phospholipase D activity along with acidicsphingomyelinase activation to induce matrix metalloproteinase-9 expression in mouse metastatic melanoma Yasumasa Kato1,2, Shigeyuki Ozawa1,3, M
Trang 1phospholipase D activity along with acidic
sphingomyelinase activation to induce matrix
metalloproteinase-9 expression in mouse metastatic
melanoma
Yasumasa Kato1,2, Shigeyuki Ozawa1,3, Mamoru Tsukuda2, Eiro Kubota3, Kaoru Miyazaki4,
Yves St-Pierre5and Ryu-Ichiro Hata1
1 Department of Biochemistry and Molecular Biology, Kanagawa Dental College, Yokosuka, Japan
2 Department of Biology and Function in the Head and Neck, Yokohama City University Graduate School of Medicine, Japan
3 Department of Oral and Maxillofacial Surgery, Kanagawa Dental College, Yokosuka, Japan
4 Division of Cell Biology, Kihara Institute for Biological Research, Yokohama City University, Japan
5 INRS-Institut Armand-Frappier, Universite´ du Que´bec, Laval, Que´bec, Canada
Keywords
acidic sphingomylinase; Ca 2+ influx;
extracellular acidic pH; MMP-9
Correspondence
Y Kato, Department of Biochemistry and
Molecular Biology, Kanagawa Dental
College, Yokosuka 238-8580, Japan
Fax: +81 46 822 8839
Tel: +81 46 822 8840
E-mail: yasumasa@kdcnet.ac.jp
(Received 23 January 2007, revised 23 April
2007, accepted 27 April 2007)
doi:10.1111/j.1742-4658.2007.05848.x
Acidic extracellular pH is a common feature of tumor tissues We have reported that culturing cells at acidic pH (5.4–6.5) induced matrix metallo-proteinase-9 expression through phospholipase D, extracellular signal regu-lated kinase 1⁄ 2 and p38 mitogen-activated protein kinases and nuclear factor-jB Here, we show that acidic extracellular pH signaling involves both pathways of phospholipase D triggered by Ca2+ influx and acidic sphingomyelinase in mouse B16 melanoma cells We found that
BAPTA-AM [1,2-bis(2-aminophenoxy)-ethane-N,N,N¢,N¢-tetraacetic acid tetrakis (acetoxymethyl) ester], a chelator of intracellular free calcium, and the voltage dependent Ca2+ channel blockers, mibefradil (for T-type) and nimodipine (for L-type), dose-dependently inhibited acidic extracellular pH-induced matrix metalloproteinase-9 expression Intracellular free cal-cium concentration ([Ca2+]i) was transiently elevated by acidic extracellular
pH, and this [Ca2+]i elevation was repressed by EGTA and the voltage dependent Ca2+ channel blockers but not by phospholipase C inhibitor, suggesting that acidic extracellular pH increased [Ca2+]i through voltage dependent Ca2+channel In contrast, SR33557, an L-type voltage depend-ent Ca2+ channel blocker and acidic sphingomyelinase inhibitor, attenu-ated matrix metalloproteinase-9 induction but did not affect calcium influx
We found that acidic sphingomyelinase activity was induced by acidic extracellular pH and that the specific acidic sphingomyelinase inhibitors (perhexiline and desipramine) and siRNA targeting aSMase⁄ smpd1 could inhibit acidic extracellular pH-induced matrix metalloproteinase-9 expres-sion BAPTA-AM reduced acidic extracellular pH-induced phospho-lipase D but not acidic sphingomyelinase acitivity The acidic
Abbreviations
aSMase, acidic sphingomyelinase; BAPTA-AM, 1,2-bis(2-aminophenoxy)-ethane-N,N,N¢,N¢-tetraacetic acid tetrakis (acetoxymethyl) ester; CM, conditioned medium; [Ca2+] i , intracellular Ca2+concentration; DAG, diacylglycerol; ERK, extracellular signal regulated kinase; IL, interleukin;
IP3, inositol 1,4,5-triphosphate; JNK, c-Jun NH2-terminal kinase; MAPK, mitogen-activated protein kinase; MMP, matrix metalloproteinase; NF-jB, nuclear factor-jB; nSMase, neutral sphingomyelinase; PC, phosphatidylcholine; pHe, extracellular pH; PKCf, protein kinase Cf; PLC, phospholipase C; PLD, phospholipase D; SM, sphingomyelin; SMase, sphingomyelinase; TNF-a, tumor necrosis factor a; TPA, 12-O-tetradecanoylphorbol 13-acetate; VDCC, voltage dependent Ca 2+ channel; VEGF, vascular endothelial growth factor.
