Inactivation of SAG E3 ubiquitin ligase blocks embryonic stem cell differentiation and sensitizes leukemia cells to retinoid acid.

Abstract

Sensitive to Apoptosis Gene (SAG), also known as RBX2 (RING box protein-2), is the RING component of SCF (SKP1, Cullin, and F-box protein) E3 ubiquitin ligase. Our previous studies have demonstrated that SAG is an anti-apoptotic protein and an attractive anti-cancer target. We also found recently that Sag knockout sensitized mouse embryonic stem cells (mES) to radiation and blocked mES cells to undergo endothelial differentiation. Here, we reported that compared to wild-type mES cells, the Sag(-/-) mES cells were much more sensitive to all-trans retinoic acid (RA)-induced suppression of cell proliferation and survival. While wild-type mES cells underwent differentiation upon exposure to RA, Sag(-/-) mES cells were induced to death via apoptosis instead. The cell fate change, reflected by cellular stiffness, can be detected as early as 12 hrs post RA exposure by AFM (Atomic Force Microscopy). We then extended this novel finding to RA differentiation therapy of leukemia, in which the resistance often develops, by testing our hypothesis that SAG inhibition would sensitize leukemia to RA. Indeed, we found a direct correlation between SAG overexpression and RA resistance in multiple leukemia lines. By using MLN4924, a small molecule inhibitor of NEDD8-Activating Enzyme (NAE), that inactivates SAG-SCF E3 ligase by blocking cullin neddylation, we were able to sensitize two otherwise resistant leukemia cell lines, HL-60 and KG-1 to RA. Mechanistically, RA sensitization by MLN4924 was mediated via enhanced apoptosis, likely through accumulation of pro-apoptotic proteins NOXA and c-JUN, two well-known substrates of SAG-SCF E3 ligase. Taken together, our study provides the proof-of-concept evidence for effective treatment of leukemia patients by RA-MLN4924 combination.

Full text

Page 1

Inactivation of SAG E3 Ubiquitin Ligase Blocks Embryonic
Stem Cell Differentiation and Sensitizes Leukemia Cells
to Retinoid Acid
Mingjia Tan1., Yun Li1,2., Ruiguo Yang3, Ning Xi3, Yi Sun1*
1Division of Radiation and Cancer Biology, Department of Radiation Oncology, University of Michigan, Ann Arbor, Michigan, United States of America, 2Key Laboratory of
Marine Drugs, Chinese Ministry of Education, and School of Medicine and Pharmacy, Ocean University of China, Qingdao, China, 3Department of Electrical and Computer
Engineering, Michigan State University, East Lansing, Michigan, United States of America
Abstract
Sensitive to Apoptosis Gene (SAG), also known as RBX2 (RING box protein-2), is the RING component of SCF (SKP1, Cullin,
and F-box protein) E3 ubiquitin ligase. Our previous studies have demonstrated that SAG is an anti-apoptotic protein and an
attractive anti-cancer target. We also found recently that Sag knockout sensitized mouse embryonic stem cells (mES) to
radiation and blocked mES cells to undergo endothelial differentiation. Here, we reported that compared to wild-type mES
cells, the Sag2/2mES cells were much more sensitive to all-trans retinoic acid (RA)-induced suppression of cell proliferation
and survival. While wild-type mES cells underwent differentiation upon exposure to RA, Sag2/2mES cells were induced to
death via apoptosis instead. The cell fate change, reflected by cellular stiffness, can be detected as early as 12 hrs post RA
exposure by AFM (Atomic Force Microscopy). We then extended this novel finding to RA differentiation therapy of leukemia,
in which the resistance often develops, by testing our hypothesis that SAG inhibition would sensitize leukemia to RA.
Indeed, we found a direct correlation between SAG overexpression and RA resistance in multiple leukemia lines. By using
MLN4924, a small molecule inhibitor of NEDD8-Activating Enzyme (NAE), that inactivates SAG-SCF E3 ligase by blocking
cullin neddylation, we were able to sensitize two otherwise resistant leukemia cell lines, HL-60 and KG-1 to RA.
Mechanistically, RA sensitization by MLN4924 was mediated via enhanced apoptosis, likely through accumulation of pro-
apoptotic proteins NOXA and c-JUN, two well-known substrates of SAG-SCF E3 ligase. Taken together, our study provides
the proof-of-concept evidence for effective treatment of leukemia patients by RA-MLN4924 combination.
Citation: Tan M, Li Y, Yang R, Xi N, Sun Y (2011) Inactivation of SAG E3 Ubiquitin Ligase Blocks Embryonic Stem Cell Differentiation and Sensitizes Leukemia Cells
to Retinoid Acid. PLoS ONE 6(11): e27726. doi:10.1371/journal.pone.0027726
Editor: Wafik S. El-Deiry, Penn State Hershey Cancer Institute, United States of America
Received July 5, 2011; Accepted October 23, 2011; Published November 15, 2011
Copyright: ? 2011 Tan et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted
use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This work is supported by the National Cancer Institute grants CA111554, CA118762 and CA156744 to YS. This work is also supported by National
Science Foundation grant IIS0713346 and National Institutes of Health grant R43 GM084520 to NX. YL was supported in part by the fellowship from China
Scholarship Council and Ocean University of China. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the
manuscript.
Competing Interests: The authors have declared that no competing interests exist.
* E-mail: sunyi@umich.edu
. These authors contributed equally to this work.
Introduction
SAG, also known as RBX2, ROC2 (Regulator of Cullins) or
RNF7 (RING finger protein-7), was originally cloned in our
laboratory as a redox-inducible antioxidant protein [1], and later
characterized as the second RING family member of the SCF E3
ubiquitin ligase (for review, see [2]). We and others have previously
shown that in cell culture systems, SAG overexpression inhibited
apoptosis induced by various agents, including redox [1,3], nitric
oxide [4], ischemia/hypoxia [5], heat shock [6], neurotoxins and
1-methyl-4-phenylpyridinium [7], and UV-irradiation [8]. SAG
over-expression also promoted the S-phase entry and cell growth
under serum starved conditions [9]. Furthermore, SAG transgenic
expression in mouse skin inhibited tumor formation at the early
stage, but enhances tumor growth at the later stage in a DMBA-
TPA carcinogenesis model [10]. On the other hand, SAG
knockdown by anti-sense or siRNA oligoes inhibited tumor cell
growth [11], and enhanced apoptosis induced by etoposide and
TRAIL [12]. The Sag knockout in mouse caused embryonic
lethality, which is associated with growth retardation and other
developmental defects [13], whereas Sag knockout in mES cells
induced radiosensitization [14], and blocked their endothelial
differentiation [13]. These cellular functions were mediated via its
antioxidant activity by scavenging ROS [1,4,6], and via its E3
ubiquitin ligase activity by promoting the degradation of p27, c-
Jun, pro-caspase-3, IkBa, HIF-1a, and NOXA in a cell context
dependent manner [9,10,12,15,16,17]. Importantly, SAG was
overexpressed in carcinomas of lung, colon, stomach and liver,
which was associated with poor prognosis in lung cancer patients
[11,17,18]. The findings suggest that SAG may play a role in
human tumorigenesis and could serve as an anticancer target.
Retinoic acids (RAs) are natural and synthetic derivatives of
vitamin A [19]. The binding of RAs to their nuclear receptors,
including RAR or RXR, PPAR in the homo- and hetero-dimer
formats [20], caused transactivation of more than 500 genes [21]
that regulates many signaling pathways [22], and cellular processes
including embryonic development, organogenesis, cell homeosta-
sis, cell growth, differentiation and apoptosis [23,24,25,26]. Given
PLoS ONE | www.plosone.org1November 2011 | Volume 6 | Issue 11 | e27726

