migratory ability is expressed as the mean distance between the two sides of the scratch. The mean distance in the control group was set as 100%. (B) Representative adhesion images of each cell lines were shown, and the adhesion ability of cells is expressed as an adhesion ratio. The number of adherent cells in the control group was set as 100%. All of the results were expressed as the mean ± SD; n = 6. a P < 0.05 vs. the control group, b P < 0.05 vs. the 2 μM PT-treated group, c P < 0.05 vs. the 4 μM PT-treated group.
ScienTiFic REPoRTS | 7: 8091 | DOI:10.1038/s41598-017-08547-0 Compared with the control group, THA treatment alone slightly decreased the cell viability to 82.13 ± 2.69% in PC9 cells and 83.36 ± 2.09% in A549 cells; PT treatment alone decreased the cell viability to 61.81 ± 6.54% in PC9 cells and 58.76 ± 5.96% in A549 cells. However, when PT was combined with THA, the cell viability was significantly decreased to 36.82 ± 3.41% in PC9 cells and 38.23 ± 3.11% in A549 cells. Compared with either PT treatment alone or THA treatment alone, the co-treatment of PT and THA significantly increased Caspase 3 activity ( Fig. 7B) and ROS generation (Fig. 7C) (P < 0.05). As shown in Fig. 7D, THA treatment increased CHOP expression in both cell lines (P < 0.05 compared with the control), and PT co-treatment with THA in PC9 and A549 cells significantly upregulated CHOP level compared with either treatment alone (P < 0.05). Moreover, co-treatment of PT and THA also significantly promoted Bax, Caspase 3 and p53 upregulation and downregulated Bcl2 level (P < 0.05 compared with either PT or THA treatment alone), thus the promotion of ERS by THA co-treatment enhanced p53 expression via PT treatment.

THA co-treatment enhanced inhibition of tumor xenograft growth by PT treatment via promoting ERS signaling and activating apoptosis-related proteins in vivo. To verify our in vitro
outcomes and whether PT treatment can inhibit tumor growth in vivo studies, we established PC9 xenografts in athymic nude mice and measured their tumor volumes. We found that the nude mice in all of the treatment groups developed subcutaneous tumors, and PT 50 mg/kg treatment or THA 1 mg/kg treatment significantly inhibited tumor growth (P < 0.05, compared with the control group, Fig. 8A), which were consistent with previously reported studies 8,27 . Further, treatment of THA 1 mg/kg sensitized the anticancer effect of PT 50 mg/kg on tumor growth (P < 0.05, compared with the PT 50 mg/kg or THA 1 mg/kg group), which indicated that PT inhibited the NSCLC cells via activating ERS in vivo. Moreover, western blot or immunohistofluorescence analyses showed that PT 50 mg/kg or THA 1 mg/kg treatment promoted the upregulation of ERS signaling molecules (p-PERK and CHOP), and regulation of apoptosis-related proteins (upregulation of Bax, Caspase 3 and p53 levels, and downregulation of Bcl2 protein) (P < 0.05, compared with the control group, Fig. 8B and C), and these effects were significantly enhanced by PT 50 mg/kg and THA 1 mg/kg co-treatment (P < 0.05, compared with the PT 50 mg/kg or THA 1 mg/kg group).

Discussion
PT, a naturally derived polyphenol compound with multiple properties (e.g., anti-inflammation, anti-obesity, and anti-oxidation), is an analogue of the well-studied resveratrol (RSV), but is reported significantly more bioavailable when ingested [28][29][30] . Moreover, PT is also a potent anticancer agent against various cancers by different mechanisms 9, 10, 12-18 . Meanwhile, several studies suggested that PT exerts anti-NSCLC activities via the induction of apoptosis-related cell death 8,21,27 . However, the actions of PT on NSCLC and the detailed mechanisms responsible for these activities are still not fully elucidated. Consistent with previous studies, PT treatment significantly repressed the cell viability and induced cell death in both PC9 and A549 cell lines with a dose-and time-dependent manner. Furthermore, we found that PT treatment suppressed the migratory and adhesive abilities of NSCLC cells, both of which are major actions associated with tumor metastasis. Additionally, PT treatment obviously inhibited tumor growth in the PC9 xenografts.
