Synergistic effects of combined treatment with histone deacetylase inhibitor suberoylanilide hydroxamic acid and TRAIL on human breast cancer cells

Previous studies showed that either histone deacetylase (HDAC) inhibitors or tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) can induce apoptosis in tumor cells including breast cancer. However, the underling mechanisms of combining HDAC inhibitors with TRAIL in the treatment of breast cancer are poorly understood. In this study, we determined the ability of SAHA and TRAIL as single agents or in combination to inhibit the growth and survival of MCF-7 and MDA-MB-231 breast cancer cells. Our results demonstrate that the distinct effects of SAHA or TRAIL individually and in combination on the proliferation, cell viability, apoptosis, cell cycle distribution, and morphological changes of MDA-MB-231 and MCF-7 cells. We further determined the different effects of SAHA or TRAIL alone and combining SAHA with TRAIL on the expression of a number of apoptosis-related molecules, cell cycle, growth factors and their receptors in cancer cells. Our results demonstrated that the combinatorial treatment of SAHA and TRAIL may target multiple pathways and serve as an effective therapeutic strategy against breast cancer. An improved understanding of the molecular mechanisms may facilitate either SAHA or TRAIL targeted use and the selection of suitable combinations.

large number of ongoing clinical trials to evaluate its utility in treating various solid tumors. Studies have shown that SAHA can induce apoptosis and growth arrest in breast cancer cell lines including MCF-7, MDA-MB-231, MDA-MB-435, MDA-MB-468, and SKBr-3 [15][16][17][18][19] . On the other hand, due to rapid hepatic glucuronidation, SAHA has a short half-life of 2 hrs, making it difficult to provide the level of drug exposure necessary for durable therapeutic efficacy on solid tumors. Adverse side effects, which become more severe at escalated doses, and intrinsic and acquired resistance to vorinostat also present significant clinical challenges 20,21 .
Tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) has been recognized as having a key role in body's natural defense mechanism and in inducing apoptosis in a variety of tumor cells, but its clinical utility has been limitated [22][23][24][25] . TRAIL mediated apoptosis is initiated by the binding of two agonistic death receptors, DR4 (TRAIL-RI) and DR5 (TRAIL-RII) in a p53-independent manner [26][27][28] . Conversely, TRAIL activity can be specifically inhibited by two decoy receptors, DcR1 (TRAIL-R3, LIT or TRID) or DcR2 (TRAIL-R4 or TRUNDD) thereby blocking its signaling of cell death 29 . TRAIL can also bind to osteoprotegerin (OPG), a soluble receptor for TRAIL, to attenuate apoptosis 30,31 . TRAIL preferentially induces apoptosis in tumor cell lines that lack DcR1, DcR2, but not in normal cells which express DcR1, DcR2, suggesting that TRAIL could potentially represent a powerful cancer therapeutic 32,33 . In recent years, TRAIL-based combinatorial therapies are emerging paradigms for cancer treatment since synergistic activation of TRAIL-induced apoptosis by chemotherapeutic drugs can generally overcome tumor cell resistance, while monotherapies are often fail. Preclinical studies and clinical trials are introducing promising results, supporting the potential effects of these combined approaches 34,35 .
A number of preclinical studies combining HDAC inhibitors with TRAIL have shown synergistic effects in inhibition of proliferation and induction of apoptosis in tumor cells 36 . SAHA was reported to induce expression of TRAIL by directly activating its promoter and triggering TRAIL-mediated apoptosis in acute myeloid leukemia cells 37 . Antisense ablation of TRAIL in the sensitive HL60 cells significantly reduced SAHA-mediated apoptotic and cytotoxic effects, indicating that TRAIL signaling pathway was important for SAHA pharmacological action 38 . In breast cancer cells, several HDAC inhibitors have been shown to enhance TRAIL-mediated apoptosis 39,40 . For example, SAHA can sensitize TRAIL-resistant breast cancer cells 17,41 . However, the underlying mechanisms of combining HDAC inhibitors with TRAIL in the treatment of breast cancer are poorly understood. The purpose of this study was to determine the ability of combining SAHA with TRAIL to selectively target the breast cancer cells, assessed by their combined effects on the growth and survival of a representative panel of breast cancer cells. We also sought to characterize the effects of combining SAHA with TRAIL on the regulation of breast cancer genes, related signaling pathways, and morphology.

