Cell 134, September 5, 2008 ©2008 Elsevier Inc. 703
It is hard to begin a discussion of cancer
cell metabolism without first mentioning
Otto Warburg. A pioneer in the study of
respiration, Warburg made a striking dis-
covery in the 1920s. He found that, even
in the presence of ample oxygen, cancer
cells prefer to metabolize glucose by gly-
colysis, a seeming paradox as glycolysis,
when compared to oxidative phosphory-
lation, is a less efficient pathway for pro-
ducing ATP (Warburg, 1956). The War-
burg effect has since been demonstrated
in different types of tumors and the con-
comitant increase in glucose uptake has
been exploited clinically for the detection
of tumors by fluorodeoxyglucose posi-
tron emission tomography (FDG-PET).
Although aerobic glycolysis has now
been generally accepted as a metabolic
hallmark of cancer, its causal relationship
with cancer progression is still unclear. In
this Essay, we discuss the possible driv-
ers, advantages, and potential liabilities
of the altered metabolism of cancer cells
(Figure 1). Although our emphasis on the
Warburg effect reflects the focus of the
field, we would also like to encourage a
broader approach to the study of cancer
metabolism that takes into account the
contributions of all interconnected small
molecule pathways of the cell.
The Tumor Microenvironment
Selects for Altered Metabolism
One compelling idea to explain the War-
burg effect is that the altered metabo-
lism of cancer cells confers a selective
advantage for survival and proliferation
in the unique tumor microenvironment.
As the early tumor expands, it outgrows
the diffusion limits of its local blood sup-
ply, leading to hypoxia and stabilization
of the hypoxia-inducible transcription
factor, HIF. HIF initiates a transcrip-
tional program that provides multiple
solutions to hypoxic stress (reviewed in
Kaelin and Ratcliffe, 2008). Because a
decreased dependence on aerobic res-
piration becomes advantageous, cell
metabolism is shifted toward glycolysis
by the increased expression of glyco-
lytic enzymes, glucose transporters, and
inhibitors of mitochondrial metabolism.
In addition, HIF stimulates angiogenesis
(the formation of new blood vessels) by
upregulating several factors, including
most prominently vascular endothelial
growth factor (VEGF).
Blood vessels recruited to the tumor
microenvironment, however, are disor-
ganized, may not deliver blood effec-
tively, and therefore do not completely
alleviate hypoxia (reviewed in Gatenby
and Gillies, 2004). The oxygen levels
within a tumor vary both spatially and
temporally, and the resulting rounds
of fluctuating oxygen levels potentially
select for tumors that constitutively
with the possible exception of tumors
that have lost the von Hippel-Lindau
protein (VHL), which normally mediates
degradation of HIF, HIF is still coupled
to oxygen levels, as evident from the
heterogeneity of HIF expression within
the tumor microenvironment (Wiesener
et al., 2001; Zhong et al., 1999). There-
fore, the Warburg effect—that is, an
uncoupling of glycolysis from oxygen
levels—cannot be explained solely by
upregulation of HIF. Other molecular
mechanisms are likely to be important,
such as the metabolic changes induced
by oncogene activation and tumor sup-
Oncogene Activation Drives
Changes in Metabolism
Not only may the tumor microenviron-
ment select for a deranged metabolism,
but oncogene status can also drive
metabolic changes. Since Warburg’s
time, the biochemical study of cancer
metabolism has been overshadowed
by efforts to identify the mutations
that contribute to cancer initiation and
progression. Recent work, however,
has demonstrated that the key compo-
nents of the Warburg effect—increased
glucose consumption, decreased oxi-
dative phosphorylation, and accom-
panying lactate production—are also
distinguishing features of oncogene
activation. The signaling molecule Ras,
a powerful oncogene when mutated,
promotes glycolysis (reviewed in Dang
and Semenza, 1999; Ramanathan et al.,
2005). Akt kinase, a well-characterized
downstream effector of insulin signaling,
reprises its role in glucose uptake and
utilization in the cancer setting (reviewed
in Manning and Cantley, 2007), whereas
the Myc transcription factor upregulates
the expression of various metabolic
genes (reviewed in Gordan et al., 2007).
