at SciVerse ScienceDirect

Biochimie 95 (2013) 518e530

Contents lists available

Biochimie

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

Research paper

Biological activities of the LXRa and b agonist, 4b-hydroxycholesterol,and of its isomer, 4a-hydroxycholesterol, on oligodendrocytes: Effectson cell growth and viability, oxidative and inflammatory status

Thomas Nury a, Mohammad Samadi b, Alexis Varin c, Tatiana Lopez c, Amira Zarrouk a,d,Mohamed Boumhras a,e, Jean-Marc Riedinger f, David Masson c, Anne Vejux a,Gérard Lizard a,*

aUniversité de Bourgogne, Equipe ‘Biochimie du peroxysome, inflammation et métabolisme lipidique’ EA 7270/INSERM, Faculté des Sciences Gabriel,6 Bd Gabriel, 21000 Dijon, FrancebDépartement de Chimie, LCP e A2MC, Université de Lorraine, Metz, FrancecCentre de Recherche INSERM866, ‘Lipides, Nutrition, Cancer’, Equipe ‘Protéines de transfert des lipides et métabolisme des lipoprotéines’, Dijon, FrancedUniversité de Monastir, Faculté de Médecine, Avenue Avicenne, Laboratoire de Biochimie- UR ‘Nutrition Humaine et Désordres Métaboliques’,Monastir, Tunisiae Laboratoire de Biochimie Métabolique et Neurosciences, Université Hassan 1er, Settat, MoroccofCentre de Lutte Contre le Cancer GF Leclerc, Laboratoire de Biologie Médicale, Dijon, France

a r t i c l e i n f o

Article history:Received 30 April 2012Accepted 15 November 2012Available online 7 December 2012

Keywords:Cell death4a-hydroxycholesterol4b-hydroxycholesterolInflammationLXROligodendrocyte (158N and primaryculture)OrganellesOxidation

* Corresponding author. Tel.: þ33 380 39 62 56; faE-mail address: Gerard.Lizard@u-bourgogne.fr (G.

0300-9084/$ e see front matter � 2012 Elsevier Mashttp://dx.doi.org/10.1016/j.biochi.2012.11.013

a b s t r a c t

The biochemical and biological properties of 4b-hydroxycholesterol and of its isomer, 4a-hydrox-ycholesterol, are not well known. So, we determined the ability of 4a- and 4b-hydroxycholesterol to reactwith LXRa and LXRb, and we characterized the activities of these oxysterols on oligodendrocytes whichare myelin synthesizing cells. The effects of 4a- and 4b-hydroxycholesterol were studied on 158N murineoligodendrocytes to assess their activities on cell growth and viability, oxidative and inflammatory status.To this end different parameters were used: cell counting with trypan blue; identification of dead cellsand cell cycle analysis with propidium iodide; evaluation of mitochondrial depolarization, lysosomalmembrane integrity, actin depolimerization, nuclear morphology, and superoxide anion production afterstaining with JC-1, acridine orange, rhodamine-phalloidin, Hoechst 33342, and dihydroethidium,respectively; evaluation of ultrastructural changes by transmission electron microscopy, and cytokinequantification with a cytometric bead array. Only 4b-hydroxycholesterol is a LXRa and b agonist. Nocytotoxic effects were found with 4a-hydroxycholesterol except a slight inhibition of cell growth atelevated concentrations. At high concentrations, 4b-hydroxycholesterol was not only able to inhibit cellgrowth, but also to induce cell death associated with a loss of mitochondrial transmembrane potential,dysfunctions of lysosomal membrane integrity, and superoxide anion overproduction. These side effectswere lower than those observed with 7-ketocholesterol and 25-hydroxycholesterol used as positivecontrols. On oligodendrocyte murine primary cultures, only lysosomal membrane integrity was slightlyaffected under treatment with 4a- and 4b-hydroxycholesterol. So, 4a- and 4b-hydroxycholesterol havedifferent biological activities. Their ability to induce cytotoxic effects on oligodendrocytes can beconsidered as weak comparatively to 7-ketocholesterol and 25-hydroxycholesterol.

� 2012 Elsevier Masson SAS. All rights reserved.

1. Introduction

The brain is the most cholesterol-rich organ in the body andcontains about 25% of the total amount: 70% of this cholesterolportion is present in myelin, a lamellar arrangement of

x: þ33 380 39 62 50.Lizard).

son SAS. All rights reserved.

oligodendrocyte plasma membrane which acts as an electricinsulator along the length of axons, 20% in glial cells, especiallyastrocytes and microglia, 10% in neurons [1]. Since the bloodebrainbarrier efficiently prevents cholesterol uptake from the circulationinto the brain, de novo synthesis is responsible for almost allcholesterol present in this organ [2]. Therefore, cholesterol leveland turnover, and cholesterol oxide derivatives (also named oxy-sterols), which are 27 carbons derivatives of cholesterol containingadditional oxygen atoms on the steroid’s nucleus or on the side

Fig. 1. Planar structures of 4b-hydroxycholesterol (4bOHC) and 4a-hydroxycholesterol(4aOHC).

T. Nury et al. / Biochimie 95 (2013) 518e530 519

chain are suspected to play major roles in neurological disorderssuch as Alzheimer disease, Multiple Sclerosis, and Huntingtondisease [3,4]. In the brain, oxysterols can be formed enzymatically,mostly by cytochrome P450 enzymes [5,6], and also non-enzymatically by reaction with oxygen [6,7]. Oxysterols haveattracted great interest because of their numerous biologicalactivities depending on the oxidation site [8]. They can be impli-cated in gene regulation via the nuclear receptor liver X receptor(LXR) [9,10], and in cell death induction associated or not withinflammatory and/or oxidative processes [11e14]. It has also beenreported that LXR and oxysterols can promote ventral midbrainneurogenesis in vivo and in human embryonic stem cells, under-lying the important role of LXR and oxysterols in brain develop-ment [15]. Moreover, in human brain, 24S-hydroxycholesterol(24SOHC) is a major oxysterol, which is formed from cholesterol bya specific cytochrome P450 enzyme, cholesterol 24-hydroxylase(CYP46A1) [1,16]. It has also been shown by gas chromatography/mass spectrometry (GC/MS) that Schwann cells and sciatic nervescontain 24SOHC, 25-hydroxycholesterol (25OHC), and 27-hydroxycholesterol (27OHC) [17]. As 24SOHC and 25OHC, whichare potent LXRa agonists [18] can induce cell death on neural cells[19e22], it cannot be excluded that they could also favor cell deathof myelin synthesizing cells (oligodendrocytes in the centralnervous system, and Schwann cells at the peripheral level), andtherefore contribute to demyelination. In human plasma, thequantitatively dominating oxysterols are 27OHC, 24SOHC, 7a-hydroxycholesterol and 4b-hydroxycholesterol (4bOHC) also foundin the brain [23e25]. The average plasma concentration of 4bOHCin 125 healthy Swedish volunteers was 30 ng/mL with a range of10e60 ng/mL [26]. The mean half-life of injected deuterium-labelled 4bOHC is 62 h, which is significantly longer than forother related oxysterols [27], and it was demonstrated that 4bOHC(but not its isomer 4a-hydroxycholesterol (4aOHC)) was formedfrom cholesterol by cytochrome P4503A4/5 (CYP3A4/5) [26,28].There are also interethnic differences in CYP3A4/5 activity [29], andplasma concentration of 4bOHC is highly increased (250e600 ng/mL) by anti-epileptic drugs (carbamazepine, phenytoin, pheno-barbital), which are potent inducers of CYP3A4/5 [26,30]. In vitroexperiments have also shown that 4bOHCwas able to activate LXRa[18,31]. However, the biochemical and biological characteristics of4aOHC and 4bOHC are still not well known.

Therefore, in the present study, we precised the capability of4aOHC and 4bOHC to activate LXRa and LXRb. To this end,comparatively with molecules which are known to activate LXRaand LXRb (T0901317, 25OHC) [18,31], we determined the ability of4bOHC and of its isomer (4aOHC) to react with LXRa and LXRbusing reporter assays. Moreover, in neurodegenerative diseases, asincreased permeability of the bloodebrain barrier can favor theentrance of plasmatic compounds into the brain [32,33], whichcould in turn induce oligodendrocytes damages [3], we character-ized the biological activity of 4aOHC and 4bOHC on oligodendro-cytes. To this end, comparatively with oxysterols (7KC, 25OHC)known as potent inducer of cell death on various cell types [12],such as those of the nervous system [19,22], we studied on 158Nmurine oligodendrocytes, which have some characteristics ofdifferentiated oligodendrocytes [34], and on murine oligodendro-cyte primary cultures, the effects of 4aOHC and 4bOHC on cellgrowth and viability, oxidative and inflammatory status.

2. Materials and methods

2.1. Materials

The following oxysterols, 7-ketocholesterol (7KC) and 25-hydroxycholesterol (25OHC) were from Sigma. T0901317, a potent

synthetic LXRa and LXRb agonist [35], was from Sigma. Thefollowing fluorochromes, propidium iodide, rhodamine-phalloidin,acridine orange, dihydroethidium, and Hoechst 33342 were fromSigma; JC-1 was from Invitrogen/Life Technologies. The polyclonalantibody raised against proteolipid protein (PLP), which is themajormyelinprotein [34], was fromAbcam. The secondary antibody (488-Alexa goat anti-rabbit) was from Invitrogen/Life Technologies.

2.2. Synthesis of 4a- and 4b-hydroxycholesterol

4a-hydroxycholesterol (4aOHC) and 4b-hydroxycholesterol(4bOHC) (Fig. 1) were prepared as previously described [36,37]. Theproducts were purified by column chromatography on silica gelusing hexaneeethyl acetate as eluent. The purity of these oxy-sterols was determined to be over 99% by H1eC13 NMR.

