Vitamin E δ-Tocotrienol Induces p27Kip1-Dependent Cell-Cycle Arrest in Pancreatic Cancer Cells via an E2F-1-Dependent Mechanism

Vitamin E δ-tocotrienol has been shown to have antitumor activity, but the precise molecular mechanism by which it inhibits the proliferation of cancer cells remains unclear. Here, we demonstrated that δ-tocotrienol exerted significant cell growth inhibition pancreatic ductal cancer (PDCA) cells without affecting normal human pancreatic ductal epithelial cell growth. We also showed that δ-tocotrienol-induced growth inhibition occurred concomitantly with G1 cell-cycle arrest and increased p27Kip1 nuclear accumulation. This finding is significant considering that loss of nuclear p27Kip1 expression is a well-established adverse prognostic factor in PDCA. Furthermore, δ-tocotrienol inactivated RAF-MEK-ERK signaling, a pathway known to suppress p27Kip1 expression. To determine whether p27Kip1 induction is required for δ-tocotrienol inhibition of PDCA cell proliferation, we stably silenced the CDKN1B gene, encoding p27Kip1, in MIAPaCa-2 PDCA cells and demonstrated that p27Kip1 silencing suppressed cell-cycle arrest induced by δ-tocotrienol. Furthermore, δ-tocotrienol induced p27Kip1 mRNA expression but not its protein degradation. p27Kip1 gene promoter activity was induced by δ-tocotrienol through the promoter's E2F-1 binding site, and this activity was attenuated by E2F-1 depletion using E2F-1 small interfering RNA. Finally, decreased proliferation, mediated by Ki67 and p27Kip1 expression by δ-tocotrienol, was confirmed in vivo in a nude mouse xenograft pancreatic cancer model. Our findings reveal a new mechanism, dependent on p27Kip1 induction, by which δ-tocotrienol can inhibit proliferation in PDCA cells, providing a new rationale for p27Kip1 as a biomarker for δ-tocotrienol efficacy in pancreatic cancer prevention and therapy.


Introduction
Pancreatic cancer is one of the most lethal cancers in the United States, ranking fourth among the leading causes of cancer-related deaths [1]. Despite treatment developments, the death rate for patients with pancreatic cancer has overall remained unchanged for decades. Investigations into novel therapies and chemopreventive agents are clearly warranted.
Studies have suggested that increased intake of dietary fruits, vegetables, and cereal grains may decrease pancreatic cancer risk [2,3,4]. Tocotrienols, found in cereal grains, comprise one of the most compelling groups of anti-tumor bioactive compounds [5]. Tocotrienols are a group of four (a-, b-, d-, c-) unsaturated, naturally occurring vitamin E compounds that not only inhibit the proliferation of a variety of human tumor cells, including breast, colon, lung, and hepatocellular [6,7,8], but also exhibit chemopreventive properties [9,10]. However, how tocotrienols attenuate tumor proliferation is poorly understood.
We previously demonstrated that d-tocotrienol exhibits the most potent anti-tumor activity among the four tocotrienol isoforms in pancreatic cancer cells [11,12]. In an ongoing phase I doseescalation clinical trial in pancreatic cancer patients, preliminary findings revealed that d-tocotrienol had no obvious toxicity at up to 3200 mg/day, which is 5 times the predicted biologically active clinical dose [13]. These findings underscore the promise of dtocotrienol for pancreatic cancer intervention. To further translate these findings in the clinic, it is important to identify relevant biomarkers of d-tocotrienol activity for early-phase hypothesesdriven clinical trials.
To this end, we investigated how d-tocotrienol inhibits pancreatic cancer cell growth and identified the cyclin-dependent kinase (CDK) inhibitor p27 Kip1 as a molecular target of dtocotrienol. p27 Kip1 functions as a tumor suppressor by its ability to block cell proliferation. p27 Kip1 is an atypical tumor suppressor because mutations of its gene are extremely rare. Nevertheless, tumor cells have evolved other mechanisms to inactivate p27 Kip1 , including increased proteolytic degradation and exclusion from the nucleus. In fact, p27 Kip1 loss has been associated with pancreatic cancer progression and poor prognosis [14,15,16,17]. Here, we report for the first time that p27 Kip1 plays a central role in d-tocotrienol-induced G 1 arrest. We also observed that induction of p27 Kip1 by d-tocotrienol occurs at the transcription level involving E2F-1-mediated promoter activation and mRNA induction.
siRNA Knockdown of p27 Kip1 in MIAPaCa-2 Cells Pre-designed, siRNA to CDK inhibitor 1B (p27 Kip1 , #118714) and nonspecific siRNA (#4611) were purchased from Ambion (Austin, TX). MIAPaCa-2 cells were plated overnight in 12-well plates without antibiotic. Transient transfection of siRNA was carried out using Oligofectamine reagent (Invitrogen, Carlsbad, CA), per manufacturer's instructions. In brief, 5 nM p27 Kip1 siRNA or control siRNA was mixed with Opti-MEM medium (Invitrogen) to a total volume of 90 mL and then complexed with 2 mL of Oligofectamine and 8 mL of Opti-MEM (total volume of complex was 100 mL). Before transfection, old medium was discarded, cells were washed with fresh Opti-MEM, and 400 mL of fresh Opti-MEM were placed into each well before adding the RNA-Oligofectamine complex. After 8 hours, 500 mL of DMEM containing 30% FBS and no antibiotics were added to each well, and cells were further incubated for 40 hours. After 40-hour transfection, cells were treated with d-tocotrienol for an additional 24 hours and harvested for trypan blue and Western blot analysis.

