Downregulation of Fer induces PP1 activation and cell-cycle arrest in malignant cells

ORIGINAL ARTICLE

Downregulation of Fer induces PP1 activation and cell-cycle arrest in

malignant cells

O Pasder, S Shpungin, Y Salem, A Makovsky, S Vilchick, S Michaeli, H Malovani and U Nir

The Mina and Everard Goodman Faculty of Life Sciences, Bar-Ilan University, Ramat-Gan, Israel

Fer is a nuclear and cytoplasmic intracellular tyrosinekinase. Herein we show that Fer is required for cell-cycleprogression in malignant cells. Decreasing the level of Ferusing the RNA interference (RNAi) approach impeded theproliferation of prostate and breast carcinoma cells andled to their arrest at the G0/G1 phase. At the molecularlevel, knockdown of Fer resulted in the activation of theretinoblastoma protein (pRB), and this was reflected byprofound hypo-phosphorylation of pRB on both cyclin-dependent kinase CDK4 and CDK2 phosphorylation sites.Dephosphorylation of pRB was not seen upon the directtargeting of either CDK4 or CDK2 expression, and wasonly partially achieved by the simultaneous depletion ofthese two kinases. Amino-acid sequence analysis revealedtwo protein phosphatase 1 (PP1) binding motifs in thekinase domain of Fer and the association of Fer with thepRB phosphatase PP1a was verified using co-immuno-precipitation analysis. Downregulation of Fer potentiatedthe activation of PP1a and overexpression of Fer decreasedthe enzymatic activity of that phosphatase. Our findingsportray Fer as a regulator of cell-cycle progression inmalignant cells and as a potential target for cancerintervention.Oncogene (2006) 25, 4194–4206. doi:10.1038/sj.onc.1209695;published online 29 May 2006

Keywords: malignant cells; Fer kinase; cell-cycle; PP1;pRB

Introduction

Fer is an intracellular tyrosine kinase, which was foundto reside in both the cytoplasm and nucleus ofmammalian cells (Letwin et al., 1988; Hao et al., 1989;Ben-Dor et al., 1999). Together with c-Fes, Ferconstitutes a distinct subfamily of intracellular tyrosinekinases that share a unique structure. Both kinases bearan extended N-terminal tail that contains an Fps/Fes/Fer/CIP4 homology domain (FCH), followed by threecoiled-coil-forming regions. Whereas the FCH domain

could mediate the association of Fer and c-Fes withmicrotubular structures, the coiled-coil domains wereshown to direct the oligomerization of these kinases.The kinase domain of the two enzymes resides at theircarboxy-terminal part and is 70% identical in the twoproteins (Greer, 2002).

Fer is activated by growth factors such as epidermalgrowth factor and platelet-derived growth factor infibroblastic cells (Kim and Wong, 1998), and byoccupation of the Fcg receptor in mast cells (Penhallowet al., 1995). Activation of Fer in these systems could belinked to the modulation of cell–cell and cell–substra-tum interactions, as Fer was shown to associate withand to phosphorylate adherence molecules (Kim andWong, 1998; Rosato et al., 1998; Craig et al., 2001;Greer, 2002; Kogata et al., 2003; Miravet et al., 2003;Piedra et al., 2003). Fer was also shown to associate withkey cellular regulatory proteins like phosphatidylinosi-tol-3 kinase (Iwanishi et al., 2000), signal transducer andactivator of transcription-3 (Stat3) (Priel-Halachmiet al., 2000) and with the cytoskeletal linker proteinplectin (Lunter and Wiche, 2002).

Although present in a wide variety of tissues and cells,the functional role of Fer has been elucidated mainly incells that carry out innate immune response (Greer,2002; Craig and Greer, 2002). Interestingly, mice devoidof an active Fer develop normally and the proliferationof fibroblasts derived from these mice is not impaired invitro (Craig et al., 2001).

However, several lines of evidence suggest that Ferplays a functional role in the progression and growth ofmalignant tumors. The kinase was detected in all humanmalignant cell lines analysed (Hao et al., 1989; Orlovskyet al., 2000) and its levels in malignant prostate tumorsare significantly higher than those detected in benigntumors (Allard et al., 2000). Furthermore, downregula-tion of Fer impaired the proliferation and abolished theability of prostate carcinoma PC3 cells to form coloniesin soft agar (Allard et al., 2000). However, the molecularpathways that underlie the function of Fer in malignantcells have not been elucidated. Intervention with theproliferation of malignant cells by the targeting of asingle kinase is intriguing considering the redundantcell-cycle regulation in malignant cells (Ortega et al.,2003). To unravel the mechanisms that underlie the roleof Fer in the proliferation of cancer cells, we specificallyknocked down Fer in retinoblastoma protein (pRB)-positive malignant cell lines. This revealed a key role of

Received 4 April 2005; revised 22 March 2006; accepted 18 April 2006;published online 29 May 2006

Correspondence: Professor U Nir, Faculty of Life Sciences Bar-IlanUniversity, Ramat-Gan 52900, Israel.E-mail: [email protected]

Oncogene (2006) 25, 4194–4206& 2006 Nature Publishing Group All rights reserved 0950-9232/06 $30.00

www.nature.com/onc

Fer in maintaining G1–S transition during cell-cycleprogression in malignant cells.

Results

Downregulation of Fer arrests malignant cells at theG0/G1 phaseTo dissect the molecular role of Fer, the cellular level ofthis kinase was specifically reduced in two malignant celllines that harbor a functional pRB: PC3, a prostatecarcinoma cell line, and MDA-MB-231, a breastcarcinoma cell line. The two cell lines were exposed for

72 h to a synthetic double-stranded small interferingRNA (siRNA) directed toward unique 21 nt in thehuman fer mRNA (Hao et al., 1989).

In parallel, the cells were treated with a non-relevantsiRNA directed toward the luciferase (luc.) mRNA. fer-siRNA led to about five and three fold specific decreasein the level of Fer in PC3 and MDA-MB-231 cells,respectively, but did not affect the level of actin, whichwas used as a control (Figure 1a and c). Decrease in thelevel of Fer lowered the number of cells in PC3- andMDA-MB-231-treated cultures by approximately 50and 40% respectively, as was determined using the3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazoliumbromide (MTT) assay (Figure 1b and d). This could

