Involvement of MINK, a Ste20 family kinase, in Ras oncogene-induced growth arrest in human ovarian surface epithelial cells.

Abstract

The ability of activated Ras to induce growth arrest of human ovarian surface epithelial (HOSE) cells via induction of the cyclin-dependent kinase inhibitor p21(WAF1/CIP1) has been used to screen for Ras pathway signaling components using a library of RNA interference (RNAi) vectors targeting the kinome. Two known Ras-regulated kinases were identified, phosphoinositide 3-kinase p110alpha and ribosomal protein S6 kinase p70(S6K1), plus the MAP kinase kinase kinase kinase MINK, which had not previously been implicated in Ras signaling. MINK is activated after Ras induction via a mechanism involving reactive oxygen species and mediates stimulation of the stress-activated protein kinase p38 MAPK downstream of the Raf/ERK pathway. p38 MAPK activation is essential for Ras-induced p21(WAF1/CIP1) upregulation and cell cycle arrest. MINK is thus a distal target of Ras signaling in the induction of a growth-arrested, senescent-like phenotype that may act to oppose oncogenic transformation in HOSE cells.

Full text

Page 1

Molecular Cell, Vol. 20, 673–685, December 9, 2005, Copyright ª2005 by Elsevier Inc.DOI 10.1016/j.molcel.2005.10.038
Involvement of MINK, a Ste20 Family Kinase,
in Ras Oncogene-Induced Growth Arrest in Human
Ovarian Surface Epithelial Cells
Barbara Nicke,1,5,7Julie Bastien,1,7Sophia J. Khanna,1
Patricia H. Warne,1Victoria Cowling,1Simon J. Cook,4
Gordon Peters,2Oona Delpuech,3Almut Schulze,3
Katrien Berns,6Jasper Mullenders,6
Roderick L. Beijersbergen,6Rene ´ Bernards,6
Trivadi S. Ganesan,5Julian Downward,1,2,*
and David C. Hancock1
1Signal Transduction Laboratory
2Molecular Oncology Laboratory
3Gene Expression Analysis Laboratory
Cancer Research UK London Research Institute
44 Lincoln’s Inn Fields
London WC2A 3PX
United Kingdom
4Laboratory of Molecular Signalling
The Babraham Institute
Babraham Research Campus
Cambridge CB2 4AT
United Kingdom
5Ovarian Cancer Group
Cancer Research UK Molecular Oncology Laboratory
Weatherall Institute of Molecular Medicine
John Radcliffe Hospital
Oxford OX3 9DS
United Kingdom
6Division of Molecular Carcinogenesis and
Center for Biomedical Genetics
The Netherlands Cancer Institute
Plesmanlaan 121
1066 CX Amsterdam
The Netherlands
Summary
The ability of activated Ras to induce growth arrest of
human ovarian surface epithelial (HOSE) cells via in-
duction of the cyclin-dependent kinase inhibitor
p21WAF1/CIP1has been used to screen for Ras pathway
signaling components using a library of RNA interfer-
ence (RNAi) vectors targeting the kinome. Two known
Ras-regulated kinases were identified, phosphoinosi-
tide 3-kinase p110a and ribosomal protein S6 kinase
p70S6K1, plus the MAP kinase kinase kinase kinase
MINK, which had not previously been implicated in
Ras signaling. MINK is activated after Ras induction
via a mechanism involving reactive oxygen species
and mediates stimulation of the stress-activated pro-
tein kinase p38 MAPK downstream of the Raf/ERK
pathway. p38 MAPK activation is essential for Ras-
induced p21WAF1/CIP1upregulation and cell cycle ar-
rest. MINK is thus a distal target of Ras signaling in
the induction of a growth-arrested, senescent-like
phenotype that may act to oppose oncogenic transfor-
mation in HOSE cells.
Introduction
Activating mutations in the RAS oncogenes are involved
in the formation of as many as a third of all human tu-
mors (http://www.sanger.ac.uk/genetics/CGP/cosmic/).
Whereas the early signaling events triggered by Ras
protein in its active, GTP bound state have been studied
in great detail (Downward, 2003), later steps in the Ras-
regulated pathways are less well understood. In order to
identify components of signaling pathways that play es-
sential functions in changing cellular phenotype, RNAi
(Downward, 2004a) has been employed recently as a
tool for genetic screens in mammalian cells in tissue
culture (Downward, 2004b). Two large-scale retroviral
libraries of RNAi vectors have been created that target
sizeable fractions of the human genome (Berns et al.,
2004; Paddison et al., 2004). These have been employed
in a high throughput, gene-by-gene analysis of the func-
tionoftheproteasome(Paddisonetal.,2004)andinase-
lective screen by using large pools of vectors targeting
different genes to seek novel regulators of p53 signaling
(Berns et al., 2004). Synthetic short interfering RNA oli-
gonucleotide collections have also been used in high
throughput screens, for example to study apoptosis
regulation (Aza-Blanc et al., 2003).
It has recently proved possible to identify potential
negative regulators of transformation by the application
of selective RNAi screens to normal immortalized cells,
looking for appearance of cells capable of anchorage-
independent growth (Kolfschoten et al., 2005; West-
brook et al., 2005). We decided to use a different end
result of the Ras signaling system as the basis for a
selective screen: oncogene-induced growth arrest. In
some untransformed cell types in tissue culture, ectopic
expression of activated mutant Ras can lead not to
transformation but to a profound growth arrest (Serrano
et al., 1997). This can show some similarity to replicative
senescence and has on occasion been termed pre-
mature senescence or stasis (Campisi, 2005).
Ras oncogene-induced growth arrest has been re-
ported to involve upregulation of various cell cycle regu-
lators and checkpoint proteins, including p53, p16INK4A,
ARF, and p21WAF1/CIP1(Lowe et al., 2004). The exact
mechanism involved varies significantly between cell
types and between species, particularly human and
mouse (Drayton et al., 2004; Serrano and Blasco, 2001).
Thereexistscontinuing controversyastowhetheronco-
gene-inducedgrowtharrestisaphysiologicalprotective
mechanismforreducingtheincidenceofcancerinmam-
mals by disabling cells that have begun to acquire some
of the multiple mutations required to establish tumori-
genesis,orwhetheritisjustanartifactofoverexpression
of potent signaling molecules in serum and oxygen-rich
tissue culture systems (Campisi, 2005), although recent
evidence has strengthened the case for its in vivo sig-
nificance (Braig et al., 2005; Chen et al., 2005; Collado
etal.,2005;Michaloglouetal.,2005).Whicheverscenario
turns out to be correct, Ras oncogene-induced growth
arrest can be used as a powerful selection to allow
screening for novel molecules that may be involved
*Correspondence: julian.downward@cancer.org.uk
7These authors contributed equally to this work.