The most parsimonious route to tumori-
genesis may be activation of key onco-
genic nodes that execute a proliferative
program, of which metabolism may be
one important arm. Moreover, regula-
tion of metabolism is not exclusive to
oncogenes. Loss of the tumor suppres-
sor protein p53 prevents expression of
Cancer Cell Metabolism:
Warburg and Beyond
Peggy P. Hsu1,2 and David M. Sabatini1,2,3,*
1Whitehead Institute for Biomedical Research and Massachusetts Institute of Technology Department of Biology, Cambridge, MA 02142, USA
2Broad Institute, Cambridge, MA 02142, USA
3Koch Institute for Integrative Cancer Research at MIT, Cambridge, MA 02139, USA
Described decades ago, the Warburg effect of aerobic glycolysis is a key metabolic hallmark of
cancer, yet its significance remains unclear. In this Essay, we re-examine the Warburg effect and
establish a framework for understanding its contribution to the altered metabolism of cancer cells.
704 Cell 134, September 5, 2008 ©2008 Elsevier Inc.
the gene encoding SCO2 (the synthesis
of cytochrome c oxidase protein), which
interferes with the function of the mito-
chondrial respiratory chain (Matoba et
al., 2006). A second p53 effector, TIGAR
(TP53-induced glycolysis and apop-
tosis regulator), inhibits glycolysis by
decreasing levels of fructose-2,6-bis-
phosphate, a potent stimulator of glyc-
olysis and inhibitor of gluconeogenesis
(Bensaad et al., 2006). Other work also
suggests that p53-mediated regulation
of glucose metabolism may be depen-
dent on the transcription factor NF-κB
(Kawauchi et al., 2008).
It has been shown that inhibition of lac-
tate dehydrogenase A (LDH-A) prevents
the Warburg effect and forces cancer
cells to revert to oxidative phosphoryla-
tion in order to reoxidize NADH and pro-
duce ATP (Fantin et al., 2006; Shim et
al., 1997). While the cells are respiratory
competent, they exhibit attenuated tumor
growth, suggesting that aerobic glycoly-
sis might be essential for cancer progres-
sion. In a primary fibroblast cell culture
model of stepwise malignant transfor-
mation through overexpression of telom-
erase, large and small T antigen, and the
H-Ras oncogene, increasing tumorige-
nicity correlates with sensitivity to glyco-
lytic inhibition. This finding suggests that
the Warburg effect might be inherent to
the molecular events of transformation
(Ramanathan et al., 2005). However, the
introduction of similar defined factors into
human mesenchymal stem cells (MSCs)
revealed that transformation can be asso-
ciated with increased dependence on
oxidative phosphorylation (Funes et al.,
2007). Interestingly, when introduced in
vivo these transformed MSCs do upreg-
ulate glycolytic genes, an effect that is
reversed when the cells are explanted
and cultured under normoxic conditions.
Figure 1. The Altered Metabolism of Cancer Cells
Drivers (A and B). The metabolic derangements in cancer cells may arise either from the selection of cells that have adapted to the tumor microenvironment or
from aberrant signaling due to oncogene activation. The tumor microenvironment is spatially and temporally heterogeneous, containing regions of low oxygen
and low pH (purple). Moreover, many canonical cancer-associated signaling pathways induce metabolic reprogramming. Target genes activated by hypoxia-
inducible factor (HIF) decrease the dependence of the cell on oxygen, whereas Ras, Myc, and Akt can also upregulate glucose consumption and glycolysis.
Loss of p53 may also recapitulate the features of the Warburg effect, that is, the uncoupling of glycolysis from oxygen levels.
Advantages (C–E). The altered metabolism of cancer cells is likely to imbue them with several proliferative and survival advantages, such as enabling cancer cells
to execute the biosynthesis of macromolecules (C), to avoid apoptosis (D), and to engage in local metabolite-based paracrine and autocrine signaling (E).
Potential Liabilities (F and G). This altered metabolism, however, may also confer several vulnerabilities on cancer cells. For example, an upregulated metabo-
lism may result in the build up of toxic metabolites, including lactate and noncanonical nucleotides, which must be disposed of (F). Moreover, cancer cells may
also exhibit a high energetic demand, for which they must either increase flux through normal ATP-generating processes, or else rely on an increased diversity
of fuel sources (G).
Cell 134, September 5, 2008 ©2008 Elsevier Inc. 705
These contrasting models suggest that
the Warburg effect may be context depen-
dent, in some cases driven by genetic
changes and in others by the demands
of the microenvironment. Regardless of
whether the tumor microenvironment or
oncogene activation plays a more impor-
tant role in driving the development of a
distinct cancer metabolism, it is likely that
the resulting alterations confer adaptive,
proliferative, and survival advantages on
the cancer cell.