Anchorage-Independent Growth Assays
For soft agar growth assays, cells were seeded at 1610 3 cells/ well in triplicate in 12-well culture dishes in 0.35% agar over a 0.6% agar layer. Various concentrations of d-tocotrienol or vehicle were included in the 0.3% agar layer of cells. Cultures were fed and treated with compound or vehicle weekly until colonies grew to a suitable size for observation (,3-4 weeks). Colonies were photographed after incubation with 1 mg/mL MTT overnight and counted. Growth of d-tocotrienol-treated colonies was compared to vehicle-treated colonies (control). Three separate experiments were performed.

FACS and Cell Proliferation Assay
Exponentially growing pancreatic cells were grown to 70% confluency in 96-well plates and incubated with increasing concentrations of d-tocotrienol or vehicle for 24-72 hours. Wells were examined for cell growth and proliferation using the MTT colorimetric assay. IC 50 results for each cell line were determined for each 24-hour time point. For cell-cycle analysis, pancreatic cells were grown to 70% confluency in 100-mm plates and then serum starved for 48 hours to allow for synchronization. After 48 hours, medium was aspirated and fresh medium with dtocotrienol (IC 50 ) or vehicle was added for 24 hours. Treated medium was then collected, monolayers were washed with cold PBS, cells were trypsinized, and cell pellets were collected. Cell pellets were washed twice with PBS, fixed in cold methanol, and rewashed with PBS to remove methanol. After resuspension in 300-500 mL PBS, cells were digested with 20 mg/mL RNase and cellular DNA was stained with propidium iodide (50 mg/mL) by 3hour incubation at room temperature in the dark. Cell-cycle distribution was analyzed by flow cytometry using a fluorescenceactivated cell sorting (FACS) system (Becton Dickinson, Franklin Lakes, NJ).