Page 2

eitherinRas-inducedsignalingorintheestablishmentof
senescence after cellular stress.
WereporthereaHOSEcellsysteminwhichtheinduc-
ible expression of oncogenic Ras leads to irreversible
growth arrest. These cells have been used in a selective
RNAi screen for resistance to Ras-induced growth ar-
rest by using a subset of the Netherlands Cancer Insti-
tuteRNAilibrary(Bernsetal.,2004)targeting 654human
genes, including all known protein kinases. Of three ver-
ified hits, two represent known direct or indirect targets
of Ras signaling: phosphoinositide 3-OH (PI 3) kinase
p110a and ribosomal protein S6 kinase p70S6K1. The
third hit, MINK, a Ste20/germinal center kinase family
MAP kinase kinase kinase kinase, has not been impli-
cated in Ras signaling previously. A pathway is eluci-
dated in ovarian epithelial cells whereby Ras activates
MINK via Raf/ERK with delayed kinetics through mech-
anisms involving reactive oxygen species. MINK then
upregulates the activity of the p38 stress-activated
kinase through MAP3K5/8 and MKK3/6, leading to
p53-independent induction of p21WAF1/CIP1and perma-
nent cell cycle arrest. Whereas PI 3-kinase and S6 ki-
nase are both involved in Ras-induced transformation
as well as growth arrest, MINK appears to play a role
onlyinthegrowtharrestresponseandthuscouldpoten-
tially have tumor suppressor function.
Results
Expression of Oncogenic Ras Induces Permanent
Growth Arrest in HOSE Cells
HOSE cells expressing the human papilloma virus onco-
genes E6 and E7 have a lifespan of about 25 population
doublings (PD) (Mok et al., 1996; Tsao et al., 1995). To
prolong this lifespan, we introduced hTERT into HOSE
642-1 cells at passage five. The resulting cells grew con-
tinuously for more than 100 PD (data not shown). We co-
transfected these cells with pcDNA6/TR and pcDNA4/
TO-V12H-Ras and thereby obtained cellclones inwhich
the expression of V12 H-Ras was under the control of
a tetracycline-inducible promoter (referred to as HOSE
V12 H-Ras cells). Unexpectedly, inducible expression
of constitutively active H-Ras in HOSE cells expressing
hTERT,E6,andE7ledtoaflat,enlarged,andvacuolated
cell morphology within 48 hr in the majority of the cells
(Figure 1A). These are known signs of a senescent-like
phenotype. This change in morphology is accompanied
byaprofoundcellproliferationarrest,withaccumulation
of cells in both G1 and G2/M phases of the cell cycle
(Figure 1B). Long-term treatment of HOSE V12 H-Ras
cells with doxycycline (dox) to induce expression of
V12 H-Ras caused a permanent growth arrest as shown
by counting cell numbers and by clonogenic assays
(Figure 1C). Removal of dox did not result in a significant
numberofarrestedHOSEV12H-Rascellsreentering the
cell cycle (data not shown).
p42/44 ERK, Akt, and GSK3 are all phosphorylated af-
ter a similar time course to the induction of Ras expres-
sion (Figure 1D), suggesting that expression of V12
H-Ras indeed activates known Ras signaling pathways
such as Raf/ERK and PI 3-kinase/Akt in HOSE cells. To
investigate themechanism oftheRas-induced cellcycle
arrest, the expression of various cell cycle regulators
was studied after induction of Ras expression (Fig-
ure 1E). V12 H-Ras expression caused dephosphoryla-
tion of pRb, a decrease in cyclin A, and an increase in
p21WAF1/CIP1levels, an effect first reported with Raf in
murine fibroblasts (Sewing et al., 1997) and Schwann
cells (Lloyd et al., 1997). Other cell cycle regulators
known to play a role in oncogene-induced growth arrest
in diploid fibroblasts include p53 and p16 (Lloyd et al.,
1997; Serrano et al., 1997). However, in the HOSE cell
system used here, p53 levels are low, presumably due
to the presence of HPV E6 and E6AP; if anything, induc-
tion of V12 H-Ras expression leads to a slight decrease
in p53 protein levels. There are also no significant
changes in the levels of p16 and p15, whereas expres-
sion of another known Cdk inhibitor, p27, is decreased
after induction of V12 H-Ras in HOSE cells.
Inhibition of the Ras-Induced Increase
in p21WAF1/CIP1Overcomes Growth Arrest
in HOSE Cells
Todetermine the role ofp21WAF1/CIP1inthe Ras-induced
growtharrestinHOSEcellsandtoexplorethepossibility
that it could be used as a positive control in an RNAi li-
braryscreenforescapefromRas-inducedgrowtharrest,
we infected HOSEV12 H-Rascells with pRETROSUPER
(pRS) retroviruses encoding short hairpin (sh) RNAs tar-
geting p21WAF1/CIP1mRNA. After selection for retroviral
integration, resistant cells were treated with dox to in-
duce Ras expression for 14 days and analyzed after an
additional 3 weeks of dox treatment. We found that
pRS-p21WAF1/CIP1-infected cells could bypass Ras-
induced growth arrest relative to mock-infected cells, as
seen by analyzing cell number and colony formation (Fig-
ures 2A and 2B). On analysis of the level of p21WAF1/CIP1
mRNA and protein levels, it was clear that basal expres-
sion of p21WAF1/CIP1was only slightly reduced by the
pRS-p21WAF1/CIP1, with mRNA levels down by about
18% (Figure2C). However, the p21WAF1/CIP1RNAivectors
did prevent the increase of p21WAF1/CIP1mRNA and pro-
tein levels induced after V12 H-Ras expression (Figures
2C and 2D). Presumably, this effect on Ras-induced, but
not basal, p21WAF1/CIP1levels reflects the selective pres-
sure on cells expressing Ras, which favors the outgrowth
of cells having significant knockdown of p21WAF1/CIP1.
Thus, we conclude that the induction and maintenance
of the V12 Ras-induced permanent growth arrest in
HOSEcellsisatleastpartiallydependentontheupregula-
tion of p21WAF1/CIP1through a p53-independent pathway.
A Selective RNAi Screen for Abrogation
of Ras-Induced Growth Arrest in HOSE Cells
We used a retroviral RNAi library, constructed in pRS,
with the HOSE V12 H-Ras cells as a system to identify
genes required for Ras-induced growth arrest. The li-
braryusedwasasubsetoftheNetherlandsCancerInsti-
tute (NKI) library (Berns et al., 2004) targeting 654 human
genes, including all known protein kinases (see Table S1
available in the Supplemental Data with this article on-
line). Each gene was targeted by three different shRNA
sequences. HOSE V12 H-Ras cells were infected with
pools of retroviral mixtures targeting 96 mRNAs each.
Cells were selected for expression of the shRNA vector
by using puromycin for 7 days, and resistant cells were
then treated with a single dose of 1 mg/ml dox. Whereas
emptyvector-infectedcontrolcellsremainedcompletely
Molecular Cell
674

Page 3

growth arrested 3 weeks after treatment, we found five
single cell clones growing out of the nonproliferating
background. Sequencing of the retroviral shRNA inserts
in these cells revealed that three of the five clones con-
tained single retroviral integrations targeting the genes
RPS6KB1 (encoding ribosomal protein S6 kinase
p70S6K1), PIK3CA (encoding phosphoinositide 3-OH ki-
nase catalytic subunit p110a), and MINK (encoding the
Ste20/germinal center kinase family MAP kinase kinase
kinase kinase MINK). The other clones contained multi-
ple integrations, including the same sequence targeting
RPS6KB1 in one case. p110a is a known direct target of
Ras (Rodriguez-Viciana et al., 1994), whereas p70S6K1is
an indirect target that is regulated by phosphoinositide
3-OH kinase (PI 3-kinase) (Chung et al., 1994). MINK
has not previously been implicated in Ras signaling
(Dan et al., 2000, 2001).
To validate this result, we subcloned the identified
shRNAsintopMig,apRSderivativeinwhichEGFP isex-
pressed from an IRES after the puromycin resistance
gene. These vectors were used to infect HOSE V12 H-
Ras cells with viruses containing only the relevant se-
quences found in the Ras-resistant colonies. Puromy-
cin-selected cell populations were FACS sorted, and
high GFP fluorescence expressing cells were collected
and analyzed for the efficiency of knockdown of their re-
spective gene targets. A quantitative RT-PCR analysis
(Figure3A)showedthattheshRNAsequences identified
in the screen are indeed able to suppress target mRNA
expression, although in the case of p110a and MINK,
C
0
1
2
3
4
5
0714 2128
cell number (x106)
control
dox
days
control dox
D
dox [hours]
E
V12 Ras
-+
dox
p21
pRb
cyclinA
p53
p16
p15
p27
0
1
1
1
p21
0.5
1.5
2
2.5
control
dox
cyclin A
fold expression of
untreated control
A
B
0
3 days 7 days

% of gated events
condoxcondox
20
40
60
80
G1SG2/M
7224 8420
P-ERK
P-Akt
P-GSK3α/
loading
pan ras
V12 Ras
0 72
mock
controldox
Figure 1. Expression of V12 H-Ras Induces Cell Cycle Arrest in Human Ovarian Surface Epithelial Cells
(A) Phase contrast micrographs of HOSE V12 H-Ras cells after 72 hr treatment with or without 1 ng/ml doxycycline (dox).
(B)CellcycledistributioninHOSEcellsafterRasinduction.HOSEV12H-Rascellsweretreatedwith1ng/mldoxforthe indicatedtimes,andDNA
content was analyzed by FACS. (Data shown are representative of three independent experiments.)
(C)HOSEV12 H-Rascellsweretreated withorwithout 1ng/ml dox,andcellnumbers werecountedatindicatedtime points. Inset, 500HOSEV12
H-Ras cells were plated and treated with or without 1 ng/ml dox. After 21 days, cells were stained with crystal violet.
(D) HOSE V12 H-Ras cells were treated with or without 1 ng/ml dox for indicated times. Total extracts were immunoblotted with the indicated
antibodies.
(E)HOSEV12 H-Rascellsweretreatedwith1ng/ml doxfor48hr.Totalcellextractswereimmunoblotted.TotalRNA wasusedinquantitativeRT-
PCR analysis by using p21WAF1/CIP1and cyclin A-specific primers.
Error bars show standard deviation of the population from the arithmetic mean of at least three independent experiments.
Role of MINK in Ras-Induced Senescence
675

