SPARC functions as an anti-stress factor by inactivating p53 through Akt-mediated MDM2 phosphorylation to promote melanoma cell survival

Article (PDF Available)inOncogene 30(49):4887-900 · June 2011with23 Reads
DOI: 10.1038/onc.2011.198 · Source: PubMed
Abstract
Aberrant expression of Secreted Protein Acidic and Rich in Cysteine (SPARC)/osteonectin has been associated with an invasive tumor cell phenotype and poor outcome in human melanomas. Although it is known that SPARC controls melanoma tumorigenesis, the precise role of SPARC in melanoma cell survival is still unclear. Here, we show that SPARC has a cell-autonomous survival activity, which requires Akt-dependent regulation of p53. Suppression of SPARC by RNA interference in several human melanoma cells and xenografted A375 tumors triggers apoptotic cell death through the mitochondrial intrinsic pathway and activation of caspase-3. Cell death induced by depletion of SPARC is dependent on p53 and induction of Bax, and results in the generation of ROS. Stabilization of p53 in SPARC-depleted cells is associated with a decrease in Akt-mediated activating phosphorylation of MDM2. Inhibition of Akt signaling pathway is important for the observed changes as overexpression of constitutively active Akt protects cells against apoptosis induced by SPARC depletion. Conversely, increased expression of SPARC stimulates Akt and MDM2 phosphorylation, thus facilitating p53 degradation. Finally, we show that overexpression of SPARC renders cells more resistant to the p53-mediated cytotoxic effects of the DNA-damaging drug actinomycin-D. Our study indicates that SPARC functions through activation of Akt and MDM2 to limit p53 levels and that acquired expression of SPARC during melanoma development would confer survival advantages through suppression of p53-dependent apoptotic pathways.
ORIGINAL ARTICLE
SPARC functions as an anti-stress factor by inactivating p53 through
Akt-mediated MDM2 phosphorylation to promote melanoma cell survival
N Fenouille
1
,
2
, A Puissant
2
,
3
, M Tichet
1
,
2
, G Zimniak
1
,
2
, P Abbe
1
,
2
, A Mallavialle
2
,
4
, S Rocchi
1
,
2
,
5
,
J-P Ortonne
1
,
2
, M Deckert
2
,
4
,
6
, R Ballotti
1
,
2
,
5
and S Tartare-Deckert
1
,
2
,
5
1
INSERM, U895, Centre Me
´
diterrane
´
en de Me
´
decine Mole
´
culaire (C3M), Biology and Pathologies of Melanocytes, Nice,
France;
2
University of Nice—Sophia Antipolis, Faculte
´
de me
´
decine, Institut Signalisation et Pathologie (IFR50), Nice, France;
3
INSERM, U895, C3M, Cell Death Differentiation and Cancer, Nice, France;
4
INSERM, U576, Nice, France;
5
Centre Hospitalier
Universitaire (CHU) de Nice, Ho
ˆ
pital Archet, Department of Dermatology, Nice, France and
6
CHU de Nice, Ho
ˆ
pital Archet,
Department of Clinical Hematology, Nice, France
Aberrant expression of Secreted Protein Acidic and Rich
in Cysteine (SPARC)/osteonectin has been associated
with an invasive tumor cell phenotype and poor outcome
in human melanomas. Although it is known that SPARC
controls melanoma tumorigenesis, the precise role of
SPARC in melanoma cell survival is still unclear. Here,
we show that SPARC has a cell-autonomous survival
activity, which requires Akt-dependent regulation of p53.
Suppression of SPARC by RNA interference in several
human melanoma cells and xenografted A375 tumors
triggers apoptotic cell death through the mitochondrial
intrinsic pathway and activation of caspase-3. Cel l death
induced by depletion of SPARC is dependent on p53 and
induction of Bax, and results in the generation of ROS.
Stabilization of p53 in SPARC-depleted cells is associated
with a decrease in Akt-mediated activating phosphory-
lation of MDM2. Inhibition of Akt signaling pathway is
important for the observed changes as overexpression of
constitutively active Akt protects cells against apoptosis
induced by SPARC depletion. Conversely, increased
expression of SPARC stimulates Akt and MDM2
phosphorylation, thus facilitating p53 degradation. Final-
ly, we show that overexpression of SPARC renders cells
more resistant to the p53-mediated cytotoxic effects of the
DNA-damaging drug actinomycin-D. Our study indicates
that SPARC functions through activation of Akt and
MDM2 to limit p53 levels and that acquired expression of
SPARC during melanoma development would confer
survival advantages through suppression of p53-dependent
apoptotic pathways.
Oncogene (2011) 30, 4887–4900; doi:10.1038/onc.2011.198;
published o nline 20 June 2011
Keywords: melanoma; p53; signaling; tumor biology
Introduction
Melanoma is the leading cause of skin cancer-related
deaths and its incidence has increased worldwide faster
than any other cancer. Despite intensive research, the
overall 5-year survival for metastat ic melanoma patients
remains less than 10%. Hallmarks of melanoma are its
propensity to rapidly spread to the lymph system and
internal organs, and its unresponsiveness to radiation
and chemotherapy. Local forms of the disease can be
cured by surgical excision, but there are no effective
treatment options available for advanced melanomas.
Melanomas arise from malignant transformation of
epidermal melanocytes and develop as multistep process
through ac cumulation of genetic and epigenetic changes
in growth and survival pathways, as well as alterations
in cell–cell communication and extracellular matrix
interactions (Miller and Mihm, 2006). Matricellular pro-
teins are found in the extra- and pericellular matrix, an d
are potent modulators of cellular functions and tumor–
stroma interactions (Bornstein and Sage, 2002). Expression
of some members of the matricellular proteins family is
altered during the development of melanoma (Fukunaga-
Kalabis et al., 2008). For example, Secreted Protein Acidic
and Rich in Cysteine (SPARC) (also called osteonectin
and BM-40) is abundantly produced by melanoma cells
but is not expressed by normal melanocytes (Ledda
et al., 1997a; Robert et al., 2006). Its expression corre-
lates with the aggressiveness of melanomas and adverse
clinical outcome (Massi et al., 1999). SPARC was shown
to be multifunctional, with activities in differentia-
tion, apoptosis, cancer cell migration and regulation
of immune cell response (Arnold and Brekken, 2009).
Mechanistically, SPARC binds to several components
of the extracellular matrix and interacts with specific
adhesion receptors such as integrin-b1 ( Nie et al., 2008;
Weaver et al., 2008). The intracellular signaling path-
ways downstream from SPARC are beginning to be
identified. They include major mediators of integrin
signaling such as integrin-linked kinase (ILK) and
focal adhesion kinase (FAK) (Barker et al., 2005;
Shi et al., 2007; Weaver et al., 2008), as well as b-catenin
and phosphatidylinositol 3-kinase (PI3K)–Akt signaling
Received 21 November 2010; revised and accepted 19 April 2011;
published online 20 June 2011
Correspondence: Dr S Tartare-Deckert, INSERM, Unite
´
895, Centre
Me
´
diterrane
´
en de Me
´
decine Mole
´
culaire (C3M), Biology and Patholo-
gies of Melanocytes, Equipe 1, 151 Route de Saint-Antoine de Ginestie
`
re,
BP 23194, Nice Ce
´
dex 3 06204, France.
E-mail: tartare@unice.fr
Oncogene (2011) 30, 4887–4900
&
2011 Macmillan Publishers Limited
All rights reserved 0950-9232/11
www.nature.com/onc
pathways (Shi et al. , 2004; Nie and Sage, 2009; Chang
et al., 2010). In cancer, SPARC may function as either a
tumor suppressor or a pro-invasive factor depending on
the tumor type and local extracellular milieu.
Several lines of evidence have strengthened the notion
that production of SPARC by tumor cells is critical
for melanoma progression. We and others have shown
that SPARC promotes migratory and invasive abilities
of human melanoma cells (Robert et al., 2006; Smit
et al., 2007). Moreover, SPARC was shown to control
melanoma growth in xenograft assays and cell-cycle
progression in cultured cells (Ledda et al., 1997b; Horie
et al., 2010; Fenouille et al., 2011). We recently showed
that RNA interference-mediated depletion of SPARC in
melanoma cells activates p53 and induces p21
Cip1/Waf1
-
dependent cell-cycle arrest, raising the possibility that
SPARC might have a role in p53 tumor-suppressor
regulation in melanoma cells (Fenouille et al. , 2011).
The acquisition of survival pathways promoting
resistance to apoptosis is a common feature of tumor
cells and contributes to tumorigenesis, metastasis and
drug resistance (Johnstone et al., 2002). p53-mediated
apoptosis has a central role in suppression of tumor-
igenesis (Vogelstein et al., 2000). Under normal condi-
tions, intracellular levels of p53 are controlled by the E3
ubiquitin ligase MDM2, which binds to p53 and
promotes its ubiquitination and degradation by the
proteasome (Honda et al., 1997; Marine et al., 2006).
