Download Resveratrol induces colon tumor cell apoptosis

Int. J. Cancer: 94, 615– 622 (2001)
© 2001 Wiley-Liss, Inc.
Publication of the International Union Against Cancer
RESVERATROL INDUCES COLON TUMOR CELL APOPTOSIS INDEPENDENTLY
OF p53 AND PRECEDED BY EPITHELIAL DIFFERENTIATION,
MITOCHONDRIAL PROLIFERATION AND MEMBRANE POTENTIAL COLLAPSE
Mojgan MAHYAR-ROEMER1, Alice KATSEN2, Pedro MESTRES2 and Klaus ROEMER3*
1
Internal Medicine IV, University of Saarland Medical School, Homburg/Saar, Germany
2
Institute for Anatomy, University of Saarland Medical School, Homburg/Saar, Germany
3
Department of Virology, University of Saarland Medical School, Homburg/Saar, Germany
Resveratrol, a polyphenol present in wine and grapes, can
inhibit tumor cell growth in vitro and tumorigenesis in vivo.
Some of its effects have been linked to activation of the p53
tumor suppressor; however, p53 is frequently mutated in
tumors, particularly in the common and often therapy-resistant colon cancers. Using the human wild-type p53-expressing HCT116 colon carcinoma cell line and HCT116 cells with
both p53 alleles inactivated by homologous recombination,
we show in the current study that resveratrol at concentrations comparable to those found in some foods can induce
apoptosis independently of p53. The cell death is primarily
mitochondria-mediated and not receptor-mediated. No cells
survived in cultures continuously exposed to 100 ␮M resveratrol for 120 hr. When compared with 5-FU, resveratrol stimulated p53 accumulation and activity only weakly and with
delayed kinetics and neither the increased levels nor the
activity affected apoptosis detectably. The apoptosis agonist
Bax was overproduced in response to resveratrol regardless
of p53 status, yet the kinetics of Bax expression were influenced by p53. Remarkably, apoptosis was preceded by mitochondrial proliferation and signs of epithelial differentiation.
Thus, resveratrol triggers a p53-independent apoptotic pathway in HCT116 cells that may be linked to differentiation.
© 2001 Wiley-Liss, Inc.
Key words: apoptosis; differentiation; mitochondria; resveratrol; p53
The antifungal phytoalexin resveratrol (3,5,4⬘-trihydroxy-transstilbene) is a constituent of many plant species and present at
particularly high levels in grapes and wine.1 As a polyphenol, it is
not only a potent inhibitor of radical formation (ED50 27 ␮M) but
acts as a pleiotropic effector molecule to inhibit initiation, promotion and progression of malignant transformation. Resveratrol inhibits the cyclooxygenase (ED50 15 ␮M) and hydroperoxidase
(ED50 3.7 ␮M) activities of COX-1 and the hydroperoxidase
activity of COX-2 (ED50 85 ␮M).2 Furthermore, it functions as an
inhibitor of platelet aggregation, a modulator of lipid and lipoprotein metabolism3 and an effective blocker of ribonucleotide reductase (ED50 100 ␮M) that provides deoxyribonucleotides for DNA
synthesis.4 It also inhibits DNA polymerase5 and mimics estradiol.6 Several of these activities constitute the basis for the chemopreventive action of resveratrol, exemplified by its antimutagenic
activity in the Ames assay, inhibition of tumorigenesis in a 2-stage
murine skin-cancer model2 and inhibition of cell proliferation in
vitro.7,8
The presence of the functional wild-type form of the p53 tumor
suppressor correlates with the sensitivity of mouse tumors and at
least some human tumors to therapeutic agents.9 –11 p53 is stabilized and activated as a transcription factor in response to a variety
of cellular stresses,12 the result being either cell-cycle arrest,
mostly through transcriptional activation of the p21WAF/Cip1 inhibitor of cyclin-dependent kinases, or apoptosis, depending on the
cell context. Apoptosis often involves changes to mitochondria.13
One of the major predictive parameters indicating commitment to
cell death appears to be selective MMP. In the normal mitochondrion, protons are pumped out of the matrix through the almost
impermeable inner mitochondrial membrane, resulting in a proton
gradient that establishes the transmembrane potential ⌬⌿m. Transient or permanent permeabilization of the inner mitochondrial
membrane entails transient or permanent ⌬⌿m collapse and, consequently, respiratory chain disturbances. This, then, favors the
production of ROS. ROS often come up late in apoptosis and
appear to accompany rather than precede cell death; however, they
have also been implicated in the early initiating events.14 p53 has
been documented to be required for the induction of apoptosis by
resveratrol in mouse cells15 and linked to the suppression of bovine
pulmonary artery endothelial cells by the drug.8 Here, we studied
the effect of resveratrol on the survival of a human colon-carcinoma cell line and a derivative in which both p53 alleles have been
disrupted by targeted homologous recombination.16
MATERIAL AND METHODS
Reagents and cell culture
Resveratrol and [6-Hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid] (Trolox威) were purchased from Alexis (San Diego,
CA); JC-1, MitoTracker green and rhodamine-123 were from
Molecular Probes (Eugene, OR); DCFH-DA, 5-FU, DAPI and PI
were from Sigma (St. Louis, MO); p53 MAb DO-1 was from
Calbiochem (San Diego, CA); p21, Bax, Fas receptor, FasL and
villin MAbs were from Transduction Laboratories/Pharmingen
(San Diego, CA); and ␤-actin and FITC-labeled anti-mouse MAbs
were from Sigma. A liquid alkaline phosphatase assay kit was
purchased from Sigma and used as recommended. Stock solutions
of resveratrol, rhodamine-123, MitoTracker green and DCFH-DA
were prepared in DMSO; JC-1 was dissolved in methanol; and
Trolox, 5-FU, DAPI and PI were dissolved in water. The caspase-3
inhibitor DEVD-CHO and the caspase-8 inhibitor IETH-CHO
were purchased from Calbiochem and dissolved in DMSO.