Trang 2Acidic extracellular pH (pHe) has been frequently
observed in solid tumors, due to excess amounts of
anaerobic glucose metabolites Acidic pHe has been
reported to affect the efficacy of chemotherapy,
inclu-ding reducing the cytotoxicity of bleomycin,
doxorubi-cin, daunorubidoxorubi-cin, epirubidoxorubi-cin, mitoxantrone, and vinca
alkaloids, but potentiating 5-fluorouracil [1] A recent
study demonstrated that acidic pHe is a predictor of
metastasis-free survival in canine soft tissue sarcomas
treated with thermoradiotherapy [2] The acidic
micro-environment may also regulate tumor angiogenesis
using a signal pathway different from that of hypoxia
[3–7] Hypoxia was recently reported to affect
expres-sion of matrix metalloproteinases (MMPs) [8], which
are important in inflammation, tumor invasion, and
metastasis
We have reported that acidic pHeinduced the
expres-sion of MMP-9⁄ gelatinase B (EC 3.4.24.35) in highly
metastatic mouse B16 melanoma cell lines, while not
affecting the expression of MMP-2⁄ gelatinase A [9]
We have also reported that acidic pHe-induced MMP-9
expression was mediated via the phospholipase D
(PLD)–mitogen-activated protein kinase (MAPK)
[extracellular signal regulated kinase (ERK)1⁄ 2 and
p38] pathway, at least in part through acidic pHe
signa-ling through nuclear factor-jB (NF-jB) [10]
Acidic pHe has been shown to increase intracellular
Ca2+ concentration ([Ca2+]i) in fibroblasts [11],
endo-thelial cells [12], and smooth muscle cells [13–15] In
addition, increased [Ca2+]i has been found to activate
PLD [16,17], which is also involved in the acidic pHe
induction of MMP-9 expression [10] [Ca2+]i elevation
can be divided into major two pathways: Ca2+ influx
through specific channels and release of Ca2+from the
endoplasmic reticulum by inositol 1,4,5-triphosphate
(IP3), a product of phospholipase C (PLC) Voltage
dependent Ca2+channels (VDCC) have been classified
into low (T-type) and high (L-type) voltage types,
which can be blocked by mibefradil and nimodipine,
respectively SR33557, which is another type of the
L-type VDCC blocker, can also inhibit mRNA
expres-sion of acidic sphingomyelinase (aSMase)⁄ acid
lyso-somal sphingomyelin phosphodiesterase 1 (smpd1) in
the signal transduction pathways of interleukin (IL)-1
and tumor necrosis factor a (TNF-a) [18,19]
MMP-9 can be up-regulated by various stimuli, including IL-1 and TNF-a, which trigger the ceramide-signaling pathway [20] Ceramide, which is generated
by the hydrolysis of sphingomyelin (SM) acts as a lipid second messenger for apoptotic signaling [21] Both aSMase and neutral sphingomyelinase (nSMase) can activate MAPKs, such as ERK1⁄ 2, Jun-N-terminal kin-ase (JNK), and p38, in various cell types [22–25] More-over, ceramide can induce MMP expression [26,27] Here, we report that PLD, which is activated by Ca2+
influx and aSMase, mediates the acidic pHeinduction of MMP-9, at least in part through NF-jB activation
Results
Acidic pHeincreases Ca2+influx through VDCC Increased [Ca2+]i has been shown to activate PLD [16,17] and acidic pHe has been shown to elevate [Ca2+]i in fibroblasts [11], endothelial cells [12], and smooth muscle cells [13–15] To determine the involve-ment of [Ca2+]i, in acidic pHe signaling, we treated cells with the calcium chelator BAPTA-AM [1,2-bis(2-aminophenoxy)-ethane-N,N,N1,N1-tetraacetic acid tetrakis (acetoxymethyl) ester] We found that BAPTA-AM dose-dependently attenuated the acidic
pHe-induced MMP-9 expression with an IC50 of 5.1 lm (Fig 1A) When we tested the effects of VDCC blockers on acidic pHe-induced MMP-9 expression, we found that the L-type VDCC blockers SR33557 [28,29] and nimodipine and the T-type blocker mibefradil dose-dependently inhibited acidic pHe-induced MMP-9 expression, with an IC50 of 13.7 lm, 3.0 lm, and 1.0 lm, respectively (Fig 1B,C) These agents at the same concentrations showed neither cellular toxicity nor any other gelatinolytic activity
Using Fluo4-AM, a fluorescent probe used to meas-ure [Ca2+]i, we observed a transient increase in [Ca2+]i
in the presence, but not in the absence, of extracellular
Ca2+(Fig 2A) The calcium chelator, EGTA, but not the broad PLC inhibitor U73122, attenuated the acidic
pHe-induced transient increase in [Ca2+]i, suggesting that [Ca2+]iis increased by Ca2+ entry not by inositol 1,4,5-triphosphate (IP3)-induced Ca2+ release from the endoplasmic reticulum (Fig 2B) Mibefradil and
sphingomyelinase inhibitors did not affect the phosphorylation of extracel-lular signal regulated kinase 1⁄ 2 and p38, but they suppressed nuclear factor-jB activity These data suggest that the calcium influx-triggered phospholipase D and acidic sphingomyelinase pathways of acidic extracel-lular pH induced matrix metalloproteinase-9 expression, at least in part, through nuclear factor-jB activation
Trang 3nimodipine prevented acidic pHe-induced Ca2+ influx
(Fig 2B), suggesting that Ca2+ influx, which occurred
through T-type and L-type VDCCs, triggered acidic
pHe-induced MMP-9 expression SR33557 (25 lm) did
not affect acidic pHe-induced Ca2+influx (Fig 2B) but
suppressed MMP-9 expression (Fig 1B), suggesting
that aSMase may be involved in acidic pHesignaling
aSMase mediates acidic pHe-induced MMP-9 expression
To investigate the involvement of aSMase in acidic
pHe signaling, we tested the effects of the aSMase specific inhibitors perhexiline [30–32] and desipramine [32–35] Both dose-dependently inhibited acidic pHe -induced MMP-9 expression, with an IC50 of 0.5 lm
A
B
Fig 2 [Ca 2+ ]iis increased through VDCC but not from the endo-plasmic reticulum Cells (40 000) cells were incubated overnight with serum-containing growing medium and with serum-free med-ium (pH 7.3) for 4 h and loaded with Fluo-4-AM (0.9 l M ) in NaCl ⁄ Pi containing 0.495 m M MgCl 2 for 30 min at room temperature (A) After washing, the cells were simulated by overlaying an acidic pH buffer [horizontal gray bar; NaCl ⁄ Pi (pH 5.9) supplemented with
15 m M Hepes, 4 m M phosphoric acid, and 0.495 m M MgCl 2 ] in the presence (open circle) or absence (closed circle) of 0.901 m M
CaCl2 [Ca 2+ ]i was measured at 490 nm excitation and 535 nm emission wavelengths, at 0.26 s intervals (B) Cells were treated with EGTA (5 m M ), mibefradil (2.5 l M ), nimodipine (5 l M ), SR33557 (25 l M ), and U73122 (50 l M ) on [Ca 2+ ]ifor 15 min and stimulated
as above Bars indicate SD.