One of the major regulators of MDM2 activity is Akt,
which phosphorylates MDM2 at Ser166 and Ser186
(Mayo and Donner, 2001; Zhou et al. , 2001; Ogawara
et al., 2002). Akt is an important pro-survival molecule,
and phosphorylation of MDM2 may serve to protect
cells from p53-induced apoptotic cell death.
In response to various cellular stresses, p53 transcrip-
tionally induces the expression of specific target genes,
which critically regulate the mitochondrial intrinsic
apoptotic pathway and mediate tumor-suppressive
functions (Vogelstein et al., 2000). An important down-
stream mediator of this pathway is Bax (Miyashita and
Reed, 1995). The apoptotic stress causes Bax transacti-
vation and its translocat ion from cytosol to mitochon-
dria, leading to the efflux of cytochrome c from
mitochondria, the assembly of apoptosome and cas-
pase-9 activation, followed by activation of effector
caspases (Schuler and Green, 2001). Consequently, the
p53 apoptosis pathway is disrupted in most human
cancers by mutation in p53, or altered expression of
downstream effectors or upstream regulators of p53
(Vogelstein et al., 2000). Unlike other cancers, p53 is
rarely mutated in melanoma (Weiss et al., 1993; Albino
et al., 1994). Thus, melanoma cells typically harbor
functional p53 protein but acquire capacities to inacti-
vate components of the p53 pathway and to evade
stress-induced cell death (Soengas and Lowe, 2003).
In this study, we investigated whether SPARC can
provide mela noma cells with survival advantage and
suppress the ability of p53 to activate the cell death
machinery. By combining gene silencing (through RNA
interference strategies) and overexpression studies in human
melanoma cells, followed by functional rescue experiments,
we now show that SPARC can override p53-mediated
apoptosis by activating Akt and MDM2. We not only
define a contribution of SPARC in melanoma cell survival,
but also determine a novel mechanism by which SPARC
promotes resistance to p53-mediated cell death.
Results
SPARC deficiency reduces clonogenicity and promotes
time-dependent apoptosis
We began by addressing the effect of SPARC knock-
down in melanoma cell proliferation and cell-cycle
distribution using small interfering RNA (si RNA).
Treatment of A375, 1205Lu and 501mel cells with
SPARC siRNA (siSPARC) for 6 days resulted in a
dramatic decrease in cellu lar proliferation as measured
by XTT assay (Supplementary Figure S1A). Cell-cycle
profiles were examined with A375 cells, which were
transfected with control or SPARC siRNA, and as we
described (Fenouille et al., 2011), SPARC depletion
induced a significant increase in the cell population in
the G
2
/M phase after 4 days (Figure 1a). However, the
increasing number of sub-G
1
cells at day 5 and 6 indi-
cated an occurrence of subsequent cell death. To assess
the effects of SPARC knockdown on cell-cycle progression
and cell death beyond 6 days, we performed clonogenic
survival assays. We used A375 clones that stably express
a doxycycline-inducible short-hairpin RNA (shRNA)
construct targeting another sequence in the SPARC-
coding region (shSPARC #3F9 and #5H9 cells). SPARC
protein was knocked down approximately 90% after
doxycycline exposure of cells (Supplementary Figure S1B),
and this resulted in decreased ability of A375 cells to
form foci and soft agar colonies, showing that SPARC is
required for the clonogenic survival of melanoma cells
(Figure 1b and Supplementary Figure S1C).
To confirm and extend the findings on cell death, we
used Annexin-V binding and flow cytometry to analyze
apoptosis in A375 cells following downregulation of
SPARC expression by siRNA or shRNA (Figure 1c).
TRAIL treatment served as a positive control for
apoptosis. Fluorescence-activated cell sorting analys is
clearly showed an increase in the number of cells that
stained posit ive for the apoptotic marker Annexin-V,
5 days following SPARC knockdown. After 5 days of
depletion by siR NA, the Annexin-V-positive cells
increased to 15% to reach 25% at day 6, which showed
that SPARC-knockdown A375 cells activate apoptotic
death. Apoptosis was also evidenced in 1205Lu and
501mel cells following treatment with siSPARC for 6
days (Figure 1d). Immunoblot analysis confirmed PARP
cleavage, a marker of cells undergoing apoptosis,
following siSPARC treatment or induction of SPARC
shRNA in melanoma cells (Figures 1e and f).
SPARC deficiency activates the mitochondria-dependent
apoptotic pathway
To further characterize the apo ptotic response, we
measured activity of the effector caspase-3 by fluori-
Antiapoptotic function of SPARC in melanoma cells
N Fenouille et al
4888
Oncogene
metric assay. TRAIL was used as an inducer of caspase
activities. As shown in Figures 2a and b, transient or
inducible SPARC knockdown in A375 cells induced a
time-dependent increase in caspase-3 activity. Caspase-3
activation was also detected in 1205Lu and 501mel
melanoma cells 6 days after transfection with siSPARC
(Figure 2c). To distinguish the apoptotic signaling
pathways activated in SPARC-depleted cells, caspase-8
Figure 1 Depletion of SPARC induces a G
2
/M cell-cycle arrest that precedes apoptosis in melanoma cells. (a) A375 cells were
transfected with 50 n
M siCTRL for 6 days, or siSPARC for the indicated time periods. Cell-cycle profiles were then analyzed for DNA
content by flow cytometry of PI-stained cells. The histograms represent the percentage of cells in the different phases of the cell cycle.
The immunoblots show the SPARC levels in the protein lysates and supernatants of these cells (bottom panel). HSP60 and Ponceau-S-
stained bands were used as loading controls. (b) A375 TRex cells stably expressing a doxycycline-inducible SPARC shRNA
(shSPARC; clone #3F9) were grown in soft agar (left panel) or on tissue culture plates (right panel), in doxycycline-free medium (open
bars) or doxycycline-containing medium (filled bars). After 15 days, colonies were detected by crystal violet staining and quantified.
The results are expressed as the percent of the number of colonies in doxycycline-treated cells compared with that in untreated cells.
The columns indicate the mean of two independent experiments performed in duplicate; the errors bars indicate the s.d. Photographs of
representative plates are shown. Immunoblot showing SPARC levels in protein lysates 7 days after doxycycline treatment is shown in
panel f.(c) Flow cytometric analysis of A375 cells after depletion of SPARC by siRNA transfection or stable expression of
doxycycline-inducible shRNA. At different time points after transfection (left panel) or after doxycycline treatment (right panel), cells
were stained with PI and Annexin-V-fluos, and analyzed by flow cytometry. The histograms represent the percentage of Annexin-V-
positive/PI-negative cells (apoptotic population). Treatment with 100 ng/ml TRAIL for 10 h served as a positive control of apoptosis.
(d) 1205Lu and 501mel were transfected with 50 n
M siCTRL or siSPARC for 6 days before being stained with Annexin-V-fluos, and
analyzed by flow cytometry as above. (e) A375 cells were transfected for various time periods with 50 n
M siCTRL or siSPARC
(left panel). A375 shlacZ or shSPARC (clone #3F9) were grown for 7 days in doxycycline-free medium ( Dox) or Doxycycline-
containing medium ( þ Dox) (right panel). Protein lysates from the resulting cells were analyzed by immunoblotting using the indicated
antibodies. Detection of PARP cleavage (cl. PARP) was used as a marker for apoptosis and HSP60 as a loading control. (f) 1205Lu
and 501mel cells were transfected with 50 n
M siCTRL or siSPARC for 6 days and protein lysates were analyzed for PARP cleavage by
immunoblotting. PI, propidium iodide; siRNA, small interfering RNA; shRNA, short-hairpin RNA; SPARC, secreted protein acidic
and rich in cysteine.
Antiapoptotic function of SPARC in melanoma cells
N Fenouille et al
4889
Oncogene
and caspase-9 activities were determ ined by specific
enzymatic assays. Caspase-9 activation, the key initiator
of the mitochondrial intrinsic pathway occurred from
day 5 to day 6 following SPARC siRNA treatment
(Figure 2d). By contrast, caspase-8, the initiator caspase
typically involved in receptor-mediated apoptosis, was
not active under these conditions, whereas TRAIL
administration led to its effective activation. This
strongly suggests that SPARC depletion initiates the
mitochondrial apoptotic pathway.