2.3. Cells and cell treatments

Murine oligodendrocytes (158N), with some characteristics ofdifferentiated oligodendrocytes, were immortalized with the SV40large T-antigen derived from Tabby male (Ta/Y) control mice[34,38,39]. Cells were seeded at 120,000 cells per well in 12 wellsmicroplates or at 240,000 cells per well in 6 wells microplatescontaining 1 and 2 mL of culture medium, respectively, constitutedby Dulbecco’s modified Eagle medium (DMEM) (Lonza) supple-mented with 5% (v/v) heat-inactivated fetal bovine serum (FBS)(Pan Biotech) and 1% antibiotics (100 U/mL penicillin, 100 mg/mLstreptomycin) (Pan Biotech). Cells were incubated at 37 �C ina humidified atmosphere containing 5% CO2, and passaged twicea week. At each passage, 158N cells were trypsinized with a (0.05%trypsine0.02% EDTA) solution (Pan Biotech).

The conditions of treatment of 158N cells with the oxysterols(4aOHC, 4bOHC, 7KC, 25OHC) were realized as previously defined[40]. Initial concentrations of oxysterols (800 mg/mL ¼ 2 mM) wereprepared as previously described for 4bOHC, 25OHC and 7KC: 1 mgoxysterols þ 50 mL absolute ethanol þ 1.2 mL culture medium[41,42]. As 4aOHC was more difficult to solubilize than the otheroxysterols used, initial concentration of 4aOHC (400 mg/mL ¼ 1 mM) was prepared as follows: 1 mg oxysterol þ 100 mLabsolute ethanol þ 2.4 mL culture medium. Briefly, after plating158N cells for 24 h, cells were further treated for 24 h with variousconcentrations of these compounds.

2.4. LXRa and LXRb reporter assays

To evaluate the potency of 4bOHC and 4aOHC to activate LXRaand/or LXRb comparatively to the LXR agonists (25OHC, T0901317)[18,31,35], HEK293T cells were seeded in triplicate in 96 well platesusing the Lipofectamine 2000 lipofection protocol (Invitrogen/LifeTechnologies).

For LXRa activation assay, each well was transfected with 60 ngof reporter plasmid (TkpGL3 LXRE, LXRE x2), 60 ng of pCMV-b-

T. Nury et al. / Biochimie 95 (2013) 518e530520

Galactosidase as a reference plasmid, and 60 ng of pSG5 humanLXRa for a total DNA content of 180 ng per well.

For LXRb activation assay, each well was transfected with 110 ngof reporter plasmid (TkpGL3 LXRE-RARa, LXRE x2), 50 ng of pCMV-b-Galactosidase as a reference plasmid, and 160 ng of pCMX LXRbfor a total DNA content of 320 ng per well.

The day following transfection, cells were treated withincreasing concentrations of 4aOHC, 4bOHC, and 25OHC (0.78, 7.81,and 31.25 mM) and T0901317 (1 nM,10 nM, and 10 mM) for 24 h, andthen lysed using 20 mL of Passive Lysis Buffer (Promega) per well forLXRa activation assay and 50 mL of Lysis Buffer (Thermo Scientific)per well for LXRb activation assay. Luciferase activity was assessedin 10 mL of each lysate with 100 mL of reconstituted Luciferase AssayBuffer (Promega) for LXRa activation assay and 30 mL of each lysatewith 50 mL of reconstituted Luciferase Assay Buffer (ThermoScientific) for LXRb activation assay. The luciferase activity wasdetermined by using a luminometer (Wallac 1420 MultilabelCounter; Perkin Elmer). b-Galactosidase activity was determined in10 mL or 20 mL (for LXRa and LXRb activation assay respectively) ofeach lysate by measuring hydrolysation of Chlorophenolred-b-D-Galactopyranoside (CPRG; Roche Diagnostics) at 560 nm usinga spectrophotometer (Wallac 1420 Multilabel Counter; PerkinElmer) and used to normalize the luciferase activity.

2.5. Cell counting and determination of IC50 and cell morphology

After trypsinization, 158N cells were centrifuged and resus-pended in culture medium. The total number of viable cells wasdetermined in the presence of trypan blue after 24 h of treatmentwith 4aOHC, 4bOHC, and 25OHC, and 7KC in a range of concen-trations from 2.5 mM to 160 mM under an inverted phase contrastmicroscope Diaphot (Nikon). This range of concentrations waschosen in order to calculate with accuracy the half maximalinhibitory concentration (IC50) corresponding to the concentrationrequired to reduce by 50% the number of viable cells [8,41].

2.6. Evaluation of cell death and determination of LD50

At the end of the incubation times, with or without 4aOHC,4bOHC, 25OHC, and 7KC in a range of concentrations from 2.5 mM to160 mM, 158N cells were stained with 5 mg/mL of propidium iodide(PI), which enters dead cells only [43]. The concentrations of thecompounds were chosen in order to determine with accuracy the‘lethal dose 50%’ (LD50) corresponding to the amount of compoundrequired to kill 50% of the test population. Fluorescence of PI wascollected on a logarithmic scale of fluorescence using a 590/10 nmbandpass filter on a Galaxy flow cytometer (Partec). Ten thousandcells were acquired for each sample. Data were analyzed withFlomax software (Partec).

2.7. Flow cytometric analysis of the cell cycle

Flow cytometric analysis of the cell cycle was realized onadherent and non-adherent cells collected by trypsinization. Cellswere pooled, washed with PBS, and 1e2 � 106 cells were stainedwith PI as previously described [44]. Briefly, cells were resuspendedin 80% cold ethanol (2 h, �20 �C), washed in PBS, and resuspendedin 300 mL of PBS containing 80 mg/mL PI and 200 mg/mL RNase A.After 1 h of incubation (37 �C), 1e2 mL of PBS were added, and flowcytometric analyses were performed on a Galaxy flow cytometer(Partec). Fluorescence of PI was collected using a 590 � 10 nmbandpass filter, and measured on a linear scale. 10,000 cells wereacquired and data were analyzed with Flomax software (Partec).

2.8. Measurement of transmembrane mitochondrial potential withJC-1

Loss of transmembranemitochondrial potential early occurs in thechronology of event leading to cell death, and it can have variousconsequences on cell metabolism. This parameter is thereforeimportant to consider for the characterization of cytotoxic activities.Variations of the mitochondrial transmembrane potential (DJm)were measured with 5,50,6,60-tetrachloro-1,10,3,30-tetraethylbenzi-midazolylcarbocyanine iodide (JC-1; Invitrogen/Life Technologies),which allows the percentage of cellswith lowDJm to be determined.JC-1 is a cationic dye that exhibits potential-dependent accumulationin mitochondria, indicated by a fluorescence emission shift fromgreen (w525 nm) to red (w590 nm). Consequently, mitochondrialdepolarization is indicated bya decrease in redfluorescence andbyanincrease in green fluorescence. In viable cells, the potential-sensitivecolor shift is due to concentration-dependent formation of red fluo-rescent J-aggregates in mitochondria [45]. JC-1 stock solution wasprepared at 1mg/mL in DMSO, stored in the dark at�20 �C, and usedat 1 mg/mL on the cell suspension adjusted at 1�106 cells/mL of PBS.Flowcytometric analyseswere performedon aGalaxyflowcytometer(Partec). The green fluorescence of JC-1 was collected ona 520� 10 nm bandpass filter, and the red fluorescence on a 630 nmlong pass filter. Ten thousand cells were acquired for each sample.Data were analyzed with FlowJo software (Tree Star Inc.).

2.9. Measurement of lysosomal membrane integrity with acridineorange

The analysis of lysosomalmembrane integrity permits to evaluatefunctional lysosomeswith active protons pumps. In these lysosomes,active proteases are unable to diffuse in the cytoplasm and to inducevarious cell damages. To measure lysosomal membrane integrity,acridine orange (AO) is widely used [46]. AO is a weak base whichaccumulates in its charged form within lysosomes of living cellsbecause of the low lysosomal pH, and which produces an orange/redfluorescence when excited by a blue light [47]. During prolongedexposure to cytotoxic agents, the orange/red fluorescence of AOdecreased markedly [48]. The shift in AO fluorescence from orange/red to green reflects leakage and redistribution of AO from thelysosomes, indicating impairment of the lysosomal membranes and/or of the ability of the lysosomes tomaintain low pH. So, AO iswidelyused to perform lysosomal integrity measurements, and thepercentage of AO-negative cells (which do not emit an orange/redfluorescence) permits quantification of the percentage of cells withdestabilized lysosomes [48]. After 24 h of culture in the absence orpresence of 4aOHC, 4bOHC, 25OHC, and 7KC, cells were stained withAO (2 mg/mL in PBS) for 15min at 37 �C, washed, resuspended in PBS,and directly analyzed by flow cytometry. The fluorescent signalsweremeasured on a Galaxy flow cytometer (Partec), and the orange/red fluorescence of AO was collected through a 580 � 10 nm band-pass filter. For each sample, fluorescence was quantified on 10,000cells, and data were analyzed with the Flomax software (Partec).

2.10. Evaluation of cytoskeletal actin organization andpolymerization with rhodamine-phalloidin

The organization of the cytoskeleton was evaluated by fluores-cence microscopy and flow cytometry after staining withrhodamine-phalloidin, which stains F-actin [49].

Microscopical observations were performed on 158N cellsgrown on glass coverslips. At the end of the treatment, cells werewashed with PBS, fixed with a 2% paraformaldehyde/PBS solution(10 min, room temperature), washed in PBS, and permeabilizedwith PBS/0.05% saponin/10% FCS (30 min, room temperature) and

T. Nury et al. / Biochimie 95 (2013) 518e530 521

incubated (30 min, room temperature, in the dark) in a 0.1 mMrhodamine-phalloidin solution prepared in PBS/0.05% saponin/10%FCS. Cell nuclei were counterstainedwith Hoechst 33342 (1 mg/mL).The slides were then mounted in fluorescent mounting medium(Dako), and stored in the dark at 4 �C until examination with anAxioskop fluorescent Zeiss microscope. Digitalized images wereobtained with an AxioCam digital camera (Zeiss).