Confocal Microscopy Analysis
Treated MiaPaCa-2 cells (50,000) per 250 mL of 20% FBS-PBS were added to each cytofunnel slot and spun at 570 rpm for 5 minutes at high acceleration. Slides were removed and air dried at room temperature. Cells were fixed in 4% paraformaldehyde for 10 minutes, washed three times with 1X PBS with agitation for 5 minutes/wash, and then permeabilized in 0.5% Triton X-100 for 5 minutes at room temperature. Cells were blocked with 2% PBS-BSA (100 mL) for 30 minutes, and p27 Kip1 antibody (1 :500) dilution prepared in 2% BSA was directly applied. Slides were incubated for 1 hour at room temperature in a closed humid chamber. Cells were then washed three times with 1X PBS with agitation for 5 minutes/wash. Secondary antibody (Alexa Fluor 594) in PBS (1:500) was prepared and added to fixed cells. Slides were again incubated for 1 hour at room temperature in closed humid chamber and then washed three times with 1X PBS for 5 minutes/wash. VectaShield (50 mL) with DAPI was added, and coverslips were applied. After incubation at 4uC in dark conditions, slides were examined under a confocal microscope.

Cytosolic and Nuclear Protein Extraction
Proteins were extracted using NE-PER Nuclear and Cytoplasmic Extraction Reagent Kit (Pierce). Treated MIAPaCa-2 cells were isolated as 20-mL packed cell volume (40 mg) in a 1.5-mL microcentrifuge tube by 5-minute centrifugation at 5006g. Supernatant was discarded, cell pellets were dried, and ice-cold CER-1 containing protease inhibitors (200 mL) was added. Tubes were vortexed vigorously for 15 seconds and incubated on ice for 10 minutes. Next, ice-cold CERII (11 mL) was added to tubes, which were vortexed for 5 seconds and incubated on ice for 1 minute, vortexed again for 5 seconds, and then centrifuged at 14,0006g for 5 minutes. Supernatant (cytosolic extract) was transferred to fresh pre-chilled tubes and stored at 280uC. We resuspended the insoluble pellet containing nuclei in 100 mL of ice-cold NER containing protease inhibitors. Tubes were vortexed for 15 seconds and kept on ice for 10 minutes. This step was repeated four times (40 minutes). Tubes were centrifuged at 14,0006g for 10 minutes, and supernatant (nuclear extract) was transferred to pre-chilled tubes and stored at 280uC.

Luciferase Reporter Assay
MIAPaCa-2 cells were seeded in 6-well plates at 2610 5 cells/ well. Each well was transfected the following day with 2 mg of p27 Kip1 luciferase reporter plasmid (full-length p27 Kip1 -1609, different 59 deletion mutants of mouse p27 Kip1 promoter luciferase reporter, or pGL-3 base cDNA empty vector; all plasmids provided by Dr. Pledger) along with siRNA E2F-1 or non-target siRNA (Santa Cruz). Metafectene was used as the transfection reagent, per manufacturer's instructions. Lysates were collected 24 hours after d-tocotrienol treatment (IC 50 50 mM). Luciferase activity was measured by the luciferase assay system kit (Promega).
For normalization of transfection efficiency, 200 ng of Renilla reniformis luciferase expression plasmid (pRL-TK vector, Promega) was included in the transfection.

RT-PCR Analysis
MIAPaCa-2 cells were seeded in 6-well plates and treated with d-tocotrienol (IC 50 50 mM) or vehicle (as control) for 24 hours. Cells were then harvested in 1X lysis buffer, and RNA was isolated using Allprep RNA/protein kit according to manufacturer's instructions (Qiagen, Valencia, CA). Reverse transcription of total RNA was performed using the SuperScript III kit (Invitrogen). The following forward and reverse primers, respectively, were used for PCR reaction: for p27 Kip1 , 59-TAACCCGGGACTTG-GAGAAG and 59-GCTTCTTGGGCGTCTGCTC to amplify a 450-bp product; for actin, 59-GCTCGTCGTCGA-CAACGGCT and 59-CAAACATGATCTGGGTCATCTTCTC

Cyclohexamide Blocks Protein Synthesis
After 12-hour d-tocotrienol treatment (IC 50 50 mM), d-tocotrienol was removed by rinsing MIAPaCa-2 cells three times with PBS; cyclohexamide (40 mg/mL) was then used to block protein synthesis. p27 Kip1 turnover rate was examined by Western blot.

