Glibenclamide inhibits cell growth by inducing G0/G1 arrest in the human breast cancer cell line MDA-MB-231.

1.6 MB·English

by  Nunez, Mariel

Núñezetal.BMCPharmacologyandToxicology2013,14:6
http://www.biomedcentral.com/2050-6511/14/6
RESEARCH ARTICLE Open Access
Glibenclamide inhibits cell growth by inducing
G0/G1 arrest in the human breast cancer cell line
MDA-MB-231
Mariel Núñez1, Vanina Medina1, Graciela Cricco1, Máximo Croci2, Claudia Cocca1, Elena Rivera1, Rosa Bergoc1,3*
and Gabriela Martín1
Abstract
Background: Glibenclamide (Gli) binds to the sulphonylureareceptor (SUR) that is a regulatory subunit of
ATP-sensitive potassium channels (K channels). Binding of Gli to SURproduces theclosure ofK channels and
ATP ATP
theinhibitionof their activity. This drug is widely used for treatmentof type 2-diabetes and it has been signaled as
antiproliferative inseveraltumor cell lines. In previous experiments we demonstrated the antitumoral effect of Gli in
mammary tumors induced inrats. The aim of the present workwas to investigate theeffect of Gli onMDA-MB-231
breast cancer cell proliferation and to examine the possible pathways involved in this action.
Results: The mRNA expression ofthe different subunits that compose the K channels was evaluatedin
ATP
MDA-MB-231 cells byreversetranscriptase-polymerase chain reaction. Results showed the expression ofmRNA for
both pore-forming isoforms Kir6.1 and Kir6.2 and for theregulatory isoformSUR2B in this cell line. Gli inhibited cell
proliferation assessed by a clonogenic method ina dose dependent manner, with anincrementin thepopulation
doubling time. The K channel opener minoxidil increased clonogenic proliferation, effectthat was counteracted
ATP
byGli.When cell cycle analysis was performed by flow cytometry,Gli induced a significant cell-cyclearrest inG0/G1
phase, together with anup-regulation ofp27 levels and a diminution incyclin E expression,both evaluated by
immunoblot. However, neither differentiation evaluated by neutral lipid accumulation nor apoptosis assessed by
different methodologies were detected. The cytostatic,non toxic effect oncell proliferation was confirmed by
removal of thedrug.
Combination treatment of Gli withtamoxifen or doxorubicin showed an increment inthe antiproliferativeeffect
onlyfor doxorubicin.
Conclusions: Our data clearly demonstrated a cytostatic effect ofGli in MDA-MB-231cells that may be mediated
through K channels, associated to the inhibition oftheG1-S phase progression.In addition, aninteresting
ATP
observation about the effect of the combination of Gli withdoxorubicin leads to future research for a potential
novel rolefor Gli as an adjuvant in breast cancer treatment
Keywords: Glibenclamide, Potassium channels,MDA-MB-231,Cytostaticeffect
*Correspondence:[email protected]
1RadioisotopesLaboratory,SchoolofPharmacyandBiochemistry,University
ofBuenosAires,BuenosAires,Argentina
3InstituteofHealthSciencesBarceló,BuenosAires,Argentina
Fulllistofauthorinformationisavailableattheendofthearticle
©2013Núñezetal.ThisisanOpenAccessarticledistributedunderthetermsoftheCreativeCommonsAttributionLicense
(http://creativecommons.org/licenses/by/2.0),whichpermitsunrestricteduse,distribution,andreproductioninanymedium,
providedtheoriginalworkisproperlycited.
Núñezetal.BMCPharmacologyandToxicology2013,14:6 Page2of13
http://www.biomedcentral.com/2050-6511/14/6
Background with cyclophosphamide, taxanes and/or 5-fluorouracil;
Sulphonylureas are used to increase insulin secretion in however, resistance to doxorubicin is common [16,17].
patient with type 2-diabetes due to their direct action on The search for effective drugs with low side effects is
pancreatic β cells. These drugs bind to the β cell sulpho- still achallengetoresearchers.
nylurea receptor (SUR) that is a regulatory subunit of MDA-MB-231celllinederivedfromahumanbreastcar-
ATP-sensitive potassium channels (K channels) [1-3]. cinomathatdonotexpressERα,isoftenusedasanexperi-
ATP
K channels regulate the transport of potassium ions mental non hormone-dependent tumor model [18,19].
ATP
through cell membranes. A diverse group of compounds Theobjectiveofthepresentworkwastoinvestigatetheef-
can bind to K channels causing them to open or fect of Gli on MDA-MB-231 cells proliferation and to
ATP
close. The opening of potassium channels in the cell examinethepossiblepathwaysinvolvedinthisaction.
membrane produces a hyperpolarization of membrane
potential. These channels are hetero-octameric com- Materialandmethods
plexes that consist of two rings: an inner ring of four in- Materials
wardly rectifying K+ channels (Kir6.X) that forms the MBA-MB-231 cells were obtained from American
pore through which potassium ions pass, and an outer Type Culture Collection (ATCC). Gli was kindly pro-
ring that comprises four SUR subunits. Two isoforms vided by Investi-Farma SA, Buenos Aires, Argentina.