We next examined whether apoptosis upon SPARC
depletion can be observed in a mouse xenograft model
of melanoma. We injected nude mice subcutaneously
with A375 shSPARC cells and treated the mice with
doxycycline to switch off SPARC in tumor tissues. As
we reported previously by Fenouille et al. (2011),
inhibiting SPARC expression in A375 shSPARC tumors
yielded slower-growing tumors compared with the
untreated group of mice (Figure 2e). At the end of the
study, analysis of caspase-3 activities of tumor tissues
showed that caspase-3 was significantl y activated in
SPARC-depleted tumors but not in SPARC-positive
tumors. Immunoblot analysis of xenograft protein
extracts confirmed effective depletion of SPARC in
tumors derived from doxycycline-treated mice. These
data indicate that the antitumor effects of SPARC
depletion can be caused by induction of apoptotic death.
p53 induces apoptosis and ROS production in SPARC-
depleted melanoma cells
A main known function of p53 is to elicit mitoc hondrial
apoptotic events (Schuler and Green, 2001). As we
showed recently that p53 is upregulated and activated in
melanoma cells following siRNA-mediated SPARC
depletion (Fenouille et al., 2011), we asked whether
p53 is involved in the above-described apoptotic res-
ponse. Immunoblot analysis showed a rapid and
Figure 2 Depletion of SPARC promotes apoptosis in melanoma cells and in tumors through the intrinsic mitochondrial pathway.
(a) A375 cells were transfected with 50 n
M siCTRL (open bars) or siSPARC (filled bars) for various time periods as indicated. The cells
were then harvested and the activity of the effector caspase-3 was determined by fluorometric assay. Treatment with 100 ng/ml TRAIL
for 10 h served as a positive control. The results are expressed as arbitrary units (AU). The columns indicate the means of four
independent experiments performed in quadruplicate; the error bars indicate the s.d. (b) A375 shSPARC cells were grown in
doxycycline-free medium (Dox) or doxycycline-containing medium ( þ Dox). At different time points after doxycycline treatment,
caspase-3 activity was evaluated as described above. (c) Detection of caspase-3 activity in 1205Lu and 501mel cells transfected with
50 n
M siCTRL or siSPARC for 6 days. (d) A375 cells were transfected with 50 nM siCTRL (open bars) or siSPARC (filled bars) for
various time periods as indicated. The cells were then harvested and the activities of initiator caspases-8 and -9 were determined by
fluorometric assay. Treatment with 100 ng/ml TRAIL for 10 h served as a positive control. The results are expressed as arbitrary units
(AU). The columns indicate the means of four independent experiments performed in quadruplicate; the error bars indicate the s.d.
(e) The left panel shows caspase-3 activity levels in A375 shSPARC tumor samples depleted (filled circles) or not (open circles) for
SPARC by 15-day doxycycline treatment in mice. The horizontal line indicates the mean of six independent measures. *Po0.005
(Student’s test). The right panel shows the growth curves of shSPARC cells in nude mice that received vehicle or 1 mg/ml doxycycline
in their drinking water (n ¼ 6). The points indicate the mean of six independent measurements; the error bars indicate the s.d.
*Po0.005; **Po0.01 (Student’s test). The immunoblot shows efficient inducible SPARC knockdown in A375 shSPARC tumors.
SPARC, secreted protein acidic and rich in cysteine.
Antiapoptotic function of SPARC in melanoma cells
N Fenouille et al
4890
Oncogene
significant time-dependent increase of p53 protein levels
in A375 cells exposed to siSPARC (Figure 3a). p53
levels were also elevated in A375 shSPARC tumors from
mice that received doxycycline treatment (Figure 3b).
We next investigated whether knockdown of p53 by
siRNA can rescue SPARC-depleted cells from apopto-
sis. Simultaneous knockdown of SPARC and p53 did
not cause caspase-3 activation and PARP degradat ion,
Figure 3 Depletion of SPARC induces apoptosis and ROS generation in a p53-dependent manner. (a) A375 cells were transfected
with 50 n
M siCTRL for 6 days, or siSPARC for various time periods. Proteins were then extracted and immunoblotting was performed
using the indicated antibodies. (b) Visualization of p53 expression levels in A375 shSPARC tumor samples depleted or not for SPARC
by doxycycline treatment in mice. (c) A375 cells were transfected with control siRNA (siCTRL), p53 siRNA (sip53), SPARC siRNA
(siSPARC) alone or in combination at final concentration of 50 n
M. The cells were harvested on day 6 after siRNA transfection.
Caspase-3 activity was determined by fluorometric assay and the expression levels of SPARC and p53 as well as detection of cleaved
PARP were analyzed by immunoblotting. (d) Quantification of apoptosis in A375 cells (TP53 wt) transfected as described above. The
percentage of apoptotic cells was determined by staining the cells with Annexin-V-fluos and flow cytometry. (e) SKmel-28 melanoma
cells (TP53 mut) were transfected with 50 n
M siCTRL or siSPARC for 6 days and subjected to flow cytometric analysis of Annexin-V-
fluos staining. TRAIL treatment was used as a positive control for apoptotic cell death. See Supplementary Figure S2 for other data
obtained in SKmel 28 cells. (f) Measurement of ROS levels using the redox-sensitive fluorescent dye DCF-DA. A375 cells were
transfected with 50 n
M siCTRL or siSPARC for the indicated time periods (top panel), or with siCTRL, sip53, siSPARC alone or in
combination at 50 n
M for 6 days (bottom panel). After incubation with 10 mM DCF-DA, the cells were analyzed by flow cytometry.
Similar results were obtained in TP53 wt 1205Lu and 501mel cells (see Supplementary Figure S3). ROS, reactive oxygen species;
siRNA, small interfering RNA; SPARC, secreted protein acidic and rich in cysteine.
Antiapoptotic function of SPARC in melanoma cells
N Fenouille et al
4891
Oncogene
compared with SPARC knockdown alone (Figure 3c).
Accordingly, the concurrent depletion of p53 pre-
vented the increase of Annexin-V-positive cells observed
in SPARC-depleted cells (Figure 3d), indicating that
apoptosis induced by siSPARC is p53-dependent.
Further evidence for the importance of p53 came from
experiments performed in SKmel28 melanoma cells
containing mutant p53. Annexin-V labeling, PARP-
cleavage and XTT assays showed that these cells were
resistant to apoptosis induced by SPARC knockdown
(Figure 3e and Supplementary Figure S2).
It is known that generation of reactive oxygen species
(ROS) represents a primary event in the regulation of
p53-induced apoptosis (Johnson et al., 1996). Conse-
quently, ROS levels were monitored by cytometric
analysis in melanoma cell s exposed to siSPARC using
the fluorescent probe DCF-DA. The fluorescent inten-
sity increased gradually in a time-dependent manner
upon SPARC depletion in A375, 1205Lu and 501mel
cells (Figure 3f and Supplementary Figures S3A and B),
suggesting that generation of ROS is involved in the
apoptotic process. This increase in ROS was also
dependent on p53, as depletion of p53 by siRNA
inhibited SPARC siRNA-induced ROS production
observed 4 days after transfection (Figure 3f). Finally,
we examined the functional contribution of ROS to
apoptosis induced by SPARC knockdown and found
that treatment with the radical scavenger N-acetyl-
L-
cysteine prevented ROS generation and attenuated
siSPARC-mediated caspase-3 activation (Supplem en-
tary Figure S3C), suggesting that ROS generation is
partially involved in apoptosis. Also, our observation
that the pan-caspase inhibitor z-VAD-fmk was unable
to prevent ROS production indicates that ROS produc-
tion occurs upstream from caspase activation in p53-
dependent apoptotic cell death induced by SPARC
depletion (Supplementary Figure S3D).
p53 drives Bax upregulation to mediate apoptosis
in SPARC-depleted melanoma cells
To identify mediators of p53-driven ROS production
and apoptosis in SPARC-depleted cells, we analyzed by
real-time quantitative–polymerase chain reaction (Q–PCR)
the temporal response of SPARC siRNA treatment on
the expression level of several transcripts involved in
p53-dependent cell-cycle arrest, apoptosis and oxidative
stress regulation (Supplementary Figure S4A). As we
reported previously, genes implicated in G
2
/M cell-cycle
arrest such as CDKN1A (p21
Cip1/Waf1
) were found to be
upregulated by SPARC depletion, whereas others such
as CCNB1, CDC2 and CDC25C were repressed. Relevant
genes also modulated by SPARC depletion included
members of the Bcl-2 proapoptotic family (Bax, PUMA
and Bim), the anti-apoptot ic Survivin gene (BIRC5) as
well as TIGAR and DRAM, two regulators of glycolysis
and autophagy, respectively. In addition, SPARC depletion
induced the expression of genes with antioxidant proper-
ties, such as those encoding Sestrin1, MnSOD and Foxo1.
Bax, PUMA and Survivin had a similar expression pattern
in both SPARC-depleted A375 and 1205Lu cells (Supple-
mentary Figure S4B). The expression profile that we
observed in the SPARC-depleted cells is consistent with
induction of p53-dependent programmed death.