Page 2

its activity in inducing differentiation, apoptosis, and growth arrest,
RA has been long used for the treatment and prevention of various
types of human cancers [19,27,28,29,30], particularly as a
differentiation therapy agent for the treatment of AML (Acute
Myeloid Leukemia) [31,32]. Significantly, RA is very effective for
the treatment of one type AML, acute promyelocytic leukemia
(APL) with a cure rate in more than 75% of patients [26].
However, the complete success of this therapy was hindered by the
development of RA resistance with subsequent disease relapses
[30], demanding combinational therapies, most effectively with
arsenic trioxide [33], and other chemotherapeutic drugs, as well as
targeted therapies, including TRAIL [26,27,30].
Here we report that cellular sensitivity to RA is significantly
increased upon inactivation of SAG E3 ubiquitin ligase via
genetic deletion or pharmacological inhibition. Specifically,
instead of inducing differentiation in wild type mES cells, RA
induced apoptosis in Sag2/2mES cells. In two SAG high-
expressing AML lines, HL60 and KG-1, MLN4924, a small
molecule inhibitor of SAG E3, effectively sensitized otherwise
resistant cells to RA via induction of apoptosis which is associated
with accumulation of pro-apoptotic proteins, c-JUN and NOXA.
Thus, our study reveals that SAG is a RA sensitizing target and
provides the first proof-of-concept evidence that MLN4924 can
be further developed as a RA chemo-sensitizer for the treatment
of AML.
Results
Sag deletion sensitized mouse embryonic stem cell to RA
Our recent study showed that Sag2/2mES cells failed to
undergo endothelial differentiation to form cystic embryoid bodies
upon the withdrawal of LIF (leukemia inhibitory factor) from
culture media [13]. Here we determined if Sag2/2mES cells were
also defective in RA-induced differentiation. Using two indepen-
dent pairs of mES clones with Sag+/+vs. Sag2/2background [14],
we first determined the effect of Sag deletion on cell proliferation
and survival in the absence and presence of RA. While Sag
deletion had no effect on monolayer and clonogenic growth, it
significantly sensitized mES cells to RA. The IC50 value (the
concentration that caused 50% growth inhibition), determined by
ATPlite-based cell proliferation assay, was 3.4- to 5.7-fold lower in
Sag2/2mES cells than Sag+/+mES cells (Fig. 1A). Likewise, the
clonogenic survival assay showed a similar 3- to 4-fold sensitization
to RA upon Sag deletion (Fig. 1B, 1C). Thus, Sag deletion
sensitizes mES cells to RA.
Sag deletion impaired ES cell differentiation induced by
RA
We next used one representative pair of mES cells (AB1/Sag2/2
vs. AB2/Sag+/+) to determine if Sag2/2mES cells failed to undergo
RA-induced differentiation. We plated mES cells in low density in
Figure 1. Sag deletion sensitizes mES cells to all-trans Retinoid Acid (RA). Two independent pairs of mES cells with genotypes of Sag+/+and
Sag2/2were seeded in 96-well plate in triplicate and treated with various concentrations of RA for 24 hrs. Cell viability was measured by ATP-lite
assay. The IC50 curve was generated and IC50 value calculated by the Prism software (A). Cells were seeded in 6-well plate in duplicate and treated
with DMSO vehicle control or RA (0.1 and 1 mM) for 6 days. The plates were stained (B) and the colonies with more than 50 cells were counted and
plotted (C). Shown is x 6 SEM from three independent experiments.
doi:10.1371/journal.pone.0027726.g001
RA Sensitization by Inactivation of SAG E3 Ligase
PLoS ONE | www.plosone.org 2November 2011 | Volume 6 | Issue 11 | e27726