Ethics Statement
This study was carried out in strict accordance with recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. The protocol was approved by the University of South Florida Institutional Animal Care and Use Committee (Application 2805).

Anti-tumor Activity in Nude Mouse Xenograft Model
We used female athymic nude (nu/nu) mice, 5-6 weeks old (Charles River, Wilmington, MA), for our animal studies. Animals were kept in clean cages limited to 4 mice per cage. The mice were cared for with ample food and water and examined on a daily basis. If tumors interfered with ambulation, caused signs of weight loss .10%, caused respiratory distress, or grew to .2 cm in size, the mice were humanely euthanized by exposure to increasing concentrations of carbon dioxide.
MIAPaCa-2 cells were harvested and resuspended in PBS (1610 6 cells/50 mL) and an equal volume of Matrigel (BD Biosciences). Cell samples (100 mL) were then injected subcutaneously into the right flank. Once tumors reached between 250 and 300 mm 3 , mice were randomized into groups of 10 and dosed by oral gavage with 0.1 mL of vehicle (purified olive oil) or d-tocotrienol (100 mg/kg) daily for 3 weeks. Tumor volumes were determined twice per week by measuring length (l) and width (w) and by calculating the volume: V = (l + w)/2 6 (l 6 w) 6 0.5236. Statistical significance between control and treated animals was determined using Student's t-test.

Immunohistochemistry and Slide Quantitation
Tumors from xenograft experiments were fixed in 4% paraformaldehyde, pH 7.2. After fixation, the tissue samples were processed into paraffin blocks. The primary antibodies used in this study were mouse monoclonal (p27, p-MAPK, and Ki-67) antibodies raised against the corresponding antigens of human origin in paraffin-embedded sections. Immunohistochemical staining was performed on a Ventana BenchMark XT (Tucson, AZ) automated slide stainer, using 4-mm thick paraffin sections from each of the representative tumor blocks selected. The sections were deparaffinized, rehydrated, and incubated with 3% H 2 O 2 to block endogenous peroxidase. After antigen unmasking using proprietary CC1 solution for 60 minutes online (standard) at 100uC, the sections were incubated with antibodies to p27 (Kip 1,

Evaluation of the Stains
The immunohistochemical expression of p27, p-MAPK, and Ki-67 proteins were determined as the product of immunostain intensity and percent of cells stained. These were scored on a 0-3 scale, with 3 being maximal. The immunostain intensity was scored with no staining being 0, light staining as 1, moderate staining as 2, and heavy staining as 3. The percent of cell stained was measured with no detectable staining as 0, 1-33% as 1, 34-66% as 2, and 67-100% as 3. The final IHC score was the product of the percent of cells stained score multiplied by the intensity score, allowing for a maximal score of 9 and a minimal score of 0.

Statistical Analysis
Data, expressed as means 6 SEM, were analyzed statistically using unpaired t-tests or one-way analysis of variance (ANOVA) where appropriate. We used GraphPad Prism version 5.04 for our analyses. Statistical significance was set at P,0.05.

