at SciVerse ScienceDirect

Biochimie 95 (2013) 496e503

Contents lists available

Biochimie

journal homepage: www.elsevier .com/locate/b iochi

Research paper

Synthesis and assessment of the relative toxicity of the oxidisedderivatives of campesterol and dihydrobrassicasterol in U937and HepG2 cells

Yvonne O’Callaghan a, Olivia Kenny a, Niamh M. O’Connell b, Anita R. Maguire c,Florence O. McCarthy b, Nora M. O’Brien a,*

a School of Food and Nutritional Sciences, University College Cork, Cork, IrelandbDepartment of Chemistry, Analytical and Biological Chemistry Research Facility, University College Cork, Cork, IrelandcDepartment of Chemistry and School of Pharmacy, Analytical and Biological Chemistry Research Facility, University College Cork, Cork, Ireland

a r t i c l e i n f o

Article history:Received 19 December 2011Accepted 18 April 2012Available online 26 April 2012

Keywords:CampesterolDihydrobrassicasterolOxidation productCell cultureCytotoxicity

* Corresponding author. Tel.: þ353 21 4902884.E-mail address: nob@ucc.ie (N.M. O’Brien).

0300-9084/$ e see front matter � 2012 Elsevier Masdoi:10.1016/j.biochi.2012.04.019

a b s t r a c t

The cytotoxic effects of the oxidised derivatives of the phytosterols, stigmasterol and b-sitosterol, havepreviously been shown to be similar but less potent than those of the equivalent cholesterol oxides in theU937 cell line. The objective of the present study was to compare the cytotoxic effects of the oxidisedderivatives of synthetic mixtures of campesterol and dihydrobrassicasterol in both the U937 and HepG2cell lines. The parent compounds consisted of a campesterol: dihydrobrassicasterol mix at a ratio of 2:1(2CMP:1DHB) and a dihydrobrassicasterol:campesterol mix at a ratio of 3:1 (3DHB:1CMP). The2CMP:1DBH oxides were more cytotoxic in the U937 cells than the 3DBH:1CMP oxides but the differencein cytotoxicity was less marked in the HepG2 cells. The order of toxicity of the individual oxidationproducts was found to be similar to that previously observed for cholesterol, b-sitosterol and stigmas-terol oxidation products in the U937 cell line. There was an increase in apoptotic nuclei in U937 cellsincubated with the 7-keto and 7b-OH derivatives of both 2CMP:1DHB and 3DHB:1CMP and also in thepresence of 3DHB:1CMP-3b,5a,6b-triol and 2CMP:1DHB-5b,6b-epoxide. An additional oxidation productsynthesised from 2CMP:1DHB, 5,6,22,23-diepoxycampestane, was cytotoxic but did not induceapoptosis. These results signify the importance of campesterol oxides in the overall paradigm ofphytosterol oxide cytotoxicity.

� 2012 Elsevier Masson SAS. All rights reserved.

1. Introduction

Campesterol is one of the most prevalent phytosterols found infoods, along with b-sitosterol and stigmasterol [1]. Although b-sitosterol is present in thehumandiet at higher concentrations, it hasbeen demonstrated that campesterol ismore readily absorbed by ratjejunal villus cells and brush bordermembranes [2] and is present athigher concentrations in human serum, both at baseline and alsofollowing dietary supplementation with sterol enriched spreads[3,4]. In contrast to the heavily promoted health benefits of dietaryphytosterol supplementation, a number of groups have identifiedphytosterols with adverse health effects: induction of endothelialdysfunction and increased size of ischaemic stroke [5]; inhibition ofcell growth [6]; aggressive vascular disease in sitosterolaemicpatients [7]. Themethyl groupat theC24positionof campesterolmay

son SAS. All rights reserved.

exist as either of two epimers, 24a-methylcholesterol (campesterol)or 24b-methylcholesterol (dihydrobrassicasterol) (Fig. 1). It is esti-mated that campesterol in terrestrial plants is comprised ofapproximately 65% 24a-methylcholesterol and 35% 24b-methyl-cholesterol, however, certain species of phytoplankton have beenfound to contain campesterol comprised entirely of one or other ofthe two epimers [8]. The epimers are difficult to isolate and studiesinvestigating thebiological activityof campesterolhavenotgenerallydistinguished between the two forms quoting instead a total cam-pesterol content.

Similar to cholesterol, campesterol has a double bond at the C5-6 position and therefore oxidation of campesterol may occur duringfood processing and storage. Grandgirard et al. [9] demonstratedthat campesterol oxides can be absorbed in the intestine of the ratand the 3b,5a,6b-triol derivative of campesterol has been detectedin the plasma of healthy human subjects [10]. It is also possible thatcampesterol may be oxidised in vivo by both enzymatic and non-enzymatic mechanisms [10]. The cytotoxic effects of cholesteroloxides have been extensively reported [11] however, data relating

Fig. 1. Structures of cholesterol and related phytosterols.

Y. O’Callaghan et al. / Biochimie 95 (2013) 496e503 497

to the biological effects of phytosterol oxides is limited [12].Previous research conducted in our laboratory has demonstratedthat the oxidised products of b-sitosterol and stigmasterol are toxicto cells in a similar manner but to a lesser extent than theircholesterol oxide counterparts in the U937 cell line [13,14].Phytosterol oxides have also demonstrated cytotoxic effects innormal cell lines [15]. This may have significant clinical implica-tions for individuals consuming phytosterol enriched products andwarrants further investigation.

Minor differences in the chemical structure of cholesterol oxidescan significantly impact on their biological effects. Massey andPownall [16] found that there was a relationship between the 3-dimensional structure of cholesterol oxides, their membranebiophysical properties and ultimately their ability to induceapoptosis. A recent study conducted by Paillasse et al. [17] foundthat neither cholesterol-5a,6a-epoxide or cholesterol-5b,6b-epoxide reacted with nucleophiles in non-catalytic conditionsalthough chemicals bearing epoxide groups are generally potentalkylating agents. The authors also found that cholesterol-5a,6a-epoxide reacted with nucleophiles under catalytic conditions butcholesterol-5b,6b-epoxide remained unreactive. Rimner et al. [18]also examined the relationship between oxysterol structure andtheir cytotoxic effects and found that the b-isomers of oxysterols(7b-hydroxycholesterol, 5b,6b-epoxide) were up to 10-fold morecytotoxic than their equivalent a-iosmers in human arterial endo-thelial cells. The authors stated that the stereospecific effects of theoxysterols could have been mediated either at the level of the cellmembrane or through a putative oxysterol receptor. Carvalho et al.[15] have also reported a relationship between the structure andbiological activity of oxysterols in a variety of both normal andcancer cell lines.