HCT116 cells and the p53-negative derivative HCT116 p53–/–16
were cultured as monolayers at 37°C in a humidified 7% CO2
atmosphere in McCoy’s 5A medium supplemented with 10% FCS.
FACS analysis
Cells were seeded in 6-well dishes to approximately 30% confluence 1 day prior to drug treatment. At various times, asynchroAbbreviations: COX-1, cyclooxygenase-1; DAPI, 4,6-diamidino-2-phenylindole; DCFH-DA, 2,7-dichlorofluorescin diacetate; 5-FU, 5-fluorouracil; MAb, monoclonal antibody; ⌬⌿m, mitochondrial membrane potential;
MMP, mitochondrial membrane permeabilization; PI, propidium iodide;
ROS, reactive oxygen species.
Grant sponsor: German Research Foundation; Grant number: SFB399
(B5).
*Correspondence to: Department of Virology, Bldg. 47, University of
Saarland Medical School, D-66421 Homburg/Saar, Germany.
Fax: ⫹49-6841-163980. E-mail: [email protected]
Received 25 January 2001; Revised 19 April 2001; Accepted 29 May
2001
616
MAHYAR-ROEMER ET AL.
FIGURE 1 – Effects of resveratrol
on the survival of HCT116 and
HCT116 p53–/– cells, measured by
flow cytometry and inspected by
light microscopy. (a) Cultures
were either mock-treated (with
DMSO, the solvent of resveratrol)
or treated with 100 ␮M resveratrol
for 32 hr. Cells were then fixed,
stained with PI and FACSscanned. The sub-G1 peak represents apoptotic cells. Note that
HCT116 cultures contained more
cells with a 2n-DNA content (G1
cells, arrows). (b) Overlay of
FACS profiles from cultures treated
with 100 ␮M resveratrol for 12 hr
(solid line) and 32 hr (dotted line),
showing the time dependence of
apoptosis induction. Again, note
the stronger decrease of the G1
peak with time in the HCT116
p53–/– cultures, suggesting a partial
G1 delay in the cultures harboring
p53. (c) Exposure of cells to
DMSO only (no) or to 10, 50 or
100 ␮M resveratrol for 48 hr. Error
bars denote SDs from 3 experiments. (d) Treatment of cells with
100 ␮M drug or DMSO only (control) and progress of apoptosis
during a 72 hr time course. (e)
Mock-treated (C), permanently
resveratrol-treated (100 ␮M for
120 hr) and temporary drug-treated
cultures (100 ␮M for 48 ⫹ 72 hr in
normal medium) were followed for
120 hr.
nously growing cultures were harvested by trypsinization, combined with the cells floating in medium, washed in PBS,
resuspended in 200 ␮l of 0.9% NaCl, squeezed through a 23.5gauge needle into 1.8 ml of methanol and fixed for 30 min at
–20°C. After repeated washes in PBS, cells were resuspended in
PBS supplemented with RNase A (25 ␮g/ml) at approximately 106
cells/ml and stained with PI (25 ␮g/ml) for several hours at 4°C.
DNA fluorescence was measured using a Becton Dickinson (Bedford, MA) FACScan; data acquisition and analysis were performed
with the Cell Quest software (Becton Dickinson). Villin expression
was examined by fixing cultures with 4% paraformaldehyde, followed by staining with the antivillin MAb and a secondary FITClabeled antimouse antibody according to standard procedures.
Immunoblotting
Cells were seeded in 10 cm dishes to approximately 50% confluence 24 hr before drug treatment. At the appropriate time points, cells
were lysed in 150 ␮l of lysis buffer heated to 85°C and containing 50
mM Tris-HCl (pH 6.8), 100 mM DTT, 2% SDS and 20% glycerol.
Samples containing 15 ␮g of total cellular protein were subjected to
12% SDS-PAGE and transferred to a nitrocellulose membrane (Immobilon-P; Millipore, Bedford, MA). For signal detection, membranes were incubated overnight with p53 MAb DO-1 (1:1,000),
p21 antibody (1:1,000), ␤-actin antibody (1:5,000) and villin antibody (1:1,000). Finally, they were incubated with a peroxidaseconjugated antimouse antibody and the complexes visualized with
Renaissance Enhanced Luminol Reagents (NEN, Boston, MA).
DIFFERENTIATION AND APOPTOSIS BY RESVERATROL
Analysis of mitochondrial changes
Relative mitochondrial mass was determined by flow cytometry
using either JC-1 at 10 ␮g/ml, MitoTracker green at 50 nM (both
analyzed for green fluorescence) or rhodamine-123 at 0.5 ␮g/ml.