d-Tocotrienol Inhibits Anchorage-Dependent and -Independent Cell Growth and Induces G 1 Arrest in Pancreatic Cancer Cells
The effects of d-tocotrienol ( 1A) on cell growth in MIAPaCa-2, BxPC-3, and SW1990 human pancreatic cancer cells were examined by 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) cell proliferation assay. Cells were treated with 0-50 mM d-tocotrienol for 24, 48, and 72 hours. The effects of dtocotrienol were also evaluated in the non-transformed human pancreatic ductal epithelial cell line HPDE6 C7 to rule out possible cytotoxic effects of d-tocotrienol on non-neoplastic cells. d-Tocotrienol treatment inhibited anchorage-dependent cell proliferation in both a time-and concentration-dependent manner in human pancreatic cancer cells ( Figure 1B); however, no significant growth inhibitory effects were noted in HPDE6 C7 cells. d-Tocotrienol treatment also significantly inhibited colony formation in MIAPaCa-2 cells grown in soft agar from 2.5 mM (P = 0.02) ( Figure 1C).
d-Tocotrienol treatment also had a significant selective effect on cell-cycle progression, as demonstrated by an increase in the percentage of pancreatic cancer cells but not HPDE6 C7 cells in the G 1 phase (P,0.001 vs. P = 0.92) ( Figure 1D). We found that the pancreatic cancer cells accumulated in the G 1 phase at the expense of a decrease in the S-phase population.

d-Tocotrienol Induces p27 Kip1 Expression and Inhibits RAS-MEK-ERK Signaling
Regulation of intracellular signaling pathways is central to the ability of oncogenes to promote cell-cycle progression. Two major pathways intimately involved in the G 1 -to-S traverse are the RASactivated RAF-MEK-ERK and PI3-AKT pathways. These pathways influence the expression, activity, or subcellular localization of key components of the cell-cycle machinery such as cyclins, CDKs, and CDK inhibitors, leading to the appropriate activation of E2F-1 transcription factors. Several agents have been described that regulate the G 1 traverse and transition into the S-phase in pancreatic cancer cells, and p27 Kip1 has been reported to be increased by these agents [22,23,24,25,26]. We therefore determined the kinetics of p27 Kip1 levels and RAF-MEK-ERK pathway activity in pancreatic cancer cells and in HPDE6 C7 cells exposed to d-tocotrienol. We found that d-tocotrienol significantly increased p27 Kip1 levels by 6 hours in the pancreatic cancer cells (Figure 2A). This increased level was sustained and increased in all cell lines by 48 hours and associated with corresponding inhibition of the activity of the activated RAF-MEK-ERK signaling pathway, as measured by decreased phosphorylated MEK and ERK levels in the pancreatic cancer cells (Figure 2A). In contrast, the PI3-AKT pathway was initially induced at 12 hours followed by subsequent inhibition at 24 hours, as measured by phosphorylated AKT levels (Figure 2A and 2B). Consistent with the effect on proliferation, d-tocotrienol suppressed p27 Kip1 levels in the nontransformed human pancreatic ductal cell line, HPDE6 C7, and had no effect on MEK, ERK, or AKT. We also showed that the effects of d-tocotrienol were specific to malignant cells, as dtocotrienol did not induce p27 Kip1 levels or alter MEK, ERK, or AKT expression in the nontransformed pancreatic ductal epithelial cells. The mechanisms of this need further investigation.
Besides modulation of its expression levels, subcellular localization is also important in governing p27 Kip1 function. To act as a cell-cycle inhibitor, p27 Kip1 must be located in the nucleus, whereas its cytoplasmic mislocalization favors cell-cycle progression and may contribute to cellular transformation [27,28]. Figure 2C shows that p27 Kip1 levels are increased in both the cytosolic and nuclear compartments of d-tocotrienol-treated MIAPaCa-2 cells compared to vehicle. In Figure 2D, we observed p27 Kip1 cytoplasmic accumulation in vehicle-treated MIAPaCa-2 cells versus increased p27 Kip1 levels in the nuclear compartment of serum-starved quiescent MiaPaCa-2 cells and d-tocotrienoltreated MIAPaCa-2 cells. Some authors have suggested that tocotrienols modulate HMG-CoA reductase activity via posttranscriptional actions, thereby inhibiting farnesylation of downstream oncogenic signaling targets, similar to the effects of lovastatin [29,30,31]. To determine whether d-tocotrienol inhibits proliferation through mevalonate pathway inhibition, we treated MIAPaCa-2 cells with d-tocotrienol or lovastatin and with increasing concentrations of mevalonate, a downstream product of HMG-CoA reductase. Addition of mevalonate to d-tocotrienoltreated cells resulted in no significant rescue from the growth inhibitory effects of d-tocotrienol on MIAPaCa-2 cells ( Figure 2E); however, lovastatin's ability to inhibit proliferation was rescued by mevalonate.
p27 Kip1 Is Required for d-Tocotrienol-Induced G 1 Arrest To determine the role of p27 Kip1 induction in d-tocotrienol pancreatic cancer cell growth inhibition, we transiently reduced p27 Kip1 expression in MIAPaCa-2 cells using p27 Kip1 small interfering RNA (siRNA; confirmed by Western blot). At 24 hours, absence of p27 Kip1 expression significantly abrogated d-tocotrienol-induced cell growth inhibition ( Figure 3A). To confirm these findings, stable cell lines expressing p27 Kip1 short hairpin RNA (shRNA) or vector were created using MIAPaCa-2 cells. Reduced p27 Kip1 expression was confirmed by Western blot analysis. Figure 3B shows that depletion of p27 Kip1 resulted in resistance to d-tocotrienol treatment. To investigate whether these observations could be generalized beyond MIAPaCa-2 cells, we analyzed stable MEFs [19,20] exhibiting p27 Kip1 knock-out or wild-type cells after d-tocotrienol treatment. We found that the lack of p27 Kip1 rendered MEFs partially resistant to the antiproliferative effects of d-tocotrienol in this cell line ( Figure 3C). Together, these studies underline the importance of p27 Kip1 in dtocotrienol-induced G 1 arrest and show that the contribution of this CDK inhibitor is not cell line specific.