The objective of the present study was to compare the cytotoxiceffects of a series of oxides synthesised from two parent compounds,the first parent compound was predominantly campesterol, 2 cam-pesterol:1 dihydrobrassiacasterol (2CMP:1DHB) 1c and the secondwas predominantly dihydrobrassicasterol, 3 dihydrobrassicasterol:1campesterol (3DHB:1CMP) 1d.The individual oxides of 2CMP:1DHBand 3DHB:1CMP were subsequently synthesised and characterisedaccording to themethods previously outlined for stigmasterol oxides[19]. The oxides were 2CMP:1DHB-5a,6a-epoxide 8c, 2CMP:1DHB-5b,6b-epoxide 7c, 7-Keto-2CMP:1DHB 4c, 7b-Hydroxy-2CMP:1DHB5c, 2CMP:1DHB -3b,5a,6b-triol 9c, 3DHB:1CMP-5a,6a-epoxide 8d,3DHB:1CMP-5b,6b-epoxide 7d, 7-keto-3DHB:1CMP 4d, 7b-Hydroxy-3DHB:1CMP 5d and 3DHB:1CMP-3b,5a,6b-triol 9d. The oxidisedderivative of 2CMP:1DHB, 5,6,22,23-Diepoxycampestane (4-epoxides) 12c was also investigated. Cell viability and apoptoticnucleiweredetermined foreachof thephytosterol oxidationproducts(POP) at 30 mM(12.5 mg/ml), 60 mM(25 mg/ml) and 120 mM(50 mg/ml)in the U937 humanmonocytic lymphoma cell line. The U937 cell lineis commonly used as a macrophage reference model in studiesinvestigating the cytotoxic effects of cholesterol oxidation products[20] and phytosterol oxidation products [13]. The cytotoxic effects ofcholesterol oxides and phytosterol oxides have previously been

shown to be cell line specific [15] therefore, the toxicity of the POPswas also investigated in the HepG2, human hepatoma cell line.Previous studies have demonstrated that the cholesterol oxides arecytotoxic but do not induce apoptosis in HepG2 cells [21].

2. Materials and methods

2.1. Materials

All chemicals and cell culture reagents were obtained fromSigma Chemical Co. (Poole, UK) unless otherwise stated. Cell lineswere obtained from the European Collection of Animal CellCultures (Salisbury, UK).

Samples of campesterol enriched 2CMP:1DHB 1c and dihydro-brassicasterol enriched 3DHB:1CMP 1d were synthesized as perO’Connell et al. [22] and were subsequently used in the oxidationsbelow.

Description of synthesis of CMP:DHB mixtures.CMP:DHB mixtures were synthesised in 6 steps starting from

stigmasterol (>95% purity, Sigma Aldrich). Stigmasterol was pro-tected via tosylate formation, solvolysed withmethanol, and cleavedunder ozonolysis conditions to yield an aldehyde. This aldehyde wasconverted via Wittig reaction to the required alkene for CMP:DHBwhich could then be hydrogenated and deprotected to yield theCMP:DHB mixtures (2CMP:1DHB and 3DHB:1CMP; both conclu-sively assigned by 1H and 13C NMR). All the CMP:DHB oxidesdescribed in this manuscript were synthesised from these materials.

2.2. Synthesis of 2CMP:1DHB oxides

The key oxides targeted in this study are 7-keto 4c and 4d, 7-b-OH 5c and 5d, b-epoxide 7c and 7d, a-epoxide 8c and 8d and triol9c and 9d to form a comparative study with other phytosteroloxides (Scheme 1) [19,23].

2CMP:1DHB 1c was converted to its corresponding acetylate 2cusing acetic anhydride and pyridine. Allylic oxidation employingchromium trioxide forms the acetate of 7-keto-2CMP:1DHB 3c. Theacetate group is cleaved using base-catalysed hydrolysis conditionsfurnishing 7-keto-2CMP:1DHB 4c. Synthesis of the other 7-positionoxide 7-b-OH 5cwas continued via stereoselective reduction of thecarbonyl group. The keto-steroid was reduced using sodium boro-hydride in the presence of cerium chloride heptahydrate whichpromotes attack of hydride from the axial position allowingformation of the equatorial alcohol functional group [24].

Oxidation was then focused on the alkene bond betweencarbons 5 and 6. Stereoselective epoxidation was achieved usinga biphasic system of potassium permanganate and copper sulphatein order to enhance oxidation on the more hindered b-face [25].Removal of the acetate group affords b-epoxide 7c.

Exposure of 2CMP:1DHB 1c to mCPBA yields the a-epoxide 8c.The peracid approaches from the less hindered side of the doublebond producing a mixture of epimers where the a-epoxidepredominates over the b-epoxide in a ratio of 4.2:1 evident from 1HNMR analysis (dH 2.90, d, a, 6-H and dH 3.06, d, b, 6-H). Theseisomers were inseparable by chromatography. Completing thispanel of oxides, synthesis of triol 9c was accomplished by acid-catalysed ring-opening of a-epoxide 8c.

The panel of 2CMP:1DHB oxides were extended to the bisep-oxide 12c as seen in previous studies [19]. Following deprotectionof 10c [22], excess mCPBA resulted in epoxidation of both doublebonds of 11c on the side chain and in the B ring, forming theepoxide on both upper and lower faces of the molecule (Scheme 2).This produced a mixture of diastereomers 5a,6a,22b,23b- 12c1,5a,6a,22a,23a- 12c2, 5b,6b,22b,23b- 12c3 and 5b,6b,22a,23a- 12c4in a ratio of 3.7:3.7:1:1 deduced from 1H NMR analysis. The 3.7:1

Scheme 1. Synthesis of campesterol and dihydrobrassicasterol oxides (c ¼ 2CMP:1DHB and d ¼ 3DHB:1CMP) (i) Ac2O, pyridine (2c ¼ 98%; 2d ¼ 96%); (ii) CrO3, dimethylpyrazole,CH2Cl2 �20 to 5 �C (3c ¼ 44%; 3d ¼ 56%); (iii) K2CO3, MeOH, H2O (4c ¼ 97%; 4d ¼ 95%); (iv) CeCl3.7H2O, NaBH4, MeOH, (5c ¼ 59%; 5d ¼ 45%); (v) KMnO4�CuSO4.5H2O, t-BuOH, H2O,CH2Cl2 (6c ¼ 32%; 6d ¼ 74%); (vi) Na2CO3, MeOH (7c ¼ 78%; 7d ¼ 77%); (vii) mCPBA, CH2Cl2 (8c ¼ 88%; 8c ¼ 49%); (viii) H2SO4, THF�H2O, (9c ¼ 77%; 9d ¼ 56%).

Y. O’Callaghan et al. / Biochimie 95 (2013) 496e503498

ratio was calculated from dH 2.90 (1H, bd, J 4.5, 6-H,12c1 and 12c2)and dH 3.07 (1H, d, J 2.2, 6-H, 12c3 and 12c4). The 1:1 ratio wascalculated from dH 2.76 (1H, dd, J 6.0, one of 23-H,12c3 or 12c4) anddH 2.82 (1H, dd, J 6.0, one of 23-H, 12c3 or 12c4). The mixtures ofdiastereomers could not be separated by column chromatography(Fig. 4).