Page 4

the knockdown in the puromycin and GFP-selected
pools at this stage is relatively modest, possibly reflect-
ing heterogeneity within the population. To ensure that
treatmentofemptyvectorcontrolandshRNAexpressing
cell populations with dox led to induction of comparable
levels of Ras expression and subsequent phosphoryla-
tion of ERK1/2, we performed a Western blot analysis
(Figure 3B). Next, we dox treated the selected cell pop-
ulations and analyzed cell growth by clonogenic assay
or cell number counting. We found that only about 1%
of the control empty vector-infected cells had been able
to bypass Ras-induced growth arrest, whereas 44% 6
15% of the p70S6K1knockdown, 28% 6 11% of the
p110a knockdown, and 43% 6 6% of the MINK knock-
down cells proliferated despite activated Ras expres-
sion (Figures 3C and 3D). Checking the shRNA-induced
knockdown of MINK after 5 weeks of dox treatment, we
found that MINK expression is induced by Ras activa-
tion by about 2-fold and this upregulation is completely
preventedbyMINK-specificshRNA(Figure3E).Fordox-
treated cells, the MINK shRNA containing cells express
26%ofthe level ofMINK mRNA ofcontrolcells. Thepro-
cess of Ras induction has thus markedly selected for
cells with more efficient knockdown of MINK, relative
to the starting population of MINK shRNA expressing
cells analyzed in Figure 3A.
To ensure that the observed phenotypic changes pro-
duced by the shRNA vectors were not caused by off-
target effects, we employed predesigned synthetic dou-
ble-stranded RNA oligonucleotides targeting p70S6K1,
p110a, and MINK. These siRNAs wereable totransiently
reduce the level of their target mRNAs (Figure 3F). We
studied the effect of these siRNAs on Ras-induced inhi-
bition of cell cycle progression in a short-term ELISA for
BrdU incorporation. HOSE V12 H-Ras cells were trans-
fected with the specific siRNAs or an siRNA targeting
GFP as a negative control, and 48 hr posttransfection,
cells were treated with dox for an additional 72 hr. In
control cells, Ras induction caused an inhibition of BrdU
incorporation of 65% 6 4% (Figure 3G). Cells trans-
fected with siRNA targeting p70S6K1, p110a, or MINK
had significantly less inhibition of BrdU incorporation
A
fold of untreated control
(cell counts)
0
0.2
0.4
0.6
0.8
1
7 14
days
pRS
p21 shRNA
pRSp21
pRS p21
C
D
0.5
1
1.5
2
2.5
fold expression of
untreated pRS-control
control
dox
p21 expression
5
10
15
20
25
fold expression of
untreated pRS-control
ras expression
control
dox
B
# of plated cells
505005000
p21
shRNA
pRS
control
dox
control
dox
p21
P-ERK
pRS
- +
p21
- +
dox
shRNA
Figure 2. Reduction of p21WAF1/CIP1Expression Partially Prevents Ras-Induced Growth Arrest in HOSE Cells
HOSE V12 H-Ras cells were infected with a retroviral mixture targeting p21WAF1/CIP1, selected for shRNA expression, then treated with 1 mg/ml
doxfor14days.Cellswerethenreplatedasappropriate:eitherat200cells/6wellforcellcountingassay,atindicatedcellnumbersforclonogenic
assays, or at 20,000 cells/10 cm plate for quantitative RT-PCR and Western blot analysis.
(A and B) HOSE V12 H-Ras cells expressing p21WAF1/CIP1siRNA were treated with or without 1 ng/ml dox, and cell numbers were counted at the
indicated time points (A), or cells were stained with crystal violet after 21 days (B).
(C and D) HOSE V12 H-Ras cells were treated with 1 ng/ml dox for 21 days, and total RNA was used to perform quantitative RT-PCR (C) and total
cell lysates to perform Western blot analysis (D) to detect p21WAF1/CIP1expression and ERK phosphorylation.
Error bars show standard deviation of the population from the arithmetic mean of at least three independent experiments.
Molecular Cell
676

Page 5

(p70S6K1, 23% 6 9%; p110a, 31% 6 5%; and MINK,
32% 6 8%). Furthermore, we found that inhibiting
Ras-controlled signaling pathways by pretreating the
cells with the chemical inhibitors LY294002, PD98059,
or Rapamycin could also almost completely prevent
the Ras-induced inhibition of BrdU incorporation
(Figure 3H). These inhibitors implicate PI 3-kinase
(LY294002), S6K (LY294002 and Rapamycin), and MEK
(PD98059) in Ras-induced growth arrest in HOSE cells,
independently confirming two of the three genes identi-
fied in the screen. The effects of the drugs in countering
the short term growth arrest induced by Ras are more
complete than with the siRNAs, probably reflecting re-
dundancy at the protein level due to the presence of
other isoforms of the genes targeted.
MINK belongs to a subgroup of GCK/STE20-like
kinases, the other very closely related kinases being
MAP4K4/HGK and TNIK. TNIK is 65% identical to
MINK, with 89% and 90% identity in the kinase and cit-
ron-homology domains, respectively. Knocking down
expression of TNIK by using siRNA oligos also greatly
reduces Ras-induced growth arrest (Figure S1). It there-
fore appears highly likely that there is some redundancy
between MINK and TNIK in the activation of Ras onco-
gene-induced arrest in HOSE cells.
Expression and Activation of MINK Is Induced
by Ras Expression
WhereasPI3-kinasep110aandp70S6K1areknown tobe
activated downstream of Ras, MINK has not previously
been associated with Ras signaling. Therefore, we ex-
amined the effects of expression of activated Ras on
MINK mRNA levels and kinase activity in HOSE cells.
Dox treatment of HOSE V12 H-Ras cells to induce Ras
expressionled toanincrease inMINK mRNAexpression
and protein levels by about 2-fold within 72 hr (Figures
4A and 4B). Additionally, in vitro kinase assays of MINK
from HOSE cells exogenously expressing HA-tagged wt
full-length MINK showed that V12 H-Ras expression
also increases MINK kinase activity 2- to 3-fold (Figures
4C and 4D). Immunoprecipitates of a kinase-inactive
mutant of MINK failed to phosphorylate the MBP sub-
strate(datanotshown).Thus,theremaybeindependent
stimulatory effects of Ras on both the expression and
the kinase activity of MINK. To elucidate which signaling
pathways might be involved in the Ras-induced activa-
tion of MINK, we performed in vitro kinase assays by us-
ing lysates from cells pretreated with the chemical in-
hibitors that we had found to prevent Ras-induced
blockage of cell cycle progression (Figure 3H). UO126,
another inhibitor of MEK, abrogated Ras-induced acti-
vation of MINK (Figure 4E), but both LY294002 and Ra-
pamycin did not significantly affect the activation of
MINK by Ras. Thus, activation of MINK after expression
of activated Ras is mediated principally through the Raf/
MEK/ERK signaling pathway in HOSE cells.
To determine whether activation of the Raf pathway
was sufficient to induce MINK activation, we created
HOSE cells with an inducible Raf construct, DRaf:ER*,
the kinase activity of which is stimulatable by 4HT, but
not natural estrogens (Bosch et al., 1997). Activation of
Raf in HOSE cells increased MINK kinase activity signif-
icantlyinatime-dependentmanner(Figure4F),support-
ing the model of Ras-induced activation of MINK via the
Raf/MEK/ERK signaling pathway.
MINK Controls Ras-Induced Growth Arrest through
p38 Stress-Activated Kinase
How might MINK be involved in Ras-induced growth
arrest in HOSE cells? MINK has been shown to be able
to activate the p38 stress-activated protein kinase (Dan
etal.,2000).Activationofp38MAPKhasbeenimplicated
in mediating premature senescence in response to high
intensity oncogenic Ras signals in primary human fibro-
blasts (Deng et al., 2004). In our cell system, we could
detect a significant increase in phosphorylation of p38
MAPK after activated Ras expression for 3 days, which
is slightly delayed relative to the kinetics of ERK phos-
phorylation (Figure 5A). The induction of p38 MAPK
phosphorylation followed the gradual increase in MINK
proteinlevels(Figure5A)andwasinhibitedbytheknock-
down of MINK by using MINK-specific siRNAs (Fig-
ure 5B). Knocking down MINK also markedly reduced
the phosphorylation of the p38 MAPK activating kinases
MKK3/6(Figure5B),butnotERKphosphorylationorRas
expression. Looking at the effects of MINK knockdown
on the expression of cell cycle regulators, there was
partial inhibition of both the Ras-induced increase in
p21WAF1/CIP1and decrease in cyclin A levels (Figure 5B).
These data implicate a role for MINK in the inhibition of
proliferation by activated Ras in HOSE cells through its
ability to activate the p38 MAPK stress kinase and in-
crease p21WAF1/CIP1expression.
To elucidate further the role of p38 MAPK activation in
the Ras-induced growth arrest in HOSE cells, we trans-
fected HOSE V12 H-Ras cells with specific siRNAs tar-
geting both MKK3 and MKK6, the two MAP kinase
kinases responsible for activating the p38 MAP kinases.
Knockdown of MKK3/6 completely abolished the ability
of activated Ras to induce phosphorylation of p38
MAPK, increase p21WAF1/CIP1expression, and decrease
cyclin A levels (Figure 5C). Furthermore, in cells where
MKK3/6 expression had been knocked down, we ob-
served markedly reduced inhibition of BrdU incorpora-
tion into DNA after 3 days of dox treatment compared
to control cells transfected with GFP siRNA (Figure 5D).
It has been shown that activation of p38 MAPK by
a 4HT inducible form of the MAP kinase kinase kinase
MEKK3, DMEKK3:ER*, induces cell cycle arrest in
CCl39 cells and Rat-1 fibroblasts (Garner et al., 2002).
We therefore tested the effect of MEKK3 activation in
HOSE cells: activation of DMEKK3:ER* by 4HT also
induced phosphorylation of p38 MAPK that was accom-
panied by induction of p21WAF1/CIP1expression and de-
creaseofcyclinAproteinlevels(FigureS2A).Wealsode-
tected a more than 90% inhibition of BrdU incorporation
in HOSE DMEKK3:ER* cells after 48 hr of 100 nM 4HT
treatment compared to no effect on cell cycle progres-
sion of 4HT on HOSE cells expressing a kinase-inactive
form of DMEKK3:ER* (Figure S2B). Furthermore, we
found that activation of DRaf:ER* by 4HT in the HOSE
DRaf:ER* cells induced p38 MAPK phosphorylation
within 72 hr of treatment (Figure S2C). Like oncogenic
Ras, activation of Raf increased p21WAF1/CIP1protein
levelsanddecreasedcyclinAlevelsinHOSEcells(Figure
S2C)andinducedacellcyclearrest,measuredbyinhibi-
tion of BrdU incorporation, of about 50% (Figure S2D).
Role of MINK in Ras-Induced Senescence
677

