Reductive glutamine metabolism by IDH1 mediates lipogenesis under hypoxia

Department of Chemical Engineering, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139, USA.
Nature (Impact Factor: 41.46). 11/2011; 481(7381):380-4. DOI: 10.1038/nature10602
Source: PubMed
ABSTRACT
Acetyl coenzyme A (AcCoA) is the central biosynthetic precursor for fatty-acid synthesis and protein acetylation. In the conventional view of mammalian cell metabolism, AcCoA is primarily generated from glucose-derived pyruvate through the citrate shuttle and ATP citrate lyase in the cytosol. However, proliferating cells that exhibit aerobic glycolysis and those exposed to hypoxia convert glucose to lactate at near-stoichiometric levels, directing glucose carbon away from the tricarboxylic acid cycle and fatty-acid synthesis. Although glutamine is consumed at levels exceeding that required for nitrogen biosynthesis, the regulation and use of glutamine metabolism in hypoxic cells is not well understood. Here we show that human cells use reductive metabolism of α-ketoglutarate to synthesize AcCoA for lipid synthesis. This isocitrate dehydrogenase-1 (IDH1)-dependent pathway is active in most cell lines under normal culture conditions, but cells grown under hypoxia rely almost exclusively on the reductive carboxylation of glutamine-derived α-ketoglutarate for de novo lipogenesis. Furthermore, renal cell lines deficient in the von Hippel-Lindau tumour suppressor protein preferentially use reductive glutamine metabolism for lipid biosynthesis even at normal oxygen levels. These results identify a critical role for oxygen in regulating carbon use to produce AcCoA and support lipid synthesis in mammalian cells.

Full-text

Available from: Paulo A Gameiro
LETTER
doi:10.1038/nature10602
Reductive glutamine metabolism by IDH1 mediates
lipogenesis under hypoxia
Christian M. Metallo
1
{, Paulo A. Gameiro
1,2,3,4
, Eric L. Bell
5
, Katherine R. Mattaini
5,6
, Juanjuan Yang
3,4
, Karsten Hiller
1
{,
Christopher M. Jewell
6
, Zachary R. Johnson
6
, Darrell J. Irvine
6,7
, Leonard Guarente
5
, Joanne K. Kelleher
1
, Matthew G. Vander
Heiden
5,6,8
, Othon Iliopoulos
3,4
& Gregory Stephanopoulos
1
Acetyl coenzyme A (AcCoA) is the central biosynthetic precursor
for fatty-acid synthesis and protein acetylation. In the conventional
view of mammalian cell metabolism, AcCoA is primarily generated
from glucose-derived pyruvate through the citrate shuttle and ATP
citrate lyase in the cytosol
1–3
. However, proliferating cells that
exhibit aerobic glycolysis and those exposed to hypoxia convert
glucose to lactate at near-stoichiometric levels, directing glucose
carbon away from the tricarboxylic acid cycle and fatty-acid
synthesis
4
. Although glutamine is consumed at levels exceeding
that required for nitrogen biosynthesis
5
, the regulation and use of
glutamine metabolism in hypoxic cells is not well understood.
Here we show that human cells use reductive metabolism of
a-ketoglutarate to synthesize AcCoA for lipid synthesis. This
isocitrate dehydrogenase-1 (IDH1)-dependent pathway is active in
most cell lines under normal culture conditions, but cells grown
under hypoxia rely almost exclusively on the reductive carboxyla-
tion of glutamine-derived a-ketoglutarate for
de novo
lipogenesis.
Furthermore, renal cell lines deficient in the von Hippel–Lindau
tumour suppressor protein preferentially use reductive glutamine
metabolism for lipid biosynthesis even at normal oxygen levels.
These results identify a critical role for oxygen in regulating carbon
use to produce AcCoA and support lipid synthesis in mammalian
cells.
Althoughhypoxic cellsexhibita shift towardsglycolytic metabolism
4
,
a functional electron transport chain and glutamine-derived carbon
are required for proliferation of most transformed cells
6
. In line with
these studies, we observed increased glutamine consumption when
A549 cells were cultured at approximately 1% oxygen while glutamate
secretion remained unchanged (Supplementary Fig. 1a), indicating
that net glutamine consumption was elevated and suggesting that
glutamine carbon is used for biosynthesis. Notably, glucose consump-
tion and lactate secretion also increased in these experiments. Con-
sistent with this observation, we found that proliferating cells under
both normoxia and hypoxia incorporate glutamine-derived carbon
into lipids (Supplementary Fig. 1b). Glutamine can contribute carbon
to lipogenic AcCoA through two distinct pathways. Cells can oxida-
tively metabolize glutamine-derived a-ketoglutarate (aKG) in the
tricarboxylic acid (TCA) cycle and generate pyruvate from malate
by glutaminolysis
5
. Alternatively, some tissues can reductively carbo-
xylate aKG to generate citrate
7,8
, and recent studies have indicated that
the IDH reaction is highly reversible
9–11
. To determine which pathway
cells use to incorporate glutamine carbon into lipids we used stable
isotopic tracers
12–14
. We first cultured several cancer cell lines with
[1-
13
C]glutamine under normoxia and quantified the isotopic label
present in metabolite pools along this pathway (Fig. 1a, red carbon
atoms). All cells tested with this tracer retained significant label
from [1-
13
C]glutamine in citrate and metabolites downstream of the
irreversible ATP citrate lyase reaction, indicating that the reductive
flux contributes to the cytosolic AcCoA pool (Supplementary Fig. 2).
Additional evidence for activity along this pathway was obtained with a
uniformly
13
C-labelled ([U-
13
C
5
]) glutamine tracer (Supplementary
Fig. 3).
To quantify the specific contributions of oxidative and reductive
glutamine metabolism to fatty-acid synthesis, we cultured cells in the
presence of tracers for several days and performed isotopomer spectral
analysis (ISA; Supplementary Fig. 4)
15
. Use of [5-
13
C]glutamine spe-
cifically allows estimation of the flux of glutamine to lipids through
reductive carboxylation (Fig. 1a, blue carbon atoms). Virtually all
cell lines cultured with this tracer generated labelled fatty acids,
metabolizing glutamine reductively in the TCA cycle to supply 10–
25% of their lipogenic AcCoA (Fig. 1b and Supplementary Fig. 5).
Consistent with these data, we were able to detect 98 6 5 c.p.m. per 10
6
cells in hexane extracts of A549 cells cultured with [5-
14
C]glutamine.
Next we directly compared the contribution of glutamine to fatty acids
through the reductive flux as a fraction of the total, the latterdetermined
by using [U-
13
C
5
]glutamine. In all cell lines tested, including those
derived from lung, mammary, colon and squamous cell carcinoma as
well as melanoma, glioblastoma and leukaemia, [5-
13
C
5
]glutamine
labelled most of the AcCoA derived from glutamine (Fig. 1b),
highlighting the general use of reductive carboxylation as the primary
route through which glutamine, glutamate and aKG carbon are con-
verted to lipids in cultured cells (Fig. 1b and Supplementary Figs 2, 3
and 5). The glutaminolysis pathway is also an important means of
glutamine catabolism and can be characterized by quantifying the
contribution of glutamine carbon to lactate. Consistent with published
reports
5
, glutamine-derived
13
C label was also detected in lactate,
and the amount of
13
C-labelled lactate produced was highest in
glioblastoma-derived cells compared with other cells cultured with
[U-
13
C
5
]glutamine (Supplementary Fig. 6).