2.3. Synthesis of 3DHB:1CMP oxides

The oxidation products were directly produced as above froma racemic sample of dihydrobrassicasterol and campesterol 1 ina ratio of 3:1 synthesised previously. The proportion of a- and b-epoxides differs slightly from the previous 2CMP:1DHB epoxides asfollows: b-epoxide 7d predominates over the a-epoxide in a ratio of7.9:1 evident from 1HNMR analysis (dH 3.06, d, b, 6-H and dH 2.90, d,a, 6-H) and a-epoxide 8d predominates over the b-epoxide ina ratio of 8.0:1 evident from 1H NMR analysis (dH 2.90, d, a, 6-H anddH 3.06, d, b, 6-H). In both cases, these isomers were inseparable bychromatography.

2.4. Cell maintenance

Human monocytic, U937 cells were grown in Royal ParkMemorial Institute (RPMI)-1640 medium supplemented with 10%(v/v) foetal bovine serum (FBS). The human, hepatoma cell line,HepG2, was maintained in Dulbecco’s modified Eagle’s medium(DMEM) supplemented with 10% (v/v) FBS and 1% non-essentialamino acids. The cells were grown at 37 �C and 5% (v/v) CO2 ina humidified incubator. Cells were cultured in the absence ofantibiotics.

Scheme 2. Synthesis of bisepoxide of 2CMP:1DHB (i) 2.5 M

2.5. Cell treatment with phytosterol oxides

Cells (U937 or HepG2) were adjusted to a density of 1�105cells/ml in media supplemented with 2.5% (v/v) FBS. HepG2 cells wereallowed to adhere overnight prior to the addition of the oxides. Theoxides were dissolved in ethanol and were added to the cells toa final concentration of 30, 60 and 120 mM. Equivalent quantities ofethanol were added to control cells and samples were incubated for24 h.

2.6. Cytotoxicity, cell viability and apoptosis

The cytotoxic and apoptotic effects of the oxides were assessedas previously described in O’Callaghan et al. [14]. Briefly, theviability of U937 cells was determined by staining the cells usinga solution containing 0.1 mM fluorescein diacetate and 0.02 mMethidium bromide (FDA-EtBr) which causes live cells to fluorescegreen and non-viable cells to fluoresce red. Cells were mixed (1:1)with FDA-EtBr and incubated at 37 �C for 5 min. Two hundred cellswere scored at 200x magnification on a Nikon Optiphot-2 fluo-rescence microscope and the viable cells were expressed asa percentage of the total cell count.

U937 cell viability was determined using the MTT cell prolifer-ation Kit (Roche, Mannheim, Germany). The MTT assay is based onthe cleavage of the yellow tetrazolium salt MTT to purple formazanby metabolically active cells. Cells were seeded in the wells of a 96well plate and were exposed to POP for 24 h. MTT (3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyl tetrazolium bromide) (10 ml)was added to each of the wells and cells were incubated for 4 hprior to the addition of 100 ml solubilisation solution (10% SDS in

H2SO4, THF�H2O (58%); (ii) mCPBA, CH2Cl2, 0 �C (54%).

Table 1Percent viable U937 cells following exposure to 3DHB:1CMP or 2CMP:1DHB oxidesat concentrations of 30 mM, 60 mM or 120 mM for 24 h, as determined by the fluo-rescein diacetate (FDA/EtBr) assay.

3DHB:1CMP (d)mean � se

2CMP:1DHB (c)mean � se

Control 95.3 � 0.3 97.5 � 0.430 mM a-epoxide (8) 87.5 � 4.3 95.6 � 0.960 mM a-epoxide (8) 89.4 � 3.0 96.0 � 1.5120 mM a-epoxide (8) 79.2 � 2.9 88.7 � 2.030 mM b-epoxide (7) 84.7 � 5.9 96.0 � 0.660 mM b-epoxide (7) 92.3 � 1.6 94.6 � 0.4120 mM b-epoxide (7) 85.8 � 3.4 62.3 � 5.4*

30 mM 7-keto (4) 90.3 � 3.1 95.0 � 1.260 mM 7-keto (4) 77.3 � 5.5 82.8 � 3.6120 mM 7-keto (4) 30.2 � 2.8* 68.8 � 1.4*

30 mM 7b-OH (5) 88.5 � 4.6 63.0 � 3.1*

60 mM 7b-OH (5) 69.0 � 8.8* 36.2 � 1.6*

120 mM 7b-OH (5) 38.5 � 8.5* 27.3 � 2.2*

30 mM 3,5,6-triol (9) 62.9 � 2.2* 63.2 � 4.4*

60 mM 3,5,6-triol (9) 34.0 � 12.0* 45.5 � 2.8*

120 mM 3,5,6-triol (9) 11.5 � 3.5* 32.6 � 3.1*

The effect of compounds 8d, 8c, 7d, 7c, 4d, 4c, 5d, 5c, 9d and 9c on U937 cell viabilityat concentrations of 30 mM, 60 mM and 120 mM was evaluated following a 24 hincubation by the fluorscein diacetate:ethidium bromide (FDA:EtBr) stainingmethod. The control represents the percentage viability in untreated cells. U937cells weremixed 1:1 with FDA:EtBr stain. A total of 200 cells were scored per sampleand the percent viable cells were determined. Data represent the mean of fourindependent experiments� standard error of themean. *P< 0.05, indicates that thevalue obtained for samples treated with either 3DHB:1CMP or 2CMP:1DHB differedsignificantly from their respective control, ANOVA followed by Dunnetts.

Y. O’Callaghan et al. / Biochimie 95 (2013) 496e503 499

0.01M HCl). The absorbance was determined at 570 nm usinga platereader (Spectrafluorplus, Tecan) and the data wereexpressed as a percentage of the control.

The Neutral Red Uptake assay was used to test the viability ofHepG2 cells exposed to POP for 24 h. Following incubation, the cellswere washed with KrebseRinger buffer (115 mMNaCl, 4.7 mM KCl,1.2 mM KH2PO4, 1.2 mM MgSO4e7H2O, 24 mM NaHCO3, 20 mMHepes, 1.1 mM glucose, 1.28 mM CaCl2e6H2O) and the neutral reddye (50 mg/ml KrebseRinger) was added. Cells were then incubatedfor 2 h at 37 �C and 5% (v/v) CO2. Cells were subsequently lysed byadding 1% glacial acetic acid and the absorbance was determined ata wavelength of 540 nm using a platereader (Spectrafluorplus,Tecan).

2.7. Apoptosis

To measure apoptosis, U937 cells were incubated with POP for24 h. Cells were harvested by centrifugation and the morphology ofthe cell nuclei was examined following staining with 200 mlHoechst 33342 (0.008 mM). Stained samples were placed ona microscope slide and examined under UV light on the NikonOptiphot-2 fluorescence microscope. A total of 300 cells per samplewere counted and the percentage of fragmented and condensednuclei was calculated.

