Effects of Caspase Inhibitors (z-VAD-fmk,

z-VDVAD-fmk) on Nile Red Fluorescence Pattern

in 7-Ketocholesterol-Treated Cells: Investigation

by Flow Cytometry and Spectral Imaging Microscopy

Anne Vejux,1 G�erard Lizard,1� Yves Tourneur,2 Jean-Marc Riedinger,3

Fr�ed�erique Frouin,4 Edmond Kahn4�*

� AbstractThe 7-ketocholesterol (7KC)-induced cell death has some characteristics of apoptosisand is associated with polar lipid accumulation. So, we investigated the effects of thebroad-spectrum caspase inhibitor z-VAD-fmk and of the caspase-2 inhibitor z-VDVAD-fmk on lipid profile evaluated by staining with Nile Red (NR). The 7KC-trea-ted human monocytic U937 cells were cultured in the absence or in the presence of thecaspase inhibitors z-VAD-fmk or z-VDVAD-fmk. When staining with NR is performed,neutral and polar lipids have yellow and orange/red emission, respectively, and fluores-cence was then analyzed by flow cytometry (FCM) and by confocal laser scanning mi-croscopy (CLSM) combined with subsequent image processing. The 3D-imagesequences were obtained by means of CLSM using spectral analysis, and were analyzedby the factor analysis of medical image sequences algorithm to differentiate spectrainside mixed fluorescence emission and get corresponding specific images. By FCM,comparatively to untreated cells, higher percentages of red fluorescent cells were identi-fied in 7KC-treated cells. Factor curves and images reveal orange and red fluorescenceemissions in 7KC-treated cells and show yellow, orange, and red fluorescence emissionsin 7KC-treated cells cultured in the presence of z-VAD-fmk or z-VDVAD-fmk. Ourdata support that investigation by FCM and by spectral analysis in CLSM associatedwith subsequent image processing provides useful tools to determine the effect of cas-pase inhibitors on lipid content evaluated with NR. They also favor the hypothesis ofrelationships between caspase activity and polar lipid accumulation. ' 2007 International

Society for Analytical Cytology

� Key terms7-ketocholesterol; caspase-2; confocal microscopy; flow cytometry; Nile Red; spectralimaging; FAMIS; unmixing

CHOLESTEROL is the most abundant sterol in human and animal tissues and it can

be oxidized to oxysterols either by specific enzymes or spontaneously when it is

exposed to heat, air, light, and oxidizing agents (1). Oxysterols, which are identified

in animal and in human tissues, have different biological activities. It includes effects

on cholesterol homeostasis, sphingolipid metabolism, radical oxygen species produc-

tion, cytokines secretion, and cellular viability (2,3), which are supposed to play key

roles in slow degenerative diseases such as atherosclerosis (4). Among these choles-

terol oxide derivatives, 7-ketocholesterol (7KC) is one of the most prominent oxy-

sterols present in atheromatous plaques (5). We previously reported that 7KC-

induced cytotoxic effects include numerous features of apoptosis such as activation

of caspases-3, -7, -8, and -9, which are associated to the formation of multilamellar

cytoplasmic structures named myelin figures (6,7). These ultrastructural features

1INSERM UMR 866 (Lipides, Nutrition etCancer; �equipe Biochimie M�etaboliqueet Nutritionnelle), IFR Sant�e STIC,Universit�e de Bourgogne-Facult�e desSciences Gabriel, 21000 Dijon, France2Centre Commun de Quantim�etrie,Universit�e Lyon-1, 69373 Lyon, France3Centre de Lutte contre le Cancer GFLeclerc, Laboratoire de BiologieM�edicale, 1 bis rue du Pr Marion, 21000Dijon, France4INSERM UMR S 678 UPMC, CHUPiti�e-Salpetri�ere, 75634 Paris Cedex 13,France

Received 16 November 2006; RevisionReceived 12 March 2007; Accepted 13March 2007�G. Lizard (gerard.lizard@u-bourgogne.fr)and E. Kahn contributed equally to thiswork.

Grant sponsors: INSERM, the LigueContre Le Cancer (Comit�e de Cote d’Or),the Conseil R�egional de Bourgogne.

*Correspondence to: Dr. Edmond Kahn,INSERM UMR S 678 UPMC, CHU Piti�e-Salpetri�ere, 75634 Paris Cedex 13,France.

Email: kahn@imed.jussieu.fr

Published online 25 April 2007 inWiley InterScience (www.interscience.wiley.com)

DOI: 10.1002/cyto.a.20410

© 2007 International Society forAnalytical Cytology

Original Article

Cytometry Part A � 71A: 550�562, 2007

which can be stained with Nile Red (NR) (a metachromatic

fluorescent probe which puts yellow and orange/red colors

on neutral and polar lipids, respectively) (8–10) are rich in

7KC, accumulate polar lipids (phosphatidylcholine and

sphingomyelin) and contain cholesterol (7). These data,

which provide experimental evidence that 7KC-induced cell

death is associated with altered lipid metabolism, are in

agreement with previous investigations performed on J774

A.1 murine macrophage like cells cultured in the presence of

oxidatively modified LDL (11). Therefore, the aim of the

present investigation was to show the interest of the com-

bined use of NR, flow cytometry, and spectral imaging mi-

croscopy to study the link between lipid accumulation and

caspase activity, by using caspase inhibitors.

The analysis of this relationship was supported by differ-

ent reports, which demonstrate some connections between

lipid metabolism and cell death. Thus, when apoptosis is

induced by the protein kinase inhibitor staurosporine, it was

demonstrated that caspase-3 cleaves SREBP-1 and SREBP-2

(Sterol-Regulatory Element Binding Proteins 1 and 2) liber-

ating a transcriptionally active fragment (12). Similarly, in

sterol-deprived cells, SREBP-1 and SREBP-2 are cleaved and

induce the release of a N-terminal fragment that enters the

nucleus and activates transcription of numerous genes. This

includes genes for the low density lipoprotein receptor as

well as genes implied in lipid homeostasis such as those

encoding enzymes involved in the cholesterol or triacylglyc-

erol/phospholipid synthesis pathway, such as 3-hydro-3-

methylglutaryl coenzyme A reductase, farnesyl diphosphate

synthase, and fatty acid synthase (13–15). It has also been

recently demonstrated that caspase-2, which can contribute

to the activation of numerous apoptotic events (16), is a

member of the SREBP-responsive gene battery in human

cells, and that modulation of caspase-2 expression via

SREBP-2 modulates the cellular lipid content (17).

To analyze the relations between lipid accumulation

and caspase activity, human monocytic U937 cells were cul-

tured in the absence or in the presence of 7KC (40 lg/ml)

associated or not with the broad-spectrum caspase inhibitor

z-VAD-fmk (200 lM) (18), or with the caspase-2 inhibitor

z-VDVAD-fmk (100 lM) (19). The effects of these inhibi-

tors on 7KC-induced lipid accumulation was measured after

staining with NR which allows to distinguish neutral and

polar lipids (8,9) both by flow cytometry (FCM), and by

spectral analysis (20) by means of confocal laser scanning

microscopy (CLSM). Three-dimensional (3D) sequences of

images obtained by spectral analysis of cells were analyzed

by the factor analysis of medical image sequences (FAMIS)

algorithm, which summarizes image sequences into a

reduced number of images called factor images, and curves

called factor curves (21–23). Image analysis was performed

on spectral sequences of images to get the specific color

expressions of NR corresponding to neutral and polar

lipids, which are yellow and orange/red, respectively.

Sequences of images were investigated by FAMIS to provide

factor images corresponding to each color expression.

