Enhanced elimination of oxidized guanine nucleotides inhibits oncogenic
RAS-induced DNA damage and premature senescence
P Rai1, JJ Young2,3,5, DGA Burton1, MG Giribaldi1, TT Onder2,3,6and RA Weinberg2,3,4
1Division of Gerontology and Geriatric Medicine, Department of Medicine, University of Miami, Miller School of Medicine, Miami,
FL, USA;2Whitehead Institute for Biomedical Research, Cambridge, MA, USA;3Department of Biology, Massachusetts Institute
of Technology, Cambridge, MA, USA and4Ludwig Center for Molecular Oncology, Cambridge, MA, USA
Approximately 20% of tumors contain activating muta-
tions in the RAS family of oncogenes. As tumors progress
to higher grades of malignancy, the expression of
oncogenic RAS has been reported to increase, leading to
an oncogene-induced senescence (OIS) response. Evasion
of this senescence barrier is a hallmark of advanced
tumors indicating that OIS serves a critical tumor-
suppressive function. Induction of OIS has been attributed
to either RAS-mediated production of reactive oxygen
species (ROS) or to induction of a DNA damage response
(DDR). However, functional links between these two
processes in triggering the senescent phenotype have not
been explicitly described. Our previous work has shown
that, in cultured untransformed cells, preventing elimina-
tion of oxidized guanine deoxyribonucleotides, which was
achieved by suppressing expression of the cellular 8-oxo-
dGTPase, human MutT homolog 1 (MTH1), sufficed to
induce a DDR as well as premature senescence. Here, we
demonstrate that overexpression of MTH1 can prevent
the oncogenic H-RAS-induced DDR and attendant
premature senescence, although it does not affect the
observed elevation in ROS levels produced by RAS
oncoprotein expression. Conversely, we find that loss of
MTH1 preferentially induces an in vitro proliferation
defect in tumorigenic cells overexpressing oncogenic RAS.
These results indicate that the guanine nucleotide pool is a
critical target for intracellular ROS produced by onco-
genic RAS and that RAS-transformed cells require robust
MTH1 expression to proliferate.
Oncogene (2011) 30, 1489–1496; doi:10.1038/onc.2010.520;
published online 15 November 2010
Keywords: cellular senescence; DNA damage; oxidative
stress; 8-oxoguanine; MTH1; RAS oncogene
Oncogene-induced senescence (OIS) (Di Micco et al., 2006),
which occurs in response to increases in oncogenic RAS
levels during tumor progression (Quintanilla et al., 1986;
Algarra et al., 1998; Sarkisian et al., 2007) is believed to
function as a major tumor-suppressor barrier. The senes-
cence response to the RAS oncoprotein has been ascribed to
its production of reactive oxygen species (ROS) (Irani et al.,
1997; Lee et al., 1999) and to the resulting induction of
a DNA damage response (DDR) (Di Micco et al., 2006;
Mallette et al., 2007). However, the connection between
these two RAS-induced effects has been unclear.
We had previously found that increasing purine
oxidation products in the nucleotide pool by suppres-
sion of the cellular 8-oxo-dGTP triphosphatase, human
MutT homolog 1 (MTH1) permits increased incorpora-
tion of this oxidized dNTP into DNA, increases DNA
double-strand breaks (DSBs) and leads to rapid induc-
tion of senescence (Rai et al., 2009). MTH1 removes
ROS-induced 8-oxoguanine from the dNTP pool,
preventing its incorporation into DNA (Nakabeppu,
2001). Others recently reported that RAS OIS in IMR90
fibroblasts leads to increased levels of cellular 8-oxo-7,
8-dihydro-20-deoxyguanosine, a signature of genomi-
cally incorporated 8-oxo-dGTP (Moiseeva et al., 2009).
Accordingly, in this study, we have determined whether
MTH1 overexpression is able to protect fibroblasts from
RAS-induced DSBs and premature senescence.
To do so, we introduced a retroviral H-RASV12 expres-
sion construct into early-passage BJ human skin control
fibroblasts or their derivatives overexpressing MTH1 and
determined the ability of H-RAS to induce premature
senescence in these cell lines. Using lentiviral short hairpin
RNA constructs, we also suppressed MTH1 expression
in isogenic cell lines to assess whether oncogenic RAS-
containing cells were sensitized to loss of MTH1 over
their non-oncogenic RAS-infected counterparts. Our stu-
dies demonstrate that cells sustaining oncogenic RAS
activation require robust MTH1 expression to proliferate.
