Nutrition and Cancer, 61(6), 864–874Copyright © 2009, Taylor & Francis Group, LLCISSN: 0163-5581 print / 1532-7914 onlineDOI: 10.1080/01635580903285130

Induction of Caspase-Independent Programmed Cell Deathby Vitamin E Natural Homologs and Synthetic Derivatives

Constantina ConstantinouYasoo Health Ltd., Nicosia, Cyprus

John Anthony HyattYasoo Health Inc., Johnson City, Tennessee, and East Tennessee State University, Johnson City,Tennessee, USA

Panayiota S. VrakaUniversity of Cyprus, Nicosia, Cyprus

Andreas Papas and Konstantinos A. PapasYasoo Health Inc., Johnson City, Tennessee, USA

Constantinos Neophytou and Vicky HadjivassiliouYasoo Health Ltd., Nicosia, Cyprus

Andreas I. ConstantinouUniversity of Cyprus, Nicosia, Cyprus

Current observations in the literature suggest that vitamin Emay be a suitable candidate for cancer chemotherapy. To investi-gate this further, we examined the ability of the vitamin E naturalhomologs [α-, β-, γ -, δ-tocopherols (α-TOC, β-TOC, γ -TOC, δ-TOC) and α-, β-, γ -, δ-tocotrienols (α-TT, β-TT, γ -TT, δ-TT)]and their corresponding succinate synthetic derivatives [α-, β-, γ -,δ-tocopheryl succinates and α-, β-, γ -, δ-tocotrienyl succinates (α-TS, β-TS, γ -TS, δ-TS)] to induce cell death in AR– (DU145 andPC3) and AR+ (LNCaP) prostate cancer cell lines. The most ef-fective of all the natural homologs of vitamin E was determined tobe δ-TT, whereas δ-TS was the most potent of all the natural andsynthetic compounds of vitamin E examined. Both γ -TT and δ-TTinduced caspase activity selectively in AR+ LNCaP cells, suggest-ing a possible role for AR for the activation of caspase-dependentprogrammed cell death (CD-PCD). More important, however, γ -TT, δ-TT, γ -TS, and δ-TS activated dominant caspase–independentprogrammed cell death (CI-PCD) in all prostate cancer cell linesexamined. Thus, vitamin E homologs and synthetic derivativesmay find applications in the treatment of prostate tumors that areresistant to caspase-activating therapeutic agents.

Submitted 3 May 2009; accepted in final form 13 August 2009.Address correspondence to Andreas I. Constantinou, Laboratory of

Cancer Biology and Chemoprevention and, Department of BiologicalSciences, Faculty of Pure and Applied Sciences, University of Cyprus,P.O. Box 20537, 1678 Nicosia, Cyprus. Fax: 00-35722-892881. E-mail:andreasc@ucy.ac.cy

INTRODUCTIONProstate cancer is the most common cancer in males, and it is

only second to lung cancer with respect to mortality (1). Studiesthat have been performed using cancer cell lines, animal models,and clinical trials have shown that vitamin E has a possible rolein the chemotherapy of prostate cancer (2–7). Vitamin E is animportant micronutrient consisting of 8 homologs with strongantioxidant activities: 4 tocopherols (TOC; α-, β-, γ - and δ-TOC) and 4 tocotrienols (TT; α-, β-, γ -, and δ-TT; Fig. 1A and1B) (8,9).

Recent studies have shown that the anticancer activities ofvitamin E homologs can be attributed to their ability to induceapoptosis (10–19). However, the efficacy and mechanism ofapoptosis modulated by the 8 homologs had not been evalu-ated previously in a single study. Although α-TOC is capable ofinducing cell cycle block, several studies have shown that thishomolog is not a strong proapoptotic inducer (6–7, 10). The roleof β-TOC in the induction of programmed cell death (PCD) hasnot been investigated thoroughly due to the low abundance ofthis compound in natural sources (11,12). γ -TOC and δ-TOCare both capable of inducing apoptosis (4–6). The induction ofapoptosis by γ -TOC is dependent on the cellular microenviron-ment (4–6). On the other hand, even though the proapoptoticpotency of δ-TOC has not been thoroughly investigated, thishomolog has been shown to be effective in all carcinogenic celllines tested (13, 14). Generally, the order of apoptotic efficiency

864

VITAMIN E INDUCED CASPASE INDEPENDENT CELL DEATH 865

FIG. 1. The structures of vitamin E natural homologs and synthetic derivativesexamined. The structures of A: tocopherols, B: tocotrienols, C: tocopherylsuccinate derivatives, and D: tocotrienyl succinate derivatives.

of the three vitamin E tocopherols is considered to be δ-TOC> γ -TOC > α-TOC (14).

Tocotrienols, generally, display a greater antitumor activitythan tocopherols without affecting normal cell growth and via-bility (13,15–18). Overall, the apoptotic potency of tocotrienolsis considered to be as follows: δ-TT> γ -TT> α-TT (19). Sim-ilarly to β-TOC, the role of β-TT in apoptosis has not beeninvestigated because it is present at very low levels in naturalsources, thereby making its extraction and use in research verydifficult (20). Although there is 1 report for total synthesis ofβ-TT, the synthetic process is considered difficult (21).

In the past few years, most research has focused on structuralvariations of vitamin E with the aim to improve the proapop-totic potency of these agents. The compounds developed in thismanner were shown to be efficient against a variety of malig-nancies (22). The most studied member of these compoundsis α-tocopheryl succinate (α-TOS). α-TOS has an ester-linked,succinic acid moiety attached to the position-6 oxygen atom ofthe phenolic ring of the chroman head (Fig. 1C). The conversionof α-TOC to α-TOS greatly improves its anticancer action intumorigenic cell lines and animal models while causing no tox-icity in normal cells (23–25). However, one major disadvantageof α-TOS is that oral administration of this compound may notbe effective due to the hydrolysis of the ester linkage by cellu-lar esterases of the intestinal tract yielding α-TOC and succinicacid, neither of which exhibits anticancer properties (23–25).Interestingly, there is very limited evidence in the literature re-garding the proapoptotic activities of the three other succinatederivatives of the tocopherols (i.e., β-TOS, γ -TOS, and δ-TOS)(22) or the more potent tocotrienols (i.e., α-TS, β-TS, γ -TS,and δ-TS) (22,26).

