ORIGINAL PAPER

Isocitrate dehydrogenase 1 mutant R132H sensitizes glioma cellsto BCNU-induced oxidative stress and cell death

Isabelle Vanessa Mohrenz • Patrick Antonietti • Stefan Pusch • David Capper •

Jorg Balss • Sophia Voigt • Susanne Weissert • Alicia Mukrowsky •

Jan Frank • Christian Senft • Volker Seifert • Andreas von Deimling •

Donat Kogel

Published online: 26 June 2013

� Springer Science+Business Media New York 2013

Abstract Isocitrate dehydrogenase 1 (IDH1) decarboxy-

lates isocitrate to a-ketoglutarate (a-KG) leading to gen-

eration of NADPH, which is required to regenerate reduced

glutathione (GSH), the major cellular ROS scavenger.

Mutation of R132 of IDH1 abrogates generation of a-KG

and leads to conversion of a-KG to 2-hydroxyglutarate. We

hypothesized that glioma cells expressing mutant IDH1

have a diminished antioxidative capacity and therefore may

encounter an ensuing loss of cytoprotection under condi-

tions of oxidative stress. Our study was performed with

LN229 cells stably overexpressing IDH1 R132H and wild

type IDH1 or with a lentiviral IDH1 knockdown. Quanti-

fication of GSH under basal conditions and following

treatment with the glutathione reductase inhibitor BCNU

revealed significantly lower GSH levels in IDH1 R132H

expressing cells and IDH1 KD cells compared to their

respective controls. FACS analysis of cell death and ROS

production also demonstrated an increased sensitivity of

IDH1-R132H-expressing cells and IDH1 KD cells to

BCNU, but not to temozolomide. The sensitivity of IDH1-

R132H-expressing cells and IDH1 KD cells to ROS

induction and cell death was further enhanced with the

transaminase inhibitor aminooxyacetic acid and under

glutamine free conditions, indicating that these cells were

more addicted to glutaminolysis. Increased sensitivity to

BCNU-induced ROS production and cell death was con-

firmed in HEK293 cells inducibly expressing the IDH1

mutants R132H, R132C and R132L. Based on these find-

ings we propose that in addition to its established pro-

tumorigenic effects, mutant IDH1 may also limit the

resistance of gliomas to specific death stimuli, therefore

opening new perspectives for therapy.

Keywords Brain tumor � Glutaminolysis �Cytoprotection � Oxidative stress � Glutathione

Introduction

Despite modern therapeutic regimens, patients suffering

from malignant gliomas still have a dismal prognosis.

Histopathological characterization of gliomas allows

classification into different subgroups (grade I–IV) [1, 2].

Standard therapy of glioblastoma multiforme (grade IV),

the most malignant glioma, consists of surgical resection

of the tumor following chemotherapy and radiation [3,

4]. Intrinsic resistance to cell death plays a fundamental

role for the therapy resistance of malignant gliomas [5].

A multitude of different genetic lesions and expression

changes in key components of cell signaling pathways

I. V. Mohrenz � P. Antonietti � S. Voigt � A. Mukrowsky �D. Kogel (&)

Experimental Neurosurgery, Neuroscience Center, Goethe

University Hospital, Theodor-Stern-Kai 7, 60590 Frankfurt,

Germany

e-mail: koegel@em.uni-frankfurt.de

S. Pusch � D. Capper � A. von Deimling

Clinical Cooperation Unit Neuropathology, German Cancer

Research Center (DKFZ), 69120 Heidelberg, Germany

D. Capper � J. Balss � S. Weissert � A. von Deimling

Department of Neuropathology, Institute of Pathology,

Ruprecht-Karls-University Heidelberg, INF 224,

69120 Heidelberg, Germany

J. Frank

Institute of Biological Chemistry and Nutrition, University

of Hohenheim, 70593 Stuttgart, Germany

C. Senft � V. Seifert

Department of Neurosurgery, Goethe University Hospital,

Theodor-Stern-Kai 7, 60590 Frankfurt, Germany

123

Apoptosis (2013) 18:1416–1425

DOI 10.1007/s10495-013-0877-8

regulating cell death/cell survival has been causally

implicated in mediating this resistance acquired during

the course of tumor progression of gliomas [5]. Hetero-

zygous mutations in IDH1 are early events in glioma-

genesis resulting in a loss of original enzyme function

and a gain of function leading to production of 2-hy-

droxyglutarate [6–8]. Somatic mutations of IDH1 are also

found in acute myeloid leukemias, chondrosarcomas as

well as other tumor entities [9, 10]. IDH1 mutations are

associated with a relatively good prognosis in grade III

and IV gliomas [6, 7, 11–13]. Wild type IDH1 catalyzes

conversion of isocitrate to a-ketoglutarate which is

required for the activity of a-ketoglutarate-dependent

dioxygenases such as prolyl hydroxylases regulating the

turnover of hypoxia-inducible transcription factors 1aand 2a (HIF1a/HIF2a), members of the ten-eleven

translocation (TET) family of 5-methylcytosine hydrox-

ylases and the Jumonji-C domain-containing histone de-

methylases (JHDMs) [14–17]. Mutation of R132H of

IDH1 evokes a gain of function effect associated with a

neomorphic enzymatic function of IDH1. It is well

established that mutant IDH1 promotes generation of the

oncometabolite 2-hydroxyglutarate which acts as a

competitive inhibitor of a-ketoglutarate-dependent diox-

ygenases, thereby driving tumorigenesis [8, 16, 17].