Results and discussion
Using a retroviral MTH1 expression vector, we devel-
oped a population of early-passage human BJ skin
fibroblasts that expresses MTH1 at levels approximately
Received 21 March 2010; revised 4 October 2010; accepted 5 October
2010; published online 15 November 2010
Correspondence: Dr P Rai, Division of Gerontology and Geriatric
Medicine, Department of Medicine, University of Miami, Miller
School of Medicine, 1600 NW 10th Avenue, RMSB 7094, Locator
Code D-503, Miami, FL 33136, USA.
5Current address: Department of Medicine, University of California,
San Francisco, CA, USA.
6Current address: Children’s Hospital Boston, Harvard Medical
School, Boston, MA, USA.
Oncogene (2011) 30, 1489–1496
& 2011 Macmillan Publishers Limited All rights reserved 0950-9232/11
10-fold higher than normally expressed by these cells
(Figure 1a). We then infected control BJ fibroblasts
and these MTH1-overexpressing derivatives with an
oncogenic H-RAS vector (expressing RAS protein by
approximately sevenfold higher than endogenous levels)
or a control vector (Figure 1a) and monitored the
proliferation of these cells over the next few weeks
(Figure 1b). We found that this level of MTH1
overexpression sufficed to prevent the senescent pheno-
type that is normally induced by this level of RAS
oncoprotein in BJ cells (Figures 1a–c).
Thus, the control H-RAS-expressing cells stopped
proliferating after approximately 3 weeks, showed
senescence-associated b-galactosidase positivity as well
as upregulated levels of the p53/p21 tumor suppressor
proteins (Figures 1a–c). In contrast, the comparable
cells overexpressing both MTH1 and oncogenic H-RAS
did not display any of these senescent features (Figures
1a–c). Elevated p16INK4a levels have been reported to
be essential for the induction of RAS-induced senes-
cence (Benanti and Galloway, 2004). We found that
introductionof oncogenicRAS led toelevated
10% fetal calf serum, 100units/ml penicillin, 100mg/ml streptomycin and 2mM L-glutamine at 371C in either 21% oxygen/5% CO2
or, where specified, in 3% oxygen/5% CO2. All media reagents were from Gibco/Invitrogen (Carlsbad, CA, USA). The MTH1
overexpression construct was cloned into the retroviral pBabe vector from a plasmid expressing full-length wildtype human MTH1
(pcDEB.MTH1), a gift from Dr Yusaku Nakabeppu at Kyushu University, as described previously (Rai et al., 2009). BJ PD28 cells
were infected with either the empty retroviral pBABE.puro (pBp) or the pBp.MTH1 overexpression vector and continuously selected in
2mg/ml puromycin for the next 10 days. At PD34, these two sets were infected with either empty pBabe.hygro (pBh) vector or an
oncogenic H-RASV12-expressing vector, pBh.H-RAS. These cells were continuously selected in 200mg/ml hygromycin thereafter. The
hygromycin-selectable version of the H-RAS oncoprotein-expressing vector expresses a lower level of the protein relative to the more
commonly used puromycin-selectable (pBabe.puro) H-RAS oncoprotein expression vector. (a) Cells were collected approximately 26
days after the initial infection and selection period. Immunoblotting was carried out on 60mg of protein from the indicated samples,
using antibodies against MTH1, p53, p21 and H-RAS. Actin was use as the loading control. The following antibodies were used: p53
(Santa Cruz, FL-393), p21 (Santa Cruz, sc-817), MTH1 (Novus Biologicals, Littleton, CO, USA; NB 100–109), H-RAS (Santa Cruz,
sc-520), p16INK4a (BD Biosciences, 554079) and actin (Abcam, ab8226, Cambridge, MA, USA). Blots were developed using the ECL
Plus Chemiluminescent Detection kit (GE Healthcare, Piscataway, NJ, USA). (b) Population-doubling curves were established for the
indicated samples. Cells were selected in hygromycin for the duration of the experiment. To determine the average rate of population
doubling (PD), 4?105cells were plated in duplicate and the number of cells was counted every 3 days using a hemocytometer, with
4?105cells being re-plated for the next count. The numbers were converted into population doublings using the following formula:
[log (no. of cells counted)?log (no. of cell plated)]/log (2). Note that although the H-RAS expressing cells enter senescence, the H-RAS/
MTH1 expressing cells develop a proliferative advantage over the empty vector and MTH1-expressing cells by week 3 of the growth
curve. (c) Detection of SA b-galactosidase activity. The assay was carried out approximately 26 days after the initial infection and
selection period as described previously (Dimri et al., 1995; Rai et al., 2009). The percentage of cells exhibiting SA-b-galactosidase
activity is indicated beneath the representative images.