d-Tocotrienol
Regulates p27 Kip1 Transcription in an E2F-1-Dependent Manner CDK inhibitor levels are intricately regulated at the level of mRNA transcription/protein synthesis and/or degradation. To determine the effects of d-tocotrienol on p27 Kip1 protein stability, we conducted cyclohexamide chase experiments. As shown in Figure 4A and 4B, the pattern of p27 Kip1 degradation in dtocotrienol-treated cells was similar to corresponding controls. Protein degradation at 4 hours in vehicle-treated cells was 45% versus 41% in d-tocotrienol-treated cells. These results indicate that mechanisms involved in d-tocotrienol-induced p27 Kip1 expression were not due to protein stabilization.
We next investigated whether d-tocotrienol induced p27 Kip1 expression at the transcriptional level. Real-time PCR showed that d-tocotrienol treatment significantly increased p27 Kip1 mRNA levels compared to vehicle in MIAPaCa-2 cells, suggesting that the d-tocotrienol-induced increase in p27 Kip1 protein levels is through induction of p27 Kip1 /mRNA transcription ( Figure 4C). d-Tocotrienol's effect on p27 Kip1 transcription was confirmed with promoter reporter assays. d-Tocotrienol treatment significantly induced (,50%) -1609 p27 Kip1 promoter activity, demonstrating that d-tocotrienol regulates p27 Kip1 protein expression at the transcription level. Moreover d-tocotrienol-dependent p27 Kip1 promoter activation was maintained upon 59-deletion to -615 bp but not upon further deletion to -326 bp. The -615/-326 DNA region contains three E2F-1 binding sites, which we have previously identified as crucial elements for the regulation of this gene ( Figure 4D and 4E) [21]. To determine whether d-tocotrienol E2F-1 binding to the p27 Kip1 promoter was important in dtocotrienol-induced p27 Kip1 transcription, we performed Western blot ( Figure 4F) and reporter assays ( Figure 4G) with knock-down of E2F-1 protein. As shown in Figure 4G, suppression of E2F-1 expression significantly inhibited d-tocotrienol's ability to induce p27 Kip1 .