At the protein level, we confirmed upregulation of
Bax and decrease of Survivin upon siRNA-mediated
SPARC depletion in A375, 501mel and 1205Lu cells,
whereas PUMA did not change significantly (Figure 4a
and Supplementary Figure S4C). In addition, levels of
Bcl-2, a critical factor for melanoma viability, were not
consistently affected by siSP ARC in the three melanoma
lines (Supplementary Figure S4C). Interestingly, Bax
induction was preceded by progressive accumulation of
p53 and induction of p21
Cip1/Waf1
in SPARC-depleted
A375 cells (Figure 4a). Notably, levels of Bax were
elevated in A375 tumors after inducible knockdown of
SPARC (Figure 4b). Finally, we showed that SPARC
siRNA-mediated upregulation of Bax was inhibited in
A375 cells treated with siRNA to p53 and was not
evidenced in p53-mutated SKmel28 cells after SPARC
depletion (Figure 4c). Figure 4d shows that increased
level of Bax in response to SPARC depletion was
accompanied by changes of its subcellular localization.
Five days after siSPARC transfection, Bax was found to
be redistributed from the cytosol fraction (F1) to the
mitochondria-containing fraction (F2) concomitantly
with the release of Smac/DIAB LO into the cytosol.
We conducted another rescue experiment to test the
contribution of Bax in apoptosis induced by SPARC
depletion. We inhibited siSPARC-induced Bax expres-
sion by co-transfecting A375 and 501mel cells with an
siRNA targeting Bax (Figure 4e). In A375 cells,
suppression of Bax significantly blocked SPARC siR-
NA-mediated apoptosis as assessed by caspase-3 activa-
tion, PARP degradation, Annexin-V staining and ROS
production (Figures 4e–g). Significant protection against
cell death was also observed with Bax siRNA in
SPARC-depleted 501mel cells. Thus, Bax appears to
be important for mediating the death signal induced by
SPARC knockdown.
p53 is preferentially stabilized in SPARC-knockdown
melanoma cells
We undertook to determine how SPARC knockdown
induces p53 accumulation and activati on. We first
examined whether SPARC deficiency upregulates p53
by a transcriptional or a posttranslational mechanism.
Knockdown of SPARC in A375 cells caused a modest
but non-significant increase in the levels of p53
transcripts as compared with that in control cells as
determined by real-time Q–PCR (Figure 5a), whereas in
1205Lu cells, p53 mRNA levels decreased following
SPARC depletion (Supplementary Figure S4B). How-
ever, in the presence of the translational inhibitor
cycloheximide, stabilization of p53 was observed with
SPARC depletion by siRNA (Figure 5b). When control
cells were treated with cycloheximide, p53 shows a half-
life of about 20 min, whereas in SPARC-depleted cells,
the pre-existing p53 protein was increased, with a half-
life of about 60 min. These findings support the
hypothesis that p53 upregulation in SPARC-depleted
Antiapoptotic function of SPARC in melanoma cells
N Fenouille et al
4892
Oncogene
Figure 4 The proapoptotic BH3 family member Bax mediates apoptosis in SPARC-depleted A375 cells. (a) A375 cells were
transfected with 50 n
M siCTRL for 6 days, or siSPARC for various time periods. 1205Lu and 501mel cells were transfected with 50 nM
siCTRL or siSPARC for 6 days. Proteins were then extracted and immunoblotting was performed using the indicated antibodies.
HSP60 was used as a loading control. (b) Bax expression levels were analyzed by immunoblotting in A375 shSPARC tumor samples
depleted or not for SPARC by doxycycline treatment in mice. (c) A375 cells were transfected for 6 days with control siRNA (siCTRL),
p53 siRNA (sip53), SPARC siRNA (siSPARC) alone or in combination at a final concentration of 50 n
M. SKmel-28 melanoma cells
(TP53 mut) were transfected with 50 n
M siCTRL or siSPARC for 6 days. Protein extracts from resulting cells were analyzed by
immunoblotting using the indicated antibodies. (d) A375 cells were transfected with 50 n
M siCTRL or siSPARC for various time
periods. Translocation of Bax from cytoplasm to mitochondria was then analyzed using subcellular fractionation and immunoblotting.
Cytoplasmic (F1) and mitochondrial (F2) fractions were purified as described under the Materials and methods. The release of Smac/
DIABLO from the mitochondria into the cytoplasm was shown as a control. g-Tubulin and LAMP-2 were used as purity controls for
the cytoplasmic and mitochondrial fraction, respectively. (e) A375 and 501mel cells were transfected with control siRNA (siCTRL),
Bax siRNA (siBax), SPARC siRNA (siSPARC) alone or in combination at a final concentration of 50 n
M. The cells were harvested on
day 6 after siRNA transfection. Caspase-3 activity was determined by fluorometric assay and cell proliferation by an XTT assay. The
expression levels of SPARC and Bax as well as detection of cleaved PARP were analyzed by immunoblotting. (f) Quantification of
apoptosis in A375 cells transfected as described above. The percentage of apoptotic cells was determined by staining the cells with
Annexin-V-fluos and flow cytometry. (g) A375 cells were transfected for 6 days with siCTRL, siBax, siSPARC alone or in combination,
before measurement of ROS production by flow cytometry using the redox-sensitive dye DCF-DA. ROS, reactive oxygen species;
siRNA, small interfering RNA; SPARC, secreted protein acidic and rich in cysteine.
Antiapoptotic function of SPARC in melanoma cells
N Fenouille et al
4893
Oncogene
cells is because of increased protein stability rather than
a change in mRNA levels.
Posttranslational modification of p53 by phospho-
rylation is critical in regulating its stabilization and
activation (Lavin and Gueven, 2006). This prompted us
to examine the effects of SPARC depletion on the Ser15
and Ser46 phosphorylation of p53. These sites are most
commonly phospho rylated in response to DNA damage
and Ser46 phosphorylation was shown to be important
for p53-mediated apoptosis. A375 cells were treated with
actinomycin-D as a positive control. Incubation with
actinomycin-D induced the accumulation of p53 and
phosphorylation at both Ser15 and Ser46 residues
(Figure 5c). By contrast, no phosphorylation was
detected at any time point in cells treated with siSPARC.
This suggests that SPARC knockdown causes the
stabilization and activation of p53 without promoting
Ser15 or Ser46 phosphorylation of p53.
SPARC regulates MDM2 and p53 through the Akt
signaling pathway
Another possibility that could lead to the stability of p53
would be the regulation of MDM2 by phosphorylation.
Previous studies have shown that p53 can be regu-
lated by Akt-mediated phosphorylation of MDM2
(Mayo and Donner, 2001; Zhou et al. , 2001; Ogawara
et al., 2002). Phosphorylation of MDM2 pro motes its
localization to the nucleus, where it targets p53 for
degradation and prevents p53-dependent apoptosis. As
targeting SPARC expression was shown to reduce Akt
activity in glioma cells (Shi et al., 2004), we reasoned
that apoptosis of SPARC-depleted melanoma cells may
be caused by reduced Akt signaling. We found that
transfection of A375 cells with siSPARC strongly
reduced the basal activating phosphorylation of Akt
on Ser473 and phosphorylation of two downst ream
targets glycogen synthase kinase 3b (GSK3b) and
MDM2 in a time-dependent manner (Figure 6a).
Total MDM2 level also decreased in SPARC-depleted
cells, which is consistent with the previous observation
that phosphorylation of MDM2 by Akt leads to
increased MDM2 stability (Ashcroft et al., 2002). We
confirmed these results in 1205Lu and 501mel cells that
were depleted for SPARC (Figure 6a). Of note,
depletion of SPARC did not significantly affect extra-
cellular signal-regulated kinase, p38, protein kinase C
(PKC) or S6 ribosomal protein phosphorylation as well
as b-catenin protein levels (Supplementary Figure S5).
We next tested whether overexpre ssion of SPARC in
melanoma cells would result in enhanced Akt signaling.
We ectopically expressed an Myc-tagged SPARC con-
struct in 501mel cells and found that overexpression of
SPARC elevated phos pho-Akt and phospho-MDM2
levels, indicating that SPARC promotes Akt activity in
melanoma cells. Predictably, levels of p53 wer e lower in
Figure 5 SPARC depletion increases p53 protein stability in the absence of Ser15 or Ser46 phosphorylation of p53. (a ) RNAs were
extracted from A375 cells transfected with 50 n
M siCTRL or siSPARC for 6 days. p53 mRNA expression was measured by SYBR
green-based real-time Q-PCR and averaged from five independent experiments. The relative expression level of p53 mRNA was
normalized for RNA concentrations using four different housekeeping genes. The t-test was performed between the mean of the two
sample groups and yielded no significant result. (b) A375 cells were transfected with the indicated siRNA. After 4 days, the cells were
treated with 5 mg/ml cycloheximide (CHX, an inhibitor of protein synthesis) and harvested at the indicated time points. The protein
levels of p53 were analyzed by immunoblotting. The lower panel shows the quantification of the immunoblot. The amount of p53
protein relative to HSP60 was plotted against the time course of CHX treatment. The calculated half-life of the p53 protein is shown.
