Page 1
©2007 L
ANDES B
IOSCIENC
E. DO NO
T DISTRIB
UTE.
Research Paper
Proteasome Inhibition Specifically Sensitizes Leukemic Cells
to Anthracyclin-Induced Apoptosis through the Accumulation of Bim
and Bax Pro-Apoptotic Proteins
Arnaud Pigneux1-4,*
François-Xavier Mahon2,3,4
François Moreau-Gaudry3,4
Maialene Uhalde2
Hubert de Verneuil3,4
Francis Lacombe2
Josy Reiffers5
Noel Milpied1
Vincent Praloran2
Francis Belloc2,3,4
1CHU Bordeaux; Hôpital Haut-Lévêque; Service d’Hématologie; 2Service des
Maladies du Sang; Pessac, France
3INSERM; E876; Bordeaux, France
4Univ V Segalen; Bordeaux, France
5Institut Bergonié; Bordeaux, France
*Correspondence to: Arnaud Pigneux; Service des Maladies du Sang; Hôpital Haut-
Lévêque; Avenue de Magellan; Pessac 33604 France; Tel.:+ 33.5.57.65.65.11;
Fax: +33.5.57.65.65.14; Email: arnaud.pigneux@chu-bordeaux.fr
Original manuscript submitted: 08/21/06
Manuscript accepted: 01/19/07
Previously published online as a Cancer Biology & Therapy E-publication:
http://www.landesbioscience.com/journals/cc/abstract.php?id=3890
Key woRds
AML, proteasome, Idarubicin, apoptosis,
BIM
AcKNowLedGeMeNts
This work was supported by grants from
Ligue Nationale contre le Cancer, Comité
Aquitaine-Charentes. We thank Pr. Didier
Bouscary for reading and criticizing the
manuscript and Janssen Laboratories for
providing Bortezomib.
ABstRAct
Proteasome inhibitors are a novel class of compounds that might increase sensitivity to
chemotherapy for acute myeloid leukemia (AML). We quantified apoptosis in THP‑1 cells
incubated with idarubicin (IDA) alone or together with a low concentration of MG132 or
bortezomib. The combination of both drugs yielded a percentage of apoptotic cells that
was significantly higher than the additive effect of both drugs administered separately
(p < 0.01). Isobologram analysis showed that both MG132 and bortezomib interacted
synergistically with IDA to induce apoptosis of THP1 cells. Western blot analysis of Bax
and Bim show an acumulation of these pro‑apoptotic proteins in THP1 treated cells.
This increase in Bim preceded the induction of apoptosis and participated in idarubi‑
cin‑induced apoptosis. Proteasome inhibition also potentiated IDA‑induced apoptosis
in primary blast cells from 22 AML patients while no such effect was found on normal
lymphocytes, PHA‑stimulated lymphocytes, normal cord blood CD34+ cells or bone
marrow normal myeloid cells. These data show that MG132 and bortezomib specifically
sensitize leukemic cells to IDA through an increase in BIM and Bax pro‑apoptotic Bcl‑2
family proteins.
INtRodUctIoN
Although the development of better chemotherapy regimens has improved remission
induction and overall survival in acute myeloid leukemia (AML), relapse remains a
common problem especially among older patients and/or patients with poor prognostic
cytogenetics.1,2 Therefore, new anticancer drugs are needed. The ubiquitin-proteasome
system represents a major mechanism by which unwanted cellular proteins are degraded
and cellular3 homeostasis is maintained.4 The 26S proteasome and more specifically its
20S catalytic core proteolytically cleave a wide variety of proteins targeted for degradation
through ubiquitination by multiple ubiquitin ligases.5 The intracellular availability of
proteins involved in diverse cellular processes, including cell cycle control, signal trans-
duction, differentiation, and survival, among numerous others, can thus be regulated.