have been described for Kir6.X (Kir6.1 and Kir6.2) and Tamoxifen (Tam) was a gift from Gador Laboratories
also for SUR (SUR1 and SUR2; SUR2 also has two splice SA, Buenos Aires, Argentina. RPMI 1640 medium
variants, SUR2A and SUR2B) [4]. In pancreatic β cells and fetal bovine serum (FBS) were purchased from
binding of sulphonylureas to SURs produce the closure GIBCO, Invitrogen, CA, USA. Ribonuclease, propidium
0
of K channels reducing cellular potassium efflux thus iodine, 3,3dihexyloxacarbocyanine iodide (DiOC6),
ATP
favoring membrane depolarization, the induction of Ca2+ saponine, FITC-labeled anti-rabbit, 5-bromo-20-deoxyuri-
influx,andinsulinsecretion[1-3]. dine (BrdU), mouse anti-BrdU monoclonal antibody,
0
Glibenclamide (Gli), a diarylsulphonilurea that blocks FITC-conjugated anti-mouse IgG, 4,6-Diamidino-2-
specifically K channels, is widely employed in the phenyindole, dilactate (DAPI), 5-bromo-4-chloro-3-indo-
ATP
treatment of diabetic patients [5], but various reports lyl-β-d-galactoside (X-gal), mouse anti-β-actin polyclonal
also described its antiproliferative effect in different neo- antibody, p27Kip1 monoclonal antibody and Nile-red stain
plastic cell lines [6,7]. Additionally, the inhibition of were purchased from Sigma, St Louis, MO, USA. Apop-
W
other classes of potassium channels also leads to a de- tag PLUSPeroxidaseInSituDetectionKitS701wasfrom
crease ofproliferation innormalandcancercells[8,9]. Chemicon International, CA, USA. Rabbit polyclonal anti-
Breast cancer is the neoplastic disease most frequently bodies against human Bax, Bcl-2 and Bcl-x were from
L/S
observed in women all over the world [10-12]. A high SantaCruzBiotechnologies,SantaCruz,CA,USA.Mouse
proportion ofmammarytumors are positive for estrogen anti-cyclin B1 monoclonal antibodies, mouse anti-cyclin E
receptors α (ERα) and consequently antihormonal ther- monoclonalantibodies,andrabbitanti-cyclinD1monoclo-
apyisindicated.Theselectiveestrogenreceptormodula- nal antibodies were from Cell Signaling Technology, Inc.,
tor tamoxifen continues to be the drug regularly used in Danvers, MA, USA. Annexin V-FITC was from Bios-
patients harboring this kind of tumors due to its efficacy ciences, USA. Chemiluminiscence system (ECL) was from
and low toxicity [13]. However, approximately 30% of AmershamBiosciencesArgentinaSS(Argentina).Nitrocel-
ERα (+) tumors do not respond to tamoxifen or develop lulose membranes were from Santa Cruz Biotechnologies,
resistance in the course of the treatment. In addition, it SantaCruz,CA,USA.MultiwellswerefromTPP,Switzer-
is known that approximately 30% of tumors do not ex- land.Allotherreagentswereofanalyticalgrade.
press ERα [14]. Although a large number of drugs have
been developed for the treatment of ERα (−) tumors, Methods
most of them give rise to important toxic effects. In Reversetranscriptasepolymerasechainreaction(RT-PCR)
order to attain better therapeutic effectiveness, combin- analysis
W
ation cytotoxic treatments for aggressive cancers have Total cellular RNA was extracted using Trizol accord-
been employed in clinics. The simultaneous use of drugs ing to the instructions of the manufacturer (GIBCO, Life
with different molecular targets can delay the emergence Technologies, USA). Total RNA (2 μg) was added to the
of chemoresistance whereas when drugs are directed to reverse transcription (RT) reaction mixture (20 μl) in
the same cellular pathway they could work synergically the presence of oligo-dT primers and samples were
for higher efficacy and selectivity. However, combination incubated at 37°C for 60 minutes. The quality of each
therapy may also increase toxicity [15]. Doxorubicin is individual’s cDNA was confirmed by PCR with primers
considered a highly effective agent in the treatment of for β-actin producing bands of the expected size (data
aggressive breast cancer patients sometimes combined not shown).The primers used were employed previously
Núñezetal.BMCPharmacologyandToxicology2013,14:6 Page3of13
http://www.biomedcentral.com/2050-6511/14/6
by other workers for Kir6.1, Kir6.2, SUR1 [4], SUR2A treatment and each time. The following formula was
and SUR2B [20]. PCR conditions were as follows: Kir6.1 usedtocalculatethedoublingtime:N =N xekxt,where
t 0
forward (5′-CATCTTTACCATGTCCTTCC-3′) and re- N was the initial number of cells that increased expo-
0
verse (5′-GTGAGCCTGAGCTGTTTTCA-3′), 336 bp; nentially with a rate constant, k. The doubling time (T)
Kir6.2 forward (5′-GCTTTGTGTCCAAGAAAGG1-3′) was calculated as: T=ln 2/k. The GraphPad Prism 5.0
and reverse (5′-CCAAAGCCAATAGTCACTTG-3′), software (GraphPad Software Inc., Philadelphia, U.S.A.)