Fer

Actin

Fer

Actin

1

0

120

100

80

60

40

20

0

120

100

80

60

40

20

0

2 3 1 2 3

12000

10000

8000

6000

4000

2000

0

7000

6000

5000

4000

3000

2000

1000

control luc.-siRNA

fer-siRNA

control luc.-siRNA

fer-siRNA

Treatment

PC3 cells

control luc.- siRNA

fer- siRNA

control luc.- siRNA

fer- siRNA

MDA-MB-231 cells

Arb

itrar

y U

nits

Per

cent

age

of v

iabl

e ce

lls

Per

cent

age

of v

iabl

e ce

llsA

rbitr

ary

Uni

ts

Treatment

cont

rol

luc-

siR

NA

fer-

siR

NA

cont

rol

luc-

siR

NA

fer-

siR

NAa

b d

c

**

**

Figure 1 Reduced number of viable cells in fer-siRNA-treated prostate and breast carcinoma cells. (a) Whole cell lysates (WCL) fromPC3 cells that were either left untreated or exposed for 72 h to luc.-siRNA, or fer-siRNA were reacted with anti-Fer or anti-actinantibodies in a Western blot analysis. Scanning values of radiograms from three independent experiments are presented as means7s.e.(*Po0.001 compared to luc.-siRNA, was calculated using a Student’s t-test). (b) Reduced number of viable cells in fer-siRNA-treatedculture. PC3 cells untreated or exposed to luc.-siRNA or fer-siRNA were subjected to an MTT viability assay. The values representpercentage of viable cells compared to the control and are means7s.e. (*Po0.01 (N¼ 5) compared to luc.-siRNA-treated cells).(c) WCL from MDA-MB-231 cells untreated or treated with luc.-siRNA or fer-siRNA were reacted with anti-Fer and anti-actinantibodies in a Western blot analysis. Scanning values of radiograms from three independent experiments are presented as means7s.e.(*Po0.005 compared to luc.-siRNA). (d) MDA-MB-231 cells untreated or exposed to luc.-siRNA or fer-siRNA were subjected to anMTT viability assay. Values represent means7s.e. (*Po0.005 (N¼ 5) compared to luc.-siRNA-treated cells).

Fer maintains G1–S transition in malignant cellsO Pasder et al

4195

Oncogene

either result from a diminished viability of the treatedcells or could reflect their impaired proliferation rate.Flow-cytometry analysis revealed that whereas luc.-siRNA did not affect the cell-cycle profile of PC3 cells,fer-siRNA increased the percentage of the G0/G1 cellsand markedly decreased the percentage of cells residingin the S phase from 19 to 7% (Figure 2a). However,neither luc.-siRNA nor fer-siRNA significantly affectedthe percentage of the sub-G0/G1 apoptotic cells in thetreated cultures (Figure 2b). Moreover, annexin V andlactate dehydrogenase assays corroborated the lack ofapoptotic or necrotic death induction, in the fer-siRNA-

treated cells (data not shown). This implies thatdepletion of Fer affects the transition of cells from theG1 to the S phase rather than affecting their viability.

To confirm the essential role of Fer in G1–Stransition, untreated and siRNA-treated PC3 cells weresubjected to bromodeoxyuridine (BrdU) incorporationassay. Whereas the percentage of S phase cells wassimilar in the untreated and in luc.-siRNA-treated cells,fer-siRNA treatment reduced the percentage of BrdU-incorporating cells from 20 to 6% (Figure 2c). Hence,reduction in the cellular level of Fer interferes with thetransition of PC3 cells from the G0/G1 to the S phase.

control

luc.-siRNA

fer-siRNA6972

82

18

G0/G1

control luc.-siRNA

S G2/M

19

7

139

11

cell

cycl

e di

strib

utio

n (%

)

320

240

160

80

0

320

400

240

160

80

320

240

160

80

0 50 100 150 200 25 0 50 100 150 200 250

00 50 100 150 200 25

� � � �

� �

2n 4n 2n 4n

2n 4n

fer-siRNA

104

103

102

101

100

104

103

102

101

100

104

103

102

101

100

0 200

2n 4n

2n 4n

2n 4n

400 600 800 1000

0 200 400 600 800 1000

0 200 400 600 800 1000

G0/G1-64.2%S-23%G2/M-12.8%

G0/G1-66.4%S-20%G2/M-11.6%

G0/G1-84.6%S-6%G2/M-10.4%

fer-siRNA

Propidium lodide

Brd

U-F

ITC

fluo

resc

ence

luc.-siRNA

controla c

b

Figure 2 Depletion of Fer arrests malignant cells at the G0/G1phase. (a) PC3 cells untreated or exposed to luc.-siRNA or fer-siRNAwere subjected to flow-cytometry analysis. Percentage of cells in each cell-cycle phase is shown as histogram. (b) Flow-cytometrydistribution profiles of the values presented in (a). The peaks of 2n and 4n DNA-containing cells are indicated by arrows. (c) Controland treated cells were incubated with BrdU and the percentage of cells incorporating BrdU (ellipses) was determined. The valuesrepresent one out of three independent experiments that gave similar results.


4196

Oncogene

Similar reduction in S phase population was alsoobserved in fer-siRNA-treated MDA-MB-231 cells(data not shown).

Knockdown of Fer leads to the downregulation of CDK2in prostate carcinoma cellsThe arrest of Fer-deprived cells in the G0/G1 phasesuggests the involvement of Fer in maintaining ormodulating the function of G1–S regulatory proteins.The cellular levels of both negative and positiveregulators of G1 progression and G1–S transition(Vermeulen et al., 2003) were therefore determined inuntreated and in siRNA-treated cells. PC3 cells weresubjected to the fer-siRNA and the levels of early G1cyclins like cyclin D1, the late G1 cyclin–cyclin E andthe S-phase cyclin–cyclin A (Vermeulen et al., 2003)were examined. Of the three cyclins analyzed, only thelevel of cyclin A was slightly decreased in fer-siRNA-treated cells (Figure 3a). We also examined the levels ofcyclin-dependent kinases (CDKs) that serve as acatalytic constituent of the cyclin/CDK complexes(Vermeulen et al., 2003). Whereas the levels of CDK4,

CDK6 and CDK1 were not changed in cells subjected tofer-siRNA, the level of CDK2 was significantly reducedin PC3 and in another prostate carcinoma cell lineDU145, which was subjected to fer-siRNA (Figure 3aand b). Interestingly, the level of CDK2 was notchanged in fer-siRNA-treated MDA-MB-231 cells(Figure 3c). In parallel, the levels of the CDK-activatingkinase (CAK) components – CDK7 and cyclin H - werealso not affected (Figure 3a), suggesting a lack of aneffect on the functional activation of CDKs (Kaldis,1999) upon ‘knock-down’ of Fer. In accordance withthis conclusion, we did not see any reduction in the invitro kinase activity of CDK4 or CDK6 in Fer-depletedcells (data not shown).

The levels of the negative regulators of CDKs –p21Cip1 and p27Kip1 (Coqueret, 2003), were also notaltered by the downregulation of Fer (Figure 3a).

To determine whether Fer regulates the accumulationof CDK2 at the transcriptional or post-transcriptionallevel, reverse transcription–polymerase chain reaction(RT–PCR) analysis was carried out and the levels of theCDK2 mRNA were determined in luc.-siRNA- and infer-siRNA-treated cells. This did not reveal a significant

Figure 3 Downregulation of CDK2 in Fer-depleted prostate carcinoma cells. (a) PC3 cells were left untreated or were subjected toluc.-siRNA or fer-siRNA. Levels of the indicated proteins were determined in a Western blot analysis. (b) DU145 cells were leftuntreated or treated with luc.-siRNA or fer-siRNA. Levels of the indicated proteins were determined using Western blot analysis.These profiles represent one out of three independent experiments that gave similar results. (c) The level of CDK2 is not changed infer-siRNA-treated MDA-MB-231 cells. MDA-MB-231 cells were either left untreated or were treated with luc.-siRNA or fer-siRNA.Levels of Fer, CDK2 and actin were determined using Western blot analysis. These profiles represent one out of three independentexperiments that gave similar results.