Altered Metabolism Provides
Substrates for Biosynthetic Pathways
Although studies in cancer metabolism
have largely been energy-centric, rap-
idly dividing cells have diverse require-
ments. Proliferating cells require not
only ATP but also nucleotides, fatty
acids, membrane lipids, and proteins,
and a reprogrammed metabolism may
serve to support synthesis of macro-
molecules. Recent studies have shown
that several steps in lipid synthesis are
required for and may even actively pro-
mote tumorigenesis. Inhibition of ATP
citrate lyase, the distal enzyme that
converts mitochondrial-derived citrate
into cytosolic acetyl coenzyme A, the
precursor for many lipid species, pre-
vents cancer cell proliferation and tumor
growth (Hatzivassiliou et al., 2005).
Fatty acid synthase, expressed at low
levels in normal tissues, is upregulated
in cancer and may also be required for
tumorigenesis (reviewed in Menendez
and Lupu, 2007). Furthermore, can-
cer cells may also enhance their bio-
synthetic capabilities by expressing a
tumor-specific form of pyruvate kinase
(PK), M2-PK. Pyruvate kinase cata-
lyzes the third irreversible reaction of
glycolysis, the conversion of phospho-
enolpyruvate (PEP) to pyruvate. Sur-
prisingly, the M2-PK of cancer cells is
thought to be less active in the conver-
sion of PEP to pyruvate and thus less
efficient at ATP production (reviewed in
Mazurek et al., 2005). A major advan-
tage to the cancer cell, however, is that
the glycolytic intermediates upstream
of PEP might be shunted into synthetic
processes. Recent work has found that
the cancer-specific M2-PK causes an
increase in the incorporation of glucose
carbons into lipids and, expanding the
connection between growth factor sig-
naling and cancer metabolism, may be
regulated by phosphotyrosine binding
(Christofk et al., 2008a, 2008b).
Making the building blocks of the cell,
however, incurs an energetic cost and
cannot fully explain the Warburg effect.
Biosynthesis, in addition to causing an
inherent increase in ATP demand in order
to execute synthetic reactions, should
also cause a decrease in ATP supply
as various glycolytic and Krebs cycle
intermediates are diverted. Lipid syn-
thesis, for example, requires the coop-
eration of glycolysis, the Krebs cycle,
and the pentose phosphate shunt. As
pyruvate must enter the mitochondria in
this case, it avoids conversion to lactate
and therefore cannot contribute to gly-
colysis-derived ATP. Moreover, whereas
increased biosynthesis may explain the
glucose hunger of cancer cells, it can-
not explain the increase in lactic acid
production originally described by War-
burg, suggesting that lactate must also
result from the metabolism of non-glu-
cose substrates. Recently, it has been
demonstrated that glutamine may be
metabolized by the citric acid cycle in
cancer cells and converted into lactate,
producing NADPH for lipid biosynthesis
and oxaloacetate for replenishment of
Krebs cycle intermediates (DeBerardinis
et al., 2007).
Metabolic Pathways Regulate
In addition to involvement in proliferation,
altered metabolism may promote another
cancer-essential function: the avoidance
of apoptosis. Loss of the p53 target
TIGAR sensitizes cancer cells to apopto-
sis, most likely by causing an increase in
reactive oxygen species (Bensaad et al.,
2006). On the other hand, overexpression
of glyceraldehyde-3-phosphate dehydro-
genase (GAPDH) prevents caspase-inde-
pendent cell death, presumably by stimu-
lating glycolysis, increasing cellular ATP
levels, and promoting autophagy (Colell
et al., 2007). Whether or not GAPDH plays
a physiological role in the regulation of
cell death remains to be determined.
Intriguingly, Bonnet et al. (2007) have
reported that treating cancer cells with
dichloroacetate (DCA), a small molecule
inhibitor of pyruvate dehydrogenase
kinase, has striking effects on their sur-
vival and on xenograft tumor growth.
DCA, a currently approved treatment
for congenital lactic acidosis, activates
oxidative phosphorylation and pro-
motes apoptosis by two mechanisms.