Effects of d-Tocotrienol on Pancreatic Tumor Cell Growth In Vivo
Using nude mouse xenografts, we found significant inhibition of tumor growth in mice treated with d-tocotrienol ( Figure 5A). Furthermore, immunohistochemical analysis demonstrated decreased Ki67, confirming inhibition of proliferation, decreased phosphorylated MAPK expression, and induced expression of p27 Kip1 ( Figure 5B). These findings correlate with our in vitro results and support the concept that Ki67 and p27 Kip1 expression levels are important potential biomarkers in d-tocotrienol treatment of pancreatic cancer.

Discussion
The novel and central finding in this study is that p27 Kip1 is indispensable in causing growth arrest in human pancreatic cancer cells by d-tocotrienol in vitro. In this study, MEFs [19,20] in which p27 Kip1 was knocked out and human pancreatic cancer MIA-PaCa-2 cell variants in which p27 Kip1 was depleted were resistant to d-tocotrienol-induced cell growth inhibition. Furthermore, dtocotrienol up-regulation of p27 Kip1 expression in MIAPaCa-2 cells was through induction of E2F-1-dependent p27 Kip1 transcription. These results establish increased p27 Kip1 levels as a significant molecular biomarker of d-tocotrienol anti-tumor efficacy against pancreatic cancer.
A key property of cancer cells and their progeny is their ability to reproduce in defiance of the normal restraints on cell division [32]. Limitless replicative potential is a hallmark of pancreatic oncogenesis. The mechanism of limitless replicative potential is linked to G 1 -to-S cell-cycle progression because this cell-cycle phase is linked to external stimuli and also commits the cell to DNA replication and subsequent mitosis. G 1 checkpoint abrogation is a common phenomenon in carcinogenesis, giving tumor cells limitless replicative potential. In animal cells, entry of cells into a new cell cycle is controlled by regulation of CDK activity in G 1 . One of the critical mechanisms by which animal cells suppress CDK activity in G 1 is through accumulation of the CDK inhibitor p27 Kip1 . p27 Kip1 physically interacts with CDK via its aminoterminal domain, inhibiting CDK activity. Thus, activating the G 1 checkpoint by upregulating the expression of p27 Kip1 is a logical approach for controlling cancer cell proliferation. In this study, we showed that p27 Kip1 induction is a crucial event in d-tocotrienolinduced G 1 arrest and inhibition of cell proliferation in MIAPaCa-2 pancreatic cancer cells. More importantly, this study provides evidence that p27 Kip1 can function in the G 1 -to-S transition checkpoint and mediate d-tocotrienol-induced G 1 arrest and cell growth inhibition in both MIAPaCa-2 pancreatic cancer cells and MEFs.
Once we identified that up-regulation of p27 Kip1 plays a central role in d-tocotrienol-induced G 1 arrest, our next aim was to identify the mechanism by which d-tocotrienol induced p27 Kip1 expression. p27 Kip1 can have opposing roles in the process of malignant transformation, dependent on its cellular localization. When localized in the nucleus, it functions as a tumor suppressor by acting as a negative regulator of the G 1 -S transition by binding and inhibiting the cyclin E-CDK2 complex. Cytoplasmic localization of p27 Kip1 supports the assembly and nuclear import of cyclin D-CDK4/6, thus promoting cell proliferation. Moreover, cytoplasmic mislocalization of p27 Kip1 seems to contribute to the progression of many cancers by increasing cell motility and metastases and, importantly, inhibiting apoptosis [33,34]. Therefore, p27 Kip1 can be considered a nuclear tumor suppressor and a cytoplasmic oncogene. In evaluating whether the d-tocotrienolinduced increase in p27 Kip1 levels in pancreatic cancer cells occurred in a specific cellular compartment, we observed increased p27 Kip1 in the nucleus of pancreatic cancer cells in vitro and in vivo. These findings suggest that d-tocotrienol treatment favors the tumor-suppressive function of p27 Kip1 .