JC-1 allows simultaneous analysis of mitochondrial mass (green
fluorescence) and mitochondrial membrane potential ⌬⌿m (red/
green fluorescence). For mitochondrial mass and ⌬⌿m analysis,
2 ⫻ 105 cells were trypsinized, resuspended in McCoy’s 5A
medium supplemented with 0.2% FCS and incubated at 37°C for
30 min. Observation wavelengths were 530 nm for green fluorescence and 585 nm for red fluorescence. Superoxide production was
monitored with the ROS-sensitive fluorescent probe DCFH-DA.
Cells were trypsinized, resuspended as for JC-1 analysis and
incubated at 37°C for 30 min with 10 ␮M DCFH-DA. Again,
fluorescence was measured by flow cytometry.
Electron microscopy
For scanning electron microscopy, cells were cultured on graded
glass slides, fixed with 2% glutaraldehyde in 0.12 M cacodylate
buffer and finally treated with 2% OsO4, 1% tannic acid and 1%
uranyl acetate, following standard procedures. After critical point
drying and coating with gold, samples were analyzed with a
Camscan (Cambridge, UK) S2 microscope.
RESULTS
Induction of p53-independent apoptosis
The effect of resveratrol, an antioxidant and antimutagen from
grapes and wine previously reported to inhibit the development of
preneoplastic lesions and tumorigenesis,2 was studied on an isogenic set of human colon adenocarcinoma cell lines with microsatellite instability: HCT116 cells expressing wild-type p53 and
HCT116 p53–/– cells derived from the parental line by targeted
disruption of both p53 alleles.16 HCT116 is a poorly differentiated
and growth factor-insensitive cell type that is, however, responsive
to various cellular stresses, including DNA damage and spindle
disruption.16,17 HCT116 cells and the p53-negative derivatives
were grown in culture with an approximately equal doubling time
(30 hr). Incubation with resveratrol at concentrations between 10
and 100 ␮M resulted in the appearance, under the electron microscope and the light microscope after DAPI staining, of condensed
nuclei and apoptotic bodies as early as 24 hr after treatment,
regardless of p53 status (data not shown).
For quantitation of apoptosis, cells cultured for 24 hr were either
treated with increasing concentrations of resveratrol and analyzed
by flow cytometry 48 hr after treatment or exposed to 100 ␮M
resveratrol and FACS-scanned at different times. Figure 1 shows
that in both HCT116 and HCT16 p53–/– cultures resveratrol induces, in a concentration- and exposure time-dependent fashion,
the appearance of a cell population with a sub-G1 DNA content,
indicative of p53-independent apoptosis. In contrast, Trolox,
which is a cell-permeable vitamin E derivative and, like resveratrol, a powerful antioxidant, induced no apoptosis in either cell line
up to a concentration of 1 mM. Remarkably, low concentrations of
resveratrol (10 ␮M) were more toxic to HCT116 p53–/– cells than
to the parental cells (Fig. 1c), suggesting that p53 has a limited
protective effect against low resveratrol quantities, possibly
through the induction of cell-cycle arrest (see below and Discussion). Exposure to 100 ␮M resveratrol for 120 hr resulted in the
complete eradication of HCT116 and HCT116 p53–/– cell monolayers, with no viable cells left, whereas treatment for 48 hr
followed by another 72 hr in the absence of drug left cells that
recovered and proliferated (Fig. 1e).
Treatment with the cell-permeable caspase-3 inhibitor DEVDCHO at concentrations between 2.5 and 20 ␮M inhibited apoptosis
up to 50% (Fig. 2), demonstrating the observed cell death to be
apoptotic. In contrast, incubation of cultures with the caspase-8
inhibitor IETH-CHO had no significant effect on apoptosis induction by resveratrol (Fig. 2), indicating that the cells do not die
through the major death receptor–mediated pathway involving
617
FIGURE 2 – Effect of caspase inhibitors on resveratrol-induced apoptosis. Cultures treated with 100 ␮M resveratrol were, in addition,
exposed to different concentrations of the caspase-3 inhibitor DEVDCHO or the caspase-8 inhibitor IETH-CHO and analyzed by flow
cytometry. Twenty micrometers of the caspase-3 inhibitor reduced
apoptosis by approximately 50%.
caspase-8 activation. Analysis of cell-cycle distribution revealed
that only p53-deficient cultures lose almost completely the 2nDNA peak in response to resveratrol and instead produce a strong
4n-DNA peak indicative of a delay or arrest in late S and G2
phases of the cell cycle. In contrast, the parental HCT116 cultures
always contained more cells in G1 than the p53-negative derivatives (32% vs. 19%, arrows in Fig. 1a), irrespective of the approximately equal apoptosis levels. This suggests that a fraction of the
HCT116 cell population responds with p53-mediated G1 arrest.
Accumulation and activation of p53
A large body of work has documented that p53 protein is
posttranslationally stabilized and activated as a transcription factor
in response to various forms of cellular stress.12 The chemotherapeutic drugs doxorubicin (Adriamycin) and 5-FU stabilize and
activate p53 in HCT116 cells.10 To study whether resveratrol
affects p53 stability and expression of p53 target genes, HCT116
cultures were exposed to 100 ␮M resveratrol or, as a control, to
375 ␮M 5-FU for 3, 6 or 12 hr. Immunoblotting of total cell
extracts revealed that resveratrol induced p53 accumulation only
weakly, whereas 5-FU induced high steady-state levels (Fig. 3a,c).