MTH1 overexpression prevents oncogenic RAS-induced senescence. BJ cells were maintained in DMEM supplemented with
MTH1 overexpression prevents RAS OIS
P Rai et al
p16INK4a levels, regardless of MTH1 expression
(Figure 1a), suggesting that, at least in BJ human
fibroblasts, the p16INK4a pathway does not suffice, on
its own, to induce senescence in response to RAS
In addition, we found that the ability of the RAS/
MTH1 coexpressing cells to continue proliferating in the
presence of overexpressed oncogenic RAS was not due
to inhibition of RAS-induced intracellular total ROS
levels, as measured by chloromethyl-dichlorofluorescein
diacetate (CM-DCF-DA). These levels were similar
between the RAS and the RAS/MTH1-expressing cells,
and as expected, were increased relative to the non-
oncogenic RAS-expressing cells (Figure 2a). Because
RAS signaling activates NADPH oxidases, which
produce superoxide radicals (Mitsushita et al., 2004),
cellular superoxide levels were also measured using the
superoxide radical-specific fluorophore, hydroethidine
(HEt). Consistent with the total ROS data, superoxide
levels were higher in the presence of oncogenic H-RAS,
and were independent of MTH1 expression levels
(Figure 2a). The relative difference in superoxide levels
between H-RAS-transduced and nontransduced cells is
not as great as that observed for total ROS levels
(Figure 2a). This quantitative discrepancy presumably
reflects both the highly efficient dismutation of super-
oxide into hydrogen peroxide in BJ cells (Serra et al.,
2003) and the contribution from elevated mitochondrial
hydrogen peroxide production (Lee et al., 1999;
Moiseeva et al., 2009), which is detected by chloro-
methyl-dichlorofluorescein diacetate but not hydro-
As RAS-induced senescence is accompanied by
increased DSBs (Di Micco et al., 2006; Mallette et al.,
2007), we assessed the effects of MTH1 overexpression
on RAS-induced DNA damage. In the cell populations
expressing H-RAS alone, a majority of cells (B85%)
stained positive for three or more DSB foci, as detected
by costaining cells with 53BP1 and gH2AX antibodies
(Figures 2c and d). In contrast, a smaller fraction
(B30%) of MTH1/RAS coexpressing cells exhibited
three or more DSB foci. The decreased number of DSB
foci in MTH1/H-RAS coexpressing cells correlated
with reduced intensity of staining for total cellular
8-oxoguanine levels in these cells, relative to the cells
expressing the H-RAS oncoprotein alone (Figure 2b).
This observation is consistent with our previous study
in which we demonstrated that suppression of MTH1
led to elevated total cellular 8-oxoguanine levels and
increased DSB foci (Rai et al., 2009).
An earlier study indicated that exogenously added 8-
oxo-GTP can stimulate RAS signaling through the
ERK pathway to a greater extent than can GTP (Yoon
et al., 2005), ostensibly because of the ability of 8-oxo-
GTP bound by the RAS protein to place it in an
activated, signal-emitting state. This suggested that
overexpression of MTH1 might prevent H-RAS-in-
duced senescence by inhibiting the RAS signaling
pathway, doing so by reducing cellular 8-oxo-GTP
levels (Nakabeppu, 2001). Accordingly, to evaluate the
possible occurrence of such a mechanism, we monitored
phospho-ERK1/2 levels—a downstream indicator of
RAS signaling—and did not find any clear-cut correla-
tions between MTH1 expression, total cellular 8-
oxoguanine levels (Figures 1a and 2b) and the extent
of ERK signaling (Figure 3). Hence, suppression of the
RAS signaling pathway does not appear to be respon-
sible for the MTH1-mediated prevention of oncogenic
RAS-induced senescence. The observation that MTH1
overexpression had no effect on RAS-induced ROS
production itself (Figure 2a), when taken together
with our results in Figures 2b–d and the fact that
MTH1 exerts its detoxification effects only on oxidized
guanine nucleotides rather than on genomic 8-dihydro-
20-deoxyguanosine (Nakabeppu, 2001), allowed us to
conclude that the observed oncogenic RAS-induced
DNA damage derived from oxidation of guanine deoxy-
ribonucleotides and their incorporation into chromo-
RAS-transformed tumor cells upregulate ROS levels
and are more susceptible to oxidative stress relative to
their untransformed counterparts (Trachootham et al.,
2006; Yagoda et al., 2007), likely through imbalances in
their antioxidant pathways (Trachootham et al., 2009)
(Supplementary Figure S1). In BJ cells, oncogenic RAS
expression upregulates MTH1 levels (Figure 1a). Simi-
larly, when we compared MTH1 expression levels in
MCF7 human breast cancer cells versus their trans-
formed derivatives, MCF7-RAS (Kasid et al., 1985), as
well as in HMLE versus HMLE-RAS cells (experimen-
tally immortalized and transformed human mammary
epithelial cells) (Elenbaas et al., 2001), we found that in
each case MTH1 levels were higher in the oncogenic
RAS-transformed counterpart cells (Figure 4a). Thus,
we undertook to determine whether suppression of
MTH1 expression would affect the proliferation of
RAS-transformed cells relative to the corresponding
non-oncogenic RAS-expressing cell line. Accordingly,
we infected the two sets of isogenic cell lines (MCF7/
MCF7-RAS and HMLE/HMLE-RAS) with either an
shGFP control vector or an shMTH1 construct
(Figure 4a). We have previously characterized this
shMTH1 construct and found that it reduces MTH1
expression by greater than 90% relative to the control
shGFP-transduced cells (Rai et al., 2009).