Mammalian cells express three IDH enzymes encoded by separate
genes: IDH1 (cytosolic, NADP
1
-dependent), IDH2 (mitochondrial,
NADP
1
-dependent) and the multi-subunit enzyme IDH3 (mitochon-
drial, NAD
1
-dependent). IDH3 is allosterically regulated and is
thought to operate in the oxidative direction. The NADP
1
-dependent
isozymes are capable of catalysing the reductive reaction; however, the
specific enzyme responsible for this flux is not definitively known
16
.As
measurements of metabolite pools in subcellular compartments and
labelling therein cannot yet be reliably obtained, we used RNA inter-
ference to knockdown expression of IDH1 and IDH2 selectively in
A549 cells. Using labelling from [1-
13
C]glutamine as a readout, we
measured a significant and robust decrease in reductive carboxylation
when IDH1 messenger RNA (mRNA) was targeted using short hairpin
RNAs (shRNAs) (Fig. 1c, d). These changes were consistent with results
using [U-
13
C
5
]glutamine (Supplementary Fig. 7) and reproduced
1
Department of Chemical Engineering, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139, USA.
2
Department of Life Sciences, University of Coimbra, 3004-517 Coimbra, Portugal.
3
Massachusetts General Hospital Cancer Center, Boston, Massachusetts 02114, USA.
4
Massachusetts General Hospital Center for Cancer Research, Charlestown, Massachusetts 02129, USA.
5
Department
of Biology, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139, USA.
6
Koch Institute for Cancer Research, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139,
USA.
7
Howard Hughes Medical Institute, Chevy Chase, Maryland 20815, USA.
8
Dana-Farber Cancer Institute, Boston, Massachusetts 02115, USA. {Present addresses: Department of Bioengineering,
University of California at San Diego, La Jolla, California 92093, USA (C.M.M.); Luxembourg Centre for Systems Biomedicine, University of Luxembourg, L-1511, Luxembourg (K.H.).
00 MONTH 2011 | VOL 000 | NATURE | 1
Macmillan Publishers Limited. All rights reserved
©2011
Page 1
using several cell lines (Supplementary Fig. 8). Finally, we used
13
C
metabolic flux analysis
17,18
to quantify intracellular fluxes using
[U-
13
C
5
]glutamine in A549 cells. The fitted data suggested that reductive
IDH flux significantly decreased when IDH1 protein levels were
decreased, and this change was the primary alteration observed in the
network (Fig. 1e; see Supplementary Tables 1–5 for complete results
and metabolic flux analysis model description).
Our results suggest that IDH1 can convert NADPH,aKG and CO
2
to
isocitrate and NADP
1
in the cytosol. Enzymatic analysis using recom-
binant protein indicated that IDH1 is indeed capable of consuming
NADPH and is responsive to physiological levels of CO
2
(Sup-
plementary Fig. 9). Importantly, the proliferation rate of all cell lines
with IDH1 knockdown was impaired (Fig. 1f and Supplementary
Fig. 8), indicating that reductive metabolism of aKG in the cytosol
may be necessary for robust growth. In contrast to our results with
IDH1 shRNAs, we detected no significant change in reductive flux
when targeting IDH2 mRNA in A549, MDA-MB-231 and HCT116
cells (Supplementary Fig. 10). Although IDH2 may promote reductive
carboxylation in some tissues or conditions, these results are consistent
with the proposed role of IDH2 as an oxidative TCA cycle enzyme
19
.
Intriguingly, we detected a significant increase in reductive carbox-
ylation activity when culturing cells under hypoxia (Supplementary
Fig. 11). As glucose is usually the primary carbon source for mam-
malian tissues
1–3
, we next compared the contributions of reductive
glutamine metabolism and glucose oxidation to fatty-acid synthesis
by culturing cells with either [5-
13
C]glutamine or uniformly labelled
[U-
13
C
6
]glucose under normal tissue culture conditions or hypoxia.
Strikingly, oxygen levels influenced fatty-acid labelling from both
tracers (Fig. 2a, b). Cellspreferentially used glucose carbon for palmitate
synthesis under normoxic conditions; however, fatty acids produced
under hypoxia were primarily synthesized from glutamine carbon
through the reductive pathway. In fact, the reductive carboxylation of
glutamine-derived aKG accounted for approximately 80% of the carbon
used for de novo lipogenesis in A549 cells growing under hypoxia
(Fig. 2c). Conversely, we detected a concomitant decrease in the contri-
bution of [U-
13
C
6
]glucose to fatty-acid synthesis under this condition.
Significant increases in the relative use of this pathway were observed in
all cell lines tested, including non-transformed cells (Supplementary
Fig. 12). In addition, T lymphocytes freshly isolated from a mouse
spleen preferentially used reductive glutamine metabolism over glucose
oxidation for fatty-acid synthesis when activated under hypoxia (Sup-
plementary Fig. 13). Although proliferation rates and relative de novo
lipogenesis were lower under hypoxia (Supplementary Fig. 14a, b), the
net flux of reductive glutamine metabolism to palmitate synthesis was
significantly increased in hypoxic cultures (Fig. 2d). Knockdown of
IDH1 protein mitigated the use of reductive glutamine metabolism
for lipogenesis underhypoxia (Supplementary Fig. 14c). Although most
human cells require glutamine for nucleotide and hexosamine bio-
synthesis, some cell lines can grow in the absence of exogenous sources
of glutamine
20
. Remarkably, we found that hypoxia increases the
dependence of such cells on glutamine, as evidenced by decreased pro-
liferation in the absence of glutamine and increased reductiveglutamine
metabolism under hypoxia when glutamine is present (Fig. 2e and
Supplementary Fig. 15).
To gain insight into the mechanisms controlling this switch to
reductive glutamine metabolism, we analysed changes in the labelling
a
0
5
10
15
20
25
30
A549 H460 MDA231 SKMel5 HCT116 A431
Contribution to lipogenic AcCoA (%)
b
0
5
10
15
20
Cit Asp Mal Fum Suc
Control
IDH1a
IDH1b
*
*
**
**
*
*
Cit
αKG
Suc
Fum
Mal
Oac
Pyr
Glu
Gln
Cit
CO
2
CO
2
CO
2
AcCoA
Oac
AcCoA
Palmitate
Mal
Glc
CO
2
Fum
Asp
1 5
CO
2
CO
2
Reductive
metabolism
Oxidative
metabolism
ACL
PDH
ME
e
0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
Flux (fmol per cell per hour)
Control
IDH1a
IDH1b
IDH (oxidative) IDH (reductive)
*
*
*
0
50
100
150
200
0 24 48 72 96 120
Relative cell number
Time (h)
Control
IDH1a
IDH1b
f
Control
IDH1a
IDH1b
IDH1
Tub
cd
Reductive glutamine metabolism
M1 label from [1
-13
C]glutamine (%)
[5-
13
C]glutamine (reductive)
[U-
13
C
5
]glutamine (all paths)
*
Figure 1
|
Reductive carboxylation is the primary route of glutamine to
lipids. a, Schematic of carbon atom (circles) transitions and tracers used to
detect reductive glutamine metabolism. Isotopic label from [1-
13
C]glutamine
(red) is lost during oxidative conversion to succinate (Suc) but retained on
citrate (Cit), oxaloacetate (Oac), aspartate (Asp), malate (Mal) and fumarate
(Fum) in the reductive pathway (green arrows). [5-
13
C]glutamine (blue)
transfers label to AcCoA through reductive metabolism only. Molecular
symmetry is shown for oxidative metabolism. Glc, glucose; Pyr, pyruvate; ME,
malic enzyme; PDH, pyruvate dehydrogenase; Glu, glutamate; Gln, glutamine.