Indeed, FAMIS makes it possible to unmix global fluores-

cence emission expressions, and provides images corre-

sponding to yellow, orange, and red mixed emissions in

spite of spectral overlaps, according to experiments carried

out on models for single spots inside spheres (24) to take

full advantage of 3D information such as spectral patterns

(25), or depth profiles (26). Hence, the reconstructed 3D

image resulted in a distorted sphere, itself containing small

spheres. As a result each yellow spot (neutral lipids) inside

a flattened cellular background can be visualized (7,10). As

a consequence also, unmixing is provided when the distance

between two yellow spots makes it possible to differentiate

them visually or according to resolution criteria (7,10).

So, the present investigation establishes the interest of the

simultaneous analysis of NR-stained cells by FCM and by

spectral analysis in CLSM associated with subsequent image

processing with FAMIS algorithm, to provide useful tools to

analyze the effects of caspase inhibitors on cellular lipid con-

tent. These complementary methods clearly show that the

broad-caspase inhibitor z-VAD-fmk and the caspase-2 inhibi-

tor z-VDVAD-fmk are capable to counteract the red fluores-

cence increase of NR triggered by 7KC.

MATERIALS ANDMETHODS

Cells and Treatments

Human promonocytic U937 leukemia cells were grown

in suspension in culture medium consisting of RPMI 1640

with GlutaMAX I (Gibco, Eragny, France), and antibiotics

(100 U/ml penicillin, 100 lg/ml streptomycin) supplemented

with 10% (v/v) heat-inactivated fetal calf serum (Gibco). Cells

were seeded at 500,000 per ml of culture medium, and

passaged twice a week. They were cultured in a 5% CO2/95%

air-humidified atmosphere at 37�C.The 7-ketocholesterol (7KC) was provided by Sigma

(L’Isle d’Abeau Chesnes, France), and its purity was deter-

mined to be 100% by gaseous phase chromatography coupled

with mass spectrometry. For all experiments, a stock solution

of 7KC was prepared at a concentration of 800 lg/ml as pre-

viously described (27). To obtain the initial solution, 800 lg of7KC were dissolved in 50 ll of absolute ethanol, 950 ll of cul-ture medium were added, and the suspension was sonicated.

To obtain 20–40 lg/ml final concentrations (corresponding to

50 and 100 lM, respectively), 25 or 50 ll of this initial solutionwas added per ml of culture medium. In all experiments, 7KC

was introduced into the culture medium at the beginning of

the culture, and treatments on U937 cells were carried out at 4,

6, 14, 18, and/or 24 h. When the cells were simultaneously trea-

ted with 7KC, and with the broad spectrum caspase inhibitor

z-VAD-fmk (200 lM) (Bachem, Voisins le Bretonneux, France), or

with the caspase-2 inhibitor z-VDVAD-fmk (100 lM) (R&D

systems, Minneapolis, MN), these inhibitors (diluted in

DMSO) (Sigma) were introduced in the medium at the begin-

ning of the culture and the oxysterols were added 1 h later. In

these conditions, the final concentration of DMSO in the cul-

ture medium was 0.2%. When the cells are treated with etopo-

side (VP16; Sigma), a stock solution was prepared at 50 mM in

DMSO, further dilutions were carried out in culture medium

ORIGINAL ARTICLE

Cytometry Part A � 71A: 550�562, 2007 551

to obtain a 50 lM final concentration, and treatments were

carried out at 1 and 4 h. In these conditions, the final concen-

tration of DMSO in the culture medium was 0.1%.

In Situ Detection of Activated Caspases with

Fluorochrome-Labeled Inhibitors of Caspase

Caspase total and caspase-2 activities were measured with

fam-VAD-fmk (Trevigen, Gaithersburg, MD) and fam-

VDVAD-fmk (Immunochemistry Technologies, Bloomington,

MN), respectively, using specifically dedicated kits according

to manufactural procedures. Fluorochrome-labeled inhibitors

of caspase (FLICA) reagents (fam-VAD-fmk; fam-VDVAD-

fmk) are cell permeant and noncytotoxic compounds, which

are widely used in FCM and microscopy to investigate caspase

activities during apoptosis (28,29). Briefly, fam-VAD-fmk is a

carboxyfluorescein derivative of VAD-fmk, and fam-VDVAD-

fmk is a carboxyfluorescein derivative of VDVAD-fmk. These

FLICA reagents were dissolved in DMSO to obtain a 1503concentrated solution. Prior to use, 303 working solutions of

fam-VAD-fmk or fam-VDVAD-fmk were prepared by diluting

the stock solution 1:5 in phosphate buffer saline (PBS). Cells

were distributed into 300-ll aliquots containing 1 3 106 cells,

and fam-VAD-fmk or fam-VDVAD-fmk working solutions

were added to obtain a 13 final concentration. After 1 h of

incubation in the dark in a 5% CO2/95% air-humidified

atmosphere at 37�C, cells were washed twice in 13 wash

buffer. The cell pellet was resuspended in 400 ll of PBS, andHoechst 33342 was added to obtain a 5 lM final concentra-

tion. After incubation for 30 min at 37�C, cells were immedi-

ately analyzed by FCM. In addition, 30 ll of the cellular sus-

pension adjusted at 106 cells/ml were applied to glass slides,

air dried, mounted in a fluorescent mounting medium (Dako,

Coppenhagen, Denmark), coverslipped, and stored in the dark

at 4�C before microscopical examinations. Flow cytometric

analyses were performed on a GALAXY flow cytometer

(Partec, M€unster, Germany) equipped with a blue laser emit-

ting at 488 nm and working at 20 mW. The green emission of

fam-VAD-fmk and fam-VDVAD-fmk were collected at 520 610 nm, and measured on a logarithmic scale. About 10,000

cells were acquired for each sample, and data were analyzed

with the FlowMax software (Partec). Observations by conven-

tional fluorescence microscopy were made with an Axioskop

fluorescent microscope (Zeiss). For each sample, 300 cells

were examined.

Protein Extraction and Western Blot Analyses

Two isoforms of caspase-2 exist: caspase-2L and caspase-

2S. As the long isoform of caspase-2 (caspase-2L) is implicated

in the triggering of apoptosis, whereas the short isoform (cas-

pase-2S) can antagonize cell death, (30) the activation of cas-

pase-2L was investigated by Western blot analysis. Briefly,

analyses of caspase-2L were performed in Ripa buffer (150 mM

NaCl, 50 mM Tris-HCl, pH 7.4, 0.1% sodium dodecyl

sulfate (SDS), 1% igepal, 0.5% Na deoxycholate, containing a

mixture of protease and phosphatase inhibitors (0.1 mM phe-

nylmethanesulfonyl fluoride, 2.5 lg/l aprotinin, 10 lg/l pep-statin A, 2.5 lg/l trypsin inhibitor, 2.5 lg/l leupeptin, 0.1 mM

orthovanadate, 40 mM b-glycerophosphate, 100 mM NaF))

extracts of untreated or treated U937 cells with 7KC (40 lg/ml

corresponding to 100 lM) for 18 h or VP16 (50 lM) for 4 h.

After a 30-min incubation at 4�C in the lysis buffer, the cell

debris were eliminated by centrifugation (20 min, 10 000g)

and the supernatant was collected. The protein concentrations

were measured by using bicinchoninic acid reagent (Pierce,

Rockford, IL) according to the method of Smith et al. (31).