4197

Oncogene

difference between the levels of the CDK2 mRNA inthe two analysed samples (Figure 4a). Furthermore,simultaneous exposure of PC3 cells to both fer-siRNAand the translation inhibitor cycloheximide did notexacerbate the effect of fer-siRNA on the level ofCDK2, but rather alleviated this effect (Figure 4b, lanes2 and 4). These findings suggest that CDK2 is regulatedby Fer at the post-translational level in prostatecarcinoma cells.

pRB is hypophosphorylated in fer-siRNA-treated cellsCDK2 associates with cyclin E and cyclin A toconstitute kinase active complexes during G1- andS-phase progression (Brugarolas et al., 1999). These twocomplexes together with CDK4–6/cyclin D1 (Ezhevskyet al., 1997) are engaged in the onset and maintenance ofthe phosphorylation state of pRB. Phosphorylation ofpRB releases E2Fs (Reed, 1997; Yamasaki, 2003), whichin turn activate transcription of genes involved in theonset and regulation of DNA replication (Herrera et al.,1996). We therefore checked the phosphorylation stateof pRB in untreated and in siRNA-treated PC3 andMDA-MB-231 cells. Whereas in untreated and in luc.siRNA-treated cells, unphosphorylated and a pool ofphosphorylated pRB forms were clearly detected(Figure 5a), in fer-siRNA-treated cells, the phosphory-lated fraction of pRB (ppRB) was dramatically reduced(Figure 5a). The nearly complete dephosphorylation of

pRB in Fer-depleted PC3 cells indicated that bothCDK2 and CDK4/6 sites are dephosphorylated intreated cells. To directly examine this possibility, thephosphorylation levels of several sites were determined.Ser795 is a key regulatory site in pRB (Connell-Crowleyet al., 1997) and it is preferentially phosphorylated byCDK4 (Pan et al., 1998). Thr821, on the other hand, is apreferential CDK2 phosphorylation target (Zarkowskaand Mittnacht, 1997). The phosphorylation levels ofthese two sites were determined using specific anti-phospho antibodies. Whereas the phosphorylation levelof these two sites was only slightly changed upondepletion of either CDK4 or CDK2 alone (Figure 5b),both sites were dephosphorylated in the fer-siRNA-treated cells (Figure 5b). Moreover, simultaneousdepletion of CDK4 and CDK2 affected the phosphor-ylation level of only Ser795 but not the phosphorylationstate of Thr821 (Figure 5b). Similarly, knockdown ofFer significantly decreased the phosphorylation level oftwo other CDK2 phosphorylation sites Ser807 andSer811 (Connell-Crowley et al., 1997), whose phosphor-ylation was shown to be necessary for G1–S transition(Brugarolas et al., 1999) (Figure 5c). To substantiate thepivotal role of Fer in maintaining the phosphorylationstate of pRB in malignant cells, a similar analysis wasconducted in MDA-MB-231 cells. The phosphorylationlevels of Ser780, which is a primary phosphorylation siteof CDK4 (Kitagawa et al., 1996), and of Ser807/811were examined.Whereas knockdown of Fer significantlyaffected the phosphorylation states of these CDK4 andCDK2 phosphorylation targets, simultaneous depletionof CDK4 and CDK2 hardly affected the phosphoryla-tion levels of these sites (Figure 5d). Interestingly, thephosphorylation levels of Thr821 and Ser795 were notaffected by the knockdown of Fer or CDK4 and CDK2in MDA-MB-231 cells (data not shown), indicating thevariations in pRB regulation in different cell types.Hence, downregulation of Fer leads to the hypopho-sphorylation of pRB.

Knockdown of Fer leads to the hypophosphorylation ofPP1aMaintaining the phosphorylation state of pRB, inde-pendently of CDK6, CDK4 and CDK2, implied theinvolvement of Fer in an alternative pRB regulatingpathway. Whereas G1/S CDKs phosphorylate pRB ondefined sites (Kitagawa et al., 1996; Connell-Crowleyet al., 1997; Zarkowska and Mittnacht, 1997), theprotein phosphatase 1 (PP1) isoforms (Wang et al.,2001) de-phosphorylate these sites in pRB (Rubin et al.,2001). The balance between these two opposingregulatory mechanisms determines the phosphorylationand activation state of pRB (Berndt et al., 1997).Phosphorylation of Thr320 in PP1a suppresses thephosphatase activity of that enzyme and increased phos-phorylation level of Thr320 is inversely correlated withthe phosphatase activity of PP1a in vivo (Dohadwalaet al., 1994; Kwon et al., 1997; Liu et al., 1999).Conversely, dephosphorylation of the Thr320 siteinduces the phosphatase activity of PP1a toward pRB

Figure 4 The cellular level of CDK2 is regulated post-transcrip-tionally by Fer in prostate carcinoma cells. (a) Relative levels ofCDK2 RNA in total RNA from untreated, luc.-siRNA-treated andfrom fer-siRNA-treated cells were determined using quantitativeRT–PCR analysis. 18S ribosomal RNA served as an internalcontrol. (b) PC3 cells were treated with luc.-siRNA or with fer-siRNA in the absence or presence of cycloheximide. Levels of Ferand CDK2 in WCL were determined using Western blot analysis.These data represent one out of three independent experiments thatgave similar results.


4198

Oncogene

(Berndt et al., 1997). We therefore examined thephosphorylation state of PP1a on Thr320 in Fer-depleted cells. Knockdown of Fer in PC3 cells did notchange the cellular level of PP1a, but led to a profounddecrease in the phosphorylation level of PP1a on Thr320(Figure 6a). Similarly, knockdown of Fer decreased thephosphorylation level of PP1a in MDA-MB-231 cells(Figure 6b).

To corroborate the regulatory link between Ferand PP1a, a reciprocal experiment was carried out.Hemaglutinin (HA)-tagged Fer (HA-Fer) was stablyoverexpressed in MCF-7 breast carcinoma cells andthe effect of this ectopic expression on the PP1aThr320 phosphorylation level was examined. Over-expression of Fer upregulated the phosphorylation ofPP1a on Thr320 (Figure 6c), but did not affect thecellular level of PP1a. This implies a regulatory linkbetween Fer and PP1a.

PP1a associates with Fer and is downregulated inFer-overexpressing cellsThe effect of Fer depletion on the activation state ofPP1a prompted us to examine a direct interactionbetween Fer and PP1. The PP1 consensus binding sitehas been thoroughly characterized in plethora ofproteins and was accurately defined (Egloff et al.,1997). Presence of two PP1 docking motifs wasidentified in various PP1 binding proteins (Ayllonet al., 2002; Bennett and Alphey, 2002; Sugiyamaet al., 2002; Garcia et al., 2003). Analysis of the murineand human Fer amino-acid sequence revealed thepresence of two PP1 docking motifs in the kinasedomain of the two kinases and these motifs areseparated by 42 aa from each other (Figure 7a). Asexpected, the two PP1 binding motifs are present also inthe kinase domain of the meiotic testes specific Fervariant – FerT (Figure 7a) (Fischman et al., 1990).