Although MTH1 knockdown had minimal effect on
the proliferation of the non-oncogenic RAS-expressing
counterparts in each set (that is, the MCF7 and HMLE
cells), it evoked a significant decrease in the growth rate
of the corresponding RAS-transformed cell populations
(Figures 4b and c). In the case of MCF7-RAS cells,
which exhibit intact p53/p21 function, this slowing of
proliferation was accompanied by an increase in
p21Cip1/Waf1 protein levels (Figure 4a) and senes-
cence-associated b-galactosidase staining (Figure 4d) as
well as increased total cellular 8-oxoguanine staining
(Supplementary Figure S2c), suggesting an MTH1 loss-
dependent activation of the senescence pathway in the
MCF7-RAS cells but not in the MCF7 cells (Figure 4d,
Supplementary Figure S2b). Although HMLE and
HMLE-RAS cells contain SV40 large T antigen and
are, therefore, protected from senescence, introduction
MTH1 overexpression prevents RAS OIS
P Rai et al
cellular ROS levels. The indicated cells were collected at equivalent confluency through trypsination, washed in ice-cold 1X Hank’s
buffered saline solution (HBSS) and incubated with freshly prepared 10mM hydroethidine (HEt, Molecular Probes/Invitrogen, D11347)
or 5-(and -6)-chloromethyl-20,70-dichlorofluorescein diacetate (CM-DCF-DA; Molecular Probes/Invitrogen, C6827) for 20min at 371C.
The cells were then washed and resuspended in 1X HBSS before detection of FITC signal through fluorescence-activated cell sorting
(FACS) on a FACScalibur machine. The abscissa, FL1-H or FL2-H, represents signal intensity from the PI (HEt) or FITC (DCF-DA)
channel and the ordinate represents the cell counts. Roughly equal number of cells was assayed for all samples. ROS levels were measured
approximately 3 weeks after the initial infection. Note that, although the RAS-transformed cells have a higher intensity of fluorescence
than the untransformed cells, MTH1 expression does not substantially alter ROS levels in the background of RAS oncoprotein
expression. (b) Detection of total cellular 8-oxoguanine. Approximately 30000–60000 cells were plated in four-well chamber slides (BD
Biosciences, Franklin Lakes, NJ, USA), 24–48h before fixation. Total cellular 8-oxoguanine was detected as previously described
(Struthers et al., 1998; Radisky et al., 2005; Rai et al., 2009). Staining procedures were carried out on cells at approximately 3 weeks after
the initial infection. The FITC/DAPI-merged images of 8-oxoG staining were generated by the AxioCam AxioVision/Zeiss (Thornwood,
NY, USA) software. Representative fields are shown. (c) Detection of DSB foci. Sample preparation was carried out as in (b) and staining
as previously described (Rai et al., 2009). All images were acquired with identical acquisition parameters between different samples using
the AxioCam Axiovision software. The merged DSB foci images were generated by overlaying separately photographed 53BP1 (green)
and gH2AX (red) channels of the same field in Photoshop and equivalently adjusting the contrast for all images to improve image quality.
Representative images are shown. (d) Quantitation of the DSB foci in Figure 2c. Cells were counted as foci-positive, if they had three or
more distinct 53BP1 or gamma-H2AX (gH2AX) foci. Five different fields comprising 30–50 cells were scored for each sample.
MTH1 overexpression prevents oncogenic RAS-induced DNA damage without affecting ROS levels. (a) Measurement of
MTH1 overexpression prevents RAS OIS
P Rai et al
of shMTH1 into HMLE-RAS, but not HMLE cells,
nonetheless reduced their proliferation rate significantly
(Figures 4c and d) and also increased their total cellular
8-oxoguanine levels (Supplementary Figures S2e and f).