The DNA fragmentation assay was carried out to confirm thepresence of apoptotic nuclei in the samples. Briefly, 2 � 106 cellswere harvested. The pellets were lysed and 10 ml RNase A (1 mg/ml)and 10 ml Proteinase K (20 mg/ml) were added. Samples wereloaded to the wells of a 1.5% (w/v) agarose gel and electrophoresiswas carried out (110 V, 3 h). The gel was visualised under UV lighton a transilluminator following staining with ethidium bromideand was photographed using the Genegenius Bioimaging system.

2.8. Statistical analysis

All data points are the mean and standard error values of at leastthree independent experiments. Data were analysed by ANOVAfollowed by Dunnett’s test. The software employed for statisticalanalysis was Graphpad Prism, version 4.00 for windows (Graphpadsoftware, San Diego, California, USA).

3. Results and discussion

3.1. Toxicity of oxides in U937 cells

The FDA-EtBr assay was used to compare the toxicity of theoxidised derivatives of 2CMP:1DHB 1c and 3DHB:1CMP 1d in theU937 cell line (Table 1) following a 24 h incubation period. The a-epoxide derivatives of both 2CMP:1DHB 8c and 3DHB:1CMP 8dwere not cytotoxic in the U937 cell line at the concentrationsinvestigated in the present study. Similarly, the a-epoxide deriva-tives of cholesterol, sitosterol and stigmasterol have been shownnot to be cytotoxic in the U937 cell line [10,11]. b-Epoxide deriva-tives of 3DHB:1CMP 7d were not toxic, following 24 h incubation,as has previously been shown for the b-epoxide derivative ofstigmasterol but cell viability was significantly reduced (P < 0.05)to 62.3%, at the highest concentration (120 mM) of 2CMP:1DHB-b-epoxide 7c. The b-epoxide derivatives of both cholesterol and b-sitosterol have also been shown to cause a significant reduction inU937 cell viability [13]. The 7-keto derivative of both 3DHB:1CMP4d and 2CMP:1DHB 4c induced a significant (P < 0.05) reduction incell viability at the 120 mM concentration. 7-Keto-b-sitosterol wassignificantly cytotoxic at the 120 mM concentration [13] while the7-keto derivative of stigmasterol was not cytotoxic in the U937 cellline [14]. The 7b-OH derivative was significantly (P < 0.05)

cytotoxic at all three concentrations of 7beOHe2CMP:1DHB 5c andat the two higher concentrations (60 mM and 120 mM) of7beOHe3DHB:1CMP 5d following a 24 h incubation. The triolderivative was the most cytotoxic of all the derivatives as previ-ously demonstrated for cholesterol, b-sitosterol and stigmasterol.The triol derivatives of both 3DHB:1CMP 9d and 2CMP:1DHB 9csignificantly reduced cell viability at all three concentrations. Tosummarise, the order of toxicity of the 2CMP:1DHB and3DHB:1CMP oxides was similar to that previously demonstrated forcholesterol, b-sitosterol and stigmasterol in the U937 cell line. In allcases higher concentrations of the phytosterol oxides wererequired in order to induce similar levels of toxicity to thoseinduced by cholesterol oxides. Therefore, the introduction of anethyl group as in b-sitosterol and stigmasterol or a methyl group asin campesterol and 22-dihydrobrassicasterol at the C24 position onthe sterol molecule reduced the cytotoxicity of these compounds inU937 cells compared to their equivalent COPs. It has previouslybeen established that oxidised phytosterols are approximatelythree-fold less cytotoxic than their equivalent cholesterol oxides inU937 cells [13,26]. Carvalho et al. [15] also found that cholesteroloxides were more cytotoxic than 7beOHesitosterol in cancer celllines, however the level of cytotoxicity induced by7beOHesitosterol was almost equal to that of cholesterol oxides innormal cells derived from the eye (ARPE-19) and the skin (BJ). Theauthors suggested that phytosterol oxides may therefore be asharmful as cholesterol oxides in vivo. In the present study, the2CMP:1DHB oxides were significantly more cytotoxic (P < 0.05) atlower concentrations than the 3DHB:1CMP oxides.

The FDA-EtBr staining assay is a membrane integrity assay andtheMTTassay determines cell viability bymeasuringmitochondrialactivity and can be a more sensitive measure of cytotoxicity. All ofthe oxidised derivatives of 2CMP:1DHB 4-9c significantly (P< 0.05)reduced viability of U937 cells as determined by the MTT assay(Table 2), the observed reduction in cell viability was dose depen-dent. U937 cell viability as determined by MTT was less affected bythe addition of the oxides of 3DHB:1CMP 4-9d in comparison with

Table 3Viability in HepG2 cells following exposure to 3DHB:1CMP or 2CMP:1DHB oxides atconcentrations of 30 mM, 60 mM or 120 mM for 24 h expressed as percent control, asdetermined by the neutral red uptake assay.

3DHB:1CMP (d)mean � se

2CMP:1DHB(c)mean � se

30 mM a-epoxide (8) 97.1 � 5.3 97.9 � 4.160 mM a-epoxide (8) 99.3 � 5.8 90.9 � 4.0120 mM a-epoxide (8) 99.4 � 4.2 77.9 � 1.9*

30 mM b-epoxide (7) 105.8 � 4.4 87.5 � 2.060 mM b-epoxide (7) 100.3 � 4.2 92.0 � 1.3120 mM b-epoxide (7) 87.5 � 4.9 89.5 � 4.830 mM 7-keto (4) 92.7 � 3.6 79.8 � 4.2*

60 mM 7-keto (4) 82.6 � 4.3* 80.0 � 10.1*

120 mM 7-keto (4) 65.7 � 3.4* 78.3 � 4.8*

30 mM 7b-OH (5) 94.6 � 3.0 77.0 � 3.0*

60 mM 7b-OH (5) 88.4 � 1.6 95.6 � 4.8120 mM 7b-OH (5) 80.7 � 1.7* 45.8 � 4.3*

30 mM 3,5,6-triol (9) 79.2 � 5.4* 52.3 � 7.4*

60 mM 3,5,6-triol (9) 69.3 � 4.1* 48.3 � 5.5*

120 mM 3,5,6-triol (9) 45.0 � 4.2* 41.1 � 5.0*

The effect of compounds 8d, 8c, 7d, 7c, 4d, 4c, 5d, 5c, 9d and 9c on HepG2 cellviability at concentrations of 30 mM, 60 mM and 120 mM was evaluated followinga 24 h incubation by the neutral red uptake assay. Absorbance of the samples wasdetermined at 540 nm following a 2 h incubation with neutral red dye. The resultsare expressed as a percentage of the control value (100%). Data represent the meanof four independent experiments � standard error of the mean. *P < 0.05, indicatesthat the value obtained for samples treated with either 3DHB:1CMP or 2CMP:1DHBdiffered significantly from their respective control, ANOVA followed by Dunnetts.