Figure 5 pRB is underphosphorylated in fer-siRNA-treated cells. (a) WCL prepared from untreated or luc.-siRNA or fer-siRNA-treated PC3 cells were resolved in a 7–17% gradient SDS-PAGE. Levels of Fer, pRB, ppRB and actin were determined. (b and c) PC3were left untreated or were treated with a luc.-siRNA, with a combination of cdk2 and cdk4 siRNAs, with a cdk4-siRNA, with a cdk2-siRNA or with a fer-siRNA. Levels of the indicated proteins or pRB phosphorylation sites were determined using Western blotanalysis. The data represent one out of three independent experiments that gave similar results. (d) WCL were prepared fromuntreated, luc.-siRNA-treated, cdk2 and cdk4-siRNA-treated or fer-siRNA-treated MDA-MB-231 cells. Levels of indicated proteinsand pRB phosphorylation sites were determined in a Western blot analysis. The data represent one out of three independentexperiments that gave similar results.


4199

Oncogene

To examine whether Fer and PP1 can indeed associatewith each other in vivo, MDA-MB-231 cells weretransiently transfected with a vector encoding a Myc-tagged PP1a catalytic subunit or with the pCMVexpression vector alone. Whereas the endogenous Ferco-precipitated with Myc-PP1a from cells expressing theectopic phosphatase, it was not precipitated by the anti-Myc antibody from lysates of cells transfected with theexpression vector alone (Figure 7b). To corroborate thisfinding, MCF-7 cells stably overexpressing HA-Fer weretransiently transfected with the Myc-PP1a encodingvector and Myc-PP1a was immunoprecipitated from celllysates using anti-Myc antibody. As was found inMDA-MB-231 cells, Fer co-precipitated with the Myc-PP1a protein (Figure 7c), implying the association of

Fer with PP1a in vivo. To further establish the role ofFer in PP1 regulation, MCF-7 cells were transiently co-transfected with a vector expressing Myc-PP1a togetherwith a vector encoding native HA-Fer, or with a vectorencoding a mutated Fer in which a highly conservedphenylalanine residue in one PP1 binding motif(Katayama et al., 2001) was replaced with alanine(HA-FerF606A). This mutation compromised the abilityof Fer to bind Myc-PP1a (Figure 7d) and ectopic expres-sion of the Fer mutant downregulated the phosphoryla-tion of PP1a on Thr320 (Figure 7e). This decrease wasonly seen following the expression of the HA-FerF606Amutant but not with the native enzyme (Figure 7e) orwith a kinase-inactive Fer mutant, FerY715F, whichlacks an autophosphorylation and substrate phosphor-ylation activity (Ben-Dor et al., 1999; Salem et al., 2005),but retains the ability to bind PP1 (Figure 7d). Theseresults corroborate the regulatory link between Ferand PP1.

To examine directly the effect of Fer on thephosphatase activity of PP1a, Myc-PP1a was transientlyexpressed in MCF-7 cells that ectopically express HA-Fer and in MCF-7 cells that harbor the expressionvector pBabe alone (Figure 8a). These two cell popula-tions accumulate different levels of Fer (Figure 8a) butexhibit similar patterns of cell-cycle progression (datanot shown). Myc-PP1a was immunoprecipitated fromcell lysates and was then subjected to an in vitrophosphatase assay. Incubation of the precipitatedenzymes for 90min in the presence of a syntheticphosphatase substrate revealed a twofold decrease inthe phosphatase activity of PP1a that was precipitatedfrom MCF-7-HA-Fer extracts (Figure 8b). Thus, over-expression of Fer decreases the phosphatase activityof PP1a.

Discussion

G1 progression is a regulatory phase wherein cellsrespond to and process growth and proliferationstimulatory signaling that lead to G1–S transition(Yamasaki, 2003). These steps are frequently deregu-lated in malignant cells and are therefore potentialtargets for cancer intervening drugs (Vermeulen et al.,2003). However, recent findings suggested a highfunctional redundancy of the G1–S regulatory compo-nents, like CDK4 and CDK2 (Berthet et al., 2003;Ortega et al., 2003; Tetsu and McCormick, 2003;Malumbres et al., 2004; Sherr and Roberts, 2004). Thisnew insight raises the need for identification of newregulators of the G1–S transition phase.

In the current work, we showed that decreasing thelevel of Fer in prostate and breast carcinoma cells,which bear a functional pRB, led to their cell-cyclearrest at the G0/G1 phase. No effect was seen on thepercentage of apoptotic cells in the treated cultures. Ferwas shown to activate the pro-oncogenic transcriptionfactor Stat3 (Priel-Halachmi et al., 2000). However, theessential role of Fer in the currently studied malignant

Figure 6 Fer modulates the phosphorylation state of PP1a. (a)Hypophosphorylation of PP1a in fer-siRNA-treated PC3 cells.PC3 cells were either left untreated or treated with luc.-siRNA orfer-siRNA. WCL were subjected to anti-Fer, anti-PP1a or anti-pThr320 PP1a in a Western blot analysis. (b) Hypophosphorylationof PP1a in fer-siRNA-treated MDA-MB-231 cells. MDA-MB-231cells were either left untreated or were treated with luc.-siRNA orfer-siRNA. WCL were reacted with anti-Fer, anti-PP1a or anti-pThr320 PP1a in a Western blot analysis. (c) Increased phosphor-ylation of PP1a in Fer-overexpressing cells. WCL from MCF-7cells infected with the pBabe vector and from MCF-7 cellsectopically expressing HA-Fer were reacted with anti-pThr320PP1a, anti-PP1a, anti-Fer and anti-HA antibodies in a Westernblot analysis. The data represent one out of three independentexperiments that gave similar results.


4200

Oncogene

cells may not be associated with the activation ofStat3, as that regulator does not accumulate in itsactivated form in PC3 and MDA-MB-231 cells (data notshown).

The proliferation arrest of Fer-depleted PC3 andMDA-MB-231 cells resulted most probably from asustained suppressive activity of pRB (Reed, 1997;Yamasaki, 2003), whose phosphorylation level (Chenget al., 1999) was significantly reduced in the treated cells.

Hypophosphorylation of pRB reflected the concomitantdephosphorylation of both CDK4 (Ser795 and Ser780)and CDK2 sites (Ser807, Ser811 and Thr821) in the Ferdeprived cells. Interestingly, whereas knockdown of Ferin PC3 cells affected all the analyzed pRB phosphoryla-tion sites, in MDA-MB-231 cells only some of theexamined sites (Ser780, Ser807 and Ser811) wereaffected. This implies a diverse regulation of pRB indifferent cell types.