First, increased flux through the elec-
tron transport chain causes depolar-
ization of the mitochondrial membrane
potential (which the authors found to
be hyperpolarized specifically in cancer
cells) and release of the apoptotic effec-
tor cytochrome c. Second, an increase
in reactive oxygen species generated by
oxidative phosphorylation upregulates
the voltage-gated K+ channel, leading to
potassium ion efflux and caspase acti-
vation. Their work suggests that can-
cer cells may shift their metabolism to
glycolysis in order to prevent cell death
and that forcing cancer cells to respire
aerobically can counteract this adapta-
tion. Although this preliminary work has
prompted some cancer patients to self-
medicate with DCA, a controlled clini-
cal trial will be essential to demonstrate
unequivocally the safety and efficacy of
DCA as an anti-cancer agent.
Cancer Cells May Signal Locally in
the Tumor Microenvironment
Cancer cells may rewire metabolic path-
ways to exploit the tumor microenviron-
ment and to support cancer-specific
signaling. Without access to the central
circulation, it is possible that metabolites
can be concentrated locally and reach
suprasystemic levels, allowing cancer
cells to engage in metabolite-mediated
autocrine and paracrine signaling that
does not occur in normal tissues. So-
called androgen-independent prostate
cancers may only be independent from
exogenous, adrenal-synthesized andro-
gens. Androgen-independent prostate
cancer cells still express the androgen
receptor and may be capable of autono-
mously synthesizing their own andro-
gens (Stanbrough et al., 2006).
Perhaps the more provocative but as
yet untested idea is that metabolites in
the diffusion-limited tumor microenviron-
ment could be acting as paracrine signal-
ing molecules. Traditionally thought of as
a glycolytic waste product, lactate may
be one such signal. As noted above, it
has been found that inhibition of lactate
dehydrogenase can block tumor growth,
most likely by multiple mechanisms. Much
of the evidence for lactate as a multifunc-
706 Cell 134, September 5, 2008 ©2008 Elsevier Inc.
tional metabolite comes from work in exer-
cise physiology and muscle metabolism
(reviewed in Philp et al., 2005). Transported
by several monocarboxylate transporters,
lactate may be shared and metabolized
among cells, although the idea is still con-
troversial (Hashimoto et al., 2006; Yoshida
et al., 2007). The interconversion of lactate
and pyruvate might alter the NAD+/NADH
ratio in cells, and lactate exchange may
serve to coordinate the metabolism of a
group of cells. The tumor-stroma inter-
action may therefore have a metabolic
component (Koukourakis et al., 2006).
Cancer cells respond cell-autonomously
to hypoxia to initiate angiogenesis, and so
it would be exciting if a metabolite such as
lactate could positively amplify this angio-
genic program, a process that requires a
semicoordinated effort among multiple
cells. Indeed, acidosis often precedes
angiogenesis, and lactate may stimulate
HIF expression independently of hypoxia
(Fukumura et al., 2001; Lu et al., 2002; Shi
et al., 2001). Cancer cells, by participating
in a kind of quorum sensing and coordi-
nating their metabolism, may therefore act
as a pseudo-organ.
Metabolism as an Upstream
Modulator of Signaling Pathways
Not only is metabolism downstream
of oncogenic pathways, but an altered
upstream metabolism may affect the
activity of signaling pathways that nor-
mally sense the state of the cell. Individu-
als with inherited mutations in succinate
dehydrogenase and fumarate hydratase
develop highly angiogenic tumors, not
unlike those exhibiting loss of the VHL
tumor suppressor protein that acts
upstream of HIF (reviewed in Kaelin and
Ratcliffe, 2008). The mechanism of tum-
origenesis in these cancer syndromes is
still contentious. However, it has been
proposed that loss of succinate dehydro-
genase and fumarate hydratase causes
an accumulation of succinate or fumar-
ate, respectively, leading to inhibition of
the prolyl hydroxylases that mark HIF for
VHL-mediated degradation (Isaacs et al.,
2005; Pollard et al., 2005; Selak et al.,
2005). In this rare case, succinate dehy-
drogenase and fumarate hydratase are
acting as bona fide tumor suppressors.