Table 2Cell viability in U937 cells following exposure to 3DHB:1CMP or 2CMP:1DHB oxidesat concentrations of 30 mM, 60 mM or 120 mM for 24 h, as determined by the MTTassay.

3DHB:1CMP (d)mean � se

2CMP:1DHB (c)mean � se

30 mM a-epoxide (8) 0.96 � 0.09 0.74 � 0.03*

60 mM a-epoxide (8) 0.93 � 0.07 0.67 � 0.05*

120 mM a-epoxide (8) 0.52 � 0.05* 0.35 � 0.06*

30 mM b-epoxide (7) 0.71 � 0.07 0.86 � 0.04*

60 mM b-epoxide (7) 0.68 � 0.05* 0.81 � 0.04*

120 mM b-epoxide (7) 0.66 � 0.04* 0.35 � 0.01*

30 mM 7-keto (4) 0.80 � 0.06 0.55 � 0.02*

60 mM 7-keto (4) 0.60 � 0.07* 0.11 � 0.00*

120 mM 7-keto (4) 0.12 � 0.03* 0.04 � 0.01*

30 mM 7b-OH (5) 0.92 � 0.14 0.14 � 0.07*

60 mM 7b-OH (5) 0.55 � 0.08* 0.02 � 0.02*

120 mM 7b-OH (5) 0.20 � 0.04* 0.00 � 0.01*

30 mM 3,5,6-triol (9) 0.49 � 0.11* 0.25 � 0.04*

60 mM 3,5,6-triol (9) 0.15 � 0.03* 0.04 � 0.02*

120 mM 3,5,6-triol (9) 0.05 � 0.01* 0.03 � 0.02*

The effect of compounds 8d, 8c, 7d, 7c, 4d, 4c, 5d, 5c, 9d and 9c on U937 cell viabilityat concentrations of 30 mM, 60 mM and 120 mM was evaluated following a 24 hincubation by the MTT assay. Absorbance of the samples was determined at 570 nmfollowing a 4 h incubation with MTT (3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide). The results are expressed relative to the control value(1.00). Data represent the mean of four independent experiments � standard errorof the mean. *P < 0.05, indicates that the value obtained for samples treated witheither 3DHB:1CMP or 2CMP:1DHB differed significantly from their respectivecontrol, ANOVA followed by Dunnetts.

Y. O’Callaghan et al. / Biochimie 95 (2013) 496e503500

the 2CMP:1DHB oxides. The a-epoxide derivative of 3DHB:1CMP8d caused a significant (P < 0.05) reduction in cell viability at the120 mM concentration only, the b-epoxide 7d, 7-keto 4d and 7b-OH5d derivatives significantly (P < 0.05) reduced cell viability at thetwo higher concentrations (60 mM and 120 mM) and the triol 9dderivative significantly (P < 0.05) reduced cell viability at allconcentrations employed in the present study. The data obtainedby the MTT assay demonstrated that the oxidised derivatives of2CMP:1DHB 4-9c were more cytotoxic than the oxidised deriva-tives of 3DHB:1CMP 4-9d in the U937 cell line.

3.2. Toxicity of oxides in HepG2 cells

Studies have demonstrated that the effects of oxysterols are cellline specific [15,27]. Therefore, we also investigated the toxicity ofthe 2CMP:1DHB and 3DHB:1CMP oxides in the HepG2, humanhepatoma cell line, using the neutral red uptake assay. The order oftoxicity of the 2CMP:1DHB and 3DHB:1CMP oxides were found tobe similar in the HepG2 and U937 cell lines (Table 3). However, theb-epoxide 7c, 7d derivative was not significantly (P < 0.05) cyto-toxic in the HepG2 cell line (Table 3). The a-epoxide derivative of2CMP:1DHB 8c was toxic at the 120 mM concentration only. Ryanet al. [13] demonstrated that a- and b- epoxysitosterol were notcytotoxic in the HepG2 cell line. 7-Keto-2CMP:1DHB 4c signifi-cantly (P < 0.05) reduced cell viability at all three concentrations(Table 3) however, no dose response was observed as the level ofcytotoxicity did not increase at concentrations above 30 mM. HepG2cell viability was significantly (P < 0.05) decreased in the presenceof 60 mM and 120 mM 7-keto-3DHB:1CMP 4d and a dose responsewas observed for this compound. 7beOHe2CMP:1DHB 5c signifi-cantly (P < 0.05) decreased HepG2 cell viability at the 30 mMconcentration and to less than 50% of the control level at the120 mM concentration. 7beOHe3DHB:1CMP 5d was only signifi-cantly cytotoxic at the highest concentration (120 mM) in theHepG2 cell line. 7-Keto-and 7beOHeb-sitosterol have previouslybeen shown to be more toxic in HepG2 cells, reducing viability to7.1% and 4.6% of the control, respectively, at 120 mM [13] however,

Koschutnig et al. [28] found that 7beOHeb-sitosterol was notcytotoxic in HepG2 cells and cell viability was 50% at 120 mM 7-keto-b-sitosterol. Again, the triol derivatives 9c, 9d were the mostcytotoxic of all the derivatives tested, reducing cell viability to lessthan 50% at 120 mM3DHB:1CMP-triol 9d and at both the 60 mM and120 mM concentrations of 2CMP:1DHB-triol 9c. The difference intoxicity between the derivatives of 3DHB:1CMP and 2CMP:1DHBwas not as evident in the HepG2 cell line as in the U937 cell line.

3.3. Apoptosis induced by oxides

The ability of a compound to induce apoptosis in cancer celllines may be used as an indication of its potential as a chemother-apeutic agent. Certain novel oxides of stigmasterol (5,6,22,23-diepoxystigmastone) have previously demonstrated an ability toinduce cell death which was almost exclusively apoptotic at certainconcentrations [14]. One of the objectives of the present study wasto differentiate between the apoptogenicity of the oxidised deriv-atives of 3DHB:1CMP and 2CMP:1DHB. The a-epoxide derivative ofboth 3DHB:1CMP 8d and 2CMP:1DHB 8c did not induce apoptosis(Table 4). The b-epoxide derivative of 2CMP:1DHB 7c causeda significant increase (P < 0.05) in apoptotic nuclei to greater than7-fold of the untreated control, at the 120 mM concentration. Therewas also a significant decrease in cell viability at this concentrationto 62.3% therefore, apoptotic cell death accounted for approxi-mately 50% of total cell death. The 7-keto derivatives 4c, 4d alsoincreased apoptotic nuclei relative to control samples. The 7b-OHderivatives 5c, 5d increased apoptotic nuclei significantly (P< 0.05)at the 60 mM and 120 mM concentration of 7beOHe3DHB:1CMP 5dand at all three concentrations of 7beOHe2CMP:1DHB 5c. The triolderivatives of 2CMP:1DHB 9c did not significantly (P < 0.05)increase the apoptotic nuclei in U937 cells. It has previously beenobserved that the predominant mode of cell death induced by thetriol derivative of b-sitosterol is necrosis [13]. The ability of theoxides to induce cell death by apoptosis was confirmed by theappearance of a ladder like pattern, representing cleavage of theDNA to fragments of 200 base pairs, in an agarose gel (Figs. 2 and 3).The oxidised derivatives of b-sitosterol have previously

Fig. 3. DNA fragmentation: U937 cells were incubated with 60 mM campesterol oxidesfor 24 h. Cells were harvested, lysed and incubated with proteinase K and RNAse A.Samples were loaded to the wells of an agarose gel and a current of 110 V was appliedfor 3 h. Apoptotic cells form a ladder like pattern representative of the cleavage of DNAto 200 base pair units which is the hallmark of apoptosis. Lanes: MW: molecularweight marker (200 bp); Control: untreated cells, a-E: 2CMP:1DHB-5a,6a-epoxide(8c); 2CMP:1DHB-5b,6b-epoxide (7c), 7-K: 7-Keto-2CMP:1DHB (4c), 7-OH: 7b-Hydroxy-2CMP:1DHB 5c, Triol: 2CMP:1DHB -3b,5a,6b-triol (9c).