Figure 7 Interaction between Fer and PP1a in malignant cells. (a) Consensus PP1a binding motifs in the kinase domains of the humanFer, the murine Fer and FerT proteins. Numbers denote the position of each motif in the Fer proteins (see text for details). (b) Ferassociates with PP1a in malignant cells. MDA-MB-231 cells were transfected with a Myc-PP1a expression vector (lanes 1 and 3) orwith the expression vector alone (lanes 2 and 4). Myc-PP1a was immunoprecipitated from cell lysates using anti-Myc antibodies.Precipitated proteins (lanes 1 and 2) and WCL (lanes 3 and 4) were reacted with anti-Fer antibodies in a Western blot analysis.Membranes were then re-probed with anti-Myc antibody. (c) MCF-7 cells overexpressing HA-Fer were transfected with a Myc-PP1aplasmid (lanes 1 and 3) or with the expression vector alone (lanes 2 and 4). Myc-PP1a was immunoprecipitated from cell lysates usinganti-Myc antibodies. Precipitated proteins (lanes 1 and 2) and WCL (lanes 3 and 4) were reacted with anti-Fer antibodies in a Westernblot analysis. Membranes were then re-probed with anti-Myc and anti-HA antibody. (d) MCF-7 cells were transiently co-transfectedwith a pcDNA3 vector encoding Myc-PP1a together with a vector encoding the HA-FerY715F mutant (lane 1), with a pcDNA3 vectorand a vector encoding HA-FerY715F (lane 2), with vectors encoding PP1a and the HA-FerF606A mutant (lane 3) or with vectorsencoding Myc-PP1a and the native HA-Fer (lane 4). Myc-PP1a was immunoprecipitated from WCL and the co-precipitation of theHA-Fer variants was determined. (e) MCF-7 cells were transiently transfected with a pCDNA3 vector (lane 1), with a vector encodingHA-Fer (lane 2), a vector encoding HA-FerY715F (lane 3) or with a vector encoding the HA-FerF606A mutant (lane 4). Thephosphorylation level of PP1a on Thr320 was determined in a Western blot analysis. The data represent one out of three independentexperiments that gave similar results.


4201

Oncogene

Depletion of either CDK4 or 2 alone did notsignificantly affect the phosphorylation state of thepRB target sites. Moreover, simultaneous knockdownof these two key cell-cycle regulators in PC3 cells led tothe dephosphorylation of Ser795 in pRB, but did notprofoundly affect Thr821, which is a preferential targetfor CDK2 (Zarkowska and Mittnacht, 1997). Theunchanged phosphorylation level of Thr821 in CDK4/CDK2-depleted PC3 cells, and of Ser780, Ser807 andSer811 in CDK4/CDK2-depleted MDA-MB-231 cells,could result from the activity of the CDK1/cyclin Acomplex, which would compensate for the lack ofCDK2/cyclin E and CDK2/cyclin A kinases in thesetreated cells (Chow et al., 2003).

Hypophosphorylation on pRB in Fer-depleted cellscould result from the malfunctioning of either positive

or negative regulators of G1 progression (Cheng et al.,1999). Of the known pRB regulators that were analysedin this work, only the level of CDK2 was significantlyaffected by the downregulation of Fer in PC3, andDU145 prostate carcinoma cells. However, the level ofCDK2 was not affected in Fer-depleted MDA-MB-231cells. The level of CDK2 is not significantly changedalong the cell cycle (Vermeulen et al., 2003); thus, thedependence of the CDK2 level on the presence ofFer is unique and could be specific to prostatecarcinoma cells.

Fer was shown to modulate the cytoplasmic andnuclear levels of b-catenin (Miravet et al., 2003; Piedraet al., 2003), an effect that could link this kinase to theupregulation of the key cell-cycle regulator cyclin D1(Shtutman et al., 1999; Sellin et al., 2001). However, thelack of an effect on the level of cyclin D1 in Fer-depletedPC3 cells excludes the role of the b-catenin pathway inmediating the activity of Fer in these cells. Similarly, wedid not detect an effect of Fer depletion on the kinaseactivity of CDK4/cyclin D1 or CDK2/cyclin E com-plexes (data not shown). This coincides with the lack ofa change in the cellular levels of the CDK activatorsCDK7 and cyclin H (Kaldis, 1999) or the inhibitorsp21Cip1 and p27Kip1 (Coqueret, 2003) in Fer-depletedcells.

All these observations prompted us to examine therole of Fer in modulating other activities that directlyregulate the functioning of pRB. PP1 is a family ofserine/threonine phosphatases that regulate a widevariety of cellular processes. Among these, PP1 activatespRB by removing neutralizing phosphate groups fromkey regulatory moieties in that tumor suppressingprotein (Alberts et al., 1993; Durfee et al., 1993; Ludlowet al., 1993). Indeed, knockdown of Fer led to thedephosphorylation and activation of PP1a, one of thePP1 variants that were shown to dephosphorylate andactivate pRB (Berndt et al., 1997). The pRB-associatedPP1a pool is inactivated by CDK mediated phosphor-ylation on Thr320 (Liu et al., 1999) and depletion of Ferled to the hypophosphorylation of PP1a on thatregulatory site. Conversely, overexpression of Ferincreased the phosphorylation of PP1a on Thr320.Thus, Fer sustains the phosphorylation level of PP1aon Thr320, and thereby supports the downregulation ofthat phosphatase. The ability of Fer to suppress theactivity of PP1a was corroborated by the decreased invitro phosphatase activity of PP1a from Fer-overexpres-sing cells. This downregulation of PP1a contributesmost probably to the neutralization of the pRBsuppressive activity, during G1 progression in malignantcells.

The two PP1 docking motifs that were revealed in thekinase domain of Fer most probably mediate theinteraction of this kinase with PP1a in vivo. Inaccordance with this notion, mutation in one of thePP1 binding motifs turned Fer into a dominant-negativemutant that led to a decrease in the phosphorylationlevel of PP1a. This effect of the HA-FerF606A mutantmight be attributed to the ability of an exogenousdominant-negative Fer to form hetero-oligomers with

Figure 8 Fer decreases the phosphatase activity of PP1a. (a)MCF-7 cells overexpressing HA-Fer or infected with the pBabevector alone were transiently transfected with a Myc-PP1aexpression plasmid. WCL were reacted with anti-Fer, anti-HAand anti-Myc antibody in a Western blot analysis. (b) Myc-PP1awas immunoprecipitated from the lysates in (a) using anti-Mycantibody. Precipitates were subjected to phosphatase in vitro assayfor the indicated periods of time. Squares represent MCF-7-pBabecells and triangles denote MCF-7-HA-Fer cells. Values representmeans7s.e. (Po0.05 (n¼ 3) when HA-Fer is compared to pBabeafter 90min incubation at 301C).


4202

Oncogene

the endogenous Fer (Craig et al., 2001) and to interferewith its activity (Orlovsky et al., 2000). The associationof Fer with PP1a might enhance the CDK dependentphosphorylation of PP1a on Thr320. Alternatively, Fercould attenuate the auto- or trans-dephosphorylation ofPP1a on Thr320 (Dohadwala et al., 1994). It should benoted that we did not detect tyrosine-phosphorylatedPP1a in Fer-expressing cells (data not shown), suggest-ing that the regulatory effect of Fer on PP1a is kinaseactivity independent, and may result from the physicalinteraction between these two proteins. Similarly, Ferbinds ERK1/2 and maintains the phosphorylation statesof these kinases, independently of the Fer tyrosinekinase activity (Salem et al., 2005).

The functioning of the PP1 phosphatases is regulatedby a wide and diverse set of proteins in eucaryotic cells(Ceulemans et al., 2002; Ceulemans and Bollen, 2004).However, Fer is the first intracellular tyrosine kinaseshown to associate with PP1 and modulate its activity.The ability of the Fer kinase to bind and modulate PP1seems to be harnessed by malignant cells for manipulat-ing the balance between ‘signaling’ and ‘anti-signaling’cascades in which PP1 phosphatases are involved(Honkanen and Golden, 2002).