During solid cancers progression, cancer cells require a high rate of protein folding and assembly in the ER, but the nutrient requirements eventually exceed the capacity of the existing vascular bed 25 . Therefore, cancer cells usually encounters stressful microenvironments, such as hypoxia and nutrient deprivation, which eventually impairs the generation of ATP and compromises ER protein folding, thus leading to ERS 22,25 . Meanwhile, UPR is subsequently activated, which aims at restoring ER homeostasis, leading to adaptions, and sustaining cell survival 24,25 . However, aside from its pro-survival effects, prolonged UPR activation owing to sever or unresolved ERS eventually triggers cell death 22,23 . Therefore, targeting the ERS signaling activation is a promising anticancer strategy, proved by bortezomib, the first proteasome inhibitor in clinical application for multiple myeloma and Burkiit lymphoma, of which the mechanisms are at least in part via ERS activation 25,26 . Furthermore, evidence for promoting ERS signaling as the anti-NSCLC therapy, especially activating the death effector CHOP via the PERK-ATF4 cascade, is compelling 7, 31, 32 . In our study, we explored the actions of ERS in the anti-NSCLC activities of PT. ER is the main storage site of Ca 2+ , and any disruption in its accumulation can promote ER stress 23,33 . Our study found PT treatment could immediately disturb the Ca 2+ homeostasis in ER and cytoplasm, and significantly upregulate cytosolic Ca 2+ levels in PC9 cells. Knockdown of CHOP by siRNA in vitro thus effectively inhibited the PT-induced Ca 2+ dyshomeostasis in ER and upregulation of cytosolic Ca 2+ levels in PC9 cells. Ca 2+ released from ER has irreversible effects on cell functioning and ultimately leads to apoptotic or necrotic cell death 34   depletion through irreversible inhibition of the ER Ca 2+ -ATPase pump which was responsible for the influx of Ca 2+ from cytosol 33,35 . Therefore, these outcomes suggest that the anti-NSCLC actions of PT is regulated, at least in part, by the stimulation of ERS signaling pathway.
CHOP is a transcription factor whose induction represses cell proliferation and induces apoptosis via suppressing anti-apoptotic proteins (e.g., Bcl2 and Bcl-xL, Mcl-1) and upregulating pro-apoptotic proteins (e.g., Bax, Bak, Bim), thus connecting ERS and mitochondrial-initiated apoptosis 2, 7, 10, 25 . Examples include inactivation of Bcl2, translocation of Bax and Bak from ER to mitochondrial, enhanced ER Ca 2+ flux, and caspases activation in response to overwhelmed ERS 2, 7, 10, 22, 36 . In present study, we found PT treatment dose-dependently decreased Bcl2 expression, increased Bax expression, and upregulated Caspsase 3 protein level and activity both in vivo and in vitro. Furthermore, our in vivo studies confirmed that the use of CHOP siRNA reversed the anti-cancer effects mentioned above, whereas treatment with THA thus enhanced those effects. Moreover, pre-incubation with z-DEVD-fmk, a Caspase 3 specific inhibitor, completely inhibited PT-induced increase of Caspase 3 activity and abolished PT-induced cell death in NSCLC cells. These results indicate that PT induces NSCLC cells apoptosis via regulation of Bcl2 family proteins, which are connected with the activation of ERS signaling and Caspase 3-dependent.
The master tumor suppressor p53 is a key protein in preventing cell transformation and tumor progression 14 . Numerous studies have revealed the connection between ERS and p53 signaling pathway 14,37 . ERS triggers p53-dependent suppression of p21 thus induction of apoptosis 14 . Moreover, instead of functioning as a transcription factor, p53 also locates at the ER and mitochondria-associated membranes, and mediates Ca 2+ signal-dependent apoptosis during ERS 14 . In present study, we found PT treatment or THA treatment could increase p53 levels in vitro and in vivo, and co-treatment of PT and THA significantly enhanced this effect, suggesting ERS activation involved in PT-induced increase of p53 protein levels. However, above mentioned results were not interfered by CHOP siRNA, suggesting PT-induced p53 upregulation is CHOP independent, and CHOP is not the only mediator for ERS-related apoptotic induction by PT treatment.