Despite this pro-oncogenic function of mutant IDH1, the

consequences of IDH1 mutation may be more complex

because a dominant negative effect of mutant IDH1 has

also been proposed [14]. There is considerable evidence

suggesting that wt IDH1 can act in an anti-apoptotic

manner and represents an important component of the

antioxidative defense machinery of cells [18–21]. Altered

enzyme functions of IDH1 may therefore evoke contrary

effects in glioma cells. On the one hand, they may exert

pro-tumorigenic effects by formation of 2-hydroxygluta-

rate, modulation of a-ketoglutarate-dependent dioxygen-

ases, epigenetic changes and a subsequent shift in gene

expression profiles. On the other hand, there may be a

loss of the cytoprotective, antioxidative function of wt

IDH1 [22]. Loss of the original enzyme function by

IDH1 mutation may therefore drive gliomagenesis, but

may also sensitize gliomas to oxidative stress and cell

death, thereby increasing tumor sensitivity to therapy and

providing a rationale for the overall better survival of

patients with IDH1 mutations.

Here, we studied the consequences of IDH mutations on

glutamine addiction and the cellular sensitivity to ROS-

induced cell death induced by the clinically relevant glu-

tathione reductase inhibitor BCNU. We demonstrate that

IDH1 mutant-expressing glioma cells are more prone to

BCNU-induced oxidative stress and cell death which is

further enhanced by inhibition of the glutaminolysis

pathway.

Materials and methods

Cell culture and lentiviral transduction

To study the cytoprotective function of IDH1, a lentiviral

knockdown was performed in LN229 glioma cells. IDH1

was silenced by transduction-ready shRNA lentiviral

particles (SHCLNV NM_005896, Sigma-Aldrich, Deis-

enhofen, Germany) according to the manufacturers’

instructions. The target sets included five sequences for

different small hairpins. The pLKO.1-puro control trans-

duction particles (SHC001 V) did not contain a hairpin

insert and were used as a negative control. For the

transduction, cells were plated in 96 well-plates and

transduced the following day at a multiplicity of infection

of 10. New medium was added to a final volume of

100 ll containing hexadimethrine bromide (Sigma-

Aldrich) at a final concentration of 8 lg/ml. Cells were

incubated for 24 h before changing the medium. After

overnight incubation, cells were washed, trypsinated, and

transferred to six-well plates, after which puromycin

(Merck Millipore, Darmstadt, Germany) was added at a

final concentration of 5 lg/ml. Generation of IDH1 wt

overexpressing cells and IDH1 R132H overexpressing

cells was done by Gateway cloning as recommended by

the manufacturer (Life Technologies, Darmstadt, Ger-

many). The IDH1 ORF, in a pDONR221 plasmid, was

obtained from the DKFZ Clone Respository and the

IDH1R132H variant was generated by site directed

mutagenesis. These ENTRY vectors were used for

transfer of the cDNA in the destination vector pDEST26

(N-terminal 69 His Tag). LN229 cells were transfected

with pDEST vectors by Fugene 6 (Promega, Madison,

USA) followed by picking of single cell clones.

HEK293T cells were transfected with pcDNA6/TR by

Fugene 6 (Promega). After applying selection pressure

with 8 lg/ml blasticidin (Invitrogen, Darmstadt, Ger-

many), single cell clones were picked. Clones were

transfected with GFP in pTREx-DEST30 for quality

analysis. Clones revealing GFP fluorescence in the

absence of tetracycline induction were discarded, whereas

clones showing a strong signal by tetracycline induction

were processed. These clones were transfected with the

following IDH1 variants: IDH1 R132H, IDH1 R132L,

IDH1 R132C, IDH1wt, Tet repressor in pTREx-DEST31.

Selection for the TetR plasmid was carried out with

8 lg/ml blasticidin and for the expression plasmid with

1 mg/ml geneticin (G-418) (GIBCO, Darmstadt, Germany).

Reagents and antibodies

BCNU, N-acetylcystein, temozolomide and propidium

iodide were purchased by SIGMA. The Annexin-V-FLUOS

Apoptosis (2013) 18:1416–1425 1417

123

staining kit was from Roche (Roche Applied Science,

Mannheim, Germany).

Immunoblots

For Western Blot analysis, cells were lysed with SDS lysis

buffer containing protease and phosphatase inhibitors.

Protein content was quantified with the BC Assay Kit

(Uptima, Montlucon Cedex, France). 50 lg of protein were

applied on a 12 % gel followed by electrophoresis. Proteins

were transferred to nitrocellulose membranes which were

incubated over night at 4 �C with a monoclonal anti-IDH1

R132H antibody [23] or an anti-IDH1 antibody [23]. As

secondary antibodies IRDye 800CW goat anti-mouse or

peroxidase-conjugated anti-Rat IgG (Cell Signaling, Dan-

vers, USA) were employed followed by IR-detection with

an Odyssey Imaging System (LI-COR Biosystems, Bad

Homburg, Germany) or Pierce ECL detection (Thermo

Fisher Scientific, Rockford, USA).

Glutathione assay

For glutathione assays, 6,000 cells were plated on a white

96-well polystyrene microtest plate (Greiner Bio-One,

Frickenhausen, Germany) and treated with BCNU or

hydrogen peroxide in the presence or absence of N-ace-

tylcysteine. For quantification of reduced glutathione, the

GSH-Glo Glutathione Assay (Promega, Madison, USA)

was used according to the manufacturers’ instructions.

Luminescence was quantified with a Centro LB 960

Luminometer (Berthold Technologies, Bad Wildbad,

Germany).

Quantification of 2-hydroxyglutarate

D2-HG concentrations were measured in supernatants with

an enzymatic assay. This assay is based on the enzyme D-2-

hydroxyglutarate dehydrogenase from Acidaminococcus

fermentans (HGDH) which catalyzes the conversion of D2-

HG to alpha ketoglutarate (a-KG) in the presence of nic-

otinamide adenine dinucleotide (NAD?). Determination of

D2-HG concentration is based on the detection of stoi-

chiometrically generated NADH. Diaphorase converts the

non-fluorescent resazurin to the fluorescent resorufin under

consumption of NADH. Fluorometric detection was carried

out on a plate reader with excitation at 540 ± 10 nm and

emission of 610 ± 10 nm.