In search of natural homologs and/or synthetic derivativesof vitamin E with strong antitumorigenic potency, we exam-ined the death promoting properties of the 8 natural forms ofvitamin E (α-TOC, β-TOC, γ -TOC, δ-TOC and α-TT, β-TT,γ -TT, δ-TT) and their succinate synthetic derivatives (α-TOS,β-TOS, γ -TOS, δ-TOS and α-TS, β-TS, γ -TS, δ-TS) in an-drogen receptor (AR)– (DU145 and PC-3) and AR+ (LNCaP)prostate cancer cell lines (Table 1, Fig. 1A–1D). To our knowl-edge, this is the first time that the efficacy and mechanism ofapoptosis modulated by the 8 vitamin E homologs have been

TABLE 1List of vitamin E homologs and synthetic derivatives

Category of Vitamin E Acronym

NaturalTocopherols α-tocopherol α-TOC

β-tocopherol β-TOCγ -tocopherol γ -TOCδ-tocopherol δ-TOC

Tocotrienols α-tocotrienol α-TTβ-tocotrienol β-TTγ -tocotrienol γ -TTδ-tocotrienol δ-TTSynthetic

Tocopheryl succinates α-tocopheryl succinate α-TOSβ-tocopheryl succinate β-TOSγ -tocopheryl succinate γ -TOSδ-tocopheryl succinate δ-TOS

Tocotrienyl succinates α-tocotrienyl succinate α-TSβ-tocotrienyl succinate β-TSγ -tocotrienyl succinate γ -TSδ-tocotrienyl succinate δ-TS

866 C. CONSTANTINOU ET AL.

evaluated in a single study. Furthermore, it is also the first timethat the tocopheryl- and tocotrienyl-succinate derivatives havebeen synthesized and their apoptotic properties have been de-termined and compared to those of vitamin E homologs.

Overall, our results have shown the following: 1) δ-TT is themost effective of all the natural homologs of vitamin E, 2) δ-TSis the most potent of all the natural and synthetic compounds ofvitamin E investigated, 3) γ -TT and δ-TT induce low caspaseactivity selectively in LNCaP cells, and 4) γ -TT, δ-TT, γ -TS,and δ-TS activate dominant caspase-independent programmedcell death (CI-PCD) in all prostate cancer cell lines examined.The latter finding is noteworthy since compounds triggeringCI-PCD could find medical applications in the adjuvant treat-ment of cancers with resistance to agents inducing apoptosis viaactivation of pathways of caspase-dependent programmed celldeath (CD-PCD).

MATERIALS AND METHODS

ReagentsDMEM, fetal bovine serum, antibiotic/antimycotic, and

trypsin used in cell culture were purchased from Gibco, Invitro-gen (Carlsbad, CA). The caspase inhibitor benzyloxycarbonyl-Val-Ala-Asp-fluoromerthyletone (z.VAD.fmk) was purchasedfrom Calbiochem (Darmstadt, Germany). The Caspase-8/FLICE Fluorometric Protease Assay was purchased fromBiosource International Inc. (Camarillo, CA), and the Caspase-3/CPP32 Fluorogenic substrate was obtained from Kamiya(Seattle, WA). All other reagents were purchased from Sigma(St. Louis, MO).

Commercial Vitamin E Natural Homologs and SyntheticDerivatives

The compounds listed on Table 1 were purchased, isolated,or synthesized as described in this section (for structures ofcompounds, see Fig. 1). d-α-tocopherol (α-TOC), and d-δ-tocopherol (δ-TOC) were obtained from commercial sources(Sigma or ADM) in minimum 95% purity. d-γ -tocopherol (γ -TOC) was purchased from Eisai Corp (Tokyo, Japan) in >95%purity. Small samples of d-β-tocopherol (β-TOC) and d-β-tocotrienol (β-TT) were generously donated by Tama Biochem-ical Co., Ltd. (Tokyo, Japan) d-α-tocopheryl succinate (α-TOS)was purchased from Sigma.

Isolation of Natural Homologs of Vitamin EIsolation of d-α-tocotrienol (α-TT) and d-γ -tocotrienol (γ -

TT). α-TT and γ -TT were isolated from approximately 50%total tocol-containing concentrate of palm oil (Tocomin-50R)obtained from Carotech, Inc. (Ipoh, Malaysia). This concen-trate contained approximately 20% γ -TT, 11% α-TT, and atotal of 19% other tocopherols and tocotrienols. The remain-der of the concentrate was largely comprised of triglycerides,fatty acids, fatty alcohols, carotenoids, and sterols. Approxi-

mately 60-g samples of Tocomin-50 were chromatographed onopen columns containing 1.5 kg of silica gel. The chromatog-raphy was followed using thin-layer silica gel chromatography(read using UV fluorescence and p-anisaldehyde spray reagent)and proton NMR spectroscopy of concentrated fractions. Elu-tion was with a gradient from pure hexanes to 12% acetone inhexanes using approximately 5 to 8 l increments containing,respectively, 0.5, 1.0, 2.0, 4.0, 7.0, and 12% acetone in hexanes.Tocotrienol-containing fractions (7% acetone and above) werestripped of solvent and rechromatographed on silica gel in thesame way to give two fractions that contained approximately60–75% of α-TT and γ -TT, respectively. To remove nontocolimpurities, which co-eluted with the two desired tocotrienols,the enriched fractions were acetylated. The fractions enrichedin α-TT and γ -TT were each stripped of solvent on the rotovap,redissolved in 20 ml each of pyridine, and treated with 10 ml ofacetic anhydride. After stirring overnight at room temperature,the two acetylation mixtures were quenched by addition of a fewml of water followed by stirring for 2 h. The reaction mixtureswere then poured into 500 ml of water, extracted with ethylacetate, and the ethyl acetate layers washed with 5% HCl (to re-move pyridine), brine, and dried over sodium sulfate. The ethylacetate was then removed on the rotovap, and the resulting twosamples (d-α- and d-γ -tocotrienyl acetates) each were carefullychromatographed on silica gel using flash-column techniques,0–8% acetone in hexane elution. This afforded samples (ap-proximately 5–7 g) of the two acetates, which were judged to beabout 85–90% pure by proton NMR spectroscopy. The purifiedacetates were reconverted to free tocotrienols by transesterifi-cation in anhydrous methanol containing a catalytic amount ofpotassium carbonate (reflux, 1 h). The reaction mixtures werestripped to about 5 ml on the rotovap, diluted with water, neu-tralized with dilute HCl, and extracted with ethyl acetate. Theorganic phases were washed with brine, dried over sodium sul-fate, and stripped on the rotovap. The resulting two tocotrienolswere flash-chromatographed on silica gel (acetone-hexane gra-dient elution) to give about 3.5 g each of > 95% pure (NMR,HPLC) α-TT and γ -TT as pale yellow viscous oils. The com-pounds had IR and NMR spectra in agreement with literaturevalues.

Isolation of d-δ-tocotrienol (δ-TT). Multigram quantities ofδ-TT were obtained in 95% purity using the methods describedabove for α-TT and γ -TT, the principle difference being that thestarting feedstock was a d-δ-tocotrienol-rich fraction of annattooil (DeltagoldR 50) obtained from American River Nutrition(Hadley, MA).