These selective effects were not due to differences in
MTH1 knockdown between the RAS-transformed and
nontransformed cell lines (Figure 4a) nor were they due
to increased cell death on MTH1 suppression (Supple-
mentary Figure S4a). Consistent with our previous
observations (Rai et al., 2009), the shMTH1-induced
proliferation defect in both MCF7-RAS and HMLE-
RAS cells could be prevented by culturing the cells at
3% oxygen before short hairpin RNA transduction
(Supplementary Figures S3, S4), a condition that
reduces cellular 8-oxoguanine formation in these cells
(Supplementary Figures S3a, S4b), and thus presumably
the requirement for MTH1 function.
These results demonstrate that loss of MTH1 creates
a proliferation defect in established oncogenic RAS-
expressing tumor cells, either through reactivation of a
senescence-like phenotype, as is the case for MCF7-RAS
cells, or through as-yet unidentified effects on cell
cycle progression as observed in the HMLE-RAS cells.
Collectively, our data suggest that robust MTH1
expression facilitates proliferation of cells sustaining
oncogenic RAS expression, regardless of whether they
are normal or transformed cells.
Oncogenic RAS-mediated production of ROS has
been demonstrated to be essential to its ability to induce
senescence (Lee et al., 1999). Chemically, the types of
ROS generated by oncogenic RAS, that is, hydrogen
peroxide and superoxide radicals, have very low
oxidation potential and are unable to directly damage
DNA. However, superoxide can potentiate formation
of the highly-damaging hydroxyl radical by the reaction
of hydrogen peroxide with iron in a Fenton reaction
(reviewed in Kawanishi et al., 2001). Oncogenic RAS
has further been reported to upregulate cellular levels of
iron (Yang and Stockwell, 2008).
pool is relatively unprotected against cellular oxidants
(Haghdoost et al., 2006), which are generated in the
mitochondria or the cytoplasm. Furthermore, guanine
deoxynucleotides and deoxynucleosides in particular
can readily associate with labile intracellular iron
(Gackowski et al., 2002; reviewed in Kruszewski,
2003), thus rendering them especially vulnerable to the
Fenton reaction-generated hydroxyl radical. Although
the precise degree of damage preference is difficult to
measure because of the dynamic nature of the nucleotide
pool, solution studies indicate that the extent of 8-
oxoguanine formed on free nucleotides is approximately
10-fold greater than that formed on intact calf thymus
DNA (Kasai and Nishimura, 1984). Thus, given the
extremely reactive nature of the hydroxyl radical, it is
much more likely that any such short-lived species will
encounter and react with DNA nucleotides before
damaging genomic DNA.
As nuclear DNA damage is a triggering factor in
oncogenic RAS-induced senescence, our results suggest
that the deoxyribonucleotide pool is an Achilles heel for
cells under oxidative stress, and that incorporation of
these oxidized DNA precursors eventually leads to
genomic damage, through generation of DNA abasic
sites and strand breaks (Rai et al., 2009). Mouse models
of OIS implicate DNA damage as a likely agent for
triggering a proliferation arrest in vivo (Bartkova et al.,
2006; Di Micco et al., 2006). Our study indicates that
were collected B10 days after infection to allow cells to recover from selection and grow to equivalent confluency. For measurements
of steady-state ERK1/2 activation, cells growing in DME/10% serum at equivalent confluency were collected and 50mg of their protein
lysates were immunoblotted against the indicated antibodies. For measuring kinetics of ERK1/2 activation, cells at equivalent
confluency (B80%) were grown under low serum (DME/1% serum) conditions for 48h and were then serum-stimulated with DME/
10% serum for the indicated time-periods before collecting. Protein (40–50mg) was probed against the equivalent antibodies (rabbit
polyclonal p-ERK1/2 antibody. Cell Signaling, Danvers, MA, USA, cat#4376; rabbit polyclonal total ERK1/2 antibody, Cell
Signaling, cat#9102). (a) Kinetics of ERK1/2 activation in oncogenic RAS-expressing cells with and without coexpression of MTH1.
Similar ERK1/2 activation kinetics were observed for the RAS oncoprotein-expressing and oncogenic RAS/MTH1-coexpressing cells,
as well as for the empty vector and MTH1-overexpressing cells at the 10-min serum stimulation time point (data not shown).
(b) Steady-state activated ERK1/2 levels in oncogenic RAS-expressing cells with and without coexpression of MTH1.