Flow cytometric analyses were realized in order to quantify F-actin content. To this end, cells were trypsinized, washed andmixedin PBS andfixed in 2% paraformaldehyde/PBS solution (15min, roomtemperature). After washing in PBS, cells were permeabilized withPBS/0.05% saponin/10% FCS (30 min, room temperature). Further,cells were incubated in a 0.1 mM rhodamine-phalloidin solutionprepared in PBS/0.05% saponin/10% FCS (30 min, room temperature,in the dark). Afterwashing in PBS, cellswere resuspended in PBS andimmediately analyzedbyflowcytometryon aGalaxyflowcytometer(Partec). Fluorescencewas collected trough a 580� 10 nm bandpassfilter. For each sample, fluorescence was quantified on 10,000 cells,and data were analyzed with the Flomax software (Partec).

2.11. Analysis of nuclear morphology by staining with Hoechst33342

Hoechst staining was used to determine the nuclearmorphology by fluorescence microscopy under ultraviolet light. Inthese conditions, dead cells characterized by condensed and/orfragmented nuclei as well as by swollen nuclei, are considered asapoptotic and necrotic cells, respectively [50], whereas living cellshave round and regular nuclei [43]. Hoechst 33342 (1 mg/mL indistilled water) was added on cell deposits at 1 mg/mL of approxi-mately 40,000 cells applied to glass slides by cytocentrifugation(5 min, 15,000 rpm) with a cytospin 2 (Shandon). The morpho-logical aspect of the nuclei was observed with an Axioskop fluo-rescent Zeiss microscope; 300 cells were examined per sample.

2.12. Transmission electron microscopy

Transmission electron microscopy is a valuable tool for charac-terizing various cytotoxic effects [43]. It was used to visualize theeventual ultrastructural modifications induced by 4a- and 4b-hydroxycholesterol on 158N cells, and was conducted as previouslydescribed [51].

2.13. Measurement of superoxide anion production withdihydroethidium

The whole intracellular production of superoxide anion (O2.�) was

detected with dihydroethidium (DHE; Invitrogen/Life Technologies)which is a non-fluorescent compound which can diffuse through cellmembranes, and which is rapidly oxidized in ethidium under theaction of O2

.� [52]. DHE was initially prepared at 10 mM in dimethylsulfoxide (DMSO), andusedat2mMfinal concentrationoncell samplesof 1 � 106 cells/mL of PBS. Cells were incubated for 15 min at 37 �Cbeforeflowcytometric analysis performed on aGalaxyflowcytometer(Partec). Fluorescence of ethidiumwas collected using a 590 � 10 nmbandpass filter, and measured on a logaritmic scale. 10,000 cells wereacquired and data analyzed with Flomax software (Partec).

2.14. Multiplexed flow cytometric analysis of inflammatorycytokines

The production of inflammatory cytokines was investigated inthe culture medium of 158N cells which were either untreated ortreated for 24 hwith different concentrations of oxysterols (4aOHC,4bOHC, 25OHC, and 7KC) (2.5e100 mM) by using a flow cytometric

bead-based assay, i.e., the multiplexed Cytometric Bead Array(CBA), as previously described [53]. The CBA human inflammationkit (BD Biosciences) consists of pools of beads (diameter: 7.5 mm)dyed with six different concentrations of the FL3 fluorophore(excitation and emission wavelengths at 488 and above 600 nm,respectively) and coupled to antibodies raised against one of sixcytokines (IL-1b, IL-6, IL-8, IL-10, TNFa and IL-12p70), so that eachdiscrete population of specific antibody coupled bead is unique byits FL3 intensity. Captured cytokines are detected via a directimmunoassay using six different antibodies coupled to phycoery-thrin emitting at 585 nm (FL2). Data were acquired with Flomaxsoftware (Partec) on a Galaxy flow cytometer (Partec) equippedwith a laser emitting at 488 nm, and analyzed using FCAP arraysoftware (BD Biosciences). Forward vs. side-scatter gating wasemployed to exclude any sample particles other than the 7.5 mmbeads. Data were displayed as two colors dot plots [FL2 (PE):bandpass 580 � 10 nm vs. FL3 (beads): long pass 665 nm] so thatthe six discrete FL3 microparticles dye intensities were distributedalong the Y-axis. Nine points standard curves ranging from 20 to5000 pg/mL were obtained by serial dilution of the reconstitutedlyophilized cytokines standard, prepared fresh daily. Cytokinesconcentrations were determined from these standard curves.

2.15. Analysis of LXRa and LXRb expression by RT-qPCR

Total RNA from 158N cells was extracted and purified using theRNeasy Mini Kit (Qiagen) with a 20 min DNAse treatment (Qiagen).Total RNA concentration was measured with TrayCell (Hellma Paris)and the purity of nucleic acids was controlled by the ratio of theabsorbance 260 nm/280 nm (1.8e2.2). Onemicrogram total RNAwasused for reverse transcription with the iScript� cDNA Synthesis Kit(Bio-Rad) according to the following reaction protocol: 5 min at25 �C,1 h at 42 �C, 5min at 85 �C andhold at 4 �C. cDNAwas amplifiedusing the MESA GREEN qPCR MasterMix Plus for SYBR� Assay w/fluorescein (Eurogentec). All PCR reactions were performed on anApplied Biosystem Step One QPCR machine (Life Science Technolo-gies). Primerswere designed to generate a PCR amplification productof 100e200 bp and were selected according to indications providedby PrimerBank (http://pga.mgh.harvard.edu/primerbank/) for 36B4and LXRb and by Primer3 (http://biotools.umassmed.edu/bioapps/primer3_www.cgi) for LXRb. The primer sequences (36B4: forward,50- CGACCTGGAAGTCCAACTAC-30 and reverse, 50- ATCTGCTG-CATCTGCTTG-30; LXRa: forward, 50- GCTCTGCTCATTGCCATCAG-30

and reverse, 50- GTTGCAGCCTCTCTACTTGGA-30; LXRb: forward, 50-ATGTCTTCCCCCACAAGTTCT-30 and reverse 50- GACCACGATGTAGG-CAGAGC-30). All primers were checked by DNA calculator (http://www.sigma-genosys.com/calc/DNACalc.asp) in case of any dimer orsecondary structure formation. The 36B4 housekeeping gene wasselected as a reference gene. PCR reaction mix was realized with300 nM of each primer, 12.5 mL MESA GREEN qPCR MasterMix Plusfor SYBR� Assay w/fluorescein and 5 mL of DNA in a 25 mL volumeadjustedwith RNase freewater. Samples aliquots for each timeswerepooled together to realize a serial dilution for establish standardcurve. The specificity of the amplicon was evaluated by meltingcurves (from 60 �C to 95 �C) and the efficiency (E) was assessed bystandard curve (E¼ 10(�1/slope)) calculated by StepOne software froma standard curve. All PCR reactions were performed on an AppliedBiosystems Step One QPCR (MeterTaq polymerase activation 20 s at90 �C, 40 cycles 3 s at 95 �C and 30 s at 60 �C).

2.16. Evaluation of 4a-hydroxycholesterol and 4b-hydroxycholesterolcytotoxicity on murine oligodendrocyte primary cultures

Primary cultures and conditions of treatments of murineoligodendrocytes were performed as follows.

Fig. 2. Effects of 4a-hydroxycholesterol and 4b-hydroxycholesterol on the activationof LXRa- and LXRb-dependent transcription. HEK293T cells were seeded in triplicatein 96 well plates using the Lipofectamine 2000 lipofection protocol (Invitrogen/LifeTechnologies). To evaluate LXRa activation, each well was transfected with a reporterplasmid (TkpGL3 LXRE, LXRE x2), pCMV-b-Galactosidase as a reference plasmids, andpSG5 human LXRa. To evaluate LXRb activation, each well was transfected witha reporter plasmid (TkpGL3 LXRE-RARa, LXRE x2), pCMV-b-Galactosidase as a refer-ence plasmid, and pCMX LXRb. The day following transfection, cells were treatedwith increasing concentration of 25-hydroxycholesterol (25OHC), 4b-hydrox-ycholesterol (4bOHC), 4a-hydroxycholesterol (4aOHC) and T0901317 for 24 h. Aftercell lysis, luciferase activity was assessed and normalized using b-Galactosidaseactivity. Data shown are mean � SD from three independent experiments. Signifi-cance of the difference between untreated and treated cells (ManneWhitney test; *P < 0.05 or less).

T. Nury et al. / Biochimie 95 (2013) 518e530522

One to two-days-old neonatal BALB/cJRj mice pups (Janvier-Europe) were washed with 70% ethanol and quickly beheaded.Briefly, the brains were removed and placed in a Petri dish con-taining 10 mL DMEM medium. Meninges and blood vessels wereremoved by rolling on a sterilized filter paper. The cleaned brainswere removed and placed in a Petri dish containing 10 mL of freshmedium (DMEM/10% FCS/1% antibiotics). Brain tissue was gentlydiced into small pieces and crushed on a 100 mm cellular sieve. Cellswere harvested in 10 mL of DMEM/10% FCS/1% antibiotics ina 50 mL tube. These 10 mL were used to seed one 12 wells plate(800 mL/well). Primary cultures were incubated in wet atmosphereat 37 �C, 5% CO2. After 4 days of culture, the mediumwas renewed,and then changed twice a week.

After 10 days of culture, different treatments were realized.Oxysterol concentrations were chosen based on data obtained on158N cells. Oligodendrocyte primary cultures were realized in 12wells plate for 24 h in the absence or in the presence of 4aOHC and4bOHC (25, 50 mM), or in the presence of 25OHC and 7KC (50 mM)used as positive controls. Cell counting in the presence of trypanblue, measurement of transmembrane mitochondrial potentialwith JC-1, of lysosomal membrane integrity with acridine orange(AO), and of superoxide anion production with dihydroethidium(DHE) were performed as described on 158N cells.