MTT: cell viability assay

MTT working solution (5 mg/ml) was prepared by dis-

solving 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazo-

lium bromide (Sigma) in PBS following sterile filtration.

One day prior treatment 2,000 cells/well were plated in a

96-well tissue culture plate in a total volume of 100 ll/

well. At least 8 technical replicates were used for each

condition. Following treatment 20 ll of MTT solution

were added (final concentration: 0,83 mg/ml) and samples

were incubated for 3 h at 37 �C, 5 % CO2. After carefully

removing the medium, 2-propanol was added to dissolve

the formazan salt. Absorbance was measured at 560 nm

with a fluorescence plate reader.

Flow cytometry

For flow cytometric detection of cell death, 20,000 cells per

well were seeded in 24-well microtiter plates. Following

respective treatments, cells were washed with PBS, tryp-

sinized and centrifuged. The cell pellet was resuspend in

50 ll HEPES-Buffer (10 mM Hepes/NaOH, pH 7.4,

140 mM NaCl, 5 mM CaCl2) after which propidium iodide

or propidium iodide in combination with Annexin V

(Sigma-Aldrich) were added. After 10 min of incubation,

samples were run on a FACS Canto II (BD Biosciences,

Heidelberg, Germany) followed by analysis with the FACs

Diva Software (BD Biosciences, Heidelberg, Germany).

All cells that were positive for Annexin and/or PI [i.e., cells

from all quadrants except the bottom left one (Q3)] were

considered dead. For FACS measurement of reactive

oxygen species (ROS), hydroethidine (HE) was applied.

For ROS detection, cells were washed in Hanks balanced

salt solution (HBSS), trypsinized and centrifuged. The

pellet was resuspend in 50 ll HBSS and hydroethidine was

added at a final concentration of 10 lM following 30 min

incubation at 37 �C. Samples were run on a FACS Canto II

(BD Biosciences, Heidelberg, Germany) followed by

analysis with the FACs Diva Software (BD Biosciences,

Heidelberg, Germany).

Determination of caspase-3-like protease activity

For measuring effector caspase-activity, treated cells were

lysed in 200 ll lysis buffer [10 mM HEPES, pH 7.4,

42 mM KCl, 5 mM MgCl2, 1 mM Phenylmethylsulfonyl

Fluoride (PMSF), 0.1 mM EDTA, 0.1 mM EGTA, 1 mM

Dithiothreitol (DTT), 1 lg/ml Pepstatin A, 1 lg/ml Leu-

peptin, 5 lg/ml Aprotinin, 0.5 % 3-(3-cholamidopropyldi-

methylammonio)-1-propane sulfonate (CHAPS)]. 50 ll of

this lysate were added to 150 ll reaction buffer (25 mM

HEPES, 1 mM EDTA, 0.1 % CHAPS, 10 % sucrose,

3 mM DTT, pH 7.5). The fluorigenic substrate Ac-DEVD-

AMC was added at a final concentration of 10 lM.

Accumulation of AMC fluorescence was monitored over

2 h using a HTS fluorescent plate reader (excitation

380 nm, emission 465 nm). Protein content was deter-

mined using the Pierce Coomassie Plus Protein Assay

1418 Apoptosis (2013) 18:1416–1425

123

reagent (KMF, Cologne, Germany). The caspase-activity is

expressed as a change in fluorescence units per lg protein

and hour.

Results

To analyze the potential effects of the IDH1 R132H

mutation on the antioxidative and anti-apoptotic function

of IDH1, we established LN229 glioma cells stably

overexpressing plasmid-encoded His-tagged wild type

IDH1 (LN229-IDH1wt, Fig. 1a, left panel) or His-tagged

IDH1 R132H (LN229-R132H, Fig. 1b, right panel). Due

to the addition of the His-tag, His-IDH1 R132H and His-

IDH1 wt migrate slightly slower in Western blots than

endogenous IDH1 (Fig. 1a, left panel) or a transiently

expressed untagged version of IDH1 R132H as detected

with a monoclonal R132H-specific antibody (Fig. 1a,

ctrl, right panel). In addition, we also established LN229

cells with a stable lentiviral knockdown of IDH1

(LN229-IDH1 KD, Fig. 1a, left panel) and empty vector-

transduced control cells. In all subsequent experiments,

G418-resistant IDH1 wt overexpressing cells (LN229-

IDH1wt) were used as a control for G418-resistant IDH1

R132H overexpressing cells (LN229-R132H) and

Puromycin-resistant empty vector-transduced cells were

used as a control for Puromycin-resistant lentiviral IDH1

KD cells.

High levels of reduced glutathione (GSH) correlate with

an increased anti-oxidative capacity. To gain insights into

the potential effects of mutant IDH1 and IDH1 KD on

cellular redox homeostasis, glutathione levels were mea-

sured under basal and stress conditions. To induce oxida-

tive stress, we initially employed H2O2 as a model

substance. In addition, we applied the alkylating agent/

glutathione reductase inhibitor BCNU (Fig. 1b) which is

already clinically used as biodegradable wafers after sur-

gical resection. Interestingly, in comparison to their

respective controls, GSH levels were significantly

decreased in LN229-R132H cells and LN229-IDH1 KD

cells under basal conditions. Treatment with BCNU and

H2O2 decreased GSH levels which again were further

depleted in LN229-R132H cells and LN229-IDH1 KD

cells. This reducing effect of IDH1-R132H and the IDH1

knockdown could be rescued by addition of the ROS

scavenger N-acetylcystein (Fig. 1b).