MTH1 expression does not show clear-cut effects on downstream oncogenic RAS signaling through the ERK pathway. Cells
MTH1 overexpression prevents RAS OIS
P Rai et al
maintained in DMEM supplemented with 10% fetal calf serum, 100units/ml penicillin, 100mg/ml streptomycin and 2mM L-glutamine
at 371C in either 21% oxygen/5% CO2or, where specified, in 3% oxygen/5% CO2. HMLE and HMLE-RAS breast epithelial cell lines
were derived from human mammary epithelial cells (HMECs) from Clonetics and maintained in supplemented MEGM media
(Clonetics/Lonza, Walkersville, MD, USA), as described (Elenbaas et al., 2001). The shRNA design, lentivirus production and
infection was done as described (Stewart et al., 2003; Rai et al., 2009). The target sequence used in this study is common to all known
transcript variants of MTH1: shMTH1: 50GAAATTCCACGGGTACTTCAA 30(Rai et al., 2009). The control shRNA was targeted
against GFP. (a) Western blotting. Immunoblotting was carried out on 50mg of protein from the indicated samples, using the noted
antibodies with actin as the loading control. Note that in each set, baseline MTH1 expression (in the shGFP controls) is higher in the
RAS-transformed cells relative to the non-oncogenic RAS-expressing cells. Note also that the differences in the proliferation rates
observed in (c, d) are neither because of discrepancies in MTH1 suppression between the RAS-transformed and untransformed
counterpart lines nor because of unequal expression of RAS oncoprotein in the shGFP-infected versus shMTHI-infected cells. (b) Rate
of proliferation in MCF7 and MCF7-RAS cells. MCF7 and MCF7-RAS cells were transduced with either the shGFP (circles) or
shMTH1 (triangles) vector and kept under puromycin selection for the duration of the experiment. At 4 days after infection, 2?105
cells were plated for proliferation curves in a set of 12 plates for a 4-day growth curve and on each subsequent day, plates were counted
in triplicate with fresh media being added to the remaining plates. **A two-tailed Student’s t-test indicated Po0.01. (c) Rate of
proliferation in HMLE and HMLE-RAS cells. HMLE cells were transduced with pWZL.blast.H-RAS and selected continuously in
10mg/ml blasticidin-containing media. HMLE and HMLE-RAS cells were infected with either the shGFP (circles) or shMTH1
(triangles) vector and kept under puromycin selection for the duration of the experiment. At 4 days after infection, 1?105cells were
plated for proliferation curves as described in (b). **A two-tailed Student’s t-test indicated Po0.01. (d) Cell density and morphology.
MCF7 and MCF7-RAS cells infected with shGFP or shMTH1 were stained for SA-b-galactosidase activity on day 4 of the
proliferation curve in (b). Note that the shMTH1 MCF7-RAS cells show a noticeable increase in SA-b-galactosidase staining relative
to control shGFP cells, whereas the shMTH1 MCF7 cells do not (left). Percentage positive cells with s.d.’s are indicated below the
respective contrast images. Representative phase-contrast image fields (right) show cell density of the indicated samples at day 4 of the
HMLE and HMLE-RAS proliferation curve in (c) to indicate the substantial proliferation defect imparted by MTH1 suppression on
the latter cells.
Suppression of MTH1 selectively inhibits proliferation of RAS-transformed cells. MCF7 and MCF7-RAS cell lines were
MTH1 overexpression prevents RAS OIS
P Rai et al
RAS-transformed tumorigenic cells have a greater
dependence on robust MTH1 expression than their
non-oncogenic RAS-containing counterparts, ostensibly
because of the greater oxidative stress sustained by the
transformed cells (Lee et al., 1999) (Supplementary
Although replication stress, defined as the deleterious
effects of partially replicated DNA persisting in the
nucleus, has been proposed as a mechanism for the
observed DNA damage (Bartkova et al., 2006; Di Micco
et al., 2006), our results would suggest that a significant
part of the DNA damage and genetic instability
observed in RAS-transformed cells originates from the
oxidation products in the deoxynucleotide pool created
by oncogenic RAS-induced ROS. It remains possible
that oncogenic RAS-induced hyperproliferation (Di
Micco et al., 2006) coupled with increased levels of
oxidized guanine deoxynucleotides, leads to an even
greater incorporation of these products into nuclear
DNA during replication, in which they can contribute to
replication stress. In support of this idea, our previous
study has shown that increasing oxidation of guanine
nucleotide DNA precursors by MTH1 suppression leads
to a DDR and induction of cell senescence in a DNA
replication-dependent manner (Rai et al., 2009).