Table 4Apoptotic nuclei in U937 cells following exposure to 3DHB:1CMP or 2CMP:1DHBoxides at concentrations of 30 mM, 60 mM or 120 mM for 24 h, as determined bystaining with Hoechst 33342.

3DHB:1CMP (d)mean � se

2CMP:1DHB (c)mean � se

Control 1.00 � 0.08 1.00 � 0.0830 mM a-epoxide (8) 1.48 � 0.20 0.86 � 0.3260 mM a-epoxide (8) 1.85 � 0.30 1.59 � 0.23120 mM a-epoxide (8) 2.33 � 0.16 2.77 � 0.4530 mM b-epoxide (7) 1.54 � 0.40 1.50 � 0.2760 mM b-epoxide (7) 1.65 � 0.22 2.23 � 0.27120 mM b-epoxide (7) 1.93 � 0.30 7.23 � 0.91*

30 mM 7-keto (4) 2.12 � 0.37 3.04 � 0.3660 mM 7-keto (4) 3.47 � 0.58* 3.95 � 0.64120 mM 7-keto (4) 3.53 � 0.42* 4.86 � 1.73*

30 mM 7b-OH (5) 1.97 � 0.34 5.86 � 2.00*

60 mM 7b-OH (5) 2.62 � 0.24* 6.54 � 1.86*

120 mM 7b-OH (5) 3.15 � 0.36* 6.82 � 0.18*

30 mM 3,5,6-triol (9) 2.05 � 0.12 2.91 � 0.3660 mM 3,5,6-triol (9) 2.65 � 0.50* 4.45 � 0.27120 mM 3,5,6-triol (9) 3.55 � 0.12* 3.23 � 0.73

The effect of compounds 8d, 8c, 7d, 7c, 4d, 4c, 5d, 5c, 9d and 9c on the quantity ofapoptotic nuclei in U937 cells exposed to 30 mM, 60 mM and 120 mM of thecompounds for 24 h was evaluated. U937 cells were harvested and nuclei werestained with Hoechst 33342. A total of 300 nuclei were scored per sample and thepercent condensed/fragmented nuclei was determined. Data are expressed as foldincrease in condensed/fragmented nuclei relative to control cells. The control cellsrepresent untreated cells and were 2.2% apoptotic. Data represent the mean of fourindependent experiments� standard error of themean. *P< 0.05, indicates that thevalue obtained for samples treated with either 3DHB:1CMP or 2CMP:1DHB differedsignificantly from their respective control, ANOVA followed by Dunnetts.

Y. O’Callaghan et al. / Biochimie 95 (2013) 496e503 501

demonstrated higher levels of apoptotic cell death [13] than theoxides investigated in the present study. Four of the five oxidesinvestigated induced apoptosis, the highest level of apoptosis wasfound in U937 cells exposed to 7-keto-b-sitosterol which caused

Fig. 2. DNA fragmentation: U937 cells were incubated with 60 mM dihydro-brassicasterol oxides for 24 h. Cells were harvested, lysed and incubated withproteinase K and RNAse A. Samples were loaded to the wells of an agarose gel anda current of 110 V was applied for 3 h. Apoptotic cells form a ladder like patternrepresentative of the cleavage of DNA to 200 base pair units which is the hallmark ofapoptosis. Lanes: MW: molecular weight marker (200 bp); Control: untreated cells, a-E: 3DHB:1CMP-5a,6a-epoxide (8d); b-E: 3DHB:1CMP-5b,6b-epoxide (7d), 7-K: 7-keto-3DHB:1CMP (4d), 7-OH: 7b-Hydroxy-3DHB:1CMP (5d), Triol: DHB:1CMP-3b,5a,6b-triol (9d).

a 20 fold increase in apoptotic nuclei, relative to the control sample.7b-OH Stigmasterol was the only one of the five oxides (a-epoxide,b-epoxide, 7-keto, 7b-OH, triol) of stigmasterol which inducedapoptosis in U937 cells causing a 10-fold increase at 60 mM [14].

3.4. Cytotoxic effects of 5,6,22,23-diepoxycampestane

The bisepoxide derivative of 2CMP:1DHB (5,6,22,23-diepoxycampestane 12c) was found to the most cytotoxic of allthe compounds investigated in the present study (Table 5) reducingcell viability to 11.7% at the 30 mM concentration therefore, thepresence of an additional epoxide group on the side chain greatlyincreased the toxicity of the compound in comparison to the a-epoxide, b-epoxide derivatives of campesterol. However, the levelof apoptosis was low and it must be concluded that cell deathinduced by this compoundwas largely necrotic. A similar derivativeof stigmasterol (5,6,22,23-diepoxystigmastane) was less cytotoxicthan 5,6,22,23-diepoxycampestane but induced cell death almostexclusively by apoptosis at the 30 mM concentration [14].

The concentrations of phytosterol oxide employed in thepresent study (30 mM, 60 mM and 120 mM) were selected to allowa comparison with cholesterol oxides which have typically beeninvestigated at concentrations ranging from 10 mM to 150 mM[19,20,27]. In general to induce a similar level of toxicity theconcentrations of phytosterol oxides required were approximatelythree-fold higher than their cholesterol oxide counterpart. Thedifference in the cytotoxicity of the various phytosterol andcholesterol oxides (a-epoxide, b-epoxide, 7-keto, 7b-OH, triol) maybe caused by differences in their uptake and metabolism by cells.Few studies have investigated the uptake of phytosterol oxides incells, however it has been found that cholesterol-5,6a-epoxide,cholesterol-5,6b-epoxide and 3b,5a,6b-cholestane-triol are alltaken up at a similar rate by rabbit aortic endothelial cells [29].Cholesterol oxides with similar structures have demonstrated

Fig. 4. Structures of the mixture of bisepoxides of 2CMP:1DHB in 12c

Y. O’Callaghan et al. / Biochimie 95 (2013) 496e503502

different metabolisms [17]. Cholesterol-5,6a-epoxide is a directmodulator of Liver-X-Receptors (LXR) [30] and bis 5,6a-24(S),25-didiepoxycholesterol is a selective LXRa modulator [31]. The 3b-sulphated form of cholesterol-5,6a-epoxide is an antagonist of LXR[32] whereas the 3b sulphated form of cholesterol-5,6b-epoxide isinactive. Both cholesterol-5,6a-epoxide and cholesterol-5,6b-epoxide can be converted to 3b,5a,6b-cholestane triol by action ofthe enzyme cholesterol epoxide hydrolase (ChEH) which may beactive in tumour cells [33]. ChEH has demonstrated a preference tohydrate Cholesterol-5,6b-epoxide over cholesterol-5,6a-epoxide[34] which may account for the greater cytotoxicity of cholesterol-5,6b-epoxide in U937 cells. To date, there has been no investigationinto the interaction of ChEH and phytosterol oxides.