Page 3

gelatin-coated plate, treated with RA at 0, 0.1 and 1 mM for 6 days,
and then stained with alkaline phosphatase (AP), a marker for
undifferentiated ES cells [34]. In the absence of RA treatment,
Sag2/2and Sag+/+mES cells showed no difference in colonic
growth and in the maintenance of the undifferentiated status
(Fig. 2A). In the presence of RA treatment (at both 0.1 and 1.0 mM),
however, the proliferation of Sag2/2mES cells were severely
inhibited with remarkable reduction in the number of undifferen-
tiated colonies (Fig. 2A). Examination microscopically revealed
many differentiated cells from Sag+/+mES cells, whereas only
scarce cell debris were observed in Sag2/2mES cells (Fig. 2B),
suggesting a robust cell killing. We also used AFM (Atomic Force
Microscopy), which measures the cell mechanics, to determine
potential early changes during RA-induced differentiation. As
shown inFigure2C,the cellularstiffness was similarbetweenSag+/+
andSag2/2mEScellsat about5–7 kPaand wasremained constant
for 36 hrs in the presence of LIF (to maintain undifferentiated
status). Upon addition of RA to induce differentiation, the cellular
stiffness increased in Sag+/+mES cells, starting at 12 hrs and
reaching the peak of 12 kPa at 24 hr. In contrast, the cellular
stiffness inSag2/2mES cells decreased by 50%to 2.5 kPa at 12 hrs
and fell beyond detection level after 18 hrs, likely due to the
initiation of cell death process. The result clearly demonstrated that
physical changes, as reflected by cellular stiffness, occur at the very
early stage during RA-induced differentiation, long before any
morphological changes which take days to occur.
Sag deletion enhanced RA induced apoptosis
We next used several independent assays to determine the
nature of RA-induced cell death in Sag2/2mES cells. Both trypan
blue exclusion assay and TUNEL showed that about 30% of
Sag2/2mES cells were dead or stained TUNEL positively after
36 hrs of RA treatment, compared to ,3% of Sag+/+mES cells
(Fig. 3A, 3B). DNA fragmentation assay, a hallmark of apoptosis,
showed a remarkable induction of DNA ladders in Sag2/2, but
not in Sag+/+mES cells (Fig. 3C). Likewise, caspase-3 activity
assay also showed a significantly higher enzymatic activation in
Sag2/2than in Sag+/+mES cells (Fig. 3D). Taken together, our
data suggest that Sag deletion triggers mES cells to undergo
apoptosis, instead of differentiation in response to RA.
Sensitization of AML cells to RA by MLN4924, a SAG-SCF
E3 ligase inhibitor
RA was used as a differentiation therapy for AML patients and
the disease relapsed upon development of RA resistance [30,31].
We therefore tested our working hypothesis that inactivation of
SAG E3 ligase may sensitize AML cells to RA. We first
Figure 2. Sag deletion blocks RA-induced differentiation. Sag+/+and Sag2/2mES cells were seeded in 6-well plate and treated with RA at
indicated concentrations for 6 days. Colonies were stained with AP (A) and counter-stained with hematoxylin (B), followed by photography. Bar
size=50 nm. Cells were seeded in geletin-coated glass coverslips and cellular stiffness was measured by AFM nanomechanical analysis at 0, 12, 18, 24
and 36 hrs post RA (1 mM) treatment. Shown is x 6 SD from three independent experiments (C).
doi:10.1371/journal.pone.0027726.g002
RA Sensitization by Inactivation of SAG E3 Ligase
PLoS ONE | www.plosone.org3November 2011 | Volume 6 | Issue 11 | e27726

Page 4

determined SAG protein levels in seven AML lines and found
SAG expression varied among these lines (Fig. 4A). We then
determined a potential correlation between SAG levels and RA
sensitivity using two SAG high-expressing lines, HL-60 and KG-1
and one SAG low-expressing line, MV4-11 (Fig. 4A), and found a
good correlation with SAG high-expressing cells being much
more resistant to RA (IC50.50 mM in HL-60 and KG-1 cells vs.
IC50,0.1 mM in MV4-11 cells) (Fig. 4B, left panel). We then
used MLN4924, a small molecule inhibitor of NEDD8-Activating
Enzyme (NAE), that inactivates SAG-SCF E3 ligase by blocking
cullin neddylation [35] for potential RA sensitization. Before the
combination study, we determined the cytotoxicity of MLN4924
as a single agent against three AML lines and found again, SAG
high-expressing lines were more resistant to MLN4924 with IC50
of 0.17 mM (HL-60), 2.32 mM (KG-1) and 0.07 mM (MV4-11),
respectively (Fig. 4B, right panel). In RA sensitization experiment,
we used two concentrations of MLN4924 that caused ,20–30%
cytotoxicity when used alone (0.05–0.1 mM for HL-60, and 0.5–
1 mM for KG-1 cells) in combination with various concentrations
of RA, ranging from 5 nM to 50 mM. We observed a dose-
dependent additive or synergetic effect on growth suppression of
HL-60 cells (Fig. 4C) or KG-1 cells (Fig. 4D). Thus, SAG
inhibitor MLN4924 could sensitize otherwise resistant AML cells
to RA.
RA-MLN4924 combination enhanced apoptotic cell
death
Since Sag deletion sensitized mES cells to RA via apoptosis
(Fig. 3), we reasoned that apoptosis would likely be the underlying
mechanism for RA sensitization by MLN4924. Indeed, the sub-G1
apoptotic population by FACS analysis was significantly increased
by the single treatment of MLN4924, but not RA. MLN4924-RA
combination treatment further enhanced apoptosis in both lines
(Fig. 5A). Likewise, DNA fragmented ladders (Fig. 5B) as well as
cleavages of PARP and caspase-3 (Fig. 3C), the hallmarks of
apoptosis, were more evidently seen in the combination group than
in MLN4924 single treatment group, although RA alone failed to
induce apoptosis. Thus, both HL-60 and KG-1 cells are resistant to
RA, and MLN4924 confers RA sensitization by inducing apoptosis.
RA-MLN4924 combination enhanced accumulation of
pro-apoptotic proteins c-JUN and NOXA
Since SAG is the RING component of SCF E3 ubiquitin ligase
required for its ligase activity [36], we expected that SAG-SCF E3
inactivation by MLN4924 via cullin deneddylation would cause
accumulation of SAG-SCF E3 ligase substrates. We further expected
that MLN4924-RA combination would cause further increase in the
levels of critical substrates, responsible for enhanced apoptosis. To
Figure 3. Sag deletion sensitizes mES cells to RA-induced apoptosis. mES cells with genotype of Sag+/+and Sag2/2were treated with
DMSO control or 1 mM RA for indicated time periods, followed by trypan blue staining (A), TUNEL staining at 36 hrs (B, left panel), with quantification
of TUNEL positive cells graphed (B, right panel), DNA fragmentation assay at 36 hrs (C), and caspase-3 activity assay at the indicated time point (D). *,
p,0.05.
doi:10.1371/journal.pone.0027726.g003
RA Sensitization by Inactivation of SAG E3 Ligase
PLoS ONE | www.plosone.org4November 2011 | Volume 6 | Issue 11 | e27726