Real-time monitoring of the effects of SAHA on the cell index (CI) of MDA-MB-231 and MCF-7 cells. Preliminary experiments were performed to evaluate the effects exerted by SAHA in MDA-MB-231
and MCF-7 cells. The cells were seeded in 96-well plates, incubated with 0-50 μ M SAHA, then evaluated through real-time monitoring of cell growth by xCELLigence Real-time Cell Analyzer (RTCA) DP system. Since the changes of cell index (CI) directly correlate to cell proliferation, cytotoxicity, adhesion, and viability, the changes of CI can directly reflect the effects of drugs. In the control group without SAHA treatment, both MDA-MB-231 and MCF-7 cells showed robust growth; CI continued to increase over time and reached a plateau at 48 h (Fig. 1A,C). In contrast, in the group with SAHA treatment, both cell lines showed a time-and dose-dependent CI values. Initially treatment of SAHA induced a remarkable increase of CI values which peaked at around 10 hours. After the spike, SAHA decreased CI values of both cell lines over progression of time. Large doses of SAHA (20-50 μ M) showed significant effects even at 24 h compared to lower doses of SAHA (0.5 μ M-2.0 μ M) that showed minimal effects on the cell condition. When the dose of SAHA reached 5.0 μ M, both cell lines displayed statistically significant decrease in CI values compared to the control group. We also examined the effects of DMSO on the growth of MDA-MB-231 and MCF-7 cells. DMSO did not show significant cytotoxicity, and CI values continued to increase in both cell lines, although CI curves appear different from those of untreated cells (Fig. 1B,D).
Real-time monitoring of the effect of SAHA/TRAIL combination on breast cancer cell proliferation. In order to investigate the effect of TRAIL on breast cancer cell proliferation, MDA-MB-231 or MCF-7 cells was individually co-cultured with 0-100 ng/ml of varying concentrations of TRAIL in xCELLigence RTCA system. The results revealed that the effects of TRAIL on MDA-MB-231 and MCF-7 cell lines are entirely different. The inhibitory effect of TRAIL, from low to high-dose on MDA-MB-231, is obvious. In contrast, MCF-7 cell line is insensitive to TRAIL except at a high concentration, 100 ng/ml, of TRAIL when more pronounced growth inhibition effect was observed ( Fig. 2A,E).
We next determined the combined effects of 0-100 ng/ml TRAIL and 5 μ M SAHA on MDA-MB-231 and MCF-7 cells. Although the combined effects on MDA-MB-231 cells did not show dose-dependency, the CI dropped significantly. TRAIL at 50 ng/ml concentration, reaction time at 48 h, showed the strongest results in the combined trial (Fig. 2B). The combined effect of SAHA and TRAIL on MCF-7 cells showed more pronounced growth inhibition. Even at 24 h, the CI fell to the bottom of the baseline (Fig. 2F).
The treatments of TRAIL alone and combined TRAIL with SAHA on MDA-MB-231 cells showed similar inhibiting effect when comparing the cell response curves. However, the effects of combined TRAIL and SAHA treatment on MCF-7 cells resulted in a sharp decline of CI, indicating that the combination of two drugs may increase the sensitivity of MCF-7 cells.
We further confirmed the results using the CellTiter 96 Aqueous One Solution Cell Proliferation Assay (Fig. 2C,D,G,H). Finally, we determined the optimal treatment time and concentration of combined TRAIL and SAHA where maximum effect is observed. We found that, for MDA-MB-231 cells, the combination of 50 ng/ml TRAIL and 5 μ M SAHA for 48 h treatment yielded the most effective inhibition of cell growth. For MCF-7 cells, 100 ng/ml TRAIL and 5 μ M SAHA for 24 hrs of treatment was the ideal combination.