A
B
C
Fig 1 Intracellular Ca 2+ chelator and VDCC blockers reduce acidic
pHe-induced MMP-9 expression Nearly confluent cells in a 24-well
culture plate were serum-starved overnight and cultured with acidic
medium (pH 5.9) in the presence of the indicated concentrations of
(A) BAPTA-AM, or (B) SR33557 for 48 h, or (C) mibefradil or
nimodi-pine for 24 h Proteins in the medium were ethanol concentrated,
and gelatinolytic activity was detected by gelatin zymography.
Experiments were performed three times; one representative
experiment is shown accompanied with the induction rate, which
was estimated by densitometry Concentration dependent
reduc-tion was seen with P-values less than 0.01 for SR33557, mibefradil
nimodipine and 0.001 for BAPTA-AM Molecular markers are
indica-ted in kDa Arrowheads indicate pro-MMP-9.
Trang 4and 6.0 lm, respectively (Fig 3A) Incubation of cells
at acidic pHe increased aSMase activity 2.0-fold but
had no effect on nSMase activity (Fig 3B) When
aSMase blockers were added to the cultures, they
sig-nificantly inhibited acidic pHe-induced aSMase activity
(Fig 3B), at concentrations sufficient to inhibit acidic
pHe-induced MMP-9 expression (Fig 3A) To prove
the contribution of aSMase in this signaling cascade
to induce MMP-9, small interfering RNA (siRNA)
technology was used Introduction of siRNA
oligo-nucleotide targeting smpd-1⁄ aSMase mRNA reduced
the acid induction of MMP-9 expression (Fig 4A)
concomitantly with the decrease in smpd-1⁄ aSMase
mRNA (Fig 4A) and its activity (Fig 4B) and also
in vivo ceramide production that is a metabolite of
aSMase from SM (Fig 4C) Interestingly, it was also found that acid induction of smpd-1⁄ aSMase mRNA expression, suggesting that the elevation of aSMase activity by acidic pHe, as shown in Fig 3B, is mainly due to an increase in the mRNA level rather than its activation These data showed a significant contribu-tion of aSMase in acidic pHe signaling to induce MMP-9 expression
Chelation of [Ca2+]iinhibits acidic pHe-induced PLD activation and MMP-9 expression but not aSMase activation
To determined the effect of [Ca2+]i elevation on PLD activity, cells were cultured with the [Ca2+]i chelater BAPTA-AM We found that this reagent dose-depend-ently reduced acidic pHe-induced PLD activity, but had no effect on aSMase activity (Fig 5), suggesting that acidic pHetriggers Ca2+influx, which is followed
by PLD activation independent of the aSMase path-way
[Ca2+]ielevation by thapsigargin at neutral pHe mimics acidic pHe-induced PLD activation and MMP-9 expression
If Ca2+ influx triggers PLD activation and this is fol-lowed by MMP-9 expression, we expected that we could mimic this effect at neutral pHe by increasing [Ca2+]i pharmacologically When the cells were cul-tured at neutral pHe with thapsigargin, a releaser of intracellular free Ca2+ from the endoplasmic reticu-lum, PLD activity was increased and MMP-9 was expressed [36,37] (Fig 6) 12-O-tetradecanoylphorbol 13-acetate (TPA) did not induce MMP-9 expression in B16 melanoma cells [9,10,38], but did so, through PLD activation, in HT1080 cells [39] Here, we found that TPA could not increase PLD activity (Fig 6), suggest-ing a reason that TPA could not induce MMP-9 expression in this model Besides, acid induction of MMP-9 expression was found without activation of AP-1 [10], generally known as the responsible factor for MMP-9 transcription which could be activated by TPA
In contrast, we found that exogenous addition of SMase dose-dependently stimulated the level observed
in the presence of thapsigargin (Fig 7A) Similarly,
C2-ceramide, a cell permeable ceramide analogue, increased MMP-9 expression in the presence, but not
in the absence, of thapsigargin at neutral pHe
(Fig 7B), suggesting that both SM and PC (phosphat-idylcholine) metabolites are important in acidic pHe induction of MMP-9 expression
A
B
Fig 3 aSMase mediates acidic pHeinduction of MMP-9
expres-sion Nearly confluent cells in a 24-well culture plate were
serum-starved overnight and cultured for 2 days in acidic medium (pH 5.9)
in the presence of the indicated concentrations of aSMase
inhibi-tors perhexiline maleate (perhexiline) and desipramine hydrochloride
(desipramine) (A) Proteins in CM were concentrated and analyzed
by gelatin zymography The arrowhead indicates MMP-9 activity.