(c) A375 cells were incubated with 1 m
M actinomycin-D (ActD) or transfected with 50 nM siCTRL for 6 days, or siSPARC for various
time periods. Total p53 expression and levels of Ser15 and Ser46 phosphorylation were determined by immunoblotting. siRNA, small
interfering RNA; SPARC, secreted protein acidic and rich in cysteine.
Antiapoptotic function of SPARC in melanoma cells
N Fenouille et al
4894
Oncogene
Figure 6 SPARC activates the Akt-dependent signaling pathway in melanoma cells to promote their survival. (a) A375 cells were
transfected with 50 n
M siCTRL for 6 days, or siSPARC for the indicated time periods. 1205Lu and 501mel were transfected with 50 nM
siCTRL or siSPARC for 6 days. The immunoblots show the phosphorylation status of Akt, GSK3b and MDM2, and the expression
levels of SPARC and p53. HSP60 was used as a loading control. (b) 501mel cells overexpressing Myc-tagged human SPARC or
carrying an empty expression cassette of pcDNA3 vector (pcCTRL) were analyzed by immunoblotting using the indicated antibodies.
The expression of exogenous SPARC was confirmed using an anti-Myc tag antibody. (c) A375 cells were transfected with the indicated
siRNA for 6 days. The subcellular localization of p53, MDM2 and SPARC was analyzed by subcellular fractionation and
immunoblotting. Cytoplasmic (F1), mitochondrial (F2) and nuclear (F3) fractions were prepared as described under Materials and
methods. g-Tubulin, LAMP-2 and histone-H1 were used as purity controls for the cytoplasmic, mitochondrial and nuclear fraction,
respectively. (d) A375 cells were infected with a control empty adenovirus (AdCMV, multiplicity of infection of 2) or with adenovirus
encoding the constitutively active mutant Myr-HA-Akt1 (AdAkt
ca
, multiplicity of infection of 2), and transfected 6 h later with 50 nM
siCTRL or siSPARC for 6 days. The cells were then harvested and proteins were analyzed by immunoblotting using the indicated
antibodies. (e) Quantification of apoptosis in A375 cells treated as described above. The percentage of apoptotic cells was determined
by staining the cells with Annexin-V-fluos and flow cytometry. SPARC, secreted protein acidic and rich in cysteine.
Antiapoptotic function of SPARC in melanoma cells
N Fenouille et al
4895
Oncogene
SPARC-overexpressing cells compared with those in
parental cells (Figure 6b). We next asked whether
SPARC knockdown alters the subcellular localization
of p53 and MDM2 (Figure 6c). Subcellular fractiona-
tion showed that, in control cells MDM2 was found in
the nuclear fraction (F3). Depletion of SPARC pro-
moted the nuclear accumulation of p53 and reduction in
nuclear MDM2 levels, indicative of reduced phospho-
rylation and stability of MDM2. SPARC, being a
secreted protein in A375 cells (Fenouille et al., 2010),
was detected predominantly associated with the micro-
somal fraction (F2). Taken together, these results
suggest that SPARC signaling downregulates p53 levels
through the Akt/MDM2 pathway.
To further probe the contribution of Akt, we analyzed
whether a constitutively active form of Akt can rescue cells
from apoptosis-mediated by SPARC depletion. A375 cells
transfected with control or SPARC siRNAs were infected
with an adenovirus control (AdCMV) or expressing
Myr-Akt (AdAkt
ca
). Expression of Myr-Akt induced an
increase in the Ser166 phosphorylation of MDM2
concomitantly with a reduction in p53 levels, indicative
of activation of Akt signaling (Figure 6d). When
constitutively active Akt was expressed together with
siSPARC, no induction of p53 and Bax protein levels was
observed upon SPARC knockdown. Consistently, as
shown in PARP and Annexin-V assays, expression of
constitutively active Akt provided protection from apop-
tosis induced by SPARC knockdown (Figures 6d and e).
SPARC suppresses p53-mediated apoptosis in response to
the chemotherapeutic drug actinomycin-D
The RNA synthesis inhibitor actinomycin-D stimulates
apoptosis mostly through activation of the p53 pathway
(Choong et al., 2009). To establish whether SPARC can
modulate the p53 response, control and SPARC-over-
expressing 501mel cells were treated with 1 m
M actino-
mycin-D and cell death/survival was monitored at
various time points by XTT assay (Figure 7a). The
concentration of actinomycin-D was selected after pilot
experiments to determine the dose of the drug that
showed survival differences between the two popula-
tions of cells. We also confirmed that at 1 m
M the
apoptotic effect of actinomycin-D in 501mel cells was
mediated by p53 (Supplementary Figure S6). We found
that increasing SPARC expression greatly decreased the
cytotoxicity of the drug, resulting in resistance to cell
death compared with control cells. In accord, p53
induction in response to actinomycin-D treatment was
inhibited by SPARC expression. To determine whether
this phenomenon was mediated by abnormal degrada-
tion of p53, control and SPARC-overexpressing cells
were treated with the proteasome inhibitor MG132, to
prevent p53 degradation, in combination with actino-
mycin-D. MG132 pre-treatment led to better accumula-
tion of p53 in SPARC-overexpressing cells in response
to actinomycin-D and to a lesser extent also in the
absence of actinomycin-D (Figure 7b). Consistent with
the stabilization result, the pro tective effect of SPARC
on actinomycin-D-mediated apoptosis was reversed in
presence of MG132 as shown by Annexin-V assays.
These results support the notion that expression of
SPARC promotes melanoma cell survival by increasing
the proteosomal degradation of p53 under basal
conditions and in response to stress.
Discussion
Unlike other solid tumors, melanomas often retain
wild-type p53 protein. However, melanoma is arguably
Figure 7 Overexpression of SPARC suppresses p53 activation and
protects cells from actinomycin-D treatment. (a) Control and
SPARC-overexpressing 501mel cells were treated with 1 m
M actino-
mycin-D for the indicated time periods. Cell proliferation was then
measured by an XTT assay. The results are expressed in percent of
control. The columns indicate the mean of four independent
determinations; the error bars indicate the s.d. The immunoblots
show exogenous SPARC–Myc and p53 levels in the protein lysates
isolated from these cells. HSP60 was used as a loading control.
(b) Control and SPARC-overexpressing 501mel cells were pretreated
with the proteasome inhibitor MG132 (2 m
M) before treatment with
actinomycin-D (1 m
M) for 9 h. The percentage of apoptotic cells was
determined by staining the cells with Annexin-V-fluos and flow
cytometry. The immunoblots show exogenous SPARC–Myc and
p53 protein levels in the lysates isolated from these cells. PI,
propidium iodide; ROS, reactive oxygen species; siRNA, small
interfering RNA; shRNA, short-hairpin RNA; SPARC, secreted
protein acidic and rich in cysteine.
Antiapoptotic function of SPARC in melanoma cells
N Fenouille et al
4896
Oncogene
one of the most chemoresistant tumors. This paradox
indicates that p53 may be disabled and several mechan-
isms have been implicated in the alterations of the p53
pathway. This feature of melanoma cells might also be
related to their cell-autonomous survival signals such as
those mediated by the Ras–Raf–mitogen-activated pro-
tein kinase (MAPK) and PI3K–Akt pathways (Soengas
and Lowe, 2003; Tuveson et al. , 2003; Madhunapantula
and Robertson, 2009). Identification of novel pathways
that alter p53 functions and counteract cell death
induction is of critical importance in our current
understanding of the biology of melanomas.
Production of the matricellular protein SPARC is
regarded as a critical event in melanoma malignancy.
We recently revealed a link between SPARC and p53,
and proposed that some pro-tumorigenic functions
of SPARC may be because of relaxation of a p53-
dependent G
2
/M checkpoint (Fenouille et al., 2011). The
signaling pathways whereby SPARC controls p53,
however, remain to be identified. In this study, we show
that SPARC produced autonomously by melanoma
cells activates the antiapoptoti c Akt signaling pathway
to limit p53 levels through phosphorylation of MDM2.
This process protects cells that retained wild-type p53
from apoptosis induced by treatment with the anticancer
drug actinomycin-D.
Our data showed that wild-type, p53-expressing
melanoma cells are sensitive to SPARC depletion as
compared with melanoma cells that harbor mutant
copies of p53. Knockdown of SPARC resulted in
mitochondrial apoptotic death associated with p53
protein stabilization, Bax transcriptional induction and
its translocation to mitochondria, followed by ROS
generation and activation of caspase-9 and caspase-3.
We also provided evidence that induction of apoptosis
in SPARC-depleted cells is p53-dependent and that one
of the main downstream mediators of p53 in this death
process is Bax. However, despite the evidence presented
for the role of Bax in apoptosis, we cannot completely
exclude the possibility that other p53 targets such as
Survivin that were found to be regulated by SPARC
knockdown contribute to this process.