b, Contribution of [5-
13
C]glutamine and [U-
13
C
5
]glutamine to lipogenic
AcCoA in cell lines. c, IDH1 levels in A549 cells expressing IDH1-specific
(IDH1a and IDH1b) or control shRNAs. d, Metabolite labelling from
[1-
13
C]glutamine from cells in c. e, IDH flux estimates from
13
C metabolic flux
analysis model in control or IDH1-knockdown A549 cells cultured with
[U-
13
C
5
]glutamine. f, Cell proliferation of A549 cells expressing control or
IDH1-shRNAs. Error bars, 95% confidence intervals (b, e) and s.e.m. (n 5 3)
(d, f). *P , 0.05.
RESEARCH LETTER
2 | NATURE | VOL 000 | 00 MONTH 2011
Macmillan Publishers Limited. All rights reserved
©2011
Page 2
and abundances of TCA cycle metabolites in cells grown under
hypoxia. Using [U-
13
C
6
]glucose, we observed a significant decrease
in relative flux through the pyruvate dehydrogenase (PDH) complex
(Fig. 3a and Supplementary Fig. 16a, b). In addition, the citrate pool
became depleted, which would be expected to increase reductive
carboxylation flux through mass action (Fig. 3b and Supplementary
Fig. 16c). On the other hand, we detected increased amounts of
isotopic label in TCA cycle intermediates when using labelled
glutamine tracers under hypoxia (Fig. 3c and Supplementary Fig. 16d,
e). Reductively metabolized glutamine accounted for as much as 40–
70% of the intracellular citrate, aspartate, malate and fumarate pools
when cells were cultured in low oxygen (Fig. 3c). Given the marked
reduction in PDH flux observed in hypoxia, we tested the ability of
dichloroacetate (DCA) to restore PDH activity and mitigate the
contribution of reductive metabolism to lipogenesis. DCA inhibits
pyruvate dehydrogenase kinases (PDKs)
21
, and PDK1 is a known
target of HIF-1a that inhibits the activity of PDH through phosphor-
ylation
22,23
. Although DCA treatment had no observable effect on
carbon use under normoxia, reductive glutamine metabolism was
inhibited and glucose oxidation was partly restored in A549 cells
cultured with DCA under hypoxia (Fig. 3d and Supplementary Fig. 16f),
suggesting that hypoxia-induced PDK1 expression contributes to the
use of reductive carboxylation for fatty-acid synthesis.
The von Hippel–Lindau (VHL) tumour suppressor protein is fre-
quently lost in renal cell carcinoma (RCC) and results in a state of
‘pseudohypoxia’ by activating HIF signalling
24–26
. To understand further
the role of this pathway in promoting the switch to reductive TCA
metabolism we tested RCC cells deficient in VHL using ISA.
Remarkably, VHL-deficient RCC cell lines preferentially used reductive
glutamine metabolism for lipogenesis, even when cultured under
normal oxygen levels, whereas those expressing wild-type (WT)
VHL behaved similarly to other carcinoma cell lines (Fig. 4a and
Supplementary Fig. 17). Re-expression of WT VHL in previously
VHL-deficient cell lines resulted in a switch back to oxidative glucose
metabolism as the source of carbon for lipid synthesis (Fig. 4b),
reduced extracellular fluxes of glucose, lactate and glutamine (Sup-
plementary Fig. 18a), and increased the pool of intracellular citrate rela-
tive to aKG (Supplementary Fig. 18b) under normoxia. Furthermore,
shRNA-mediated knockdown of HIF-2a partly restored glucose-
mediated lipogenesis in 786-O cells (Fig. 4c, d). Consistent with
VHLand HIF influencing the switch to reductiveglutamine metabolism
during hypoxia, glucose entryinto the TCA cycle by PDH was increased
under normoxia upon introduction of WT VHL or knockdown of HIF-
2a in 786-O cells (Supplementary Fig. 18c). Similar changes were
observed after knockdown of the HIF-a dimerization partner ARNT
(aryl hydrocarbon receptor nuclear translocator) in VHL-deficient
normoxic UMRC2 cells, which express both HIF-1a and HIF-2a
(ref. 27), or after ARNT knockdown in hypoxic A549 and 143B cells
(Supplementary Fig. 19).
Our results highlight an important role for reductive TCA metabolism
of glutamine in cell proliferation at physiological oxygen levels
0
20
40
60
80
100
Normoxia Hypoxia
Glucose oxidation
Glutamine reduction
c
0
5
10
15
65
70
75
80
M0 M1 M2 M3 M4 M5 M6 M7 M8
Palmitate MID from [5-
13
C]glutamine (%)
a
Contribution to lipogenic AcCoA (%)
0
10
20
30
40
45
75
80
Palmitate MID from [U-
13
C
6
]glucose (%)
Number of isotopes per molecule
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
Glucose
oxidation
Glutamine
reduction
Palmitate per hour 10
6
cells (pmol)
Normoxia
Hypoxia
0.0
0.2
0.4
0.6
0.8
1.0
1.2
Normoxia Hypoxia
Cell count (millions)
Gln+ Gln–
e
*
*
*
**
**
***
*
b
d
Normoxia
Hypoxia
Normoxia
Hypoxia
M2M0 M4 M6 M8 M10 M12 M14 M16
Number of isotopes
per molecule
Figure 2
|
Hypoxia reprograms cells to rely on reductive glutamine
metabolism for lipid synthesis. a, b, Labelling of palmitate extracts from A549
cells cultured under normoxia or hypoxia with [5-
13
C]glutamine (a)or
[U-
13
C
6
]glucose (b). Similar results were observed in myristate, oleate and
stearate pools (not shown). c, Relative contribution of glucose oxidation
([U-
13
C
6
]glucose) or glutamine reduction ([5-
13
C]glutamine) to lipogenic
AcCoA in A549 cells under normoxia and hypoxia. d, Absolute fluxes of
[U-
13
C
6
]glucose and [5-
13
C]glutamine to palmitate in A549 cells. Error bars,
95% confidence intervals from model (c, d); *P , 0.05. e, Huh7 cell
proliferation after 4 days in the presence or absence of glutamine. Error bars,
s.e.m. (n 5 3) (a, b, e). **P , 0.005 comparing glutamine-free cultures.