Page 6

con
H
fold of untreated control
(absorbance @ 450nm)
0
0.2
0.4
0.6
0.8
1
1.2
F
GFP S6K p110 MINK
fold expression of
GFP siRNA control
0
0.2
0.4
0.6
0.8
1
1.2
siRNA
0
0.2
0.4
0.6
0.8
1
GFPp110 S6K MINK
G
fold of untreated control
(absorbance @ 450nm)
siRNA
p110
shRNA
MINK
shRNA
control
dox
control
dox
20200 2000
# of plated cells
A
0
0.2
0.4
0.6
0.8
1
1.2
shRNA
fold expression of
pMig control
S6KMINKpMigp110
D
E
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
pMig S6Kp110 MINK
fold of untreated control
(cell counts)
shRNA
pMigMINK shRNA
0
0.5
1
1.5
2
2.5
fold expression of
pMig control
control
dox
MINK expression
C
S6K
shRNA
20 2002000
pMig
PD LY Rapa
pan ras
B
P-ERK
pMig
- +
shRNA
dox
S6K
- +
p110
- +
MINK
- +
loading
Figure 3. RNAi Library Screen for Escape from Ras-Induced Growth Arrest in HOSE Cells
(A) HOSE V12 H-Ras cells were infected with pMIG retroviruses targeting p70S6K1, p110a, and MINK using the shRNA sequence identified in the
RNAi library screen. Total RNA was used to perform quantitative RT-PCR by using specific primers for p70S6K1, p110a, and MINK. (Data shown
are representative of three independent experiments.)
(B–E) Puromycin-resistant HOSE V12 H-Ras cell populations expressing the shRNA sequences identified in the RNAi library screen were FACS
sorted for GFP expression and 106of these selected cells/10 cm dish were subsequently treated with 1 mg/ml dox for 14 days. Cells were then
platedeitheratindicatedcellnumbersforclonogenicassays,at200cells/6wellforcellcountingassay,orat20,000cells/10cmplateforWestern
blotand quantitative RT-PCR analysis and treatedwith orwithout 1 ng/ml dox.After 21days, total celllysates wereusedin Western blot analysis
todetectRasandphospho-ERK(B),cellswerestainedwithcrystalviolet(C),cellnumberswerecounted(D),ortotalRNAusedinquantitativeRT-
PCR to detect MINK expression (E).
(Fand G)HOSE V12H-Rascellsweretransfected with thespecificsiRNA oligos (62.5nM), cultured incomplete media,and retransfected after24
hr. Forty-eight hours after the first transfection, cells were plated to be treated with or without 1 ng/ml dox for 72 hr and assayed by using
Molecular Cell
678

Page 7

To address whether the involvement of MINK in Ras-
induced growth arrest is limited to ovarian epithelial
cells, we also studied human diploid fibroblasts. In Hs68
cells overexpressing hTERT, activation of Raf1 using a 4
hydroxytamoxifen (4HT)-inducible estrogen receptor fu-
sion protein leads to a major decrease in cell prolifera-
tion that is largely reversed by knockdown of MINK
using RNAi (Figure S3). Raf induction of p38 activation
is also reversed by MINK RNAi in these fibroblasts. It
is therefore possible that MINK plays a role in p38-
mediated growth arrest downstream of Ras and Raf in
human primary fibroblasts as well as ovarian epithelial
cells. In a different epithelial cell system, human lung
smallairway epithelial cellsimmortalizedwith SV40 large
T and telomerase (Lundberg et al., 2002), the induction
of expression of V12 K-Ras also leads to an inhibition
of cell proliferation which, although rather slower than
in HOSE cells, is reversed by reduction in MINK expres-
sion (data not shown).
To analyze the pathway linking MINK and p38 MAPK,
systematic RNAi analysis was carried out on the 14
known members of the MAP3K family in the human
genome, MAP3K1–MAP3K14 (HUGO nomenclature).
As shown in Figures S4A and S4B, knockdown of two
different MAP3Ks, MAP3K8 (Cot/Tpl2) and MAP3K5
(Ask1), relieves Ras-induced growth arrest. Neither of
these appears to work in the upstream part of the path-
way, as their knockdown does not result in reduction of
ERK activation by Ras, although it does block p38 in-
duction(FigureS4C).Ask1hasbeenreportedtoactivate
both p38 and JNK stress-activated kinases and to be in-
fluenced by oxidative stress through a number of mech-
anisms, making it a good candidate for linking MINK
with p38 activation. By contrast, Cot/Tpl2 has not
quantitative RT-PCR for expression of the target genes (F) or BrdU ELISA for proliferation (G).
(H) HOSE V12 H-Ras cells were treated with or without the indicated chemical inhibitors (PD98059- PD, LY294002- LY, Rapamycin-Rapa) for 30
min prior to start of treatment with 1 ng/ml dox for further 72 hr. Cells were assayed by using an ELISA to detect BrdU incorporation.
Error bars show standard deviation of the population from the arithmetic mean of at least three independent experiments.
fold expression of
untreated control
0.5
1
1.5
2
2.5
MINK expression
controldox
0
A
< MBP
phosphorylation
input
10 0.11000 0dox [ng/ml]
1.01.21.92.9 Rel. signal
C
< MBP
phosphorylation
input
dox
control
UOLYRapa
Rel. signal0.82.7 1.02.2 1.02.20.9 0.7
++++
----
E
MINK
+-dox
loading
B
< MBP
phosphorylation
input
0123 dox [days]
Rel. signal 1.00.8 1.5 2.6
D
< MBP
phosphorylation
input
031 4HT [days]
Rel. signal 1.01.03.0
F
Figure 4. Activated Ras Expression Induces Expression and Kinase Activation of MINK
(A and B) HOSE V12 H-Ras cells were treated with or without 10 ng/ml dox for 72 hr. Total RNA was used in quantitative RT-PCR by using MINK-
specific primers (A), or total cell lysates were used in Western blot analysis to detect MINK (B).
(C–E) HOSE V12 H-Ras cells were transfected with an HA-tagged wild-type (wt) MINK expression construct and treated with different concentra-
tions of dox for 72 hr (C), with or without 10 ng/ml dox for the indicated times (D) or for 72 hr with or without 10 ng/ml dox after pretreatment with
10mMoftheMEKinhibitorUO126(UO),10mMofthePI3-KinhibitorLY294002(LY),or50nMofthemTORinhibitorRapamycin(Rapa)for30min(E).
Cellswerethenlysed,andinvitrokinaseassayswereperformedbyusingmyelinbasicprotein(MBP)asasubstrateinthepresenceof[g-32P]ATP.
The reaction products were subjected to SDS-PAGE and autoradiography (top panels). Bottom panels, anti-HA antibody Western blot to detect
tagged MINK. The radiolabeled MBP bands were analyzed by scanning and densitometry, the signal normalized to the loading control, and the
ratio of stimulated versus unstimulated sample given below lane.
(F)HOSEV12H-RascellswereinfectedwiththeDRaf:ER*inducibleRafconstruct.Puromycin-resistantcellsweretransfectedwithHA-taggedwt
MINK and treated with or without 100 nM 4-hydroxytamoxifen (4HT) for the indicated times. In vitro kinase assay was then performed as above.
Error bars show standard deviation of the population from the arithmetic mean of at least three independent experiments.
Role of MINK in Ras-Induced Senescence
679