The generation of ROS in SPARC-depleted melano-
ma cells is consistent with a recent study showing that
SPARC-null epithelial cells have constitutive increased
basal levels of ROS compared with wild-type cells
(Weaver et al., 2008). ROS are important downstream
mediators of apoptosis triggered by p53 (Johnson et al.,
1996). The requirement of Bax to ROS increase that we
observed is consistent with the notion of a mitochon-
drial role in ROS production after upregulation of p53.
Finally, our observations that melanoma cells over-
expressing SPARC are more resistant to p53-mediated
apoptosis induced by actinomycin-D provide proof-of-
concept evidence that the pathological elevations of
SPARC frequently observed in melanomas might be
important for maintaining cell survival through inacti-
vation of p53 functions.
We confirmed our previous findings that p53 is acti-
vated following SPARC depletion in cells and xenograft
tumors (Fenouille et al., 2011). Activation of p53 in
SPARC-depleted cells promoted an early induction of
p21
Cip1/Waf1
and G
2
/M cell-cycle arrest that is followed by
Bax induction and apoptosis. It is not fully understood
how melanoma cells undergo apoptosis at the G
2
/M
checkpoint. However, our present data show that the
duration of p53 accumulation following SPARC defi-
ciency then resulted in the upregulation of Bax, whereas
p21
Cip1/Waf1
levels remained constant. Such sequential
kinetics of p21
Cip1/Waf1
and Bax expression might explain
the cell fate decision in SPARC-knockdown cells.
We went further to address how SPARC modulates
p53 and observed that SPARC knockdown med iates
p53 activation in the absence of markers of DNA
damage (that is, Ser15 or Ser46 phosphorylat ion of p53).
Instead, we found that SPARC activates Akt and
phosphorylation of MDM2 at Ser166, which was
identified as a site of activating phosphorylation by
Akt (Zhou et al., 2001; Ogawara et al., 2002).
Conversely, SPARC depletion lowered the amount of
Akt and MDM2 phosphorylation, and reduced the
levels of nuclear MDM2, which is in agreement with the
notion that Akt-mediated MDM2 phosphorylation
increases its stability (Ashcroft et al., 2002). These
observations combined with our data obtained using
MG132 make it very likely that SPARC signaling
through Akt downregulates p53 by activation of
MDM2-mediated ubiquitination and degradation of
p53. Importantly, we showed that activated Akt
prevents the apoptosis hallmarks of SPARC- depleted
cells. It is thus likely that SPA RC knockdown-induced
inhibition of Akt signaling is responsible for the p53-
mediated apoptosis reported here. Although a previous
work showed that SPARC activates Akt to mediate
its antiapoptotic function in glioma cells (Shi et al.,
2004), the observation that SPARC downregulates p53
through Akt-mediated activating phosphorylation of
MDM2 is a novel and unprecedented finding.
Integrins have emerged as potential cellular receptors
for SPARC. SPARC was shown to interact with
integrin-b1 (Nie et al. , 2008; Weaver et al., 2008), and
reports suggested that integrins avb3 and avb5 can
mediate some effects of SPARC (De et al., 2003;
Sangaletti et al., 2008). Integrin-avb3 has been linked
to melanoma progression and SPARC expression
(Sturm et al., 2002), and is required for melanoma
cell survival (Montgomery et al., 1994). Interestingly,
integrin-avb3 ligation can counteract p53 activation in
vascular endothelial cells (Stromblad et al., 1996).
Further research is thus required to determine whether
SPARC inactivates p53 through binding to melanoma
integrin-avb3.
The role of SPARC in cell survival and death is
complex. SPARC was originally identified as a stress
response gene (Sage et al., 1986), and subsequently
described as a c-Jun-responsive target gene that can be
repressed or induced depending on cell type (Mettouchi
et al., 1994; Briggs et al., 2002). Consistent with a role of
SPARC in cellular stress, recent studies showed that
SPARC protects lens epithelial cells from stress-induced
apoptosis and suppresses apoptosis in pulmonary
fibroblasts (Weaver et al., 2008; Chang et al., 2010).
Antiapoptotic function of SPARC in melanoma cells
N Fenouille et al
4897
Oncogene
In addition, a pro-survival activity for SPARC in
gliomas was documented (Shi et al., 2004). Our findings
agree with these observations and indicate that SPARC
functions to prevent stress-induced apoptosis in mela-
noma cells. Interestingly, the Drosophila homolog of
SPARC was recently shown to mediate a self-protective
signal during the cell competition process (Portela et al.,
2010). Thus, the stress protection function of SPARC
appears to be evolutionarily conserved between flies and
humans. On the other hand, there is evidence in some
contexts that SPARC induces apoptosis in ovaria n
cancer cells (Yiu et al., 2001) and modulates sensitivity
to chemotherapy in colon cancer cells by enhancing
apoptosis (Tang and Tai, 2007).
SPARC may exert its pro- or anti-tumorigeni c effects
through multiple mechanisms, but our study shows
that the antiapoptotic function of SPARC is mediated
through p53. Hence, some features of SPARC on tumor
biology might be related to the p53 status of cancer cells.
It is therefore possible that its paradoxical role in
tumorigenesis is a result of SPARC’s inhibitory effect
toward p53. Interestingly, it has been shown that p53
activation leads to the inhibition of SPARC secretion
in human glioma cells (Khwaja et al., 2006). This,
combined with our observation that SPARC negatively
regulates p53 functio n, argues in favor of a regulatory
feedback loop between SPARC and p53. It is tempting
to speculate that in melan omas, this loop would result
in increased SPARC secretion in the tumor micro-
environment and, in turn, further inhibition of the p53
pathway.
In conclusion, we propose that SPARC can promote
melanoma survival through tumor cell-aut onomous
signaling that involves Akt/MDM2-mediated p53 de-
gradation. Importantly, the sensitivity of wild-type p53-
carrying melanoma cells to SPARC removal indicates
that their survi val is reliant on SPARC levels. Reactiva-
tion of p53-dependent cell death programs has been
proposed previously for therapeutic intervention for
melanomas wi th a wild-type p53 status (Soengas et al.,
2001; Smalley et al., 2007). Our study shows the efficacy
of SPARC RNA interference to activate a p53 response
leading to apoptosis. In mela noma cases that retain
intact p53, targeting SPARC may thus be a promising
strategy to impede the progression of this lethal form of
skin cancer.
Materials and methods
Cell lines, siRNA, antibodies and reagents
Human melanoma cell lines with known p53 status were
maintained as described by Fenouille et al. (2011). A375 cells
stably expressing doxycycline-inducible SPARC shRNA and
501mel cells expressing an Myc-tagged human SPARC
(501mel SPARC) were described previously by Fenouille
et al. (2011). The list of antibodies used and details of siRNA
are included in the Supplementary information. Adenoviruses
carrying an empty expression cassette of pcDNA3 vector, used
as control or expressing the constitutively active mutant of
Akt1, Myr-HA-Akt1, were purchased from Vector Biolabs
(Philadelphia, PA, USA). Adenoviruses were amplified as described
earlier by Gaggioli et al. (2005). Chemicals were obtained from
Sigma-Aldrich (St Louis, MO, USA).
Cell proliferation and cell death analysis
Cell-cycle profiles and sub-G
1
analysis were performed by flow
cytometric analysis of propidium iodide (PI)-stained cells.
Cells were permeabilized and stained with 40 mg/ml PI before
analysis using a FACScan (Becton Dickinson, Le Pont de
Claix, France) and the CellQuest software. For flow cytometric
analysis of apoptosis, cells were stained with the Annexin-V-
fluos staining kit (Roche Applied Science, Indianapolis, IN,
USA). Cell proliferation assays were performed using the
XTT assay (Roche Diagnostic, Indianapolis, IN, USA) and
are described in the Supplementary information.
Colony formation and soft agar assays
Colony growth assay was performed as described by Bailet
et al. (2009). Details of soft agar assays are provided in the
Supplementary information.
Immunoblotting
Immunoblotting was performed as described previously by
Fenouille et al. (2011).
Subcellular fractionation assay
Cytoplasmic, microsomal and nuclear fractions were prepared
using the ProteoExtract-Subcellular Proteome Extraction kit
(Calbiochem, Merck Chemicals, Nottingham, UK) according to
the manufacturer’s instructions. The purity of each fraction was
analyzed by immunoblotting using antibodies against g-tubulin
(cytoplasmic fraction), LAMP-2 (microsomal fraction) and
histone-H1 (nuclear fraction).