***P , 0.001 comparing normoxia and hypoxia.
0
10
20
30
40
50
60
70
80
Cit Asp Mal Fum Suc
Reductive glutamine metabolism
M1 label from [1-
13
C]glutamine (%)
Normoxia
Hypoxia
c
b
a
0
10
20
30
40
50
60
Cit aKG Suc Fum Asp
Glucose oxidation
M2 label from [U-
13
C
6
]glucose (%)
Normoxia
Hypoxia
0
20
40
60
80
100
120
Citrate aKG
Relative metabolite abundance (%)
Normoxia
Hypoxia
0
20
40
60
80
100
Control DCA+ Control DCA+
Contribution to lipogenic AcCoA (%)
Normoxia H
y
poxia
Glucose oxidation
Glutamine reduction
**
**
**
**
**
**
**
***
***
**
**
**
*
d
Figure 3
|
Reductive TCA metabolism increases under hypoxia. MRC5 cells
were cultured under normoxia or hypoxia for 3 days in the presence of tracer.
a, Relative level of glucose oxidation as determined by M2 labelling from
[U-
13
C
6
]glucose (see Supplementary Fig. 16a for atom transition map). M2
isotopologues were the most abundant labelled metabolites in mass spectra.
b, Relative abundance of citrate and aKG. c, Relative contribution of reductive
glutamine metabolism to TCA metabolites, determined by M1 labelling from
[1-
13
C]glutamine. d, Contribution of glucose oxidation and glutamine
reduction to lipogenesis in A549 cells cultured with or without 5 mM DCA.
Error bars, s.e.m. (ac)(n 5 3) and 95% confidence intervals (d). *P , 0.01 and
**P , 0.001 comparing normoxia to hypoxia. ***P , 0.05 comparing control
with DCA in hypoxia.
LETTER RESEARCH
00 MONTH 2011 | VOL 000 | NATURE | 3
Macmillan Publishers Limited. All rights reserved
©2011
Page 3
(Fig. 4e). Given the almost exclusive use of reductive carboxylation for
lipogenesis under hypoxia, a redundant or contributing role of mito-
chondrial IDH2 in this pathway is probable; however, our data provide
evidence that the reductive pathway involves IDH1-mediated catalysis
in the cytoplasm. Although the carbon source that cells use for
lipid synthesis appears to be determined, at least partly, by HIF-
mediated regulation of PDK1, additional hypoxia-associated changes
may also promote reductive glutamine metabolism. For example,
HIF-2a enhances c-MYC activity
28
, which in turn drives glutamine
catabolism through the regulation of numerous genes including
glutaminase
29
. This metabolic reprogramming provides an effective,
glucose-independent means of generating AcCoA for biosynthesis.
Because glucose is also delivered to cells through the vasculature, it may
be limited in microenvironments with decreased oxygen availability
30
.
Reductively metabolizing amino acids for lipid synthesis under these
conditions would allow cells to conserve glucose for production of
ribose and other biosynthetic precursors (for example, one carbon pool,
hexosamines) that are not typically generated from other nutrients.
Thus, reductive metabolism may allow cells to distribute available
nutrients more efficiently in poorly vascularized microenvironments.
These results add a new dimension to our understanding of cell
metabolism and suggest potential therapeutic targets along the reduc-
tive carboxylation and glutamine catabolic pathways that could
mitigate hypoxic tumour growth.
METHODS SUMMARY
For determination of steady-state labelling of polar metabolites, cells were cultured
for approximately 24 h in the presence of
13
C-labelled glutamine or glucose before
extraction. For experiments involving stable isotopic labelling of lipid biomass,
cells were grown for approximately 3–5 days in the presence of tracer before
extraction. Details of the extraction and derivatization methods are described in
the Supplementary Methods. Computational determination of metabolic fluxes,
confidence intervals, de novo lipogenesis and the contribution of tracers to fatty-
acid carbon was accomplished using an in-house software package, Metran
18
.
Details of the metabolic networks and gas chromatography/mass spectrometry
(GC/MS) measurements used for modelling and complete results are described in
Supplementary Information. The generation of cells stably expressing control
shRNAs or those targeted IDH1 or IDH2 is described in Supplementary
Methods; all experiments were conducted within four passages of initial selection.
Hypoxic microenvironments were generated by feeding incubators with a pre-
mixed gas composed of 1% O
2
,5%CO
2
and 94% N
2
, and O
2
levels were confirmed
to range between 1 and 3% using a Fyrite combustion analyser. For details of
recombinant IDH1 production and enzyme assays, T-cell activation, medium
analysis, [5-
14
C]glutamine experiment and western blotting, see Supplementary
Methods.
Full Methods and any associated references are available in the online version of
the paper at www.nature.com/nature.
Received 17 March; accepted 28 September 2011.
Published online 20 November 2011.
1. Hatzivassiliou, G. et al. ATP citrate lyase inhibition can suppress tumor cell growth.
Cancer Cell 8, 311–321 (2005).
2. Wellen, K. E. et al. ATP-citrate lyase links cellular metabolism to histone acetylation.
Science 324, 1076–1080 (2009).
3. Migita, T. et al. ATP citrate lyase: activation and therapeutic implications in non-
small cell lung cancer. Cancer Res. 68, 8547–8554 (2008).
4. Semenza, G. L. HIF-1: upstream and downstream of cancer metabolism. Curr.
Opin. Genet. Dev. 20, 51–56 (2010).
5. DeBerardinis, R. J. et al. Beyond aerobic glycolysis:transformedcells can engage in
glutamine metabolism that exceeds the requirement for protein and nucleotide
synthesis. Proc. Natl Acad. Sci. USA 104, 19345–19350 (2007).
6. Weinberg, F. et al. Mitochondrial metabolism and ROS generation are essential for
Kras-mediated tumorigenicity. Proc. Natl Acad. Sci. USA 107, 8788–8793 (2010).
7. Des Rosiers,C. et al. Isotopomer analysis of citricacid cycle and gluconeogenesis in
rat liver. Reversibility of isocitrate dehydrogenase and involvement of ATP-citrate
lyase in gluconeogenesis. J. Biol. Chem. 270, 10027–10036 (1995).
8. Yoo, H., Antoniewicz, M. R., Stephanopoulos, G. & Kelleher, J. K. Quantifying
reductive carboxylation flux of glutamine to lipid in a brown adipocyte cell line.
J. Biol. Chem. 283, 20621–20627 (2008).
9. Metallo, C. M., Walther, J. L. & Stephanopoulos, G. Evaluation of
13
C isotopic tracers
for metabolic flux analysis in mammalian cells. J. Biotechnol. 144, 167–174
(2009).
10. Lemons, J. M. et al. Quiescent fibroblasts exhibit high metabolic activity. PLoS Biol.
8, e1000514 (2010).
11. Ward, P. S. et al. The common feature of leukemia-associated IDH1 and IDH2
mutations is a neomorphic enzyme activity converting a-ketoglutarate to
2-hydroxyglutarate. Cancer Cell 17, 225–234 (2010).
12. Boros, L. G. et al. Defective RNA ribose synthesis in fibroblasts from patients with
thiamine-responsive megaloblastic anemia (TRMA). Blood 102, 3556–3561
(2003).