In subsequent experiments, we studied the effect of

IDHR132H expression/IDH1 knockdown on the sensitivity

to BCNU-induced cell death. In a dose response experi-

ment, 50–250 lM BCNU were applied for 48 h after

B

GS

H(r

elat

ive

to c

on

tro

l)

ctrl BCNU H2O2 BCNU H2O20

20

40

60

80

100

120

0

20

40

60

80

100

120Ø IDH1 KD

**

* *

*

* ***

#

# #

NAC

GS

H(r

elat

ive

to c

on

tro

l)

ctrl BCNU H2O2 BCNU H2O2

IDH1 wt IDH1 R132H

**

**

*

**

*

##

#

NAC

321ctrl

IDH1 R132HØ

IDH1 KDIDH1 wt

GAPDH

1 2

GAPDH

A

His-R132HR132H

His-IDH1 wtIDH1 wt

Fig. 1 Reduced glutathione (GSH) levels in glioma cells with a

stable knockdown of IDH1 or expressing IDH1 mutant R132H.

a Establishment of LN229 glioma cell lines with stable transduction

of empty vector (Ø), stable lentiviral IDH1 knockdown (IDH1 KD) or

overexpressing His-tagged wt IDH1 (IDH1 wt, left panel, clones 1

and 2), and cell lines stably overexpressing equal amounts of His-

tagged IDH1 R132H (IDH1 R132H, right panel, clones 1, 2 and 3).

Transfection control (ctrl): LN229 cells transiently expressing

untagged IDH1 R132H. b Cultures were left untreated (ctrl) or

treated with ROS inducing agents, e.g. the glutathione reductase

inhibitor BCNU (200 lM) for 48 h and 1 mM H2O2 for 2 h after

which cellular GSH contents of LN229 IDH1 KD cells in comparison

to empty vector-transduced LN229 control cells (left panel) or LN229

IDH1 R132H cells in comparison to LN229 IDH1 wt-transfected cells

(right panel) were analyzed by an enzymatic GSH assay. Where

indicated, cultures were pre-treated for 2 h with the antioxidant N-

acetylcysteine with subsequent addition of 200 lM BCNU for 48 h or

1 mM H2O2 for 2 h. Graphs represent means of n = 8 cul-

tures ? SEM.*p \ 0.05, significant difference to untreated controls;#p \ 0.05, significant difference to respective control cell line.

Experiments were repeated at least 2 times with similar results

Apoptosis (2013) 18:1416–1425 1419

123

which cell death was quantified by FACS analysis of

Propidium iodide uptake (Fig. 2a). The obtained data

indicate a dose-dependent increase in cell death in all cell

lines. In line with the findings obtained in the GSH assays,

the sensitivity in response to BCNU was enhanced in

LN229 IDH1 KD cells in comparison to LN229 empty

vector control cells (Ø) (Fig. 2a, left panel) and in LN229

IDH1 R132H cells in comparison to LN229 IDH1 wt cells

(Fig. 2a, right panel) at concentrations of 50, 100 and

200 lM. At 250 lM BCNU-induced cell death reaches a

plateau indicating a cytotoxic concentration at which no

difference between the cell lines can be observed. To

correlate the cellular sensitivity to BCNU-induced cell

death with the extent of oxidative stress, we also mea-

sured ROS levels at two time points (3 and 48 h) by

Hydroethidine (HE) staining by FACS (Fig. 2b). Indeed,

B

A

cell

dea

th [

%]

ctrl TMZ0

20

40

60 ØIDH1 KD

**

cell

dea

th [

%]

ctrl TMZ0

10

20

30 IDH1 wtIDH1 R132H

**

incr

ease

RO

S [

%]

ctrl 3hBCNU 48hBCNU BCNU+NAC0

20

40

60

80 ØIDH1 KD

#

#

**

*

*

incr

ease

RO

S [

%]

ctrl 3hBCNU 48hBCNU BCNU+NAC0

20

40

60

80 IDH1 wtIDH1 R132H

*

* *

*##

n.s.C n.s.

ctrl 24hBCNU

48h TMZ0

10

20

30

40

50

incr

ease

RO

S [

%]

**

*#

BCNU[µM]

cell

dea

th [

%]

ctrl 50 100 200 2500

20

40

60

80

100 ØIDH1 KD

**

*

* *

** *

#

#

#

BCNU[µM]

cell

dea

th [

%]

ctrl 50 100 200 2500

20

40

60

80

100 IDH1wtIDH1 R132H

#

* *

*

***

Fig. 2 IDH1 R132H-expressing cells and IDH1 knockdown cells are

significantly more sensitive to cell death induced with the glutathione

reductase inhibitor BCNU and exhibit higher reactive oxygene

species (ROS) levels. a Cultures were left untreated (ctrl) or treated

with increasing concentrations (50–250 nM) of the glutathione

reductase inhibitor BCNU for 48 h. Total cell death was quantified

by propidium iodide staining followed by flow cytometry. b Cultures

were treated with 500 lM BCNU for 3 h or 200 lM BCNU for 48 h.

Cells were pre-treated for 2 h with the antioxidant N-acetylcysteine

prior to addition of BCNU for 48 h where indicated. Quantitative

measurement of ROS was done by hydroethidine (HE) staining

followed by flow cytometry. (C) Cultures were treated with the

alkylating agent temozolomide (TMZ) at a final concentration of

100 lM for 96 h (left and middle panels). Total cell death was

quantified by propidium iodide staining followed by flow cytometry.

All graphs represent means of n = 12 cultures ?SEM; *, p \ 0.05,

significant difference to untreated controls; #, p \ 0.05, significant

difference to respective control cell line. Experiments were repeated

at least 3 times. c) right panel: empty vector-transduced cultures were

treated with 200 lM BCNU for 24 and 48 h or with 100 lM TMZ for

48 h. Quantitative measurement of ROS was done by hydroethidine

(HE) staining followed by flow cytometry. All graphs represent

means of n = 4 cultures ? SEM; *p \ 0.05, significant difference to

untreated controls; #p \ 0.05, significant difference to TMZ treatment

1420 Apoptosis (2013) 18:1416–1425

123

BCNU-triggered ROS production was further enhanced in

LN229-IDH1 KD cells and LN229-R132H cells at both

time points in comparison to their controls (Fig. 2b). In

contrast to their response to BCNU, LN229-IDH1 KD cells

and LN229-R132H did not display a statistically significant

increase in the sensitivity to the alkylating agent TMZ

which is not an equally potent inducer of oxidative stress in

comparison to BCNU (Fig. 2c). The catalytic reaction

exerted by wt IDH1 is not the only metabolic pathway used

for a-ketoglutarate production, since a-ketoglutarate can

also be synthesized via the glutaminolysis pathway.