Compounds which induce apoptosis can provide targeted anti-cancer therapies as the apoptotic program is altered in tumour cells[35] therefore phytosterol oxides such as 5,6,22,23-diepoxystigmastane may have the potential to be developed as chemothera-peutic agents. Cholesterol oxides havedemonstrated selective toxicityto cancer cells and are relatively non-toxic in normal cells, howevercertain phytosterol oxides have demonstrated cytotoxic effects innormal cells [15] therefore it requires further study to elucidate thepotential benefits of these compounds as anticancer therapies.

Table 5Cytotoxicity in U937 and HepG2 cells and apoptotic nuclei in U937 cells followingexposure to 5,6,22,23-Diepoxycampestane (Bisepoxide 12c) at concentrations of30 mM, 60 mM or 120 mM for 24 h.

Viability(U937 cells)mean � se

Viability(HepG2 cells)mean � se

Apoptosis(U937 cells)mean � se

30 mM Bisepoxide (12c) 11.7 � 5.2* 48.4 � 8.2* 4.45 � 0.91*

60 mM Bisepoxide (12c) 1.3 � 1.8* 0.0 � 5.6* 0.59 � 0.27120 mM Bisepoxide (12c) 0.0 � 0.0* 0.0 � 5.4* 0.41 � 0.14

The effect of 5,6,22,23-Diepoxycampestane on U937 and HepG2 cell viability atconcentrations of 30 mM, 60 mM and 120 mM was evaluated following a 24 h incu-bation. Cell viability in U937 cells was determined by fluorescein diacetate:ethidiumbromide staining, data are percent viable cells. Cell viability was determined by theneutral red uptake assay in the HepG2 cells, data are expressed as a percentage ofthe control. Apoptotic nuclei were determined following staining with Hoechst33342, data are expressed as fold increase in condensed/fragmented nuclei relativethe control cells. The control cells represent untreated cells andwere 2.2% apoptotic.Data represent the mean of four independent experiments � standard error of themean. *P < 0.05, significantly different from the control, untreated cells, ANOVAfollowed by Dunnetts.

4. Conclusion

Elucidation of the potential detrimental health effects ofphytosterol oxidation products has hitherto been hampered by thelack of standard compounds. The present study which investigatesthe cytotoxic effects of campesterol oxides and dihydro-brassicasterol oxides in U937 and HepG2 cells adds to previousstudies which reported the cytotoxic effects of b-sitosterol oxidesand stigmasterol oxides [13,14]. Cholesterol oxides have greatercytotoxicity than phytosterol oxides in both the U937 and HepG2cell lines. In relation to the phytosterol oxides, the oxidised deriv-atives of b-sitosterol were the most cytotoxic and the mostapoptotic, followed by the campesterol (67%) oxides, stigmasteroloxides and dihydrobrassicasterol (75%) oxides, although it ispossible by enhancing the purity of campesterol, the toxicity of theoxides may be increased. These results signify the importance ofcampesterol oxides in the overall paradigm of phytosterol oxidecytotoxicity.

Acknowledgements

Funding for this research was provided under the NationalDevelopment Plan, through the Food Institutional ResearchMeasure, administered by the Department of Agriculture, Food andFisheries, Ireland.

Appendix A. Supplementary material

Supplementary material related to this article can be foundonline at doi:10.1016/j.biochi.2012.04.019.

References

[1] J. Plat, R.P. Mensink, Plant stanol and sterol esters in the control of bloodcholesterol levels: mechanism and safety aspects, Review, Am. J. Cardiol. 96(2005) 15e22.

[2] P. Child, A. Kuksis, Uptake of 7-dehydro derivatives of cholesterol, campes-terol, and b-sitosterol by rat erythrocytes, jejunal villus cells and brush bordermembranes, J. Lipid Res. 24 (1983) 552e565.

[3] T.A. Miettinen, M. Nissinen, M. Lepäntalo, A. Albäck, M. Railo, P. Vikatmaa,M. Kaste, S. Mustanoja, H. Gylling, Non-cholesterol sterols in serum andendarterectomized carotid arteries after a short-term plant stanol and sterolester challenge, Nutr. Metab. Cardiovasc. Dis. 21 (2011) 182e188.

Y. O’Callaghan et al. / Biochimie 95 (2013) 496e503 503

[4] E.R. Kelly, J. Plat, R.P. Mensink, T.T. Berendschot, Effects of long term plantsterol and -stanol consumption on the retinal vasculature: a randomizedcontrolled trial in statin users, Atherosclerosis 214 (2011) 225e230.

[5] O. Weingartner, D. Lutjohann, S. Ji, N. Weisshoff, F. List, T. Sudhop, K. vonBergmann, K. Gertz, J. Konig, H.J. Schafers, M. Endres, M. Bohm, U. Laufs,Vascular effects of diet supplementation with phytosterol, J. Am. Coll. Cardiol.51 (2008) 1553e1561.

[6] P.G. Bradford, A.B. Awad, Phytosterols as anticancer compounds, Mol. Nutr.Food Res. 51 (2007) 161e170.

[7] K. Bhattacharyya, W.E. Connor, Beta-sitosterolemia and xanthomatosis. Anewly described lipid storage disease in two sisters, J. Clin. Invest. 53 (1974)1033e1043.

[8] P.K. Gladu, G.W. Patterson, G.H. Wikfors, D.J. Chitwood, W.R. Lusby, Sterols ofsome Diatoms, Phytochemistry 30 (1991) 2301e2303.

[9] A. Grandgirard, J.P. Sergiel, M. Nour, J. Demaison-Meloche, C. Giniès,Lymphatic absorption of phytosterol oxides in rats, Lipids 34 (1999) 563e570.

[10] A. Grandgirard, L. Martine, L. Demaison, C. Cordelet, C. Joffre, O. Berdeaux,E. Semon, Oxyphytosterols are present in plasma of healthy human subjects,Br. J. Nutr. 91 (2004) 101e106.

[11] A. Vejux, L. Malvitte, G. Lizard, Side effects of oxysterols: cytotoxicity, oxida-tion, inflammation, and phospholipidosis, Braz. J. Med. Biol. Res. 41 (2008)545e556.