Cit
αKG
Suc
Oac
AcCoA
AcCoA
Oac
Fum
Cit
αKG
CO
2
ACO1
IDH1
Pyr
Glucose
Lactate
Mal
Lipid synthesis
PDK1
HIF/ARNT
PDH
DCA
CO
2
CO
2
CO
2
Hypoxia
VHL loss
Glutamine
Glu
ACL
Cytosol
Mitochondrion
Glu
Glutamine
Pyr
GLS
IDH3
IDH2
Nitrogen
metabolism
ACO2
Fum
Asp
Mal
e
0
20
40
60
80
100
786-O
UMRC2
ACHN
SN12C
Contribution to lipogenic
AcCoA under normoxia (%)
(VHL+)(VHL–)
Glucose oxidation
Glutamine reduction
d
c
a
0
20
40
60
80
100
PRC3
(VHL–)
WT8
(VHL+)
Contribution to
lipogenic AcCoA (%)
Glucose oxidation
Glutamine reduction
0
20
40
60
80
100
pTV
Control
pTR
HIF2a kd
Contribution to
lipogenic AcCoA (%)
Glucose oxidation
Glutamine reduction
HIF2α
Actin
WT8
PRC3
pTV
pTR
*
*
*
*
*
*
*
*
b
Figure 4
|
HIF/ARNT/VHL signalling regulate carbon use for lipogenesis.
ac, Contribution of glucose oxidation ([U-
13
C
6
]glucose) and glutamine
reduction ([5-
13
C]glutamine) to lipogenesis in RCC lines (a), parental control
(PRC3) and VHL1 (WT8) cells derived from 786-O line (b) or vector control
(pTV) or HIF2a shRNA (pTR) cells derived from 786-O line (c). d, Western
blot to determine HIF-2a levels for cells in b, c. Error bars, 95% confidence
intervals obtained from ISA model. *P , 0.05. e, Model depicting the metabolic
reprograming of mammalian cells by hypoxia or VHL loss to use reductive
glutamine metabolism for lipogenesis. HIF stabilization drives transcription of
PDK1, which decreases PDH activity and subsequently intracellular citrate
levels. IDH1 and ACO1 reductively generate lipogenic citrate from glutamine-
derived aKG. DCA can inhibit PDKs, forcing increased glucose oxidation in
hypoxic cells.
RESEARCH LETTER
4 | NATURE | VOL 000 | 00 MONTH 2011
Macmillan Publishers Limited. All rights reserved
©2011
Page 4
13. Munger, J. et al. Systems-level metabolic flux profiling identifies fattyacid synthesis
as a target for antiviral therapy. Nature Biotechnol. 26, 1179–1186 (2008).
14. Maier, K., Hofmann, U., Reuss, M. & Mauch, K. Identification of metabolic fluxes in
hepatic cells from transient
13
C-labeling experiments: part II. Flux estimation.
Biotechnol. Bioeng. 100, 355–370 (2008).
15. Kharroubi, A. T., Masterson, T. M., Aldaghlas, T. A., Kennedy, K. A. & Kelleher, J. K.
Isotopomer spectral analysis of triglyceride fatty acid synthesis in 3T3–L1 cells.
Am. J. Physiol. 263, E667–E675 (1992).
16. Siebert, G., Carsiotis, M. & Plaut, G. W. The enzymatic properties of isocitric
dehydrogenase. J. Biol. Chem. 226, 977–991 (1957).
17. Sauer, U. Metabolic networks in motion:
13
C-based flux analysis. Mol. Syst. Biol. 2,
62 (2006).
18. Young, J. D., Walther, J. L., Antoniewicz, M. R., Yoo, H. & Stephanopoulos, G. An
elementary metabolite unit (EMU) based method of isotopically nonstationary flux
analysis. Biotechnol. Bioeng. 99, 686–699 (2008).
19. Hartong, D. T. et al. Insights from retinitis pigmentosa into the roles of isocitrate
dehydrogenases in the Krebs cycle. Nature Genet. 40, 1230–1234 (2008).
20. Cheng, T. et al. Pyruvate carboxylase is required for glutamine-independent
growth of tumor cells. Proc. Natl Acad. Sci. USA 108, 8674–8679 (2011).
21. Bonnet, S. et al. A mitochondria-K
1
channel axis is suppressed in cancer and its
normalization promotes apoptosis and inhibits cancer growth. Cancer Cell 11,
37–51 (2007).
22. Kim, J. W., Tchernyshyov, I., Semenza, G. L. & Dang, C. V. HIF-1-mediated
expression of pyruvate dehydrogenase kinase: a metabolic switch required for
cellular adaptation to hypoxia. Cell Metab. 3, 177–185 (2006).
23. Papandreou, I., Cairns, R. A., Fontana, L., Lim, A. L. & Denko, N. C. HIF-1 mediates
adaptation to hypoxia by actively downregulating mitochondrial oxygen
consumption. Cell Metab. 3, 187–197 (2006).
24. Kaelin, W. G. Jr. Molecular basis of the VHL hereditary cancer syndrome. Nature
Rev. Cancer 2, 673–682 (2002).
25. Ivan, M. et al. HIFa targetedfor VHL-mediated destruction by proline hydroxylation:
implications for O
2
sensing. Science 292, 464–468 (2001).
26. Jaakkola, P. et al. Targeting of HIF-a to the von Hippel-Lindau ubiquitylation
complex by O
2
-regulated prolyl hydroxylation. Science 292, 468–472 (2001).
27. Zimmer, M. et al. Small-molecule inhibitors of HIF-2a translation link its 59UTR
iron-responsive element to oxygen sensing. Mol. Cell 32, 838–848 (2008).
28. Gordan, J. D., Bertout, J. A., Hu, C. J., Diehl, J. A. & Simon, M. C. HIF-2a promotes
hypoxic cell proliferation by enhancing c-myc transcriptional activity. Cancer Cell
11, 335–347 (2007).
29. Gao, P. et al. c-Myc suppression of miR-23a/b enhances mitochondrial
glutaminase expression and glutamine metabolism. Nature 458, 762–765 (2009).
30. Sutherland, R. M. Cell and environment interactions in tumor microregions: the
multicell spheroid model. Science 240, 177–184 (1988).
Supplementary Information is linked to the online version of the paper at
www.nature.com/nature.
Acknowledgements We thank N. Vokes and P. Ward for discussions. We also thank
S. Gross and Agios Pharmaceuticals for providing the IDH1 construct. We acknowledge
support from National Institutes of Health grant R01 DK075850-01. C.M.M. is
supported by a postdoctoral fellowship from the American Cancer Society. K.H. is
supported by the German Research Foundation (DFG) grant HI1400. L.G. is supported
by the NIH and the Glenn Foundation for Medical Research. M.G.V.H. is supported by
the Burrough’s Wellcome Fund, the Smith Family, the Damon Runyon Cancer
Research Foundation and the National Cancer Institute. D.J.I. is an investigator of the
Howard Hughes Medical Institute. O.I. is supported by R01 CA122591 and the Dana
Farber/Harvard Cancer Center Kidney SPORE Grant Developmental Award.
Author Contributions C.M.M., P.A.G., E.L.B., K.R.M., J.Y., K.H. and C.M.J. performed
cellular experiments and isotope tracing. C.M.M. and P.A.G. performed metabolite
profiling and analysed data. K.R.M. and M.G.V.H. performed enzyme assays and
14
C
experiments. E.L.B., J.Y. and Z.R.J. generated western blots. D.J.I. and L.G. provided
support and reagents. J.K.K., M.G.V.H., O.I. and G.S. provided conceptual advice. C.M.M.,
J.K.K., M.G.V.H., O.I. and G.S. wrote and edited the paper.