Therefore, we assessed whether IDH1 R132H-expressing

cells and IDH1 KD cells may be more dependent on glu-

taminolysis. Firstly, we employed MTT assays (Fig. 3) to

study the cell proliferation which was measured over a

period of 96 h in 24 h intervals starting with time point 0

and could observe that the LN229 IDH1 knockdown cells

and LN229 R132H cells grew significantly slower in full

medium (3.65 mM glutamine, Fig. 3) compared to the

LN229 empty vector control cells and the LN229 IDH1 wt

cells, respectively. As expected, glutamine withdrawal

leads to decreased proliferation in all four cell lines.

Indeed, in LN229 IDH1 KD cells and LN229 R132H cells,

proliferation was almost absent after 96 h whereas the

control cell lines partially retained their ability to prolif-

erate. Next we investigated whether glutamine withdrawal

also would further sensitize LN229 IDH1 KD cells and

LN229 R132H cells to BCNU-induced cell death. To this

end, LN229 cell lines again were cultivated in glutamine

free medium and in full medium (3,65 mM glutamine) and

treated with 200 lM BCNU for 48 h after which cell death

was quantified by FACS analysis of Annexin V/Propidium

iodide staining (Fig. 4a, b). The obtained data revealed that

glutamine deprivation significantly enhanced BCNU-

induced cell death in all four cell lines, but this potentiating

effect was much more dramatic in LN229 IDH1 KD cells

and LN229 R132H cells in comparison to the controls. We

also could also observe a shift of the cell populations from

Annexin-positive/PI negative cells to double positively

labeled cells (Fig. 4c). In order to further scrutinize the

potential therapeutic significance of these observations, we

employed amino oxyacetate (AOA), a transaminase and

glutaminolysis inhibitor (Fig. 5). Combined treatment with

BCNU and AOA again lead to a significant increase in cell

death compared to BCNU alone with 78 and 62 % dead

cells in LN229 IDH1 KD cultures and LN229 R132H

cultures, respectively.

To confirm the results obtained in glioma cells in a

second independent cell model, we employed HEK 293

cells inducibly expressing wt IDH1 and the three IDH1

mutants R132H, R132C, R132L. Figure 6a shows the time-

dependent induction of wt IDH1 (left panel) and IDH1

R132H (right panel) over a time course of 48 h. Interest-

ingly, induced expression of all three IDH1 mutants led to

a very similar potentiation of BCNU-induced cell death

(Fig. 6b) and ROS production (Fig. 6c) versus control

cells. In HEK293 R132H, HEK293 R132C and HEK293

R132L cells, BCNU-induced cell death and ROS levels

were also slightly enhanced in the absence of tetracycline,

suggesting residual expression of the IDH1 mutants under

non-induced conditions (Fig. 6b, c). In addition to quanti-

fication of cell death and ROS levels, we also measured the

amount of 2-HG produced by the different cell lines

(Fig. 6d). In contrast to their almost identical effects on cell

death and ROS induction, expression of the three IDH1

mutants R132H, R132C, R132L was associated with

drastically different amounts of 2-HG production, sug-

gesting that 2-HG may play a negligible role for the

potentiating effects of mutant IDH1 on cell death.

Discussion

In recent years, considerable evidence for a pro-tumori-

genic role of mutant IDH1 and its oncometabolite 2-HG,

2-HG-dependent epigenetic changes and activation of HIFs

time [h]

MT

T a

ctiv

ity

MT

T a

ctiv

ity

0 24 48 72 960.0

0.5

1.0

1.5

2.0Ø + GlnIDH1 KD + GlnØ - GlnIDH1 KD - Gln

**

*

*

time [h]

0 24 48 72 960.0

0.5

1.0

1.5IDH1 wt + GlnIDH1 R132H + GlnIDH1 wt - GlnIDH1 R132H - Gln

**

**

Fig. 3 LN229 IDH1 R132H and IDH1 KD cells are more addicted to

glutaminolysis. LN229 IDH1 KD cells in comparison to empty

vector-transduced LN229 control cells (left panel) or LN229 IDH1

R132H cells in comparison to LN229 IDH1 wt-transfected cells (right

panel) were cultivated in the presence of 3.65 mM glutamine or under

glutamine free conditions prior to determination of MTT activity.

Absorbance was monitored at 560 nm in 24 h intervals for up to 96 h.

Data represent means of n = 6 cultures ? SEM. Experiments were

repeated at least 3 times with similar results.*p \ 0.05, significant

difference to glutamine-containing medium

Apoptosis (2013) 18:1416–1425 1421

123

has been provided [8, 15–17]. In contrast, wild type IDH1

has an established antioxidative and anti-apoptotic role

[18–21] and loss of its original enzyme function may

compromise this function, thereby sensitizing IDH1-

mutated tumor cells to oxidative stress and cell death.

