Biochimica et Biophysica Acta 1813 (2011) 1438–1445

Contents lists available at ScienceDirect

Biochimica et Biophysica Acta

j ourna l homepage: www.e lsev ie r.com/ locate /bbamcr

Arginine vasopressin controls p27Kip1 protein expression by PKC activation andirreversibly inhibits the proliferation of K-Ras-dependent mouse Y1 adrenocorticalmalignant cells

Fabio L. Forti a,⁎, Hugo A. Armelin a,b

a Departamento de Bioquimica, Instituto de Quimica, Universidade de Sao Paulo, Av. Prof. Lineu Prestes, 748-Bl 09 Inf., Sl 922, Cidade Universitaria, Cep 05508-900, Sao Paulo-SP, Brazilb Instituto Butantan, CATcepid, Av. Vital Brasil, 1500, CEP 05503-900, São Paulo, SP, Brazil

⁎ Corresponding author at: Av. Prof. Lineu Prestes,05508-000, São Paulo-SP, Brazil. Tel.: +55 11 3091 217

E-mail address: flforti@iq.usp.br (F.L. Forti).

0167-4889/$ – see front matter © 2011 Elsevier B.V. Adoi:10.1016/j.bbamcr.2011.04.007

a b s t r a c t

a r t i c l e i n f o

Article history:Received 25 October 2010Received in revised form 4 March 2011Accepted 27 April 2011Available online 5 May 2011

Keywords:Arginine vasopressinp27Kip1

Mouse Y1 adrenocortical cellsK-RasProtein kinase C

The neurohypophyseal hormone arginine vasopressin (AVP) is a classic mitogen in many cells. In K-Ras-dependent mouse Y1 adrenocortical malignant cells, AVP elicits antagonistic responses such as the activationof the PKC and the ERK1/2 mitogenic pathways to down-regulate cyclin D1 gene expression, which inducessenescence-associated β-galactosidase (SA-βGal) and leads to cell cycle arrest. Here, we report that in themetabolic background of Y1 cells, PKC activation either by AVP or by PMA inhibits the PI3K/Akt pathway andstabilises the p27Kip1 protein even in the presence of the mitogen fibroblast growth factor 2 (FGF2). Theseresults suggest that p27Kip1 is a critical signalling node in the mechanisms underlying the survival of the Y1cells. In Y1 cells that transiently express wild-type p27Kip1, AVP caused a severe reduction in cell survival, asshown by clonogenic assays. However, AVP promoted the survival of Y1 cells transiently expressing mutantp27-S10A or mutant p27-T187A, which cannot be phosphorylated at Ser10 and Thr187, respectively. Inaddition, PKC activation by PMA mimics the toxic effect caused by AVP in Y1 cells, and inhibition of PKCcompletely abolishes the effects caused by both PMA and AVP in clonogenic assays. The vulnerability of Y1cells during PKC activation is a phenotype conditioned upon K-ras oncogene amplification because K-Rasdown-regulation with an inducible form of the dominant-negative mutant H-RasN17 has resulted in Y1 cellsthat are resistant to AVP's deleterious effects. These data show that the survival destabilisation of K-Ras-dependent Y1 malignant cells by AVP requires large quantities of the p27Kip1 protein as well asphosphorylation of the p27Kip1 protein at both Ser10 and Thr187.

748-Bloco 09i, Sala 922, Cep2; fax: +55 11 3091 2186.

ll rights reserved.

© 2011 Elsevier B.V. All rights reserved.

1. Introduction

Arginine vasopressin (AVP) is a vasoactive, antidiuretic, neuroen-docrine peptide hormone synthesised in the supraoptic (SON) andparaventricular (PVN) nuclei of the hypothalamus [1,2]. The effectscaused by AVP are mediated by three G protein-coupled receptors(GPCR): V1aR, which is found in the brain, liver, peripheral tissue andvessel smooth muscle [3]; V1bR, which is found in pituitarycorticotrophs and the brain [4]; and V2R, which is found in renalcollecting ducts [5]. In addition to its classical endocrine functions(e.g., regulation of antidiuresis, vasoconstriction and stress response),AVP also regulates mitogenesis, cell survival and death [6,7]. Evidenceof the mitogenic activity of AVP comes mainly from studies of eitherprimary cultures or cell lines derived from peripheral tissues thatexpress V1aR receptors [8]. However, a recent report [9] shows thatelevated AVP levels in rats caused the proliferation of kidney tubularcells via V2R and ERK phosphorylation while V1aR-expressing cells

from other tissues were not affected. On the other hand, other studieshave quantitatively analysed the phosphoproteome of cultured V2R-expressing renal collecting duct cells and found that AVP stimulationled to a reduction in the phosphorylation of mitogenic kinases, such asERK and cyclin-dependent kinases [10]. Thus, the roles of both AVPand its receptors V1aR and V2R in mitogenic signalling are still notcompletely understood and deserve more investigation from differentpoint of view.

Over the last decade, studies have found that AVP controls thegrowth and proliferation of lung, adrenal, pancreatic, breast cancercells and normal cardiac cells [11–15]. In V1aR-expressing mouse Y1adrenal malignant cells, we previously reported that AVP elicits twoantagonistic responses: first, AVP down-regulates cyclin D1 mRNAand protein, which leads to cell cycle arrest; second, AVP activatesPI3K, PKC and ERK1/2 mitogenic pathways [16]. In addition, wedemonstrated that the Y1 cell's anti-proliferative response to AVP ismediated by RhoA GTPase, which causes the down-regulation ofcyclin D1 and induction of senescence-associated β-galactosidase[14].