(B) Membrane fractions (50 lg), prepared using a 0.2% Triton
X-100 buffer, were incubated for 60 min at 37 C in 250 m M sodium
acetate, 1 m M EDTA (pH 5.0) for aSMase or 250 m M Tris ⁄ HCl
(pH 7.4) for nSMase, each containing 0.05 lCi [choline methyl- 14
C]-SM Radioactive phosphorylcholine was extracted with
chloro-form ⁄ methanol (2 : 1, v ⁄ v) and the radioactivities in the aqueous
phase were determined by liquid scintillation counting Closed and
open columns indicate aSMase and nSMase activities, respectively.
Bars indicate SD.
Trang 5Inhibition of aSMase activity has no effect on
ERK1⁄ 2 and p38 phosphorylations
To assess the contribution of MAPKs to the
down-stream signaling of aSMase at acidic pHe, we
meas-ured the levels of the phosphorylated (active) forms of
MAPKs in these cultures We previously showed that
phosphorylation of ERK1⁄ 2 and p38 MAPKs was
significantly decreased by the PLD inhibitor
(1-buta-nol), whereas the total amounts of ERK1⁄ 2 and
p38 MAPKs were not affected [10] We found that
perhexiline and desipramine inhibition of aSMase did
not affect the activation of ERK1⁄ 2 and p38 MAPKs
(Fig 8) Similar findings were observed with the other aSMase inhibitor, SR33557 (data not shown) The JNK phosphorylation level was not affected by acidic
pHe [10] and aSMase inhibitors had no effect on its basal phosphorylation level (data not shown) These data suggested that ERK1⁄ 2 and p38 MAPKs were
A
B
C
Fig 4 Knockdown of aSMase ⁄ smpd1 expression reduces acidic
pH e -induced MMP-9 expression Cells, which have been
transfect-ed with siRNA oligonucleotide targeting aSMase ⁄ smpd1, were treated with acidic or neutral pHe (A) Proteins in CM were ethanol concentrated, and gelatinolytic activity was detected by gelatin zymography Total RNA was extracted and mmp-9, aSMase ⁄ smpd1 and b-actin gene expressions were analyzed by RT-PCR using spe-cific primer sets (B) Membrane fractions (50 lg), prepared using a 0.2% Triton X-100 buffer, were incubated for 60 min at 37 C in
250 m M sodium acetate, 1 m M EDTA (pH 5.0) containing 0.05 lCi [choline methyl- 14 C]-SM Radioactive phosphorylcholine was extrac-ted with chloroform ⁄ methanol (2 : 1, v ⁄ v) and the radioactivities in the aqueous phase were determined by liquid scintillation counting (C) The aSMase siRNA-transfected cells were labelled with 0.5 lCiÆmL)1[9,10-3H]-palmitic acid and then stimulated with acidic
pH medium for 24 h Lipids were extracted from the cells with chloroform ⁄ methanol and analyzed by thin layer chromatography The [3H]-ceramide formed was identified by comigration of N-palmi-toyl- D -erythro-sphingosine The spots, which were identified as [ 3 H]-ceramide, were scrapped off and the radioactivities were coun-ted by liquid scintillation counting Bars indicate SD *P < 0.05;
***P < 0.001 (Student’s t-test).
Fig 5 [Ca2+] i chelation reduces acidic pH e -induced PLD but not aSMase activity Nearly confluent cells in a 60 mm culture dish were serum-starved overnight and cultured for 2 days in acidic medium (pH 5.9), in the presence or absence of the indicated con-centrations of BAPTA-AM The membrane fractions (50 lg) were prepared, and aSMase activity (closed column) was measured by incubation for 60 min at 37 C in 250 m M sodium acetate, 1 m M
EDTA (pH 5.0) containing 0.05 lCi [choline methyl-14C]-SM, fol-lowed by scintillation counting of the aqueous phase PLD activity (open column) of the membrane fractions was measured using an AmplexTMRed PLD assay kit and a fluorescence microplate reader, with an excitation wavelength of 535 nm and a detection wave-length of 590 nm Bars indicate SD.
Trang 6not downstream targets of aSMase in acidic pHe
signaling
Inhibition of aSMase activity attenuates acidic
pHe-induced NF-jB and MMP-9 promoter
activities
The MAPK kinase inhibitor PD098059 and the p38
inhibitor SB203580 have been shown to inhibit acidic
pHe-induced NF-jB activity [10] We found that the aSMase inhibitors reduced wild-type MMP-9 promoter activity, as well as altering NF-jB-mutant MMP-9 promoter activity (Fig 9A) Moreover, aSMase inhibi-tion partially reduced acidic pHe-induced NF-jB activ-ity (Fig 9B), suggesting that acidic pHe-induced NF-jB is coregulated by the Ca2+⁄ PLD ⁄ MAPK and aSMase pathways These cascades proposed were sche-matically summarized in Fig 10 We found, however, that a mutant MMP-9 promoter lacking the NF-jB binding site (DNF-jB) still showed inducibility at acidic pHe and that this induction was attenuated by the aSMase inhibitors Although acidic pHe-induced NF-jB activity was down-regulated by these inhibitors
at the same concentrations, this inhibition was only partial, suggesting that other transcription factor(s) may be the downstream target(s) of aSMase Some candidates were considered
factors known within the minimal MMP-9 promoter region, Ets1 and SP1 were potentially involved in the aci-dic pHesignaling Indeed, using transcription factor-decoy and siRNA technologies, we found that Ets1 and SP1 were responsible for acid induction of MMP-9 expression (Y Kato, S Ozawa and R I Hata, unpublished data)
The upstream signaling cascade leading to their activa-tions (e.g MAPKs and aSMase) is currently under inves-tigation
Fig 6 Thapsigargin increased [Ca2+] i induces PLD activity and
MMP-9 expression at neutral pHe Nearly confluent cells were
serum-starved and incubated with thapsigargin (Thap, 2.5 l M ), TPA
(80 n M ) or vehicle at pH e 7.3 Gelatinolytic activity in CM was
ana-lyzed by zymography (inset) The cells were lysed with 0.2% Triton
X-100, and the lysates were subjected to Amplex TM Red PLD
assay Bars indicate SD *P < 0.05; ***P < 0.001 (Student’s t-test).