Synthesis of Vitamin E DerivativesSynthesis of d-β-, γ -, δ-tocopheryl- (β/γ /δ/-TOC) and d-

α-, β-, γ -, δ-tocotrienyl succinates (α/β/γ /δ/-TT). The to-copheryl and tocotrienyl hydrogen succinates are well-knowncompounds, described previously (27–29). For our research, the7 commercially unavailable tocol succinates were all preparedby the same method, described as follows. A solution of 0.50 g

VITAMIN E INDUCED CASPASE INDEPENDENT CELL DEATH 867

(0.00126 mole) of > 95% pure β/γ /δ/-TOC or α/β/γ /δ/-TTin 10 ml of anhydrous pyridine was stirred at 20◦C under nitro-gen atmosphere. There was added a catalytic quantity (ca. 50mg) of 4-dimethylaminopyridine and 1g (0.010 mole) of suc-cinic anhydride. The reaction mixture was stirred for 20 h, atwhich time TLC analysis indicated consumption of the startingtocopherol/tocotrienol and formation of a single polar product.There was added to the mixture about 1g of water, and stirringwas continued for 2 h to insure hydrolysis of the excess suc-cinic anhydride employed. The mixture was then poured into250 ml of water and extracted with ethyl acetate. The extractwas washed with 5% aq. HCl (to remove pyridine) and withbrine. The solution was dried over anhydrous sodium sulfateand stripped of solvent on the rotovap to give a crude productas pale yellow viscous syrup, completely homogeneous by TLCand HPLC. Final traces of solvent were removed by storingthe compound on a high-vacuum line at 22◦C for 7 days. Allsuccinates had the expected NMR and IR spectra.

Cell Lines and Culture ConditionsDU-145, PC-3, and LNCaP cell lines were obtained from the

American Type Culture Collection (ATCC) (Manassas, VA).DU145 (AR–), PC3 (AR–), and LNCaP (AR+) cells were cul-tured in DMEM supplemented with 10% fetal bovine serum and1% antibiotic/antimycotic. The cells were maintained at 37◦Cand passaged 2 to 3 times/wk.

Crystal Violet StainingA total of 1 × 104 DU-145, PC3, or LNCaP cells were seeded

per well of a 96-well plate and incubated for 24 h. At the endof 24 h, the growth medium was removed and replaced withfresh DMEM or DMEM supplemented with the various vitaminE compounds and synthetic derivatives (10, 20, 40, or 100 µM)and/or 20 µM z.VAD.fmk, and the plates were incubated forthe time periods described in the figure legends. At the end ofeach incubation period, the medium was removed, and 100 µl of10% formalin was added in each well for 5 min. Subsequently,formalin was removed; the wells were washed with PBS andincubated with 100 µl of 0.2% crystal violet for 10 min. Thewells were then washed with distilled water several times toensure removal of the dye, and the plates were allowed to dry atroom temperature. After drying, 100 µl of acetic acid was addedper well, and the plates were incubated at room temperature for5 min. At the end of the incubation period, 100 µl of acetic acidwere added again to each well, and the plates were immediatelyread on a microplate reader at 620 nm. The absorbance obtainedin untreated control cells for each of the three cell lines and timepoints (24 h, 48 h, 72 h) was considered as 100% viability. TheIC50value (i.e., the concentration of each compound at which50% of cell death was observed) for the different cell lines, andtime points were calculated using Prism software version 5.0(Graphpad, San Diego, CA). Whenever 50% of cell death wasnot achieved with up to 100 µM, the IC50 value was designatedas >100.

DAPI StainingCircular TC-treated cover slips were placed in the bottom of

each well of a 24-well plate. A total of 1 × 105 of DU-145, PC-3,or LNCaP cells were seeded in each well and incubated for 24 h.At the end of 24 h, the growth medium was removed and re-placed with fresh DMEM or DMEM supplemented with 20 µMof the appropriate vitamin E homolog or synthetic derivative.Following an incubation of 24 h, the medium was removed, andthe cells were incubated for 5 min with DAPI stain (1 µg/ml).Each cover slip was then removed, covered with glycerol, andthe morphology of the cells’ nuclei was observed using a flu-orescence microscope (Leica, Wetzlar, Germany) at excitationwavelength 350 nm.

Caspase AssaysCaspase-8 activity assay. Caspase-8 activity assays were

performed using the Caspase 8/FLICE Fluorometric Proteaseassay kit (Biosource International Inc., Camarillo, CA). Thecells in 10 mm2 plates were treated with the vitamin E natu-ral compounds, synthetic derivatives, or etoposide as describedin the figure legends, in the presence or absence of 20 µMof the caspase inhibitor z.VAD.fmk (Calbiochem, Darmstadt,Germany). At the end of the incubation period, the cells werewashed with PBS, the PBS was removed, and the cells werelysed with 150 µl of cell lysis buffer (provided in Caspase 8Apotarget kit) per plate. The plates were placed on ice, and cellswere allowed to lyse for 10 min. The cells were mixed in wellwith pipette to ensure lysis. The lysate was transferred into afresh eppendorf, and the latter was spun for 5 min at 11,000 rpmat 4◦C. The supernatant was removed and placed into a neweppendorf, leaving the pellet behind. The lysates were storedat –70◦C until needed for the performance of the assay. Whenneeded, the samples were thawed and used in a Bradford assayto determine protein concentration. Subsequently, 120 µg ofeach sample was added per well of a 96-well plate. The DTTwas then added to the provided reaction buffer mix (DTT finalconcentration: 10 mM). Fifty µl of reaction buffer mix was thenadded per sample in a well; 5 µl of 1 mM of the caspase substrateIETD-AFC was then added per well and incubated at 37◦C for1 to 2 h. At the end of the incubation period, the 96-well platewas read on a fluorescence reader (excitation 400 nm, emission505, slit width 15).