Here, we report that activation of PKC with PMA mimics the anti-proliferative effect of AVP in Y1 adrenal cells, which is conditioned on

1439F.L. Forti, H.A. Armelin / Biochimica et Biophysica Acta 1813 (2011) 1438–1445

the over-expression of the K-ras oncogene. According to themetabolicbackground of Y1 cells, activation of PKC seems to inhibit the PI3K/Aktpathway, leading to the stabilisation of p27Kip1 irrespective of a strongmitogen presence, and inhibitors of PKC completely protect cells fromboth PMA and AVP's toxic effects. Thus, p27Kip1 is the target ofmitogenic action in the mechanisms underlying the survival of the K-Ras-dependent Y1 malignant cell. In addition, the effects of AVP onp27Kip1 signalling are more complex than the effects of PMA onp27Kip1 signalling; does not act solely on p27Kip1 protein stabilisationmediated by PKC activation. AVP was able to differentially regulateendogenous levels of p27Kip1 phosphorylation at Ser10 or Thr187 andalso increase the survival rate of Y1 cells that transiently expresseither of the mutants p27-S10A or p27-T187A. Each of these mutantsimpaired nuclear translocation and degradation of p27 due to theirinability to be phosphorylated at Ser10 and Thr187.

2. Materials and methods

2.1. Chemicals

Synthetic arginine vasopressin (AVP) peptide was purchased fromSigma. Geneticin (G418), bisindolylmalimide (GF109203X), Go6983,H-89, LY294002, dexamethasone, DMSO, lipofectamine, buffers,protein ladders and other chemicals were obtained from Calbiochem,Invitrogen and Sigma. Recombinant bovine FGF2 was prepared in thelaboratory of Dr. Angelo Gambarini, Instituto de Química (USP, SãoPaulo).

2.2. Cell culture

Stocks from the parental mouse Y1 adrenocortical tumour cell line[17] and Y1-transfectant clonal sub-lines carrying the dominant-negative mutant H-rasN17 (Y1-RasN17 conditional clones) [18] weregrown in 10% FCS-DMEwithout or with 100 μg/mL G418, respectively.The H-RasN17 mutant protein was induced in Y1-RasN17 conditionalclones with 0.5 μM dexamethasone, added 12 h before starting theexperiment, as previously described [18]. To arrest the cell cycle at theG0/G1 boundary, exponentially growing cells were incubated for 48 hin serum-free medium (SFM).

2.3. Phosphorylation analysis of ERK1/2 and AKT proteins

Cells were lysed in cold 62.5 mM Tris–HCl pH 6.8, 2% w/v SDS, 10%glycerol, 50 mM DTT and 1% w/v bromophenol blue. Lysates weresonicated for 2 min, boiled for 5 min and clarified by centrifugation(14000 rpm, 5 min, 4 °C). Then, 150 μg protein aliquots were loadedon 10% SDS-PAGE gels. After electroblotting onto Hybond-C nitrocel-lulose membranes using a semi-dry Bio-Rad apparatus, total ERK1/2or phospho-ERK1/2 (Thr202/Tyr204) and total AKT or phospho-AKT(Ser473) were detected with polyclonal specific rabbit antibodies(Cell Signalling), followed by a secondary peroxidase-conjugatedanti-rabbit polyclonal antibody for chemiluminescent detection (ECL,Amersham-Pharmacia). Ponceau staining of the electroblotted mem-brane was used to visually confirm the efficiency of protein transfer.

2.4. Levels of p27Kip1 protein

Cells were lysed in 20 mM Tris–HCl pH 8.0, 135 mM NaCl, 10%glycerol, 1% Nonidet P-40, 1 mM sodium orthovanadate, 2 μg/mLleupeptin, 2 μg/mL aprotinin, and 2 μg/mL pepstatin. After quantita-tion by the Bradford assay, aliquots of 150 μg proteinweremixedwithSDS-PAGE sample buffer, loaded onto 10% SDS-PAGE gels andprocessed for Western blotting for the AKT/PKB protein using a rabbitpolyclonal antibody monospecific for mouse p27Kip1 as well asphosphorylated p27Kip1–Ser10 and p27Kip1–Thr187 (Santa Cruz).

Ponceau staining of the electroblotted membrane was used to visuallyconfirm an efficient protein transfer.

2.5. Clonogenic assays

Parental Y1 cells and Y1 transfectant clonal lines were plated in60 mm plates (1–5×103 cells/plate) containing 10% FCS/DME (withor without 100 μM G418, as required) and maintained in cultureovernight. Afterwards, the cells were submitted to AVP, PMA and/orFGF2 stimulation with or without 1 h of pre-treatment with specificPKC inhibitors. Triplicate plates from each cell line were then washed2× with PBS, 1× with DMEM, reefed with 10% FCS/DME medium andreturned to the incubator for 10–15 days or until colonies werevisualised. The medium was changed every 2–3 days and cells werefixed by PBS/10% formaldehyde solution at room temperature for15 min and then washed 3× with PBS. The cell colonies were stainedwith a 0.5% crystal violet solution for 4 h and then washed instreaming water. Results are presented as the average of threeindependent experiments.

2.6. Transient transfection of Y1 cells with p27Kip1 constructs

A total of 104 Y1 cells were seeded in 12 multiwell plates (~75%confluence) containing DME/10% FCS and incubated at 37 °C/5% CO2.Cells were transfected 24 h later with 10 μg of plasmid DNAharbouring the p27Kip1 wild-type, constructs S10A or T187A (kindlydonated by Dr. Christopher McAndrew from Prof. Dr. DanielDonoghue's laboratory, Department of Chemistry and Biochemistry,UCSD, San Diego, CA) added to 500 μL of DME medium withoutantibiotics or serum. Then, 100 μL of DMEmediumwithout antibioticsor serum plus 10 μL of Lipofectin was prepared. Both solutions wereleft at room temperature for 30–45 min. They were then mixedtogether and incubated for an additional 15 min. The plasmid plusLipofectin mixture was added to previously washed cell platesdropwise and placed in an incubator at 37 °C/5% CO2 for 6 h for celltransfection. DME/10% FCS was added to the medium and plates wereincubated for another 24 h. Afterwards, the medium in the plates waschanged to fresh DME/10% FCS plus 10−9 M AVP (final concentration)and incubated for another period of 24 h. Plates were then washedthree times by PBS, refed with DME/10%FCS and changed every 3 daysuntil visualisation of countable colonies. The surviving colonies werethen treated and counted according to the previous protocol for theclonogenic assay.