NADPH which is generated during IDH1-catalyzed

decarboxylation of isocitrate is an essential component of

the cellular antioxidative defense which is continuously

required to regenerate reduced glutathione (GSH) by glu-

tathione reductase, especially under conditions of enhanced

oxidative stress. Increased generation of reactive oxygen

species (ROS) and an altered redox status are general

A B

cell

dea

th [

%]

ctrl BCNU ctrl BCNU0

20

40

60

80

100 ØIDH1KD

*

*

#

*

*

#

cell

dea

th [

%]

ctrl BCNU ctrl BCNU0

10

20

30

40

50 IDH1wtIDH1R132H

*

*

#

**

#

Gln Gln GlnGln

C

ctrl BCNU

IDH1 R132H

Annexin V

iod

ide

Pro

pid

ium

+ - + -

Gln

-

1,3%0,2%

93,9% 4,6%

+

0,1% 4,4%

70,5% 25,0%

0,1% 6,0%

76,5% 17,4%

0,3% 14,8%

68,7% 16,2%

IDH1 wt

1,0%0,2%

94,6% 4,3%

Gln

0,2% 1,4%

91,3% 7,1%

0,2% 6,3%

16,5%77,0%

ctrl BCNU2,6%0,2%

81,4% 15,9%

D

ctrl 24hBCNU 48hBCNU STS0

10

20

30

40

50

DE

VD

clea

vag

e[A

.U./h

/µg

pro

tein

] *

Fig. 4 Glutamine withdrawal sensitizes LN229 IDH1 R132H and

IDH1 KD cells to BCNU. Cultures were cultivated in the presence of

3.65 mM glutamine or in glutamine free medium ±200 lM BCNU

for 48 h. Cell death of LN229 IDH1 KD cells in comparison to empty

vector-transduced LN229 control cells a or LN229 IDH1 R132H cells

in comparison to LN229 IDH1 wt-transfected cells b was analyzed by

Annexin V/propidium iodide staining followed by flow cytometry.

Data represent means of n = 12 cultures ? SEM. *p \ 0.05, signif-

icant difference to untreated controls; #p \ 0.05, significant differ-

ence to respective control cell line. Experiments were repeated at least

3 times with similar results. c Representative histograms of ±BCNU-

treated (200 lM BCNU for 48 h) IDH1 R132H cells in comparision

with IDH1 wt control cells are shown in the presence of 3.65 mM

glutamine or in glutamine free medium. d Empty vector-transduced

Cultures were treated with 200 lM BCNU for 24 and 48 h or with

3 lM Staurosporine (STS) for 6 h. Quantitative measurement of

effector caspase activation was done by a caspase-3-like activity

assay. All graphs represent means of n = 4 cultures ? SEM;

*p \ 0.05, significant difference to untreated controls

1422 Apoptosis (2013) 18:1416–1425

123

hallmarks of cancer and the GSH antioxidant system plays

a central role in the adaptation of cancer cells to oxidative

stress [24]. As mutated IDH1 consumes rather than pro-

duces NADPH [8], it may further lower the antioxidative

capacity of IDH1-mutated tumors [22]. In line with this

hypothesis, our study demonstrates profound effects of

mutant IDH1 on the cellular sensitivity to ROS-induced

cell death induced by the clinically relevant glutathione

reductase inhibitor BCNU. In particular, our FACS anal-

yses revealed an increased ROS-generation in IDH1

R132H-expressing cells and IDH1 KD cells after BCNU

treatment and an enhanced sensitivity to cell death induced

by BCNU, but not by TMZ. These data suggest that mutant

IDH1-R132H may particularly sensitize glioma cells to

ROS-inducing cancer drugs, with comparatively minor

effects on cell death sensitivity to TMZ.

Our study also provides support for the notion that

mutant IDH1 propagates the glutamine addiction of glioma

cells [25]. To fulfil the cellular demand for a-ketoglutarate,

cells with mutant IDH1 may shift to the glutaminolysis

pathway for a-ketoglutarate production [25]. Therefore,

tumors with IDH1 mutation likely have an increased

demand for glutamine and this glutamine addiction may

represent an Achilles heel [26] of these tumors. In our

experiments, depletion of glutamine further increased the

sensitivity of IDH1 R132H-expressing cells and IDH1 KD

cells to BCNU, indeed suggesting a switch to the gluta-

minolysis pathway for a-KG generation in these cells. Of

note, a-ketoglutarate production via the glutaminolysis

pathway can occur either via reactions catalyzed by glu-

tamate oxaloacetate transaminase (GOT) or glutamate

dehydrogenase (GLUD) [25]. These alternative metabolic

pathways explain why the sensitizing effects of the GOT

inhibitor AOA were less pronounced in comparison to

glutamine withdrawal. Despite these limitations of AOA, it

was previously shown to exert cytotoxic effects in gluta-

mine-dependent glioma cells [26, 27]. Our data support the

hypothesis that the glutamine-to-glutamate conversion

could be a metabolic bottleneck for IDH-mutated cells

([26, 28] and from a clinical standpoint, glutaminase

inhibitors [25] may represent the most realistic approach to

block a-ketoglutarate production in glutamine-addicted

cancer cells.

To confirm our findings made in glioma cells, we

additionally quantified increased ROS production and

sensitivity to BCNU in HEK293 cells inducibly expressing

three mutants of IDH1. Interestingly, our results obtained

in this cell model demonstrate that increased ROS levels

and cell death clearly were not correlated with the vastly

different levels of 2-HG generated by the different mutants,

but rather very similar for all three mutants (R132H,

R132C, R132L). Of note, metabolic profiling of glioma

cells either expressing IDH1 R132H or treated with 2-HG

had previously revealed 2HG-independent effects on the

levels of cellular metabolites including glutamate and other

metabolites directly or indirectly derived from glutamate,

including GSH [29]. Therefore, abrogation of the original

enzyme function of IDH1 may play a dominant role in the

sensitizing effects of IDH1 mutants to ROS-dependent cell

death.