Caspase-3 activity assay. Caspase-3 activity assays wereperformed using the Kamiya caspase 3 fluorogenic substrate(Kamiya, Seattle, WA). The cells in 10 mm2 plates were treatedwith the vitamin E natural compounds or synthetic derivatives asdescribed in the figure legends in the presence or absence of thecaspase inhibitor z.VAD.fmk (Calbiochem). At the end of theincubation period, the cells were washed with PBS, the PBS wasremoved, and the cells were lysed with 150 µl of cell lysis buffer(provided in Caspase 8 Apotarget kit) per plate. The plates wereplaced on ice, and cells were allowed to lyse for 10 min. Thecells were mixed in well with pipette to ensure lysis. The lysatewas transferred into a fresh eppendorf, and the latter was spun

868 C. CONSTANTINOU ET AL.

for 5 min at 11,000 rpm at 4◦C. The supernatant was removedand placed into a new eppendorf, leaving the pellet behind. Thelysates were stored at –70◦C until needed for the performanceof the assay. When needed, the samples were thawed and usedin a Bradford assay to determine protein concentration. Sub-sequently, 40 to 80 µg of each sample was added per well ofa 96-well plate. The Assay Buffer Mix was then prepared byadding 1 ml Caspase Assay Fluorometric Buffer (2.4 g Hepes,20 g sucrose, 0.2 g CHAPS, dissolved in 200 ml water and pH7.4 with NaOH) per 1 µl Caspase 3 substrate (AC-DEVD-AFC,used at a final concentration of 2.5 µM) per 10 µl of 1 M DTT(final concentration 10 mM). Two hundred µl of assay bufferwas then added per sample in a well, and the plate was coveredin foil and incubated at 37◦C for 1 to 3 h. At the end of the timepoints, the plate was read on plate reader (excitation 400 nm,emission 505, slit width 15).

Statistical AnalysesStatistical analyses were conducted using Prism software

version 5.0 (Graphpad, San Diego, CA).

RESULTS

Preparation of Vitamin E Homologs and SyntheticDerivatives

The 8 homologs of vitamin E (α-TOC, β-TOC, γ -TOC, δ-TOC and α-TT, β-TT, γ -TT, δ-TT) and their succinate syntheticderivatives (α-TOS, β-TOS, γ -TOS, δ-TOS and α-TS, β-TS,γ -TS, δ-TS; Table 1, Fig. 1) were prepared as described in Mate-rials and Methods for a subsequent evaluation and comparisonof their death promoting potencies.

Antiproliferative Effects of Vitamin E Homologs andSynthetic Derivatives

Using the crystal violet cell proliferation assay, cell viabilitywas monitored with a range of concentrations between 0 and100 µM for the 8 vitamin E homologs and 8 synthetic deriva-tives shown on Table 1 and Fig. 1 as described in Materialsand Methods. The results were used to determine the IC50 val-ues (Table 2). Tocotrienols were generally more effective thantocopherols. δ-TT was the most effective of the natural vitaminE homologs, with IC50 values of 20 µM, 25 µM, and 11 µMfollowing a 72-h incubation in DU145, PC3, and LNCaP cells,

TABLE 2The IC50 values (µM) of vitamin E homologs and synthetic derivatives in DU145, PC3, and LNCaP cells

DU145 PC3 LNCaP

24 h 48 h 72 h 24 h 48 h 72 h 24 h 48 h 72 h

Tocopherolsα-TOC >100 >100 >100 >100 >100 >100 >100 >100 >100β-TOC >100 >100 >100 >100 >100 >100 >100 >100 >100γ -TOC >100 >100 >100 >100 >100 >100 >100 >100 >100δ-TOC >100 >100 >100 >100 89 89 62 43 56

Tocopherol succinatesα-TOS >100 >100 >100 >100 >100 57 17 40 60β-TOS >100 >100 68 >100 >100 91 >100 >100 95γ -TOS >100 >100 >100 >100 >100 >100 8 7 38δ-TOS 60 51 56 64 22 37 10 12 16

Tocotrienolsα-TT >100 >100 >100 >100 >100 >100 >100 >100 >100β-TT 60 53 52 59 48 44 70 52 29γ -TT 34 30 35 33 24 23 21 21 20δ-TT 20 27 20 32 25 25 26 19 11

Tocotrienol succinatesα-TS >100 >100 >100 >100 >100 >100 >100 >100 >100β-TS 60 56 47 72 68 55 >100 >100 86γ -TS 36 19 36 41 31 28 21 15 11δ-TS 20 11 15 20 18 28 10 9 8

DU145, PC3, and LNCaP cells were incubated in 96 well plates with the vitamin E homologs and synthetic derivatives (range of concentration0–100 µM) for a period of 24–72 hours. At the end of the incubation period the cells were stained with crystal violet and the IC50 values weredetermined as described in the Materials and Methods section. The values are the means of three determinations from three independentexperiments. Error <10%.

VITAMIN E INDUCED CASPASE INDEPENDENT CELL DEATH 869

FIG. 2. Examination of morphological features of apoptosis with DAPI stain-ing. A: DAPI staining of DU145 cells either left untreated (Control) or treatedfor 24 h with 20 µM γ -TT, δ-TT, or δ-TS. B: DAPI staining of DU145, PC3,or LNCaP cells left untreated (Control) or treated for 48 h with 20 µM δ-TS.The images are representative of three different experiments. Arrows indicateapoptotic bodies.

respectively. The results of our study propose that the antipro-liferative ability of the tocotrienols increases as follows: α-TT< β-TT< γ -TT < δ-TT. Of the 8 homologs and 8 syntheticderivatives examined, δ-TS was the most effective, with IC50

values of 15 µM, 28 µM, and 8 µM following a 72-h incu-bation in DU145, PC3, and LNCaP cells, respectively. δ-TSwas extremely more potent than the popular α-TOS, which wefound ineffective in the three cell systems examined. The AR+LNCaP cell line was the most sensitive of the three cell linestested, suggesting that the AR may play a role in this process(Table 2).

In order to investigate whether the antiproliferative effects in-duced by the most effective vitamin E homologs γ -TT and δ-TTand the most potent synthetic derivative δ-TS could be attributedto apoptosis, the cells were treated with these compounds andtheir morphology was examined with DAPI staining. All threecompounds induced morphological features indicative of apop-tosis in DU145 cells since apoptotic bodies were clearly visiblewithin 24 h of incubation (Fig. 2A). Similar cell morphologywas visible in PC3 and LNCaP cells treated with δ-TS (Fig. 2B),γ -TT, δ-TT, or γ -TS (data not shown).

�-TT and �-TT Induce Synergistic Cell Death in DU145and PC3 Cells

To examine a possible synergy in the induction of cell deathby the tocotrienols, the three cell lines were incubated with ei-ther β-TT, γ -TT, or δ-TT or with one of the three tocotrienolcombinations (β-TT +γ -TT, β-TT +δ-TT, or γ -TT +δ-TT) atequimolar concentrations for 72 h. α-TT was not included in thisinvestigation since this homolog was incapable of inducing celldeath in any of the three cell lines at lower than 100 µM con-centration (Table 2). Only the combination of γ -TT +δ-TT pro-duced synergistic induction of cell death in the cell lines DU145and PC3, whereas the two other combinations (β-TT +γ -TT andβ-TT +δ-TT) did not show any synergistic effects (Table 3).