Author Information Reprints and permissions information is available at
www.nature.com/reprints. The authors declare no competing financial interests.
Readers are welcome to comment on the online version of this article at
www.nature.com/nature. Correspondence and requests for materials should be
addressed to G.S. (gregstep@mit.edu) or O.I. (iliopoul@helix.mgh.harvard.edu).
LETTER RESEARCH
00 MONTH 2011 | VOL 000 | NATURE | 5
Macmillan Publishers Limited. All rights reserved
©2011
Page 5
METHODS
Cell culture, isotopic labelling and hypoxia. All cell lines were cultured in
Dulbecco’s modified Eagle medium (DMEM; Mediatech) containing 10% fetal
bovine serum (FBS; Invitrogen) and 100 U ml
21
penicillin/streptomycin
(Mediatech) unless otherwise mentioned. Cell lines were obtained from ATCC
unless otherwise noted. The VC3 glioma cell line was provided by
T. Lautenschlaeger, MDA-MB-231 and HCT116 cell lines were provided by
F. Chiaradonn a, MRC5 cells were provided by S. Lippard, SN12C, ACHN and
786-O cells were provided by K. Courtney and L. Cantley, MCF10A cells were
provided by J. Brugge and Huh7 cells were provided by M. Hemann. MCF10A
cells were cultured in custom DMEM/F12 (Hyc lone) containing 5% horse
serum, 20 ng ml
21
EGF, 10 mgml
21
insulin, 100 ng ml
21
cholera toxin, 0.5 mg
ml
21
hydrocortisone, 100 U ml
21
penicillin/streptomycin , and labelled/unla-
belled glucose and glutamine at 18 mM and 2.5 mM, respectively. PRC3, WT8,
pTV and pTR cell li nes were subcloned from the 786-O cell line as previously
described
31,32
. For isotopic labelling experiments, cells were cultured in six-well
plates in glucose- and gl utamine-free DMEM (Sigma) containing 10% dialysed
FBS (Invitrogen), 100 U ml
21
penicillin/streptomycin, naturally labelled glucos e
or glutamine, and the appropriate tracer, including [U-
13
C
5
]glutamine (Isotec),
[5-
13
C]glutamine (C/D/N Isotopes), [1-
13
C]glutamine, [U-
13
C
6
]glucose and
[1,2-
13
C
2
]glucose (all from Cambridge Isotope Labs). Steady-state labelling of
organic and amino acids was accomplished by culturing sub-confluent cells in
tracer medium for 24 h (Supplem entary Fig. 20). Labelling of fatty acids for ISA
was conducted over 3–5 days of culture in an excess of tracer medium (3–4 ml
per we ll in a six-well plate) to prevent nutrient depletion. Hypoxic culture was
conducte d by feeding a custom mixture of 1% O
2
,5%CO
2
and 94% N
2
to a
standard incubator co ntrolled at 5% CO
2
. HEPES (Mediatech) was added to the
culture medi um at 20 mM to maintain pH between normoxic and hypoxic
cultures. The internal gas content was monito red using Fyrite gas analysers
(Bacharach) for CO
2
and O
2
.O
2
levels were confirmed at 1–3% during hypoxic
culture.
Isolation and culture of primary CD8
1
T lymphocytes. Whole spleens were
isolated from OT-1 mice (Jackson Laboratories) expressing ovalbumin
(SIINFEKL)-specific T-cell receptors on CD8
1
T cells
33
. Spleens were passed
through 40-mm cell strainers, and red blood cells were removed by lysis (ACK
buffer, Gibco). Splenocytes were magnetically enriched for CD8
1
T cells by nega-
tive selection (Stem Cell Technologies) then suspended in DMEM containing 10%
dialysed serum, 100 U ml
21
penicillin/streptomycin and 13 MEM non-essential
amino-acid supplement, and labelled/unlabelled glucose (25 mM) and glutamine
(4 mM). Cells were cultured for 6 days in 12-well plates under hypoxia and
expanded by addition of 10 mgml
21
SIINFEKL peptide. Media was supplemented
with 30 U ml
21
of recombinant interleukin-2 every 48 h, and cells were main-
tained at a concentration between 1 3 10
6
and 2.5 3 10
6
cells per millilitre by
adding additional medium. To verify proliferation, cells were incubated with
2 mM carboxyfluorescein succinimidyl ester (CSFE) in PBS 1 2% FBS for 5 min,
washed three times in buffer, re-suspended in media and cultured as above.
Viability and T-cell enrichment were determined by flow cytometry after staining
with 49,6-diamidino-2-phenylindole (DAPI) and fluorescent anti-CD8 antibody.
T-cell proliferation was quantified by flow cytometry using a FACSCanto II
equipped with an High Throughput system (Becton Dickinson) and analysed
using FlowJo 7.5.5 (TreeStar).
Lentiviral-mediated generation of cells with knockdown of IDH1/2. Stable cell
cultures with decreased IDH1 and IDH2 expression were generated by lentiviral-
mediated shRNA expression. pLKO.1 lentiviral vectors targeting IDH1 had
shRNA sequences of CCGGGCTGCTTGCATTAAAGGTTTACTCGAGTAAA
CCTTTAATGCAAGCAGCTTTTT (IDH1a; TRCN0000027298) and CCGGCG
AATCATTTGGGAATTGATTCTCGAGAATCAATTCCCAAATGATTCGTTT
TT (IDH1b; TRCN0000027289), IDH2 had shRNA sequence CCGGGTGGA
CATCCAGCTAAAGTATCTC GAGATACTTTAGCTGGATGTCCAC TTTTT
(TRCN0000027225). For controls, either non-targeting control shRNA (SHC002;
Sigma) or pLKO.1 scrambled control vector
34
(Addgene) were used. pLKO.1 vector
targeting ARNT with shRNA sequence CCGGGAGAAGTCAGATGGTTTATTT
CTCGAGAAATAAACCATCTGACTTCTCTTTTT (TRCN0000003819) was
obtained from Open Biosystems. HEK293T cells were co-transfected with
pLKO.1 vectors and packaging plasmids to produce lentivirus. Filtered superna-
tants were used for infection, and cells were selectedwith puromycin (2 mgml
21
)for
at least two passages before initiating tracer and flux experiments.
Metabolite extraction and GC/MS analysis. At the conclusion of culture, cells
were rinsed with 1 ml ice-cold PBS and quenched with 0.4 ml ice-cold methanol.
An equal volume of water was added, and cells were collected in tubes by scraping
with a pipette. One volume of ice-cold chloroform was added to each tube, and the
extracts were vortexed at 4 uC for 30 min. Samples were centrifuged at 14,000g for
5 min at room temperature. For analysis of polar metabolites, the aqueous phase
was transferred to a new tube for evaporation under airflow. For ISA experiments,
the non-polar fraction was collected and evaporated under airflow. In some ISA
experiments, cells were trypsinized, counted and pelleted before lipid extraction as
described above.