Caspase activity measurement
Samples were lysed for 30 min at 4 1C in lysis buffer (50 m
M
HEPES (pH 8), 150 mM NaCl, 20 mM EDTA, 1 mM phenyl-
methylsulphonyl fluoride, 10 mg/ml leupeptin, 10 mg/ml apro-
tinin and 0.2% Triton X-100). Each assay was performed in
quadruplicate using 50 mg of protein. Briefly, extracts were
incubated for various time periods in the presence of 0.2 m
M
fluorogenic caspase substrates in a 96-well plate. The substrates
for each of the different caspases activities were as follows:
Caspase-3, caspase-6 and caspase-7, Ac-DEVD-AMC; Caspase-8,
Ac-IETD-AMC; Caspase-9, Ac-LEHD-AMC.
Detection of ROS by flow cytometry
ROS levels were measured using the redox-sensitive dye CM-
H
2
DCFDA (Molecular Probes, Invitrogen, Cergy Pontoise,
France). The green fluorescence was collected using a FACScan
flow cytometer and the CellQuest software for acquisition and
analysis.
Xenograft studies
The inducible xenograft model used to evaluate in vivo melanoma
growth has been described previously by Fenouille et al. (2011).
Statistical analysis
Details of statistical analysis have been provided in the
Supplementary information.
Conflict of interest
The authors declare no conflict of interest.
Antiapoptotic function of SPARC in melanoma cells
N Fenouille et al
4898
Oncogene
Acknowledgements
This work was supported in part by INSERM (Institut
National de la Sante
´
et de la Recherche Me
´
dicale) and
grant 1136 from ARC (Association pour la Recherche
sur le Cancer). S Tartare-Deckert is a recipient of a
Contrat d’Interface Clinique, Service de Dermatologie, CHU
de Nice.
References
Albino AP, Vidal MJ, McNutt NS, Shea CR, Prieto VG, Nanus DM
et al. (1994). Mutation and expression of the p53 gene in human
malignant melanoma. Melanoma Res 4: 35–45.
Arnold SA, Brekken RA. (2009). SPARC: a matricellular regulator of
tumorigenesis. J Cell Commun Signal 3: 255–273.
Ashcroft M, Ludwig RL, Woods DB, Copeland TD, Weber HO,
MacRae EJ et al. (2002). Phosphorylation of HDM2 by Akt.
Oncogene 21: 1955–1962.
Bailet O, Fenouille N, Abbe P, Robert G, Rocchi S, Gonthier N et al.
(2009). Spleen tyrosine kinase functions as a tumor suppressor in
melanoma cells by inducing senescence-like growth arrest. Cancer
Res 69: 2748–2756.
Barker TH, Baneyx G, Cardo-Vila M, Workman GA, Weaver M,
Menon PM et al. (2005). SPARC regulates extracellular matrix
organization through its modulation of integrin-linked kinase
activity. J Biol Chem 280: 36483–36493.
Bornstein P, Sage EH. (2002). Matricellular proteins: extracellular
modulators of cell function. Curr Opin Cell Biol 14: 608–616.
Briggs J, Chamboredon S, Castellazzi M, Kerry JA, Bos TJ. (2002).
Transcriptional upregulation of SPARC, in response to c-Jun
overexpression, contributes to increased motility and invasion of
MCF7 breast cancer cells. Oncogene 21: 7077–7091.
Chang W, Wei K, Jacobs SS, Upadhyay D, Weill D, Rosen GD.
(2010). SPARC suppresses apoptosis of idiopathic pulmonary
fibrosis fibroblasts through constitutive activation of beta-catenin.
J Biol Chem 285: 8196–8206.
Choong ML, Yang H, Lee MA, Lane DP. (2009). Specific activation of
the p53 pathway by low dose actinomycin D: a new route to p53
based cyclotherapy. Cell Cycle 8: 2810–2818.
De S, Chen J, Narizhneva NV, Heston W, Brainard J, Sage EH et al.
(2003). Molecular pathway for cancer metastasis to bone. J Biol
Chem 278: 39044–39050.
Fenouille N, Puissant A, Dufies M, Robert G, Jacquel A, Ohanna M
et al. (2010). Persistent activation of the Fyn/ERK kinase signaling
axis mediates imatinib resistance in chronic myelogenous leukemia
cells through upregulation of intracellular SPARC. Cancer Res 70:
9659–9670.
Fenouille N, Robert G, Tichet M, Puissant A, Dufies M, Rocchi S
et al. (2011). The p53/p21(Cip1/Waf1) pathway mediates the effects
of SPARC on melanoma cell cycle progression. Pigment Cell
Melanoma Res 24: 219–232.
Fukunaga-Kalabis M, Santiago-Walker A, Herlyn M. (2008).
Matricellular proteins produced by melanocytes and melanomas:
in search for functions. Cancer Microenviron 1: 93–102.
Gaggioli C, Deckert M, Robert G, Abbe P, Batoz M, Ehrengruber
MU et al. (2005). HGF induces fibronectin matrix synthesis in
melanoma cells through MAP kinase-dependent signaling pathway
and induction of Egr-1. Oncogene 24: 1423–1433.
Honda R, Tanaka H, Yasuda H. (1997). Oncoprotein MDM2 is a
ubiquitin ligase E3 for tumor suppressor p53. FEBS Lett 420: 25–27.
Horie K, Tsuchihara M, Nakatsura T. (2010). Silencing of secreted
protein acidic and rich in cysteine inhibits the growth of human
melanoma cells with G arrest induction. Cancer Sci 101
: 913–919.
Johnson TM, Yu ZX, Ferrans VJ, Lowenstein RA, Finkel T. (1996).
Reactive oxygen species are downstream mediators of p53-
dependent apoptosis. Proc Natl Acad Sci USA 93: 11848–11852.
Johnstone RW, Ruefli AA, Lowe SW. (2002). Apoptosis: a link
between cancer genetics and chemotherapy. Cell 108: 153–164.
Khwaja FW, Svoboda P, Reed M, Pohl J, Pyrzynska B, Van Meir EG.
(2006). Proteomic identification of the wt-p53-regulated tumor cell
secretome. Oncogene 25: 7650–7661.
Lavin MF, Gueven N. (2006). The complexity of p53 stabilization and
activation. Cell Death Differ 13: 941–950.
Ledda F, Bravo AI, Adris S, Bover L, Mordoh J, Podhajcer OL.
(1997a). The expression of the secreted protein acidic and rich in
cysteine (SPARC) is associated with the neoplastic progression of
human melanoma. J Invest Dermatol 108: 210–214.
Ledda MF, Adris S, Bravo AI, Kairiyama C, Bover L, Chernajovsky
Y et al. (1997b). Suppression of SPARC expression by antisense
RNA abrogates the tumorigenicity of human melanoma cells. Nat
Med 3: 171–176.
Madhunapantula SV, Robertson GP. (2009). The PTEN–AKT3
signaling cascade as a therapeutic target in melanoma. Pigment
Cell Melanoma Res 22: 400–419.
Marine JC, Francoz S, Maetens M, Wahl G, Toledo F, Lozano G.
(2006). Keeping p53 in check: essential and synergistic functions of
Mdm2 and Mdm4. Cell Death Differ 13: 927–934.
Massi D, Franchi A, Borgognoni L, Reali UM, Santucci M. (1999).
Osteonectin expression correlates with clinical outcome in thin
cutaneous malignant melanomas. Hum Pathol 30: 339–344.
Mayo LD, Donner DB. (2001). A phosphatidylinositol 3-kinase/Akt
pathway promotes translocation of Mdm2 from the cytoplasm to
the nucleus. Proc Natl Acad Sci USA 98: 11598–11603.
Mettouchi A, Cabon F, Montreau N, Vernier P, Mercier G, Blangy D
et al. (1994). SPARC and thrombospondin genes are repressed by the
c-jun oncogene in rat embryo fibroblasts. EMBO J 13: 5668–5678.
Miller AJ, Mihm Jr MC. (2006). Melanoma. N Engl J Med 355: 51–65.
Miyashita T, Reed JC. (1995). Tumor suppressor p53 is a direct
transcriptional activator of the human bax gene. Cell 80: 293–299.
Montgomery AM, Reisfeld RA, Cheresh DA. (1994). Integrin alpha v
beta 3 rescues melanoma cells from apoptosis in three-dimensional
dermal collagen. Proc Natl Acad Sci USA 91: 8856–8860.
Nie J, Chang B, Traktuev DO, Sun J, March K, Chan L et al. (2008).
IFATS collection: combinatorial peptides identify alpha5beta1
integrin as a receptor for the matricellular protein SPARC on
adipose stromal cells. Stem Cells 26: 2735–2745.
Nie J, Sage EH. (2009). SPARC functions as an inhibitor of
adipogenesis. J Cell Commun Signal 3: 247–254.
Ogawara Y, Kishishita S, Obata T, Isazawa Y, Suzuki T, Tanaka K
et al. (2002). Akt enhances Mdm2-mediated ubiquitination and
degradation of p53. J Biol Chem 277: 21843–21850.