301 bp; 5 min 95°C, 35 cycles of 95°C for 20 s, 52°C for wasemployed.
45 s, and 72°C for 1 min. SUR1 forward (5′-CGATGC
CATCATCACAGAAG-3′) and reverse (5′-CTGAG Cellcycleanalysisbyflowcytometry
CAGCTTCTCTGGCTT-3′), 291 bp; SUR2A foward Cells were cultured for 24 h without FBS. Synchro-
(5′-ATGCGGTTGTCACTGAAGG-3′) and reverse (5′- nized cells were then treated with IC Gli (25 μM) or
50
AATAGAAGAGACACGGTGAGC-3′), 215 bp; SUR2B vehicle immediately after release from the block and
forward (5′-GATGCGGTTGTCACTGAAGG-3′) and re- harvested for up to 72 h. Then cells were collected by
verse (5′-TCATCACAATAACCAGGTCTGC-3′), 244 bp; trypsinization, fixed with ice cold methanol, centri-
5 min 95°C, 35 cycles of 95°C for 1 min, 55°C for 1 min, fuged and resuspended in 0.5 ml of propidium iodide
and 72°C for 2 min. Reactions were terminated by final (PI) staining solution (50 μg/ml PI in PBS containing
elongation step of 7 min at 72°C (Gene Amp PCR System 0.2 mg/ml of DNase-free RNase A). After incubation
2400,PerkinElmer,MA,USA).Negativecontrolswereper- for 30 min at 37°C, samples were evaluated by flow
formed with water instead of cDNA. PCR products were cytometry (Becton Dickinson, USA). Cell cycle distri-
subjected to gel electrophoresis and detected by gel docu- bution was analyzed using Cylchred 1.0.2 software
mentation system LumiBis DNR (Bio-Imaging Systems, (Cardiff University, UK).
Jerusalem, Israel).
QuantificationofDNAsynthesis
Cellculture The quantification of cellular DNA synthesis was per-
MBA-MB-231 cels were cultured in RPMI1640 medium formed on cells by the addition of 30 μM BrdU for 2 h.
supplemented with 10% FBS, 0.3 g/l L-glutamine and Cells grown on sterile slides were washed with PBS and
40 mg/l gentamicine and in the presence of Gli,Tam or fixed with 10% formaldehyde in PBS. To denature the
vehicle. Cells were maintained at 37°C in a humidified DNA into single-stranded molecules, cells were incu-
atmosphere containing5% CO . bated with 3 N HCl for 30 min at room temperature.
2
Then cells were washed in 1 ml of 0.1 M Na B O ,
2 4 7
Cellproliferationassay pH 8.5 to neutralize the acid and were then incubated
For clonogenic assay, cells were seeded in 6-well plates overnight at 4°C with mouse anti-BrdU monoclonal anti-
(1.5×103 cells/well) and incubated in the presence or ab- body(1:50).Then,cellswereincubatedfor2hat37°Cwith
sence of drugs for 10 days. Cells were treated with 10 to FITC-conjugated anti-mouse IgG (1:100). After washed
50 μM Gli, 0.005 to 5 μM minoxidil; 0.1 to 5 μM tam- with PBS, cells were stained with DAPI (1:8000). Fluores-
oxifen; 0.01 to 0.1 nM doxorubicin or with the concen- cence was further visualized by fluorescence microscope.
tration of Gli that inhibited the proliferation to the 50% Atleast1000cellswerescoredforeachdetermination.
(IC ) plus different doses of minoxidil, tamoxifen or
50
doxorubicin to assess the combined action of both Apoptosis
drugs. Cells were then fixed with 10% formaldehyde in Apoptotic MDA-MB-231 cells were detected after treat-
phosphate-buffered saline, PBS, stained with 1% tolui- ment with Glior vehicle for 72h.Phosphatidylserineex-
dine blue in 70% ethanol. The clonogenic proliferation posure on the surface of apoptotic cells was detected by
was evaluated by counting the colonies with 50 cells or flow cytometry after staining with Annexin V-FITC and
more. The results are expressed as a percentage of con- PI (50 μg/ml). Data were analyzed using WinMDI 2.8
trolvalues. software(ScrippsInstitute,CA,USA).
Determinationofdoublingtime Mitochondrialtransmembranepotential
For doubling time determination cells were seeded in Variations of the mitochondrial transmembrane poten-
6-well plates (4×104 cells/well), starved for 24 h and tial of the cells, ΔΨ , were studied by means of the up-
m
then incubated with IC Gli (25 μM) or vehicle for take of DiOC6, a specific fluorochrome that has been
50
up to 96 h. Cells were trypsinized at 0, 1, 2, 3 and widely used in monitoring ΔΨ . Cells were plated and
m
4 days and counted using a hemocytometer. All treated 24 h after with 25 μM Gli for different incuba-
experiments were performed in the logarithmic phase tionperiods(24,48and72hs).Thediluted dyeatafinal
of cell growth. Triplicate plates were analyzed for each concentration of 40 nM in PBS was applied to cells for
Núñezetal.BMCPharmacologyandToxicology2013,14:6 Page4of13
http://www.biomedcentral.com/2050-6511/14/6
15 min at 37°C. Cells were then washed twice with PBS, optical microscopy. At least 1000 cells were scored for
harvested and then analyzed by flow cytometry (Becton eachdetermination.