Furthermore, expression of the p53 target p21 was only weakly
stimulated by resveratrol, reflecting not only the lack of strong
early stabilization of p53 but the weak activity as a transcription
factor of the p53 protein. In comparison, 5-FU treatment induced
high levels of p21. At later times after resveratrol treatment (24
and 48 hr), p53 levels rose more significantly compared to mocktreated cells, while p21 remained essentially at background levels
(Fig. 3b,c). However, in the p53-negative cultures, p21 expression
was inhibited by resveratrol, suggesting that the background levels
of p21 observed in the parental cells require functional p53 and
that the S/G2 phase arrest in HCT116 p53–/– cultures (Fig. 2a) is
transient since sustained G2 arrest requires p21.16
Levels of the apoptosis agonist Bax, which is regulated through
p53 itself as well as through p53-independent mechanisms, rose
under resveratrol in both HCT116 and HCT116 p53–/– cells but
with different kinetics (Fig. 3b). In HCT116 cells, initially almost
undetectable levels of Bax increased significantly during the first
24 hr and accumulated further during the 48 hr of treatment. In the
p53-negative cells, the basal level of Bax appeared to be slightly
elevated and Bax expression reached a maximum at around 24 hr,
only to fall off again thereafter. The importance of Bax for resveratrol-induced apoptosis in unclear at present. Neither Fas receptor nor FasL expression was affected by resveratrol (data not
shown), which together with the ineffectiveness of the caspase-8
inhibitor indicates that resveratrol cytotoxicity does not require Fas
signaling. Combined, the data show that resveratrol, in contrast to
5-FU, induces p53 accumulation and activation late and only
weakly in HCT116 cells. Levels of the proapoptotic protein Bax
618
MAHYAR-ROEMER ET AL.
rise through mechanisms that are in part independent of p53. This
in combination with the similar apoptosis profiles produced by
HCT116 and HCT116 p53–/– cells during the 72 hr time course
(Fig. 1b,d) underscores that apoptosis under resveratrol is p53independent. Nevertheless, the (limited) stimulation of p53 and
p21 may be responsible for the maintenance of a G1 peak in
HCT116 cultures (Fig. 1a, arrows) and for the relatively lower
sensitivity of these cells to apoptosis provoked by low resveratrol
quantities (Fig. 1c).
Mitochondrial changes in response to resveratrol
Apoptosis is frequently accompanied by complex mitochondrial
changes. To monitor the mitochondrial mass, mitochondria were
stained with the J aggregate-forming lipophilic cation JC-1, which
as a monomer emits green fluorescence and in a reaction driven by
the mitochondrial transmembrane potential ⌬⌿m turns into a red
fluorescence-emitting dimer, thereby allowing the simultaneous
analysis of the total mitochondrial mass per cell (green fluorescence) and of ⌬⌿m (the quotient of red/green fluorescence).
Treatment of cells with 100 ␮M resveratrol for different times led
to an increase of the mitochondrial mass relative to mock-treated
controls by approximately a factor of 2 within 24 to 36 hr,
regardless of p53 status. At later time points, the relative mitochondrial mass decreased slightly in HCT116 cells (Fig. 4a– c).
SDs were derived from 3 separate experiments. Electron microscopy suggested that the mitochondrial mass increase was mostly
the result of mitochondrial proliferation and not primarily of a
volume increase as larger numbers as well as dividing mitochondria were observed (not shown). Even the basal mitochondrial
mass (in the absence of treatment) varied between the 2 sister cell
lines: HCT116 cells lacking p53 contained a higher basal number
of mitochondria than the parental cells (Fig. 4d). This correlates
with the higher basal degree of differentiation of HCT116 p53–/–
cells (see below). The mitochondrial mass changes measured by
JC-1 were confirmed by fluorescence staining with another mitochondrion-selective dye, MitoTracker green (Fig. 4c).
When cells were analyzed for ⌬⌿m integrity upon exposure to
100 ␮M resveratrol, the incremental increase in JC-1 green fluorescence was not accompanied by a corresponding increase in JC-1
red fluorescence in both HCT116 and HCT116 p53–/– cells. This
indicated a transient decrease of ⌬⌿m, detectable as early as 3 to
6 hr and peaking 24 hr after the beginning of treatment (Fig. 4e).
Such membrane potential collapse is the result of a permeability
increase of the inner mitochondrial membrane and is frequently
associated with the initial phases of apoptosis. To confirm the
observed ⌬⌿m changes, cells were stained with MitoTracker
green, which stains mitochondria regardless of energy state, or
rhodamine-123 at a concentration of 0.5 ␮g/ml for 30 min at 37°C,
which stains cells according to ⌬⌿m.18 Again, we observed a
decrease in membrane potential, i.e., an increase in mitochondrial
mass without a concomitant increase in rhodamine fluorescence,
during the first 24 hr after resveratrol treatment, though the SDs
were higher than in the JC-1 study (Fig. 4f). The mitochondrial
mass increase and ⌬⌿m collapse were not affected by treatment of
cultures with caspase inhibitor (Fig. 5a and data not shown),
indicating that caspase activation was downstream to these mitochondrial changes.
FIGURE 3 – p53, p21 and Bax expression in response to resveratrol
or 5-FU, determined by immunoblotting. (a) HCT116 cultures were
treated with 100 ␮M resveratrol or 375 ␮M 5-FU for 3, 6 or 12 hr; and
15 ␮g protein extracts were subjected to Western blot analysis on a
12% SDS gel. p53 and p21 antibodies were used at 1:1,000 dilution;
␤-actin antibody was diluted 1:5,000. (b) Cultures were either mocktreated (C) or treated with 100 ␮M resveratrol for 24 or 48 hr.