Page 5

this end, we first confirmed that MLN4924 treatment indeed caused
NEDD8 removal from cullin-1 (Fig. 6A). We then determined
potential accumulation of a subset of SAG-SCF E3 substrates known
to regulate cell growth, cell cycle progression and apoptosis [37,38],
and found expected accumulation by MLN4924 of CDT1, b-
catenin, WEE1, p27, phosphor-IkBa, c-JUN and NOXA, but not
pro-apoptotic proteins BIM EL and IkBa (Fig 6B, 6C). Among all
these substrates tested, only c-JUN in KG-1 cells and NOXA in HL-
60 cells were further increased upon MLN4924-RA combination
(Fig.6D),indicatingacorrelationbetweenfurtheraccumulationsofc-
JUN/NOXA and enhanced induction of apoptosis. In addition, we
found that although c-Jun was not detectable in mES cells regardless
of RA treatment (data not shown), Noxa, but not another pro-
apoptotic protein Bim EL, known as SCF E3 substrate [39], did
accumulate in Sag2/2, but not Sag+/+mES cells upon RA treatment
(Fig. 6D). These results further suggest an involvement of NOXA, a
newly identified SAG substrate [17], in RA-induced apoptosis.
Discussion
Our recent work showed that during mES cell differentiation
induced by LIF withdrawal, Sag deletion had no effect on the
formation of embryoid bodies, but prevented the endothelial
differentiation to form blood island structure [13]. Here we
determined the effect of Sag deletion on differentiation induced by
RA, a powerful agent that triggers terminal differentiation of mES
to many cell types [40]. We found that Sag deletion prevented
mES cells from undergoing differentiation, rather triggered
apoptosis instead. Although the molecular mechanism that
governs the switch from RA-induced differentiation to apoptosis
in the absence of Sag E3 ligase merits a future investigation,
accumulation of its substrates, such as pro-apoptotic protein Noxa,
is likely involved.
In this study, we also found for the first time that AFM, a valid
tool in measurement of cell mechanics with incremental
application in cell biology [41], can be used to monitor the
process of RA-induced mES cell differentiation at very early stage.
In wild type mES cells, we observed an increased cellular stiffness,
starting at 12 hrs post RA treatment, long before any measurable
change in cell morphology. In contrast, Sag2/2mES cells showed
a decrease in cellular stiffness, followed by its complete loss after
18 hrs, likely at the point when apoptosis is committed. This strike
difference, observed at the very early stage of RA differentiation,
suggests that the Sag-dependent ‘‘decision’’ of differentiation vs.
Figure 4. Correlation between SAG overexpression and RA resistance in AML cell lines and their sensitivity to RA and/or MLN4924.
Proteins were extracted from 7 AML lines and subjected to immunoblotting using antibody against SAG with b-actin as the loading control (A). Cells
were seeded in 96-well plate in triplicate and treated with various concentrations of RA (B, left panel), MLN4924 (B, right panel) or in combination at
indicated concentrations of each drug for 48 hrs (C, HL-60 cells) and (D, KG-1 cells), followed by ATP-lite cell viability assay. Values were normalized to
the untreated control. Shown is x 6 SEM from three independent experiments with IC50 curve generated and IC50 value calculated using the Prism
software.
doi:10.1371/journal.pone.0027726.g004
RA Sensitization by Inactivation of SAG E3 Ligase
PLoS ONE | www.plosone.org5November 2011 | Volume 6 | Issue 11 | e27726