Dickinson, USA). Results were expressed as the percent-
ageofmeanfluorescence ofrespective controls. Lipidaccumulation
InordertoexaminethepossibleactionofGliinlipidac-
Determinationofcellcycleandapoptosisrelatedproteins cumulation, a classic terminal differentiation marker in
Flowcytometry mammarycells,thelevelsofneutrallipidweremeasured
After treatment with Gli or vehicle for 72 h MDA-MB- by flow cytometry using Nile-red staining [23]. Cells
231 cells were harvested, fixed with 4% formaldehyde treated with Gli or vehicle for 2, 3 or 7 days were fixed
and permeabilized with saponine. To evaluate intracellu- and then incubated with Nile-red at a final concentra-
lar protein content, cells were incubated with rabbit tion of 1 μg/ml in PBS for 20 min at room temperature.
anti-Bcl-2 and rabbit anti-Bax antibodies. After washing, Cells were then analyzed by flow cytometry (Becton
cells were incubated for 20 min with FITC-labeled anti- Dickinson,USA).
rabbit. The samples were analyzed with a FACScalibur Data analysis was performed using WinMDI 2.8 soft-
flow cytometer (Becton Dickinson, USA). Data analysis ware (Scripps Institute, CA, USA). The results were
was performed using WinMDI 2.8 software. For each expressedasapercentageofcontrol values.
sample 20,000 events were collected. The results are
expressedasthepercentage ofrespective controlvalues. Cytostaticeffectofglibenclamide
MDA-MB-231 cells (1x104 cells/well) were plated into
Westernblotassay 6-well plates. After 24 h, cells were treated with 25 μM
Cells were placed on ice and washed twice with cold Gli or vehicle and detached with trypsin 3 or 7 days
PBS. Cellswerethenscrapedinto alysis buffer(100 mM later. Cells were then counted and seeded in 6-well
Tris/HCl buffer, pH 8, containing 1% Triton X-100 and plates (1.2×103 cells/well) in triplicates. After 10 days in
protease inhibitors) and incubated for 15 min on ice. culture, colonies were fixed with 10% buffered formalin
After centrifugation at 6000 rpm for 10 min, the super- and stained with 1% toluidine blue in 70% ethanol. The
natants were used for protein determination according number of colonies was determined and normalized to
to Bradford assay [21]. For Western blot, loading buffer thenumberofcoloniesincontrols.
(100 mM Tris/HCl buffer, pH 8,containing 1.7% sodium
dodecyl sulfate (SDS), 0.02% bromophenol blue, 1.5% Statisticalanalysis
dithiotreitol, and 5% of glycerol) was added to samples In all cases the data shown are the means±SEM of at
andtheywereboiledfor3minutes.Equalamountsofpro- least three independent experiments. Statistical analysis
teins(50μg)werefractionatedonSDS-polyacrylamidegels is indicated in each legend. Data were analyzed using the
(12%) and transferred electrophoretically onto nitrocellu- GraphPadPrism5.0(GraphPadSoftwareInc.,Philadelphia,
lose membranes. Membranes were blocked and probed U.S.A.)andPvalueslessthan0.05wereconsideredstatisti-
overnight with primary mouse anti-cyclin D (1:100), callysignificant.
1
mouse anti-cyclin E (1:500), mouse anti-cyclin B (1:200),
mouse anti-p27Kip1 (1:200), rabbit anti-Bax (1:500), rabbit Results
anti-Bcl-x (1:500) and mouse anti-β-actin (1:1000) anti- ExpressionofK channelsinMDA-MB-231cells
L/S ATP
bodies. Immunoreactivity was detected by using horserad- Though many reports describe the presence of different
ish peroxidase-conjugated anti-mouse or anti-rabbit IgG, potassium channels in diverse human cancer cell lines,
as appropriate, and visualized by enhanced chemilumines- at present there is little evidence about the expression of
cence. Densitometric analyses were performed using the K channels in breast cancer cells. Since Gli is a spe-
ATP
softwareImageJ1.32J(NIH,USA). cific blocker of K channels, the expression of mRNA
ATP
for different subunits (Kir6.1, Kir6.2 and SURs) was
Evaluationofsenescence-likephenotype examined by RT-PCR in MDA-MB-231 cells. As shown
Senescence-associated β-galactosidase (SA-β-Gal) activ- in Figure 1A bands of 336 and 301 bp for Kir6.1 and
itywasdetectedincellsaspreviouslydescribed byDimri Kir6.2 respectively were detected after electrophoresis,
[22] with some modifications. Cells treated with Gli indicating gene expression of pore components for at
(25 μM) or vehicle for 72 hs were fixed in 3.0% formal- least two channel types in this cell line. Furthermore, a
dehyde for5min,washedinphosphatebufferedsolution predicted band of 244 bp was found indicating the ex-