Importantly, production of superoxide radicals by
oncogenic RAS is also essential to its tumorigenic
function, as RAS-mediated transformation can be
abrogated by suppression of the NADPH oxidase
(Nox1) that is upregulated by RAS and represents the
source of RAS-induced superoxide (Mitsushita et al.,
2004). Hence, RAS-transformed cells must deal with the
DNA damage-inducing effects of ROS without elim-
inating ROS production entirely. This would suggest
that such cells may be compensating for high levels of
RAS oncoprotein signaling by increasing MTH1 levels
to enhance detoxification of an oxidative lesion that
contributes to the DDR and senescence induction
(Moiseeva et al., 2009; Rai et al., 2009). Indeed, analysis
of the ONCOMINE data sets indicates that MTH1
levels are significantly elevated in pancreatic adeno-
carcinomas and in nonsmall cell lung carcinomas, two
human cancer types characterized by endogenous RAS
mutations (Garber et al., 2001; Buchholz et al., 2005).
Taken together, these various observations indicate that
the oxidative stress and senescence response elicited
by the RAS oncoprotein acts in large part through
oxidation of substrates within the guanine deoxynucleo-
tide pool, and that the resulting incorporation of
oxidized dNTPs explains the DDR evoked as part of
the resulting oncogenic RAS-induced senescence.
Conflict of interest
The authors declare no conflict of interest.
We are grateful to Anisleidys Munoz for experimental
assistance. We thank Dr Carlos Perez-Stable and Dr Ramiro
Verdun for helpful comments on this manuscript. RAW is a
Professor at American Cancer Society Research and at Daniel
K Ludwig Cancer Research. This work was supported by a
Leukemia and Lymphoma Society Postdoctoral fellowship
and a James and Esther King Florida Biomedical Research
Program New Investigator grant (to PR), a Howard Hughes
Medical Institute summer undergraduate fellowship (to JJY)
and grants from the Ellison Medical Foundation for Aging
Research, The Ludwig Center for Molecular Oncology and the
Breast Cancer Research Fund (to RAW).
Algarra I, Perez M, Serrano MJ, Garrido F, Gaforio JJ. (1998).
c-K-ras overexpression is characteristic for metastases derived from
a methylcholanthrene-induced fibrosarcoma. Invasion Metastasis
Bartkova J, Rezaei N, Liontos M, Karakaidos P, Kletsas D, Issaeva N
et al. (2006). Oncogene-induced senescence is part of the tumori-
genesis barrier imposed by DNA damage checkpoints. Nature 444:
Benanti JA, Galloway DA. (2004). Normal human fibroblasts are
resistant to RAS-induced senescence. Mol Cell Biol 24: 2842–2852.
Buchholz M, Braun M, Heidenblut A, Kestler HA, Kloppel G,
Schmiegel W et al. (2005). Transcriptome analysis of microdissected
pancreatic intraepithelial neoplastic lesions. Oncogene 24: 6626–6636.
Di Micco R, Fumagalli M, Cicalese A, Piccinin S, Gasparini P,
Luise C et al. (2006). Oncogene-induced senescence is a DNA
damage response triggered by DNA hyper-replication. Nature 444:
Dimri G, Lee X, Basile G, Acosta M, Scott G, Roskelley C et al.
(1995). A biomarker that identifies senescent human cells in
culture and in aging skin in vivo. Proc Natl Acad Sci USA 92:
Elenbaas B, Spirio L, Koerner F, Fleming MD, Zimonjic DB,
Donaher JL et al. (2001). Human breast cancer cells generated by
oncogenic transformation of primary mammary epithelial cells.
Genes Dev 15: 50–65.
Gackowski D, Kruszewski M, Bartlomiejczyk T, Jawien A, Ciecierski
M, Olinski R. (2002). The level of 8-oxo-7,8-dihydro-20-deoxygua-
nosine is positively correlated with the size of the labile iron pool in
human lymphocytes. J Biol Inorg Chem 7: 548–550.
Garber ME, Troyanskaya OG, Schluens K, Petersen S, Thaesler Z,
Pacyna-Gengelbach M et al. (2001). Diversity of gene expression
in adenocarcinoma of the lung. Proc Natl Acad Sci USA 98:
Haghdoost S, Sjolander L, Czene S, Harms-Ringdahl M. (2006).
The nucleotide pool is a significant target for oxidative stress.
Free Radic Biol Med 41: 620–626.
Irani K, Xia Y, Zweier JL, Sollott SJ, Der CJ, Fearon ER et al. (1997).
Mitogenic signaling mediated by oxidants in Ras-transformed
fibroblasts. Science 275: 1649–1652.
Kasai H, Nishimura S. (1984). Hydroxylation of deoxyguanosine
at the C-8 position by ascorbic acid and other reducing agents.
Nucleic Acids Res 12: 2137–2145.