Eighty micrograms of protein were incubated in loading buffer

(125 mM Tris/HCl, pH 6,8, 10% (w/v) mercaptoethanol, 4,6%

(w/v) SDS, 20% (v/v) glycerol, 0,003% (w/v) Bromophenol

blue), boiled during 3 min, separated by SDS-PAGE, and elec-

troblotted onto a polyvinylidine difluoride membrane (Bio-

Rad, Marnes-la-Coquette, France). After blocking nonspecific

binding sites during 2 h at room temperature in TPBS (NaCl/

Pi, 0,1 % Tween-20), the membranes were incubated overnight

at 4�C with the mouse monoclonal primary antibody (anti-

caspase-2L; BD-Biosciences, San Jose, CA) diluted in TPBS.

After three 10 min washes with TPBS, the membranes were

incubated (1 h, room temperature) with rabbit anti-mouse

horseradish peroxidase-conjugated secondary antibody (Bio-

Rad) used at a dilution of 1:2500, and washed three times (10

min) in TPBS. The membranes were revealed using an

enhanced chemoluminescence detection kit (Amersham, Les

Ulis, France) according to manufacturer’s procedure. Each

experiment was repeated three times with identical results.

The blots were probed with a mouse anti-human Hsc-70

monoclonal antibody (Santa Cruz Biotechnology, Santa Cruz,

CA) to confirm that they had been equally loaded.

Colorimetric Assay for Caspase-2 Protease Activity

To assess caspase-2 activity, the cleavage of VDVAD-p-

nitroaniline (VDVAD-pNA) was measured according to the

manufacturer procedure with the caspase-2 colorimetric assay

kit (R&D Systems, Minneapolis, MN) in U937 cells cultured

in the absence or in the presence of 7KC (40 lg/ml) during 4,

6, 14, 18, and/or 24 h, or of VP16 (50 lM) during 1 and 4 h.

Briefly, at the end of the incubation time, U937 cells were

counted and collected by centrifugation (10 min, 250g). The

supernate was removed while the cell pellet was lysed by

the addition of lysis buffer (2.106 cells/50 ll of lysis buffer). Thecell lysate was incubated on ice (10 min), centrifuged (1 min,

4�C, 10,000g), and the supernate was collected. For each assay,

the cell lysate was mixed with the reaction buffer and with the

caspase-2 colorimetric substrate (VDVAD-pNA). The enzy-

matic reaction for caspase-2 activity was carried out in a 96-

well flat bottom microplate (2 h, 37�C). The cleavage of the

peptide VDVAD-pNA releases the chromophore pNA, which is

quantified spectrophotometrically at a wavelength of 405 nm.

The plate was read on a microplate reader (Biomek plata reader,

Beckman, Fullerton, CA), and the level of caspase-2 enzymatic

activity in the cell lysate is directly proportional to the

color reaction. In these conditions, at least three independent

experiments were performed in duplicate.

ORIGINAL ARTICLE

552 Effects of Caspase Inhibitors on Nile Red Fluorescence Pattern

Biochemical Characterization of Lipid Content

U937 cells were cultured during 24 h in the absence or in

the presence of 7KC (20 lg/ml corresponding to 50 lM). The

cells were collected by centrifugation, washed twice in PBS,

and enumerated with a hematocytometer. Total lipids were

further extracted by the method of Folch et al. (32). The lipid

contents of cholesterol (esterified and unesterified), phosphati-

dylcholine, and sphingomyelin were determined. To this end,

lipids were extracted and analyzed by capillary gas chromatog-

raphy and mass spectrometry (CGC/MS) as previously

described (33).

Staining Conditions with Nile Red

NR is a phenoxazine dye that can be used on living cells

to localize and quantify neutral and polar lipids (Sigma). This

dye stains neutral lipids in yellow (570–590 nm), and polar

lipids in orange/red (590 nm and above) when it is excited at

488 nm (8,9). When it is excited at 532 nm, NR can be used to

identify polar lipids, which are colored in orange/red (10). In

the present investigation, NR was prepared at 100 lg/ml in

DMSO (Sigma). After 18 h of treatment with 7KC (40 lg/ml),

NR was added to the culture medium at a final concentration

of 0.1 lg/ml on a cellular suspension adjusted to 106 cells/ml.

After 15 min of incubation at 37�C, flow cytometric analyses

of NR stained cells were immediately performed on a

CYFLOW green flow cytometer (Partec, M€unster, Germany)

as previously described (10). The orange/red fluorescence of

NR was acquired with a 630 nm long pass filter, and the mean

fluorescence intensities of untreated and 7KC-treated cells

were measured on a logarithmic scale. About 10,000 cells were

acquired for each sample, and data were obtained with the

FlowMax software (Partec).

Characterization of Nuclear Morphology by Means of

Hoechst 33342 Staining

Nuclear morphology was studied by means of Hoechst

33342 staining, using fluorescence microscopy with an Axios-

kop microscope (Zeiss, Jena, Germany) and ultra violet (UV)

light excitation. Apoptotic cells were characterized by con-

densed and/or fragmented nuclei, and oncotic cells were iden-

tified by swollen nuclei (34,35). For each sample, 300 cells

were examined.

Laser Scanning Confocal Microscopy, Spectral

Imaging, and Image Analysis

For microscopic analyses, 30 ll of a cellular suspension

stained with NR and adjusted at 106 cells/ml were applied to

glass slides, air dried, mounted in a fluorescent mounting me-

dium (Dako), coverslipped, and stored in the dark at 4�C before

microscopical examinations with a CLSM Leica TCS SPL

equipped with a UV/visible laser (Spectra Physics 2018, Spectra

Physics, Montain View, CA). Preliminary studies were per-

formed to determine the most convenient excitation line (488,

514, and 532 nm) and the excitation at 488 nm was selected.

Similarly, preliminary experiments were also performed to deter-

mine the most convenient band pass filter settings able to differ-

entiate emissions. As a result, yellow, orange, and red fluores-

cence emissions could not be differentiated properly and

advanced spectral analysis was therefore processed (7,10).

Images at 0.1 lm (x,y) pixel sizes were then obtained in

512 3 512 matrices depending on required sizes of the nuclei at

633 magnification of the CLSM. Advanced acquisitions per-

formed in the spectral analysis mode (20,36) via the RT30/70

dichroic mirror of the CLSM resulted in spectral sequences of 30

images selected inside successive 10-nm filters in the 525–

695 nm emission range. Sequences of images were further pro-

cessed by the FAMIS algorithm (21), available at Apteryx under

the name of Pixies (www.apteryx.fr). FAMIS synthesizes image

sequences into a reduced number of images called factor images

and curves called factor curves corresponding to fluorescent

stains (22,23,37). To improve interpretation of the results, factor

images are superimposed. Each factor image is coded in a differ-

ent color. Regular acquisitions were also performed in the

orange/red range (570–620 nm) and joined to the results to

facilitate the interpretation. When inhibitors are involved, four

factors were required to cope with the number of possible color

expressions and the presence of residuals. In other cases, three

factors only were required.

Thebasic idea of FAMIS is to process the curves that repre-

sent the evolution of fluorescence intensity of each pixel (x, y) in

the image spectral sequence S(x, y, k) (Fig. 1). FAMIS assumes

that each pixel is a mixture of different fluorescent patterns and

aims unmixing them. It assumes that each curve is a positive

weighted sum of fundamental curves called factors. Factors cor-

respond to spectral emission profiles Pk(k) of the different fluor-ochromes inside the slide. For each factor, the set of positive

weights computed for each curve yields one image called factor

image ak (x,y).

Factors are estimated in a two-step procedure from the

image sequence.

Pixels of the images are combined into small square clus-

ters. Fluorescent image intensity curves are computed for each

cluster and define the columns of an N 3 P rectangular matrix,

denoted X, where N is the total number of images in the

sequence, and P, the total number of clusters.