(PBS) and stained in 1 mg/ml X-gal solution at pH 6.0 pression of SUR2B gene. The expression of SUR1 and
for 4 hs at 37°C. SA-β-Gal positive cells were stained in SUR2A genes was not detected (291 bp and 215 bp, re-
blue. To visualize the cell architecture, the slides were spectively). Altogether these results indicate that pore-
counter-stained by haematoxylin and quantified by forming and regulatory subunits are expressed in this
Núñezetal.BMCPharmacologyandToxicology2013,14:6 Page5of13
http://www.biomedcentral.com/2050-6511/14/6
Figure1EffectofGlibenclamide,aspecificblockeroftheK
ATP
channels,oncellgrowth.ThefigureshowsthemRNAexpression
ofK channelscomponentsandtheeffectofGliandminoxidil
ATP
(Min)oncellproliferationinMDA-MB-231cells.Proliferationwas
evaluatedbycountingofcolonieswith50cellsormoreand
expressedaspercentageofvaluesobtainedwithvehicle(means±
SEMofthreeexperimentsonparallel).PanelA:mRNAexpressionof
K channelsinMDA-MB-231cellsbyRT-PCRanalysis.Agarosegel
ATP
electrophoresisofPCRproductsshowedbandscorrespondingto:
Kir6.1(336bp),Kir6.2(301bp),SUR2B(312bp).Nobandswere
detectedforSUR2A(215bp)orSUR1(291bp).CN:negativecontrol
PanelB:Inhibitionofproliferationobtainedwithdifferent
concentrationsofGli(10,20,30or50μM).Insertshowsthedose–
responsecurveusedtodetermineIC (IC =25μM).PanelC:
50 50
IncreaseofproliferationobtainedwithMin0.05;0.5or5μM.Panel
D:ResultsobtainedwithIC GliplusdifferentconcentrationofMin.
50
PanelBandC:*p<0.05vs.control;**p<0.01vscontrol;***p<
0.001vs.control,OnewayANOVAandDunnetposttest.PanelD:
***p<0.01vs.Min0.05μM;vs.Min0.5μM;vs.Min5μM.###p<
0.001vs.Min0.05μM;vs.Min0.5μM;vs.Min5μM.Oneway
ANOVAandTuckeyposttest.
cell line. ERα (+) MCF-7 breast cancer cells were also
analyzed and Kir6.1, Kir6.2 and SUR1 were found
expressedinthis cellline(data not shown).
Effectofglibenclamideoncellproliferation
The effect of Gli on cell proliferation was tested by
means of a clonogenic assay. A significant concentration
dependent inhibition on cell growth was observed when
Gli was added to cell cultures in concentrations over 10
μM; the IC value was 25.6±3.2 μM (Figure 1B). The
50
increased doubling time (T value, Table 1) obtained in
thepresenceof25μMGliisinconcordancewith thein-
hibition of proliferation previously demonstrated using
the clonogenicassay.
In order to support the hypothesis of K channels
ATP
involvement in MDA-MB-231 cell proliferation we used
minoxidil, a well known specific opener of these chan-
nels. The results showed an increase in cell clonogenic
growth for concentrations over 0.05 μM, which became
significant at 5 μM (Figure 1C). Figure 1D shows that
the increment in proliferation produced by the channel
opener wastotally reversed by25μMGli.
The analysis of cell cycle phase distribution demon-
strated that Gli produces a significant increase in the
number of cells in G1 phase at 24, 48 and 72 h post
treatment, clearly demonstratinga significant G0/G1 cell
cycle arrest (Figure 2A). A consequent decrease in cells
Table1Determinationofcelldoublingtime
Treatment DuplicationTime(hs)
Control 24.0±4.3
Gli 34.6±4.5*
Gli:25μMgliblenclamide.*p<0.05vscontrol,ttest.
Núñezetal.BMCPharmacologyandToxicology2013,14:6 Page6of13
http://www.biomedcentral.com/2050-6511/14/6
Figure2(Seelegendonnextpage.)
Núñezetal.BMCPharmacologyandToxicology2013,14:6 Page7of13
http://www.biomedcentral.com/2050-6511/14/6
(Seefigureonpreviouspage.)
Figure2EffectofGlibenclamideoncellcycleprogression.PanelA:SynchronizedMDA-MB-231cellsweretreatedwithIC Gli(25μM)or
50
vehiclefor24,48or72handthefractionofcellsineachphaseofcellcyclewasevaluatedbyflowcitometry.Glitreatmentclearlyarrestedcells
atG0/G1phase.Resultsareexpressedaspercentageofthevalueobtainedwithvehicle(means±SEMofthreeexperimentsonparallel).°p<0.01
vscontrol;*p<0.001vscontrol,ttest.Leftbars:control;rightbars:Gli-treatedcells.PanelB:AdecreaseinBrdUincorporationtoDNAwas
observedwhencellsweretreatedwith25μMGlifor48h.Resultsareexpressedasthemeans±SEMofthreeexperimentsonparallel.*p<0.05
vs.control,ttest.PanelC:ExpressionofG1-SregulatoryproteinsinMDA-MB-231cellstreatedwithGliorvehiclefor72hwasanalyzedby
Westernblot.GlidecreasedthelevelofcyclinEandincreasedp27Kip1.RepresentativeimmunoblotimagesofcyclinsD1,B1,Eandp27Kip1are
illustrated.Relativequantificationwasperformedbydensitometricanalyses.Actindensitometricvalueswereusedtostandardizeforprotein
loading.Barsrepresentthemean±SEMofthreeindependentexperiments.**p<0.01vscontrol;***p<0.001vscontrol,ttest.