NS, not significant Arrowhead indicates pro-MMP-9.
A
B
Fig 7 SM hydrolysis contributes to MMP-9 expression Nearly
confluent cells were serum-starved and incubated for 48 h with the
indicated concentrations of bacterial SMase (Staphylococcus
aure-us) (A) or 25 l M C 2 -ceramide (B) in the presence or absence of
2.5 l M thapsigargin at pHe7.3 CM was collected, concentrated,
and MMP-9 activity was assayed by zymography Arrowheads
indi-cate pro-MMP-9.
Fig 8 aSMase inhibitors do not affect acidic pHe-induced phosphorylation of ERK1 ⁄ 2 and p38 Nearly confluent cells were serum-starved and incubated with or without 10 l M desipramine hydrochloride (desipramine) or 10 l M perhexiline maleate (perhexi-line) or at pHe5.9 for 48 h The cells were lysed and MAPK phos-phorylation was analyzed by western blotting using phospho-specific ERK1 ⁄ 2 or p38 polyclonal antibodies The induction rate of phosphorylated ratio was estimated by the densitometry and expressed as the relative values for the ratio of vehicle control at
pH e 7.3 p-ERK1 ⁄ 2, phosphorylated ERK1 ⁄ 2; p-p38, phosphorylated p38.
Trang 7Acidic pHe, a common feature of solid tumors, is thought to decrease the efficacy of chemotherapy regi-mens [40–44] Angiogenesis-related gene expression was found to be induced by acidic pHethrough hypoxia independent pathways involving platelet-derived endothelial cell growth factor⁄ thymidine phosphorylase
in human breast tumor cells [45], the inducible isoform
of nitric oxide synthase in macrophages [46], vascular endothelial cell growth factor in glioma [6] and gliobla-stoma [7] cells and IL-8 expression in human pancre-atic adenocarcinoma [3,47,48] and ovarian carcinoma cells [4] In addition, we have reported that expression
of MMP-9 in mouse metastatic B16 melanoma cells was induced by acidic pHe (pHe6.5–5.4) and that, among B16 clones, the rate of induction was correlated with metastatic potential [9] Most recently, acidic pHe
was reported to enhance the metastatic potential of human melanoma cells, accompanied by elevation of proteinases and proangiogenic factors such as MMP-9, MMP-2, cathepsin B, cathepsin L, vascular endothelial growth factor (VEGF)-A, and IL-8 [49] We also reported that acidic pHeinduction of MMP-9 expres-sion was mediated through the PLD–MAPK pathway [10] Here, we further examined whether increased [Ca2+]i and SM metabolism contributed to the acidic
pHe signaling induction of MMP-9 expression These contributions were also investigated in human lung adenocarcinoma cell line A549 Perhexiline (aSMase inhibitor) and nimodipine (L-type VDCC blocker) reduced acidic pHe-induced MMP-9 expression in A549 but mibefradile (T-type VDCC blocker) had no effect on this induction (data not shown), suggesting that the contribution of aSMase and Ca2+ influx is essential for acidic pHe signaling but the majority of the VDCC type
3 involved in this signaling is cell type specific
NF-jB is a transcription factor responsible for MMP-9 expression [50] and can mediate acidic pHe signaling [10] Acidic pHe-induced activity of PLD, but not aSMase, was suppressed by chelating [Ca2+]i, sug-gesting that Ca2+ influx activated PLD, but not aSMase It has been reported that aSMase activity could be induced by PC-derived diacylglycerol (DAG) through PC-PLC but not by phosphatidylinositol 4,5-biphosphate-derived DAG through PLD followed by phosphatidate phosphatase Because U73122 had little effect on [Ca2+]i, IP3is not likely to be involved in aci-dic pHeinduced [Ca2+]ielevation PC, a metabolite of PC-PLC, decreased after pHe dropped and D609, an inhibitor of PC-PLC, did not dose-dependently inhibit acidic pHe-induced MMP-9 expression [10] Thus,
Fig 10 Schematic representation of a proposed acidic pHe
signa-ling to induce MMP-9 expression.
Fig 9 aSMase inhibitors inhibit acidic pHe-induced NF-jB activity
and MMP-9 promoter activity Cells cultured overnight with 10%
fetal bovine serum in six-well plates were transfected with 1 lg
of mouse MMP-9 promoter-luciferase reporter construct (A) or
PathoDetect NF-jB-luciferase reporter construct (B) using
TransfectinTM in serum-free DMEM ⁄ F12 at pH e 7.3 After 18 h,
the cells were washed twice and cultured for 24 h with or
without perhexiline maleate (perhexiline) or desipramine
hydro-chloride (desipramine) at pH e 7.3 or 5.9 The cells were lysed
and subjected to dual luciferase assay; and transfection efficiency
was normalized by cotransfecting a Renilla luciferase reporter
construct WT, pGL3MMP9 (wild-type MMP-9 promoter
con-struct); DNF-jB, pGL3MMP9DNF-jB (MMP-9 promoter
con-struct mutated at the NF-jB binding site) **P < 0.05;