Mutations in metabolic genes, how-
ever, need not be a cancer-causing
event. More subtly, the activation of vari-
ous metabolic pathways might modulate
the activity of downstream pro-cancer
factors. Whereas it is well-accepted that
growth factor signaling is commonly
dysregulated in cancer, the involvement
of nutrient or energy signaling in cancer
remains unclear. In prokaryotes, various
metabolites are sensed directly by the
signaling machinery. The mammalian
pathways that respond to energy and
nutrient status may also interface with
metabolites directly. It is well established
that AMP-kinase senses the AMP/ATP
ratio (reviewed in Hardie, 2007), whereas
mTOR (the mammalian target of rapamy-
cin) senses cellular amino acid con-
centrations (Kim et al., 2008; Sancak et
al., 2008). Both AMP-kinase and mTOR
have been linked to tumor syndromes.
It is possible that one way to upregulate
these pro-growth signaling pathways
is to increase the levels of the normal
metabolites that they sense.
Metabolism Upregulation Generates
Although altered metabolism confers
several advantages on the cancer cell, it
does not come without disadvantages.
As a consequence of a deranged or sim-
ply overactive metabolism, cancer cells
may be burdened with toxic byproducts
that require disposal. So far, there is rela-
tively little evidence for this hypothesis in
the existing literature, but a few exam-
ples do suggest that cancer cells require
detoxification mechanisms to maintain
survival. Although there are enzymes
that detoxify exogenous toxins, sev-
eral “house-cleaning” enzymes, a term
coined from studies in bacteria, deal with
endogenous toxic metabolites (reviewed
in Galperin et al., 2006). The best exam-
ple of “house-cleaning” enzymes are
the NUDIX (noncanonical nucleoside
diphosphate linked to some other moiety
X) hydrolases, a family of enzymes that
act on the nucleotide pool and remove
noncanonical nucleoside triphosphates.
When incorporated into the DNA, these
aberrant nucleotides can lead to mis-
matches, mutations, and eventually
cell death. The dUTP pyrophosphatase
(DUT), which hydrolyzes dUTP to dUMP
and prevents the incorporation of uracils
into DNA, may play a role in resistance
to thymidylate synthase inhibitors. Sup-
pression of DUT sensitizes some can-
cer cells to pyrimidine antimetabolites,
suggesting that inhibition of these cel-
lular house-cleaning enzymes may be
an effective adjunct chemotherapeutic
strategy (Koehler and Ladner, 2004).
The lactate production associated with
the shift to a glycolytic metabolism is
thought to contribute to the acidification
of the microenvironment. Able to adapt
to and even benefit from an acidic envi-
ronment, cancer cells have been shown
to upregulate vacuolar H+-ATPases,
Na+-H+ antiporters, and H+-linked mono-
carboxylate transporters (reviewed in
Gatenby and Gillies, 2004). Inhibition of
these adaptive mechanisms can lead to
decreased viability of cancer cells and
increased sensitivity to chemotherapeu-
tic agents (reviewed in Fais et al., 2007;
Fang et al., 2006).
Many mysteries remain unsolved in our
understanding of even normal human
metabolism, let alone that of cancer cells.
The metabolic pathways of the mamma-
lian cell and their many interconnections
are incomplete, as many enzymes remain
unannotated in the human genome.
Although we have guesses by homology,
the identities of the human enzymes that
catalyze reactions we know must occur
are still elusive. In addition to annotating
all human metabolic genes, the “ins” and
the “outs” (i.e., the metabolites that enter
and exit cells) should be measured and
cataloged. It is also entirely unclear what
percentage of the cellular fuel is normally
used for ATP generation, biosynthesis, or
other processes. And with few exceptions
surprisingly little is known about intercel-
lular metabolism. Much of our understand-
ing of metabolism has been inherited from
work in simple organisms; the compart-
mental nature of human metabolism is an
exciting area of potential exploration.
Although aerobic glycolysis is the
most characterized, although still puz-
zling, metabolic phenomenon in cancer,
many other aspects of cancer metabo-
lism are likely to be derangements of
normal metabolism and ought to be elu-
cidated. The nutrient conditions of the
tumor microenvironment have not yet
been carefully examined. Cancer cells,
despite engaging in energy-costly pro-
cesses, must still be able to maintain ATP
levels, by either relying on increased flux
Cell 134, September 5, 2008 ©2008 Elsevier Inc. 707
through glycolysis or utilizing a diversity
of fuel sources. Several hypotheses exist
as to why a fraction of tumors are refrac-
tory to imaging by FDG-PET. One pos-
sibility is that certain cancer cells may
not be primarily glucose-metabolizers
but may rely on alternative fuel sources,
the detailed characterization of which
may lead to the detection and treatment
of “PET-negative” tumors. Furthermore,
there are more complex questions to
be answered: Is it possible that cancer
cells exhibit “metabolite addiction”? Are
there unique cancer-specific metabolic
pathways, or combinations of pathways,
utilized by the cancer cell but not by nor-
mal cells? Are different stages of meta-
bolic adaptations required for the cancer
cell to progress from the primary tumor
stage to invasion to metastasis? How
malleable is cancer metabolism?