in S and G2 phase versus control was also observed. mitochondrial transmembrane potential (ΔΨ ) that is
m
Consistent with these observations, Gli inhibited the ac- associated with mitochondrial dysfunction and linked to
tive DNA synthesis when it was evaluated by BrdU in- celldeathandlossofcellviability(Table2).
corporation(Figure 2B). It is known that the Bcl-2 family of mitochondrial pro-
Theexpression of proteins implicated in the control of teins is strongly linked to the process of apoptotic cell
different phases of the cell cycle was investigated by death; some members of the family act as antiapoptotic
Western blot analysis. Studies of proteins specifically proteinssuchasBcl-2,Bcl-x ,whileothersactasinductors
L
related with phase G1 of cell cycle demonstrated that 25 ofcelldeathasBcl-x andBax[24-26].Wedeterminedthe
S
μM Gli reduced expression of cyclin E whereas cyclin expression of these proteins by flow cytometry and West-
D1 remained unchanged after 72 h of treatment. Fur- ernblot.Bybothmethodologiesweshowedthat72hafter
thermore, p27Kip1 levels were up-regulated in the same treatment with 25 μM Gli, the level of expression of pro-
experimental conditions. In addition, the level of cyclin apoptotic protein Bax was slightly increased in relation to
B1 expression, which is involved in the control of G2-M control at the same time, although this increase was not
transition,wasnot modified byGlitreatment. statistically significant (Figure 4A and 4B). On the other
hand, antiapoptotic Bcl-2 protein did not modify its ex-
Effectofglibenclamideoncelldeath pressionwhen cells weretreated withGli (Figure4B). The
To determine if the decrease in proliferation exerted by pro-apoptotic isoform, Bcl-xS, showed a very low expres-
Gli could be due to an apoptotic effect, we assessed sion while the antiapoptotic isoform Bcl-xL expression
apoptosis by two different methodologies. Results levels were higher but did not significantly change with
showedthatGlididnotincreasethenumberofapoptotic Gli-treatment(Figure4A).
cells by Annexin-V staining (3.66±0.62% in control vs We also evaluated the induction of cell senescence as
3.70±0.69%) after 72 h of treatment (Figure 3). In ac- a mechanism of cell death. To identify the senescent
cordance, neither it produced the disruption of the cells, senescence-associated β-galactosidase (SA-β-GAL)
Figure3EvaluationofapoptosisbyAnnexin-Vmethod.ApoptosiswasassessedafterincubatingthecellswithGliorvehicleby72h.For
positivecontrolcellsweretreatedwithHO (5mM)for30minutes.FluorescencewasevaluatedimmediatelyafterAnnexin-Vstainingbyflow
2 2
cytometry.Gli(25μM)didnotinduceanincrementinapoptosisofMDA-MB-231cells.PositiveAnnexin-Vcellsareshowninbothrightquadrants.
Núñezetal.BMCPharmacologyandToxicology2013,14:6 Page8of13
http://www.biomedcentral.com/2050-6511/14/6
Table2EffectofGlionmitochondrialtransmembrane moreeffectivetoinhibitcellproliferationthansingletreat-
potential(ΔΨm) ments(Figure7D).
Time(hs) ΔΨm(%ofcontrol)
24 99.7±7.3 Discussion
Different subtypes of potassium channels have been
48 97.0±3.9
showntobedirectlyimplicatedinnormalandmalignant
72 94.3±12.7
cell proliferation [9]. Some of these channels are overex-
Gli:25μMglibenclamide.p:NSvs.control,OnewayANOVA.
pressed in tumors and therefore they are potential tar-
getsforanticancertherapies[28].
We have previously demonstrated that Gli exerts an
activity was assessed. MDA-MB-231 cells treated with
25 μM Gli showed an increase in the percentage of sen- antitumoral action on NMU-induced mammary tumors
inrats,whichisanexperimentalmodelsimilarinERex-
escent cells versus those treated with vehicle (3.5±0.4%
pression and hormone-dependence to human breast
vs1.2±0.2%;Figure 5).
cancer [29,30]. In these tumors Gli action was poten-
tiated by the combination with tamoxifen [31]. In this
Effectofglibenclamideoncelldifferentiation paper we analyzed the effect of Gli in MDA-MB-231
Differentiation is one possible mechanism involved in ERα(−)breastcancercellproliferation.