�-TT and �-TT Activate Caspase-8 and Caspase-3 inLNCaP Cells

Cell death induced by γ -TT and δ-TT was evident within 24h of treatment in all three cell lines (Table 2). Caspase-8 activitywas not affected by γ -TT in any of the three cell lines (Fig.3A). However, this agent produced a small increase in caspase-3 activity only in LNCaP cells after 72 h of treatment (Fig.3B). In LNCaP cells, caspase-8 activity was increased within24 h in response to δ-TT treatment (Fig. 3A), whereas caspase-3 activity was delayed, reaching its maximum after 72 h oftreatment (Figure 3B). γ -TT and δ-TT did not induce significantactivation of the two caspases in DU145 and PC3 cells.

In order to investigate whether the synergistic cell death in-duced by γ -TT and δ-TT is due to increased caspase activities,we compared the effects of the individual compounds (oncaspase-8 and caspase-3) to those of their combinations. In-terestingly, the combination of γ -TT +δ-TT did not cause fur-ther increases in any of the two caspase activities. This obser-vation held even in LNCaP cells where the individual agentsinduced these caspases (Fig. 3A and 3B). These results showthat the synergy produced by the two homologs in the inductionof cell death is not due to increased caspase activities, sug-gesting that the combinations may possibly activate CI-PCDpathways.

�-TS Does Not Activate Caspase-8 and Caspase-3 inDU145, PC3, or LNCaP Cells

The caspase-8 and caspase-3 activities of γ -TT and δ-TTwere compared to those produced by γ -TS and δ-TS, respec-tively. This comparison was made to determine if the additionof succinic acid on tocotrienols improves their proapoptotic po-tency. γ -TS produced higher levels of caspase-8 activity thanγ -TT in DU145 (72 h) and LNCaP (24 h) cells (Table 4). Withrespect to caspase-3, γ -TT was the most effective, producing a40% increase (over the control levels) after 72 h of treatment,whereas γ -TS was ineffective. Similarly, δ-TT effectively ac-tivated caspase-3 in LNCaP cells, producing almost 200% in-crease after 72 h of treatment in LNCaP cells. δ-TS was inef-fective, indicating that the addition of succinate abolishes thecaspase-3 activating effects of δ-TT.

870 C. CONSTANTINOU ET AL.

TABLE 3Examination of the ability of tocotrienols to work synergistically in inducing cell death in DU145, PC3, and LNCaP cells

Cell line Tocotrienol Combination vs Statistical significance (p value)

DU145 β-TT+γ -TT (IC50 = 39 µM) β-TT (IC50 = 38 µM)γ -TT (IC50 = 25 µM)

ns

β-TT+δ-TT (IC50 = 36 µM) β-TT (IC50 = 38 µM)δ-TT (IC50 = 21 µM)

ns

γ -TT+δ-TT (IC50 = 15 µM) γ -TT (IC50 = 25 µM)δ-TT (IC50 = 21 µM)

0.01740.014

PC3 β-TT+γ -TT (IC50 = 24 µM) β-TT(IC50 = 37 µM)γ -TT aC,n = 22 µM)

ns

β-TT+δ-TT (IC50 = 41 µM) β-TT (IC50 = 37 µM)δ-TT (IC50 = 36 µM)

ns

γ -TT+δ-TT (IC50 = 19 µM) γ -TT IC50 = 22 µM)δ-TT (IC50 = 36 µM)

0.0002 <.0001

LNCaP β-TT+γ -TT (IC50 = 24 µM) β-TT(IC50 = 18 µM)γ -TT aC,n = 15 µM)

ns

β-TT+δ-TT (IC50 = 26 µM) β-TT(IC50 = 18 µM)δ-TT (IC50 = 17 µM)

ns

γ -TT+δ-TT (IC50 = 18 µM) γ -TT (IC50 = 15 µM)δ-TT (IC50 = 17 µM)

ns

DU145, PC3, or LNCaP cells were incubated with 20 µM of each of the three tocotrienols β-TT, γ -TT, or δ-TT or with equmolorcombinations of two tocotrienols (10 µM β-TT+10 µM γ -TT, 10 µM β-TT + 10 µM δ-TT or 10 µM γ -TT+10 µM δ-TT) for 72 h. At theend of the incubation period, the cells were stained with crystal violet and the IC50 values were determined as described in the Materials andMethods section. The IC50 values obtained from each of the three possible combinations (β-TT+γ -TT, β-TT+δ-TT, or γ -TT+δ-TT) weredirectly compared to those obtained from each of the tocotrienols used in the combination to determine synergy. The statistical significancewas determined by unpaired t-test. When p value >0.05, non-significant (ns) difference.

TABLE 4Examination of caspase-8 and -3 activities induced by γ -TT, δ-TT, γ -TS, δ-TS, and etoposide

DU145 PC3 LNCap

Treatment 24 h 48 h 72 h 24 h 48 h 72 h 24 h 48 h 72 h

Caspase-8 activity (% increase over control)γ -TT — — — 2.7 ± 1.6 — — — — —γ -TS — — 41.6 ± 5 — — 43.9 ± 2.1 — —δ-TT — — — 23.7 ± 13.7 — — 41.2 ± 13 3.9 ± 2.4 —δ-TS — — — — — — 57.7 ± 6.5 21.4 ± 12.4 —Etoposide 170 ± 13 7.7 ± 5.6 35.2 ± 23.4

Caspase-3 activity (% increase over control)γ -TT — — — 9.1 ± 5.2 2.5 ± 1.6 — — 10.6 ± 5.4 40 ± 15γ -TS — 9.6 ± 2.9 6.4 ± 4.2 4.8 ± 2.7 — — 16.1 ± 9.5 — —δ-TT — — — 12.2 ± 3.4 4.5 ± 3 — — 16.3 ± 5.2 198 ± 9.2δ-TS — — — 4.6 ± 2.3 — — — — —Etoposide 485.7 ± 2.9 200 ± 43 399.2 ± 45

DU145, PC3, and LNCaP cells were incubated with 20 µM of γ -TT, δ-TT, γ -TS, δ-TS for 24 h–72 h or 100 µg/ml etoposide for 24 h. Atthe end of each time point the cells were collected, lysed, and assayed for caspase-8 or caspase-3 activities as described in the Materials andMethods section. The values are the means expressed as a percentage increase over untreated control ± SEMs of three different determinations.(−) indicates no increase from control value.

VITAMIN E INDUCED CASPASE INDEPENDENT CELL DEATH 871

FIG. 3. Investigation of induction of caspase-8 and caspase-3 activities by γ -TT and δ-TT. DU145, PC3, and LNCaP cells were incubated with 20 µM of γ -TT,20 µM of δ-TT, or with the equimolar combination of 10 µM γ -TT + 10 µM of δ-TT for 24 h to 72 h. At the end of each time point, the cells were collected,lysed, and assayed for A: caspase-8 and B: caspase-3 activities as described in Materials and Methods. The values are expressed as a percentage increase overuntreated control and are the means ± SEMs of three different determinations.