The first step, called orthogonal analysis, considers the

singular value decomposition of the matrix X, using an appro-

priate weighting of the curves. The matrix is reconstituted

using singular vectors associated with the largest singular

values. The number of vectors is determined according to the

number of factors in the model. Two, three, or four vectors are

usually required. This step filters the noise of the experimental

curves and reduces the dimension problem of the factor

search.

The second step, called oblique analysis, aims to estimate

factors representing the fundamental curves. After the orthog-

onal analysis, each curve is a linear combination of the first

singular vectors. Because of orthogonal conditions, these vec-

tors have both positive and negative values. As a consequence,

iterative algorithms have been proposed to find an appropriate

solution, i.e. factors with positive weights of the curves on

these factors. In the case of spectral analysis, factors were

selected by positive constraints. Factor images were recom-

puted in the original sampling by oblique projection on the

ORIGINAL ARTICLE

Cytometry Part A � 71A: 550�562, 2007 553

Figure 1. Basic presentation of FAMIS. To process the curves that represent the evolution of fluorescence intensity of each pixel (x, y) inthe image spectral sequence, factors are estimated in a two-step procedure from the image sequence. Pixels of the images are combined

into clusters. Fluorescent image intensity curves are computed for each cluster and define the columns of an N 3 P rectangular matrix,denoted X, where N is the total number of images in the sequence, and P, the total number of clusters. The first step, called orthogonalanalysis, considers the Singular Value Decomposition (SVD) of the matrix X. The matrix is reconstituted using singular vectors associated

with the largest singular values. Here, three vectors are required. In the second step, called oblique analysis, each curve is a linear combi-

nation of the first singular vectors. Factor images are recomputed in the original sampling by oblique projection on the factors, and the

computation is made by least squares minimization of distance D.

Figure 2. Analysis of caspase total activity, and of caspase-2 activation by conventional fluorescence microscopy, flow cytometry, colori-

metric assay and Western blot analyses. U937 were cultured during 1, 4, 6, 14, 18, and 24 h in the absence (control) or in the presence of

7KC (40 lg/ml) associated or not with z-VAD-fmk (broad-spectrum caspase inhibitor), or with z-VDVAD-fmk (caspase-2 inhibitor), or in the

presence of VP16 (50 lM) during 1 or 4 h. Detection and quantification of broad-spectrum caspase and caspase-2 activities with fluoro-

chrome-labeled inhibitors of caspases (FLICA): (A) fam-VDVAD-fmk��negative untreated cells counterstained with Hoechst 33342 (absenceof caspase-2 activity); (B) fam-VDVAD-fmk��positive 7KC-treated cells (presence of caspase-2 activity in cells with condensed, fragmented,and swollen nuclei); cn: condensed nuclei; fn: fragmented nuclei; sn: swollen nuclei; (C) quantification of fam-VAD-fmk��positive cells(broad-spectrum caspase positive cells) and of fam-VDVAD-fmk��positive cells (caspase-2 positive cells) by flow cytometry; shaded histo-grams correspond to spontaneous fluorescence of the cells. (D) Detection of caspase-2L pro-form by Western blot analyses. U937 cells

were cultured in the absence (control) or presence of 7-ketocholesterol or of VP16. At the end of the incubation time, cell extracts were sub-

ject to SDS/PAGE, and immunoblotted with an anti-caspase-2L mouse monoclonal antibody. In these conditions, a marked decrease of the

pro-form of the caspase-2L (48 kDa) is observed both in 7KC- and VP16-treated cells. (E) Caspase-2 activity measured with a caspase-2 col-

orimetric assay in U937 cells cultured in the absence (control) or in the presence of 7-ketocholesterol (7KC) or of VP16. (F) Percentage of

cells with active caspase-2 determined with the fluorochrome-labeled inhibitors of caspases (FLICA) method in U937 cells cultured in the

absence (control) or in the presence of 7-ketocholesterol (7KC) or of VP16. * indicates statistical significant differences (P < 0.05) betweenuntreated (control) and 7KC- (or VP16) treated cells.

factors, and the computation was made by least squares mini-

mization of distance D:

D ¼ RRðSðx; y; kÞ � Rakðx; yÞ � PkðkÞÞ2

Statistical Analysis

Statistical analyses were performed with SigmaStat 2.03

software (Systat Software, Richmond, CA) with the Student

ORIGINAL ARTICLE

554 Effects of Caspase Inhibitors on Nile Red Fluorescence Pattern

Figure 2.

ORIGINAL ARTICLE

Cytometry Part A � 71A: 550�562, 2007 555

t test. Data were considered statistically different at a P-value

of 0.05 or less.

RESULTS

Relationships Between Caspase Total Activity,

Caspase-2 Activity, and Polar Lipid Accumulation:

Analysis by Conventional Fluorescence

Microscopy and Flow Cytometry

In previous investigations, we demonstrated that the

maximum proportions of U937 cells simultaneously charac-

terized by a typical morphology of apoptosis (condensed and/

or fragmented nuclei) and by an important accumulation of

polar lipids measured by staining with NR, were identified at

18 h when they are cultured in the presence of 7KC used at

20–40 lg/ml (7,10). Therefore in the present study, the rela-

tionships between polar lipid accumulation and caspase activ-

ity were analyzed in the same experimental conditions. The

involvement of caspase activity occuring during 7KC-induced

cell death was confirmed by the in situ detection of activated

caspases with FLICA using fam-VAD-fmk; and the contribu-

tion of caspase-2 was assayed with FLICA using fam-VDVAD-

fmk, by Western blot and by colorimetric assay for caspase-2

protease activity (Fig. 2). Data obtained with FLICA clearly

show an absence of caspase activity in untreated cells whereas

some caspase positive cells (84% 6 4%) were detected under

treatment with 7KC (Figs. 2A–2C). They also reveal the pre-

sence of caspase-2 activity among a high percentage (77% 64%) of 7KC-treated cells corresponding either to apoptotic

cells (condensed and/or fragmented nuclei) or to oncotic cells

(swollen nuclei) (Figs. 2A–2C) (35). Moreover, under treat-

ment with 7KC and comparatively to untreated cells, a marked

reduction of the pro-form of caspase-2L (the proapoptotic

isoform of caspase-2) was detected by Western blotting. We

also observed a significant time dependent increase of caspase-

2 activity and of the percentage of cells with active caspase-2

(Figs. 2D–2F). These different data were validated by those

obtained with VP16 (Figs. 2C–2F), which is known to activate

numerous caspases, including caspase-2 (38), and which are in

agreement with previous investigations (39,40). So, taken

together, our data strongly support an activation of caspase-2

under treatment with 7KC. Interestingly, whereas the broad spec-

trum caspase inhibitor (z-VAD-fmk) strongly reduces the pro-

portions of cells with fragmented nuclei as well as the cells

with swollen nuclei, the caspase-2 inhibitor (z-VDVAD-fmk)

has no effect on nuclear morphology (Fig. 3A). However, the

important increase of the orange/red fluorescence of NR (cor-

responding to an accumulation of polar lipids) (7) observed

in the presence of 7KC was strongly counteracted when the

cells are simultaneously treated either with z-VAD-fmk or

with z-VDVAD-fmk (Fig. 3B). These observations were in

agreement with those performed by conventional fluorescence

microscopy. Thus, when the cells were cultured with 7KC asso-

Figure 3. Effects of z-VAD-fmk and z-VDVAD-fmk on nuclear mor-

phology and Nile Red fluorescence. (A) Effects of z-VAD-fmk

(200 lM) and z-VDVAD-fmk (100 lM) on themorphological aspect ofthe nuclei (condensed, fragmented and swollen) identified by stain-

ing with Hoechst 33342; (B) Effects of z-VAD-fmk (200 lM) and z-VDVAD-fmk (100 lM) on polar lipid accumulationmeasured by flowcytometry with NR. * indicates statistical significant differences (P <0.05) between 7KC- and (7KC 1 z-VAD-fmk)-treated cells, and

between 7KC- and (7KC1 z-VDVAD-fmk)-treated cells.