Kasid A LM, Papageorge AG, Lowy DR, Gelmann EP. (1985).
Transfection of v-rasH DNA into MCF-7 human breast cancer cells
bypasses dependence on estrogen for tumorigenicity. Science 228:
MTH1 overexpression prevents RAS OIS
P Rai et al
Kawanishi S, Hiraku Y, Oikawa S. (2001). Mechanism of guanine-
specific DNA damage by oxidative stress and its role in
carcinogenesis and aging. Mutat Res 488: 65–76.
Kruszewski M. (2003). Labile iron pool: the main determinant of
cellular response to oxidative stress. Mutat Res 531: 81–92.
Lee AC, Fenster BE, Ito H, Takeda K, Bae NS, Hirai T et al. (1999).
Ras proteins induce senescence by altering the intracellular levels of
reactive oxygen species. J Biol Chem 274: 7936–7940.
Mallette FA, Gaumont-Leclerc M-F, Ferbeyre G. (2007). The DNA
damage signaling pathway is a critical mediator of oncogene-
induced senescence. Genes Dev 21: 43–48.
Mitsushita J, Lambeth JD, Kamata T. (2004). The superoxide-
generating oxidase Nox1 is functionally required for Ras oncogene
transformation. Cancer Res 64: 3580–3585.
Moiseeva O, Bourdeau V, Roux A, Deschenes-Simard X, Ferbeyre G.
(2009). Mitochondrial dysfunction contributes to oncogene-induced
senescence. Mol Cell Biol 29: 4495–4507.
Nakabeppu Y. (2001). Molecular genetics and structural biology of
human MutT homolog, MTH1. Mutat Res 477: 59–70.
Quintanilla M, Brown K, Ramsden M, Balmain A. (1986). Carcino-
gen-specific mutation and amplification of Ha-ras during mouse
skin carcinogenesis. Nature 322: 78–80.
Radisky DC, Levy DD, Littlepage LE, Liu H, Nelson CM,
Fata JE et al. (2005). Rac1b and reactive oxygen species
mediate MMP-3-induced EMT and genomic instability. Nature
Rai P, Onder TT, Young JJ, McFaline JL, Pang B, Dedon PC et al.
(2009). Continuous elimination of oxidized nucleotides is necessary
to prevent rapid onset of cellular senescence. Proc Natl Acad Sci
USA 106: 169–174.
Sarkisian CJ, Keister BA, Stairs DB, Boxer RB, Moody SE, Chodosh
LA. (2007). Dose-dependent oncogene-induced senescence in vivo
and its evasion during mammary tumorigenesis. Nat Cell Biol 9:
Serra V, von Zglinicki T, Lorenz M, Saretzki G. (2003). Extracellular
superoxide dismutase is a major antioxidant in human fibroblasts
and slows telomere shortening. J Biol Chem 278: 6824–6830.
Stewart SA, Dykxhoorn DM, Palliser D, Mizuno H, Yu EY, An DS
et al. (2003). Lentivirus-delivered stable gene silencing by RNAi in
primary cells. RNA 9: 493–501.
Struthers L, Patel R, Clark J, Thomas S. (1998). Direct detection
of 8-oxodeoxyguanosine and 8-oxoguanine by avidin and its
analogues. Anal Biochem 255: 20–31.
Trachootham D, Alexandre J, Huang P. (2009). Targeting cancer cells
by ROS-mediated mechanisms: a radical therapeutic approach?
Nat Rev Drug Discov 8: 579–591.
Trachootham D, Zhou Y, Zhang H, Demizu Y, Chen Z, Pelicano H
et al. (2006). Selective killing of oncogenically transformed cells
through a ROS-mediated mechanism by [beta]-phenylethyl isothio-
cyanate. Cancer Cell 10: 241–252.
Yagoda N, von Rechenberg M, Zaganjor E, Bauer AJ, Yang WS,
Fridman DJ et al. (2007). RAS-RAF-MEK-dependent oxidative
cell death involving voltage-dependent anion channels. Nature 447:
Yang WS, Stockwell BR. (2008). Synthetic lethal screening identifies
compounds activating iron-dependent, nonapoptotic cell death in
oncogenic-RAS-harboring cancer cells. Chem Biol 15: 234–245.
Yoon S-H, Hyun J-W, Choi J, Choi E-Y, Kim H-J, Lee S-J et al. (2005).
in vitro evidence for the recognition of 8-oxoGTP by Ras, a small
GTP-binding protein. Biochem Biophys Res Commun 327: 342–348.
Supplementary Information accompanies the paper on the Oncogene website (http://www.nature.com/onc)
MTH1 overexpression prevents RAS OIS
P Rai et al