2.7. Statistics

The number of colonies per plate found in independent experi-ments was pooled and statistically analysed by the Chi square (χ2)test with pb0.05. Pairing of the AVP, PMA, and/or FGF2 treated/untreated samples or clonal/parental cell lines was used to assessstatistical significance.

3. Results

3.1. AVP or PMA irreversibly inhibits the proliferation ofK-Ras-dependent Y1 malignant adrenocortical cells by aprocess completely dependent on PKC activity

Clonogenic assays showed that 24 h treatments of either AVP or itsmimicking agent PMA, a cell-permeable and direct activator of PKC,strongly inhibited the development of colonies of mouse Y1adrenocortical cells growing in DME containing 10% FCS (Fig. 1).The specific PKC inhibitor Go6983 or bisindolylmaleimide GF109203X(not shown) completely protects Y1 cells from AVP and PMA (Fig. 1),implying that this inhibition of proliferation is entirely dependent onthe PKC pathway in both cases.

0

200

400

600

800

1000

1200

N I (Go6983) AVP I+AVP PMA PMA+I

CO

LO

NIE

S /

PL

AT

E*

**

****

Fig. 1. Clonogenic assays showing thatAVPandPMAblockcolony formationofY1cells, butpre-treatment with PKC inhibitors blocks these inhibitory effects. Parental Y1 cells wereplated (104 cells/plate of 60 mm diameter) and were pre-treated or not 24 h later with10 μM Go6983 and finally with 1 nM AVP or PMA 10 ng/mL for 24 h. The medium waschanged to fresh 10% FCS/DME every third day. Results of colonies/plate from threeindependent experiments were pooled and statistically analysed by χ2 with 1 degree offreedom, pb0.005. Thepanel histogramdisplays averagesof colonies/platewith error bars.

1440 F.L. Forti, H.A. Armelin / Biochimica et Biophysica Acta 1813 (2011) 1438–1445

The mouse Y1 adrenocortical cell line displays tight control of G0/G1→S transition of cell cycle and is a relevant and useful model forstudying cell cycle regulation by AVP because it only expresses theV1a receptor as measured by quantitative RT-PCR (P. Asprino and H.Armelin, unpublished results). However, we previously reported thatthe Y1 cell line exhibits a malignant phenotype dependent onchronically elevated levels of K-Ras-GTP, which cause constitutiveactivation of the PI3K/AKT pathway [18–20]. These high levels of K-Ras-GTP also make Y1 cells prone to inhibition of proliferation by AVPand PMA, a conclusion supported by the results of Fig. 2. A stable Y1transfectant sub-line carrying the dominant-negative mutant H-rasN17 under the control of the mouse mammary tumour virus [18]was previously treated with 0.5 μMdexamethasone for 12 h to inducethe H-rasN17 transgene, causing a reduction in endogenous levels ofK-Ras-GTP. This reduction, in turn, rendered Y1-RasN17 cells resistantto inhibition of proliferation by both AVP and PMA (Fig. 2), whileparental Y1 cells were not affected by dexamethasone treatment (notshown).

To gain insight into the molecular mechanisms underlying thesephenomena, we probed signalling pathways involving PKC, AKT andERK by Western blot analysis. AVP stimulated ERK1/2 phosphoryla-

0

100

200

300

400

500

600

700

800

900

1000

1100

Y1 Cells Y1-RasN17 3.1 Y1-RasN17 3.1 +Dexamethasone

Co

lon

ies

/ Pla

te

NTPMA 10ng/mlAVP 1nM

** ******

**

Fig. 2. Clonogenic assays showing that AVP and PMA inhibit colony formation in Y1 cellsand Y1-RasN17 mutant clones, but not in Y1-RasN17 induced by dexamethasone. Y1cells and clones were plated (104 cells/plate of 60 mm diameter) and were pre-treated(Y1-RasN17 clone 3.1) or not with 0.5 μM dexamethasone for 24 h and finally with1 nM AVP or 10 ng/mL PMA for a subsequent 24 h. The medium was changed to fresh10% FCS/DME every third day. Results of colonies/plate from three independentexperiments were pooled and statistically analysed by χ2 with 1 degree of freedom;*, pb0.001; **, pb0.005. The panel histogram displays averages of colonies/plate witherror bars.

tion, which peaked at 5 min (Fig. 3A) and began decreasing at 10 min(Fig. S1A and B); AVP stimulation of ERK1/2 was partially reduced byPKC inhibitors (Fig. 3A). PMA (10 ng/mL) also stimulated phosphor-ylation of ERK1/2, but as expected, PKC inhibitors completely blockedthis PMA activity (Fig. S1A and B). Thus, although activation of PKCwas sufficient for promoting ERK1/2 phosphorylation, both AVP andFGF2 also stimulated ERK1/2 phosphorylation by pathways that wereindependent of PKC. It is worth to mention that all effects of FGF2 inY1 cells are completely abolished by the FGFR tyrosine-kinase specificinhibitor, PD173074 [Costa et al. (2008) Cancer Res], and are verylikely mediated by the FGFR1IIIc and FGFR2IIIc receptors (detected inthese cells by both quantitative RT-PCR and sequencing analyses ofRT-PCR-amplified fragments; J. Salotti and H.A. Armelin, unpublishedresults).