Page 6

apoptosis in response to RA, must occurs rapidly, likely involving
gene transcription (RA effects) and rapid protein turn-over (Sag
effects). The data also suggests that AFM measurement of cellular
stiffness may be used as an early biomarker for cell differentiation.
Applying our finding of RA sensitization by Sag deletion in
mES cells to AML, a human malignancy responsive to RA
differentiation therapy, we observed a direct correlation of SAG
overexpression and RA resistance. To mimic SAG deletion, we
used MLN4924, a newly discovered small molecule inhibitor of
NAE [35] that inactivates SAG-SCF E3 ligase by blocking cullin
neddylation [35,42], and found that combination of RA and
MLN4924 caused significantly more apoptosis than each drug
alone in two RA resistant AML cell lines. Mechanistically, we
correlated enhanced apoptosis to higher level accumulation of c-
JUN in KG-1 cells and of NOXA in HL-60 cells in drug
combination. It is worth noting that MLN4924 was recently
shown to be an effective apoptosis-inducing agent against AML
lines, including MV4-11 and HL-60 [43].
It is well-established that c-JUN forms homo-dimer and hetero-
dimer (with c-FOS) to compose active transcription factor AP-1
[44], and that AP-1 is involved in both induction and suppression
of apoptosis in a cell context dependent manner [45]. It is also
known that c-JUN could form a hetero-dimer with ATF-2
(Activating Transcription Factor-2), a stress responsive bZIP
family of transcript factor with context-dependent proliferative as
well as pro-apoptotic activities [46,47]. Our observation that c-
JUN is further accumulated in RA-MLN4924 treated cells with a
consequent enhancement of apoptosis suggested its pro-apoptotic
role. On the other hand, NOXA is a well-known BH3 domain-
containing pro-apoptotic protein [48], which binds to anti-
apoptotic proteins MCL1 and A1 to release their inhibitory
binding with pro-apoptotic proteins BAX and BAK, thus inducing
apoptosis [49]. Our recent study revealed that NOXA is a
substrate of SAG E3 ligase that mediated apoptosis induced by
SAG knockdown [17]. The fact that NOXA accumulates upon
Sag deletion or inactivation in both mES and HL-60 cells in
response to RA, strongly argue its apoptosis-causing role in these
settings.
In summary, our study showed here that Sag E3 ligase
determines the cell fate of mES cells to undergo differentiation
or apoptosis in response to RA. We also validated that SAG, which
is often overexpressed in human cancers [17,50], is an attractive
target for RA-mediated differentiation therapy in AML. Together
with our previous work, demonstrating that SAG-SCF E3
ubiquitin ligase is an attractive anti-cancer and radiosensitizing
target (for review, see [50,51,52,53]), the small molecule inhibitor
of SAG-SCF E3 ligase, such as MLN4924, which is currently
under several phase I clinical trials [54], may have multiple
applications in the treatment of human cancers.
Materials and Methods
Ethics Statement
No human participants. All animal procedures were approved
by the University of Michigan Committee on Use and Care of
Animals with an approval number 09701. Animal care was
provided in accordance with the principles and procedures
outlined in the National Research Council Guide for the Care
and Use of Laboratory Animals.
Figure 5. MLN4924 induces RA sensitization by inducing apoptosis. Cells were treated with DMSO control, RA (1.5 mM), MLN4924 (0.1 mM
for HL-60, 1.0 mM for KG-1) alone or in combination for 24 hrs (HL-60) or 36 hrs (KG-1), followed by FACS analysis for apoptotic sub-G1 population. *,
p,0.05; **, p,0.01; ***, p,0.0001, compared to control (A). Cells were treated with indicated drugs at concentrations described above for 24 hrs,
followed by DNA fragmentation assay (B) or immunoblotting using indicated antibodies (C).
doi:10.1371/journal.pone.0027726.g005
RA Sensitization by Inactivation of SAG E3 Ligase
PLoS ONE | www.plosone.org6November 2011 | Volume 6 | Issue 11 | e27726

Page 7

Generation and maintenance of Sag embryonic stem cells
Generation of ES cells from blastocysts were performed as
described [55]. Briefly, blastocysts at the embryonic stage of E3.5
were isolated from Sag+/2females mated with Sag+/2males, and
cultured on irradiated MEFs feeder cells in DMEM medium
containing 15% fetal bovine serum, 0.1 mM b-mercaptoethanol,
50 IU/ml penicillin, 50 mg/ml streptomycin, 103u/ml ESGRO,
and 12.5 mM PD98059. Inner cell mass outgrowths were
trypsinized and cultured in 35-mm cell culture dishes in the same
medium, but without antibiotics and PD98059. Embryonic stem
cells were passaged on 0.1% gelatin-coated dishes to eliminate the
feeder cells before genotyping and experimentation.
Cells and cell culture
All AML cells (KG-1, HL-60, SigM5, ML-2, MV4-11, OCI-AML3
and OCI-AML2) were the gift from Dr. S. Malek at the University of
Michigan. All the lines were originally purchased from ATCC and
were cultured, according to ATCC suggested culture conditions.
ATPlite growth assay and IC50determination
Cells were seeded in 96-well plates in triplicates and treated with
all trans-RA (Sigma) and/or MLN4924 (a gift from Millennium) at
various concentrations for 24 hrs (mES) or 44 hrs (AML). Cell
viability was measured with ATPlite kit (Perkin Elmer).
Clonogenic assay
ES cells were seeded in six-well gelatin-coated plates in
duplicate. Cells were treated next day with RA at various
concentrations, followed by incubation at 37uC incubator for 5
to 7 days. Colonies were washed with PBS once and stained with
brilliant blue G-250. The colonies with 50 or more cells were
counted [14].
Alkaline phosphatase (AP) staining
Sag+/+and Sag2/2mES cells were seeded in 6-well gelatin-
coated plates, treated with DMSO vehicle control or RA at
0.1 mM or 1 mM for 6 days. The colonies were stained with
Alkaline Phosphatase kits (Sigma-Aldrich) according to the
manufacture’ protocol. Briefly, after PBS wash, the colonies were
fixed with fixative solution for 30 seconds, rinsed gently in water
for 45 seconds, and then stained with alkaline-dye mixture by
incubation at room temperature for 15 minutes. The colonies
were rinsed, counterstained with hematoxylin, and evaluated
microscopically.
AFM (Atomic Force Microscopy) nanomechanical analysis
All the experiments for AFM measurement were performed in
aqueous environment at room temperature on Bioscope (Bruker
AXS, Santa Barbara, CA) using a Nanoscope IV controller. Once
Figure 6. Induction of SAG-SCF E3 ligase substrates by RA and/or MLN4924. Cells were treated with indicated concentrations of MLN4924
for 24 hrs, followed by immunoblotting using cullin-1 antibody with b-actin as the loading control (A). Cells were treated with RA (1.5 mM), MLN4924
(0.1 mM for HL-60, 1.0 mM for KG-1) alone or in combination for 24 hrs, followed by immunoblotting using indicated antiobodies (B&C). mES cells
were treated with RA at indicated concentrations for 24 hrs, followed by immunoblotting using NOXA antibody with b-actin as the loading control
(D).
doi:10.1371/journal.pone.0027726.g006
RA Sensitization by Inactivation of SAG E3 Ligase
PLoS ONE | www.plosone.org 7November 2011 | Volume 6 | Issue 11 | e27726