Ras activation indirectly causes cytoplasmic localization of p27 Kip via activation of its effector pathways, such as RAF-MEK-ERK. Phosphorylation of p27 Kip1 on 3 different sites (Ser-10, Thr-157, and Thr-198) by components of these signaling pathways has been shown to result in p27 Kip1 cytoplasmic localization [33,35,36,37,38,39,40,41,42]. Ras-regulated signaling pathways play an important role in initiation and progression of human pancreatic cancer [43,44,45]. In particular, the RasRRAF-MAP-ERK kinase (MEK)RERK-MAPK pathway plays an important regulatory role in cell-cycle division [46,47,48]. Gysin et al. showed that pharmacologic inhibition of RAF-MEK-ERK signaling in pancreatic cancer cells inhibits cell growth and results in G 1 cell-cycle arrest through induced expression of p27 Kip1 [49]. Our results show that p27 Kip1 induction by d-tocotrienol is related to inhibition of the RAF-MEK-ERK pathway. The mechanism by which d-tocotrienol inhibits Ras activation signaling is poorly understood. A previous study implicated d-tocotrienol inhibition of Ras prenylation by its suppression of HMG-CoA activity [50]. We did not observe any rescue of d-tocotrienol anti-proliferative activity in MIAPaCa-2 cells with mevalonate supplementation, suggesting that d-tocotrienol inhibition of activated Ras effector signaling is not by inhibition of Ras prenylation through inhibition of HMG-CoA reductase.
Mechanisms by which tumor cells inactivate p27 Kip1 include increased proteolytic degradation and inhibition of transcription. Our results demonstrated that d-tocotrienol did not affect the proteolytic degradation compared to control at 4 hours, but induced the transcription of p27 Kip1 . However, the effects of dtocotrienol on p27 Kip1 promoter activity are modest, indicating that transcriptional stimulation might not be the sole contributor to d-tocotrienol-induced p27 Kip1 mRNA accumulation. Additional post-transcriptional mechanisms such as message stabilization by RNA-binding proteins might also be involved. Using p27 Kip1 promoter-deletion mutants and E2F-1 genetic knockdown, we demonstrated that d-tocotrienol-induced p27 Kip1 promoter activity involves E2F-1. At first glance, this induction appears paradoxical because the main function of p27 Kip1 is to inhibit E2F-1 activation by inhibiting the phosphorylation of retinoblastoma protein through inhibition of the G 1 CDK. However, induction of p27 Kip1 expression by E2F-1 may potentially function as a feedback regulatory mechanism that limits E2F-1 activity [21]. We have shown that suppression of E2F-1 expression by siRNA correlated with significantly reduced p27 Kip1 levels and, alternatively, inhibition of p27 Kip1 expression by siRNA-enhanced E2F-1 function [21]. Together, our findings suggest that d-tocotrienol may initially promote E2F-1 function early in the cell's exposure, resulting in E2F-1 binding to the p27 Kip1 promoter and upregulation of p27 Kip1 transcription and protein expression. Increased p27 Kip1 protein levels then inhibit G 1 CDK, which inhibits Rb-phosphorylation and thereby E2F-1 activation.
d-Tocotrienol has shown promising efficacy against several cancer models, including pancreatic cancer [11,12,50]. We have almost completed a phase I dose-escalation study of patients with pancreatic cancer. To advance the development of d-tocotrienol into phase II clinical trials, the molecular targets that play a central role in d-tocotrienol efficacy in pancreatic cancer need to be elaborated so that they can be used as biomarkers of p27 Kip1 efficacy. Here, we show convincingly for the first time that upregulation of p27 Kip1 is an important event in d-tocotrienolinduced G 1 arrest in human pancreatic cancer cells. In summary, our study provides the experimental justification for the use of nuclear p27 Kip1 as molecular markers of d-tocotrienol efficacy in human pancreatic tumor.