The levels of caspase activities induced by the DNA dam-aging agent etoposide were also determined for comparison.Low levels of the initiator, caspase-8, and very high levels ofthe effector, caspase-3, were visible following a 24-h incubationwith the agent (Table 4). The high levels of caspase-3 activityinduced by etoposide compared to the low levels induced byvitamin E homologs and their synthetic derivatives suggests theutilization of CI-PCD pathways by the latter.

�-TT, �-TT, �-TS, and �-TS Activate Pathways of CI-PCDDU145, PC3, and LNCaP cells were incubated with γ -TT, δ-

TT, γ -TS, or δ-TS in the presence or absence of the panspecificcaspase inhibitor z.VAD.fmk. This experiment was performedto determine if caspase activity is required for the decreasein viability introduced by these agents. Although z.VAD.fmkeffectively inhibited 100% of etoposide-induced caspase-3 ac-tivity (Fig. 4B) and caspase-8 activity (data not shown) in allthree cell lines, this panspecific caspase inhibitor was inef-

fective against δ-TT and δ-TS. In fact, the caspase inhibitor,instead of increasing cell survival when introduced togetherwith these tocotrienols, in some occasions rather produced anadditional decrease in viability. The effects of γ -TT (in de-creasing cell viability) were only partially prevented by thecaspase inhibitor in all three cell lines. γ -TS produced mixedresults; in DU-145 cells, the combination with the caspase in-hibitor produced an additional small decrease in cell viabilitycompared to γ -TS alone, whereas in PC-3 and LNCaP cells,the caspase inhibitor allowed a small increase in cell viability(Fig. 4A).

The observation that z.VAD.fmk did not cause a recoveryin the decreased cell viability produced by δ-TS in DU145,PC3, and LNCaP cells is consistent with the inability of thissynthetic derivative to activate caspase-8 and caspase-3 (Table4). A general lack of recovery in cell viability was also observedwhen lower or higher concentrations of γ -TT, δ-TT, γ -TS, andδ-TS were tested in the presence or absence of the caspaseinhibitor (data not shown).

872 C. CONSTANTINOU ET AL.

FIG. 4. Investigation of requirement of caspase-dependent cell death by γ -TT, δ-TT, γ -TS and δ-TS. A: DU145, PC3, and LNCaP cells were incubated with 40µM of γ -TT, 40 µM of γ -TS, 40 µM of δ-TT or 40 µM of δ-TS in the presence or absence of 20 µM of z.VAD.fmk in 96-well plates for 48 h. At the end ofthe incubation period, the cells were stained with crystal violet and % cell viability was determined as described in Materials and Methods. The values are themeans ± SEMs of three different determinations. B: DU145, PC3, and LNCaP cells were incubated with 100 µg/ml of etoposide in the presence or absence of20 µM of z.VAD.fmk and incubated for 24 h. At the end of the incubation, the cells were collected and lysed and assayed for caspase-3 activity as described inMaterials and Methods and expressed as a percentage increase over untreated control. The values are the means ± SEMs of three different determinations.

DISCUSSIONMost research in the past few years has focused on produc-

ing synthetic derivatives of α-TOC. Some common examplesinclude α-TOS and α-TEA (α-tocopheryl ether linked aceticacid), whereas the use of other natural homologs of vitaminE as the basis for production of new synthetic molecules hasbeen scarce (23–25,30). Nevertheless, benefits could be ob-tained from the design of molecules based on homologs otherthan α-TOC. The fact that tocotrienols are more effective thantocopherols suggests the importance of unsaturation on the tailin the induction of apoptosis. The four tocotrienols differ in theposition and number of methyl groups, and our observations and

several reports in the literature have suggested that the smallerthe number of methyl groups on the chroman head the greater theantiproliferative potency (δ-TT> γ -TT> β-TT> α-TT) (19).Furthermore, the addition of succinic acid on the hydroxyl groupof α-TOC and its conversion to α-TOS increases its proapoptoticpotency, suggesting a possible role for the addition of succinicacid on other homologs of vitamin E (23–25).

This study presents the first time that the efficacy and mecha-nism of apoptosis modulated by the 8 vitamin E homologs havebeen evaluated in a single study. Furthermore, it is also the firsttime that the tocopheryl and tocotrienyl—succinate derivativesof the 8 homologs—have been synthesized and their apoptotic

VITAMIN E INDUCED CASPASE INDEPENDENT CELL DEATH 873

properties have been determined and compared to those of vita-min E homologs. Our results have shown that δ-TT is the mostpotent homolog, and δ-TS is the most potent synthetic deriva-tive of vitamin E in the induction of cell death in prostate cancercells (Table 2). We propose that δ-TS benefits from having 1)at its basis δ-TT, the most potent of all 8 homologs of vitaminE; and 2) its hydroxyl group being modified by the additionof succinic acid. Interestingly, the increased death promotingpotency of δ-TS compared to δ-TT could not be attributed tothe induction of higher levels of caspase activity. In fact, theconversion of δ-TT to δ-TS completely eliminated the activa-tion of caspase-3 observed in LNCaP cells by the former agent,strongly suggesting that the higher killing potency of δ-TS ismediated via CI-PCD pathways (Table 4).

It has been previously proposed that γ -TT and δ-TT are ca-pable of inducing caspase activation and therefore CD-PCD inseveral cell lines (17,31). For example, γ -TT has been reportedto activate caspase-8 and caspase-3 in neoplastic mammary ep-ithelial cells and caspase-8, caspase-9, and caspase-3 in Hep3Bhepatoma cells (17,31). Contrary to these reports, in this study,we found that neither γ -TT nor δ-TT produce significant acti-vation of caspase-8 and caspase-3 in DU145 and PC3 cells atconcentrations and time points that kill 50% of the cells. Inter-estingly, under the same treatment conditions in LNCaP cells,these agents are effective, producing substantial activation ofcaspase-8 and/or caspase-3 (Fig. 3, Table 4). The results sug-gest that the tocotrienol-induced caspase activation is selective.The caspase activation by γ -TT and δ-TT only in AR+ LNCaPbut not in AR– DU145 and PC3 cells may suggest a possibleinvolvement of the androgen receptor. However, to prove thisnotion, further experiments should be performed in isogenic celllines expressing or not expressing the receptor. Alternatively, thedifferences in sensitivity of the three prostate cancer cells maybe due to differences in the biovailability and transport pro-teins of vitamin E, for example, α-tocopherol associated protein(TAP), scavenger receptor class B type I, α -tocopherol transferprotein (TTP), and ATP-binding cassette transporter A1 (12).