Scientific RepoRts | 6:28004 | DOI: 10.1038/srep28004 The effect of SAHA/TRAIL combination on cell viability and apoptosis of cancer cells. To further confirm the effects of SAHA and TRAIL on breast cancer cell proliferation, we determined the cell viability and apoptosis in MDA-MB-231 and MCF-7 cell lines. In comparison with DMSO control treatment, both cell viability and cell number decreased in MDA-MB-231 and MCF-7 cells after SAHA treatments. TRAIL alone was slightly less effective than SAHA' effect in inhibiting growth of both cell lines. However, combined TRAIL and SAHA treatment induced dramatic decreases in cell viability and cell number of both MDA-MB-231 and MCF-7 cells with over 50% reduction in the cell number ( Fig. 3B,C,E,F). As expected, in comparison with DMSO control treatments, the apoptotic cells increased in MDA-MB-231 and MCF-7 cells after the SAHA treatment. TRAIL alone only showed slight increase in apoptotic cells in the two cell lines. However, the apoptotic cells dramatically increased in MDA-MB-231 and MCF-7 cells after combining TRAIL and SAHA treatment; the apoptotic rate reached 13.15% in the early stage and 23.95% in the late stage in MDA-MB-231. The apoptotic rate reached 18.40% in the early stage and 14.75% in the late stage in MCF-7 cells (Fig. 3A,D). Treatment of TRIAL slightly increased cells in G0/G1phase, reduced cells in S phase from 39.3% to 25.16%, and increased G2/M cells from 4.78% to 11.12%. Combination treatment of TRAIL and SAHA reduced the cells in S phase from 39.3% to 9.29%, while 81.74% of total cells were confined in G0/G1 phase (Fig. 4A). For MCF-7 cells treated with SAHA, the G0/G1 phase cells increased from 36.49% to 71.93%, the cells in S phase decreased from 45.51% to 19.52% and the cells in G2/M decreased from 18.00% to 8.55%. For the treatment of TRAIL, there were slightly increased cells in G1/S phase, slightly decreased cells in S phase and no significant changes of cell numbers in G2/M. For combined SAHA and TRAIL treatment, the cell numbers showed slightly more changes in three different phases than SAHA treatment alone (Fig. 4B).
The effect of SAHA/TRAIL combination on morphological changes of cancer cells. We further examined the morphological changes of cancer cells after SAHA or TRAIL treatments using Nalge-Nunc "Lab-Tek ® " coverglass system and BioStation IM-Q. The cells grew well and occupied to 90% of the full plate after 48 hrs in the control treatment of DMSO. We observed distinct morphological changes of cancer cells after TRAIL or SAHA treatment alone. For MDA-MB-231, some cells showed apoptotic morphology characterized by cell rounding and shrinkage, appearance of membrane bubbles and several condensed or fragmented nuclei, which were further measured with Annexin V positivity even at 12 hrs after SAHA treatments. Apoptotic cells were seen at 24 hrs after TRAIL treatment. In contrast, apoptotic cells were more evident in reaching 50% of the total cells at 24 hrs after combined treatment of TRAIL and SAHA (Fig. 5A). For MCF-7, more apoptotic cells were seen at 36 hrs after SAHA treatments. Only a small number of apoptotic cells were seen at 24 hrs, but more were evident at 36 hrs after TRAIL treatment. However, apoptotic cells were observed at 12 hrs, and almost all cells were dead at 48 hrs after combined TRAIL and SAHA treatments (Fig. 5B).

The effects of SAHA and TRAIL on the expression of apoptosis-related molecules in cancer cells.
In order to clarify the mechanisms of apoptosis induction by SAHA and TRAIL in breast cancer cells, we determined a number of apoptosis-related protein expression using antibody array. SAHA alone inhibited Bcl-2, Bcl-x and Phospho P53 (s46) expression while inducing DR5 and CDKN1A expression in both cell lines. SAHA alone also significantly induced Caspase-3 expression in MDA-MB-231 but not MCF-7 cells. TRAIL alone significantly induced Bax, Caspase-3 and DR5 expression, and it inhibited Bcl-2, Bcl-x, CDKN1A expression in MDA-MB-231 cells. In MCF-7 cells, TRAIL alone induced DR5 and CDKN1A expression. As expected, combined TRAIL and SAHA treatments significantly induced Bax, DR5 and CDKN1A, and they also significantly inhibited Bcl-2, Bcl-x and Phospho P53 (s46) expression in both of the cell lines. Notably, in comparison with control treatment, SAHA alone or SAHA plus TRAIL induced CDKN1A expression accompanied by decreased Phospho P53 (s46) expression in both MCF-7 and MDA-MB-231 cells. However, TRAIL alone did not show this correlation in neither cell line (Fig. 6A,C). In addition, we observed that combined TRAIL and SAHA treatments apparently increased Bax, Caspase-3 and CDKN1A in MDA-MB-231 cells and only Bax in MCF-7 compared to TRAIL or SAHA alone. We then determined the mRNA expression of apoptosis-related molecules, Bax and Bcl-2, TRAIL related receptors, DR4, DR5 using real time PCR. We found that SAHA dramatically induced DR5, but not DR4 mRNA expression in both MDA-MB-231 and MCF-7 cell lines. SAHA inhibited Bcl-2 mRNA expression in MDA-MB-231 cell line. TRAIL or SAHA alone, as well as combined TRAIL and SAHA treatments significantly induced Bax mRNA expression in both cell lines, and they also inhibited Bcl-2 mRNA expression in MDA-MB-231 cells (Fig. 6B,D).