Collectively, our data support the major hypothesis of

this study, i.e. that mutation of IDH1 may exert opposing

effects on tumor progression and therapy resistance. They

may in part also explain the overall better survival of

patients with IDH1 mutations. Based on our own obser-

vations and on data published elsewhere, mutant IDH1 may

exert pro-tumorigenic effects via generation of 2-HG and

altered gene expression on the one hand. On the other hand,

it may also addict tumors to glutamine metabolism and

limit their resistance to specific death stimuli, therefore

cell

dea

th [

%]

Ø IDH1 KD0

20

40

60

80

100

ctrl BCNU BCNU + AOA

*

*

*

*#

#

cell

dea

th [

%]

IDH1 wt IDH1 R132H0

20

40

60

80

100

ctrl BCNU BCNU + AOA

**

*

*#

#

Fig. 5 Pharmacological inhibition of glutaminolysis enhances sensi-

tivity to BCNU-induced cell death. Cultures were cultivated in

medium containing 3.65 mM glutamine and left untreated (ctrl) or

treated with 200 lM BCNU for 48 h in the presence or absence of

10 mM aminoxyacetic acid (AOA). Cell death of LN229 IDH1 KD

cells in comparison to empty vector-transduced LN229 control cells

(left panel) or LN229 IDH1 R132H cells in comparison to LN229

IDH1 wt-transfected cells (right panel) was analyzed by propidium

iodide staining followed by flow cytometry. Data represent means of

n = 6 cultures ? SEM. Experiments were repeated at least 3 times

with similar results.*p \ 0.05, significant difference to untreated

controls; #p \ 0.05, significant difference to respective control cell

line

Apoptosis (2013) 18:1416–1425 1423

123

A

0.5 1 2 4 6 24

R132H

GAPDH

Tetracycline

48

Tetracycline-

- + +BCNU+ +-

-

IDH1 wt

GAPDH

HEK Tet on IDH1 wild type HEK Tet on IDH1R132H

Tetracycline

4 8 16 24 48time [h] time [h]

TetracyclineBCNU

+ +- + - +

incr

ease

RO

S [

%]

0

10

20

30Tet Rep

IDH1 wtIDH1 R132HIDH1 R132CIDH1 R132L

*** *

**

*

#

--

cell

dea

th [

%]

0

20

40

60Tet RepIDH1 wtIDH1 R132HIDH1 R132CIDH1 R132L

**

* *

*

*

*

*

*

*#

#B

C

D

D2-

HG

[ M

]

- Tet + Tet0

10

20

30Tet repIDH1 wtIDH1 R132HIDH1 R132CIDH1 R132L

*

*

*

#

#

µ

Fig. 6 Enhanced sensitivity to BCNU-induced cell death and ROS

production in HEK293 cells with inducible expression of mutant

IDH1. a Inducible expression of IDH1wt and IDH1 R132H as

analyzed by Western Blot in a time course experiment after induction

with 1 lg/ml tetracycline. b HEK293 cell lines expressing Tet

repressor only (Tet Rep), or inducibly expressing wt IDH1 and the

three IDH1 mutants IDH1 R132H, IDH1 R132C and IDH1 R132L

were left untreated or were treated with 200 lM BCNU for 48 h in

the presence and absence of tetracycline. Cell death was analyzed by

Annexin V/propidium iodide staining followed by flow cytometry.

c HEK293 cell lines were treated with 500 lM BCNU for 3 h.

Quantitative measurement of ROS was done by hydroethidine (HE)

staining followed by flow cytometry. All graphs represent means of

n = 4 cultures ? SEM; *p \ 0.05, significant difference to untreated

controls; #p \ 0.05, significant difference to wt IDH1-expressing

control cell line. Experiments were repeated at least 2 times with

similar results. d Quantification of D-2-HG levels secreted by cells

inducibly expressing IDH1 R132H, IDH1 R132C and IDH1 R132L in

comparison to controls (Tet Rep, IDH1 wt). Cells were left untreated

or treated with tetracycline for 48 h prior to analysis. Shown data

represent means of n = 3 cultures ? SEM; *p \ 0.05, significant

difference to non-induced controls; #p \ 0.05, significant difference

to wt IDH1-expressing control cell line

1424 Apoptosis (2013) 18:1416–1425

123

opening new perspectives for therapeutic intervention, such

as combinations of ROS-inducing cancer drugs with

inhibitors of glutaminolysis. Our study also provides a

potential rationale for the selection of patients for future

trials most likely to benefit from therapy.

Acknowledgments We thank Gabriele Kopf and Hildegard Konig

for excellent technical assistance.

Funding This study was supported by the Deutsche Krebshilfe

(Grant 108795).

Conflict of interest The authors declare no conflict of interest.

References

1. Ohgaki H, Kleihues P (2011) Genetic profile of astrocytic and

oligodendroglial gliomas. Brain Tumor Pathol 28:177–183

2. Ohgaki H, Kleihues P (2007) Genetic pathways to primary and

secondary glioblastoma. Am J Pathol 170:1445–1453

3. Stupp R, Hegi ME, Mason WP, van den Bent MJ, Taphoorn MJ,

Janzer RC et al (2009) Effects of radiotherapy with concomitant

and adjuvant temozolomide versus radiotherapy alone on survival

in glioblastoma in a randomised phase III study: 5-year analysis

of the EORTC-NCIC trial. Lancet Oncol 10:459–466

4. Stupp R, Mason WP, van den Bent MJ, Weller M, Fisher B,

Taphoorn MJ et al (2005) Radiotherapy plus concomitant and

adjuvant temozolomide for glioblastoma. N Engl J Med 352:

987–996

5. Ziegler DS, Kung AL, Kieran MW (2008) Anti-apoptosis

mechanisms in malignant gliomas. J Clin Oncol 26:493–500

6. Parsons DW, Jones S, Zhang X, Lin JC, Leary RJ, Angenendt P

et al (2008) An integrated genomic analysis of human glioblas-

toma multiforme. Science 321:1807–1812

7. Yan H, Parsons DW, Jin G, McLendon R, Rasheed BA, Yuan W

et al (2009) IDH1 and IDH2 mutations in gliomas. N Engl J Med

360:765–773

8. Dang L, White DW, Gross S, Bennett BD, Bittinger MA, Drig-

gers EM et al (2009) Cancer-associated IDH1 mutations produce

2-hydroxyglutarate. Nature 462:739–744

9. Paschka P, Schlenk RF, Gaidzik VI, Habdank M, Kronke J,

Bullinger L et al (2010) IDH1 and IDH2 mutations are frequent

genetic alterations in acute myeloid leukemia and confer adverse

prognosis in cytogenetically normal acute myeloid leukemia with

NPM1 mutation without FLT3 internal tandem duplication. J Clin

Oncol 28:3636–3643

10. Amary MF, Bacsi K, Maggiani F, Damato S, Halai D, Berisha F

et al (2011) IDH1 and IDH2 mutations are frequent events in

central chondrosarcoma and central and periosteal chondromas

but not in other mesenchymal tumours. J Pathol 224:334–343

11. Sanson M, Marie Y, Paris S, Idbaih A, Laffaire J, Ducray F et al

(2009) Isocitrate dehydrogenase 1 codon 132 mutation is an

important prognostic biomarker in gliomas. J Clin Oncol

27:4150–4154

12. van den Bent MJ, Dubbink HJ, Marie Y, Brandes AA, Taphoorn

MJ, Wesseling P et al (2010) IDH1 and IDH2 mutations are

prognostic but not predictive for outcome in anaplastic oligo-

dendroglial tumors: a report of the European Organization for

Research and Treatment of Cancer Brain Tumor Group. Clin

Cancer Res 16:1597–1604

13. Hartmann C, Hentschel B, Wick W, Capper D, Felsberg J, Simon

M et al (2010) Patients with IDH1 wild type anaplastic astrocy-

tomas exhibit worse prognosis than IDH1-mutated glioblastomas,

and IDH1 mutation status accounts for the unfavorable prognostic

effect of higher age: implications for classification of gliomas.

Acta Neuropathol 120:707–718

14. Zhao S, Lin Y, Xu W, Jiang W, Zha Z, Wang P et al (2009)

Glioma-derived mutations in IDH1 dominantly inhibit IDH1

catalytic activity and induce HIF-1alpha. Science 324:261–265

15. Sasaki M, Knobbe CB, Itsumi M, Elia AJ, Harris IS, Chio II et al

(2012) D-2-hydroxyglutarate produced by mutant IDH1 perturbs

collagen maturation and basement membrane function. Genes

Dev 26:2038–2049

16. Chowdhury R, Yeoh KK, Tian YM, Hillringhaus L, Bagg EA,

Rose NR et al (2011) The oncometabolite 2-hydroxyglutarate

inhibits histone lysine demethylases. EMBO Rep 12:463–469

17. Xu W, Yang H, Liu Y, Yang Y, Wang P, Kim SH et al (2011)

Oncometabolite 2-hydroxyglutarate is a competitive inhibitor of

alpha-ketoglutarate-dependent dioxygenases. Cancer Cell 19:

17–30

18. Jo SH, Lee SH, Chun HS, Lee SM, Koh HJ, Lee SE et al (2002)

Cellular defense against UVB-induced phototoxicity by cytosolic

NADP(?)-dependent isocitrate dehydrogenase. Biochem Bio-

phys Res Commun 292:542–549

19. Kim SY, Lee SM, Tak JK, Choi KS, Kwon TK, Park JW (2007)

Regulation of singlet oxygen-induced apoptosis by cytosolic

NADP?-dependent isocitrate dehydrogenase. Mol Cell Biochem

302:27–34

20. Koh HJ, Lee SM, Son BG, Lee SH, Ryoo ZY, Chang KT et al

(2004) Cytosolic NADP?-dependent isocitrate dehydrogenase plays

a key role in lipid metabolism. J Biol Chem 279:39968–39974

21. Lee SM, Koh HJ, Park DC, Song BJ, Huh TL, Park JW (2002)

Cytosolic NADP(?)-dependent isocitrate dehydrogenase status

modulates oxidative damage to cells. Free Radic Biol Med

32:1185–1196

22. Bleeker FE, Atai NA, Lamba S, Jonker A, Rijkeboer D, Bosch

KS et al (2010) The prognostic IDH1(R132) mutation is associ-

ated with reduced NADP?-dependent IDH activity in glioblas-

toma. Acta Neuropathol 119:487–494

23. Capper D, Zentgraf H, Balss J, Hartmann C, von Deimling A

(2009) Monoclonal antibody specific for IDH1 R132H mutation.

Acta Neuropathol 118:599–601

24. 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

25. Seltzer MJ, Bennett BD, Joshi AD, Gao P, Thomas AG, Ferraris DV

et al (2010) Inhibition of glutaminase preferentially slows growth of

glioma cells with mutant IDH1. Cancer Res 70:8981–8987

26. Wise DR, Thompson CB (2010) Glutamine addiction: a new

therapeutic target in cancer. Trends Biochem Sci 35:427–433

27. Wise DR, DeBerardinis RJ, Mancuso A, Sayed N, Zhang XY,

Pfeiffer HK et al (2008) Myc regulates a transcriptional program

that stimulates mitochondrial glutaminolysis and leads to gluta-

mine addiction. Proc Natl Acad Sci USA 105:18782–18787

28. Reitman ZJ, Yan H (2010) Isocitrate dehydrogenase 1 and 2

mutations in cancer: alterations at a crossroads of cellular

metabolism. J Natl Cancer Inst 102:932–941

29. Reitman ZJ, Jin G, Karoly ED, Spasojevic I, Yang J, Kinzler KW

et al (2011) Profiling the effects of isocitrate dehydrogenase 1 and

2 mutations on the cellular metabolome. Proc Natl Acad Sci USA

108:3270–3275

Apoptosis (2013) 18:1416–1425 1425

123