Interestingly, unlike in prostate and breast carcinomacells, the functioning and accumulation level of Fer wasfound not to be essential in proliferating primaryfibroblasts (Craig and Greer, 2002). Moreover, we failedto detect an interaction between Fer and PP1 in primarymice embryonic fibroblasts (data not shown). Thissuggests that the regulatory role of Fer in cell-cycleprogression might differ between normal and malignantcells. The role of Fer in malignant cells may be related tothe tendency of these cells to restrain phosphataseactivities in growth-promoting pathways (Honkanenand Golden, 2002).

Further studies should reveal what other cell-cyclestages are affected by the knockdown of Fer inmalignant cells, and whether Fer could serve as a novelintervening target for cancer therapy.

Materials and methods

Preparation of siRNADouble-stranded RNAs, 21 nt long, were synthesized byDharmacon Research (Lafayette, CO, USA). The targetedsequence was derived from the human fer cDNA (accessionno. J03358) 50-AAA GAA ATT TAT GGC CCT GAG-30 (nt84–104 relative to the fer mRNA translation initiation codon).Custom SMART pool siRNA to target the human cdk2mRNA and the human cdk4 mRNA was purchased fromDharmacon. Selected siRNA sequences were submitted to aBLAST search against the human genome sequence to ensurespecificity of the siRNA. A sequence targeting firefly (Photinuspyralis) luciferase (luc.) gene (accession no. X65324) was usedas a control.

Cell lines and transfection of cellsHuman PC3 and MDA-MB-231 cells were grown at 371C inRPMI and MCF-7 cells were grown in DMEM (Gibco-

Invitrogen, USA). All growth media were supplemented with10% fetal calf serum (FCS).For siRNA transfection, 1.5� 105 PC3 cells were seeded in

6-cm Petri dish 1 day before transfection. A 30ml portion ofsiRNA from 20 mM stock solution was mixed with 300 mlOptiMEM (Gibco-Invitrogen) medium, and then incubated atroom temperature for 5min. A 20ml portion of Metafectene(Biontex, Germany) was mixed with 160 ml OptiMEM in aseparate tube and then incubated at room temperature for5min. The two mixtures were then combined, gently mixedand incubated for another 15min at room temperature, for theformation of lipid–siRNA complexes. Growth medium wasremoved from the cells, which were then covered with 1290 mlOptiMEM, 200 ml FCS and 510 ml lipid–siRNA mixture. FinalsiRNA concentration in the Petri dish was 300 nM. After 8 h,1800 ml OptiMEM and 200 ml FCS were added and cells wereharvested 72 h after transfection.For siRNA transfection, 1.4� 105 MDA-MB-231 cells were

seeded in 6-cm Petri dish 1 day before transfection. OptiMEM(300 ml) was mixed with 30ml of 20 mM siRNA duplex andwith 7ml of PolyMAG (Magnetofection-Chemicell GmBH,Germany) and the mixture was left for 5min at room temperature.After addition of 1663ml OptiMEM, the entire mixture wasadded to the cells and the culture dish was laid on amanufacturer supplied magnet plate (MagnetoFACTOREplate) for an additional 15min at 371C. The transfectionmixture was then removed and the transfected cells weregrown in RPMI supplied with 10% FCS, for 72 h.For transient overexpression of PP1a, 7� 105 MDA-MB-

231 or 1� 106 MCF-7 cells were seeded in 10 cm plates 1 daybefore transfection. Each transfection was performed bymixing 12 mg DNA with 12 ml PolyMAG in 200ml OptiMEMmedium and incubated for 5min at room temperature. Afteraddition of 5ml OptiMEM medium, the entire mixture wasadded to the cells and the culture dish was put on a magnetplate for additional 15min at 371C. The transfection mixturewas then removed and the transfected cells were grown inRPMI supplemented with 10% FCS. Cells were harvested forprotein extraction 48 h after beginning of transfection.For transient overexpression of HA-Fer, HA-FerF606A or

pcDNA3 expression vector alone, PC3 cells were seeded in10 cm plates 1 day before transfection. Each transfection wasperformed by mixing 15 mg DNA in 1ml OptiMEM mediumand mixture was left for 5min at room temperature. A 30 mlportion of Lipofectamine 2000 (Invitrogen, USA) was mixedwith 1ml OptiMEM in a separate tube and incubated at roomtemperature for 5min. The two mixtures were then combined,gently mixed and incubated for another 30min at roomtemperature. After addition of 8ml OptiMEM medium, theentire mixture was added to the cells for 6 h at 371C. Thetransfection mixture was then removed and the transfectedcells were grown in RPMI supplemented with 10% FCS. Cellswere harvested for protein extraction 48 h after beginning oftransfection.MCF-7 cells overexpressing HA-Fer were established as

follows: 293T cells were co-transfected with 10 mg pHAFer(Salem et al., 2005) and 10 mg of the pCL-Ampho packagingplasmid (Naviaux et al., 1996). Production of viruses wasperformed as described before (Morgenstern and Land, 1990).For retroviral infection, 2� 105 MCF-7 cells were seeded in6 cm plates and were then infected with undiluted 293Tsupernatants, in the presence of 8mg/ml polybrene (Sigma,USA) for 6 h. This process was repeated three times insuccession, producing an MOI (multiplicity of infection). Twodays later, cells were split and selected in puromycin up to aconcentration of 2 mg/ml. Cells transfected with the pBabepuro vector alone were used as a control.


4203

Oncogene

For treatment with cycloheximide, PC3 cells transfectedwith siRNA for 72 h, were subjected to 10 mg/ml cycloheximidefor an additional 4 h at the end of the transfection process.

Quantitation of viable cellsRelative number of viable cells in a given culture wasdetermined using the MTT assay. Cultures were assayed forthe production of formazan by the reduction of MTT. At 72 hafter transfection in 96-well plates, growth medium wasremoved and cells were incubated for 2 h with 100 ml MTTsolution (5mg/ml MTT dissolved in phosphate-buffered saline(PBS)) at 371C. Cells were then lysed with 50% N,N-dimethylformamide and the levels of the formazan reduction productwere determined by measuring absorbance at 570 nm in acomputer-controlled microplate analyser.

Flow-cytometry analysisCells were rinsed twice with cold PBS, and then harvested bytreatment with trypsin. Suspended cells were washed twice in10ml cold PBS and centrifuged at 500 � g for 5min. The pelletwas resuspended in 200 ml cold PBS and cells were stained inthe dark with 200 ml propidium iodide (PI) solution (20mM

Tris, pH 8, 1mM NaCl, 0.1% (v/v) NP-40, 1.4mg/ml RNaseA, 0.05mg/ml PI) for 30min at 371C. The total cellular DNAcontent was determined using a Becton Dickinson flowcytometer (FACSCalibur), Multiple software and ModFitLT software.

BrdU incorporation assayPC3 cells were grown in 6-cm Petri dish and 72 h aftertransfection with RNAi, were incubated for 30min with 10mMBrdU (Boehringer Mannheim, Germany). Incorporation ofBrdU was determined using a Becton Dickinson anti-BrdUantibody (347580), according to the manufacturer’s protocol.Cells were also stained with 5mg/ml PI solution. The cellularDNA content was determined using a Becton Dickinson flowcytometer (FACSCalibur) and Multiple software.