2.17. Detection of PLP expression by indirect immunofluorescence

Murine oligodendrocytes primary cultures were washed withPBS, fixed with a 2% paraformaldehyde/PBS solution (10 min, roomtemperature), washed in PBS, and permeabilized with PBS/0.05%saponin/10% FCS (30 min, room temperature). Cells were thenincubated with rabbit polyclonal antibody raised against PLPdiluted 1/100 in PBS/0.05% saponin/10% FCS (1 h, room tempera-ture). Further, cells were washed in PBS and incubated withsecondary antibodies: a 488-Alexa goat anti-rabbit used at 1/400prepared in PBS/0.05% saponin/10% FCS (30 min, room tempera-ture, in the dark). After washing in PBS, nuclei were counterstainedwith 1 mg/mL Hoechst 33342, slides were mounted in fluorescentmounting medium (Dako), observations were made under anAxioskop Zeiss microscope, and digitalized images were obtainedwith an Axiocam Zeiss camera (Zeiss).

2.18. Statistical analysis

Analyses were carried out with WinSTAT� for Microsoft� Excel(version 2012.1). Since the biological variables were not normallydistributed, nonparametric ManneWhitney test was used. P values<0.05 were considered significant in all tests.

3. Results

3.1. In vitro measurement of LXRa and LXRb activation by 4a-hydroxycholesterol and 4b-hydroxycholesterol

To determine whether 4aOHC and 4bOHC activate LXRa and/orLXRb-dependent transcription, we performed a luciferase reporterassay, using a construct containing two copies of a synthetic LXREfused with a thymidine kinase promoter (Fig. 2). As positivecontrols, we used the synthetic LXR agonist: T0901317, and thenatural LXR agonist 25OHC. The LXR agonists strongly activated thereporter in a dose manner dependent concentration. Cell treatmentwith 4bOHC increased luciferase activity in a dose-dependentmanner. At the opposite, 4aOHC is unable of inducing an increase inluciferase activity. So, 4bOHC is an LXR a and b agonist but not4aOHC.

3.2. Effects of 4a-hydroxycholesterol and 4b-hydroxycholesterol oncell growth and viability of 158N murine oligodendrocytes

158N murine oligodendrocytes presenting some characteristicsof well differentiated murine oligodendrocytes (Fig. 3A) andexpressing high levels of LXRb (Fig. 3B), were cultured in 12 wellsmicroplates (initially plated at 120,000 cells/well) for 24 h withoxysterols (4aOHC, 4bOHC) in a wide range of concentrations from1 to 64 mg/mL (2.5e160 mM) in order to precisely define their IC50and LD50. In those conditions, ethanol concentration in the culturemedium did not exceed 0.64%. The effects of these oxysterols on cellgrowth and viability were compared with those of 25OHC and 7KCwhich are known to induce cell death on neural cells [19,21,40].

As determined from the curves (total number of living cellsversus oxysterol concentrations) by cell counting with trypan blue,the mean concentrations required to reduce by 50% the number ofcells (IC50) were around 20 mM for 4aOHC and 4bOHC (Fig. 4A andC), and 5e10 mM for 25OHC and 7KC (Fig. 4E and G).

In addition, comparatively to untreated cells (control), whendead cells were identified by flow cytometry after staining with PI,

Fig. 4. Effects of 4a-hydroxycholesterol, 4b-hydroxycholesterol, 25-hydroxycholesterol, andcells were cultured for 24 h in the absence or presence 4a-hydroxycholesterol (4aOHC), 4(7KC) in a range of concentrations from 2.5 to 160 mM. Cell growth, evaluated by the totaof trypan blue, and the percentage of dead cells was quantified by flow cytometry afterindependent experiments. Significance of the difference between untreated- and oxysterobetween absolute control and vehicle (ethanol: 0.64%; corresponding to the maximum fina

Fig. 3. Antigenic and LXR profiles of 158N murine oligodendrocytes. 158N cells havesome characteristics of mature oligodendrocytes (A). Whereas 158N cells only slightlyexpress LXRa mRNAs (Ct ¼ 27.74 � 0.16), they express LXRb mRNAs(Ct ¼ 23.30 � 0.08). The mRNA levels were measured using real-time RT-qPCR andnormalized to 36B4. Data shown are expressed as 2 � DCt and are the mean � SD oftwo or three experiments. Significance of the difference between LXRa and LXRbmRNA levels (ManneWhitney test; * P < 0.05 or less).

T. Nury et al. / Biochimie 95 (2013) 518e530 523

the highest proportions of cell death obtained under treatmentwith 4aOHC and 4bOHC were 20 � 10% and 35 � 15%, respectively(Fig. 4B and D). As determined from the curves (% dead cells versusoxysterol concentrations), with 4aOHC and 4bOHC, the concen-trations used do not permit to obtain 50% of dead cells (LD50)(Fig. 4B and D). LD50 was in the range of 20 mM for 25OHC, andaround 20e40 mM for 7KC (Fig. 4F and H). In agreement withprevious data [41,42], no effect of ethanol was observed on cellgrowth and viability. Flow cytometric analysis of the repartition ofthe cells in the different phases of the cell cycle did not showmarked differences between untreated cells, 4aOHC- and 4bOHC-treated cells (Table 1). Thus, at the opposite of 25OHC and 7KC,which reduce cell growth by activation of cell death (highpercentages of PI positive cells simultaneously identified), 4aOHCand 4bOHC reduce cell growthwithoutmarked effects on cell deathinduction.

3.3. Effect of 4a-hydroxycholesterol and 4b-hydroxycholesterol onmitochondrial and lysosomal activities of 158N cells

To evaluate the effects of 4aOHC and 4bOHC on the mitochon-drial and lysosomal status of 158N murine oligodendrocytes,different criteria were taken in consideration: mitochondrialtransmembrane potential (DJm) measured with JC-1, and lyso-somal membrane integrity estimated with acridine orange (AO).The effects of 4aOHC and 4bOHCwere compared to those of 25OHCand 7KC (50 mM) used as positive controls.

7-ketocholesterol on cell growth and viability of 158N murine oligodendrocytes. 158Nb-hydroxycholesterol (4bOHC), 25-hydroxycholesterol (25OHC), and 7-ketocholesteroll number of living cells (A, C, E, G), was determined by cell counting in the presencestaining with propidium iodide (B, D, F, H). Data shown are mean � SD from threel-treated cells (ManneWhitney test; * P < 0.05 or less). No difference was observedl concentration of ethanol used).

Table 1Repartition of oxysterol-treated cells in the different phases of the cell cycle.

G0/G1 (%) S (%) G2/M (%)

Control 52 � 4 17 � 4 32 � 1Vehicle 51 � 3 16 � 4 33 � 24aOHC2.5 mM 57 � 1 15 � 2 29 � 2*12.5 mM 57 � 3 12 � 3 32 � 125 mM 56 � 3 15 � 3 28 � 1*50 mM 58 � 1* 12 � 0* 25 � 8*100 mM 62 � 1* 14 � 2 25 � 1*4bOHC2.5 mM 52 � 1 17 � 1 32 � 112.5 mM 51 � 1 16 � 0 34 � 0*25 mM 53 � 1 14 � 0 33 � 050 mM 54 � 1 17 � 8 30 � 7100 mM 54 � 0 17 � 1 30 � 1*7KC (50 mM) 53 � 2 13 � 5 35 � 325OHC (50 mM) 50 � 1 20 � 2 31 � 1

Cell cycle analysis was performed by flow cytometry on untreated- (control) andoxysterol-treated cells after 24 h of culture. Significance of the difference betweenuntreated- and oxysterol-treated cells (ManneWhitney test; * P < 0.05 or less).

T. Nury et al. / Biochimie 95 (2013) 518e530524

In the range of concentration (2.5e100 mM) associated witha decrease of cell growth under treatment with 4aOHC and4bOHC no effect of 4aOHC was observed on DJm and onlysosomal functions (Fig. 5A and B). However, under treatmentwith 4bOHC significant increased of cells with depolarized mito-chondria and with destabilized lysosomes were detected at 25, 50,and 100 mM (Fig. 5C and D). Under treatment with 25OHC and 7KC,the percentages of cells with depolarized mitochondria and withdestabilized lysosomes were 3e4 times higher than in 4aOHC(50 mM), and 1.5e2 times higher than in 4bOHC (50 mM).

3.4. Effect of 4a-hydroxycholesterol and 4b-hydroxycholesterol onactin organization and polymerization

The evaluation of cytoskeleton organization was realized byfluorescence microscopy and flow cytometry after staining withrhodamine-phalloidin, which stains F-actin, on 158N cells culturedfor 24 h in the absence or in the presence of 4aOHC and 4bOHC(2.5e100 mM), 25OHC (25 mM), and 7KC (25 mM) for 24 h. Undertreatment with 4aOHC and 4bOHC (50 and 100 mM), large

Fig. 5. Effects of 4a-hydroxycholesterol, 4b-hydroxycholesterol, 25-hydroxycholesterol, amembrane integrity on 158N cells. 158N cells were cultured for 24 h in the absence or prconcentrations from 2.5 to 100 mM. 25-hydroxycholesterol (25OHC), and 7-ketocholesterdepolarized mitochondria and with destabilized lysosomes, were determined after staininindependent experiments. Significance of the difference between untreated- and oxysterobetween absolute control and vehicle (ethanol: 0.2%; corresponding to the maximum final

fluorescent spots of rhodamine-phalloidin, which might corre-spond to actin aggregates, were observed in the cytoplasm (Fig. 6A).Flow cytometric analysis of rhodamine-phalloidin-stained cellsallowing to distinguish between actin under its globular form(slightly fluorescent) and under its filamentous form (stronglyfluorescent) was used to evaluate cytoskeleton disorganization byusing the following ratio [(Mean Fluorescence Intensity of oxy-sterol-treated cells)/(Mean Fluorescence Intensity of untreated cells(control))] � 100. In these conditions, the ratio observed in treatedcells were not statistically different comparatively to untreatedcells, or vehicle (0.2% ethanol)-treated cells, excepted under treat-ment with 4bOHC (2.5 mM) (Fig. 6B) suggesting little impact of4aOHC and 4bOHC on cytoskeleton organization.