However, the interactions between AVP, PMA and FGF2 regulatethe levels of phosphorylated AKT in a complex manner. We havepreviously shown that FGF2 up-regulates the basal levels of phospho-AKT in Y1 cells [20]. This up-regulation effect of FGF2 is inhibited byPMA (Fig. S1B). Again, as expected, PMA inhibition of the FGF2 up-regulation of phospho-AKT was completely abolished by PKC in-hibitors (Fig. S1B) without interfering with basal levels of phospho-AKT. It is worth emphasising that the specific inhibitor PI3K,LY294002, totally eliminated bands of phospho-AKT (Fig. 3A),implying that phospho-AKT is under a high and constant rate ofdephosphorylation and suggesting that steady levels of phospho-AKTdepend on a balance between PI3K activity and a phosphatase activity.Thus, FGF2 very likely up-regulates phospho-AKT steady state levelsby increasing PI3K activity, a step that is antagonized by PKCactivation by PMA (Fig. S1B). On the other hand, AVP exerts a dualregulatory role on steady state levels of phospho-AKT. AVP is a stronginhibitor of phospho-AKT up-regulation by FGF2 (Fig. S2B), mimick-ing PMA. Intriguingly, this inhibition of the up-regulatory effect ofFGF2 is stronger at lower concentrations of AVP, i.e., 10−9 to 10−10 M(Fig. S2C), suggesting a specificity very likely mediated by V1Areceptors. Coherently with this interpretation, the antagonism of AVPover the up-regulatory effect of FGF2 is independent of PKA activity(Fig. 3B), a response expected from cells that do not express AVP-V1Breceptors. However, in the presence of PKC inhibitors, AVP resembledFGF2 up-regulating phospho-AKT steady state levels (Fig. 3A), aneffect additive to that of FGF2 (Fig. S2A). Altogether, these resultssuggest that AVP initiates two parallel antagonistic pathways: first,AVP activates PKC, which inhibits PI3K/Akt; second, AVP directlystimulates PI3K/Akt.

3.2. AVP both increases stability in and promotes phosphorylationof p27Kip1 in G0/G1-arrested Y1 cells stimulated by FGF2

We analysed p27Kip1 expression in Y1 cells to further investigatethemolecular basis of the antagonism between AVP andmitogens likeserum and FGF2. We previously reported that G0/G1-arrested Y1 cellsdisplayed high basal levels of p27Kip1, which were down-regulated bytreatment with serum or FGF2 [18]. Here, we show that AVPmaintained p27Kip1 at high levels for 24 h irrespective of the presenceof FGF2 (Fig. 4A), and this antagonistic effect of AVP was mimicked byPMA (Fig. 4B). In addition, inhibitors of PKC completely blocked thestabilising effect that PMA has on p27Kip1 levels but did not block theeffect that AVP has on p27Kip1 levels (Fig. 4B). Presumably, theantagonistic effects of PMA on the FGF2 down-regulation of p27Kip1

levels are mediated by the inhibition of the PI3K/Akt pathway.Conversely, AVP very likely promotes stabilisation of p27Kip1 levels bytwo parallel methods, i.e., (a) inhibition of the PI3K/Akt pathway and(b) via another pathway independent of PI3K/Akt. Thus, in thepresence of PKC inhibitors, AVP caused the up-regulation of AKT(Fig. 3A and Fig. S2A) and also enhanced the up-regulation of AKT byFGF2 (Fig. S2A). However, despite up-regulating AKT (Fig. 3A), AVPwas able to maintain high levels of p27Kip1 in the presence of PKC

Go6983 1µM30min +

AVP 1nM5min + + + +Bisindolylmaleimide 5µM 30min + + + +

+ + +LY294002 20µM 30min + + + +

+ +

+ +FGF2 5ng/ml 10min +

pAKT

pERK1pERK2

AVP 1nM5min + +FGF22ng/ml5min + +H89 2µM30min + + + +

B

pAKT

Total AKT

Total AKT

1 2 3 4 5

1 2 3 4 5 6 7 8 9 10 11 12 13

0

0,5

1

1,5

2

2,5

1 2 3 4 5

Rel

ativ

e p

AK

T le

vels

(fo

ld c

on

tro

l)

**

0

0,5

1

1,5

2

2,5

3

3,5

4

1 2 3 4 5 6 7 8 9 10 11 12 13

Rel

ativ

e P

ho

sph

o-P

rote

in

leve

ls (

fold

co

ntr

ol)

pAKTpERK1/2

****

*

*

*

****

**

*

*

**

A

Fig. 3. Activation of ERK1/2 and AKT by AVP and FGF2 in the presence of PKC, PI3K and PKA inhibitors monitored by Western blotting. Parental Y1 cells G0/G1-arrested by 48 h ofserum deprivation were pre-treated for 0.5 h with PKC (bisindolylmaleimide or Go6983) and PI3K (LY294002) inhibitors (A) or PKA (H-89) inhibitor (B) and stimulated with AVPand/or FGF2 at the indicated times; lysates obtained were quantified by Bradford assay and probed against phospho-ERK1/2 and phospho-AKT. The Westerns shown arerepresentative of three independent experiments, which were quantified by densitometry and averaged to generate the histograms displayed in the panel; statistic significance bythe F-test or ANOVA; *, pb0.01; **, pb0.05.

1441F.L. Forti, H.A. Armelin / Biochimica et Biophysica Acta 1813 (2011) 1438–1445

inhibitors (Fig. 4B). Therefore, in the presence of PKC inhibitors, AVP/V1aR and FGF2/FGFR signalling axes resemble each other in regard toactivation of both ERK1/2 and PI3K/Akt mitogenic pathways, but theireffects on p27Kip1 activity are different.

Regulation of p27Kip1 activity is known for being exerted on bothprotein stability and subcellular location. Phosphorylation of p27Kip1

at Thr187 is the only p27Kip1 covalent modification recognised andbound by ubiquitin ligase for subsequent degradation, whereasphosphorylation of p27Kip1 at Ser10, which is catalysed by bothnuclear kinases and AKT, labels p27Kip1 so that it can be exported fromthe nucleus. Therefore, it became important to analyse p27Kip1

phosphorylation in Y1 cells submitted to AVP, FGF2 or AVP+FGF2.Thus, G0/G1-arrested Y1 cells stimulated to progress across the G1phase by serum (not shown) or FGF2 displayed(a) high levels ofp27Kip1 phosphorylated at Thr187 up to 5 h, which decreasedbetween 8 and 12 h and went up again by 24 h (Fig. 5A) and (b)increased levels of p27Kip1 phosphorylated at Ser10 after 8 h (Fig. 5B).These results are consistent with the principle that FGF2 promotesAKT activation with consequent phosphorylation of p27Kip1 at bothThr187 and Ser10, leading to, respectively, degradation and relocationof p27Kip1 from the nucleus to the cytoplasm, allowing G1 phaseprogression. Noticeably, AVP also regulates p27Kip1 phosphorylationat both Thr187 (Fig. 5A) and Ser10 (Fig. 5B) but with a different,antagonistic pattern compared to FGF2.