the loss of cell proliferative ability. In mammary cells We studied the expression of mRNA for the different
the accumulation of neutral lipids in cytoplasm is a subunits that constitute K channels in MDA-MB-231
ATP
specific marker of this process. We evaluated by flow cells and determined that these cells express both pore
cytometry the content of neutral lipids using Nile-red forming subunits, Kir6.1 and Kir6.2, and the regulatory
staining and results demonstrated no differences be- subunit SUR2B. Coincidently, Bondestine et al. have re-
tween control and 25 μM Gli treated cells up to 7 days cently reported the expression of SUR2 protein in
(Table 3). Sodium butyrate, an effective differentiation MDA-MB-231cells,while SUR1could notbedetected in
agent in ERα (+) and ERα (−) breast cancer cells [27], this cell line [32]. In consequence two whole octameric
was used as positive control. functional channels could be present in MDA-MB-231
cells. In specific tissues different subunit combinations
have been detected, e.g., pancreatic β cells and neurons
Cytostaticeffectofglibenclamide
have Kir6.2/SUR1 channels, whereas skeletal muscle
To determine whether the growth inhibitory effect
expresses Kir6.2/SUR2A channels. These differences
was reversible, a characteristic of cytostatic agents,
cells were treated with 25 μM Gli or vehicle for 3 or imply that drugs may have distinct abilities to affect
K channels in diverse tissues depending on the type
7 days and then they were trypsinized and re-plated ATP
of SUR expressed [20]. Gli binds to every type of SUR
at low density in the absence of any treatment to as-
and therefore inhibits the activity of all known K
sess clonogenic proliferation. Results in Figure 6 ATP
channels [33]. When we assayed the effect of Gli on
showed that there is no significant difference between
MDA-MB-231 cell proliferation, results showed that Gli
Gli-pretreated cells for different time periods and
inhibits clonogenic ability in a concentration-dependent
control cells, suggesting that the antiproliferative ef-
way, with an increment in population doubling time. It
fect of Gli is elicited only when the drug is present
has been reported that the potassium-dependent
and it does not involve cell toxicity.
changes in membrane potential play a crucial role in the
proliferation of many types of normal and tumor cells.
Combinationtreatments The opening of potassium channels in the cell mem-
We assayed the combination of Gli with tamoxifen or brane produces a hyperpolarization of membrane poten-
doxorubicin to explore a possible increase in efficacy. The tial which is required for the progression through the
growth of MDA-MB-231 cells was inhibited by tamoxifen cell cycle. As a consequence in the presence of potas-
withanIC equalto5μM(Figure7A).Theconcentration sium channel blockers cell proliferation is inhibited
50
ofGlithatinhibitedcellgrowthina50%(25μM)wasused [34,35]. Woodfort et al. early determined that different
toevaluatethecombinedactionofGliplustamoxifen.The potassium channel antagonists, as Gli, produce a
inhibitory effect exerted by the combination of Gli plus concentration-dependent growth inhibition on MCF-7
tamoxifen was similar to that observed for Gli alone cells with a significant arrest of cells in G0/G1 phase
(Figure7B). [36]. Diverse drugs signaled as specific K channel
ATP
Doxorubicinalsoinhibitedcellproliferationinaconcen- openerswhich include minoxidil, pinacidiland diaxozide
tration dependent manner (Figure 7C). The combination augment DNA synthesis and proliferation of normal and
of 25 μM Gli plus doxorubicin in doses over 0.05 nM was tumor cells [37-39]. In the experiments using minoxidil
Núñezetal.BMCPharmacologyandToxicology2013,14:6 Page9of13
http://www.biomedcentral.com/2050-6511/14/6
Figure4Expressionofproteinsinvolvedinapoptosis.PanelA:
BaxandBcl-x expressiondeterminedbyWesternBlotemploying
L/S
specificantibodiesinMDA-MB-231cellstreatedfor72hwith25μM
Gli(Gli)orvehicle(control)cells.Thefigureshowsarepresentative
Westernblotofthreeindependentexperimentsandthe
quantificationofbandsobtainedforBaxandBcl-x protein.Bars
L/S
representthemean±SEMofthreeindependentexperiments.p:
Nonsignificant(NS),ttest.PanelB:ExpressionofBcl-2andBax
proteininMDA-MB-231cellstreatedby72hwith25μMGli(Gli)or
vehicle, obtained by flow cytometry. Bars represent the mean
fluorescence±SEM obtained by three independent experiments.
p: NS, t test.
we demonstrated an increase in MDA-MB-231 cell prolif-
eration in concentrations over 0.05 μM. In addition when
different concentrations of minoxidil were combined with
25 μM Gli the increment in proliferation was totally
reverted. Taken together, these results suggest that Gli
could reduce cell proliferation in MDA-MB-231 cells act-
ingthroughK channels.
ATP
The analysis of cell cycle progression indicated that
25 μM Gli produces an arrest in the G0/G1 phase of cell
cycleafter48hoftreatment.Afterreleaseofserumstarva-
tion,thediminutionoftheproportionofcellsinSphasein
GlitreatedcultureswasconfirmedbythedecreaseinBrdU
incorporation. The cell division cycle integrates several
processes and signal transduction pathways to commit the
progression of a cell through or its arrest in a specific cell
cyclephase.Itisgenerallyacceptedthatthecellcycleregu-
latorscyclinD1andcyclinEplayanimportantroleinearly
G1andlateG1progression.Inaddition,theprogressionis
tightlyregulatedbytherespectivecyclinactivatedsubunits
cyclin-dependent kinases (Cdks) and their inhibitors
(Cdkis). p27 is a member of the Kip/CIP family of Cdkis
known to act in the G1phase of the cell cycle, preventing
G1-S transition [40,41]. The study of proteins involved in
the progression through the phases of cell cycle by immu-
noblot, has evidenced a decrease in cyclin E expression
witharaiseincyclininhibitorp27Kip1inGlitreatedMDA-
MB-231 cells. It is known that the activity of cyclin E in
conjunction with its kinase subunit Cdk2, is limiting for
the passage of cells through the restriction point needed
for the progression of cells from G1 into S-phase [42].