From a therapeutic
knowledge of the causes, benefits, and
vulnerabilities of cancer cell metabolism
will enable the identification of new drug
targets and will facilitate the design of
metabolite mimetics that are uniquely
taken up by cancer cells or converted
into the active form by enzymes upregu-
lated in tumors. Profiling of either metab-
olites or enzymatic activities may allow
us to develop diagnostic tests of can-
cer, and metabolite derivatives can be
used for the molecular imaging of can-
cer, as exemplified by FDG-PET. We find
the possibility of a new class of cancer
therapeutics and diagnostic tools espe-
cially exciting. Therefore, we emphasize
the need to explore beyond a glucose
and energy-centric driven model of can-
cer metabolism to a broader one that
encompasses all of the metabolic needs
of a cancer cell. Perhaps it is time to step
out from under Warburg’s shadow.
We thank T. DiCesare for help with the figure.
Bensaad, K., Tsuruta, A., Selak, M.A., Vidal,
M.N., Nakano, K., Bartrons, R., Gottlieb, E., and
Vousden, K.H. (2006). Cell 126, 107–120.
Bonnet, S., Archer, S.L., Allalunis-Turner, J.,
Haromy, A., Beaulieu, C., Thompson, R., Lee,
C.T., Lopaschuk, G.D., Puttagunta, L., Harry, G.,
et al. (2007). Cancer Cell 11, 37–51.
Christofk, H.R., Vander Heiden, M.G., Harris,
M.H., Ramanathan, A., Gerszten, R.E., Wei, R.,
Fleming, M.D., Schreiber, S.L., and Cantley, L.C.
(2008a). Nature 452, 230–233.
Christofk, H.R., Vander Heiden, M.G., Wu, N.,
Asara, J.M., and Cantley, L.C. (2008b). Nature
Colell, A., Ricci, J.E., Tait, S., Milasta, S., Mau-
rer, U., Bouchier-Hayes, L., Fitzgerald, P., Guio-
Carrion, A., Waterhouse, N.J., Li, C.W., et al.
(2007). Cell 129, 983–997.
Dang, C.V., and Semenza, G.L. (1999). Trends
Biochem. Sci. 24, 68–72.
DeBerardinis, R.J., Mancuso, A., Daikhin, E.,
Nissim, I., Yudkoff, M., Wehrli, S., and Thomp-
son, C.B. (2007). Proc. Natl. Acad. Sci. USA 104,
Fais, S., De Milito, A., You, H., and Qin, W.
(2007). Cancer Res. 67, 10627–10630.
Fang, J., Quinones, Q.J., Holman, T.L., Morow-
itz, M.J., Wang, Q., Zhao, H., Sivo, F., Maris,
J.M., and Wahl, M.L. (2006). Mol. Pharmacol.
Fantin, V.R., St-Pierre, J., and Leder, P. (2006).
Cancer Cell 9, 425–434.
Fukumura, D., Xu, L., Chen, Y., Gohongi, T.,
Seed, B., and Jain, R.K. (2001). Cancer Res. 61,
Funes, J.M., Quintero, M., Henderson, S., Mar-
tinez, D., Qureshi, U., Westwood, C., Clements,
M.O., Bourboulia, D., Pedley, R.B., Moncada,
S., and Boshoff, C. (2007). Proc. Natl. Acad. Sci.
USA 104, 6223–6228.
Galperin, M.Y., Moroz, O.V., Wilson, K.S., and
Murzin, A.G. (2006). Mol. Microbiol. 59, 5–19.
Gatenby, R.A., and Gillies, R.J. (2004). Nat. Rev.
Cancer 4, 891–899.
Gordan, J.D., Thompson, C.B., and Simon, M.C.
(2007). Cancer Cell 12, 108–113.
Hardie, D.G. (2007). Nat. Rev. Mol. Cell Biol. 8,
Hashimoto, T., Hussien, R., and Brooks, G.A.