3.3. AVP promotes survival of Y1 cells ectopically expressing p27Kip1

that are mutated at Ser10 or Thr187

The mutants p27-S10A and p27-T187A as well as the wild-typep27Kip1-WT were ectopically expressed in Y1 cells to test thehypothesis that p27Kip1 stabilisation underlies AVP's inhibition of Y1cell proliferation. Plasmid constructs containing p27Kip1-WT, p27-S10A and p27-T187A were transiently transfected into Y1 cellcultures, either treated or untreated with AVP, for 24 h. Afterwards,they were incubated in fresh medium until they developed visiblecolonies (Fig. 6). When compared with mock-transfected controlcultures, the numbers of surviving colonies in the Y1 cells transfectedwith p27Kip1-WT, S10A or T187A were reduced by 56%, 70% and 83%,respectively (Fig. 6). These results were expected because transientectopic expression of p27Kip1-WT, p27-S10A and p27-T187A shouldcause respectively the inhibition of cell growth and likely reduction ofcell viability. In addition, AVP treatment reduced the number ofsurviving colonies in both mock-transfected controls and Y1 cellstransfected with p27Kip1-WT (Fig. 6). These results were alsoexpected; the results confirmed the hypothesis that AVP inhibitsAkt activation by mitogens like FCS or FGF2, changing pattern ofp27Kip1 phosphorylation at both Ser10 and Thr187 (Fig. 5) and leadingto p27Kip1 stabilisation (Fig. 4). However, AVP treatment surprisinglypromoted the survival of Y1 cells transiently transfected with either of

p27Kip1

p27Kip1

p27Kip1

AVP+FGF2

Bisindolylmaleimide + + +

+ + +

+ + +

+ + +

5 M 30 min

Bisindolylmaleimide5 M 30 min

p27Kip1

PMA10ng/ml 1h 3h 5h 1h 3h 5h

1h 3h 5h 1h 3h 5h

1h 3h 5h 1h 3h 5h

1h 3h 5h 1h 3h 5h

B

p27Kip1

AVP1nM

PMA 10ng/ml

C

FGF2 5ng/ml

AVP 1nM

FGF2 5ng/ml

actin

actin

1 2 3 4 5 6

1 2 3 4 5 6

7

1 2 3 4 5 6 7

0

0,5

1

1,5

2

2,5

3

1 2 3 4 5 6

Rel

ativ

e p

27 le

vels

(fo

ld c

on

tro

l)

Rel

ativ

e p

27 le

vels

(fo

ld c

on

tro

l)FGF2

AVP

FGF2+AVP

0

0,5

1

1,5

2

2,5

3

1 2 3 4 5 6 7

PMA+I

AVP+I

00,20,40,60,8

11,21,41,61,8

2

1 2 3 4 5 6 7

Rel

ativ

e p

27 le

vels

(fo

ld c

on

tro

l)

FGF2+AVP

FGF2+PMA

*

**

**

******

****

****

**

** ** ** ** **

** * *

*

**

**

* **

*

*

p27Kip1

p27Kip1

actin

Time (h) 0 1 5 8 12 24

FGF2 5ng/ml

AVP 1nM

A

Fig. 4. p27Kip1 levels in G0/G1-arrested Y1 cells by serum deprivation and stimulated with AVP, PMA and/or FGF2. Cellular lysates quantified by Bradford assay were probed againsttotal p27Kip1 protein and show that AVP and FGF2 differently affect the stability or expression of p27Kip1 (A), while PMA has similar effects to those of AVP (C), which are bothinhibited by PKC inhibitors (B). Blots shown are representative of three independent experiments, which were quantified by densitometry and averaged to generate the histogramsdisplayed in the panel; statistic significance by the F-test or ANOVA; *, pb0.01; **, pb0.05.

1442 F.L. Forti, H.A. Armelin / Biochimica et Biophysica Acta 1813 (2011) 1438–1445

the mutants p27-S10A or p27-T187A, which cannot be phosphory-lated at Ser10 and Thr187, respectively, as shown by the two-foldincrease in the number of surviving colonies in both cases (Fig. 6).Thus, the effect of AVP on Y1 cell survival is dependent on which typeof p27Kip1 a cell expresses.

4. Discussion and conclusion

For almost 20 years, regulation of mitogenesis and cell survival bythe AVP/V1aR signalling system has been studied in both primarycultures and cell lines derived from multiple peripheral tissues [6–8].The main purpose of this study was to investigate the interactionbetween the p27Kip1 protein and PKC and PI3K/Akt signallingpathways in both pro-senescence and anti-proliferative responses,which are triggered by the AVP/V1aR axis, in mouse malignant Y1adrenocortical cells. Thus, we hypothesized that, in addition to thedown-regulation of the cyclin D1, the increase of the cell cycleinhibitor p27Kip1, as reported in studies on the anti-proliferativeeffects of ACTH in Y1 cell line [18], might be another putative target ofAVP acting against the oncogenic ras-driven proliferation of thesecells. We have shown that AVP triggers two antagonistic responses inV1aR-expressing mouse Y1 adrenal malignant cells: (a) it activates

the PI3K/Akt, PKC and ERK1/2 mitogenic pathways [16], and (b) itconcomitantly activates RhoA GTPase, causing down-regulation ofcyclin D1, induction of senescence-associated β-galactosidase and cellcycle arrest [14,16].