Page 8

previously been reported to have a significant role in the
regulation of p38 activation, suggesting the data here
could reflect a more indirect mechanism. Because
knockdown efficiency was only tested for MAP3K5
and MAP3K8, it is possible that some of the other
MAP3Ks could influence this pathway but are not being
effectively targeted by the siRNA pools used here.
Ras-Induced Reactive Oxygen Species Promote
Activation of MINK and p38 MAPK
Reactive oxygen species (ROS) are produced in re-
sponse to Ras activation, with high ROS levels being as-
sociated with growth arrest and senescence (Campisi,
2005; Lee et al., 1999). We therefore analyzed the effect
of Ras activation upon ROS generation, as detected by
DCF-DA staining, in our cell system. V12 H-Ras HOSE
cells exhibited a marked increase in the levels of cellular
ROS within 24 hr of dox addition, and the increase be-
came greater with extended treatment time (Figure 6A).
Pretreatment of cells with the MEK inhibitor PD98059
completelyabrogatedRas-inducedROS(Figure6A),im-
plyingacriticalrolefortheMEK/ERKpathwayinthispro-
cess. Pretreatment of cells with the antioxidant N-acetyl
cysteine (NAC) also strongly inhibited Ras-induced ROS
levels (Figure 6B), together with the appearance of the
senescent cell morphology characteristic of HOSE cells
after Ras-induced arrest (Figure 6C).
To assess further the effect of antioxidant addition
upon indicators of Ras-induced growth arrest, we
treated V12 H-Ras HOSE cells with NAC for 30 min prior
to dox-induced Ras expression. Figures 6D and 6E illus-
tratethat,whereas Ras expressionandERK phosphory-
lationareunaffectedbyNACaddition,MKK3/6,p38,and
MINK kinase activation are all profoundly compromised
bysuchtreatment.Analternativeantioxidant,pyrrolidine
dithiocarbamate plus ascorbate, showed asimilar effect
(data not shown). Lastly, the direct addition of H2O2to
V12 H-Ras HOSEcells, in ordertoinduce rapidoxidative
stress, resulted in phosphorylation of p38 and MKK3/6
within1hrinadose-dependentmanner(Figure6F).Con-
sistentwiththis,MINKkinaseactivitywasalsorapidlyin-
duced in response to H2O2treatment in both V12 H-Ras
HOSE cells and COS-7 cells (Figure 6G).Taken together,
these results suggest that activated Ras-induced cell
A
C
pan ras
P-ERK
P-p38
p21
Cyclin A
dox
MKK3/6
-
siRNA
P-MKK3/6
loading
GFP
-++
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
GFPMKK3/6
fold of untreated control
(absorbance @ 450nm)
siRNA
D
P-p38
MINK
pan ras
P-ERK
loading
dox [days] 0 1 2 3 4
dox
GFPMINK
-
siRNA
-++
P-p38
P-MKK3/6
loading
pan ras
P-ERK
p21
Cyclin A
B
MINK
Figure 5. Ras-Induced Growth Arrest in
HOSE V12 H-Ras Cells Involves MINK-Medi-
ated Activation of p38 MAPK Downstream
of the Raf/ERK Pathway
(A)HOSEV12H-Rascellsweretreatedwithor
without 10 ng/ml dox for the indicated times,
and total extracts were immunoblotted with
the indicated antibodies.
(B) HOSE V12 H-Ras cells were double trans-
fected on consecutive days with 62.5 nM si-
RNAs targeting MINK and GFP as a control.
Forty-eight hours after the first transfection,
cells were treated with or without 10 ng/ml
dox for an additional 72 hr. Total cell lysates
were then prepared and Western blot analysis
performed.
(C and D) HOSE V12 H-Ras cells were double
transfected with 62.5 nM siRNA targeting
GFP as a control and MKK3 and MKK6 (with
31.25 nM each) on consecutive days. Forty-
eight hours after the first transfection, cells
were treated with 10 ng/ml dox for 72 hr and
whole cell lysates were used to perform West-
ern blot analysis (C) or assayed by using an
ELISA to detect BrdU incorporation (D).
In (D), error bars show the standard deviation
of the population from the arithmetic mean of
five independent replicates.
Molecular Cell
680

Page 9

cycle arrest may be mediated through a sequential sig-
naling pathway that involves activation of Raf/MEK/
ERK, leading to ROS-mediated activation of MINK and
thep38stresskinase,resultinginp21WAF1/CIP1induction.
MINK Is Required for Ras-Induced Senescence,
but Not Transformation
The fact that PI 3-kinase and p70S6K1were identified
here as mediators of Ras-induced growth arrest in addi-
tion to their known roles in Ras-induced transformation
raised the possibility that MINK might also function in
both context-dependent outcomes of Ras signaling.
However, we have failed to find any evidence that
MINK function is required to maintain the transformed
phenotype of several human tumor cell lines with acti-
vating mutations in their endogenous K-Ras gene. Sta-
ble knockdown of MINK expression by vector-based
RNAi in several tumor cell lines carrying activating Ras
mutations, such as A549, HCT116, HT29, and H1299,
failed to compromise their transformed phenotype as
measured by colony formation in soft agar (Figure S5),
although knockdown of p110a and S6K1 does impair
this to varying extents. Thus, reduction of MINK mRNA
expression by about 80% does not affect the trans-
formed phenotype of tumor cells with activated Ras. In
addition, the establishment of the Ras transformed phe-
notype in immortalized cell lines is not affected by ex-
pression of either kinase inactive or wild-type (wt)
MINK. Activated Ras expression constructs were co-
transfected along with those encoding kinase dead or
wt MINK, or control plasmids. Neither kinase dead
MINK, which appears to have dominant-negative activ-
ity (data not shown) nor wt MINK had a significant im-
pact on the level of Ras driven transformation of NIH
3T3 cells as measured in a focus formation assay. (V12
H-Ras + control plasmid, 29 6 3 foci; V12 H-Ras + kinase
dead MINK plasmid, 26 6 3 foci; and V12 H-Ras + wt
MINK plasmid, 29 6 4 foci, n = 3.) To the extent that con-
clusions can be reached from negative data, these ex-
periments suggest that MINK may not be required for
Ras-induced transformation.
Discussion
Ras-Induced Growth Arrest in HOSE Cells
The ovarian epithelial cells used here expressing human
papilloma virus E6 and E7 oncoproteins, the catalytic
subunit of telomerase, and tetracycline-inducible acti-
vatedRas wereoriginally madewiththeaim ofproviding
an in vitro model for ovarian cancer. However, in con-
trast to experiences in other cell systems using immor-
talizing genes together with constitutive Ras expression
(Hahn et al., 1999), induction of Ras activation caused
growth arrest rather than transformation in this ovarian
epithelial cell system. The induction of growth arrest
by Ras in the presence of E6 suggests that p53 does
not play amajor rolein the cellcycle arrest inthese cells.
This would be expected with HPV E6 present, which ap-
pears to be keeping p53 levels very low compared to the
levels in the SV40 early region and telomerase immortal-
ized IOSE 80 cells (data not shown). Indeed, Ras causes
a slight decrease in p53 protein levels (Figure 1), per-
haps through Ras induction of MDM2 via the Raf/ERK
pathway (Ries et al., 2000). The use of p53 RNAi con-
structs also failed to impair Ras-induced growth arrest
(datanotshown).Rashereinducesp21WAF1/CIP1expres-
sion in a p53-independent manner, and abrogation of
this induction by RNAi alleviated Ras-induced arrest
(Figures 1 and 2). In addition, there is no sign of p16 in-
duction. It therefore appears that inducible expression
of Ras leads to a growth arrest in this ovarian epithelial
cell system by a mechanism that does not involve either
p53 or p16, the two pathways most commonly associ-
ated with oncogene-induced growth arrest (Lloyd et al.,
1997; Serrano et al., 1997; Sewing et al., 1997). A p53-
independent Ras induction of p21WAF1/CIP1-mediated
cell cycle arrest has previously been reported in myeloid
leukemia cells (Delgado et al., 2000).
MINK Signaling in Ras-Induced Growth Arrest
After activation of Ras and Raf, MINK expression and
activity are induced relatively slowly over a few days.
There are effects both on MINK transcription and on
the activity of MINK protein expressed from an exoge-
nous promoter (Figure 4). The delayed response indi-
cates that the induction is likely to be indirect. MINK
has been reported to activate both the JNK and p38
MAPK stress kinase pathways (Dan et al., 2000), and
p38 MAPK has been implicated in cellular growth arrest
in some systems (Haq et al., 2002; Wang et al., 2002), so
it is tempting to speculate that the involvement of MINK
in Ras-induced growth arrest in HOSE cells is due to its
ability to control p38 MAPK, as confirmed in Figure 5.
MINK appears likely to be activating p38 MAPK through
downstream MAP kinase kinase kinases, such as
MAP3K5 (Ask1), which in turn activate MKK3 and
MKK6, which finally activate p38 MAPK.
Delayed induction of p38 MAPK downstream of the
Raf pathway has been previously reported to play a
roleinRas-inducedprematuresenescenceinfibroblasts
(Bulavinetal.,2003;Deng etal.,2004; Wangetal.,2002).
High,butnotmoderate,intensityprolongedactivationof
Ras, and Raf/ERK led to senescence in a p38 MAPK-
dependent manner (Deng et al., 2004; Wang et al., 2002).
p38 MAPK induction may also be important in causing
cell senescence in response to a number of stresses in
addition to expression of the Ras oncogene, such as ox-
idative stress, telomere shortening, DNA damage, and
inappropriate culture conditions (Haq et al., 2002; Iwasa
et al., 2003). Ras induction of p38 MAPK activation has
been reported to limit the ability of Ras to transform rat
intestinal epithelial cells, in contrast to Ras induction of
ERK and JNK pathways, which promote transformation
(Pruitt et al., 2002; Shields et al., 2002). Multiple mecha-
nisms exist by which activation of p38 MAPK can inhibit
cell proliferation (Bulavin and Fornace, 2004).
MINK may represent a missing link between pro-
longed Raf/ERK activation and p38 MAPK induction. Al-
though the nature of the intermediate steps linking ERK
activation and MINK induction are not known in detail at
present, they are sensitive to conditions that reduce the
level of reactive oxygen species produced after expres-
sion of activated Ras. Oxidative stress is necessary and
sufficient for the activation of MINK downstream of Ras.
ActivatedRasexpressionhaslongbeenassociatedwith
elevation of the levels of ROS, both in the induction of
cell proliferation (Irani et al., 1997) and, at higher ROS
levels, cell senescence (Lee et al., 1999). Interestingly,
Role of MINK in Ras-Induced Senescence
681