***P < 0.01 (Student’s t-test).
Trang 8although the pathway involving PC-PLC may be ruled
out, DAG derived from PC through PLD and
phosphatidate phosphatase, but not from
phosphati-dylinositol 4,5-biphosphate through
phosphatidyl-inositol specific PLC, may be involved in acidic pHe
signaling
Because thapsigargin induced MMP-9 expression
along with a 1.8-fold increase in PLD activity [16,36],
the basal activity of aSMase may be sufficient, but that
of PLD may be defective, for induction of MMP-9
expression at neutral pHe We have reported that
aci-dic pHe increased ERK1⁄ 2 and p38, but not JNK,
phosphorylation and that the former was attenuated
by 1-butanol, a PLD inhibitor [10] ERK1⁄ 2, JNK and
p38 are activated as downstream targets of nSMase
and induce MMP-1 expression in fibroblasts [26] In
B16-BL6 cells, however, aSMase inhibitors had little
effect on the phosphorylation of ERK1⁄ 2 and p38
Because ceramide can be metabolized from SM by
both SMases, this difference may be cell type specific
Further studies are needed to clarify the role of
aSMase in each cell type
We have shown here that, although NF-jB is a
downstream target of aSMase, the signaling pathway
connecting the two is still unclear One candidate
mediator is protein kinase Cf (PKCf), because
cera-mide is an activator of PKCf [51,52] and because
PKCf can directly phosphorylate the p65 (Ser311)
subunit of NF-jB [53] We found that a PKCf
pseudo-substrate can inhibit acidic pHe-induced MMP-9
expression (Y Kato, S Ozawa and R I Hata,
unpub-lished data)
Because aSMase not only contributes to apoptosis,
but also to metastatic ability, its ability to adapt and
be selected for resistance to microenvironmental stress
such as acidic pHe may be indicative of its more
aggressive phenotype, using an ‘apoptotic signal’ This
concept is supported by results showing that hypoxia
inducible factor 1a, a key transcription factor for
VEGF during angiogenesis, induces apoptosis in
nor-mal pancreatic islets [54] but prevents cell death and
even stimulates growth of pancreatic cancer cells [55]
Although the pHe of tumor tissues is acidic and
anaerobic glucose metabolites are the major source of
acidity, tumor acidity was shown to be caused by
excess amounts of CO2, regardless of pO2, through the
pentose phosphate pathway, in glycolysis-impaired
(phosphoglucose isomerase-deficient) cells [56] This
pathway provides cells with ribose 5-phosphate, which
is used to synthesize nucleic acids Thus, highly
prolif-erating cells need more ribose 5-phosphate for DNA
replication and RNA synthesis, thereby producing
excess amounts of CO2 These observations suggest
that extracellular acidity in tumors is partly regulated
by an hypoxia-independent pathway Because tumor acidity affects the response radiation therapy and che-motherapy, pharmacological blockade of VDCC
prevent tumor invasion and metastasis
In conclusion, we found that two independent path-ways; Ca2+–PLD–MAPKs (ERK1⁄ 2 and p38) and aSMase, leading to NF-jB activation, are essential in acidic pHeinduction of MMP-9 expression
Experimental procedures
Reagents SR33557 [([2-isopropyl-1-(4-[3-N-methyl-N-(3,4-dimethoxy-phenethyl) amino] propyloxy) benzenesulfonyl]) indolizine],
an aSMase specific inhibitor, was kindly provided by San-ofi-Aventis (Paris, France) BAPTA-AM and
N-palmitoyl-d-erythro-sphingosine [C16:0 (palmitoyl) ceramide] were purchased from Calbiochem (La Jolla, CA, USA), and fluo 4-AM was obtained from Dojindo (Kumamoto, Japan) DMEM and Ham’s F-12 (F-12), and TRIzol Regent were obtained from Invitrogen (Carlsbad, CA, USA); Trans-fectinTM and siLentFectTM Lipid Reagents were obtained from Bio-Rad (Hercules, CA, USA); the Dual Luciferase Reporter Assay kit was obtained from Toyo Ink (Tokyo, Japan); Staphylococcus aureus SMase, perhexiline maleate salt, and desipramine hydrochloride were obtained from Sigma (St Louis, MO, USA); fetal bovine serum was obtained from Cell Culture Technologies GmbH (Zurich, Switzerland); [choline methyl-14C]-SM was obtained from Amersham Biosciences (Piscataway, NJ, USA); [9,10-3 H]-palmitic acid (50.0 CiÆmmol)1) was obtained from Moravec Biochemicals (Brea, CA, USA); Immobilon-P [poly(vinylid-ene difluoride)] membrane was obtained from Millipore (Bedford, MA, USA); and the Nuclear Extract kit was obtained from Active Motif (Carlsbad, CA, USA) EGTA, TPA and the ImmunostarTM Western blotting detection kits, which included a chemiluminescent reagent and peroxidase-conjugated swine rabbit IgG or goat anti-mouse IgG, were obtained from Wako (Tokyo, Japan) The blocking reagent N102 was obtained from NOF Corp (Tokyo, Japan); siRNA oligonucleotide targeting aSMase⁄ smpd1 and a control oligonucleotide (scramble), and anti-bodies directed against total or phosphorylated MAPKs (sc-7976-R, sc-154, sc-7149, sc-7975-R, sc-571, sc-6254) were obtained from Santa Cruz (Santa Cruz, CA, USA) Silica Gel60 F254 plate was obtained from Merck KGaA (Darmstadt, Germany)