Here, we explored both PMA and AVP signalling in the metabolicbackground of K-Ras-dependent Y1 adrenocortical cells to show thatactivation of PKC with PMAmimics the effect of AVP, and inhibitors ofPKC completely protect cells from both PMA and AVP toxic effects inclonogenic assays (Fig. 1). The responsiveness of Y1 cells to PKCactivators is phenotypically dependent on Ki-ras oncogene amplifi-cation because K-Ras-GTP down-regulation by ectopic expression of adexamethasone-inducible form of the dominant-negative mutant H-RasN17 abolishes the toxic effects of both PMA and AVP (Fig. 2). Touncover the molecular signalling mechanisms underlying thesephenomena, we analysed the outcome of ERK, PKC and PI3K/Aktactivation in Y1 malignant cells. Thus, AVP activated ERK1/2 despitePKC activity, but inhibitors of PKC reduced the AVP's degree of ERK1/2activation (Fig. 3A); however, AVP had a dual effect on the PI3K/Aktpathway. First, AVP inhibited PI3K/Akt via PKC activation (Fig. 3B);however, if PKC was blocked by specific inhibitors, AVP activated thePI3K/Akt pathway like a bona fide mitogen (Fig. 3B). Second, PMA,which acts exclusively via the activation of PKC, promoted

p27Kip1 -phospho-T187

Time (h)

Time (h)

0 1 5 8 12 24

p27Kip1 -phospho-T187

p27Kip1 -phospho-T187

p27Kip1 -phospho-S10

0 1 5 8 12 24

B

p27Kip1 -phospho-S10

p27Kip1 -phospho-S10

actin

actin

1 2 3 4 5 6

1 2 3 4 5 6

0

0,5

1

1,5

2

2,5

3

1 2 3 4 5 6Rel

ativ

e P

ho

sph

o-p

27-T

187

leve

ls(f

old

co

ntr

ol)

0

0,5

1

1,5

2

2,5

1 2 3 4 5 6Rel

ativ

e P

ho

sph

o-p

27-S

10 le

vels

(fo

ld c

on

tro

l)

FGF2AVPFGF2+AVP

FGF2AVPFGF2+AVP

**

***

*

*

**

*

*

**

**

** ** *

**

**

FGF2 5ng/ml

AVP 1nM

AVP+FGF2

FGF2 5ng/ml

AVP 1nM

AVP+FGF2

A

Fig. 5. Phosphorylation levels of p27Kip1–Ser10 or p27Kip1–Thr187 in G0/G1-arrested Y1 cells stimulated with AVP and/or FGF2. Cellular lysates obtained at the indicated times afterAVP±FGF2 stimuli were assayed byWestern blotting against p27Kip1–phospho-Thr187 (A) or p27Kip1–phospho-Ser10 (B) and showed different patterns of protein levels as the cellstransitioned through the cell cycle phases. The Westerns shown are representative of three independent experiments, which were quantified by densitometry and averaged togenerate the histograms displayed in the panel; statistic significance by the F-test or ANOVA; *, pb0.01; **, pb0.05.

1443F.L. Forti, H.A. Armelin / Biochimica et Biophysica Acta 1813 (2011) 1438–1445

phosphorylation of ERK1/2 and inhibited PI3K/Akt activation (Fig. S1Aand B). Inhibitors of PKC abolished the effects of PMA on both ERK1/2and PI3K/Akt. These results indicate that chronic high levels of K-Ras-GTP, as found in Y1 cells [18], lead to a malignant phenotype in whichPKC is a target that inhibits the up-regulation of the PI3K/Akt pathwaywhen activated and thus reduces both cell proliferation and survival.By these molecular mechanisms, AVP stabilises and/or increases theexpression of the p27Kip1 protein (Fig. 4C) to arrest the cell cycle andto antagonize the mitogenic activity of FGF2 or serum. Therefore,p27Kip1 seems to be a critical signalling node in the molecularmechanisms underlying the survival vulnerability of K-Ras-depen-dent malignant cells. It is interesting to note that p27Kip1 is the targetof the anti-proliferative mechanisms of two distinct peptide hor-mones: AVP, as is discussed in this paper, and ACTH [18]. Moreover,similar anti-proliferative mechanisms involving both the PI3K–Aktpathway and the p27Kip1 protein were recently reported in mousemodels of K-Ras-driven urethane lung tumours [21].

We previously have shown that AVP blocked the expression ofcyclin D1 mRNA and protein induced by FGF2 in G0/G1-arrested Y1cells [16]. In this paper, we are extending this investigation to showthat AVP, under the same conditions, also maintains high levels ofp27Kip1 protein. These results suggest that AVP must simultaneouslycause both blocking of cyclin D1 expression and p27Kip1 stabilisationto achieve irreversible cell cycle arrest and the reduction of survival inY1 adrenal malignant cells. However, we also reported that ectopicover-expression of cyclin D1 was sufficient to render Y1 cellscompletely resistant to the noxious effects of AVP [16]; thisobservation does not contradict the idea that irreversible Y1 cellcycle arrest requires both the blocking of cyclin D1 expression andp27Kip1 stabilisation because the ectopically expressed cyclin D1,when coupled to CDKs, presumably interacts with p27Kip1 to yieldneutralised complexes in a cell nucleus [22–25].

However, the effects of AVP on p27Kip1 signalling seem morecomplex than simple p27Kip1 protein stabilisation mediated by PKCactivation (Fig. 4) or by differential post-transcriptional regulation ofkey phosphorylation sites (Fig. 5). In Y1 cells that transiently andectopically express wild-type p27Kip1, AVP caused a significantreduction in cell survival, as shown by clonogenic assays (Fig. 6).Conversely, AVP promoted the survival of Y1 cells by transientlyexpressing either of the mutants p27-S10A or p27-T187A, neither ofwhich can be phosphorylated at Ser10 and Thr187, respectively(Fig. 6). Thus, Y1 cells recognise AVP as a survival factor whenendogenous p27Kip1 is replaced by exogenous mutated p27Kip1 withan altered affinity for specific cyclin–CDK complexes, stability andsubcellular localisation [26,27].