The proteasome has consequently become an attractive target for therapeutic intervention
in cancer chemotherapy.6 Proteasome inhibitors are novel antitumor agents with
preclinical evidence of activity against hematological malignancies and solid tumors.7,8
Several molecular targets of proteasome inhibitors have been proposed. Inhibition of
the proteasome abrogates degradation of IkB inducing its cytoplasmic accumulation, thus
blocking the nuclear translocation and transcriptional activity of NF-kB and, in turn, the
growth and survival pathways regulated by NFkB.9,10 Initial studies of primary AML cells
have demonstrated that treatment with the proteasome inhibitor MG132 causes rapid
inhibition of NFkB and strongly induces apoptosis.11
Proteasome inhibitors could also cause the accumulation of pro-apoptotic Bcl-2 family
proteins such as Bax, Bim and Bid, thus overcoming Bcl-2 protection or directly inducing
apoptosis.12-16 By this mechanism, proteasome inhibitors induce the release of cytochrome
c from the mitochondria. The apoptosome recruits and activates caspase-9, which in turn
activates caspases 3 and 7 responsible for the biochemical and morphological changes
associated with apoptosis.17-21
Results of recent clinical trials indicate that proteasome inhibitors (Bortezomib,
Velcade®) are highly active in patients with multiple myeloma, including those with disease
refractory to more conventional agents.22 However, clinical results for AML with the same
compound are less encouraging23 and combining drugs might be of interest. A primary
in vitro report24 showed that leukemic stem cells are extremely sensitive to a combination
[Cancer Biology & Therapy 6:4, 603-611; April 2007]; ©2007 Landes Bioscience
www.landesbioscience.com Cancer Biology & Therapy 603
Page 2
©2007 L
ANDES B
IOSCIENC
E. DO NO
T DISTRIB
UTE.
Proteasome Inhibitors and Anthracycline in AML
of proteasome inhibitor (MG132) and anthra-
cyclin treatment and thus undergo rapid and
extensive apoptosis. Anthracyclin treatment
was shown to increase cellular dependency on
NFkB activity and further sensitize cells to the
loss of NFkB activity induced by proteasome
inhibitors. Other reports on several tumor
models showed that proteasome-inhibitors
can increase the sensitivity to chemotherapy,
including agents with antileukemia activity25-27
or other agents affecting protein
degradation such as 17-(allylamino)-
demethoxygeldanamycin.28
In this study, we investigated the apoptotic
effects of proteasome inhibitors currently used in
laboratory (MG132) or in clinical (bortezomib)
settings, alone and in association with idaru-
bicin, on THP-1, a monocytic leukemia cell
line.29 The study was furthermore performed
using normal blood cells and primary blast
cells from AML patients. Proteasome inhibi-
tors and IDA were found to exert a synergistic
cytotoxic activity on leukemic cells in vitro,
thereby confirming their potential interest in
AML treatment.
MAteRIALs ANd MetHods
Drugs. The proteasome inhibitor MG132
(carbobenzoxyl-L-Leucyl-L-Leucyl-L-Leucinal)
was purchased from Calbiochem (Merck,
VWR International SAS, France) and was
stored aliquoted at -20˚C as a 3 mM solution
in DMSO. Bortezomib was kindly provided
by Janssen-Cilag and was stored aliquoted at -20˚C as a 1mg/ml
solution. IDA (Zavedos®) was from Pharmacia (Guyancourt, France)
and was stored aliquoted at -20˚C as a 1 mg/ml aqueous solution.
Cells and cell culture. THP-1 cells were obtained from the
ECACC (N˚: 88081201). They were derived from the blood of a
one-year-old boy with acute monocytic leukemia. MG132-resistant
THP-1 (THP-1-RMG) was obtained by culturing THP-1 cells with
incrementally increasing concentrations of MG132. The resistant cell
line was maintained in 0.45 mM MG132. All the experiments with
cell lines were carried out using cells in exponential growth phase.
Mononuclear cells (MNC) were obtained from EDTA-anti-coagu-
lated peripheral blood or bone marrow from 22 adults with AML
(Table 1) and from normal donors by Ficoll centrifugation.
CD34+ cells were isolated from umbilical cord blood by MACS
(Myltenyi Biotec, France). They were cultured in the presence of a
cocktail of cytokines (G-CSF, MGDF, SCF at 100ng/ml and IL-3 at
0.5 ng/ml).
For some experiments, MNC were stimulated for 72H with
5 mg/ml of phytohemagglutinin (Sigma-Aldrich, St Quentin Fallavier,
France).
THP-1 cells and MNC were grown in RPMI 1640 medium
(Gibco-BRL, Eragny, France) supplemented with 10% fetal serum
(FCS) and 1 mM L-glutamine, 10 mM Hepes (Gibco-BRL), 100U/
ml penicillin, and 50 mg/ml streptomycin in a humidified 95%
O2 and 5% CO2 atmosphere at 37˚C. All apoptosis studies were
performed on leukemic cell line cultured at 3 x 105/ml or on MNC
cultured at 106/ml by incubated for 18 hours in culture medium
without drug (control) and with various concentrations of IDA,
MG132 and bortezomib or with combination of IDA and protea-
some inhibitors.