Vectors The PathoDetect NF-jB cis-reporting system (pNF-jB-Luc) was obtained from Stratagene (La Jolla, CA, USA)
Trang 9The MMP-9 promoter luciferase reporter construct and its
mutant construct of the NF-jB binding site have been
described previously [10,57] The cytomegalovirus-driven
Renilla luciferase reporter vector (pRL-CMV, Promega,
Madison, WI, USA) was used to monitor transfection
efficiency
Cells and cell culture
B16-BL6 cells were cultured in DMEM containing 15 mm
Hepes (pH 7.3) supplemented with heat-inactivated 10%
fetal bovine serum Because induction of MMP-9
expres-sion occurred from pHe6.5–5.4 [9], we fixed the pH of
the assay media at 5.9 for acidic pHe and at 7.3 for
neut-ral pHe To prepare serum-free assay media (DMEM⁄
F-12), a 1 : 1 mixture of DMEM and F-12 was
supple-mented with 15 mm Hepes and 4 mm phosphoric acid and
adjusted to pH 5.9 with HCl or to pH 7.3 with NaOH
[9,10,38]
SiRNA-mediated gene silencing
To suppress aSMase mRNA expression, siRNA technology
was used Oligonucleotide (2 nm) targeting aSMase⁄ smpd1
was transfected into cells with siLentFectTM Lipid Reagent
in a serum-free DMEM⁄ F-12 at pH 7.3 and cultured for
48 h The transfectants were stimulated with acidic medium
for 48 h The scrambled siRNA were used for a control At
the end of incubation period, proteins in conditioned
med-ium (CM) and total RNA were obtained for zymography
to detect MMP-9 activity and RT-PCR was used to detect
mmp-9gene expression
Preparation of concentrated CM for zymography
Proteins in CM were concentrated by adding three volumes
of ice-cold ethanol as described previously [10,58] The
quantity of samples was normalized for zymography assay
based on the DNA contents of the cultures (1.5 lg DNA⁄
lane), as measured using bisbenzimide [59]
Gelatin zymography
Gelatinolytic activities in the CM were analyzed by gelatin
zymography, as described previously [9,10,60,61] Briefly,
ethanol-precipitated proteins were electrophoresed in
SDS-7.5% polyacrylamide gels containing 0.1% gelatin The gels
were washed in 2.5% Triton X-100 with gentle shaking for
1 h at room temperature to remove SDS and incubated for
20 h in reaction buffer [50 mm Tris⁄ HCl (pH 7.5), 100 mm
NaCl, 10 mm CaCl2, and 0.002% NaN3] at 37C
Gela-tinolytic activity was visualized as a clear zone on a
blue background following Coomassie Brilliant Blue R250
staining
[Ca2+]imeasurements Cells were inoculated at a density of 40 000 cells⁄ well in 96-well culture plates Following overnight incubation, the cells were washed twice with Ca2+- and Mg2+-free Dulbecco’s phosphate-balanced saline (NaCl⁄ Pi) and incubated in serum-free DMEM for 4 h The cells were incubated with Fluo 4-AM (final concentration, 0.9 lm) in NaCl⁄ Pi (pH 7.3) containing 0.901 mm CaCl2and 0.495 mm MgCl2 for 30 min at room temperature and washed four times with NaCl⁄ Pi (pH 7.3) containing 0.495 mm MgCl2 The cells were overlain with NaCl⁄ Pi (pH 5.9) supplemented with
15 mm Hepes, 4 mm phosphoric acid, and 0.495 mm MgCl2
in the presence or absence of 0.901 mm CaCl2 The [Ca2+]i was measured at 490 nm excitation and 535 nm emission wavelengths, at 0.26 s intervals using Tecan GENiosProTM fluorescence plate reader (Gro¨dig, Salzburg, Austria) Where indicated, cells were incubated for 5 min with cal-cium channel blockers dissolved in the NaCl⁄ Pi (pH 7.3) containing 0.901 mm CaCl2 and 0.495 mm MgCl2, and the cells were overlain with channel blocker-containing NaCl⁄ Pi (pH 5.9) supplemented with 15 mm Hepes, 4 mm phosphoric acid, 0.495 mm MgCl2, and 0.901 mm CaCl2
PLD activiy Membrane fractions of the cells were prepared using 0.2% Triton X-100 PLD activity was detected using the AmplexTM Red PLD assay kit (Molecular Probes, Eugene,
OR, USA) [10] Whole cell lysates were incubated with
250 lm PC, 100 mUÆmL)1 Alcaligenes sp choline oxidase,
1 UÆmL)1 horseradish peroxidase, and 50 lm 10-acetyl-3,7-dihydrophenoxazine (AmplexTM Red reagent) in reaction buffer consisting of 50 mm Tris⁄ HCl (pH 8.0), 5 mm CaCl2, and 0.2% Triton X-100 PLD activity was measured with a fluorescence microplate reader using an excitation wavelength of 535 nm and detection wavelength of 590 nm
SMase activities SMase activities were measured as described previously [18] Briefly, membrane fractions (50 lg), prepared using 0.2% Triton X-100, were incubated for 60 min at 37C in 200 lL
250 mm sodium acetate, 1 mm EDTA (pH 5.0) for aSMase
or 200 lL 250 mm Tris⁄ HCl (pH 7.4) for nSMase, each con-taining 0.05 lCi [choline methyl-14C]-SM Radioactive phos-phorylcholine was extracted with 750 lL of chloroform⁄ methanol (2 : 1, v⁄ v), and the radioactivity in the aqueous phase was determined by liquid scintillation counting