Figure 4. Effects of 7-ketocholesterol on neutral and polar lipids:

biochemical analyses and microscopical observations performed

after staining with Nile Red. U937 cells were cultured during 18–24h in the absence (control) or in the presence of 7-ketocholesterol

(7KC, 20–40 lg/ml), and the effect of this oxysterol on the cellularcontent in neutral and polar lipids was determined by fluorescence

microscopy, after staining with Nile Red (NR) (0.1 lg/ml) whichemits a yellow or orange-red fluorescence in the presence of neu-

tral and polar lipids, respectively, when it is excited at 488 nm.

(A,B) Images correspond to untreated (control) and 7KC (40 lg/ml,18 h)-treated cells, respectively, observed by conventional fluores-

cence microscopy after staining with NR. Various numbers of

punctuated yellow cytoplasmic structures are observed in

untreated cells, whereas large orange-red cytoplasmic are mainly

present in 7KC-treated cells. (C) Biochemical analyses of lipids by

CGC/MS in untreated (control) and 7KC (20 lg/ml, 24 h)-treatedcells. Data are mean 6 SEM of three independent experiments. *

indicates statistical significant differences (P < 0.05) between

untreated (control) and 7KC-treated cells. (D–I) Multispectral analy-sis of untreated (control) and 7KC (40 lg/ml, 18 h)-treated U937cells stained with NR and observed by confocal laser scanning mi-

croscopy. On untreated cells, small yellow cytoplasmic fluorescent

spots are observed (D–E), and when FAMIS is processed and threefactors are required, an orange emission (594 nm) is visualized in

untreated cells in the first factor image, and a second factor image

corresponds to a yellow emission (582 nm). (F) The third factor

image corresponds to residuals. On 7KC-treated cells, only few or

no yellow fluorescent spots are detected (G,H), and when FAMIS is

processed and three factors are required, a red emission (605 nm)

is visualized in a first factor image and the second and third factor

images correspond to residuals (I). (F,I) corresponding three factor

curves associated to factor images (E), and (H), respectively.

ORIGINAL ARTICLE

556 Effects of Caspase Inhibitors on Nile Red Fluorescence Pattern

Figure 4.

ORIGINAL ARTICLE

Cytometry Part A � 71A: 550�562, 2007 557

ciated or not with z-VAD-fmk or z-VDVAD-fmk, the number

of yellow cytoplasmic fluorescent spots was higher in the pre-

sence of caspase inhibitors (Data not shown). Interestingly, no

caspase-6 activity was detected in 7KC-treated U937 cells (35),

and no effects of the caspase-6 inhibitor (z-VEID-fmk,

100 lM) was found on the orange/red fluorescence of NR on

untreated and 7KC-treated cells (Figs. 3A–3B). Noteworthy,

when the cells were cultured in the presence of DMSO (0.1–

0.2% final concentrations) used to dilute caspase inhibitors and

VP16, no cytotoxic effects were observed (the percentages of cells

with condensed/fragmented or swollen nuclei were similar in

untreated and DMSO-treated cells), no caspase activity was

detected, and no effect on caspase-2 activation was found (Data

not shown). Taken together, the present observations show that

the broad caspase inhibitor z-VAD-fmk and that the caspase-2

inhibitor z-VDVAD-fmk are capable to counteract the increase

of the orange/red fluorescence of NR triggered by 7KC, and con-

sequently that they can reduce polar lipid accumulation. So,

these results hint a possible relationship between caspase activity

and polar lipid accumulation, and they strengthen data obtained

in previous studies evocating a potential role of caspase-2 in

lipid metabolism (17).

Spectral Imaging Microscopy by Confocal Laser

Microscopy and Subsequent Image Processing by

Factor Analysis of Medical Image Sequences

In agreement with our previous investigations (7,10),

the analysis of the cellular lipid content in untreated and

7KC (20–40 lg/ml)-treated U937 cells taken at 18 and 24 h

of culture revealed important differences of lipid profile

when the investigations are performed by conventional

fluorescence microscopy after staining with NR or by CGC

coupled with MS (Fig. 4). Thus, after staining with NR, in

untreated cells, punctuated yellow cytoplasmic structures

were observed but some cellular areas emitting a slight or-

ange-red fluorescence were also detected (Fig. 4A). How-

ever, in the presence of 7KC, cells with large orange-red

cytoplasmic structures (containing yellow spots or not)

were mainly observed (Fig. 4B). These observations which

suggest an accumulation of polar lipids in 7KC-treated cells

were confirmed and validated by subsequent analyses of

cellular extracts by CGC/MS. Indeed, in 7KC-treated cells,

comparatively to untreated cells, an important increase of

polar lipids (phosphatidylcholine, sphingomyelin) as well as

of cholesterol was found (Fig. 4C).

In the case of CLSM spectral observations of human

untreated and 7KC-treated monocytic U937 cells cultured in

the absence or in the presence of z-VAD-fmk or z-VDVAD-

fmk, fluorescence emissions were collected through narrow

band-pass filters then processed before interpretation. The exci-

tation at 488 nm was performed and NR emission was col-

lected in the spectral mode in 10 nm filters from blue to red

(525 � 695 nm). The resulting spectral sequences were investi-

gated by means of FAMIS. Regular acquisitions were also per-

formed in the orange/red range (570–620 nm) and joined to

the figures in which color superimposition of factor images is

included as well as factor curves. Factor images corresponding

to color expressions were coded in yellow, orange and red and

the remaining one, which corresponds to residuals was coded

in green.

In the case of untreated monocytic U937 cells stained

with NR (Figs. 4D–4F), when FAMIS is processed and three

factors are required, an orange emission (594 nm) is visualized

in the cells in the first factor image and a second factor image

corresponds to a yellow emission (582 nm) (Fig. 4F). The

third factor image corresponds to residuals.

In the case of 7KC-treated U937 cells stained with NR

(Figs. 4G–4I), when FAMIS is processed and three factors are

required, a red emission (605 nm) is only visualized in a first

factor image. The second and third factor images correspond

to residuals (Fig. 4I).

In the case of U937 cells cultured with z-VAD-fmk and

stained with NR (Figs. 5A–5C), when FAMIS is processed and

three factors are required, an orange emission (594 nm) is

visualized in a first factor image and a yellow emission

(582 nm) is visualized in a second factor image while the third

factor image corresponds to residuals (Fig. 5C).

In the case of U937 cells cultured with z-VDVAD-fmk

and stained with NR (Figs. 5D–5F), when FAMIS is pro-

cessed and three factors are required, an orange emission

(594 nm) is visualized in a first factor image and a yellow

emission (582 nm) is visualized in a second factor image

while the third factor image corresponds to residuals

(Fig. 5F).

In the case of 7KC-treated U937 cells cultured in the

presence of z-VAD-fmk and stained with NR (Fig. 5G–5I),

when FAMIS is processed and three factors are required, a

red emission (611 nm) combined with a yellow emission

(582 nm) is visualized in a first factor image and an orange

emission (594 nm) is visualized in a second factor image

while the third factor image corresponds to residuals. When

four factors are required the yellow and red emissions are

unmixed (Fig. 5I).