Mitochondrial oxidative stress and the onset of ERS often occur together 38 , and their molecular events linking may attribute to ROS generation 38,39 . ROS has been identified as a potential anticancer target and its accumulation may represent a cause or a result of prolonged ERS 2, 40, 41 . PERK has been recently reported uniquely enriched at the mitochondria-associated ER membranes 39 . Despite the canonical signaling via the PERK-ATF4-CHOP branch of ERS induces cell apoptosis, Verfaillie and colleagues have demonstrated that PERK at the ER-mitochondria contact sites mediates the ROS-based mitochondrial apoptosis during persistent ERS 39 . In addition, several studies also suggested that activation of CHOP also contributed to ROS generation 13,23,42 . In our studies, the suppression of NSCLC cell viability by PT treatment was associated with the rapid increase in ROS generation, and decrease of GSH content and MMP. Moreover, CHOP siRNA treatment attenuated the production of ROS and upregulated the level of GSH, whereas these effects of PT were enhanced by THA co-treatment. GSH is a major non-protein cellular antioxidant that can eliminate intracellular ROS, and the degree of ROS level and perturbations in the GSH redox balance are critical in determining whether cells undergo survival or elimination processes 10,43,44 . A recent study by Benlloch et al. reported that PT treatment suppressed the anti-oxidative properties of melanoma and pancreatic cancer cells via downregulation of GSH and glucocorticoid levels, and inhibition of glucocorticoid receptor and nuclear factor (erythroid-derived 2)-like 2 (Nrf2)-dependent antioxidant defense system 16 . Based on above results, we propose that the induction of ERS-ROS-dependent cell death signaling may be an important anti-NSCLC mechanism of PT treatment.
Autophagy activation has been proposed as an alternative mechanism of cell death induced by polyphenols 45 . Consistent with other studies 18, 21, 45 , we found PT treatment promoted the induction of autophagy in NSCLC cells. Numerous studies also revealed intensive connections between the activation of ERS and autophagy involved in the induction of apoptotic cell death on cancer cells [46][47][48] . However, whether the promotion of autophagy by PT treatment antagonizing 18,21 or participating 45 the anti-NSCLC effect of PT remains controversial, thus warranting further investigation.
Collectively, our experiments provide in vivo and in vitro mechanistic evidence that PT is a potent anti-NSCLC agent via activating persistent ERS signaling. Suppression of ERS desensitizes the anticancer effects of PT on NSCLC, whereas enhanced by ERS stimulators. In addition, the PT treatment induced ERS-pro-apoptotic mechanisms on NSCLC may be related to both PERK-ATF4-CHOP cascade, activation of p53, and ERS-ROS-based cell death signaling. Above all, activation of ERS signaling is a promising anticancer therapy for NSCLC (Fig. 9). A small clinic trail with an enrollment of 80 confirmed PT is generally safe for use in humans at doses up to 250 mg per day, and has no direct side-effect on hepatic or renal function 49 . In our studies, compared with NSCLC cells, HBE cells were well tolerated with PT at relatively low concentrations. However, up to the concentration of 60 μM, PT had similar pro-apoptotic and pro-ERS effects on both NSCLC cells and HBE cells, suggesting concentration of PT should be maintained at a relatively effective level for killing cancer cells with no side-effects in clinic. Moreover, PT has also been proven to prevent lung cancer carcinogenesis at a lower concentration (5 μM) 49 , and inhibit the lung cancer stem cells generation and its associated malignancy 20 . Thus, PT displays Figure 9. Schematic diagram about the anti-NSCLC activity of PT via the stimulation of ERS. Our in vivo and in vitro studies showed that PT treatment could stimulate ERS and upregulate a series of ERS-related molecules (i.e., p-PERK, IRE1, ATF4, CHOP) and promoting the efflux of Ca 2+ from ER into cytoplasm, thus promoting apoptosis (mechanisms via upregulation of Bax, Caspase 3 and p53 expression, and decrease of Bcl2 level), enhancing ROS generation and MMP depolarization, reducing intracellular GSH levels, and inhibiting the ability of migration and adhesion in NSCLC cells. Moreover, THA treatment can enhance the anticancer activity of PT on NSCLC. PT, pterostilbene; THA, thapsigargin; ERS, endoplasmic reticulum stress; NSCLC, non-small-cell lung cancer; p-PERK, phosphorylated-PKR-like ER kinase; IRE1, inositol-requiring enzyme 1; ATF4, activating transcription factor 4; CHOP, C/EBP-homologous protein; Bcl2, B-cell lymphoma-2; Bax, Bcl2-associated X protein; ROS, reactive oxygen species; GSH, glutathione; MMP, mitochondrial membrane potential. potent anti-NSCLC properties, and its detailed anticancer mechanisms and application in clinic still warrant further investigation.