Analysis of apoptosis by flow cytometry. Blast cells and lympho-
cytes were identified by flow cytometry (FCM) on the basis of their
CD45 expression and scatter properties.30 FITC-annexin V (Beck
man-Coulter-Immunotech) was used as specified by the manufac-
turer to measure the exposure of phosphotidylserine on the external
membrane. Cells (3 x 105) were centrifuged and resuspended in 100
ml of culture medium containing 5 ml of PC5-anti-CD45 antibody
(Beckman-Coulter, Margency, France) and incubated for 10 min at
room temperature. Afterwards, 500 ml 1X binding buffer containing
5 ml FITC-annexin V were added to the cell suspension. The samples
were incubated on ice for 10 min before analysis at 525 nm with
an Epics XL cytometer (Beckman-Coulter). The FITC-annexin V
stained cells were considered as apoptotic and the negative cells as
viable after analysis of 104 blast cells.31 The cell lines did not require
CD45 staining but were otherwise analyzed in the same way for
annexin-V binding. Propidium iodide was added in the latter experi-
ments (cell lines) to differentiate apoptotic from secondary necrotic
cells. For normal bone marrow myeloid cells analysis, the MNC
were labeled with anti-CD13-PC5 and anti-CD15-PC5 before
annexin V-FITC labeling. The analysis of apoptosis was gated on
CD13/CD15-positive cells. CD45-PC5, annexin V-FITC, and PI
fluorescences were measured through 675 nm, 525 nm and 620 BP
Table 1 Details of patients examined in this study
Patient Age sex FAB Blast, % Karyotype MG132,% IdA,% MG132 +
Apoptosis Apoptosis IdA,%
Apoptosis
1 50 F RAEB‑t 10 Complex 21 0 30
2 60 F M0 66 Normal 5 5 20
3 84 F M2 66 ND 52 4 64
4 50 M M1 80 Normal 11 10 31
9 69 M M4 48 Inv 16 76 79 89
10 43 M M4 80 Normal 30 6 31
11 83 M M2 82 ND 22 17 44
12 38 M M0 80 Triso 21 0 28 49
13 74 M RAEB‑t 15 ND 16 11 40
14 45 F M4 80 Inv 16 4 11 23
15 35 F M4 41 Inv 16 1 5 0
17 22 F M3 83 15/17 56 19 90
18 77 F M1 45 ND 21 7 40
22 62 F M1 85 Complex 1 7 8
5 66 F M2 32 Complex 5 0 27
6 70 F M2 88 ND 6 1 1
7 70 F M2 28 Trisomy 8 3 0 0
8 70 M M1 89e Complex 0 5 0
16 54 M M2 70 complex 0 0 0
19 78 F M1 96 Normal 0 0 0
20 55 M M1 96 Triso 4 43 0 43
21 71 M M2 92 ND 44 0 41
F, female; M, male; FAB, French-American-British classification for AML (M0 to M7); RAEBt, RefractoryAnemia with Excess of Blast in trans-
formation (20–30% medullary blast cells ); ND, not done; Complex ≥ 3 anomalies.
604 Cancer Biology & Therapy 2007; Vol. 6 Issue 4
Page 3
©2007 L
ANDES B
IOSCIENC
E. DO NO
T DISTRIB
UTE.
Proteasome Inhibitors and Anthracycline in AML
filters respectively. The loss of mitochondrial membrane potential
was evaluated by FCM using DiOC6(3) (100 ng/mL) as a probe.
Isobologram Analysis. The combination effect of IDA with
MG132 or bortezomib was analyzed by the isobologram method as
described previously32 at the point inducing 50% apoptosis in 24 h
as assessed by the DiOC6 (3) method. The envelopes of additivity
were defined by three isoeffect lines constructed from the dose-re-
sponse curves of the single agents.