Antibodies were used as above; Bax antibody was diluted 1:500. (c)
Intensities of scanned p53, p21 and Bax signals relative to the ␤-actin
signal (set as 1), indicating an increase in p53, p21 and Bax expression
in HCT116 cells under resveratrol. Bax: ⫹, with resveratrol; –, without
resveratrol.
DIFFERENTIATION AND APOPTOSIS BY RESVERATROL
619
FIGURE 4 – Mitochondrial mass and membrane potential changes measured by flow cytometry. (a) Relative green JC-1 fluorescence intensity
indicating mitochondrial mass of mock-treated (solid line) and resveratrol-treated (100 ␮M) cells 24 hr after drug exposure. The diagram was
adjusted for differences in basal JC-1 fluorescence between HCT116 and HCT116 p53–/– cells. (b) Quantification of mitochondrial mass increase
under 100 ␮M resveratrol for 48 hr. SDs were derived from 3 experiments. (c) Mitochondrial mass determined with the mitochondrion-specific
fluorescent dye MitoTracker green. (d) Fluorescence profile showing basal differences in mitochondrial mass between the (untreated) cell lines.
(e) Quotient of JC-1 red/green fluorescence documents the temporary breakdown of the mitochondrial membrane potential (⌬⌿m) at different
times after exposure to 100 ␮M resveratrol. (f) Quotient of fluorescence obtained with MitoTracker green (which stains mitochondria
independently of energy state) and with rhodamine (which reflects ⌬⌿m), again showing a decline of ⌬⌿m. SDs were derived from 3
experiments.
The mitochondrial mass increase and ⌬⌿m decrease have been
associated with uncoupling of the respiratory chain and an elevated
production of ROS. To test whether there was any excess ROS
production in the presence of 100 ␮M resveratrol, we stained live
cells with the hydrogen peroxide-sensitive fluorescent dye DCFHDA. The initial peroxide load was up to 25% lower in the presence
of the antioxidant resveratrol compared with mock-treated control
cells, but levels of the oxidized fluorescence product DCF sharply
increased under resveratrol to approximately 200% of control
levels between 12 and 24 hr and plateaued between 24 and 36 hr
independently of p53 (Fig. 5b). Like the mitochondrial changes,
ROS production was not affected by caspase inhibitor (data not
shown). Thus, ROS production in response to resveratrol parallels
the mitochondrial mass increase and ⌬⌿m collapse and coincides
with the onset of apoptosis.
Epithelial differentiation at early times during resveratrol
treatment
HCT116 cells and the p53-negative derivatives normally grow
as flat, triangular or slightly rounded adherent cells in tissue
culture. Scanning and transmission electron microscopy revealed
that in HCT116 cells the ultrastructural features of differentiation
are only weakly developed, in accord with previous findings indicating the poor differentiation of this cell type.17 However, by 24
hr after resveratrol treatment and regardless of p53 status, cells had
begun to assume a morphology resembling that of columnar absorptive cells, with an increase in size, a more polygonal shape, the
development of a dense apical brush border and the arrangement of
cells in tight contact with the appearance of desmosomes (Fig. 6a).
At these early time points after drug exposure and increasingly at
the later time points, cultures occasionally contained cells that
showed the condensed, rounded morphology characteristic of apoptosis (not shown). Electron microscopic examination of proliferating cultures treated with resveratrol for various times was
performed independently at least 3 times and once with freshly
thawed cultures. Interestingly, untreated HCT116 p53–/– cells already carried a denser brush border and contained more mitochondria than the untreated parental cells (Fig. 4d and data not shown),
suggesting that the p53-negative cells are somewhat more differ-
620
MAHYAR-ROEMER ET AL.
entiated than the parental cells, despite the similar growth characteristics and phase-contrast appearances. Nevertheless, resveratrol
increased these features of epithelial differentiation in both cell
types.
Alkaline phosphatase was expressed at only very low levels in
both untreated cell types, in accord with previous findings,19 and
these levels were not detectably elevated upon resveratrol treatment. The 95 kDa protein villin is usually associated with microvillar actin bundles and, together with brush border myosin I, is
a tissue-specific factor with a role in intestinal microvilli morphogenesis; it constitutes a differentiation marker during early embryogenesis.19 When antihuman villin MAb was employed on extracts
from control cells or cells treated with resveratrol, a signal indicative of villin overproduction became visible at early times after
drug treatment (with a maximum at 24 hr) and disappeared by 48
hr in both HCT116 and HCT116 p53–/– cell extracts (Fig. 6b).
Similarly, when mock-treated cultures of cells treated with resveratrol for 24 hr were fixed and incubated with the villin MAb
followed by a FITC-labeled antimouse secondary antibody and
analyzed by cytometry, a fluorescence shift indicative of villin
expression was observed (Fig. 6c). The curve suggests that although most cells in the culture express villin, a subpopulation
expresses the protein more strongly. The reasons for these differences are unclear at present. In summary, HCT116 cells show
signs of terminal epithelial differentiation under resveratrol that
precede the execution of apoptosis.