In vivo ceramide production
In vivo ceramide production
previously [62,63] Cells were inoculated into six-well
Trang 10culture plate at a density of 2.5· 105
cells⁄ well Follow-ing overnight incubation, the cells were washed twice
with NaCl⁄ Pi and labelled with 1.5 lCi ⁄ well [9,10-3
H]-palmitic acid in serum-free DMEM for 18 h The cells
were then stimulated with acidic assay medium for 24 h
Lipids were extracted from the cells with
chloro-form⁄ methanol (2 : 1, v ⁄ v) Lipids in the chloroform
phase were collected and analyzed by thin-layer
chroma-tography using a Silica Gel60 F254 plate (20· 20 cm) and
ethyl acetate⁄ acetic acid ⁄ 2,2,4-trimethypentane (9 : 2 : 5)
as a solvent The spots, which were identified as [3
H]-cer-amide by comigration of
N-palmitoyl-d-erythro-sphingo-sine [C16:0 (palmitoyl) ceramide], were scrapped off and
their radioactivities were counted by liquid scintillation
counter
RT-PCR
Total RNA was extracted by using TRIsol Reagent,
reverse-transcribed by MMLV super transcriptase, and
amplified by Taq polymerase with specific primer sets:
aSMase⁄ smpd1 (26 cycles, 258 bp), 5¢-TTC CTG CCA
GAG CTT ATC-3¢ (forward) and 5¢-TCC TCA AAG
AGA TGG ACG-3¢ (Reverse); mmp-9 (28 cycles, 471 bp),
5¢-GTA TGG TCG TGG CTC TAA GC-3¢ (forward)
and 5¢-AAA ACC CTC TTG GTC TGC GG-3¢ (reverse);
b-actin (18 cycles, 555 bp) 5¢-CAT CGT GGG CCG
CTC TAG GCA CCA AG-3¢ (forward) and 5¢-GCA
CAG CTT CTC TTT GAT GTC ACG CAC-3¢ (reverse)
PCR thermal conditions used were: aSMase⁄ smpd1 and
mmp-9, 94C for 30 s; annealing, 56C for 30 s;
extention, 72C for 30 s; b-actin, denature, 94 C for
30 s; annealing, 62C for 30 s; extention, 72 C for
30 s
Western blot analysis
The active forms of MAPKs were detected by western
blotting as described previously [10,64] Cells were lysed
with the Nuclear Extract kit according to the
manufac-turer’s protocol Proteins in the cell lysate (20 lg) were
separated on SDS-containing 10% polyacrylamide gels
and transferred to Immobilon-P membranes using the
Bio-Rad western blot apparatus After blocking with
20% blocking reagent N102 in Tris-buffered saline
solu-tion [20 mm Tris⁄ HCl (pH 7.6), 137 mm NaCl] containing
0.05% Tween-20, the membrane was incubated with
primary antibody in the same buffer containing 10%
Blocking Regent N102 After sequential incubations
with biotin-conjugated secondary antibody and
horserad-ish peroxidase-conjugated avidin, the blots were incubated
with a chemiluminescent substrate using an
Immuno-starTM detection kit, and the signals were detected
with the LAS3000 imaging system (Fuji Film, Tokyo,
Japan)
Luciferase reporter assay The PathoDetect NF-jB cis-reporting system, an inducible reporter vector containing the luciferase reporter gene driven by a basic promoter element (TATA box) and the cis-enhancer NF-jB, was used to measure NF-jB activity [10] An MMP-9 promoter luciferase reporter construct and its mutant construct were used to measure MMP-9 promo-ter activity [10,57] These reporpromo-ter vectors (1 lg⁄ 35 mm dish) were transfected into B16-BL6 cells with Trans-fectinTMin six-well culture plates according to the manufac-turer’s protocol, and transfection efficiency was monitored
by cotransfection of the Renilla luciferase reporter vector (pRL-CMV) and a dual luciferase reporter assay kit
Protein concentrations Protein concentration was determined according to the Bradford method, using the Bio-Rad protein assay kit and bovine serum albumin as the standard
Statistical analysis The two-tailed Student’s t-test was used for statistical com-parisons A value of P < 0.05 was considered statistically significant
Acknowledgements
We thank Drs Charles A Lambert, Pierre Mineur, Agne´s Noe¨l, Francis Frankenne, and Jean-Michel Foidart of the Universite´ de Lie`ge, Belgium, for their critical discussions This work was supported in part
by the Grants-in-Aid for ‘High-Tech Research Center Project’ from the Ministry of Education, Culture, Sports, Science and Technology of Japan and for Scientific Research (B) and (C)
for the Promotion of Science, Japan
References
1 Raghunand N & Gillies RJ (2001) pH and chemother-apy Novartis Found Symp 240, 199–211; discussion 265–198
2 Lora-Michiels M, Yu D, Sanders L, Poulson JM, Azuma C, Case B, Vujaskovic Z, Thrall DE, Charles
HC & Dewhirst MW (2006) Extracellular pH and P-31 magnetic resonance spectroscopic variables are related
to outcome in canine soft tissue sarcomas treated with thermoradiotherapy Clin Cancer Res 12, 5733– 5740
3 Shi Q, Abbruzzese JL, Huang S, Fidler IJ, Xiong Q & Xie K (1999) Constitutive and inducible interleukin 8 expression by hypoxia and acidosis renders human