Idarubicin Incorporation Measurement. THP1 cells were incu-
bated for 1 h with increasing concentrations of IDA and analyzed by
Figure 1. Proteasome inhibition increases anthracyclin‑induced apoptosis. (A) THP‑1 cells were incubated with either (IDA) alone (15 ng/ml) or with MG132
(0.25 mM) (MG + IDA) for 18 h. The cells were then stained for annexin V binding (AnV) to label phosphatidylserine exposure and propidium iodide (PI) to
label membrane permeabilization. The samples were then analyzed by flow cytometry and bivariate analysis was performed showing an increase in annexin
V positive cells in the MG132 treated sample. Sensitive‑ (B) and MG132 resistant‑ (C) THP‑1 cells were untreated (C) or incubated with either (IDA) alone
(15 ng/ml), MG132 alone (0.25 mM) or both (MG+IDA) for 24 h. The cells were then stained with annexin V‑FITC and propidium iodide and analyzed as
in A. The percentage of apoptotic (AnV+), was scored. (D) HL60, U937 or THP‑1 cells were treated with 0.25 mM MG132 and apoptosis was determined
at the mitochondrial level using DiOC6(3) probe and flow cytometry analysis. The induced apoptosis (%) was plotted as a function of treatment duration.
(E) HL60, U937 or THP‑1 cells were treated with 0.25 mM MG132, 15 ng/ml IDA or both for 20 h and apoptosis was determined as in (D). The induced
apoptosis was calculated as indicated in material and methods and the mean ± SD of three independent determinations was shown for each experiment.
www.landesbioscience.com Cancer Biology & Therapy 605
Page 4
©2007 L
ANDES B
IOSCIENC
E. DO NO
T DISTRIB
UTE.
Proteasome Inhibitors and Anthracycline in AML
flow cytometry for the fluorescence of the cytoplasmic drug through
a 575 BP filter using an Epics XL cytometer (Beckman-Coulter).
Western blot analysis. Whole cell lysates were prepared and
proteins were separated by SDS-PAGE in 10 or 12% acrylamide gel.
After electro-transfer to nitrocellulose filters, blots were incubated
with blocking solution (5% low fat dry milk in TBS-Tween) for 1 h
at room temperature followed by overnight incubation with primary
antibodies at a 1/500 to 1/1000 dilution. The antibodies used were
monoclonal mouse anti-a-tubulin (clone B-5-1-2, Sigma-Aldrich,
France), anti- phospho-IkB-a (Ser 32/36) (Alexis, COGER, Paris,
Figure 2. Proteasome inhibition interacts synergistically with IDA to induce apoptosis of leukemic cells. (A) THP‑1 cells were incubated with increasing
concentrations of IDA in the presence (MG) or not (Control) of 0.25 mM of MG132. After 18 h treatment, the cells were stained with annexin V‑FITC and
analyzed as in Figure 1. The percentage of induced apoptosis was plotted as a function of the concentration. Mean ± SD of three independent experiments.
(B) Isobologram analysis of the combined effect of IDA with MG132 (left panel) or Bortezomib (right panel). The envelopes of additivity are represented by
the three lines. The concentrations of each drug alone that produced 50% of apoptosis as measured by the DiOC6(3) method are expressed as 1 on each
axis. The plotted data points show the relative values of concentrations producing 50% apoptosis when THP1 cells were treated with the drug combinations.
(C) After 1 h incubation a sample was taken and the orange fluorescence of IDA was measured by flow cytometry as indicated in Materials and Methods.
The mean fluorescence intensity was plotted as a function of concentration.
606 Cancer Biology & Therapy 2007; Vol. 6 Issue 4
Page 5
©2007 L
ANDES B
IOSCIENC
E. DO NO
T DISTRIB
UTE.
Proteasome Inhibitors and Anthracycline in AML
France), IkB-a (Imgenex, Clinisciences, Montrouge, France), rabbit
polyclonal anti-Bim (Sigma-Aldrich, Sr Quentin Fallavier, France),
anti-Bax (Santa Cruz, Tebu, France) and rabbit monoclonal anti-
caspase 3 (Cell Signaling Technology, Ozyme, Saint Quentin, France).