DISCUSSION
FIGURE 5 – Effect of caspase-3 inhibitor and ROS production. (a)
Example of a fluorescence profile showing mitochondrial mass increase under 100 ␮M resveratrol and measured by JC-1 green fluorescence. Mitochondrial masses do not measurably change when the
caspase-3 inhibitor is included (100 ␮M). (b) Increase with time of
DCF fluorescence upon oxidation of DCFH-DA by peroxides in response
to 100 ␮M resveratrol. SDs were calculated from 3 experiments.
Resveratrol has recently been reported to cause apoptosis in
normal mouse embryo fibroblasts but not in p53–/– mouse embryo
fibroblasts.15 We report here that, in contrast to the results obtained
with mouse embryo fibroblasts, resveratrol induces apoptosis in
human microsatellite-unstable colon carcinoma cell line HCT116
regardless of p53 status and at concentrations comparable to those
found in wine and grapes. As has been shown for THP-1 human
monocytic leukemia cells,20 the effect was reversible after removal
of the drug from the culture; nevertheless, the continuous presence
of resveratrol for 120 hr resulted in complete eradication of the
culture. Remarkably, p53 appears to have a limited protective
effect against apoptosis at low resveratrol concentrations. Since
p53 is weakly but detectably stabilized and activated as a transcription factor under resveratrol and, as a consequence, the inhibitor of cyclin-dependent kinases p21WAF/Cip1 is also weakly
stimulated and since cultures of p53-containing cells harbor more
cells with a 2n-DNA content than p53-negative cultures (Fig. 1a),
this limited protective effect may be the result of p53 exerting its
checkpoint function. Indeed and in accord with these findings,
Bunz et al.10 have shown that loss of p53-controlled checkpoints in
HCT116 cells can make the cells more sensitive to apoptosis by
the cytostatic drug doxorubicin and radiation. However, the protective effect of p53 and p21 against resveratrol is not nearly as
strong as that against doxorubicin.10,21
Several studies have suggested that the chemopreventive activity of resveratrol is largely based on its proapoptotic effects. For
instance, azomethane-induced colon carcinogenesis in rats appears
to be counteracted by apoptosis of damaged cells.22 Furthermore,
mouse and human leukemic cells selectively undergo apoptosis
under resveratrol, while hematopoietic progenitor cells and differentiated cells are largely resistant.20,23 However, different cell
types and even the same cell lines have produced conflicting
results with regard to the effects of resveratrol. In the human HL60
promyelocytic leukemia cell line, the drug has been variably
reported to induce apoptosis,24 differentiation2 and differentiation
with S/G2 arrest without apoptosis.7 Moreover, LNCaP prostatecarcinoma cells downregulated androgen receptor in one study25
but not in another.26 Breast-carcinoma cells have been variably
demonstrated to be growth-stimulated by (estrogen-mimicking)
resveratrol,6 to be growth-inhibited by resveratrol regardless of the
expression of estrogen receptor27 or to suffer Fas-mediated apo-
DIFFERENTIATION AND APOPTOSIS BY RESVERATROL
621
ptosis.24 Still others have described that resveratrol causes apoptosis independently of Fas signaling.20 In agreement with the findings showing that resveratrol-induced apoptosis is not dependent
on Fas/FasL, we present here data indicating that HCT116 cells die
a “mitochondrion type” and not a “death receptor type”, of apoptosis.
Mitochondria initiating the apoptotic cascade may have a permeabilized outer membrane through the relocalization of monomeric proapoptotic Bax protein from the cytosol to the mitochondrial membrane followed by multimerization and pore formation.
This, then, results in the release of soluble apoptosis-triggering
intermembrane proteins, including cytochrome c, and finally in the
activation of caspases 9 and 3.13 We found that Bax is transiently
overproduced in response to resveratrol independently of p53 (Fig.
3b) and that caspase-3 inhibitor blocks apoptosis (Fig. 2). Cytochrome c release uncouples the respiratory chain and thereby
favors the transfer of electrons to oxygen to generate ROS;28 thus,
ROS production is a relatively late event in this apoptosis cascade.
In accord with this, we observed ROS mostly at late times after
resveratrol treatment (Fig. 5b), when many cells already showed
signs of apoptotic breakdown. Membrane potential (⌬⌿m) collapse following the transient or permanent opening of the membrane permeability transition pores in the inner mitochondrial
membrane is frequently associated with the mitochondrion type of
apoptosis.13,28 Our results show that the beginning of ⌬⌿m breakdown under resveratrol precedes ROS production and the appearance of apoptotic cells (Figs. 1, 4, 5), suggesting that this is an
early event in the apoptotic cascade triggered by the drug. Thus,
apoptosis by resveratrol appears to be of the classical mitochondrion type, involving membrane permeabilization, ROS generation
and ⌬⌿m dissipation. The very early mechanisms that trigger the
mitochondrial changes, however, might involve inhibition of I␬B
kinase activation and thus of NF␬B-controlled transcription by
resveratrol.29,30 NF␬B constitutes an important survival factor in
some cell types.31
Chemopreventive and -therapeutic drugs often stimulate the
accumulation and activation of p53,9,12 as does resveratrol,8,15
perhaps through induction of DNA damage in the presence of
copper ions.32 In contrast to these reports,8,15 p53 accumulation
was weak in the present study and occurred late following resveratrol treatment compared to the response to 5-FU. In the absence of p53, p21 levels fell off in response to resveratrol, suggesting that at least some of the limited quantities of p53 that come
up in the parental HCT116 cells are active and stimulate p21
expression. p21 in turn may give rise to the subpopulation of cells
in HCT116 cultures with a 2n-DNA content indicative of G1 block.