[12] G. Garcia-Llatas, M.T. Rodriguez-Estrada, Current and new insights onphytosterol oxides in plant sterol-enriched food, Chem. Phys. Lipids 164(2011) 607e624.

[13] E. Ryan, J. Chopra, F.O. McCarthy, A.R. Maguire, N.M. O’Brien, Qualitative andquantitative comparison of the cytotoxic and apoptotic potential of phytos-terol oxidation products with their corresponding cholesterol oxidationproducts, Br. J. Nutr. 94 (2005) 443e451.

[14] Y.C. O’Callaghan, D.A. Foley, N.M. O’Connell, F.O. McCarthy, A.R. Maguire,N.M. O’Brien, Cytotoxic and apoptotic effects of the oxidised derivatives ofstigmasterol in the U937 human monocytic cell line, J. Agric. Food Chem. 58(2010) 10793e10798.

[15] J.F. Carvalho, M.M. Silva, J.N. Moreira, S. Simões, M.L. Sá e Melo, Sterols asanticancer agents: synthesis of ring-B oxygenated steroids, cytotoxic profileand comprehensive SAR analysis, J. Med. Chem. 53 (2010) 7632e7638.

[16] J.B. Massey, H.J. Pownall, Structures of biologically active oxysterols determinetheir differential effects on phospholipid membranes, Biochemistry 45 (2006)10747e10758.

[17] M.R. Paillasse, N. Saffon, H. Gornitzka, S. Silvente-Poirot, M. Poirot, P. deMedina, Surprising ’un’ reactivity of cholesterol-5,6-epoxides towardsnucleophiles, J. Lipid Res. 53 (2012) 718e725.

[18] S. Rimner, A.l. Makdessi, H. Sweidan, J. Wischhusen, B. Rabenstein, K. Shatat,P. Mayer, I. Spyridopoulos, Relevance and mechanism of oxysterol stereo-specificity in coronary artery disease, Free Radic. Biol. Med. 38 (2005)535e544.

[19] D.A. Foley, Y.C. O’Callaghan, N.M. O’Brien, F.O. McCarthy, A.R. Maguire,Synthesis and characterisation of stigmasterol oxidation products, J. Agric.Food Chem. 58 (2010) 1165e1173.

[20] G. Lizard, S. Gueldry, O. Sordet, S. Monier, A. Athias, C. Miguet, G. Bessede,S. Lemaire, E. Solary, P. Gambert, Glutathione is implied in the control of 7-

ketocholesterol-induced apoptosis, which is associated with radical oxygenspecies production, FASEB J. 15 (1998) 1651e1663.

[21] Y.C. O’Callaghan, J.A. Woods, N.M. O’Brien, Characteristics of 7 beta-hydroxycholesterol-induced cell death in a human monocytic blood cellline, U937, and a human hepatoma cell line, HepG2, Toxicol. In Vitro 16 (2002)245e251.

[22] N.M. O’Connell, Y.C. O’Callaghan, N.M. O’Brien, A.R. Maguire, F.O. McCarthy,Synthetic Routes to Campesterol and Dihydrobrassicasterol: a First ReportedSynthesis of the Key Phytosterol Dihydrobrassicasterol, Tetrahedron,in press.

[23] F.O. McCarthy, J. Chopra, A. Ford, S.A. Hogan, J.P. Kerry, N.M. O’Brien, E. Ryan,A.R. Maguire, Synthesis, isolation and characterisation of beta-sitosterol andbeta-sitosterol oxide derivatives, Org. Biomol. Chem. 3 (2005) 3059e3065.

[24] E. St’astna, I. Cerny, V. Pouzar, H. Chodounska, Stereoselectivity of sodiumborohydride reduction of saturated steroidal ketones utilizing conditions ofLuche reduction, Steroids 75 (2010) 721e725.

[25] J.A.R. Salvador, M.L.S.E. Melo, A.S.C. Neves, Oxidations with potassiumpermanganate e metal sulphates and nitrates. b-Selective epoxidation of D5-unsaturated steroids, Tetrahedron Lett. 37 (1996) 687e690.

[26] L. Maguire, M. Konoplyannikov, A. Ford, A.R. Maguire, N.M. O’Brien,Comparison of the cytotoxic effects of beta-sitosterol oxides and a cholesteroloxide, 7beta-hydroxycholesterol, in cultured mammalian cells, Br. J. Nutr. 90(2003) 767e775.

[27] G. Lizard, S. Monier, C. Cordelet, L. Gesquière, V. Deckert, S. Gueldry, L. Lagrost,P. Gambert, Characterisation and comparison of the mode of cell death,apoptosis versus necrosis, induced by 7beta-hydroxycholesterol and 7-ketocholesterol in the cells of the vascular wall, Arterioscler. Thromb. Vasc.Biol. 19 (1999) 1190e1200.

[28] K. Koschutnig, S. Heikkinen, S. Kemmo, A.M. Lampi, V. Piironen, K.H. Wagner,Cytotoxic and apoptotic effects of single and mixed oxides of beta-sitosterolon HepG2 cells, Toxicol. In Vitro 23 (2009) 755e762.

[29] A. Sevanian, J. Berliner, H. Peterson, Uptake, metabolism, and cytotoxicity ofisomeric cholesterol-5,6-epoxides in rabbit aortic endothelial cells, J. LipidRes. 32 (1991) 147e155.

[30] T.J. Berrodin, Q. Shen, E.M. Quinet, M.R. Yudt, L.P. Freedman, S. Nagpal,Identification of 5a, 6a-epoxycholesterol as a novel modulator of liver Xreceptor activity, Mol. Pharmacol. 78 (2010) 1046e1058.

[31] B.A. Janowski, M.J. Grogan, S.A. Jones, G.B. Wisely, S.A. Kliewer, E.J. Corey,D.J. Mangelsdorf, Structural requirements of ligands for the oxysterol liver Xreceptors LXRalpha and LXRbeta, Proc. Natl. Acad. Sci. 96 (1999) 266e271.

[32] C. Song, R.A. Hiipakka, S. Liao, Auto-oxidized cholesterol sulfates are antago-nistic ligands of liver X receptors: implications for the development andtreatment of atherosclerosis, Steroids 66 (2001) 473e479.

[33] P. de Medina, M.R. Paillasse, G. Segala, M. Poirot, S. Silvente-Poirot, Identifi-cation and pharmacological characterization of cholesterol-5,6-epoxidehydrolase as a target for tamoxifen and AEBS ligands, Proc. Natl. Acad. Sci.107 (2010) 13521e13525.

[34] A. Sevanian, L.L. McLeod, Catalytic properties and inhibition of hepaticcholesterol-epoxide hydrolase, J. Biol. Chem. 261 (1986) 54e59.

[35] D. Rossi, G. Gaidano, Messengers of cell death: apoptotic signaling in healthand disease, Haematologica 88 (2003) 212e218.