ImmunoprecipitationProtein lysates (700 mg) were incubated overnight at 41C with1:100 diluted anti-c-Myc antibody (13–2500, Zymed, USA).Antigen–antibody complexes were precipitated with protein A-Sepharose (Amersham Biosciences, UK) after 2 h incubationat 41C. Precipitates were washed three times with TGET buffer(20mM Tris-HCl, 10% glycerol, pH 7.5, EDTA 1mM, 0.1%Triton X-100). The first wash was carried out with TGETbuffer supplemented with 150mM NaCl, the second wash wasdone with TGET containing 75mM NaCl, and the last washwas with TGET devoid of NaCl.

Western blot analysisWCL were prepared as described before (Salem et al., 2005).For Western blot analysis, 20–30 mg protein from each samplewas resolved in 7–12% SDS–polyacrylamide gel electrophor-esis (SDS–PAGE). Electroblotted proteins were detected usingspecific anti-Fer antibodies (Priel-Halachmi et al., 2000), anti-CDK2 (sc-6248, Santa Cruz, USA), anti-CDK1 (sc-54, SantaCruz), anti-CDK4 (610148, BD Biosciences, USA), anti-CDK6 (554086, BD Biosciences), anti-CDK7 (sc-856, SantaCruz), anti-cyclin A (sc-751, Santa Cruz), anti-cyclin D1(sc-6281, Santa Cruz), anti-cyclin E (sc-247, Santa Cruz), anti-cyclin H (sc-1662, Santa Cruz), anti-pRB (sc-102, Santa Cruz),anti-phosphosthreonine821-pRB (pThr821) (44–582, Biosource,USA), anti-phosphoserine795-pRB (pSer795) (sc-7986-R, SantaCruz), anti-phosphoserine780-pRB (pSer780) (9307, Cell Signal-ing, USA), anti-phosphoserine807/811-pRB (pSer807/811) (9308,

Cell Signaling), anti-PP1a (2582, Cell Signaling), anti-PP1 (sc-443, Santa Cruz), anti-phosphothreonine320-PP1a (pThr320)(2581, Cell Signaling), anti-p21Cip1 (sc-6246, Santa Cruz),anti-p27Kip1 (sc-528, Santa Cruz), anti-HA (NMS-101R,Covance, USA), anti-Myc (13–2500, Zymed) and anti-actinantibody (A 4700, Sigma). Bound antibodies were visualizedusing chemiluminescence reaction (Pierce, USA). The blotswere scanned using an optical scanner (Umax Astra 3400) andthe optical density of each band was compiled using theImageJ software application.

Quantitative RT–PCR analysisWhole-cell RNA was extracted from PC3 cells using TRIReagent (Molecular Research Center, ICN, USA) followingthe manufacturer’s instructions. Total RNA (25 ng) wasreverse-transcribed using AccessQuick RT–PCR System (Pro-mega, USA) with specific oligonucleotides derived from thehuman CDK2 mRNA: forward primer 50-CTA GCT TTCTGC CAT TCT CAT C-30 and reverse primer 50-GAA GAGCTG GTC AAT CTC AGA A-30. The selected primers arederived from separated exons (4 and 6), which flank an intronof the CKD2 gene. PCR was run for 27 cycles, which wasfound to be optimal for quantitative comparison of the CDK2mRNA levels. The expected 275-bp-long PCR product isspecific to the mature CDK2 mRNA, and can be differentiatedfrom a contamination of genomic DNA PCR products. 18Sribosomal RNA primer pairs (Quantum RNA, Ambion, USA)were used to yield a 489 bp fragment derived from the 18SrRNA (ribosomal RNA) as an internal control. PCR productswere separated in 1.4% agarose gel and visualized withethidium bromide.

Construction of plasmidsThe pCMV-c-Myc-PP1a expression plasmid was constructedby amplifying a human PP1a cDNA clone (RZDP-IR-AUp969F0224D) with a forward primer 50-AACCCGAATTCTGTCCGACAGCGAGAAGC-30, which bears at its50-end an EcoR1 site, and a backward primer 50-AAGGACTCGAGCTATTTCTTGGCTTTGGCGG-30, which bears anXho1 site at is 30 end. The amplified fragment was clonedinto an EcoR1–Xho1 cut pCMV-c-Myc vector (BD ClonetechInc., USA).Mutated Fer (FerF606A) was constructed by ‘site-directed

mutagenesis’ method as follows. Phe606 was changed to Ala asfollows: primers A and B are reverse complement primers thatcontain the mutated residue: A, 50-AAAATAAAGGCTCTACAGGAAGC-30; and B, 50-GCTTCCTGTAGAGCCTTTATTTT-30. Primers C and D are the external 50 and 30

primers designed for the amplification of the mutated cDNAportions: C, 50-GCATGACTGGTACCATGGTG-30; and D,50-TTCTTGCTCTCGAGATACAACA-30. The two overlap-ping cDNA portions were separately amplified from a murinefer-cDNA and were than mixed, denatured and allowed tocomplete a mutated double-stranded fragment using Pfu UltraDNA polymerase. The obtained mutated fer fragment wasthen amplified and cloned into a corresponding fer-cDNA togive rise to a mutated FerF606A.

In vitro phosphatase assayMCF-7 cells overexpressing HA-Fer or harboring the retro-viral vector pBabe as control were transiently transfected witha Myc-PP1a expression vector. WCL were prepared from thecells 48 h after transfection in a lysis buffer that lackedphosphatase inhibitors. Myc-PP1a was immunoprecipitatedfrom 1mg lysate with anti-c-Myc antibody as described above.The immunoprecipitated Myc-PP1a was rinsed twice with


4204

Oncogene

buffer A (50mM Tris-HCl, 5% glycerol, 50mM NaCl, 1mM

dithiothreitol, 1mM MgCl2, 3mM MnCl2). The immunecomplexes (each containing 150 ng protein) were thensuspended in a phosphatase activity buffer (buffer A contain-ing 0.1mg/ml purified BSA) and were then incubated at 301Cin the presence of the synthetic substrate 5mM pNPP disodiumhexahydrate (Sigma). Dephosphorylation of the substrate wasquantitated by measuring absorbance at 405 nm in a compu-ter-controlled microplate analyser.

Acknowledgements

We thank Ms Sharon Victor for typing this manuscript andMr Uri Karo for assistance with the flow-cytometry analysis.This work was supported by grants from CaPCURE Israel, theMinistry of Health Chief Scientist’s Research Fund, DeborahSilin and Samuel-Heinrich Winograd Research Fund, theWeinkselbaum Family Research Fund, and a grant from theHorowitz Foundation.

References

Alberts AS, Thorburn AM, Shenolikar S, Mumby MC,Feramisco JR. (1993). Proc Natl Acad Sci USA 90:388–392.

Allard P, Zoubeidi A, Nguyen LT, Tessier S, Tanguay S,Chevrette M et al. (2000). Mol Cell Endocrinol 159: 63–77.

Ayllon V, Cayla X, Garcia A, Fleischer A, Rebollo A. (2002).Eur J Immunol 32: 1847–1855.

Ben-Dor I, Bern O, Tennenbaum T, Nir U. (1999). CellGrowth Differ 10: 113–129.

Bennett D, Alphey L. (2002). Nat Genet 31: 419–423.Berndt N, Dohadwala M, Liu CW. (1997). Curr Biol 7:375–386.