Page 10

dox dox
+NAC
H2O2
EtOH
0
10
20
30
% DCF-DA positive cells
0
5
10
15
20
25
30
35
control
dox PD
dox 24h
dox 72h
H2O2/10min
% DCF-DA positive cells
A
DE
F
C
G
B
P-p38
H2O2 [mM]02.510
loading
P-MKK3/6
< MBP
phosphorylation
H2O2
input
Rel.signal11.81 2.7
HOSECOS-7
5 10
< MBP
phosphorylation
dox
0
input
Rel.signal
-+
con NAC
-+
13.70.9 1.50.9 1.2
P-p38
P-MKK3/6
loading
pan ras
P-ERK
dox
NAC [mM]
control
dox
dox + NAC
NAC
++
++
+
--
--
-
Figure 6. Induction of ROS Downstream of the Raf/ERK Pathway Is Essential for MINK and p38 MAPK Activation
(A) HOSE V12 H-Ras cells were treated with or without 1 mg/ml dox for the indicated times after pretreatment or not with 10 mM of the MEK in-
hibitor PD98059 (PD), and reactive oxygen species (ROS) levels were analyzed by FACS.
Molecular Cell
682

Page 11

a recently reported genetic suppressor element screen
inratembryofibroblastsfoundtheoxidoreductaseSela-
din-1 as a necessary component of Ras-induced, p53-
dependent senescence (Wu et al., 2004). The induction
of MINK activity after Ras activation could perhaps re-
sult from inhibition of oxidation-sensitive phosphatases
(Rhee et al., 2000) or from redox-sensitive induction of
autocrine production of cytokines (Haddad et al., 2001).
Screening for Regulators of Growth Arrest
and Transformation
The screen of 654 human genes, including all protein
kinases, for those with an essential role in Ras-induced
growth arrest in HOSE cells yielded three verified hits:
PI3-kinasep110a, ribosomal
p70S6K1, and MINK. Of these, p110a and p70S6K1have
beenwellcharacterizedasplayingimportantrolesinma-
lignant transformation, acting downstream of Ras and
other oncogenes (Downward, 2003). PI 3-kinase and S6
kinase signaling have not previously been implicated in
theinductionofprematuresenescencebyRas,although
the PI 3-kinase inhibitor LY294002 has previously been
reported to block hydrogen peroxide-induced senes-
cence in human diploid fibroblasts (Wang et al., 2004)
andPTEN deletion has recently been associated within-
duction of p53-dependent senescence in mouse pros-
tateglands(Chenetal.,2005).AsearlytargetsofRassig-
naling, it is not unexpected that PI 3-kinase and p70S6K1
mightplayimportantrolesinthisresponsetoexpression
of activated Ras. Although the mechanism underlying
the involvement of PI 3-kinase and p70S6K1in Ras-
induced growth arrest was not explored in detail, it
seems very likely that the initial signaling pathways reg-
ulated by Ras may be similar to those situations where
transformation results. Nevertheless, the final pheno-
typic outcome for the cell is determined by signal inte-
gration with other inputs further down the pathways,
along with effects of signal amplitude and duration.
The screen failed to identify components of the Raf/
MEK/ERK pathway, despite the ability of PD98059 to
blockshort-termRas-inducedgrowtharrest(Figure3G).
This could reflect redundancy in this pathway, ineffec-
tive targeting of pathway components by the shRNA
vectors, or failure of inhibition of ERK to allow long-term
clonogenic proliferation despite avoiding the short-
term cell cycle arrest. Similarly, MKK3 and MKK6 were
not picked up: use of siRNA oligos indicated that both
need to be targeted simultaneously (data not shown).
proteinS6 kinase
This failure raises the question of the success rate of
the screen at picking up critical molecules in Ras-
induced growth arrest signaling. Redundancy clearly
has the potential to limit the success of the screen but
is hard to avoid short of using high throughput screen-
ing with very small pools of RNAi vectors designed to
targetcloselyrelatedproteinssimultaneously.However,
a more important issue here is likely to be the relative in-
efficiency of the RNAi: the use of large pools of vectors
and stable selection incombination with the low number
of retroviral integration events possible per cell means
that the selection for the phenotype must normally be
achieved by a single integrated retroviral vector Al-
though RNAi can certainly provide gene silencing of
this potency, it is likely that a proportion of the vectors
in the library will be unable to provide sufficient levels
of knockdown, even if they can show reasonable silenc-
ing in transient transfections of a single vector.
The fact that PI 3-kinase and p70S6K1were identified
here as mediators of Ras-induced growth arrest in addi-
tion to their known roles in Ras-induced transformation
raised the possibility that MINK might also function in
both context-dependent outcomes of Ras signaling.
On the other hand, MINK is clearly acting much further
downstream than PI 3-kinase or p70S6K1, so it might
act in pathways that are only involved in growth arrest
and not in other aspects of Ras biology. We have failed
to find any evidence that MINK function is required to
maintain the transformed phenotype of several human
tumor cell lines with activating mutations in their endog-
enous K-Ras gene. It seems, therefore, that MINK func-
tions in Ras signaling below any point of commonality
between transformation and growth arrest, or prema-
ture senescence, signaling. MINK is therefore likely to
act only to mediate the growth inhibitory effects of Ras
signaling and not the transformation promoting effects
and is thus a potential tumor suppressor that could limit
the potential of activated mutant Ras proteins to pro-
motemalignancy. Weare
whether there is evidence for mutation or reduced ex-
pression of MINK or its close relatives in human tumors.
currently investigating
Experimental Procedures
RNAi Screen
DNAzol (Invitrogen) was used to extract genomic DNA from ex-
panded colonies. pRS vector-specific primers were used to perform
PCR amplification of the shRNA inserts, and products were subcl-
oned by using the TA-cloning system (Invitrogen) followed by DNA
(B) ROS levels were assessed in cells treated for 72 hr with 10 ng/ml dox and with or without 10 mM of the antioxidant NAC for 1 hr.
In (A) and (B), as a positive control for ROS production, HOSE V12 H-Ras cells were treated with H2O2for 10 min.
(C)ThephasecontrastmicrographsofHOSEV12H-Rascellsafter48hrtreatmentwith10ng/mldoxandwithorwithout10mMoftheantioxidant
NAC.
(D) HOSE V12 H-Ras cells were treated with or without 10 mM of NAC for 30 min prior treatment with 10 ng/ml dox for further 72 hr. Whole-cell
lysates were used to perform Western blot analysis.
(E) HOSE V12 H-Ras cells were transfected with an HA-tagged wt MINK expression construct and treated with or without 10 ng/ml dox for 48 hr
after pretreatment with 5 or 10 mM of NAC for 30 minutes. In vitro kinase assays were performed with MBP as a substrate in the presence of
[g-32P]ATP. The reaction products were subjected to SDS-PAGE and autoradiography.
(F) HOSE V12 H-Ras cells were treated with indicated concentrations of H2O2for 1 hr, and total cell extracts were immunoblotted with the in-
dicated antibodies.
(G) HOSE V12 H-Ras cells and COS-7 cells were transfected with an HA-tagged wt MINK expression construct and treated with or without 5 mM
ofH2O2.Invitro kinaseassayswereperformed byusing MBPasasubstrate inthepresence of [g-32P]ATP. Thereaction products weresubjected
to SDS-PAGE and autoradiography.
In (E) and (G), relative signals were calculated as described for Figure 4C.
Error bars show standard deviation of the population from the arithmetic mean of at least three independent experiments.
Role of MINK in Ras-Induced Senescence
683