Abrogation of p21 expression under resveratrol was also observed
in normal colon mucosa of the rat22 and in LNCaP cells.25 The
proapoptotic bax gene is a further target of p53. In the current
study, however, accumulation of Bax under resveratrol was only
partially p53-dependent as the kinetics of Bax accumulation were
clearly different in the parental HCT116 and the p53–/– cells (Fig.
3b). Bax overproduction in response to resveratrol was also observed in a model of rat colon carcinogenesis.22 Whether Bax
FIGURE 6 – Signs of terminal epithelial differentiation upon treatment with 100 ␮M resveratrol. (a) Analysis by scanning electron
microscopy of HCT116 cells 24 hr after the beginning of drug exposure (low, 1,000⫻ magnification; high, 10,000⫻ magnification). Note
the increased cell size, tight cell contacts and high density of microvilli
in response to resveratrol. (b) Temporary overproduction of the differentiation marker and microvillar protein villin in response to resveratrol (24 and 48 hr) and in comparison to mock-treated cells (C).
Protein (15 ␮g) was analyzed on an 8% SDS gel with villin antibody
at 1:1,000 dilution and ␤-actin antibody at 1:5,000 dilution. (c) FACS
analysis of HCT116 cells after 24 hr of resveratrol treatment. Cultures
were either mock-treated (solid line) or treated with the drug (dotted
line), fixed, stained for villin and analyzed for green fluorescence. Note
the twin peaks produced by villin-positive cultures.
622
MAHYAR-ROEMER ET AL.
alone can account for the apoptotic response of HCT116 cells
remains to be seen.
Tumor cells frequently show phenotypes indicative of a block in
the normal differentiation pathway. Various treatments capable of
inducing differentiation have been described,33 and surprisingly,
tumor cells forced to differentiate often eventually succumb to
apoptosis. In perhaps the most successful differentiation– extinction strategy, retinoic acid was used to eradicate acute myelocytic
leukemia.34 Nevertheless, differentiation followed by death has
also been reported to occur in human colon-carcinoma cells. For
instance, the tyrosine kinase inhibitors herbimycin and genistein
induce, in analogy to resveratrol in HCT116 cells, G2 arrest, a
microvillar brush border and eventually apoptosis in the (mutant
p53-containing) Colo-205 cell line.18 Furthermore, this differenti-
ation was accompanied by a decline in ⌬⌿m, ROS production and,
unexpectedly, an increase in mitochondrial number. Mitochondrial
mass increase preceding cell death has also been observed, e.g., in
zidovudine (AZT)-caused myopathy in AIDS patients35 and upon
etoposide treatment of hematopoietic cells.36 Whether mitochondrial proliferation is a prerequisite for some forms of mitochondrion type apoptosis and how it is controlled remain to be investigated.
ACKNOWLEDGEMENTS
We are grateful to Dr. B. Vogelstein for supplying HCT116
cells and the p53-negative derivatives. KR was the recipient of a
grant from the German Research Foundation.
REFERENCES
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
Kopp P. Resveratrol, a phytoestrogen found in red wine. A possible
explanation for the conundrum of the “French paradox”? Eur J Endocrinol 1998;138:619 –20.
Jang M, Cai L, Udeani GO, et al. Cancer chemopreventive activity of
resveratrol, a natural product derived from grapes. Science 1997;275:
218 –24.
Soleas GJ, Diamandis EP, Goldberg DM. Resveratrol: a molecule
whose time has come? And gone? Clin Biochem 1997;30:91–113.
Fontecave M, Lepoivre M, Elleingand E, et al. Resveratrol, a remarkable
inhibitor of ribonucleotide reductase. FEBS Lett 1998;421:277–9.
Sun NJ, Woo SH, Cassady JM, et al. DNA polymerase and topoisomerase II inhibitors from Psoralea corylifolia. J Nat Prod 1998;61:
362– 6.
Gehm BD, McAndrews JM, Chien PY, et al. Resveratrol, a polyphenolic compound found in grapes and wine, is an agonist for the
estrogen receptor. Proc Natl Acad Sci USA 1997;94:14138 – 43.
Ragione FD, Cucciolla V, Boriello A, et al. Resveratrol arrests the cell
division cycle at S/G2 phase transition. Biochem Biophys Res Commun 1998;250:53– 8.
Hsieh T, Juan G, Darzynkiewicz Z, et al. Resveratrol increases nitric
oxide synthase, induces accumulation of p53 and p21 WAF1/CIP1
and suppresses cultured bovine pulmonary artery endothelial cell
proliferation by perturbing progression through S and G2. Cancer Res
1999;59:2596 – 601.
Lowe SW, Bodis S, McClatchey A, et al. p53 status and efficacy of
cancer therapy in vivo. Science 1994;266:807–10.
Bunz F, Hwang PM, Torrance C, et al. Disruption of p53 in human
cancer cells alters the response to therapeutic agents. J Clin Invest
1999;104:263–9.
Weinstein JN, Myers TG, O’Connor PM, et al. An informationintensive approach to the molecular pharmacology of cancer. Science
1997;275:343–9.