In the case of 7KC-treated U937 cells cultured in the

presence of z-VDVAD-fmk and stained with NR (Figs. 5J–

5L), when FAMIS is processed and three factors are

required, a red emission (611 nm) is visualized in a first

factor image and a yellow emission (576 nm) is visualized

in a second factor image while the third factor image corre-

sponds to an orange emission (594 nm). When four factors

are required, the extra factor corresponds to residuals (Fig.

5L).

In all the cases, the eye validated the correspondence

between factor curves and emission spectra, and between fac-

tor images and stained objects.

Taken together, these data show that yellow, orange, and

red fluorescence emissions are mainly observed in untreated

cells, whereas in 7KC-treated cells orange/red fluorescence is

only detected. These observations are in agreement with an

accumulation of polar lipids in 7KC-treated cells (7). In z-

VAD-fmk- or z-VDVAD-fmk-treated cells, yellow, orange, and

red fluorescence emissions are also simultaneously observed.

Interestingly, in (7KC 1 z-VAD-fmk)- or (7KC 1 z-VDVAD-

fmk)-treated cells, red fluorescence is not the principal compo-

ORIGINAL ARTICLE

558 Effects of Caspase Inhibitors on Nile Red Fluorescence Pattern

Figure 5. Multispectral analysis of 7-ketocholesterol-treated human monocytic U937 cells cultured with z-VAD-fmk or z-VDVAD-fmk,

stained with Nile Red, and observed by confocal laser scanning microscopy. U937 cells were cultured for 18 h in the presence of of z-VAD-

fmk (100 lM) or z-VDVAD-fmk (200 lM), and in the presence of 7-ketocholesterol (7KC) used at 40 lg/ml associated with z-VAD-fmk or z-VDVAD-fmk. Cells were stained with NR (0.1 lg/ml). On the cells cultured with z-VAD-fmk, when regular and/or spectral analyses by meansof confocal laser scanning microscopy (CLSM) are used, some yellow fluorescent spots are observed in the cytoplasm (A,B), and when

FAMIS is processed and three factors are required, an orange emission (594 nm) is visualized in a first factor image and a yellow emission

(582 nm) is visualized in a second factor image while the third factor image corresponds to residuals (C). On the cells cultured with z-

VDVAD-fmk, when regular and/or spectral analyses with CLSM are used, some yellow cytoplasmic structures are observed (D,E), and

when FAMIS is processed and three factors are required, a red emission (611 nm) combined with a yellow emission (582 nm) is visualized

in a first factor image, and an orange emission (594 nm) is visualized in a second factor image, while the third factor image corresponds to

residuals (F). On 7KC-treated cells cultured with z-VAD-fmk (G,H) or z-VDVAD-fmk (J,K), when FAMIS is processed on 7KC-treated cells cul-

tured with z-VAD-fmk and four factors are required, a red emission (611 nm) and a yellow emission (582 nm) are visualized in the first two

factor images and an orange emission (594 nm) is visualized in a third factor image while the fourth factor image corresponds to residuals

(I). When FAMIS is processed on 7KC-treated cells cultured with z-VDVAD-fmk, and four factors are required, a red emission (611 nm) is

visualized in a first factor image and a yellow emission (576 nm) is visualized in a second factor image while the third factor image corre-

sponds to an orange emission (594 nm). The fourth factor image corresponds to residuals (L).

nent, and as in untreated cells, yellow and orange components

are also detected. These data show that caspase inhibitors,

including caspase-2 inhibitor, are able to counteract 7KC-

induced polar lipid accumulation. They support the hypothesis

that caspases are involved in polar lipid accumulation.

DISCUSSION

Among oxysterols, those resulting from a spontaneous

oxidation of cholesterol at C7 such as 7-ketocholesterol

(7KC) are present at enhanced concentrations in the plasma

of atherosclerotic subjects and in atheromatous plaques

(1,4,5). The 7KC is one of the most prominent oxysterols

found in atherosclerotic lesion, and it has been shown that

it is capable to favor polar lipid accumulation, and to

induce a complex mode of cell death with some characteris-

tics of apoptosis (41). As these biological activities contri-

bute to the initiation and the development of atherosclero-

sis (42), it is important to precise the relationships between

these different effects, and to identify molecules capable to

counteract them in order to identify and develop new anti-

atherosclerotic treatments. So, in the present study, we

investigated the relationships between caspase activity and

polar lipid accumulation by using caspase inhibitors on

human monocytic U937 cells treated with 7KC and stained

with NR, allowing to differentiate neutral and polar lipids

(8,9), and constituting a clinical flow cytometric biomarker

to characterize phospholipidosis (43). To this end we used

FCM, and developed an original spectral method of investi-

gation by CLSM combined with subsequent image process-

ing by the FAMIS algorithm to differentiate fluorescence

emission spectra and get corresponding specific images.

Noteworthy, our data obtained by fluorescence micros-

copy and FCM with FLICA, confirm and underline that 7KC

triggers a caspase dependent mode of cell death by apoptosis

(35), and they also favor an activation of caspase-2 supported

by three different complementary methods: flow cytrometry

with FLICA, Western blot analyses, and measurement of cas-

pase-2 activity by colorimetric assays. Moreover, on 7KC trea-

ted cells stained with NR, the increase of red fluorescence in-

tensity observed by FCM, and the marked red fluorescent pro-

file detected by spectral imaging were in agreement with the

marked accumulation of phospholipids (phosphatidylcholine,

sphingomyelin) shown by biochemical analyses, which evo-

cates phospholipidosis (44). Consequently, it was of interest to

investigate the relationships between cell death and polar lipid

accumulation with the use of the wide spectrum caspase in-

hibitor (z-VAD-fmk) and of the caspase-2 inhibitor (z-

VDVAD-fmk) by regular and spectral analysis by CLSM and

by FAMIS processing of 3D-image sequences.

In this study, the ability to unmix yellow, orange, and red

fluorescence emissions of lipids stained with NR has been

shown by using CLSM and FAMIS. Problems resulting from

the superimposition of emissions are solved by means of

sequences of images obtained on one photomultiplier detector

of the CLSM by selection of emissions through 10-nm filters

in the range of 525–695 nm. Sequences of 30 images character-

ize the global fluorescence emission, and are easy to get in

CLSM. Processing these sequences is straightforward but

interpretation of the results requires some a priori knowledge

of what is expected: neutral lipids are yellow, polar lipids are

orange/red, and noise is present. Also, the number of required

factors has to be selected; it depends on the number of mixed

colors and noise. Thus, the described method is well adapted

to differentiate fluorescence signals inside a mixture of yellow

and red emissions comprising auto-fluorescence, taking into

account differences in spectral patterns of fluorochromes. We

verified that it was possible to differentiate targets, as already

shown in HeLa cells that we used as model to detect and char-

acterize in situ hybridisation spots (23). Concerning the char-

acterization of specific stainings, it was possible to differentiate

fluorescence emissions by their spectral patterns within a

reduced range of observations by combining sequences of

images into 3D sequences rather than a change of filter set-

tings, by processing these sequences through FAMIS.