Portela M, Casas-Tinto S, Rhiner C, Lopez-Gay JM, Dominguez O,
Soldini D et al
. (2010). Drosophila SPARC is a self-protective
signal expressed by loser cells during cell competition. Dev Cell 19:
562–573.
Robert G, Gaggioli C, Bailet O, Chavey C, Abbe P, Aberdam E
et al. (2006). SPARC represses E-cadherin and induces
mesenchymal transition during melanoma development. Cancer
Res 66: 7516–7523.
Sage H, Tupper J, Bramson R. (1986). Endothelial cell injury in vitro is
associated with increased secretion of an M
r
43 000 glycoprotein
ligand. J Cell Physiol 127: 373–387.
Sangaletti S, Di Carlo E, Gariboldi S, Miotti S, Cappetti B, Parenza M
et al. (2008). Macrophage-derived SPARC bridges tumor cell–
extracellular matrix interactions toward metastasis. Cancer Res 68:
9050–9059.
Schuler M, Green DR. (2001). Mechanisms of p53-dependent
apoptosis. Biochem Soc Trans 29: 684–688.
Shi Q, Bao S, Maxwell JA, Reese ED, Friedman HS, Bigner DD et al.
(2004). Secreted protein acidic, rich in cysteine (SPARC), mediates
cellular survival of gliomas through AKT activation. J Biol Chem
279: 52200–52209.
Antiapoptotic function of SPARC in melanoma cells
N Fenouille et al
4899
Oncogene
Shi Q, Bao S, Song L, Wu Q, Bigner DD, Hjelmeland AB et al. (2007).
Targeting SPARC expression decreases glioma cellular survival and
invasion associated with reduced activities of FAK and ILK
kinases. Oncogene 26: 4084–4094.
Smalley KS, Contractor R, Haass NK, Kulp AN, Atilla-Gokcumen
GE, Williams DS et al. (2007). An organometallic protein kinase
inhibitor pharmacologically activates p53 and induces apoptosis in
human melanoma cells. Cancer Res 67: 209–217.
Smit DJ, Gardiner BB, Sturm RA. (2007). Osteonectin downregulates
E-cadherin, induces osteopontin and focal adhesion kinase activity
stimulating an invasive melanoma phenotype. Int J Cancer 121:
2653–2660.
Soengas MS, Capodieci P, Polsky D, Mora J, Esteller M, Opitz-Araya
X et al. (2001). Inactivation of the apoptosis effector Apaf-1 in
malignant melanoma. Nature 409: 207–211.
Soengas MS, Lowe SW. (2003). Apoptosis and melanoma chemore-
sistance. Oncogene 22: 3138–3151.
Stromblad S, Becker JC, Yebra M, Brooks PC, Cheresh DA. (1996).
Suppression of p53 activity and p21WAF1/CIP1 expression by
vascular cell integrin alphaVbeta3 during angiogenesis. J Clin Invest
98: 426–433.
Sturm RA, Satyamoorthy K, Meier F, Gardiner BB, Smit DJ, Vaidya
B et al. (2002). Osteonectin/SPARC induction by ectopic beta(3)
integrin in human radial growth phase primary melanoma cells.
Cancer Res 62: 226–232.
Tang MJ, Tai IT. (2007). A novel interaction between procaspase 8
and SPARC enhances apoptosis and potentiates chemotherapy
sensitivity in colorectal cancers. J Biol Chem 282: 34457–34467.
Tuveson DA, Weber BL, Herlyn M. (2003). BRAF as a potential
therapeutic target in melanoma and other malignancies. Cancer Cell
4: 95–98.
Vogelstein B, Lane D, Levine AJ. (2000). Surfing the p53 network.
Nature 408: 307–310.
Weaver MS, Workman G, Sage EH. (2008). The copper binding
domain of SPARC mediates cell survival in vitro via interaction with
integrin beta1 and activation of integrin-linked kinase. J Biol Chem
283: 22826–22837.
Weiss J, Schwechheimer K, Cavenee WK, Herlyn M, Arden KC.
(1993). Mutation and expression of the p53 gene in malignant
melanoma cell lines. Int J Cancer 54: 693–699.
Yiu GK, Chan WY, Ng SW, Chan PS, Cheung KK, Berkowitz RS
et al. (2001). SPARC (secreted protein acidic and rich in cysteine)
induces apoptosis in ovarian cancer cells. Am J Pathol 159: 609–622.
Zhou BP, Liao Y, Xia W, Zou Y, Spohn B, Hung MC. (2001). HER-2/
neu induces p53 ubiquitination via Akt-mediated MDM2 phos-
phorylation. Nat Cell Biol 3: 973–982.
Supplementary Information accompanies the paper on the Oncogene website (http://www.nature.com/onc)
Antiapoptotic function of SPARC in melanoma cells
N Fenouille et al
4900
Oncogene
    • "increases the phosphorylation and activation of AKT and MDM2 to downregulate p53, which results in the inactivation of p53 by increased ubiquitination and proteasome degradation [74, 75]. Our previous studies investigated the correlation between protein expression in the AKT-MDM2-p53 pathway and H. pylori infection in chronic non-atrophic gastritis (CNAG), metaplastic atrophy (MA), gastric dysplasia (Dys) and GC patients. "
    Article · Jul 2015
    • "A similar effect on SPARC was previously described for γ-linoleic acid in breast and colon cancer cell lines [35]. Contrary to decreased cell survival and proliferation associated with SPARC silencing [36,37], our data showing reduction of extracellular SPARC in favour of its cytoplasmic form suggests some additional roles specifically played by intracellular SPARC. Abnormal accumulation of SPARC was also observed in the endoplasmic reticulum of patients with pseudo-achondroplasia [38] , reminiscent of the abnormal retention of collagens observed in other connective diseases [39]. "
    [Show abstract] [Hide abstract] ABSTRACT: A proper balance between saturated and unsaturated fatty acids (FA) is required for maintaining cell homeostasis. The increased demand of FA to assemble the plasma membranes of continuously dividing cancer cells might unbalance their ratio and critically affect tumor outgrowth. We unveiled the role of Stearoyl-CoA desaturase SCD5 in converting saturated FA into monounsaturated FA during melanoma progression. SCD5 is downregulated in advanced melanoma and its restored expression significantly reduced melanoma malignancy both in vitro and in vivo through a mechanism governing the secretion of extracellular matrix proteins, such as Secreted protein acidic and rich in cysteine (SPARC) and Ccollage IV and of their proteases, like Cathepsin B. Enforced expression of SCD5 or supplementation of its enzymatic product oleic acid reduced the intracellular pH (pHe > pHi) and, in turn, vesicular trafficking across plasma membranes as well as melanoma dissemination. This intracellular acidification appears also to depend on SCD5-induced reduction of the C2 subunit of the vacuolar H + -ATPase, a proton pump whose inhibition changes the secretion profile of cancer cells. Our data support a role for SCD5 and its enzymatic product oleic acid in protection against malignancy, offering an explanation for the beneficial Mediterranean diet. Furthermore, SCD5 appears to functionally connect tumor cells and surrounding stroma toward modification of tumor microenvironment, with consequences on tumor spread and resistance to treatment. This article is protected by copyright. All rights reserved.
    Full-text · Article · Mar 2015
    • "Cell cycle profiles were determined by flow cytometric analysis of DAPI-stained cells as described before [44]. Detection of NOx (NO and NO-derived species as nitrite and nitrate) in the supernatants of melanoma and melanocytes were detected by the Griess method using the manufacturer protocol. "
    [Show abstract] [Hide abstract] ABSTRACT: Nitric Oxide (NO) and Reactive oxygen species (ROS) are endogenous regulators of angiogenesis-related events as endothelial cell proliferation and survival, but NO/ROS defect or unbalance contribute to cancers. We recently designed a novel photoactive inhibitor of NO-Synthases (NOS) called NS1, which binds their NADPH site in vitro. Here, we show that NS1 inhibited NO formed in aortic rings. NS1-induced NO decrease led to an inhibition of angiogenesis in a model of VEGF-induced endothelial tubes formation. Beside this effect, NS1 reduced ROS levels in endothelial and melanoma A375 cells and in aorta. In metastatic melanoma cells, NS1 first induced a strong decrease of VEGF and blocked melanoma cell cycle at G2/M. NS1 decreased NOX4 and ROS levels that could lead to a specific proliferation arrest and cell death. In contrast, NS1 did not perturb melanocytes growth. Altogether, NS1 revealed a possible cross-talk between eNOS- and NOX4 –associated pathways in melanoma cells via VEGF, Erk and Akt modulation by NS1 that could be targeted to stop proliferation. NS1 thus constitutes a promising tool that modulates NO and redox stresses by targeting and directly inhibiting eNOS and, at least indirectly, NADPH oxidase(s), with great potential to control angiogenesis.
    Full-text · Article · Sep 2014
Show more