The results of this study provide strong support to two pre-vious reports that have suggested the involvement of caspase-independent pathways in the induction of apoptosis by vitaminE homologs (32,33) and help identify CI-PCD as the main modeof cell killing by γ -TT and δ-TT. Jiang et al. (32) reported thatalthough γ -TOC activates caspase-9, caspase-3 and caspase-7in LNCaP cells, apoptosis could not be completely reversed bythe pancaspase inhibitor (z.VAD.fmk), indicating the involve-ment of an alternative caspase-independent pathway. A similarobservation was made in MDA-MB-231 cells treated with γ -TTin which morphological features of apoptosis became evidentaccompanied by mitochondrial disruption and release of cy-tochrome c in the absence of poly-(ADP-ribose)-polymerasecleavage, suggesting the lack of involvement of caspases in theinduction of apoptosis (33). These results are consistent with ourhypothesis that the tocotrienols γ -TT and δ-TT kill prostate can-cer cells by activating caspase-independent pathways of PCD.

Of special interest is the observation that γ -TT and δ-TTinduce cell death synergistically only in AR– PC-3 and DU-145cell lines and in the absence of increased caspase activity (Ta-ble 3, Fig. 3). One possible explanation of this observation isthat these compounds may be activating two different caspase-independent pathways that synergistically produce higher levelsof cell death. The synergistic death-promoting potencies of thetwo tocotrienols could find practical applications if the combi-nation of γ -TT and δ-TT were used in the treatment of andro-gen independent prostate cancer. The activation of synergisticpathways of cell death by the two compounds could ensurebetter eradication of tumor cells and lower levels of tumor resis-tance often associated with classical chemotherapy. Also, thiscombination would avoid the problem of hydrolyzability oftenassociated with synthetic derivatives such as α-TOS (23–25).

Similar to the synergistic effects of γ -TT and δ-TT in the in-duction of cell death (Table 3), the increased levels of cell deathinduced by the synthetic derivative δ-TS relative to δ-TT (Table2) could also not be attributed to increased caspase activity (Fig.3, Table 4). Furthermore, the inability of the caspase inhibitorz.VAD.fmk to recover the loss of cell viability induced by γ -TT, δ-TT, γ -TS, and δ-TS (Fig. 4) provides additional supportto our hypothesis that the main mode of apoptotic cell deathinduced by vitamin E homologs and their synthetic derivativesis mediated via a caspase-independent pathway (CI-PCD). Al-though the exact pathways of CI-PCD have not been unraveled,it is known that the mitochondrion is the main organelle or-chestrating the series of events leading to CI-PCD (34,35). Theproapoptotic proteins Bax and Bid, the cathepsins (released bylysosomes), the calpains and mainly AIF have been shown tohave a central role in the induction of CI-PCD (36–40). Eventhough at present, the exact pathways activated by vitamin Ehomologs and their derivatives in prostate cancer cells remainunknown, our data strongly suggest that these agents triggerCI-PCD dominant pathways.

The basis of future research should be to develop vitaminE derivatives based on the most potent tocotrienols instead ofthe inactive homolog α-TOC. One such promising compoundis the synthetic derivative δ-TS identified in this study. Ourresults suggest that δ-TS is more effective than the popular α-TOS against prostate cancer. Consequently, the identificationof the signalling pathways and in particular, those involvedin CI-PCD regulated by δ-TS, is of vital significance. Futureexperiments should focus on investigating the possible syner-gistic potency of the effective tocotrienols or tocotrienol syn-thetic derivatives such as δ-TS in combinations with chemother-apeutic agents commonly used in the clinic. In the future,we should also progress with a more rational design of evenbetter synthetic derivatives of vitamin E. The design of anotherδ-TT-based molecule should address the issue of hydrolyzabil-ity. The production of a nonhydrolyzable molecule based onδ-TT is expected to have an even greater antitumorigenic po-tency. The ability of γ -TT, δ-TT, and δ-TS documented in thisstudy to activate CI-PCD may prove to be useful in the adjuvant

874 C. CONSTANTINOU ET AL.

treatment of prostate cancers that are resistant to traditionallyused chemotherapeutic agents whose mode of action involvesthe activation of caspases.

ACKNOWLEDGMENTSResearch in our laboratories is supported by the Cyprus

Research Promotion Foundation (grant-KINHTIKOTHTA/0506/06, grant ϒ�EIA /0104/06, and grant SYNERGIA/0505/20).

REFERENCES1. Jemal A, Siegel R, Ward E, Murra T, Xu J, et al.: Cancer statistics, 2006.

CA Cancer J Clin 56, 106–130, 2006.2. Shun MC, Yu W, Gapor A, Parsons R, Atkinson J, et al.: Pro-apoptotic

mechanisms of action of a novel vitamin E analog (α-TEA) and a naturallyoccurring form of vitamin E (δ-tocotrienol) in MDA-MB-435 human breastcancer cells. Nutr Cancer 48, 95–105, 2004.

3. Limpens J, Schroder FH, de Ridder CM, Wildhagen MF, Obermuller-JevivUC, et al.: Combined lycopene and vitamin E treatment suppresses thegrowth of PC-346C human prostate cancer cells in nude mice. J Nutr 136,1287–1293, 2006.

4. Campbell SE, Stone WL, Lee S, Whaley S, Yang H, et al.: Comparativeeffects of RRR-α and RRR-γ -tocopherol on proliferation and apoptosis inhuman colon cancer cell lines. BMC Cancer 6, 13, 2006.

5. Min J, Guo J, Zhao F, and Cai D: Effect of apoptosis induced by differentvitamin E homologous analogues in human hepatoma cells (HepG2). WeiSheng Yan Jiu 32, 343–345, 2003.

6. McIntyre BS, Briski KP, Gapor A, and Sylvester PW: Anti-proliferativeand apoptotic effects of tocopherols and tocotrienols on preneoplastic andneoplastic mouse mammary epithelial cells. Proc Soc Exp Biol Med 224,292–301, 2000.

7. Pastori M, Pfander H, and Boscoboinik D: Lycopene in association withalpha-tocopherol inhibits at physiological concentrations proliferation ofprostate carcinoma cells. Biochem Biophys Res Commun 250, 582–585,1998.

8. Burton GW and Traber MG: Vitamin E: antioxidant activity, biokinetics,and bioavailability. Annu Rev Nutr 10, 357–382, 1990.

9. Kamal-Eldin A and Appelqvist LA: The chemistry and antioxidant proper-ties of tocopherols and tocotrienols. Lipids 31, 671–701, 1996.