In these conditions, spectral analysis and subsequent

FAMIS analysis of 3D sequences of confocal images per-

mitted to show the effect of the wide caspase inhibitor (z-

VAD-fmk), and of the caspase-2 inhibitor (z-VDVAD-fmk)

in the distribution of neutral and polar lipids on 7KC-trea-

ted cells. Multispectral analysis provides differentiated yel-

low, orange, and red emissions of fluorescence of NR on

corresponding images when cell deposits were screened in

the blue (488 nm) excitation mode, which optimized the

fluorescence expression of NR and permits to distinguish

neutral and polar lipids which are stained in yellow and or-

ange/red, respectively (8,9). In agreement with data

obtained by FCM, CLSM revealed the modifications in the

fluorescence of NR (decreased orange/red fluorescence

emission, and presence of a yellow fluorescence emission)

when 7KC-treated cells are simultaneously incubated with

z-VAD-fmk and z-VDVAD-fmk, suggesting that caspases,

such as caspase-2, can be involved in the regulation of polar

lipid accumulation. As factor curves and images show the

narrow yellow, orange, and red emissions of NR, a reliable

evaluation of the effect of z-VAD-fmk or z-VDVAD-fmk is

obtained. Therefore, our data underline that NR is a useful

and suitable marker to detect polar lipids (43,45), and that

spectral analysis and subsequent image processing might

constitute a useful tool to analyze the effect of chemicals

on lipid content, especially to follow and to characterize

the lipid changes associated with oxysterols such as 7KC,

present at important levels in the atherosclerotic plaque (5).

As a consequence, our data highlight that spectral analysis

provides a new tool to analyze the in situ content and evo-

lution of lipids inside cells by means of fluorescence and

image distribution. Thus, the technology and the methodol-

ogy which we described and which allow to perform impar-

tial analyses of NR stained cells, can constitute an helpful

tool to the biologists and the pathologists to characterize

the lipid content on isolated cells or tissue sections prior

the use of any sophisticated biochemical or physical techni-

ques such as gas chromatography coupled with MS, mag-

netic resonance imaging, or Raman microscopy (46). How-

ever, additional biophysical analyses are still needed to

ORIGINAL ARTICLE

560 Effects of Caspase Inhibitors on Nile Red Fluorescence Pattern

determine with accuracy routinely whether a particular

threshold of fluorescence can be defined, especially between

the yellow and orange fluorescence of NR, in order to es-

tablish with accuracy cellular changes in neutral and polar

lipids. Consequently, a better knowledge of the values of

the wavelengths resulting from the interaction of neutral

and polar lipids with NR is required to improve the in situ

characterization of these lipids and to establish precise lipid

profiles by spectral imaging.

In conclusion, results obtained by FCM with FLICA

and NR underline that 7KC induces a caspase dependent

mode of apoptosis associated with an activation of caspase-

2, and that this oxysterol favors polar lipid accumulation.

Moreover, as FCM and spectral imaging analysis reveal that

the caspase inhitors (z-VAD-fmk, z-VDVAD-fmk) can mod-

ulate lipid accumulation, the results reinforce the hypothesis

that some relationships will exist between caspase activation

(including caspase-2) and lipid metabolism. Thus, the anal-

ysis of lipids based on NR emission spectra obtained by

CLSM and FAMIS processing of 3D-image sequences con-

stitutes a new approach to the in situ characterization of

lipids, which might have numerous biological and pharma-

cological applications.

ACKNOWLEDGMENTS

We thank Pr Eric Solary (INSERM UMR 866, Dijon,

France) for his helpful comments and suggestions, and Mrs

Anne Athias (INSERM UMR 866/Inserm IFR STIC Sant�e,Dijon, France) for the analyses by gas chromatography

coupled with mass spectrometry.

LITERATURE CITED

1. Schroepfer GJ Jr. Oxysterols: Modulators of cholesterol metabolism and other pro-cesses. Physiol Rev 2000;80:361–554.

2. Guardiola F, Codony R, Addis PB, Rafecas M, Boatella J. Biological effects of oxyster-ols: Current status. Food Chem Toxicol 1996;34:193–211.

3. Massey JB. Membrane and protein interactions of oxysterols. Curr Opin Lipidol2006;17:296–301.

4. Colles SM, Maxson JM, Carlson SG, Chisolm GM. Oxidized LDL-induced injury andapoptosis in atherosclerosis. Potential roles for oxysterols. Trends Cardiovasc Med2001;11:131–138.

5. Brown AJ, Jessup W. Oxysterols and atherosclerosis. Atherosclerosis 1999;142:1–28.

6. Miguet-Alfonsi C, Prunet C, Monier S, Bessede G, Lemaire-Ewing S, Berthier A,M�en�etrier F, N�eel D, Gambert P, Lizard G. Analysis of oxidative processes and ofmyelin figure formation before and after the loss of mitochondrial transmem-brane potential during 7b-hydroxycholesterol and 7-ketocholesterol-induced ap-optosis: comparison with various pro-apoptotic chemicals. Biochem Pharmacol2002;64:527–541.

7. Vejux A, Kahn E, Dumas D, Bessede G, M�en�etrier F, Athias A, Riedinger JM, FrouinF, Stoltz JF, Ogier-Denis E, Todd-Pokropek A, Lizard G. 7-ketocholesterol favors lipidaccumulation and colocalizes with Nile Red positive structures formed during 7-ketocholesterol-induced apoptosis: Analysis by flow cytometry, FRET biphoton spec-tral imaging microscopy, and subcellular fractionation. Cytometry Part A 2005;64A:87–100.

8. Greenspan P, Mayer EP, Fowler SD. Nile red: A selective fluorescent stain for intracel-lular lipid droplets. J Cell Biol 1985;100:965–973.

9. Klinkner AM, Bugelski PJ, Waites CR, Louden C, Hart TK, Kerns WD. A novel tech-nique for mapping the lipid composition of atherosclerotic fatty streaks by en facefluorescence microscopy. J Histochem Cytochem 1997;45:743–753.

10. Kahn E, Vejux A, Dumas D, Montange T, Frouin F, Robert V, Riedinger JM, Stoltz JF,Gambert P, Todd-Pokropek A, Lizard G. FRETmultiphoton spectral imaging micros-copy of 7-ketocholesterol and Nile Red in U937 monocytic cells loaded with 7-keto-cholesterol. Anal Quant Cytol Histol 2004;26:304–313.

11. Maior I, Mandel H, Aviram M. Macrophage uptake of oxidized LDL inhibits lysoso-mal sphingomyelinase, thus causing the accumulation of unesterified cholesterol-sphigomyelin-rich particles in the lysosomes. A possible role for the 7-ketocholes-terol. Arterioscler Thromb Vasc Biol 1995;15:1378–1387.

12. Wang X, Zelenski NG, Yang J, Sakai J, Brown MS, Goldstein JL. Cleavage of sterolregulatory element binding protein (SREBPs) by CPP32 during apoptosis. EMBO J1996;15:1012–1020.

13. Magana MM, Osborne TF. Two tandem binding sites for sterol regulatory elementbinding proteins are required for sterol regulation of fatty-acid synthase promoter.J Biol Chem 1996;271:32689–32694.

14. Ericsson J, Jackson SM, Lee BC, Edwards PA. Sterol regulatory element binding pro-tein binds to a cis element in the promoter of the farnesyl diphosphate synthase gene.Proc Natl Acad Sci USA 1996;93:945–950.

15. Wang X, Briggs MR, Hua X, Yokoyama C, Goldstein JL, Brown MS. Nuclearprotein that binds sterol regulatory element of low density lipoprotein receptorpromoter. II. Purification and characterization. J Biol Chem 1993;268:14497–14504.

16. Castedo M, Perfettini JL, Roumier T, Valent A, Raslova H, Yakushijin K, Horne D,Feunteun J, Lenoir G, Medema R, Vainchenker W, Kroemer G. Mitotic catastropheconstitutes a special case of apoptosis whose suppression entails aneuploidy. Onco-gene 2004;23:4362–4370.

17. Logette E, Le Jossic-Corcos C, Masson D, Solier S, Sequeira-Legrand A, Dugail I,Lemaire-Ewing S, Desoche L, Solary E, Corcos L. Caspase-2, a novel lipid sensorunder the control of sterol regulatory element binding protein 2. Mol Cell Biol2005;25:9621–9631.