Furthermore, we sought to investigate the potential involvement of signaling mechanisms in proliferation and inhibition, as well as apoptosis induction by SAHA and TRAIL for breast cancer cells. We used quantitative PCR arrays to determine the expression levels of 168 genes, which included the regulation of apoptosis, cell cycle, growth factors and their receptors involved in signaling downstream of the growth factors, the genes targeted by growth factor-mediated stimulation, members of the MAP kinase pathway, the AKT/PI-3K pathway, the STATs pathway, and other downstream signaling cascades. The results showed 5-fold or greater changes of mRNAs as a result of treatment with SAHA and TRAIL, listed in Tables 1 and 2. We found pronounced and significant changes of WT1 and PDGFRA levels in MDA-MB-231 cells (Fig. 7A,B), FOSB, TP53, EGF, EGR1, DUSP6 level in MCF-7 cells (Fig. 8A,B), and RHOA level in both cell lines (Figs 7B and 8B). Importantly, the PDGFRA and DUSP6 levels were increased by over 50-fold in MDA-MB-231 and MCF-7 cell line, respectively. RHOA level decreased by more than 70,000-fold in both cell lines after treatment with SAHA and TRAIL. There were also significantly   Tables 1  and 2, CDKN1A was not included in Table 2, which was upregulated 4.84 and 4.97 times upon treatment with SAHA alone or the combinatorial treatment of SAHA and TRAIL in MCF-7 cells.
In addition, we found that SAHA treatment caused significant up-regulation or down-regulation of genes closely related to cell cycle regulation in MDA-MB-231 cell, including CCNA2, CCNB2, CDK6, CDKL1, RAD21, RB1, WEE1 and WT1. In MCF-7 cells, SAHA had more positive effects on the induction of apoptosis, with major changes in the expression of apoptotic genes including JUN, FOSB, NFKB1, EGF, EGR1, LTA, and DUSP6. Furthermore, several genes associated with cell cycle regulation were found to be suppressed by TRAIL in MDA-MB-231 cell, whereas no differences in mRNA expression was seen between vehicle and TRAIL treatment in MCF-7 cells.
Finally, we assessed the relationships of 26 genes through net-walking using GNCPro 42 (Fig. 9). CDKN1A is correlated with the most number of genes in this panel of 26 genes. Other interactive genes include JUN, EGR1, NFKB1, STAT5A, CDK6, TP53, RB1.

Discussion
In this report, we first evaluated the effects exerted by SAHA in a triple negative breast cancer cell model, MDA-MB-231 and an ER+ MCF-7 cell lines. We used RTCA, which is an impedance-based technology that can be used for label-free and real-time monitoring of cell properties, such as cell adherence, proliferation, migration and cytotoxicity. RTCA is also a powerful and reliable tool that can be used in drug discovery for toxicity and pharmacology studies [43][44][45] . We found that SAHA treatment induced a time-and dose-dependent decrease in CI values in both cell lines. These results indicate that SAHA can be used as an agent in combinatorial treatment of both triple-negative and ER positive breast cancer cancers. Early study indicated that HDAC inhibitors such as trichostatin A, valproic acid, and SAHA inhibited proliferation of MCF-7 cells. When used in combination with tamoxifen, SAHA inhibited proliferation more potently than with either agent alone 46 . More recently, results from two published articles revealed that SAHA, together with Olaparib, a poly (ADP-ribose) polymerase (PARP) inhibitor 47 or the naturally occurring sesquiterpene lactone Parthenolide (PN) 48 , were effective in inhibiting triple-negative breast cancer cell proliferation.