3.5. Cytological and ultrastructural analyses of 158N cells treatedwith 4a-hydroxycholesterol and 4b-hydroxycholesterol

Cytological, ultrastructural, and antigenic analyses were per-formed on 4aOHC- and 4bOHC-treated 158N cells in order toidentify eventual modifications not revealed by the flow cytometricfunctional tests used to evaluate cell death and associateddysfunctions (staining with PI, JC-1, AO, and rhodamine-phalloidin). To this end, 158N cells were cultured with 4aOHC or4bOHC in a range of concentration from 2.5 to 100 mM for 24 h, and25OHC and 7KC (25 mM)were used as positive controls. After nucleistaining with Hoechst 33342, allowing to distinguish betweenviable cells (cells with round and regular nuclei), apoptotic cells(cells with condensed and/or fragmented nuclei), and necrotic cells(cells with irregular and diffuse nuclei from various sizes andshapes) [43], around 15e25% of apoptotic cells were observedunder treatment with 4aOHC and 4bOHC (2.5e100 mM) (Fig. 6A).Some apoptotic cells were only observed under treatment with 7KC(Fig. 7A). In addition, observations performed by transmissionelectron microscopy revealed the presence of some vacuoles ofvarious sizes and shapes, especially in the cytoplasm of 4bOHC,25OHC, and 7KC; these vacuoles contained more or less cellulardebris, mainly under treatment with 25OHC; no vacuoles weredetected in 4aOHC-treated cells (Fig. 7B). In untreated cells, and0.2% ethanol (vehicle)-treated cells, few apoptotic (or necrotic) cells(5% <) were detected, and no cells with vacuoles were found(Fig. 7A).

nd 7-ketocholesterol on mitochondrial transmembrane potential and on lysosomalesence 4a-hydroxycholesterol (4aOHC), 4b-hydroxycholesterol (4bOHC) in a range ofol (7KC), 50 mM each, were used as positive controls. The percentages of cells, withg with JC-1 and acridine orange, respectively. Data shown are mean � SD from threel-treated cells (ManneWhitney test; * P < 0.05 or less). No difference was observedconcentration of ethanol used).

Fig. 6. Effects of 4a-hydroxycholesterol, 4b-hydroxycholesterol, 25-hydroxycholesterol, and 7-ketocholesterol on cytoskeleton organization on 158N cells. 158N cells were culturedfor 24 h in the absence or presence 4a-hydroxycholesterol (4aOHC), 4b-hydroxycholesterol (4bOHC) in a range of concentrations from 2.5 to 100 mM. 25-hydroxycholesterol(25OHC), and 7-ketocholesterol (7KC), 50 mM each, were used as positive controls. The evaluation of cytoskeleton organization was realized by fluorescence microscopy (A) andflow cytometry (B) after staining with rhodamine-phalloidine, which stains F-actin. Flow cytometric evaluation of cytoskeleton disorganization was estimated by using thefollowing ratio [(Mean fluorescence intensity (MFI) of oxysterol-treated cells)/(Mean fluorescence intensity (MFI) of untreated cells (0))] � 100. Data shown are mean � SD fromthree independent experiments. Significance of the difference between untreated- and oxysterol-treated cells (ManneWhitney test; * P < 0.05 or less). No difference was observedbetween absolute control and vehicle (ethanol: 0.2%; corresponding to the maximum final concentration of ethanol used).

T. Nury et al. / Biochimie 95 (2013) 518e530 525

3.6. Effects of 4a-hydroxycholesterol and 4b-hydroxycholesterol onreactive oxygen species production and cytokine secretion on 158Ncells

To determine the impact of 4aOHC and 4bOHC (2.5e100 mM) onoxidative and inflammatory status of 158N murine oligodendro-cytes, these cells were cultured for 24 h in the absence and in thepresence of these compounds, and 25OHC (50 mM) and 7KC (50 mM)were used as positive controls. Flow cytometric evaluation ofReactive Oxygen Species (ROS) production after staining withdihydroethidium (DHE) did not revealed a stimulation of ROSproduction under treatment with 4aOHC whereas with 4bOHC (25,50, and 100 mM) the percentages of HE positive cells were 2e4times higher than in untreated cells (Fig. 8). The percentages ofHE positive cells observed with 25OHC and 7KC (50 mM) weresimilar or twice higher than with 4bOHC (50 mM) (Fig. 8). In addi-tion, when compared to untreated cells, 4aOHC and 4bOHC (2.5e100 mM) did not enhance IL-6, IL-10, MCP-1, IFNɣ, TNFa, and IL-12p70 secretion measured in the culture medium with the multi-plexed flow cytometric CBAmethod [53], and under treatment with25OHC and 7KC, no stimulation of cytokine secretion was observed(our data not shown).

3.7. Evaluation of the cytotoxic effects of 4a-hydroxycholesterol and4b-hydroxycholesterol on murine oligodendrocyte primary cultures

After 10 days of culture, murine oligodendrocyte primarycultures (Fig. 9A) were treated for 24 h in the absence or presenceof 4aOHC, 4bOHC (25, 50 mM), or with 25-hydroxycholesterol(25OHC), or 7-ketocholesterol (7KC) (50 mM) used as positivecontrols. Interestingly, the expression pattern of LXRa and LXRbobserved in murine oligodendrocyte primary cultures was similarthan in 158N cells: the mRNA level of LXRb was higher than thoseof LXRa (Fig. 9B). To investigate the effects of 4aOHC and 4bOHCon murine oligodendrocyte primary cultures, oxysterol concen-trations were chosen based on the significant effects obtainedwith 4bOHC on mitochondrial transmembrane potential, lyso-somal membrane integrity, and superoxide anions overproductionon 158N cells. In these conditions, small effects of 4aOHC and4bOHC were observed on lysosomal membrane integrity at 50 mMas revealed by an increase of the percentage of cells with desta-bilized lysosomes (Fig. 9CeE). With 7KC and 25OHC, used aspositive control, significant activities were only found on trans-membrane mitochondrial depolarization and/or lysosomal integ-rity at 50 mM (Fig. 9CeE).

Fig. 7. Nuclear morphology and ultrastructural characteristics of 158N murine oligodendrocytes cultured in the absence or in the presence of 4a-hydroxycholesterol, 4b-hydroxycholesterol, 25-hydroxycholesterol, and 7-ketocholesterol. A: The percentage of apoptotic cells (cells with condensed and/or fragmented nuclei) was determined byfluorescence microscopy after staining with Hoechst 33342 on 158N cells cultured for 24 h in the absence or presence 4a-hydroxycholesterol (4aOHC), 4b-hydroxycholesterol(4bOHC) in a range of concentrations from 2.5 to 100 mM. The oxysterols, 25-hydroxycholesterol (25OHC), and 7-ketocholesterol (7KC), 50 mM each, were used as positive controls.Data shown are mean � SD from three independent experiments. Significance of the difference between untreated- and oxysterol-treated cells (ManneWhitney test; * P < 0.05 orless). No difference was observed between absolute control and vehicle (ethanol: 0.2%; corresponding to the maximum final concentration of ethanol used). B: Transmissionelectron microscopy of 158N cells cultured for 24 h in the absence (control cells (0)) or presence of ethanol (0.1%), 4a-hydroxycholesterol (25 mM), 4b-hydroxycholesterol (25 mM),25-hydroxycholesterol (25 mM), and 7-ketocholesterol (25 mM). In untreated cells, vehicle (ethanol 0.1%), and 4a-hydroxycholesterol-treated cells no or few vacuoles (indicated bydark arrows) were observed. In 4b-hydroxycholesterol-, 25-hydroxycholesterol-, and 7-ketocholesterol-treated cells some vacuoles of various sizes and shapes, which can containmore or less cellular debris (especially under treatment with 25-hydroxycholesterol), were observed.

T. Nury et al. / Biochimie 95 (2013) 518e530526

4. Discussion

In some neurodegenerative diseases, increased permeability ofthe bloodebrain barrier can favor the accumulation of peripherallipids in the brain. Thus, in Alzheimer patients, as a consequence ofblood brain barrier damages, there are some evidences of anincreased flux from the plasma into the brain of some steroids suchas phytosterols and 27OHC, which is one of the quantitatively mostimportant oxysterol in the circulation of adults [24,31]. Conse-quently, increased levels of 27OHC in the brain of Alzheimerpatients contribute to the development of the pathology [31]. Byanalogy with 27OHC, 4aOHC and 4bOHC, which are also present atelevated levels in the plasma (especially 4bOHC), could enter intothe brain in neurodegenerative diseases associated with enhancedbloodebrain barrier permeability, and contribute to neuronaldamages. In the context of demyelinating diseases, such as multiplesclerosis and leukodystrophies, it is therefore important to precisethe effects of 4aOHC and 4bOHC on oligodendrocytes, which aremyelin synthesizing cells [54]. Currently, as biochemical and bio-logical properties of 4b-hydroxycholesterol and of its isomer, 4a-hydroxycholesterol, are not well known, we precised the ability of4a- and 4b-hydroxycholesterol to react with LXRa and LXRb, and

we characterized the activities of these oxysterols on 158N murineoligodendrocytes, and on murine oligodendrocyte primarycultures.

In agreement with data previously reported [18], we confirmthat 4bOHC, is a potent LXRa agonist, and we show that this oxy-sterol can also activate LXRb. However, we established that 4aOHCwas unable to activate LXRa and LXRb. This observation supportsthe hypothesis that 4bOHC is probably involved in the regulation ofvarious genes, which remain to be identified, and which are placedunder the control of LXRa and/or LXRb.