Page 12

sequencing. For more information on primers and pRS methodolo-
gies see http://www.screeninc.nl/.
Transfection of siRNA Oligos
A pool of four (SMART pools, Dharmacon) or three (Ambion) prede-
signed siRNA oligos per gene of interest were tested. The sequence
that induced the strongest knockdown of mRNA expression (as val-
idated by quantitative RT-PCR) was principally used for further ex-
periments, after confirmation that at least one other sequence
against the same gene had similar biological effects (sequences
canbefoundinTableS2).Cellsweretransfectedwith62.5nMsiRNA
(Ambion) or 70 nM SMART pool siRNA (Dharmacon) using Trans-
Messenger transfection reagent (Qiagen). After 24 hr, cells were re-
transfected by using the same protocol and, after a further 24 hr,
were set up and treated as appropriate.
mRNA Analysis by Real-Time RT-PCR
Total RNA was extracted by using the RNAeasy system (Qiagen),
and complementary DNA was synthesized (Invitrogen), used as
a template for TaqMan real-time PCR analysis using SYBR-Green
(Applied Biosystems) to measure relative transcript levels of target
genes, and normalized to 18S RNA levels. Sequences of PCR pri-
mers can be found in Table S3.
Measurement of Intracellular ROS Levels
Cells were harvested by trypsinization and incubated with 50 mM
DCF-DA for 30min at37ºC before analysis by FACS using propidium
iodide to counterstain and exclude dead cells. Cell Quest 3.2 (BD
Biosciences) software was used for analysis.
Cell Transformation Assays
Colony formation assays in soft agar were carried out as described
in Rodriguez-Viciana et al. (1997) by using tumor cell lines stably ex-
pressing pRS constructs targeting MINK, p70S6K1, and the PI 3-ki-
nase p110a (see Table S2). Effective knockdown was checked by
RT-PCR. Focus formation assay in NIH 3T3 cells cotransfected with
V12 H-Ras expression construct and dominant-negative or wt MINK
were carried out as described in Rodriguez-Viciana et al. (1997).
For further detail of experimental methods, see the Supplemental
Data.
Supplemental Data
SupplementalDataincludeSupplementalExperimentalProcedures,
five figures, and three tables and can befound with this article online
at http://www.molecule.org/cgi/content/full/20/5/673/DC1/.
Acknowledgments
We are grateful to Robert Weinberg, Miguel Martins, Charles Swan-
ton, Ippeita Dan, Martin McMahon, Samuel Mok, Christoph Sachse,
and Nelly Auersperg for sharing reagents and/or providing advice.
We also thank Graham Clark and the CRUK LRI Equipment Park
andFACSLaboratory staffforvaluable assistance.K.B.andJ.M.ac-
knowledge the Dutch Cancer Society for grant support. This work
was funded by Cancer Research UK.
Received: March 24, 2005
Revised: July 25, 2005
Accepted: October 12, 2005
Published: December 8, 2005
References
Aza-Blanc, P., Cooper, C.L., Wagner, K., Batalov, S.,Deveraux, Q.L.,
and Cooke, M.P. (2003). Identification of modulators of TRAIL-
induced apoptosis via RNAi-based phenotypic screening. Mol.
Cell 12, 627–637.
Berns, K., Hijmans, E.M., Mullenders, J., Brummelkamp, T.R., Velds,
A., Heimerikx, M., Kerkhoven, R.M., Madiredjo, M., Nijkamp, W.,
Weigelt, B., et al. (2004). A large-scale RNAi screen in human cells
identifies new components of the p53 pathway. Nature 428, 431–
437.
Bosch, E., Cherwinski, H., Peterson, D., and McMahon, M. (1997).
Mutations of critical amino acids affect the biological and biochem-
ical properties of oncogenic A-Raf and Raf-1. Oncogene 15, 1021–
1033.
Braig, M., Lee, S., Loddenkemper, C., Rudolph, C., Peters, A.H.,
Schlegelberger, B., Stein, H., Dorken, B., Jenuwein, T., and Schmitt,
C.A. (2005). Oncogene-induced senescence as an initial barrier in
lymphoma development. Nature 436, 660–665.
Bulavin,D.V.,andFornace,A.J.,Jr.(2004).p38MAPkinase’semerg-
ing role as a tumor suppressor. Adv. Cancer Res. 92, 95–118.
Bulavin, D.V., Kovalsky, O., Hollander, M.C., and Fornace, A.J., Jr.
(2003). Loss of oncogenic H-ras-induced cell cycle arrest and p38
mitogen-activated protein kinase activation by disruption of
Gadd45a. Mol. Cell. Biol. 23, 3859–3871.
Campisi, J. (2005). Senescent cells, tumor suppression, and organ-
ismal aging: good citizens, bad neighbors. Cell 120, 513–522.
Chen, Z., Trotman, L.C., Shaffer, D., Lin, H.K., Dotan, Z.A., Niki, M.,
Koutcher, J.A., Scher, H.I., Ludwig, T., Gerald, W., et al. (2005). Cru-
cial role of p53-dependent cellular senescence in suppression of
Pten-deficient tumorigenesis. Nature 436, 725–730.
Chung, J., Grammer, T.C., Lemon, K.P., Kazlauskas, A., and Blenis,
J. (1994). PDGF- and insulin-dependent pp70S6k activation medi-
ated by phosphatidylinositol-3-OH kinase. Nature 370, 71–75.
Collado, M., Gil, J., Efeyan, A., Guerra, C., Schuhmacher, A.J., Bar-
radas, M., Benguria, A., Zaballos, A., Flores, J.M., Barbacid, M.,
et al. (2005). Tumour biology: senescence in premalignant tumours.
Nature 436, 642.
Dan, I., Watanabe, N.M., Kobayashi, T., Yamashita-Suzuki, K., Fuka-
gaya, Y., Kajikawa, E., Kimura, W.K., Nakashima, T.M., Matsumoto,
K., Ninomiya-Tsuji, J., and Kusumi, A. (2000). Molecular cloning of
MINK, a novel member of mammalian GCK family kinases, which
is up-regulated during postnatal mouse cerebral development.
FEBS Lett. 469, 19–23.
Dan, I., Watanabe, N.M., and Kusumi, A. (2001). The Ste20 group
kinases as regulators of MAP kinase cascades. Trends Cell Biol.
11, 220–230.
Delgado, M.D., Vaque, J.P., Arozarena, I., Lopez-Ilasaca, M.A., Mar-
tinez, C., Crespo, P., and Leon, J. (2000). H-, K- and N-Ras inhibit
myeloid leukemia cell proliferation by a p21WAF1-dependent mech-
anism. Oncogene 19, 783–790.
Deng,Q.,Liao,R.,Wu,B.L.,andSun,P.(2004).Highintensityrassig-
naling induces premature senescence by activating p38 pathway in
primary human fibroblasts. J. Biol. Chem. 279, 1050–1059.
Downward, J. (2003). Targeting RAS signalling pathways in cancer
therapy. Nat. Rev. Cancer 3, 11–22.
Downward, J. (2004a). RNA interference. BMJ 328, 1245–1248.
Downward, J. (2004b). Use of RNA interference libraries to investi-
gate oncogenic signalling in mammalian cells. Oncogene 23,
8376–8383.
Drayton, S., Brookes, S., Rowe, J., and Peters, G. (2004). The signif-
icance of p16INK4a in cell defenses against transformation. Cell Cy-
cle 3, 611–615.
Garner,A.P.,Weston,C.R.,Todd,D.E.,Balmanno,K.,andCook,S.J.
(2002). Delta MEKK3:ER* activation induces a p38 alpha/beta
2-dependent cell cycle arrest at the G2 checkpoint. Oncogene 21,
8089–8104.
Haddad, J.J., Safieh-Garabedian, B., Saade, N.E., Kanaan, S.A., and
Land, S.C. (2001). Chemioxyexcitation (delta pO2/ROS)-dependent
release of IL-1 beta, IL-6 and TNF-alpha: evidence of cytokines as
oxygen-sensitive mediators in the alveolar epithelium. Cytokine
13, 138–147.
Hahn, W.C., Counter, C.M., Lundberg, A.S., Beijersbergen, R.L.,
Brooks, M.W., and Weinberg, R.A.(1999).Creation of human tumour
cells with defined genetic elements. Nature 400, 464–468.
Haq, R., Brenton, J.D., Takahashi, M., Finan, D., Finkielsztein, A.,
Damaraju, S., Rottapel, R., and Zanke, B. (2002). Constitutive
p38HOG mitogen-activated protein kinase activation induces per-
manent cell cycle arrest and senescence. Cancer Res. 62, 5076–
5082.
Molecular Cell
684