Page 8

the mES cells seeded onto geletin-coated glass coverslips grew to
confluency in LIF containing medium, they were observed under
the AFM with an inverted optical microscopy underneath. AFM
nanoindentation was performed on the center of mES cells. At
each time point (0, 12, 18, 24 and 36 hrs) post RA addition, 50
cells were randomly selected and measured to obtain the force-
displacement curves. In the AFM nanoindentation experiment, a
cone-shaped AFM tip with a silicon nitride cantilever (Bruker
AXS, Santa Barbara, CA) was used with the spring constant
0.06 nN/nm. The applied force was calculated by Hooke’s law as
F~kd, where d is the vertical deflection of the cantilever and k is
the spring constant of the cantilever. According to Hertzian
model, the relationship between indentation force and deforma-
tion defines (for a cone-shaped tip): F~2
p
E
1{n2d2tana, where a is
the half angle of the cone-shaped AFM tip, u is the Poisson ratio, d
is the indentation depth and E is the Young’s modulus value. The
half open angle of the tip in our case is 17.5u and we used 0.5 as
Poisson ratio. By fitting the force-displacement curve with this
model, Young’s modulus value can be generated. The Young’s
moduli were then processed statistically to obtain the mean and
standard deviation at each time point.
Trypan blue exclusion assay
Sag+/+or Sag2/2mES cells were plated in 24-well gelatin-
coated plate and incubated overnight at 37uC. Cells were treated
next day with 1 mM RA for 24 or 36 hrs. Cells were detached with
trypsin, stained with trypan blue dye, and counted microscopical-
ly. Ratio of dead blue cells out of total cells counted was plotted.
DNA fragmentation analysis
The mES cells were seeded in 10-cm dishes at 2.56106cells per
dish and treated next day with DMSO (control) or RA (1 mM) for
36 hrs. The AML cells were seeded in 25-cm2flasks at 0.56106
cell/mL (10 mL per flask) and treated for 24 hrs with DMSO
(control), RA (1.5 mM), MLN4924 (0.1 mM for HL-60, 1.0 mM for
KG-1), alone or in combination. Cells were harvested, pelleted,
and lysed in 600 ml of lysis buffer (5 mM Tris-HCl, pH8.0, 20 mM
EDTA, and 0.5% Triton X-100). DNA was isolated by phenol/
chloroform extraction and ethanol precipitation and treated with
RNAse and subjected to electrophoresis in a 1.8% agarose gel
[14].
Caspase activation assay
The activity of caspase3 was measured using a fluorogenic
caspase assay with Ac-DEVA-AFC as the substrate. The results
are expressed as fold activation, as compare to the control [14].
TUNEL staining
Mouse embryonic stem cells were seeded in 12-well gelatin-
coated plates, treated with DMSO vehicle control or 1 mM RA for
36 hrs. Cells were washed with cold PBS 3 times and fixed in 4%
PFA-PBS for subsequent TUNEL staining using in situ Cell Death
Detection Kit (Roche), according to manufacturer’s instruction.
Immunoblotting analysis
AML cells were harvested and lysed in a Triton X-100 lysis
buffer (20 mmol/L Tris-HCl (pH 8.0), 150 mmol/L NaCl, 1%
Triton X-100, 5 mmol/L EGTA, and 5 mmol/L EDTA) with
freshly added protease inhibitor tablet (Roche, Indianapolis, IN)
for 30 min on ice, followed by centrifugation for 20 min.
Supernatants were measured for protein concentration using a
Bio-Rad protein assay reagent (Bio-Rad, Hercules, CA) and
subjected to immunoblotting analysis with the following antibod-
ies: SAG monoclonal antibody (raised against the RING domain)
[17], Caspase-3, CUL1, CDT1, b-Catenin, WEE1, IkBa, C-JUN
(Santa Cruz Biotechnology, CA), b-Actin (Sigma, MO), p27 (BD
PharMingen, CA), Phospho-IkBa, NOXA, PARP, BIM (Cell
Signaling Technology, MA).
FACS (Fluorescence-activated Cell Sorting)
Cells seeded at 0.56106cell/mL in flasks (10 mL per 25-cm2
flask) were treated for 24 hrs (HL-60) or 36 hrs (KG-1) with
DMSO (control), RA (1.5 mM), MLN4924 (0.1 mM for HL-60,
1.0 mM for KG-1), alone or in combination and analyzed by flow
cytometry [17].
Statistical analysis
Paired two-tailed Student’s t test was used in comparisons
between two groups to determine p value for statistical
significance.
Acknowledgments
We thank Millennium Pharmaceuticals, Inc., for providing us MLN4924
and Dr. S. Malek at the University of Michigan for providing us AML cell
lines. We also thank Dr. C. Su from Bruker AXS, Inc., for his technical
advice.
Author Contributions
Conceived and designed the experiments: YS NX. Performed the
experiments: MT YL RY. Analyzed the data: MT YL RY NX YS.
Contributed reagents/materials/analysis tools: MT YL RY NX YS. Wrote
the paper: YS.
References
1. Duan H, Wang Y, Aviram M, Swaroop M, Loo JA, et al. (1999) SAG, a novel
zinc RING finger protein that protects cells from apoptosis induced by redox
agents. Mol Cell Biol 19: 3145–3155.
2. Sun Y, Tan M, Duan H, Swaroop M (2001) SAG/ROC/Rbx/Hrt, a zinc
RING finger gene family: molecular cloning, biochemical properties, and
biological functions. Antioxid Redox Signal 3: 635–650.
3. Sun Y (1999) Alteration of SAG mRNA in human cancer cell lines:
Requirement for the RING finger domain for apoptosis protection. Carcino-
genesis 20: 1899–1903.
4. Yang ES, Park JW (2006) Regulation of nitric oxide-induced apoptosis by
sensitive to apoptosis gene protein. Free Radic Res 40: 279–284.
5. Chanalaris A, Sun Y, Latchman DS, Stephanou A (2003) SAG attenuates
apoptotic cell death caused by simulated ischaemia/reoxygenation in rat
cardiomyocytes. J Mol Cell Cardiol 35: 257–264.
6. Lee SJ, Yang ES, Kim SY, Shin SW, Park JW (2008) Regulation of heat shock-
induced apoptosis by sensitive to apoptosis gene protein. Free Radic Biol Med
45: 167–176.
7. Kim SY, Kim MY, Mo JS, Park JW, Park HS (2007) SAG protects human
neuroblastoma SH-SY5Y cells against 1-methyl-4-phenylpyridinium ion
(MPP+)-induced cytotoxicity via the downregulation of ROS generation and
JNK signaling. Neurosci Lett 413: 132–136.
8. He H, Gu Q, Zheng M, Normolle D, Sun Y (2008) SAG/ROC2/RBX2 E3
ligase promotes UVB-induced skin hyperplasia, but not skin tumors, by
simultaneously targeting c-Jun/AP-1 and p27. Carcinogenesis 29: 858–865.
9. Duan H, Tsvetkov LM, Liu Y, Song Y, Swaroop M, et al. (2001) Promotion of
S-phase entry and cell growth under serum starvation by SAG/ROC2/Rbx2/
Hrt2, an E3 ubiquitin ligase component: association with inhibition of p27
accumulation. Mol Carcinog 30: 37–46.
10. Gu Q, Bowden TG, Normolle D, Sun Y (2007) SAG/ROC2 E3 ligase regulates
skin carcinogenesis by stage dependent targeting of c-Jun/AP1 and IkB/NF-kB.
J Cell Biol 178: 1009–1023.
11. Huang Y, Duan H, Sun Y (2001) Elevated expression of SAG/ROC2/Rbx2/
Hrt2 in human colon carcinomas: SAG does not induce neoplastic transforma-
tion, but its antisense transfection inhibits tumor cell growth. Mol Carcinog 30:
62–70.
12. Tan M, Gallegos JR, Gu Q, Huang Y, Li J, et al. (2006) SAG/ROC-SCFbeta-
TrCP E3 ubiquitin ligase promotes pro-caspase-3 degradation as a mechanism
of apoptosis protection. Neoplasia 8: 1042–1054.
RA Sensitization by Inactivation of SAG E3 Ligase
PLoS ONE | www.plosone.org8November 2011 | Volume 6 | Issue 11 | e27726