Analyses of cell apoptosis by TUNEL and Annexin V-FITC/PI assay. The levels of cellular apoptosis
in PC9, A549 and HBE cells were analyzed via a TUNEL assay as previously described 2 , using an in situ cell death detection kit according to the manufacturer's directions. The TUNEL assay was performed to stain the nuclei of the apoptotic cells (green), and DAPI was used to stain the nuclei of all cells (blue). Images were obtained by using an Olympus FV1000 confocal microscope (Olympus, Japan) laser confocal microscope, and the TUNEL-positive cells were counted in 5 randomly selected fields under high-power magnification. The apoptotic index was expressed as the ratio between the number of TUNEL-positive cells and the total number of DAPI-positive cells counted × 100%.
For AnnexinV/PI double staining, cells were cultured in confocal dishes for 24 h. After treatments, cells were incubated in binding buffer containing Annexin V-FITC and propidium iodide (PI) at room temperature for 10 min. Annexin V-FITC was performed to stain the membrane of the early-stage apoptotic cells (green), and PI was used to stain the nuclei of late apoptotic or necrotic cells (red). Images were obtained by using an Olympus FV1000 confocal microscope (Olympus, Japan) laser confocal microscope, and the only Annexin V-positive cells were counted in 5 randomly selected fields under high-power magnification. The apoptotic index was expressed as the ratio between the number of Annexin V-positive cells and the total number of cells counted × 100%.
Analysis of cell mitochondrial membrane potential. The changes in mitochondrial membrane potential (MMP) was reflected by the cells staining with the fluorescent dye JC-1 as previously described 8 . Briefly, lung cancer cells (2 × 10 5 cells) were cultured in confocal dishes. After different treatments, cells were incubated with JC-1 working solution in the dark at 37 °C for 20 min. Cells exhibiting red fluorescence are in normal state with high MMP, however, when the cells are in an apoptotic or necrotic state, the JC-1 is present as a monomer and MMP is decreased, thus the dye emits green fluorescence. Images were obtained by using an Olympus FV1000 (Olympus, Japan) laser confocal microscope, and the results were expressed as the proportion of cells with a low MMP.

Analyses of Caspase 3 activity, ROS generation, and GSH level.
After treatment, Caspase 3 activity was measured via using a colorimetric assay kit according to the manufacturer's instructions. The cells or tissue lysates (20 μl) were added to a buffer containing a p-nitroaniline (pNA)-conjugated substrate for Caspase 3 (Ac-DEVD-pNA) to yield a 100 μl reaction volume. The reactions were performed for 2 h at 37 °C. The released pNA concentrations were calculated based on the absorbance values at 405 nm and the calibration curve of the defined pNA solutions. The Caspase 3 activity in the control group was set as 100%.
To analyze intracellualr ROS generation, samples were trypsinized and subsequently incubated in DCFH-DA (5 μM) in PBS for 2 h at 37 °C. After incubation, the DCFH fluorescence of samples were measured using an FLX 800 fluorescence microplate reader at an excitation of 488 nm and an emission of 522 nm (Biotech Instruments, Inc., USA). A cell-free condition was used as the background, and the fluorescence intensity in the control group was defined as 100%. The generation of intracellular reduced GSH was simultaneously measured using commercial kits according to the provided instructions and as previously described 2 . Fluorescent products were measured using an FLX 800 fluorescence microplate reader at an excitation of 335 nm and an emission of 541 nm. The GSH level in the control group was set as 100%.
Analyses of cell migration and adhesion. We performed adhesion and migration analyses at lower concentrations of PT in order to do not affect cell activity. As our preliminary experiments showed that PT (lower than 6 μM) treatment for 24 h did not affect the proliferation of both PC9 and A549 cells. A cell culture wound-healing assay was performed to analyze cell migration, and 1 × 10 5 cells were seeded in the 6 well plate in basal medium containing 10% FBS at 37 °C with a humidified 5% CO 2 in incubator. When the cells were grown to confluence, a linear wound was made in the confluent monolayer using the 200 uL micropipette tip. Then cells were washed with PBS to remove the cellular debris. After treatment with PT (2 μM, 4 μM, and 6 μM) for 24 h, the movements of the wound edges were monitored under the microscope. The results are showed as the distance between the cells on either side of the scratch.