Page 13

Irani, K., Xia, Y., Zweier, J.L., Sollott, S.J., Der, C.J., Fearon, E.R.,
Sundaresan, M., Finkel, T., and Goldschmidt-Clermont, P.J. (1997).
Mitogenic signaling mediated by oxidants in Ras-transformed fibro-
blasts. Science 275, 1649–1652.
Iwasa,H.,Han,J.,andIshikawa, F.(2003).Mitogen-activatedprotein
kinase p38 defines the common senescence-signalling pathway.
Genes Cells 8, 131–144.
Kolfschoten, I.G., van Leeuwen, B., Berns, K., Mullenders, J., Bei-
jersbergen, R.L., Bernards, R., Voorhoeve, P.M., and Agami, R.
(2005). A genetic screen identifies PITX1 as a suppressor of RAS ac-
tivity and tumorigenicity. Cell 121, 849–858.
Lee, A.C., Fenster, B.E., Ito, H., Takeda, K., Bae, N.S., Hirai, T., Yu,
Z.X., Ferrans, V.J., Howard, B.H., and Finkel, T. (1999). Ras proteins
induce senescence by altering the intracellular levels of reactive ox-
ygen species. J. Biol. Chem. 274, 7936–7940.
Lloyd, A.C., Obermuller, F., Staddon, S., Barth, C.F., McMahon, M.,
and Land, H. (1997). Cooperating oncogenes converge to regulate
cyclin/cdk complexes. Genes Dev. 11, 663–677.
Lowe, S.W., Cepero, E., and Evan, G. (2004). Intrinsic tumour sup-
pression. Nature 432, 307–315.
Lundberg, A.S., Randell, S.H., Stewart, S.A., Elenbaas, B., Hartwell,
K.A., Brooks, M.W., Fleming, M.D., Olsen, J.C., Miller, S.W., Wein-
berg, R.A., and Hahn, W.C. (2002). Immortalization and transforma-
tion of primary human airway epithelial cells by gene transfer. Onco-
gene 21, 4577–4586.
Michaloglou, C., Vredeveld, L.C., Soengas, M.S., Denoyelle, C., Kuil-
man, T., van der Horst, C.M., Majoor, D.M., Shay, J.W., Mooi, W.J.,
and Peeper, D.S. (2005). BRAFE600-associated senescence-like
cell cycle arrest of human naevi. Nature 436, 720–724.
Mok, S.C., Chan, W.Y., Wong, K.K., Muto, M.G., and Berkowitz, R.S.
(1996). SPARC, an extracellular matrix protein with tumor-suppress-
ing activity in human ovarian epithelial cells. Oncogene 12, 1895–
1901.
Paddison, P.J., Silva, J.M., Conklin, D.S., Schlabach, M., Li, M., Ar-
uleba, S., Balija, V., O’Shaughnessy, A., Gnoj, L., Scobie, K., et al.
(2004). A resource for large-scale RNA-interference-based screens
in mammals. Nature 428, 427–431.
Pruitt, K., Pruitt, W.M., Bilter, G.K., Westwick, J.K., and Der, C.J.
(2002). Raf-independent deregulation of p38 and JNK mitogen-
activated protein kinases are critical for Ras transformation.
J. Biol. Chem. 277, 31808–31817.
Rhee, S.G., Bae, Y.S., Lee, S.R., and Kwon, J. (2000). Hydrogen per-
oxide: a key messenger that modulates protein phosphorylation
through cysteine oxidation. Sci. STKE 2000, PE1.
Ries, S., Biederer, C., Woods, D., Shifman, O., Shirasawa, S., Sasa-
zuki, T., McMahon, M., Oren, M., and McCormick, F. (2000). Oppos-
ing effects of Ras on p53: transcriptional activation of mdm2 and in-
duction of p19ARF. Cell 103, 321–330.
Rodriguez-Viciana, P., Warne, P.H., Dhand, R., Vanhaesebroeck, B.,
Gout, I., Fry, M.J., Waterfield, M.D., and Downward, J. (1994). Phos-
phatidylinositol-3-OH kinase as a direct target of Ras. Nature 370,
527–532.
Rodriguez-Viciana,P., Warne,P.H.,Khwaja, A.,Marte,B.M.,Pappin,
D., Das, P., Waterfield, M.D., Ridley, A., and Downward, J. (1997).
Role of phosphoinositide 3-OH kinase in cell transformation and
control of the actin cytoskeleton by Ras. Cell 89, 457–467.
Serrano, M., and Blasco, M.A. (2001). Putting the stress on senes-
cence. Curr. Opin. Cell Biol. 13, 748–753.
Serrano, M., Lin, A.W., McCurrach, M.E., Beach, D., and Lowe, S.W.
(1997). Oncogenic ras provokes premature cell senescence associ-
ated with accumulation of p53 and p16INK4a. Cell 88, 593–602.
Sewing, A., Wiseman, B., Lloyd, A.C., and Land, H. (1997). High-
intensity Raf signal causes cell cycle arrest mediated by p21Cip1.
Mol. Cell. Biol. 17, 5588–5597.
Shields, J.M., Mehta, H., Pruitt, K., and Der, C.J. (2002). Opposing
roles of the extracellular signal-regulated kinase and p38 mitogen-
activated protein kinase cascades in Ras-mediated downregulation
of tropomyosin. Mol. Cell. Biol. 22, 2304–2317.
Tsao, S.W., Mok, S.C., Fey, E.G., Fletcher, J.A., Wan, T.S., Chew,
E.C., Muto, M.G., Knapp, R.C., and Berkowitz, R.S. (1995). Charac-
terization of human ovarian surface epithelial cells immortalized by
human papilloma viral oncogenes (HPV-E6E7 ORFs). Exp. Cell
Res. 218, 499–507.
Wang, W., Chen, J.X., Liao, R., Deng, Q., Zhou, J.J., Huang, S., and
Sun, P. (2002). Sequential activation of the MEK-extracellular signal-
regulated kinase and MKK3/6-p38 mitogen-activated protein kinase
pathways mediates oncogenic ras-induced premature senescence.
Mol. Cell. Biol. 22, 3389–3403.
Wang, Y., Meng, A., and Zhou, D. (2004). Inhibition of phosphatidy-
linostol 3-kinase uncouples H2O2-induced senescent phenotype
and cell cycle arrest in normal human diploid fibroblasts. Exp. Cell
Res. 298, 188–196.
Westbrook,T.F.,Martin,E.S.,Schlabach, M.R.,Leng,Y.,Liang,A.C.,
Feng, B., Zhao, J.J., Roberts, T.M., Mandel, G., Hannon, G.J., et al.
(2005). A genetic screen for candidate tumor suppressors identifies
REST. Cell 121, 837–848.
Wu, C., Miloslavskaya, I., Demontis, S., Maestro, R., and Galaktio-
nov, K. (2004). Regulation of cellular response to oncogenic and ox-
idative stress by Seladin-1. Nature 432, 640–645.
Role of MINK in Ras-Induced Senescence
685