Page 9

13. Tan M, Zhao Y, Kim S-J, Liu M, Jia L, et al. (2011) SAG/RBX2/ROC2 E3
ubiquitin ligase is essential for vascular and neural development by targeting NF-
1 for degradation. Dev Cell;in press.
14. Tan M, Zhu Y, Kovacev J, Zhao Y, Pan ZQ, et al. (2010) Disruption of Sag/
Rbx2/Roc2 induces radiosensitization by increasing ROS levels and blocking
NF-kB activation in mouse embryonic stem cells. Free Radic Biol Med 49:
976–983.
15. Gu Q, Tan M, Sun Y (2007) SAG/ROC2/Rbx2 is a novel activator protein-1
target that promotes c-Jun degradation and inhibits 12-O-tetradecanoylphorbol-
13-acetate-induced neoplastic transformation. Cancer Res 67: 3616–3625.
16. Tan M, Gu Q, He H, Pamarthy D, Semenza GL, et al. (2008) SAG/ROC2/
RBX2 is a HIF-1 target gene that promotes HIF-1alpha ubiquitination and
degradation. Oncogene 27: 1404–1411.
17. Jia L, Yang J, Hao X, Zheng M, He H, et al. (2010) Validation of SAG/RBX2/
ROC2 E3 Ubiquitin Ligase as an Anticancer and Radiosensitizing Target. Clin
Cancer Res 16: 814–824.
18. Sasaki H, Yukiue H, Kobayashi Y, Moriyama S, Nakashima Y, et al. (2001)
Expression of the sensitive to apoptosis gene, SAG, as a prognostic marker in
nonsmall cell lung cancer. Int J Cancer 95: 375–377.
19. Dragnev KH, Petty WJ, Dmitrovsky E (2003) Retinoid targets in cancer therapy
and chemoprevention. Cancer Biol Ther 2: S150–156.
20. Okuno M, Kojima S, Matsushima-Nishiwaki R, Tsurumi H, Muto Y, et al.
(2004) Retinoids in cancer chemoprevention. Curr Cancer Drug Targets 4:
285–298.
21. Balmer JE, Blomhoff R (2002) Gene expression regulation by retinoic acid.
J Lipid Res 43: 1773–1808.
22. Theodosiou M, Laudet V, Schubert M (2010) From carrot to clinic: an overview
of the retinoic acid signaling pathway. Cell Mol Life Sci 67: 1423–1445.
23. Gudas LJ (1994) Retinoids and vertebrate development. J Biol Chem 269:
15399–15402.
24. Mark M, Ghyselinck NB, Chambon P (2009) Function of retinoic acid receptors
during embryonic development. Nucl Recept Signal 7: e002.
25. Gudas LJ, Wagner JA (2011) Retinoids regulate stem cell differentiation. J Cell
Physiol 226: 322–330.
26. Jimenez-Lara AM, Clarke N, Altucci L, Gronemeyer H (2004) Retinoic-acid-
induced apoptosis in leukemia cells. Trends Mol Med 10: 508–515.
27. Tang XH, Gudas LJ (2011) Retinoids, retinoic acid receptors, and cancer. Annu
Rev Pathol 6: 345–364.
28. Bushue N, Wan YJ (2010) Retinoid pathway and cancer therapeutics. Adv Drug
Deliv Rev 62: 1285–1298.
29. Altucci L, Gronemeyer H (2001) The promise of retinoids to fight against
cancer. Nat Rev Cancer 1: 181–193.
30. Camacho LH (2003) Clinical applications of retinoids in cancer medicine. J Biol
Regul Homeost Agents 17: 98–114.
31. Petrie K, Zelent A, Waxman S (2009) Differentiation therapy of acute myeloid
leukemia: past, present and future. Curr Opin Hematol 16: 84–91.
32. Altucci L, Wilhelm E, Gronemeyer H (2004) Leukemia: beneficial actions of
retinoids and rexinoids. Int J Biochem Cell Biol 36: 178–182.
33. de The H, Chen Z (2010) Acute promyelocytic leukaemia: novel insights into the
mechanisms of cure. Nat Rev Cancer 10: 775–783.
34. Huang S, Laoukili J, Epping MT, Koster J, Holzel M, et al. (2009) ZNF423 is
critically required for retinoic acid-induced differentiation and is a marker of
neuroblastoma outcome. Cancer Cell 15: 328–340.
35. Soucy TA, Smith PG, Milhollen MA, Berger AJ, Gavin JM, et al. (2009) An