Support for the concept of p27Kip1 dual role in both tumoursuppression and tumour promotion has been established in recentyears. Thus, the suppressor role of p27Kip1 was first demonstrated bythe increased susceptibility of p27Kip1-deficient mice to developtumours [28–31]. However, the CDKN1B gene that encodes p27Kip1 is anotable exception to the classic tumour suppressor gene paradigmbecause it is rarely mutated in tumours [32,33]. Instead, reducedlevels of the p27Kip1 protein, or cellular mislocalisation, correlate withtumour aggressiveness and poor prognosis [34]. However, specificcases of cancers in the breast, thyroid, oesophagus and colon werefound that displayed nuclear p27Kip1 exportation to the cytoplasm,suggesting that cytoplasmic mislocalisation of p27Kip1 may beimportant for tumour progression [35]. In summary, while tumoursuppressor activity is activated in normal cells in response to anti-proliferative signals, thus tilting the balance towards cell cycleblockage and inhibition of proliferation, tumour cells have developedstrategies to down-regulate these functions of p27Kip1. However,other activities that appear to be pro-tumourigenic are neitheraffected nor even enhanced, which results in a stimulation of tumour

Control p27Kip1-WT p27Kip1-S10A p27Kip1-T187A

DME/10%FCS+

AVP 1nM

DME/10%FCS

0306090

120150180210240270300

Co

lon

ies

/ Pla

te

-AVP

+ AVP

***

***

**

*

330

Fig. 6. Clonogenic assays of Y1 cells ectopically expressing wild-type p27Kip1 and its mutants show that AVP effects depend on the levels and status of phosphorylated forms ofp27Kip1. Y1 cells (2×104) were plated in six-well plates and lipofected 24 h later with WT–p27Kip1, p27Kip1–S10A and p27Kip1–T187A constructs for 6 h. After returning cells to 10%FBS-containing medium and another 24 h of recovery, the cells were stimulated with 1 nM AVP for 24 h. The mediumwas changed every third day until colonies developed. Resultsof colonies/plate from three independent experiments were pooled and statistically analysed by χ2 with 1 degree of freedom; *, pb0.05; **, pb0.01; ***, pb0.001. The panelhistogram displays averages of colonies/plate with error bars; upper panel displays averages of colonies/plate.

1444 F.L. Forti, H.A. Armelin / Biochimica et Biophysica Acta 1813 (2011) 1438–1445

development. The p27Kip1 balance is regulated by multiple post-translational modifications that affect its function by altering in-teractions between proteins, affecting subcellular localisation andmodulating protein stability. This study further demonstrates the dualroles of the p27Kip1 protein and relates the dual roles of AVP signallingto mouse tumour cells.

Supplementarymaterials related to this article can be found onlineat doi:10.1016/j.bbamcr.2011.04.007.

Disclosures

The authors fully declare that there is no financial or otherpotential conflict of interest.

Funding

This work was supported by Sao Paulo Research Foundation,Fapesp (grant numbers08/58264-5 and 08/51273-9) and NationalCouncil for Scientific and Technological Development, CNPq (grantnumber 475488/2008-3).

Author contribution statements

The first author was fully responsible for conducting theexperimental portion and for writing and editing the manuscript;the co-author and supervisor were both responsible for writing themanuscript.

Acknowledgements

The authors thank the funding agencies Fapesp and CNPq and thetechnical support from the people at the Instituto de Quimica,Universidade de Sao Paulo.

References

[1] M.J. Brownstein, J.T. Russell, H. Gainer, Synthesis, transport, and release ofposterior pituitary hormones, Science 207 (1980) 373–378.

[2] W.S. Young 3rd, H. Gainer, Transgenesis and the study of expression, cellulartargeting and function of oxytocin, vasopressin and their receptors, Neuroendo-crinology 78 (2003) 185–203.

[3] F.A. Antoni, Novel ligand specificity of pituitary vasopressin receptors in the rat,Neuroendocrinology 39 (1984) 186–188.

[4] S.J. Lolait, A.M. O'Carroll, L.C. Mahan, C.C. Felder, D.C. Button, Extrapituitaryexpression of the rat V1b vasopressin receptor gene, Proc. Natl. Acad. Sci. U. S. A.92 (1995) 6783–6787.

[5] L. Bankir, Antidiuretic action of vasopressin: quantitative aspects and interactionbetween V1a and V2 receptor-mediated effects, Cardiovasc. Res. 51 (2001)372–390.

[6] D.A. Carter, C.K. Fai, D. Murphy, Neurohypophyseal peptides as regulators ofgrowth and development. A review, J. Mol. Neurosci. 4 (1993) 11–19.

[7] H.K. Caldwell, H.J. Lee, A.H. Macbeth, W.S. Young 3rd, Vasopressin: behavioralroles of an “original” neuropeptide, Prog. Neurobiol. 84 (2008) 1–24.

[8] S. Jard, C. Barberis, S. Audigier, E. Tribollet, Neurohypophyseal hormone receptorsystems in brain and periphery, Prog. Brain Res. 72 (1987) 173–187.

[9] G. Alonso, E. Galibert, V. Boulay, A. Guillou, A. Jean, V. Compan, G. Guillon,Sustained elevated levels of circulating vasopressin selectively stimulate theproliferation of kidney tubular cells via the activation of V2 receptors,Endocrinology 150 (2009) 239–250.

[10] M.M. Rinschen, M.J. Yu, G. Wang, E.S. Boja, J.D. Hoffert, T. Pisitkun, M.A. Knepper,Quantitative phosphoproteomic analysis reveals vasopressin V2-receptor-dependentsignaling pathways in renal collecting duct cells, Proc.Natl. Acad. Sci. U. S. A. 107 (2010)3882–3887.

[11] W.G. North, Gene regulation of vasopressin and vasopressin receptors in cancer,Exp. Physiol. 85 (2000) 27S–40S.

1445F.L. Forti, H.A. Armelin / Biochimica et Biophysica Acta 1813 (2011) 1438–1445

[12] C. Pequeux, B.P. Keegan, M.T. Hagelstein, V. Geenen, J.J. Legros, W.G. North,Oxytocin- and vasopressin-induced growth of human small-cell lung cancer ismediated by the mitogen-activated protein kinase pathway, Endocr. Relat. Cancer11 (2004) 871–885.

[13] B.P. Keegan, B.L. Akerman, C. Péqueux, W.G. North, Provasopressin expression bybreast cancer cells: implications for growth and novel treatment strategies, BreastCancer Res. Treat. 95 (2006) 265–277.