After treatment, 100 μL of medium containing 1 × 10 4 cells were incubated to each well in a 96-well plate for 30 min at 37 °C. The medium in each well was then discarded for adhesion analysis. The adherent cells were stained with MTT. The stained cells were observed using an inverted phase-contrast microscope. Next, images were obtained via the 600D camera (Canon, Japan), and five fields were randomly selected for quantification. Finally, 100 μL of DMSO was added to each well, and the plates were incubated for 30 min at 37 °C with shaking. The OD value of each well at 570 nm was measured using a SpectraMax 190 spectrophotometer (Molecular Devices, Sunnyvale, CA, USA), and the OD value of the control group was normalized to 100%.
Analyses of cytosolic calcium. Free cytosolic Ca 2+ was measured by fluorescence imaging using the Ca 2+ indicator dye Fluo-3AM. PC9 cells were cultured in confocal dishes for 24 h. Cells were washed 3 times with Hank's balanced salt solution (HBSS) without Ca 2+ and Mg 2+ , then cells were incubated with 5 μM Fluo-3AM and 0.2% pluronic F-127 in HBSS for 1 h at 37 °C. Afterwards, Fluo-3AM containing solution was removed and cells were incubated for an additional 30 min in HBSS at 37 °C. Before exposed to PT 40 μM (HBSS was applied for control group), the cells were scanned for 3 min to obtain a basal fluorescence intensity level of intracellular Ca 2+ (F0) and another 12 min under the treatments to obtain the real-time fluorescenceintensity (F), the ratio was F/ F0. Fluorescence was excited at 506 nm, emission was detected at 526 nm, and the image was recorded every 30 s by Olympus FV1000 (Olympus, Japan) laser confocal microscope. The results were expressed as the ratio of F/F0. Western blot. Western blot procedures were presented as previously described 2 . Briefly, the cells or tumor samples were lysed in sample buffer [150 mM Tris (pH 6.8), 8 M urea, 50 mM DTT, 2% sodium dodecyl sulfate, 15% sucrose, 2 mM EDTA, 0.01% bromophenol blue, and 1% protease and phosphatase inhibitor cocktail], sonicated, boiled, fractionated by SDS-PAGE, transferred to PVDF membranes, and subjected to western blot analysis with various antibodies. The fluorescent signal was detected using a BioRad imaging system (BioRad, Hercules, CA, USA), and the signal was quantified using Image Lab Software (BioRad, Hercules, CA, USA).
Immunocytofluorescence and immunohistofluorescence staining. NSCLC Cells were cultured in confocal dishes for further treatments. The formalin-fixed paraffin-embedded tumor tissue sections were performed by deparaffinizing and rehydrating the tissues followed by antigen retrieval. Cells or tissue sections were successively blocked with 0.1% Triton X-100 for 15 min, blocked with goat serum for 2 h at room temperature, then incubated with primary antibody overnight at 4 °C. They were washed with PBS and incubated with Cy3, goat anti-rabbit IgG for 2 h at 37 °C. After washed with PBS, they stained with DAPI for 15 min at 37 °C. Images were acquired using Olympus FV1000 confocal microscope (Olympus, Japan).  . 16001, 16002). Nude mice were s.c. injected with 7 × 10 6 cells transduced with PC9 cells into both flanks of each animal, respectively. The body weight and tumor size of each mouse were measured every 3 days. The estimated tumor volume was calculated using the formula: volume = 0.5 × length × width 2 . Based on the data from a preliminary study, we initiated treatment when V reached approximately 100 mm 3 . Then the mice were randomly allocated to 4 groups (n = 6/group): the control group (0.05% DMSO), PT 50 mg/kg, THA 1 mg/kg, and PT 50 mg/kg and THA 1 mg/kg co-treatment groups. PT and THA were diluted with DMSO in saline and was administered intraperitoneally every day. On day 28, the tumors were excised from euthanized mice for additional analysis.

Anticancer activity of PT
Statistical analyses. All of the data are presented as the means ± standard deviation (m ± SD).
Between-group comparisons, one-way ANOVA followed by Bonferroni post hoc test was performed with SPSS 13.0 (SPSS Inc., Chicago, USA) software. A P value of less than 0.05 was considered to be significant.