inhibitor of NEDD8-activating enzyme as a new approach to treat cancer.
Nature 458: 732–736.
36. Swaroop M, Wang Y, Miller P, Duan H, Jatkoe T, et al. (2000) Yeast homolog
of human SAG/ROC2/Rbx2/Hrt2 is essential for cell growth, but not for
germination: Chip profiling implicates its role in cell cycle regulation. Oncogene
19: 2855–2866.
37. Nakayama KI, Nakayama K (2006) Ubiquitin ligases: cell-cycle control and
cancer. Nat Rev Cancer 6: 369–381.
38. Skaar JR, D’Angiolella V, Pagan JK, Pagano M (2009) SnapShot: F Box
Proteins II. Cell 137: 1358, 1358 e1351.
39. Dehan E, Bassermann F, Guardavaccaro D, Vasiliver-Shamis G, Cohen M,
et al. (2009) betaTrCP- and Rsk1/2-mediated degradation of BimEL inhibits
apoptosis. Mol Cell 33: 109–116.
40. Soprano DR, Teets BW, Soprano KJ (2007) Role of retinoic acid in the
differentiation of embryonal carcinoma and embryonic stem cells. Vitam Horm
75: 69–95.
41. Kirmizis D, Logothetidis S (2010) Atomic force microscopy probing in the
measurement of cell mechanics. Int J Nanomedicine 5: 137–145.
42. Brownell JE, Sintchak MD, Gavin JM, Liao H, Bruzzese FJ, et al. (2010)
Substrate-assisted inhibition of ubiquitin-like protein-activating enzymes: the
NEDD8 E1 inhibitor MLN4924 forms a NEDD8-AMP mimetic in situ. Mol
Cell 37: 102–111.
43. Swords RT, Kelly KR, Smith PG, Garnsey JJ, Mahalingam D, et al. (2010)
Inhibition of NEDD8-activating enzyme: a novel approach for the treatment of
acute myeloid leukemia. Blood 115: 3796–3800.
44. Angel P, Karin M (1991) The role of Jun, Fos and the AP-1 complex in cell-
proliferation and transformation. Biochim Biophys Acta 1072: 129–157.
45. Shaulian E, Karin M (2002) AP-1 as a regulator of cell life and death. Nat Cell
Biol 4: E131–136.
46. Hayakawa J, Mittal S, Wang Y, Korkmaz KS, Adamson E, et al. (2004)
Identification of promoters bound by c-Jun/ATF2 during rapid large-scale gene
activation following genotoxic stress. Mol Cell 16: 521–535.
47. Vlahopoulos SA, Logotheti S, Mikas D, Giarika A, Gorgoulis V, et al. (2008)
The role of ATF-2 in oncogenesis. Bioessays 30: 314–327.
48. Oda E, Ohki R, Murasawa H, Nemoto J, Shibue T, et al. (2000) Noxa, a BH3-
only member of the Bcl-2 family and candidate mediator of p53-induced
apoptosis. Science 288: 1053–1058.
49. Cory S, Adams JM (2002) The Bcl2 family: regulators of the cellular life-or-
death switch. Nat Rev Cancer 2: 647–656.
50. Wei D, Sun Y (2010) Small RING finger proteins RBX1 and RBX2 of SCF E3
ubiquitin ligases: the role in cancer and as cancer targets. Genes & Cancer 1:
700–707.
51. Jia L, Sun Y (2011) SCF E3 ubiquitin ligases as anticancer targets. Current
Cancer Drug Targets 11: 347–356.
52. Sun Y (2003) Targeting E3 ubiquitin ligases for cancer therapy. Cancer Biol
Therapy 2: 623–629.
53. Sun Y (2006) E3 ubiquitin ligases as cancer targets and biomarkers. Neoplasia 8:
645–654.
54. Soucy TA, Smith PG, Rolfe M (2009) Targeting NEDD8-activated cullin-RING
ligases for the treatment of cancer. Clin Cancer Res 15: 3912–3916.
55. Tan M, Davis SW, Saunders TL, Zhu Y, Sun Y (2009) RBX1/ROC1
disruption results in early embryonic lethality due to proliferation failure,
partially rescued by simultaneous loss of p27. Proc Natl Acad Sci U S A 106:
6203–6208.
RA Sensitization by Inactivation of SAG E3 Ligase
PLoS ONE | www.plosone.org9November 2011 | Volume 6 | Issue 11 | e27726