The blots were revealed with peroxidase conjugated secondary
antibody (Fab’)2 (Jackson Immuno Research Laboratories, Inc.,
Interchim, Montlucon, France) and a chemoluminescence kit
(Western Lightning, PerkinElmer Life Science, Courtaboeuf, France)
and quantified using a Kodak Image Station 440CF (PerkinElmer
Life Science, Courtaboeuf, France). The IkB phosphorylation rate
and the amount of Bim, Bax and Bcl-2 were calculated from these
quantification values.
siRNA experiments. Downregulation of Bim expression was
achieved through lentiviral expression of siRNAs. THP-1 cells were
transduced with both lentiviral vectors (C-siRNA and Bim-siRNA)
at a multiplicity of infection of 2 in RPMI 10% FCS medium with
8mg/mL protamine sulfate (Sigma, St Louis, MI). Transduction
efficiency was checked by testing for EGFP expression by flow
cytometry. After one week of culturing, the GFP-expressing cells
were sorted by flow cytometry using an Elite-EPICS cell sorter
(Beckman-Coulter). At the time of the experiments, more than 95%
of the cells were fluorescent.
Statistical analysis. The paired Student’s t-test was used to analyze
the data. Owing to the large range in spontaneous apoptosis during
the culture of patient samples, the drug-induced apoptosis was
calculated as the percentage of drug-specific apoptosis.33
(drug-induced apoptosis - apoptosis in control) X100
(100 - apoptosis in control)
ResULts
Proteasome inhibition enhance anthracyclin‑induced apop‑
tosis on THP1 leukemia cell‑line. As shown in Figure 1A, a low
concentration of IDA induced a low rate of apoptosis in THP-1
cells as assessed by annexin V binding. When the cells were
co-treated with MG132 (0.25 mM) to inhibit the proteasome and
IDA (15 ng/ml), the rate of apoptotic (annexin V positive) cells was
increased. PI-labeled cells were late apoptotic cells with a permea-
bilized membrane that could be assumed as necrosis secondary to
apoptosis. Their percentage was also increased. When apoptotic cells
were quantified (Fig. 1B), both MG132 and IDA induced a low but
significant rate of apoptotic cells (6 ± 1 and 11 ± 3% respectively).
Co-treatment with MG132 and IDA showed an additive effect
between MG132 and IDA (27 ± 3%). The percentage of apoptosis
induced by co-treatment was higher than the sum of the effects of
the drugs applied separately (p < 0.02). This suggests that protea-
some inhibition by MG132 increased the IDA-induced recruitment
of cells into the apoptotic compartment. When the experiments
were performed on MG132 resistant THP-1, only a slight effect of
IDA was observed without any increase in the presence of MG132
(Fig. 1C). This confirmed that the proteasome inhibition powered
the effect of IDA through a cell-dependent mechanism. Similar
experiments were performed on two other myeloid cell lines (HL60
and U937) which presented different sensitivities to 0.25 mM
MG132 than THP1 (Fig. 1D). The apoptosis was measured at the
mitochondrial membrane level and the potentiation of IDA effect by
MG132 was verified on THP-1 (Fig. 1E). A high cooperation was
observed in HL60 which was more sensitive than THP-1 to MG132
while in U937, which was poorly sensitive to MG132, an additive
effect was only induced (Fig. 1E). These results strongly suggest that
the cooperation between IDA and MG132 to induce apoptosis of
myeloid leukemic cells occur upstream the mitochondrial step of
apoptosis and is related to the MG132 efficiency.
It was thus interesting to investigate whether proteasome inhibi-
tion killed THP-1 cells by itself or whether it sensitized the cells
to the effect of IDA. As shown in Figure 2A, at the concentration
we used (0.25 mM), proteasome inhibition by MG132 induced
only a low amount of apoptotic cells above spontaneous apoptosis
Figure 3. Proteasome inhibition increases the intracellular concentration in
pro‑apoptotic proteins of the Bcl‑2 family. THP‑1 cells were non‑treated (C) or
incubated with either (IDA) alone (15 ng/ml), MG132 alone (0.25 mM) or
both (MG+IDA) for 18 h. The proteins were then extracted and submitted to
western blot analysis and probed with anti‑IkB, anti‑phospho‑IkB, anti‑Bax,
anti‑Bim or anti‑Bcl‑2 antibodies as indicated. The bands from western blot
analysis were quantified using a Kodak Imager and the ratios Bax/Bcl‑2,
Bim‑EL/Bcl‑2 and IkB/Bcl‑2 were calculated for each treatment and normal‑
ized to the non‑treated control (unit value). The ratios were plotted as a
function of treatment for MG132 (MG) and/or IDA (B) and for Bortezomib
(Bo) and/or IDA (C). The means + SD deviation from three to seven determi‑
nations are shown. *: p<0.05 by the paired t test.
www.landesbioscience.com Cancer Biology & Therapy 607
End of preview.