10. Boscoboinik D, Chatelain E, Bartoli GM, and Azzi A: Molecular basis ofalpha-tocopherol inhibition of smooth muscle cell proliferation in vitro.EXS 62, 164–177, 1992.

11. Speek A, Schrijver J, and Schreurs W: Vitamin E composition of someseed oils as determined by high-performance liquid chromatography withfluorescence detection. J Food Sci 50, 121–124, 1985.

12. Tan B: Palm carotenoids, tocopherols, and tocotrienols. J Am Oil Chem Soc66, 770–776, 1989.

13. Yu W, Simmons-Menchaca M, Gapor A, Sanders BG, and Kline K: In-duction of apoptosis in human breast cancer cells by tocopherols and to-cotrienols. Nutr Cancer 33, 26–32, 1999.

14. Min J, Guo J, Zhao F, and Cai D: Effect of apoptosis induced by differentvitamin E homologous analogues in human hepatoma cells (HepG2). WeiSheng Yan Jiu 32, 343–345, 2003.

15. Srivastava JK and Gupta S: Tocotrienol-rich fraction of palm oil inducescell cycle arrest and apoptosis selectively in human prostate cancer cells.Biochem Biophys Res Commun 346, 447–453, 2006.

16. Sylvester PW and Shah S: Intracellular mechanisms mediating tocotrienol-induced apoptosis in neoplastic mammary epithelial cells. Asia Pac J ClinNutr 14, 366–373, 2005.

17. Sakai M, Okabe M, Tachibana H, and Yamada K: Apoptosis induction bygamma-tocotrienol in human hepatoma Hep3B cells. J Nutr Biochem 17,672–676, 2006.

18. Kumar KS, Raghavan M, Hieber K, Ege C, Mog S, et al.: Pref-erential radiation sensitization of prostate cancer in nude mice by

nutraceutical antioxidant gamma-tocotrienol. Life Sci 78, 2099–2104,2006.

19. He L, Mo H, Hadisusilo S, Qureshi AA, and Elson CE: Isoprenoids suppressthe growth of murine B16 melanomas in vitro and in vivo. J Nutr 127, 668–674, 1997.

20. Tan B: Appropriate spectrum vitamin E and new perspectives on desmethyltocopherols and tocotrienols. JANA 8, 35–43, 2005.

21. Couladouros E, Moutsos V, Lampropoulou M, Little J, and Hyatt J: J OrgChem 72, 6735–6741, 2007.

22. Birringer M, EyTina JH, Salvatore BA, and Neuzil J: Vitamin E analoguesas inducers of apoptosis: structure–function relation. Br J Cancer 88, 1948–1955, 2003.

23. Lawson KA, Anderson K, Simmons-Menchaca M, Atkinson J, Sun L,et al.: Novel vitamin E analogue decreases syngeneic mouse mammarytumor burden and reduces lung metastasis. Mol Cancer Ther 2, 437–444,2003.

24. Fariss MW, Nicholls-Grzemski FA, Tirmenstein MA, and Zhang JG: En-hanced antioxidant and cytoprotective abilities of vitamin E succinate isassociated with a rapid uptake advantage in rat hepatocytes and mitochon-dria. Free Radic Biol Med 31, 530–541, 2001.

25. Yu W, Liao QY, Hantash FM, Sanders BG, and Kline K: Activation ofextracellular signal-regulated kinase and c-Jun-NH2-terminal kinase butnot p38 mitogen-activated protein kinases is required for RRR-α-tocopherylsuccinate-induced apoptosis of human breast cancer cells. Cancer Res 61,6569–6576, 2001.

26. Vraka PS, Drouza C, Rikkou MP, Odysseos AD, and Keramidas AD:Synthesis and study of the cancer cell growth inhibitory properties ofalpha-, gamma-tocopheryl and gamma-tocotrienyl 2-phenylselenyl succi-nates. Bioorg Med Chem 14, 2684–2696, 2006.

27. Cawley J and Stern M: U.S. Patent 2,680,749 (to Eastman Kodak), 1951.28. Goh S, Hew N, Ong A, Choo Y, and Brumby S: Tocotrienols from palm oil:

electron spin resonance spectra of tocotrienoxyl radicals. J Am Oil ChemSoc 67, 250–254, 1990.

29. Hyatt J: U.S. Patent 5,869,704 (to Eastman Chem. Co.), 1999.30. Jia L, Yu W, Wang P, Sanders BG, Kline K: In vivo and in vitro studies of

anticancer actions of α-TEA for human prostate cancer cells. Prostate 68,849–860, 2008.

31. Sylvester PW, Shah SJ, and Samant GV: Intracellular signaling mechanismsmediating the antiproliferative and apoptotic effects of gamma-tocotrienolin neoplastic mammary epithelial cells. J Plant Physiol 162, 803–810, 2005.

32. Jiang Q, Wong J, and Ames BN: Gamma-tocopherol induces apoptosis inandrogen-responsive LNCaP prostate cancer cells via caspase-dependentand independent mechanisms. Ann NY Acad Sci 1031, 399–400, 2004.

33. Takahashi K and Loo G: Disruption of mitochondria during tocotrienol-induced apoptosis in MDA-MB-231 human breast cancer cells. BiochemPharmacol 67, 315–324, 2004.

34. Mathiasen IS, Lademann U, and Jaattela M: Apoptosis induced by vitaminD compounds in breast cancer cells is inhibited by Bcl-2 but does notinvolve known caspases or p53. Cancer Res 59, 4848–4856, 1999.

35. Rudolf E, Rudolf K, and Cervinka M: Selenium activates p53 and p38pathways and induces caspase-independent cell death in cervical cancercells. Cell Biol Toxicol. 24, 123–141, 2008.

36. Okuno S, Shimizu S, Ito T, Nomura M, and Hamada E: Bcl-2 preventscaspase-independent cell death. J Biol Chem 273, 34272–34277, 1998.

37. Xiang J, Chao DT, and Korsmeyer SJ: BAX-induced cell death may notrequire interleukin 1 beta-converting enzyme-like proteases. Proc Natl AcadSci USA 93, 14559–14563, 1996.

38. Brunk UT, Neuzil J, and Eaton J: Lysosomal involvement in apoptosis.Redox Report 6, 91–97, 2001.

39. Neuzil J, Zhao M, Ostermann G, Sticha M, and Gellert N: Alpha-tocopherylsuccinate, an agent with in vivo anti-tumor activity, induces apoptosis bycausing lysosomal instability. Biochem J 362, 709–715, 2002.

40. Lorenzo HK and Susin SA: Therapeutic potential of AIF-mediated caspase-independent programmed cell death. Drug Resist Updates 10, 235–255,2007.