Berthet C, Aleem E, Coppola V, Tessarollo L, Kaldis P.(2003). Curr Biol 13: 1775–1785.

Brugarolas J, Moberg K, Boyd SD, Taya Y, Jacks T, Lees JA.(1999). Proc Natl Acad Sci USA 96: 1002–1007.

Ceulemans H, Bollen M. (2004). Physiol Rev 84: 1–39.Ceulemans H, Stalmans W, Bollen M. (2002). BioEssays 24:371–381.

Cheng M, Olivier P, Diehl JA, Fero M, Roussel MF, RobertsJM et al. (1999). EMBO J 18: 1571–1583.

Chow JP, Siu WY, Fung TK, Chan WM, Lau A, Arooz Tet al. (2003). Mol Biol Cell 14: 3989–4002.

Connell-Crowley L, Harper JW, Goodrich DW. (1997). MolBiol Cell 8: 287–301.

Coqueret O. (2003). Trends Cell Biol 13: 65–70.Craig AW, Greer PA. (2002). Mol Cell Biol 22: 6363–6374.Craig AW, Zirngibl R, Williams K, Cole LA, Greer PA.(2001). Mol Cell Biol 21: 603–613.

Dohadwala M, da Cruz e Silva EF, Hall FL, Williams RT,Carbonaro-Hall DA, Nairn AC et al. (1994). Proc Natl AcadSci USA 91: 6408–6412.

Durfee T, Becherer K, Chen PL, Yeh SH, Yang Y, KilburnAE et al. (1993). Genes Dev 7: 555–569.

Egloff MP, Johnson DF, Moorhead G, Cohen PT, Cohen P,Barford D. (1997). EMBO J 16: 1876–1887.

Ezhevsky SA, Nagahara H, Vocero-Akbani AM, Gius DR,Wei MC, Dowdy SF. (1997). Proc Natl Acad Sci USA 94:10699–10704.

Fischman K, Edman JC, Shackleford GM, Turner JA, RutterWJ, Nir U. (1990). Mol Cell Biol 10: 146–153.

Garcia A, Cayla X, Guergnon J, Dessauge F, Hospital V,Rebollo MP et al. (2003). Biochimie 85: 721–726.

Greer P. (2002). Nat Rev Mol Cell Biol 3: 278–289.Hao Q-L, Heisterkamp N, Groffen J. (1989). Mol Cell Biol 9:1587–1593.

Herrera RE, Sah VP, Williams BO, Makela TP, Weinberg RA,Jacks T. (1996). Mol Cell Biol 16: 2402–2407.

Honkanen RE, Golden T. (2002). Curr Med Chem 9:2055–2075.

Iwanishi M, Czech MP, Cherniack AD. (2000). J Biol Chem275: 38995–39000.

Kaldis P. (1999). Cell Mol Life Sci 55: 284–296.Katayama H, Zhou H, Li Q, Tatsuka M, Sen S. (2001). J Biol

Chem 276: 46219–46224.Kim L, Wong TW. (1998). J Biol Chem 273: 23542–23548.Kitagawa M, Higashi H, Jung HK, Suzuki-Takahashi I,Ikeda M, Tamai K et al. (1996). EMBO J 15:7060–7069.

Kogata N, Masuda M, Kamioka Y, Yamagishi A, Endo A,Okada M et al. (2003). Mol Biol Cell 14: 3553–3564.

Kwon YG, Lee SY, Choi Y, Greengard P, Nairn AC. (1997).Proc Natl Acad Sci USA 94: 2168–2173.

Letwin K, Yee S-P, Pawson T. (1988). Oncogene 3:621–627.

Liu CW, Wang RH, Dohadwala M, Schonthal AH,Villa-Moruzzi E, Berndt N. (1999). J Biol Chem 274:29470–29475.

Ludlow JW, Glendening CL, Livingston DM, DeCarprio JA.(1993). Mol Cell Biol 13: 367–372.

Lunter PC, Wiche G. (2002). Biochem Biophys Res Commun296: 904–910.

Malumbres M, Sotillo R, Santamaria D, Galan J, Cerezo A,Ortega S et al. (2004). Cell 118: 493–504.

Miravet S, Piedra J, Castano J, Raurell I, Franci C, Dunach Met al. (2003). Mol Cell Biol 23: 7391–7402.

Morgenstern JP, Land H. (1990). Nucleic Acids Res 18:3587–3596.

Naviaux RK, Costanzi E, Haas M, Verma IM. (1996). J Virol70: 5701–5705.

Orlovsky K, Ben-Dor I, Priel-Halachmi S, Malovany H, NirU. (2000). Biochemistry 39: 11084–11091.

Ortega S, Prieto I, Odajima J, Martin A, Dubus P, Sotillo Ret al. (2003). Nat Genet 35: 25–31.

Pan W, Sun T, Hoess R, Grafstrom R. (1998). Carcinogenesis19: 765–769.

Penhallow RC, Class K, Sonoda H, Bolen JB, Rowley RB.(1995). J Biol Chem 270: 23362–23365.

Piedra J, Miravet S, Castano J, Palmer HG, Heisterkamp N,Garcia dH et al. (2003). Mol Cell Biol 23: 2287–2297.

Priel-Halachmi S, Ben-Dor I, Shpungin S, Tennenbaum T,Molavani H, Bachrach M et al. (2000). J Biol Chem 275:28902–28910.

Reed SI. (1997) In: Kastern M (ed). Checkpoint Controls andCancer, Cancer Surveys, 29. Cold Spring Harbor LaboratoryPress: Cold Spring Harbor, NY. pp 7–23.

Rosato R, Veltmaat JM, Groffen J, Heisterkamp N. (1998).Mol Cell Biol 18: 5762–5770.

Rubin E, Mittnacht S, Villa-Moruzzi E, Ludlow JW. (2001).Oncogene 20: 3776–3785.

Salem Y, Shpungin S, Pasder O, Pomp O, Taler M, MalovaniH et al. (2005). Cell Signal 17: 341–353.

Sellin JH, Umar S, Xiao J, Morris AP. (2001). Cancer Res 61:2899–2906.


4205

Oncogene

Sherr CJ, Roberts JM. (2004). Genes Dev 18: 2699–2711.Shtutman M, Zhurinsky J, Simcha I, Albanese C, D’Amico M,Pestell R et al. (1999). Proc Natl Acad Sci USA 96:5522–5527.

Sugiyama K, Sugiura K, Hara T, Sugimoto K, Shima H,Honda K et al. (2002). Oncogene 21: 3103–3111.

Tetsu O, McCormick F. (2003). Cancer Cell 3: 233–245.

Vermeulen K, Van Bockstaele DR, Berneman ZN. (2003). CellProlif 36: 131–149.

Wang RH, Liu CW, Avramis VI, Berndt N. (2001). Oncogene20: 6111–6122.

Yamasaki L. (2003). Cancer Treat Res 115: 209–239.Zarkowska T, Mittnacht S. (1997). J Biol Chem 272:12738–12746.


4206

Oncogene

Downregulation of Fer induces PP1 activation and cell-cycle arrest in malignant cells

Documents

Transcript of Downregulation of Fer induces PP1 activation and cell-cycle arrest in malignant cells