Unlike SAHA, TRAIL sensitivity varied with the phenotype of the breast cancer cell lines. Triple-negative breast cancer cells were very sensitive to TRAIL-induced apoptosis; in contrast, ER positive cells were resistant to TRAIL-induced apoptosis across a wide range of doses 49 . Previous studies also reveal that MDA-MB-231 cells exhibit a reduced sensitivity to TRAIL, but treatment with HDAC inhibitors restore cell sensitivity to TRAIL 40,41,50 . Our results showed that TRAIL or SAHA alone had only modest effect on reducing cell viability and cell number, but combined TRAIL and SAHA treatment significantly decreased cell viability and cell number in MDA-MB-231 and MCF-7 cells. In addition, TRAIL plus SAHA treatment significantly increased apoptotic  Our results suggest that SAHA or SAHA plus TRAIL treatment plays a major role in the cell cycle G0/G1 phase, in the suppression of S phase, and consequently in the inhibition of proliferation and in inducing cell apoptosis in breast cancer cells. Treatment with SAHA caused a significant increase of cells in the G0/G1 phase and a decrease of cell numbers in the S phase in both breast cancer cell lines, whereas a much small increase of cells in G0/G1 phase was observed after treatment with TRAIL. The combination treatment of SAHA and TRAIL further increased the cells confined in the G0/G1 phase and marked decrease of cells in S phase in both cell lines. Interestingly, we also observed distinct morphological changes of cancer cells after treatment with TRAIL or SAHA alone. In MDA-MB-231, apoptotic morphology of cells was seen at 12 hrs after SAHA treatment, but not until 24 hrs after TRAIL treatment. Combination treatment of TRAIL and SAHA increased apoptotic cells to 50% of total at 24 hrs. In MCF-7, while SAHA alone was also seen as more effective than TRAIL alone in inducing apoptosis, combination treatment induced apoptosis at an earlier time (12 hrs). Additionally, combination treatment of breast cancer cells with TRAIL and SAHA caused detachment of cells followed by anoikis, a form of apoptosis which occurs after cell detachment 51 . Anoikis is induced upon cell detachment from extracellular matrix and anoikis-resistance contributes to metastasis allowing cancer cells to survive in the blood stream and invade distant sites 52,53 .
It has been reported that simultaneous administration of SAHA and TRAIL significantly upregulated the TRAIL death receptors DR4/DR5 and downregulated the anti-apoptotic members of the BCL-2 family in several tumor cell lines [54][55][56][57] . To better understand the molecular events that lead to the synergistic interaction of SAHA and TRAIL, we determined apoptosis-related molecule, BCL-2-associated X protein (BAX), Bcl-2, TRAIL related DR4, DR5 protein and mRNA expression using antibody array and real time PCR. BAX, the first BCL-2 interaction partner, regulates the critical balance between cellular life and death 58,59 . In contrast to BCL-2, BAX promotes, rather than blocks cell death after a stress stimulus 59 . We observed that combined TRAIL and SAHA treatments apparently increased Bax protein, Caspase-3 and CDKN1A in MDA-MB-231 cells and only Bax protein in MCF-7 than TRAIL or SAHA alone. These protein changes may explain why combined TRAIL and SAHA treatments are more effective in MDA-MB-231 than that in MCF-7 cells. We showed that SAHA inhibited Bcl-2 mRNA expression in MDA-MB-231 cell line. TRAIL or SAHA alone, as well as combined TRAIL and SAHA We further determined the potential involvement of signaling mechanisms underlying the effects of SAHA and TRAIL for breast cancer cells using quantitative PCR arrays. The 168 candidate genes that were examined included apoptosis, cell cycle, growth factors and their receptors, as well as downstream signaling cascades. Wilms' tumor 1 (WT1), reported to correlate with higher histological grades, ER-negative and basal-like and ERBB2 molecular subtypes in breast cancer 68 , was significantly downregulated in MDA-MB-231 cells upon treatment with SAHA or SAHA + TRAIL. The platelet-derived growth factor receptor α (PDGFRA), known to be associated with tumor aggressiveness in lymph node-negative breast cancer patients who had not received adjuvant systemic therapy 69 , after treatment of SAHA or SAHA + TRAIL was significantly upregulated MDA-MB-231 cells. In MCF-7 cells, genes that had over 5-fold changes upon treatment of SAHA and TRAIL alone or in combination include FosB, a member of the AP-1 family of transcription factors as important regulators of cell proliferation and differentiation and highly expressed in normal mammary epithelia but down-regulated in poorly differentiated breast cancer 70 , TP53, EGF, EGR1, and DUSP6. RhoA, one of the more extensively studied members of the Rho family of small GTPase, participate in tumor cell migration and invasion 71 , which was reported to be associated with progression in invasive breast duct carcinoma 72 was most prominently downregulated in both cell lines. In addition, cell cycle related genes, such as CCNA2, CCNB2, CDK6, CDKL1, RAD21, RB1, WEE1 were also found to differentially expressed as a result of treatment in MDA-MB-231 cells, while expression of JUN, FOSB, NFKB1, EGF, EGR1, LTA genes were significantly altered in MCF-7 cells. The relationships of the 26 identified genes were further accessed through net-walking using GNCPro 42 . As shown in Fig. 9, most of the genes are related to each other, with the exception of WEE1, DUSP6, EPS8 and CDKL1. CDKN1A is identified as the most interactive gene in the group. Other active genes include JUN, EGR1, NFKB1, STAT5A, CDK6, TP53, RB1. These active genes appear to be involved in the molecular mechanisms of SAHA or TRAIL action in breast   cancer cells. Our results suggest that the combinatorial treatment of SAHA and TRAIL may target multiple pathways. The exact sequence of signaling events induced by the treatment SAHA and TRAIL for breast cancer cells requires further clarification in future studies.
In conclusion, we investigated comprehensively the distinct effects of SAHA or TRAIL alone and in combination on the proliferation, cell viability, apoptosis, cell cycle distribution and morphological changes of MDA-MB-231 and MCF-7 cells. We observed that the different effects of SAHA or TRAIL alone and in combination on the expression of a number of apoptosis-related molecules, cell cycle, growth factors and their receptors in cancer cells. Our results demonstrated that the combinatorial treatment of SAHA and TRAIL may target multiple pathways and could serve as an effective therapeutic strategy against breast cancer. An improved understanding of the molecular mechanisms may facilitate either SAHA or TRAIL targeted use and the selection of suitable combinations.    To confirm the results of RTCA, 1.0 × 10 4 /well of MDA-MB-231 or MCF-7 cells was loaded into 96 well plates. Subsequently, the cells were starved without FBS for 24 hours, and the medium in each well was replaced with medium containing various concentrations of SAHA or TRAIL as described above. The same concentrations of DMSO were used as control. For an additional 48 hours incubation, the culture supernatants were used to assess drug dose-response effects using the CellTiter 96 Aqueous One Solution Cell Proliferation Assay (Promega) according to the manufacturer's protocol. The spectrophotometric absorbance of each sample was measured at 490 nm using TECAN spectora. The dose-response effect was defined by plotting the corrected absorbance versus concentration of drug.

Materials and Methods
Cell Viability, Apoptosis and Cell Cycle assay. MDA-MB-231 cell was plated in 6 well plates at a density of 5 × 10 5 cells per well. After synchronization with 5 μ M DMSO (basal medium) without FBS for 24 hours, the cells were incubated in complete culture medium containing 5 μ M SAHA or combining with 50 ng/ml TRAIL for 48 hours. MCF-7 cell was also plated in 6 well plates at a density of 5 × 10 5 cells per well. After synchronization with 5 μ M DMSO (basal medium) without FBS for 24 hours, the cells were incubated in complete culture medium containing 5 μ M SAHA or combining with 100 ng/ml TRAIL for 24 hours.