(2006). Am. J. Physiol. Endocrinol. Metab. 290,
Hatzivassiliou, G., Zhao, F., Bauer, D.E., An-
dreadis, C., Shaw, A.N., Dhanak, D., Hingorani,
S.R., Tuveson, D.A., and Thompson, C.B. (2005).
Cancer Cell 8, 311–321.
Isaacs, J.S., Jung, Y.J., Mole, D.R., Lee, S., Tor-
res-Cabala, C., Chung, Y.L., Merino, M., Trepel,
J., Zbar, B., Toro, J., et al. (2005). Cancer Cell
Kaelin, W.G., Jr., and Ratcliffe, P.J. (2008). Mol.
Cell 30, 393–402.
Kawauchi, K., Araki, K., Tobiume, K., and Tana-
ka, N. (2008). Nat. Cell Biol. 10, 611–618.
Kim, E., Goraksha-Hicks, P., Li, L., Neufeld,
T.P., and Guan, K.L. (2008). Nat. Cell Biol. 10,
Koehler, S.E., and Ladner, R.D. (2004). Mol.
Pharmacol. 66, 620–626.
Koukourakis, M.I., Giatromanolaki, A., Harris,
A.L., and Sivridis, E. (2006). Cancer Res. 66,
Lu, H., Forbes, R.A., and Verma, A. (2002). J.
Biol. Chem. 277, 23111–23115.
Manning, B.D., and Cantley, L.C. (2007). Cell
Matoba, S., Kang, J.G., Patino, W.D., Wragg,
A., Boehm, M., Gavrilova, O., Hurley, P.J.,
Bunz, F., and Hwang, P.M. (2006). Science 312,
Mazurek, S., Boschek, C.B., Hugo, F., and
Eigenbrodt, E. (2005). Semin. Cancer Biol. 15,
Menendez, J.A., and Lupu, R. (2007). Nat. Rev.
Cancer 7, 763–777.
Philp, A., Macdonald, A.L., and Watt, P.W. (2005).
J. Exp. Biol. 208, 4561–4575.
Pollard, P.J., Briere, J.J., Alam, N.A., Barwell, J.,
Barclay, E., Wortham, N.C., Hunt, T., Mitchell,
M., Olpin, S., Moat, S.J., et al. (2005). Hum. Mol.
Genet. 14, 2231–2239.
Ramanathan, A., Wang, C., and Schreiber,
S.L. (2005). Proc. Natl. Acad. Sci. USA 102,
Sancak, Y., Peterson, T.R., Shaul, Y.D., Lindquist,
R.A., Thoreen, C.C., Bar-Peled, L., and Sabatini,
D.M. (2008). Science 320, 1496–1501.
Selak, M.A., Armour, S.M., MacKenzie, E.D.,
Boulahbel, H., Watson, D.G., Mansfield, K.D.,
Pan, Y., Simon, M.C., Thompson, C.B., and
Gottlieb, E. (2005). Cancer Cell 7, 77–85.
Shi, Q., Le, X., Wang, B., Abbruzzese, J.L.,
Xiong, Q., He, Y., and Xie, K. (2001). Oncogene
Shim, H., Dolde, C., Lewis, B.C., Wu, C.S.,
Dang, G., Jungmann, R.A., Dalla-Favera, R., and
Dang, C.V. (1997). Proc. Natl. Acad. Sci. USA 94,
Stanbrough, M., Bubley, G.J., Ross, K., Gol-
ub, T.R., Rubin, M.A., Penning, T.M., Febbo,
P.G., and Balk, S.P. (2006). Cancer Res. 66,
Warburg, O. (1956). Science 124, 269–270.
Wiesener, M.S., Munchenhagen, P.M., Berger, I.,
Morgan, N.V., Roigas, J., Schwiertz, A., Jurgens-
en, J.S., Gruber, G., Maxwell, P.H., Loning, S.A.,
et al. (2001). Cancer Res. 61, 5215–5222.
Yoshida, Y., Holloway, G.P., Ljubicic, V., Hatta,
H., Spriet, L.L., Hood, D.A., and Bonen, A.
(2007). J. Physiol. 582, 1317–1335.
Zhong, H., De Marzo, A.M., Laughner, E., Lim,
M., Hilton, D.A., Zagzag, D., Buechler, P., Isaacs,
W.B., Semenza, G.L., and Simons, J.W. (1999).
Cancer Res. 59, 5830–5835.