Whereas a link between the activation of LXR-dependent genesand the ability to induce cytotoxic effects (modifications of lipidmetabolism, activation of oxidative stress and inflammatoryprocesses, induction of cell death) has been supposed [55], it is alsowell established that non-LXR agonists can also exert potent cyto-toxic activities [12]. It was therefore important to simultaneouslyevaluate the eventual side effects of 4aOHC and 4bOHC, and tocompare their effects with those of 25OHC and 7KC, which are ableto activate or not LXR [18], respectively, and to induce numerousdysfunctions on different cell types, including 158N cells [40].

Noteworthy, 4aOHC had no cytotoxic effects on 158N murineoligodendrocytes, except an inhibition of cell growth at high

Fig. 8. Effects of 4a-hydroxycholesterol, 4b-hydroxycholesterol, 25-hydroxycholesterol, and 7-ketocholesterol on superoxide anion production on 158N cells. 158N cells werecultured for 24 h in the absence or presence of 4a-hydroxycholesterol (4aOHC), 4b-hydroxycholesterol (4bOHC) in a range of concentrations from 2.5 to 100 mM. 25-hydroxycholesterol (25OHC), and 7-ketocholesterol (7KC), 50 mM each, were used as positive controls. The percentages of cells producing superoxide anions (% hydroethidine(HE) positive cells) were determined by flow cytometry after staining with dihydroethidium (DHE). Data shown are mean � SD from three independent experiments. Significance ofthe difference between untreated- and oxysterol-treated cells (ManneWhitney test; * P < 0.05 or less). No difference was observed between absolute control and vehicle (ethanol:0.2%; corresponding to the maximum final concentration of ethanol used).

T. Nury et al. / Biochimie 95 (2013) 518e530 527

concentrations. However, 4bOHC, was not only able to inhibit cellgrowth in the same range of concentrations than 4aOHC, but it alsoslightly increased cell death measured by increased permeabilityto PI.

This inhibition of cell growth, which is a commonpoint betweenthe different oxysterols studied (4aOHC, 4bOHC, 7KC, and 25OHC),was in the following range of order: 4aOHC similar than4bOHC< 25OHC< 7KC. In agreement, with previous investigationsperformed on endothelial cells [41,56], the effect of 7KC was mostpotent than those of 25OHC, and the inhibition of cell proliferationobserved with these oxysterols was associated with an increasedpermeability to PI, which entered dead cell only, when the cyto-plasmic membrane is damaged [43,57]. With 4aOHC lowpercentage of PI positive cells were also observed. However, asthese cells were identified only at very high concentrations, ourdata establish that 4aOHC is not a potent inducer of cell death.With4bOHC, whereas the percentage of PI positive was higher thanwith4aOHC, it however did not exceed 29 � 16% whereas with 7KC and25OHC more than 80% of PI positive cells were identified. Onmurine oligodendrocyte primary cultures no increased of PI posi-tive cells were observed under treatment with 4aOHC and 4bOHC.Therefore, in agreement with data obtained on tumoral cells [8],our data establish that the ability of 4aOHC and 4bOHC to inducecell death of oligodendrocytes can be considered as very weak andweak, respectively, comparatively to 7KC and 25OHC used aspositive control.

In order, to explain the inhibition of cell growth observed undertreatment with the different oxysterols studied, which is inagreement with a decreased incorporation of [3H] thymidine foundon endothelial and arterial smooth muscle cells treated with25OHC and 7KC [56,58], flow cytometric analyses of the repartitionof the cells in the different phases of the cell cycle were performed.On human monocytic cells (THP-1), an accumulation at G2/Mphase, was reported under treatment with 7b-hydroxycholesteroland 25OHC [59]. However, in our experimental conditions, with thedifferent oxysterols used, only minor modifications of the reparti-tion of the cells in the different phases of the cell cycle were foundbetween untreated cells and oxysterol-treated cells. Under treat-ment with 4aOHC and 4bOHC, based on the cytostatic effectobserved (inhibition of cell growth without or with slight increasedpermeability to PI), this could suggest that the time spent by thecells in the different phases of the cell cycle could be increased.With 7KC and 25OHC, based on the induction of cell deathobserved, this could favor the hypothesis that the ability of these

compounds to induce cell death would not depend on the phase ofthe cell cycle considered.

As more or less pronounced organelle damages, especially at themitochondrial and lysosomal level, were reported on various celltypes under treatment with different oxysterols [12,60], it was alsoof interest to precise the ability of 4aOHC and 4bOHC to inducemitochondrial and lysosomal dysfunctions which could have someconsequences on lipid trafficking and metabolism [61,62]. Whereasnone effects were observed with 4aOHC, slight increase of thepercentage of cells with depolarized mitochondria and destabilizedlysosomes were observed with 4bOHC in a range of concentrationfrom 25 to 100 mM as well as a slight increase of the number ofvacuoles revealed by transmission electron microscopy. However,these effects were lower than those induced by 7KC and 25OHC at50 mM. No significant changes were also observed under treatmentwith 4aOHC on the cytoskeleton organization and on themorphological aspect of the nuclei evaluated by staining withrhodamine-phalloidin and Hoechst 33342, respectively, and onthese parameters only minor modifications were found undertreatment with 4bOHC. With 7KC and 25OHC, it is noteworthy, thattwo different modes of cell death were induced, an apoptotic modeof cell death with 7KC, and a non-apoptotic mode of cell deathunder treatment with 25OHC, underlying the ability of oxysterolsto induce various types of cell death [42,63]. On murine oligoden-drocytes primary cultures, only slight dysfunctions of lysosomalmembrane integrity were observed under treatment with 4aOHCand 4bOHC. Thus, with the different parameters used to evaluateand characterize cell death associated events (loss of mitochondrialpotential, lysosomal destabilization, cytoskeleton disorganization,morphological nuclear changes) on 158N murine oligodendrocytesand on murine oligodendrocyte primary cultures, and compara-tively to data obtained with positive controls (7KC, 25OHC), ourresults demonstrate that 4aOHC and 4bOHC have slight impact onmajor organelles (mitochondria, lysosome) and cytoskeletonorganization in these cells even at high concentrations.

As oligodendrocyte damages can also occur through activationof oxidative stress and inflammation, which can in turn contributeto demyelination [64e66], the ability of 4aOHC and 4bOHC toinduce overproduction of ROS and cytokines secretion was impor-tant to consider. Whereas no overproduction of ROS (especiallysuperoxide anion) was found under treatment with 4aOHC, theoverproduction found with 4bOHC was in the same range thanthose established with 25OHC, known to be a potent inducer ofoxidative stress [11]. It was however lower than those observed

Fig. 9. Effects of 4a-hydroxycholesterol, 4b-hydroxycholesterol, 25-hydroxycholesterol, and 7-ketocholesterol on viability, transmembrane potential, lysosomal membraneintegrity, and superoxide anion production on murine oligodendrocyte primary cultures. After 10 days of cultures, murine oligodendrocyte primary cultures were treated for 24 h inthe absence or presence of 4a-hydroxycholesterol (4aOHC), 4b-hydroxycholesterol (4bOHC) (25, 50 mM), or with 25-hydroxycholesterol (25OHC), or 7-ketocholesterol (7KC)(50 mM). (A): Characterization of murine oligodendrocyte primary culture; oligodendrocytes (small arrows) strongly express PLP; large arrows point towards astrocytes. (B): Inmurine oligodendrocyte primary cultures higher expression of LXRb was found; LXRa mRNAs (Ct ¼ 26.9 � 0.1), LXRb mRNAs (Ct ¼ 24.8 � 0.1). The mRNA levels were measuredusing real-time RT-qPCR and normalized to 36B4. Data shown are expressed as 2 � DCt and are the mean � SD of five experiments. Significance of the difference between LXRa andLXRb mRNA levels (ManneWhitney test; * P < 0.05 or less). The biological activities of 4aOHC and 4bOHC on murine oligodendrocyte primary cultures were evaluated withdifferent functional tests. The percentages of living cells were determined after staining with trypan blue (C). The percentages of cells with depolarized mitochondria (D), withdestabilized lysosomes (E), and overproducing superoxide anions (% hydroethidine (HE) positive cells) (F) were determined by flow cytometry after staining with JC-1, AO, ordihydroethidium (DHE), respectively. Data shown are mean � SD from three independent experiments. Significance of the difference between untreated- and oxysterol-treated cells(ManneWhitney test; * P < 0.05 or less). No difference was observed between absolute control and vehicle (ethanol: 0.2%; corresponding to the maximum final concentration ofethanol used).

T. Nury et al. / Biochimie 95 (2013) 518e530528

with 7KC known to be a strong inducer of NAD(P)H-oxidase [67].The absence of superoxide anion production found on murineoligodendrocyte primary culture under treatment with 4aOHC and4bOHC underlines a need to clarify the ability of 4aOHC and 4bOHCto stimulate or not ROS production. Similarly, whereas no cytokine

overproduction (IL-1b, IL-6, IL-8, IL-10, TNFa and IL-12p70) weredetected under treatment with 4aOHC, 4bOHC, 7KC and 25OHC,complementary experiments on glial cells should be of interest todetermine with accuracy the absence of pro-inflammatory prop-erties of these different oxysterols on brain cells.

T. Nury et al. / Biochimie 95 (2013) 518e530 529

5. Conclusion

Our data bring some new information on the effects of 4aOHCand 4bOHC on the activation of LXRa and LXRb, and on their bio-logical activities on oligodendrocytes: 1) 4bOHC, but not 4aOHC,activates LXRa and LXRb; 2) 4aOHC and 4bOHCwere able to inhibitcell growth in a similar manner on 158N cells; 3) no markedcytotoxic activities of these oxysterols evaluated with variousparameters (increased permeability to PI, loss of transmembranemitochondrial potential, lysosomal destabilization, cytoskeletondisorganization, overproduction of ROS, stimulation of inflamma-tory cytokine secretion) were found on 158N cells and murineoligodendrocyte primary cultures especially with 4aOHC, compar-atively to 7KC and 25OHC. Therefore, the ability of 4aOHC and4bOHC to favor oligodendrocyte damages in vitro via cell deathinduction, organelle dysfunctions, oxidative and/or inflammatorystress can be considered as weak.