p27Kip1 binds to the cyclinE/Cdk2 complex and inhibits
the kinase thus impeding G1-S passage [43]. Altogether
our results suggest that Gli inhibits MDA-MB-231 cell
proliferation by hindering G1-S transition. Other authors
havealsoreportedthatdifferentpotassiumchannelsblock-
ers and other agents that cause cell membrane
depolarization,producethearrestofcellsinG1phasewith
theinvolvementofdifferentcyclinsandinhibitorsdepend-
ing on the cell type [34]. Eto reported that various anti-
cancer agents specifically up-regulate p27Kip1 expression
without affecting expression of the other regulatory pro-
teins of G1-S cell cycle transition in human breast cancer
Núñezetal.BMCPharmacologyandToxicology2013,14:6 Page10of13
http://www.biomedcentral.com/2050-6511/14/6
Figure6Evaluationofcytotoxicorcytostaticeffect.MDA-MB-231
cellstreatedwith25μMGliorvehiclefor3or7dayswerere-seededat
lowdensitytoevaluateclonogeniccapacity.After10daysinculture,the
numberofcolonieswasdeterminedandnormalizedtothenumberof
coloniesincontrols.BarsshowthatGlipre-treatmentdidnot
signifcantlyaffectclonogeniccapacity.p:NS,OnewayANOVA.
Figure5EvaluationofSenescence.MDA-MB-231cellswere
culturedfor48hwith25μMGliorwithvehicle.Senescencewas
assessedbytheactivityofSA-β-GAL.PanelA:representative
tumors in rats,we reported that Gliclearly produced the
photographswherepositiveSA-β-GALcellsareindicatedbyarrows.
inhibition of tumor growth through a decrease in cell
Gliproducesanincrementincellsenescence.PanelB:Percentageof
cellsSA-β-GALpositivewerecalculatedbycountingofatleast1000 proliferation and an increase in cell apoptosis and differ-
cells(630X).Barsrepresentthemean±SEMofthreeindependent entiation [31]. It has been reported that Gli produces
experiments.*p<0.001vs.control,ttest. apoptosis in malignant cell lines such as hepatoblastoma
and gastric cancer cells [6,7]. Furthermore, Iwakura and
coworkers found that Gli produces a sustained increase
cell lines. Moreover, in concordance with our work, he in the entrance of Ca2+ to the cells inducing their death
reported that the up-regulation of p27Kip1 expression in through apoptosis [46]. On the contrary, we demon-
thesecelllinesbyanti-canceragentslinearlyandpositively strated that Gli neither produced apoptosis nor triggered
correlates with the degree of growth inhibition of NMU- the early events of this mechanism of cell death in
induced mammary tumors by the same anti-cancer agent MDA-MB-231 cells. We also studied the expression of
[44]. apoptosisrelated proteinsand determined thatGlididnot
It has been suggested that chemotherapeutic agents significantly affect the ratio of the expression of Bax/Bcl-2
can prevent mammary carcinogenesis and tumor growth proteinsatanytimeevaluated.Therelationbetweenantia-
through different mechanisms that include apoptosis, poptotic and apoptotic proteins of Bcl-2 family is a better
differentiation and senescence. Regardless of the mech- determinant of the susceptibility to apoptosis than the ex-
anism, the evaluation of new antitumoral drugs has as a pressionofeachmemberseparately.
goal to stop the cell proliferation and produce the death In view that some chemotherapeutic agents are able to
of the tumor cell by apoptotic or non-apoptotic means produce cell senescence as part of their mechanism of
[45]. Previously, working with NMU-induced mammary action [45], this possible way of action was studied in
the MDA-MB-231 cells treated with Gli. Data obtained
inourexperimentsindicatethatthereisa slight increase
Table3EffectofGlionneutrallipidaccumulation in cell senescence after Gli treatment. However, Gli did
Time(days) MeanFluorescenceofNileRed(%ofcontrol) not result cytotoxic for neoplastic MDA-MB-231cellsso
Gli NaButyrate they clearly keep their clonogenic capacity after being
2 92.3±4.7 134.3±6.7*** exposedtoGliforsevendays.Ourresultsareinagreement
with the reported by Woodfork et al., showing that Gli
3 99.3±1.7 165.7±7.3***
induced a cell cycle arrest in G0/G1phase in MCF-7 cells
7 99.0±10.1 192.7±5.3***
thatcouldberevertedbytheremovalofthedrug[36].
Gli:25μMglibenclamide.NaButyrate:10mMSodiumbutyrateasthepositive
Mammary tumor cell differentiation is characterized by
control.p:NSGlivscontrol;***p<0.001butyratevscontrol,OnewayANOVA
andDunnettest. an arrest in cell proliferation, nuclear and cytoplasmatic