Jimenez GS, Khan SH, Stommel JM, et al. p53 regulation by posttranslational modification and nuclear retention in response to diverse
stresses. Oncogene 1999;18:7656 – 65.
Kroemer G, Reed JC. Mitochondrial control of cell death. Nat Med
2000;6:513–9.
Ghadourifar P, Schenk U, Klein SD, et al. Mitochondrial nitric-oxide
synthase stimulation causes cytochrome c release from isolated mitochondria— evidence for intramitochondrial peroxynitrite formation.
J Biol Chem 1999;274:31185– 8.
Huang C, Ma W, Goranson A, et al. Resveratrol suppresses cell
transformation and induces apoptosis through a p53-dependent pathway. Carcinogenesis 1999;20:237– 42.
Bunz F, Dutriaux A, Lengauer C, et al. Requirement for p53 and p21
to sustain G2 arrest after DNA damage. Science 1998;282:1497–501.
Wang H, Rajagopal S, Reynolds S, et al. Differentiation-promoting
effect of 1-O(2-methoxy)hexadecyl glycerol in human colon cancer
cells. J Cell Physiol 1999;178:173– 8.
Mancini M anderson BO, Caldwell E, et al. Mitochondrial proliferation and paradoxical membrane depolarization during terminal differentiation and apoptosis in a human colon carcinoma cell line. J Cell
Biol 1997;138:449 – 69.
Chantret I, Barbat A, Dussauix E, et al. Epithelial polarity, villin
expression and enterocytic differentiation of cultured human colon
carcinoma cells: a survey of twenty cell lines. Cancer Res 1988;48:
1936 – 42.
20. Tsan M-F, White JE, Maheshwari JG, et al. Resveratrol induces Fas
signalling-independent apoptosis in THP-1 human monocytic leukemia cells. Br J Haematol 2000;109:405–12.
21. Mahyar-Roemer M, Roemer K. p21Waf1/Cip1 can protect human colon
carcinoma cells against p53-dependent and p53-independent apoptosis induced by natural chemopreventive and therapeutic agents. Oncogene 2001;20:3387–98.
22. Tessitore L, Davit A, Sarotto I, et al. Resveratrol depresses the growth
of colorectal aberrant crypt foci by affecting bax and p21 (CIP)
expression. Carcinogenesis 2000;21:1619 –22.
23. Gautam SC, Xu YX, Dumaguin M, et al. Resveratrol selectively
inhibits leukemia cells: a prospective agent for ex vivo bone marrow
purging. Bone Marrow Transplant 2000;25:639 – 45.
24. Clement MV, Hirpara JL, Vhawdhury SH, et al. Chemopreventive
agent resveratrol, a natural product derived from grapes, triggers
CD95 signalling-dependent apoptosis in human tumor cells. Blood
1998;92:996 –1002.
25. Mitchell SH, Zhu W, Young CY. Resveratrol inhibits the expression
and function of the androgen receptor in LNCaP prostate cancer cells.
Cancer Res 1999;59:5892–5.
26. Hsieh TC, Wu JM. Grape-derived chemopreventive agent resveratrol
decreases prostate-specific antigen (PSA) expression in LNCaP cells
by an androgen receptor (AR)-independent mechanism. Anticancer
Res 2000;20:225– 8.
27. Mgbonyebi OP, Russo J, Russo IH. Antiproliferative effect of synthetic resveratrol on human breast epithelial cells. Int J Oncol 1998;
12:865–9.
28. Green DR, Reed JC. Mitochondria and apoptosis. Science 1998;281:
1309 –12.
29. Homes McNary M, Baldwin AS. Chemopreventive properties of
trans-resveratrol are associated with inhibition of activation of the IkB
kinase. Cancer Res 2000;60:3477– 83.
30. Manna SK, Mukhopadhyay A, Aggarwal BB. Resveratrol suppresses
TNF-induced activation of nuclear transcription factor NF␬B, activator protein I and apoptosis: potential role of reactive oxygen intermediates and lipid peroxidation. J Immunol 2000;164:6509 –19.
31. Liu Z, Hsu H, Goeddel DV, et al. Dissection of TNF receptor 1
effector functions: JNK activation is not linked to apoptosis while
NF-␬B activation prevents cell death. Cell 1996;87:565–76.
32. Ahmad A, Farhan-Asad S, Singh S, et al. DNA breakage by resveratrol and Cu (II): reaction mechanism and bacteriophage inactivation.
Cancer Lett 2000;154:29 –37.
33. Beere HM, Hickman JA. Differentiation: a suitable strategy for cancer
chemotherapy? Anticancer Drug Des 1993;8:299 –322.
34. Ohashi H, Ichikawa A, Takagi H, et al. Remission induction of acute
promyelocytic leukemia by all-trans-retinoic acid: molecular evidence
of restoration of normal hematopoiesis after differentiation and subsequent extinction of leukemic clone. Leukemia 1992;6:859 – 62.
35. Cupler EJ, Danon MJ, Jay C. Early features of zivudine-associated
myopathy: histopathological findings and clinical correlations. Acta
Neuropathol 1995;90:1– 6.
36. Reipert S, Berry J, Hughes MF, et al. Changes of mitochondrial mass
in the hemopoietic stem cell line FDCP-mix after treatment with
etoposide: a correlative study by multiparameter flow cytometry and
confocal and electron microscopy. Exp Cell Res 1995;221:281– 8.