Acknowledgements

This work was supported by grants from the INSERM, the Uni-versité de Bourgogne, ELA Foundation, and the Conseil Régional deBourgogne. This work was presented as oral presentation at the 1stENOR (European Network on Oxysterols Research) symposium‘Oxysterols and Related Sterols in Chemistry, Biology &Medicine’ inRome, Italia, September 22e24, 2011.

References

[1] I. Björkhem, S. Meaney, Brain cholesterol: long secret life behind a barrier,Arterioscler. Thromb. Vasc. Biol. 24 (2004) 806e815.

[2] J.M. Dietschy, S.D. Turley, Thematic review series: brain lipids. Cholesterolmetabolism in the central nervous system during early development and inthe mature animal, J. Lipid Res. 45 (2004) 1375e1397.

[3] T.M. Jeitner, I. Voloshyna, A.B. Reiss, Oxysterol derivatives of cholesterol inneurodegenerative disorders, Curr. Med. Chem. 18 (2011) 1515e1525.

[4] V. Leoni, C. Caccia, Oxysterols as biomarkers in neurodegenerative diseases,Chem. Phys. Lipids 164 (2011) 515e524.

[5] D.W. Russell, Oxysterol biosynthetic enzymes, Biochim. Biophys. Acta 1529(2000) 126e135.

[6] I. Björkhem, Do oxysterols control cholesterol homeostasis? J. Clin. Invest. 110(2002) 725e730.

[7] G.J. Schroepfer Jr., Oxysterols: modulators of cholesterol metabolism andother processes, Physiol. Rev. 80 (2000) 361e554.

[8] J.F. Carvalho, M.M. Silva, J.N. Moreira, S. Simões, M.L. Sá E Melo, Selectivecytotoxicity of oxysterols through structural modulation on rings A and B.Synthesis, in vitro evaluation, and SAR, J. Med. Chem. 54 (2011) 6375e6393.

[9] D.J. Peet, B.A. Janowski, D.J. Mangelsdorf, The LXRs: a new class of oxysterolreceptors, Curr. Opin. Genet. Dev. 8 (1998). 571e575.

[10] S. Ducheix, J.M. Lobaccaro, P.G. Martin, H. Guillou, Liver X Receptor: an oxy-sterol sensor and a major player in the control of lipogenesis, Chem. Phys.Lipids 164 (2011) 500e514.

[11] S. Lemaire-Ewing, C. Prunet, T. Montange, A. Vejux, A. Berthier, G. Bessède,L. Corcos, P. Gambert, D. Néel, G. Lizard, Comparison of the cytotoxic, pro-oxidant and pro-inflammatory characteristics of different oxysterols, Cell.Biol. Toxicol. 21 (2005) 97e114.

[12] A. Vejux, G. Lizard, Cytotoxic effects of oxysterols associated with humandiseases: Induction of cell death (apoptosis and/or oncosis), oxidative andinflammatory activities, and phospholipidosis, Mol. Aspects Med. 30 (2009)153e170.

[13] A. Jusakul, P. Yongvanit, W. Loilome, N. Namwat, R. Kuver, Mechanisms ofoxysterol-induced carcinogenesis, Lipids Health Dis. 10 (2011) 44.

[14] P. de Medina, M.R. Paillasse, G. Ségala, F. Khallouki, S. Brillouet, F. Dalenc,F. Courbon, M. Record, M. Poirot, S. Silvente-Poirot, Importance of cholesteroland oxysterols metabolism in the pharmacology of tamoxifen and other AEBSligands, Chem. Phys. Lipids 164 (2011) 432e437.

[15] P. Sacchetti, K.M. Sousa, A.C. Hall, I. Liste, K.R. Steffensen, S. Theofilopoulos,C.L. Parish, C. Hazenberg, L.A. Richter, O. Hovatta, J.A. Gustafsson, E. Arenas,Liver X receptors and oxysterols promote ventral midbrain neurogenesisin vivo and in human embryonic stem cells, Cell Stem Cell. 5 (2009) 409e419.

[16] E.G. Lund, J.M. Guileyardo, D.W. Russell, cDNA cloning of cholesterol 24-hydroxylase, a mediator of cholesterol homeostasis in the brain, Proc. Natl.Acad. Sci. U S A 96 (1999) 7238e7243.

[17] J. Makoukji, G. Shackleford, D. Meffre, J. Grenier, P. Liere, J.M. Lobaccaro,M. Schumacher, C. Massaad, Interplay between LXR and Wnt/b-catenin

signaling in the negative regulation of peripheral myelin genes by oxysterols,J. Neurosci. 31 (2011) 9620e9629.

[18] 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. U S A 96 (1999) 266e271.

[19] J.Y. Chang, L.Z. Liu, Neurotoxicity of cholesterol oxides on cultured cerebellargranule cells, Neurochem. Int. 32 (1998) 317e323.

[20] J.P. Rantham Prabhakara, G. Feist, S. Thomasson, A. Thompson, E. Schommer,O. Ghribi, Differential effects of 24-hydroxycholesterol and 27-hydroxycholesterol on tyrosine hydroxylase and alpha-synuclein in humanneuroblastoma SH-SY5Y cells, J. Neurochem. 107 (2008) 1722e1729.

[21] A. Trousson, S. Bernard, P.X. Petit, P. Liere, A. Pianos, K. El Hadri,J.M. Lobaccaro, M.S. Ghandour, M. Raymondjean, M. Schumacher, C. Massaad,25 hydroxycholesterol provokes oligodendrocyte cell line apoptosis andstimulates the secreted phospholipase A2 type IIA via LXR beta and PXR,J. Neurochem. 109 (2009) 945e958.

[22] K. Yamanaka, Y. Saito, T. Yamamori, Y. Urano, N. Noguchi, 24(S) hydrox-ycholesterol induces neuronal cell death through necroptosis, a form of pro-grammed necrosis, J. Biol. Chem. 286 (2011) 24666e24673.

[23] O. Breuer, Identification and quantitation of cholest-5-ene-3b,4b-diol in ratliver and human plasma, J. Lipid Res. 36 (1995) 2275e2281.

[24] S. Dzeletovic, O. Breuer, E. Lund, U. Diczfalusy, Determination of cholesteroloxidation products in human plasma by isotope dilution-mass spectrometry,Anal. Biochem. 225 (1995) 73e80.

[25] G.S. Tint, P. Pentchev, G. Xu, A.K. Batta, S. Shefer, G. Salen, A. Honda, Choles-terol and oxygenated cholesterol concentrations are markedly elevated inperipheral tissue but not in brain from mice with the Niemann-Pick type Cphenotype, J. Inherit. Metab. Dis. 21 (1998) 853e863.

[26] K. Bodin, L. Bretillon, Y. Aden, L. Bertilsson, U. Broomé, C. Einarsson,U. Diczfalusy, Antiepileptic drugs increase plasma levels of 4b-hydrox-ycholesterol in humans. Evidence for involvement of cytochrome P450 3A4,J. Biol. Chem. 276 (2001) 38685e38689.

[27] S. Meaney, M. Hassan, A. Sakinis, D. Lütjohann, K. von Bergmann,A. Wennmalm, U. Diczfalusy, I. Björkhem, Evidence that the major oxysterolsin human circulation originate from distinct pools of cholesterol: a stableisotope study, J. Lipid Res. 42 (2001) 70e78.

[28] U. Diczfalusy, H. Nylén, P. Elander, L. Bertilsson, 4b-Hydroxycholesterol, anendogenous marker of CYP3A4/5 activity in humans, Br. J. Clin. Pharmacol. 71(2011) 183e189.

[29] U. Diczfalusy, J. Miura, H.K. Roh, R.A. Mirghani, J. Sayi, H. Larsson, K.G. Bodin,A. Allqvist, M. Jande, J.W. Kim, E. Aklillu, L.L.L. Gustafsson, L. Bertilsson, 4Beta-hydroxycholesterol is a new endogenous CYP3A marker: relationship toCYP3A5 genotype, quinine 3-hydroxylation and sex in Koreans, Swedes andTanzanians, Pharmacogenet. Genomics 18 (2008) 201e208.

[30] K. Wide, H. Larsson, L. Bertilsson, U. Diczfalusy, Time course of the increase in4beta-hydroxycholesterol concentration during carbamazepine treatment ofpaediatric patients with epilepsy, Br. J. Clin. Pharmacol. 65 (2008) 708e715.

[31] B.A. Janowski, P.J. Willy, T.R. Devi, J.R. Falck, D.J. Mangelsdorf, An oxysterolsignalling pathway mediated by the nuclear receptor LXR alpha, Nature 383(1996) 728e731.

[32] B.S. Desai, A.J. Monahan, P.M. Carvey, B. Hendey, Blood-brain barrierpathology in Alzheimer’s and Parkinson’s disease: implications for drugtherapy, Cell. Transplant. 16 (2007) 285e299.

[33] I. Björkhem, A. Cedazo-Minguez, V. Leoni, S. Meaney, Oxysterols and neuro-degenerative diseases, Mol. Aspects Med. 30 (2009) 171e179.

[34] M. Baarine, K. Ragot, E.C. Genin, H. El Hajj, D. Trompier, P. Andreoletti,M.S. Ghandour, F. Menetrier, M. Cherkaoui-Malki, S. Savary, G. Lizard,Peroxisomal and mitochondrial status of two murine oligodendrocytic celllines (158N, 158JP): potential models for the study of peroxisomal disor-ders associated with dysmyelination processes, J. Neurochem. 111 (2009)119e131.