Dried polar metabolites were dissolved in 20 mlof2%methoxyaminehydro-
chloride in pyridine (Pierce) and held at 37 u C for 1.5 h. After dissolution and
reaction, tert-butyldimethylsilyl (TBDMS) derivatization was initiated by add-
ing 30 ml N-methyl-N-(tert-butyldimethylsilyl)trifluoroacetamide (MBTSTFA)
1 1% tert -butyldimethylchlorosilane (TBDMCS; Pierce) and incub ating at
55 u C for 1 h. Fatty-acid methyl esters were gene rated by disso lving and reacting
dried chloroform fractions in 50–100 ml of Methyl-8 reagent (Pierce ) and
incuba ting at 60 uC for 1 h. GC/MS analysis was pe rformed using an Agilent
6890 GC equipped with a 30 m DB-35MS capillary column connected to an
Agilent 5975B MS operating under electron impact ionization at 70 eV. One
microlitre of sample was injected in splitless mode at 270 uC, using helium as the
carrier gas at a flow rate of 1 ml min
21
. For measurement of organic and amino
aci ds, the GC oven temperature was held at 10 0 uC for 3 min and increased to
300 u Cat3.5u min
21
. For analysis of fatty-acid methyl esters, the GC temper-
ature was held at 100 uC for 5 min after injection, increased to 200 u Cat15u
min
21
,thento250uCat5u min
21
and finally to 300 uCat15u min
21
.TheMS
sou rce and quadrupole were held at 230 uC and 150 uC, respectively, and the
detector was run in scanning mode, recording i on abundance in the range of
100–605 m/z. Mass isotopomer distributions were determined by integrating
the appropriate ion fragments listed in Supplementary Table 1. When required,
mass isotopomer distributions were corrected for natural isotope abundan ce
using in- house algorithms adapted from ref. 35. Relative metabolite abun-
dances were measured us ing a norvaline internal standard and integrating all
potentially labelled ion s in the metabolite fragments listed in Supplementary
Table 1.
ISA and metabolic flux analysis. Compu tational estimation of fluxes or per-
centage enrichment in the AcCoA pool and their associated 95% confidence
interva ls we re accomplished using the elementary metabolite unit (EMU)-based
software Metran, executed within Matlab (Mathworks) as previously
described
36–40
. Briefly, fluxes were determined iteratively by simulating MS
measurements fro m a given flux vector and comparing with mass isotopo mer
distribution measurements (three biological replicates). Upon obtaining an
acceptable fit, confidence intervals were determined for each flux using
parameter continuation. Assumptions, network (including atom transitions),
raw data an d model fits are presented in Suppleme ntary Information. ISA was
performed in a similar manner using the simple network described in
Supplementary Fig. 4 and Supplementary Table 6 to determine the tracer enrich-
ment in lipogenic AcCoA (D value) and percentage of newly synthesized lipids
and de novo lipogenesis, g(t) (ref. 15). Uncorrected mass isotopomer distribu-
tions, fitted parameters and confidence intervals used for ISA are listed in
Supplementary Information.
Calculation of absolute flux of tracers to palmitate in biomass. The quantity of
newly synthesized palmitate was determined by multiplying the fractional newly
synthesized palmitate value (g(t) value from ISA) by the total cellular palmitate.
Total cellular palmitate was quantified by GC/MS using a triheptadecanoin
internal standard. Flux of a given tracer to palmitate was calculated by multiplying
the tracer contribution (D value from ISA) by the amount of newly synthesized
palmitate and dividing by the integral viable cell density over the course of the
experiment.
Detection of
14
C incorporation into lipids. One hundred and fifty thousand
A549 cells were plated in DMEM. After attachment, [5-
14
C]glutamine was added
to the medium at a final concentration of 1 mCi ml
21
. After 72 h of growth, lipids
were extracted from plates using two rinses of 500 ml hexane:isopropanol (3:2).
Three hundred microlitres of PBS was added to induce phase separation and
collect the non-polar phase, and an additional 300 ml was added to the hexane
fraction to rinse the non-polar fraction. The hexane fraction was dried under
nitrogen gas, and the residue was dissolved in 200 ml chloroform before quantify-
ing c.p.m. by liquid scintillation counting. Matched plates of cells were counted to
determine cell number.
Metabolite analysis of spent medium. Glucose, lactate, glutamine and glutamate
concentrations were measured in fresh and spent medium samples using a Yellow
Springs Instruments 7100. Cell number was determined using a hemocytometer.
Extracellular flux measurements were calculated by assuming exponential growth
over the culture period to determine integral viable cell density.
Purification of recombinant IDH1. H is-tagged IDH1 in pET41a was trans-
formed i nto Escherichia coli (BL21 plysS DE3) and cells were grown with
kanamycin selection to an D
600 nm
of 0.6. The cells were then moved to 18 uC
and induced with 1 mM IPTG for 16–18 h, pelleted and subjected to freeze/thaw
before re-suspension in 60 ml lysis buffer (20 mM Tris, pH 7.4, 0.1% Triton
RESEARCH LETTER
Macmillan Publishers Limited. All rights reserved
©2011
Page 6
X-100, 500 mM NaCl, 5 mM b-me rcaptoethanol, 10% glycerol , supplemented
with protease inhibitors). Cells were lysed by sonication and protein bound to
Ni-NTA agarose. The beads were batch washed three or four times with wash
buffer (20 mM Tris, pH 7.4, 500 mM NaCl, 5 mM b-mercaptoetha nol, 10%
glycerol), then eluted f rom a column in 1-ml fract ions with elution
buffer (20 mM Tris, pH 7.4 , 500 mM NaCl, 5 mM b-mercaptoethanol,
500 mM imidazol e, 10% glycerol). The first and s econd fractions containing
most of the protein were dialysed into 50 mM Tris pH 7.5, 200 mM NaCl, 5 mM
b-mercaptoethanol, 2 mM MnCl
2
, 10% glycerol and the recombinant enzyme
was stored at 280 uC.
Recombinant IDH1 enzyme assays. All reactions were performed in reaction
buffer (100 mM Tris pH 7.5, 1.3 mM MnCl
2
, 200 mM NADPH and 2 mM
a-ketoglutarate), which was equilibrated overnight at 0%, 5% or 10% CO
2
as
indicated. One hundred microlitres of reaction buffer for each CO
2
condition
was added to 10 mg rIDH1 and activity was measured by following NADPH
fluorescence (excitation at 340 nm, emission at 460 nm).
SDS–PAGE and western blotting. Cells w ere rinsed with ice-cold PBS and
lysed using RIPA buffer. Proteins were separated by SDS–PAGE and transferred
to a nitrocellulose membrane. After blocking, membranes were probed with
goat anti-IDH1 poly clonal antibody (Santa Cruz Biotec hnology, sc49996),
mouse a nti-IDH2 monoclonal antibody (Abcam, ab55271), rabb it anti-
HIF2a polyclonal antibody (Novus Biologicals, N B100-122), mouse anti-
ARNT1 monoclonal antibody (BD Biosciences, 611079), mouse anti-b-actin
mouse monoclonal antibody (Novus Biologicals, ab8226), mous e anti-tubulin
monoclonal antibody (Sigma) or rabbit anti-tubulin antibody (Sigma) . Protein
was detected using horsera dish-peroxidase-conjugate d secondary antibodies
and chemiluminescence.