[14] F.L. Forti, H.A. Armelin, Vasopressin triggers senescence in K-ras transformed cellsvia RhoA-dependent downregulation of cyclin D1, Endocr. Relat. Cancer 14 (2007)1117–1725.

[15] Y.P. He, L.Y. Zhao, Q.S. Zheng, S.W. Liu, X.Y. Zhao, X.L. Lu, X.L. Niu, Argininevasopressin stimulates proliferation of adult rat cardiac fibroblasts via proteinkinase C-extracellular signal-regulated kinase 1/2 pathway, Acta Physiol. Sin. 60(2008) 333–340.

[16] T.T. Schwindt, F.L. Forti, M.A. Juliano, L. Juliano, H.A. Armelin, Arginine vasopressininhibition of cyclin D1 gene expression blocks the cell cycle and cell proliferationin the mouse Y1 adrenocortical tumor cell line, Biochemistry 42 (2003)2116–2121.

[17] Y. Yasumura, Y. Buonassissi, G. Sato, Clonal analysis of differentiated function inanimal cell cultures. I. Possible correlated maintenance of differentiated functionand the diploid karyotype, Cancer Res. 26 (1966) 529–535.

[18] F.L. Forti, T.T. Schwindt, M.S. Moraes, C.B. Eichler, H.A. Armelin, ACTH promotionof p27(Kip1) induction in mouse Y1 adrenocortical tumor cells is dependent onboth PKA activation and Akt/PKB inactivation, Biochemistry 41 (2002)10133–10140.

[19] M. Schwab, K. Alitalo, H. Varmus, J. Bishop, D. George, A cellular oncogene (c-Ki-ras)is amplified, overexpressed, and located within karyotypic abnormalities in mouseadrenocortical tumour cells, Nature 303 (1983) 497–501.

[20] F.L. Forti, H.A. Armelin, ACTH inhibits A Ras-dependent anti-apoptotic andmitogenic pathway in mouse Y1 adrenocortical cells, Endocr. Res. 26 (2000)911–914.

[21] K.S. Kelly-Spratt, J. Philipp-Staheli, K.E. Gurley, K. Hoon-Kim, S. Knoblaugh, C.J.Kemp, Inhibition of PI3K restores nuclear p27Kip1 expression in amouse model ofK-ras driven lung cancer, Oncogene 28 (2009) 3652–3662.

[22] M. Imoto, Y. Doki, W. Jiang, E.K. Han, I.B. Weinstein, Effects of cyclin D1overexpression on G1 progression-related events, Exp. Cell Res. 236 (1997)173–180.

[23] M. James, A. Ray, D. Leznova, S.W. Blain, Differential modification of p27Kip1controls its cyclin D-cdk4 inhibitory activity, Mol. Cell. Biol. 28 (2008) 498–510.

[24] J.K. Kim, J.A. Diehl, Nuclear cyclin D1: an oncogenic driver in human cancer, J. Cell.Physiol. 220 (2009) 292–296.

[25] P. Sicinski, S. Zacharek, C. Kim, Duality of p27Kip1 function in tumorigenesis,Genes Dev. 21 (2007) 1703–1706.

[26] A. Borriello, V. Cucciolla, A. Oliva, V. Zappia, F. Della Ragione, p27Kip1metabolism:a fascinating labyrinth, Cell Cycle 6 (2007) 1053–1061.

[27] A. Besson, M. Gurian-West, X. Chen, K.S. Kelly-Spratt, C.J. Kemp, M. James, J.M.Roberts, A pathway in quiescent cells that controls p27Kip1 stability, subcellularlocalization, and tumor suppression, Genes Dev. 20 (2006) 47–64.

[28] K. Nakayama, N. Ishida, M. Shirane, A. Inomata, T. Inoue, N. Shishido, I. Horii, D.Y.Loh, K. Nakayama, Mice lacking p27Kip1 display increased body size, multipleorgan hyperplasia, retinal dysplasia, and pituitary tumors, Cell 85 (1996) 707–720.

[29] M.L. Fero, M. Rivkin, M. Tasch, P. Porter, C.E. Carow, E. Firpo, K. Polyak, L.H. Tsai, V.Broudy, R.M. Perlmutter, K. Kaushansky, J.M. Roberts, A syndrome of multiorganhyperplasia with features of gigantism, tumorigenesis, and female sterility inp27Kip1-deficient mice, Cell 85 (1996) 733–744.

[30] H. Kiyokawa, R.D. Kineman, K.O. Manova-Todorova, V.C. Soares, E.S. Hoffman, M.Ono, D. Khanam, A.C. Hayday, L.A. Frohman, A. Koff, Enhanced growth of micelacking the cyclin-dependent kinase inhibitor function of p27Kip1, Cell 85 (1996)721–732.

[31] M.L. Fero, E. Randel, K.E. Gurley, J.M. Roberts, C.J. Kemp, The murine gene p27Kip1is haploinsufficient for tumour suppression, Nature 396 (1998) 177–180.

[32] N. Kawamata, R. Morosetti, C.W. Miller, D. Park, K.S. Spirin, T. Nakamaki, S.Takeuchi, Y. Hatta, J. Simpson, S. Wilcyznski, Molecular analysis of the cyclin-dependent kinase inhibitor gene p27Kip1 in human malignancies, Cancer Res. 55(1995) 2266–2269.

[33] M.V. Ponce-Castaneda, M.H. Lee, E. Latres, K. Polyak, L. Lacombe, K. Montgomery,S. Mathew, K. Krauter, J. Sheinfeld, J. Massague, p27Kip1: chromosomal mappingto 12p12-12p13.1 and absence of mutations in human tumors, Cancer Res. 55(1995) 1211–1214.

[34] I.M. Chu, L. Hengst, J.M. Slingerland, The Cdk inhibitor p27 in human cancer:prognostic potential and relevance to anticancer therapy, Nat. Rev. Cancer8 (2008) 253–267.

[35] S.W. Blain, J. Massagué, ‘Breast cancer banishes p27 from nucleus, Nat. Med.8 (2002) 1076–1078.