[35] G. Wójcicka, A. Jamroz-Wi�sniewska, K. Horoszewicz, J. Be1towski, Liver Xreceptors (LXRs). Part I: structure, function, regulation of activity, and role inlipid metabolism, Postepy. Hig. Med. Dosw. 61 (2007) 736e759.

[36] A. Viger, A. Marquet, D.H.R. Barton, W. Motherwell, S.Z. Zard, Stereochemistryof the reduction of 3b-benzoyloxycholest-5-en-4-one with sodium borohy-dride, J. Chem. Soc. Perkin Trans. 1 (1982) 1937e1940.

[37] E. Ma, T. Choi, An efficient 4b-hydroxylation of steroidal 5-en-3b-ols and 1,4-conjugation of steroidal 4-en-3-ones using SeO2 oxidation, Bull. Korean Chem.Soc. 30 (2009) 245e248.

[38] A.C. Feutz, D. Pham-Dinh, B. Allinquant, M. Miehe, S. Ghandour, An immor-talized jimpy oligodendrocyte cell line: defects in cell cycle and cAMPpathway, Glia 34 (2001) 241e252.

[39] M.S. Ghandour, A.C. Feutz, W. Jalabi, O. Taleb, D. Bessert, M. Cypher, L. Carlock,R.P. Skoff, Trafficking of PLP/DM20 and cAMP signaling in immortalized jimpyoligodendrocytes, Glia 40 (2002) 300e311.

[40] K. Ragot, D. Delmas, A. Athias, T. Nury, M. Baarine, G. Lizard, a-tocopherolimpairs 7-ketocholesterol-induced caspase-3-dependent apoptosis involvingGSK-3 activation and Mcl-1 degradation on 158N murine oligodendrocytes,Chem. Phys. Lipids 164 (2011) 469e478.

[41] G. Lizard, V. Deckert, L. Dubrez, M. Moisant, P. Gambert, L. Lagrost, Inductionof apoptosis in endothelial cells treated with cholesterol oxides, Am. J. Pathol.148 (1996) 1625e1638.

[42] G. Lizard, S. Monier, C. Cordelet, L. Gesquière, V. Deckert, S. Gueldry, L. Lagrost,P. Gambert, Characterization and comparison of the mode of cell death,

T. Nury et al. / Biochimie 95 (2013) 518e530530

apoptosis versus necrosis, induced by 7beta-hydroxycholesterol and 7-ketocholesterol in the cells of the vascular wall, Arterioscler. Thromb. Vasc.Biol. 19 (1999) 1190e1200.

[43] G. Lizard, S. Fournel, L. Genestier, N. Dhedin, C. Chaput, M. Flacher, M. Mutin,G. Panaye, J.P. Revillard, Kinetics of plasma membrane and mitochondrialalterations in cells undergoing apoptosis, Cytometry 21 (1995) 275e283.

[44] A.K. Marel, G. Lizard, J.C. Izard, N. Latruffe, D. Delmas, Inhibitory effects oftrans-resveratrol analogs molecules on the proliferation and the cell cycleprogression of human colon tumoral cells, Mol. Nutr. Food Res. 52 (2008)538e548.

[45] S.T. Smiley, M. Reers, C. Mottola-Hartshorn, M. Lin, A. Chen, T.W. Smith,G.D. Steele Jr., L.B. Chen, Intracellular heterogeneity in mitochondrialmembrane potentials revealed by a J-aggregate-forming lipophilic cation JC-1,Proc. Natl. Acad. Sci. U S A 88 (1991) 3671e3675.

[46] M. Baarine, K. Ragot, A. Athias, T. Nury, Z. Kattan, E.C. Genin, P. Andreoletti,F. Ménétrier, J.M. Riedinger, M. Bardou, G. Lizard, Incidence of Abcd1 level onthe induction of cell death and organelle dysfunctions triggered by very longchain fatty acids and TNF-a on oligodendrocytes and astrocytes, Neuro-toxicology 33 (2012) 212e228.

[47] M. Olsson, I. Rundquist, U. Brunk, Flow cytofluorometry of lysosomal acridineorange uptake by living cultured cells, Acta Path. Microbiol. Immunol. Scand.(Sect A) 95 (1987) 159e165.

[48] X.M. Yuan, W. Li, U.T. Brunk, H. Dalen, Y.H. Chang, A. Sevanian, Lysosomaldstabilization during macrophage damage induced by cholesterol oxidationproducts, Free Radic. Biol. Med. 28 (2000) 208e218.

[49] F. Belloc, P. Vincendeau, G. Freyburger, P. Dumain, M.R. Boisseau, Flow cyto-metric study of the activation of polymorphonuclear cells, J. Leukoc. Biol. 48(1990) 353e358.

[50] G. Majno, I. Joris, Apoptosis, oncosis, and necrosis. An overview of cell death,Am. J. Pathol. 146 (1995) 3e15.

[51] E. Kahn, M. Baarine, A. Dauphin, K. Ragot, N. Tissot, A. Seguin, F. Ménétrier,Z. Kattan, C.M. Bachelet, F. Frouin, G. Lizard, Characterization of lipidmembrane organization revealed by LAURDAN color shift during cell deathinduced by 7-ketocholesterol and very long chain fatty acids on 158N murineoligodendrocytes: investigation by mono- and bi-photon spectral imagingconfocal microscopy associated with Factor Analysis of Medical ImageSequences (FAMIS), Cytometry Part A 79A (2011) 293e305.

[52] G. Rothe, G. Valet, Flow cytometric analysis of respiratory burst activity inphagocytes with hydroethidine and 20 ,70-dichlorofluorescin, J. Leukoc. Biol. 47(1990) 440e448.

[53] C. Prunet, T. Montange, A. Véjux, A. Laubriet, J.F. Rohmer, J.M. Riedinger,A. Athias, S. Lemaire-Ewing, D. Néel, J.M. Petit, E. Steinmetz, R. Brenot,P. Gambert, G. Lizard, Multiplexed flow cytometric analyses of pro- and anti-inflammatory cytokines in the culture media of oxysterol-treated human

monocytic cells and in the sera of atherosclerotic patients, Cytometry A 69(2006) 359e373.

[54] N. Baumann, D. Pham-Dinh, Biology of oligodendrocyte and myelin in themammalian central nervous system, Physiol. Rev. 81 (2011) 871e927.

[55] D. Kaul, Molecular link between cholesterol, cytokines and atherosclerosis,Mol. Cell. Biochem. 219 (2011) 65e71.

[56] G. Lizard, S. Gueldry, V. Deckert, P. Gambert, L. Lagrost, Evaluation of thecytotoxic effects of some oxysterols and of cholesterol on endothelial cellgrowth: methodological aspects, Pathol. Biol. (Paris) 45 (1997) 281e290.

[57] C.J. Yeh, B.L. Hsi, W.P. Faulk, Propidium iodide as a nuclear marker inimmunofluorescence. II. Use with cellular identification and viability studies,J. Immunol. Meth. 43 (1981) 269e275.

[58] Q. Zhou, E. Wasowicz, F.A. Kummerow, Failure of vitamin E to protect culturedhuman arterial smooth muscle cells against oxysterol-induced cytotoxicity,J. Am. Coll. Nutr. 14 (1995) 169e175.

[59] H.K. Lim, H.K. Kang, E.S. Yoo, B.J. Kim, Y.W. Kim, M. Cho, J.H. Lee, Y.S. Lee,M.H. Chung, J.W. Hyun, Oxysterols induce apoptosis and accumulation of cellcycle at G(2)/M phase in the human monocytic THP-1 cell line, Life Sci. 72(2003) 1389e1399.

[60] S. Lordan, J.J. Mackrill, N.M. O’Brien, Oxysterols and mechanisms of apoptoticsignaling: implications in the pathology of degenerative diseases, J. Nutr.Biochem. 20 (2009) 321e336.

[61] R.E. Soccio, J.L. Breslow, Intracellular cholesterol transport, Arterioscler.Thromb. Vasc. Biol. 24 (2004) 1150e1160.

[62] A. Toulmay, W.A. Prinz, Lipid transfer and signaling at organelle contact sites:the tip of the iceberg, Curr. Opin. Cell. Biol. 23 (2011) 458e463.

[63] S. Ayala-Torres, B.H. Johnson, E.B. Thompson, Oxysterol sensitive and resistantlymphoid cells: correlation with regulation of cellular nucleic acid bindingprotein mRNA, J. Steroid. Biochem. Mol. Biol. 48 (1994) 307e315.

[64] L. Haider, M.T. Fischer, J.M. Frischer, J. Bauer, R. Höftberger, G. Botond,H. Esterbauer, C.J. Binder, J.L. Witztum, H. Lassmann, Oxidative damage inmultiple sclerosis lesions, Brain 134 (Pt 7) (2011) 1914e1924.

[65] M. Kapadia, B. Sakic, Autoimmune and inflammatory mechanisms of CNSdamage, Prog. Neurobiol. 95 (2011) 301e333.

[66] E. Galea, N. Launay, M. Portero-Otin, M. Ruiz, R. Pamplona, P. Aubourg,I. Ferrer, A. Pujol, Oxidative stress underlying axonal degeneration in adre-noleukodystrophy: a paradigm for multifactorial neurodegenerative diseases?Biochim. Biophys. Acta (2012). PubMed PMID: 22353463.

[67] E. Pedruzzi, C. Guichard, V. Ollivier, F. Driss, M. Fay, C. Prunet, J.C. Marie,C. Pouzet, M. Samadi, C. Elbim, Y. O’dowd, M. Bens, A. Vandewalle,M.A. Gougerot-Pocidalo, G. Lizard, E. Ogier-Denis, NAD(P)H oxidase Nox-4mediates 7-ketocholesterol-induced endoplasmic reticulum stress andapoptosis in human aortic smooth muscle cells, Mol. Cell. Biol. 24 (2004)10703e10717.