31. Zimmer, M., Doucette, D., Siddiqui, N. & Iliopoulos, O. Inhibition of hypoxia-
inducible factor is sufficient for growth suppression of VHL
2/2
tumors. Mol. Cancer
Res. 2, 89–95 (2004).
32. Iliopoulos, O., Kibel, A., Gray, S. & Kaelin, W. G. Jr. Tumour suppression by the
human von Hippel-Lindau gene product. Nature Med. 1, 822–826 (1995).
33. Hogquist, K. A. et al. T cell receptor antagonist peptides induce positive selection.
Cell 76, 17–27 (1994).
34. Sarbassov, D. D., Guertin, D. A., Ali, S. M. & Sabatini, D. M. Phosphorylation and
regulation of Akt/PKB by the rictor-mTOR complex. Science 307, 1098–1101
(2005).
35. Fernandez, C. A., Des Rosiers, C., Previs, S. F., David, F. & Brunengraber, H.
Correction of
13
C mass isotopomer distributions for natural stable isotope
abundance. J. Mass Spectrom. 31, 255–262 (1996).
36. Antoniewicz, M. R., Kelleher, J. K. & Stephanopoulos, G. Determination of
confidence intervals of metabolic fluxes estimated from stable isotope
measurements. Metab. Eng. 8, 324–337 (2006).
37. Antoniewicz, M. R., Kelleher, J. K. & Stephanopoulos, G. Elementary metabolite
units (EMU): a novel framework for modeling isotopic distributions. Metab. Eng. 9,
68–86 (2007).
38. Noguchi, Y. et al. Effect of anaplerotic fluxes and amino acid availability on hepatic
lipoapoptosis. J. Biol. Chem. 284, 33425–33436 (2009).
39. Gaglio, D. et al. Oncogenic K-Ras decouples glucose and glutamine metabolism to
support cancer cell growth. Mol. Syst. Biol. 7, 523 (2011).
40. Grassian, A. R., Metallo, C. M., Coloff, J. L., Stephanopoulos, G. & Brugge, J. S. Erk
regulation of pyruvate dehydrogenase flux through PDK4 modulates cell
proliferation. Genes Dev. 25, 1716–1733 (2011).
LETTER RESEARCH
Macmillan Publishers Limited. All rights reserved
©2011
Page 7
  • Source
    • "To demonstrate and validate our approach, we analyzed cellular metabolism of lung cancer cells at different oxygen levels. The data-driven analysis of isotopic enrichment correctly identified enhanced reductive carboxylation of 2-oxoglutarate to isocitrate by IDH and subsequent cleavage of citrate by ACLY to produce cytosolic acetyl-CoA which flows into increased fatty acid biosynthesis [36][37][38](Fig. 3). "
    [Show abstract] [Hide abstract] ABSTRACT: Background Metabolism gained increasing interest for the understanding of diseases and to pinpoint therapeutic intervention points. However, classical metabolomics techniques only provide a very static view on metabolism. Metabolic flux analysis methods, on the other hand, are highly targeted and require detailed knowledge on metabolism beforehand. Results We present a novel workflow to analyze non-targeted metabolome-wide stable isotope labeling data to detect metabolic flux changes in a non-targeted manner. Furthermore, we show how similarity-analysis of isotopic enrichment patterns can be used for pathway contextualization of unidentified compounds. We illustrate our approach with the analysis of changes in cellular metabolism of human adenocarcinoma cells in response to decreased oxygen availability. Starting without a priori knowledge, we detect metabolic flux changes, leading to an increased glutamine contribution to acetyl-CoA production, reveal biosynthesis of N-acetylaspartate by N-acetyltransferase 8-like (NAT8L) in lung cancer cells and show that NAT8L silencing inhibits proliferation of A549, JHH-4, PH5CH8, and BEAS-2B cells. Conclusions Differential stable isotope labeling analysis provides qualitative metabolic flux information in a non-targeted manner. Furthermore, similarity analysis of enrichment patterns provides information on metabolically closely related compounds. N-acetylaspartate and NAT8L are important players in cancer cell metabolism, a context in which they have not received much attention yet. Electronic supplementary material The online version of this article (doi:10.1186/s40170-016-0150-z) contains supplementary material, which is available to authorized users.
    Full-text · Article · Dec 2016
  • Source
    • "In cancer cells, TCA cycle anaplerosis is maintained mainly by glutamine [45,46] . Glutaminederived α-ketoglutarate is reductively carboxylated by isocitrate dehydrogenase 1 or 2 (IDH1, IDH2) to isocitrate/ citrate (Figure 1) [47,48]. NADPH-linked mitochondrial isocitrate dehydrogenase 2 (IDH2) is a PE5-down- regulated enzyme. "
    [Show abstract] [Hide abstract] ABSTRACT: Ribonucleases represent a new class of antitumor RNA-damaging drugs. However, many wild-type members of the vertebrate secreted ribonuclease family are not cytotoxic because they are not able to evade the cytosolic ribonuclease inhibitor. We previously engineered the human pancreatic ribonuclease to direct it to the cell nucleus where the inhibitor is not present. The best characterized variant is PE5 that kills cancer cells through apoptosis mediated by the p21WAF1/CIP1 induction and the inactivation of JNK. Here, we have used microarray-derived transcriptional profiling to identify PE5 regulated genes on the NCI/ADR-RES ovarian cancer cell line. RT-qPCR analyses have confirmed the expression microarray findings. The results show that PE5 cause pleiotropic effects. Among them, it is remarkable the down-regulation of multiple genes that code for enzymes involved in deregulated metabolic pathways in cancer cells.
    Full-text · Article · Mar 2016 · Oncotarget
  • Source
    • "The TCA cycle is also critically important in cancer cells for the generation of biochemical intermediates to sustain high rates of proliferation (Deberardinis et al., 2008). Most cancer cells derive lipogenic acetyl-coenzyme A (CoA) from pyruvate through Ca 2+ -dependent PDH (Hatzivassiliou et al., 2005; Metallo et al., 2012), and activation of PDH can promote cancer cell senescence (Kaplon et al., 2013 ). Some cancers have defective OXPHOS but, nevertheless, rely on the TCA cycle (Mullen et al., 2012Mullen et al., , 2014). "
    [Show abstract] [Hide abstract] ABSTRACT: In the absence of low-level ER-to-mitochondrial Ca(2+) transfer, ATP levels fall, and AMPK-dependent, mTOR-independent autophagy is induced as an essential survival mechanism in many cell types. Here, we demonstrate that tumorigenic cancer cell lines, transformed primary human fibroblasts, and tumors in vivo respond similarly but that autophagy is insufficient for survival, and cancer cells die while their normal counterparts are spared. Cancer cell death is due to compromised bioenergetics that can be rescued with metabolic substrates or nucleotides and caused by necrosis associated with mitotic catastrophe during their proliferation. Our findings reveal an unexpected dependency on constitutive Ca(2+) transfer to mitochondria for viability of tumorigenic cells and suggest that mitochondrial Ca(2+) addiction is a feature of cancer cells.
    Full-text · Article · Mar 2016 · Cell Reports
Show more