Cell viability assay was performed using the Muse Count & Viability reagent (Millipore) following the manufacturer's protocols. Following trypsinization, 2 × 10 5 of harvested cells (50 ul cell suspension) were added with 450 ul Count & Viability reagent (Millipore). The results were obtained with Muse Count & Viability software module, and the statistics showed the percentage of viable cells. For the apoptotic assay, 1 × 10 6 of cells were transferred in suspension to a new tube and incubated with 100 μ l of Muse Annexin V & Dead Cell reagent (Millipore) for 20 minutes at room temperature. Muse Cell Analyzer (Millipore) determined apoptosis, and the statistics showed the percentages of the cells represented by alive, apoptosis and dead population. For cell cycle assay, 5 × 10 5 cells were plated per well into a 6-well plate. After treatment with 5 μ M SAHA or combining with 100 ng/ml TRAIL as described above, 10 μ l PI were added to the cell suspension. Samples were filtered through 200-mesh filters and then analyzed on BD FACS Calibur flow cytometer. The DNA content of the cells was analyzed for the percentage of cells in G 1 , S and G 2 /M phase.
Time-lapse live cell imaging acquisition. For our experiments, Nalge-Nunc "Lab-Tek ® " coverglass system (Rochester, NY) was used to culture cells. Briefly, a density of 5 × 10 5 cells was loaded in the each well of 4-chamber system. SAHA, TRAIL alone and combination treatments were initiated on experimental day 3 after starvation and aggregate formation. Phase contrast microscopic images of cells were obtained using the BioStation IM-Q (Nikon Corporation, Tokyo, Japan). BioStation IM-Q is an automatic cell maintenance system with a microscope and camera that enables scheduled automatic image acquisition. Image acquisition timing was set to every 12 hour (magnification = 20x) and time points were designated as time 0, 12, 24, 36 and 48, indicating each of the 12 hour imaging intervals.
Apoptotic fluorescence detection. Annexin-V-FLUOS staining kit was used for apoptosis detection following the manufacturer's instruction. Briefly, 5 × 10 5 cells were seeded onto each well of 12 well plates and allowed to adhere overnight to be treated with desired concentrations of SAHA and TRAIL for 0, 12, 24, 36 and 48 hours. Cells were added 100 μ L Annexin-V-FLUOS labeling solution and incubated for 15 minutes at 15 °C-20 °C and immediately analyzed by fluorescence microscopy.
Human apoptosis antibody array. The array is a rapid, sensitive tool to detect the relative levels of expression of apoptosis-related proteins simultaneously. The experiment was carried out in accordance with manufacturer's instructions. First, approximately 1 × 10 7 cells with SAHA and TRAIL treatment were solubilized in lysis buffer and centrifuged at 14000 g for 5 minutes. Protein concentrations of the resulting lysates were measured using a BCA protein assay kit. Next, each of antibody-coated array membranes was placed into the provided dish, and 200 ug of prepared cell lysates were added each well of the dish to incubate at 4 °C with gentle shaking overnight. The membranes were washed with wash buffer and then incubated with 1.5 ml lyophilized biotinylated antibodies for 1 hour on a rocking platform shaker. The mixture of biotin-conjugated antibodies was removed, and membranes were incubated with horseradish peroxidase -conjugated streptavidin for 30 minutes. After a final wash, membrane intensity was acquired using chemiluminescence and pixel densities can be analyzed using Gelpro Analyzer software (Media Cybernetics, Rockville, MD). Densities were measured as a percentage of the positive controls included on each membrane. After subtracting background signals and normalization to positive controls, comparison of signal intensities among array images can be used to determine relative differences in expression levels of each protein between groups.
Q-PCR array. To validate relative results of mRNA expression, human qPCR arrays for cell cycle tox and cancer related gene and EGF/PDGF signaling related gene were employed. Briefly, total RNA was collected from cells using high pure RNA isolation kit with DNase I. RNA was quantified with a Nanodrop 2000 machine (Thermo Scientific). A total of 1 μ g RNA was used for reverse transcription with transcriptor first strand cDNA synthesis kit. Q-PCR reactions were performed on a 7500 Real-time PCR system following manufacturer's instructions. Data normalization was based on correcting all C t values for the average C t values of GAPDH gene present on the array. Three independent biological replicates were performed.
To analyze the PCR-array data, an MS-Excel sheet with macros was downloaded from the website (https:// www.qiagen.com/us/products/genes%20and%20pathways/data-analysis-center-overview-page/), imported the C t values into the template sheet, uploaded the template form and chose the correct analysis factors. Relative changes of gene expression in the array were calculated using the 2 −ΔΔCt (threshold cycle) method. Data analysis. Student's t-test was used for data analysis. Data are presented as mean ± SEM. Values for p < 0.05 were considered statistically significant. The model included the main effects of treatments and replicates.