18. Rodriguez I, Matsuura K, Ody C, Nagata S, Vassalli P. Systemic injection of a tripep-tide inhibits the intracellular activation of CPP32-like proteases in vivo and fully pro-tects mice against Fas-mediated fulminant liver destruction and death. J Exp Med1996;184:2067–2072.

19. Jangi SM, Diaz-Perez JL, Ochoa-Lizarralde B, Martin-Ruiz I, Asumendi A, Perez-Yarza G, Gardeazabal J, Diaz-Ramon JL, Boyano MD. H1 histamine receptor antago-nists induce genotoxic and caspase-2-dependent apoptosis in human melanoma cells.Carcinogenesis 2006;27:1787–1796.

20. Garini Y, Young IT, McNamara G. Spectral imaging: principles and applications.Cytometry Part A 2006;69A:735–747.

21. Frouin F, Cinotti L, Benali H, Buvat I, Bazin JP, Millet P, Di Paola R. Extraction offunctional volumes from medical dynamic volumetric datasets. Comput Med Ima-ging Graph 1993;17:397–404.

22. Kahn E, Lizard G, P�el�egrini M, Frouin F, Roignot P, Chardonnet Y, Di Paola R. Four-dimensional factor analysis of confocal images sequences (4D-FAMIS) to detect andcharacterize low numbers of human papillomavirus DNA by FISH in HeLa and SiHacells. J Microsc 1999;193: 227–243.

23. Kahn E, Lizard G, Frouin F, Roignot P, Chardonnet Y, Di Paola R. Laser scanningconfocal microscopy and factor analysis of biomedical image sequences (FAMIS) todetect and characterize HPV DNA sequences by FISH in HeLa cells. Cytometry1997;28:269–279.

24. Kahn E, Hotmar J, Bernheim A. FISH 3D chromosomal studies in nuclei from cytoge-netic preparations by confocal microscopy. Anal Quant Cytol Histol 1996;18:109–114.

25. Kahn E, Hotmar J, Frouin F, Di Paola M, Bazin JP, Di Paola R, Bernheim A. Spectraland dynamic confocal fluorescence characterization of cytogenetic preparations. AnalCell Pathol 1996;12:45–54.

26. Kahn E, Frouin F, Hotmar J, Di Paola R, Bernheim A. Confocal fluorescence analysisof the depth and focus of cytogenetic preparations. Anal Quant Cytol Histol1997;19:404–412.

27. Lizard G, Gueldry S, Sordet O, Monier S, Athias A, Miguet C, Bessede G, Lemaire S,Solary E, Gambert P. Glutathione is implied in the control of 7-ketocholesterol-induced apoptosis, which is associated with radical oxygen species production.FASEB J 1998;12:1651–1663.

28. Bedner E, Smolewski P, Amstad P, Darzynkiewicz Z. Activation of caspases measuredin situ by binding of fluorochrome-labeled inhibitors of caspases (FLICA): Correla-tion with DNA fragmentation. Exp Cell Res 2000;259:308–313.

29. Grabarek J, Amstad P, Darzynkiewicz Z. Use of fluorescently labeled caspase inhibi-tors as affinity labels to detect activated caspases. Hum Cell 2002;15:1–12.

30. Droin N, Rebe C, Bichat F, Hammann A, Bertrand R, Solary E. Modulation of apo-ptosis by procaspase-2 short isoform: Selective inhibition of chromatin condensation,apoptotic body formation and phosphatidylserine externalization. Oncogene 2001;20:260–269.

31. Smith PK, Krohn RI, Hermanson GT, Mallia AK, Gartner FH, Provenzano MD, Fuji-moto EK, Goeke NM, Olson BJ, Klenk DC. Measurement of protein using bicincho-ninic acid. Anal Biochem 1985;150:76–86.

32. Folch J, Lees M, Sloane Stanley GH. A simple method for the isolation and purifica-tion of total lipids from animal tissues. J Biol Chem 1957;226:497–509.

33. Miguet C, Monier S, Bettaieb A, Bessede G, Laubriet A, Lemaire S, N�eel D, GambertP, Lizard G. Ceramide generation occuring during 7b-hydroxycholesterol- and 7-ketocholesterol-induced apoptosis is caspase-independent and is not required to trig-ger cell death. Cell Death Differ 2001;8:83–99.

34. Lizard G, Fournel S, Genestier L, Dhedin N, Chaput C, Flacher M, Mutin M, PanayeG, Revillard JP. Kinetics of plasma membrane and mitochondrial alterations in cellsundergoing apoptosis. Cytometry 1995; 21:275–283.

35. Prunet C, Lemaire-Ewing S, M�en�etrier F, N�eel D, Lizard G. Activation of caspase-3-dependent and -independent pathways during 7-ketocholesterol- and 7b-hydro-xycholesterol-induced cell death: A morphological and biochemical study. J Bio-chem Mol Toxicol 2005; 19:311–326.

36. Hiraoka Y, Shimi T, Haraguchi T. Multispectral imaging fluorescence microscopyfor living cells. Cell Struct Funct 2002;27:367–374.

37. Kahn E, Frouin F, Souchier C, Bernengo JC, Bruzzoni-Giovanelli H, Clement O, FrijaG, Di Paola R, Calvo F, Linares-Cruz G. Confocal multilaser focusing and single-lasercharacterization of UV excitable stains of cellular preparations. Cytometry 2000;40:42–49.

ORIGINAL ARTICLE

Cytometry Part A � 71A: 550�562, 2007 561

38. Robertson JD, Enoksson M, Suomela M, Zhivotovsky B, Orrenius S. Caspase-2 actsupstream of mitochondria to promote cytochrome c release during etoposide-induced apoptosis. J Biol Chem 2002;277: 29803–29809.

39. Cathelin S, Rebe C, Haddaoui L, Simioni N, Verdier F, Fontenay M, Launay S,Mayeux P, Solary E. Identification of proteins cleaved downstream of caspase activa-tion in monocytes undergoing macrophage differentiation. J Biol Chem 2006;281:17779–17788.

40. Sordet O, Rebe C, Leroy I, Bruey JM, Garrido C, Miguet C, Lizard G, Plenchette S,Corcos L, Solary E. Mitochondria-targeting drugs arsenic trioxide and lonidaminebypass the resistance of TPA-differentiated leukemic cells to apoptosis. Blood2001;97:3931–3940.

41. Lyons MA, Brown AJ. 7-ketocholesterol. Int J Biochem Cell Biol 1999;31:369–375.

42. Gau GT, Wright RS. Pathophysiology, diagnosis, and management of dyslipidemia.Curr Probl Cardiol 2006;31:445–486.

43. Casartelli A, Bonato M, Cristofori P, Crivellente F, Dal Negro G, Masotto I, Muti-nelli C, Valko K, Bonfante V. A cell based approach for the early assessment of thephospholipidogenic potential in pharmaceutical research and drug development.Cell Biol Toxicol 2003;19:161–176.

44. Reasor MJ, Kacew S. Drug-induced phospholipidosis: Are there functional conse-quences? Exp Biol Med 2001;226:825–830.

45. Halstead BW, Zwickl CM, Morgan RE, Monteith DK, Thomas CE, Kowers RK, Ber-ridge BR. A clinical flow cytometric biomarker strategy: Validation of peripheral leu-kocyte phospholipidosis using Nile Red. J Appl Toxicol 2006;26:169–177.

46. Whitley A, Adar F. Confocal spectral imaging in tissue with contrast provided byRaman vibrational signatures. Cytometry Part A 2006; 69A:880–887.

ORIGINAL ARTICLE

562 Effects of Caspase Inhibitors on Nile Red Fluorescence Pattern